Oxfordstudiesinroman Societyandlaw: General Editors . . C

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi
OXFORD STUDIES IN ROMAN

SOCIETY AND LAW
General Editors
    . . c
OXFORD STUDIES IN ROMAN SOCIETY AND LAW
The aim of this monograph series is to create an interdisciplinary forum

devoted to the interaction between legal history and ancient history, in the
context of the study of Roman law. Focusing on the relationship of law to
society, the volumes will cover the most significant periods of Roman law (up
to the death of Justinian in 565) so as to provide a balanced view of growth,
decline, and resurgence. Most importantly, the series will provoke general
debate over the extent to which legal rules should be examined in light of the
society which produced them in order to understand their purpose and
efficacy.
Roman Law and

Economics
Volume II
Exchange, Ownership, and Disputes
Edited by
GIUSEPPE DARI-MATTIACCI
AND DENNIS P. KEHOE
1
3
Great Clarendon Street, Oxford, OX2 6DP,
United Kingdom
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide. Oxford is a registered trade mark of
Oxford University Press in the UK and in certain other countries
© Oxford University Press 2020
The moral rights of the authors have been asserted
First Edition published in 2020
Impression: 1
All rights reserved. No part of this publication may be reproduced, stored in
a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by licence or under terms agreed with the appropriate reprographics
rights organization. Enquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above
You must not circulate this work in any other form
and you must impose this same condition on any acquirer
Published in the United States of America by Oxford University Press
198 Madison Avenue, New York, NY 10016, United States of America
British Library Cataloguing in Publication Data
Data available
Library of Congress Control Number: 2020934649
ISBN 978–0–19–878721–1
Printed and bound by
CPI Group (UK) Ltd, Croydon, CR0 4YY
Links to third party websites are provided by Oxford in good faith and
for information only. Oxford disclaims any responsibility for the materials
contained in any third party website referenced in this work.
Preface
In recent years, historians of Roman law and the Roman economy

have increasingly turned toward insights from economic theory to
come to a better understanding of the relationship between Roman
law and institutions and the economy of the Roman world. The
research of such scholars has borrowed tools mainly from the fields
of law and economics and new institutional economics, and the result
has been very fruitful, with an increasing number of Roman histor-
ians pursuing genuine interdisciplinary work that applies modern
theoretical perspectives to understanding the development of
Roman jurisprudence and legal institutions in numerous areas of
the law that were important to the economy, such as contracts
of sale and lease and hire, markets, credit, agency, and banking, not
to mention the vast field of Roman slavery. At the same time, modern
legal scholars and economists have found the ancient world, particu-
larly the Roman Empire, to be an intriguing test case to trace the role
of law and legal institutions in a pre-industrial society. The Roman
Empire provides a fitting test case for theories about the relationship
between law and the economy because its legal institutions are rela-
tively well documented, while at the same time the economic per-
formance of the Roman world has been an area of intensive study and
debate among economic historians. One of the principal questions
that historians of the Roman economy have been addressing in recent
years concerns the performance of the economy of the Roman
Empire: given the favorable conditions resulting from relative peace,
political unification, and a more uniform and predictable system of
courts and enforcement of contracts, to what extent did the Roman
Empire experience economic growth, and, if so, how were the fruits of
economic growth shared across society and to what extent did the
legal institutions in place contribute and change in response to it?
Roman economic historians have also focused on many other ques-
tions related to these overarching issues, such as the effects of demo-
graphic change on the economy, economic planning by Roman
property owners, and the role of the state in the economy, not to
mention the actual effectiveness of Roman legal institutions and the
vi Preface
degree to which people in the empire relied on them as they sought to
protect their own most important economic interests.
The rich debate over these and many other issues provides a very
favorable climate for genuine interdisciplinary research on the
Roman world. This collection of essays endeavors to make an import-
ant contribution to bringing together scholars with diverse research
backgrounds and thus to consolidate a new field of research, the
economic analysis of Roman law (or Roman law and economics).
The collection does so by drawing together scholars from a variety of
fields both ancient and modern, and integrating insights from legal
history, economic history, and the social sciences and, in particular,
economic theory and econometrics. We hope to achieve two sets
of goals.
First, this collection provides a novel perspective on the function,
evolution, and, possibly, rationale of Roman legal institutions. While
we have no ambition to provide a systematic analysis, the various
chapters offer a wide coverage of the law and institutions of ancient
Rome and provide an innovative perspective of often long-studied
issues.
Second, this collection contributes a radically interdisciplinary
methodological toolbox to the analysis of Roman legal institutions
(and ancient legal institutions, more generally). Through the various
chapters, the reader is exposed to a rich array of methodological
approaches. Careful historical analysis of both legal and economic
institutions is combined with cutting-edge theoretical and empirical
examination.
Next to those who, like us, are fascinated by the law and institu-
tions of ancient Rome, we hope to interest a broader community of
historians (both legal and economic), legal scholars, and social scien-
tists. On the one hand, the set of methods that are used in this book
can be applied to many more issues than we could cover in two
volumes. Students of Roman legal and economic institutions will
hopefully find some of these methods useful. On the other hand, we
hope to have demonstrated that the information available on about
one thousand years of Roman history offers an interesting natural
laboratory to test the explanatory power of modern theoretical
approaches.
This project would have not been possible without invaluable
advice, support, and encouragement we received from Barbara
Abatino, Erasmo Giambona, Henry Hansmann, Elio Lo Cascio, and
Preface vii
Peter Temin at many crucial junctures. We are also deeply indebted
to the series editors, Paul du Plessis and Thomas A. J. McGinn, and
two anonymous reviewers, who guided and advised us throughout
a long and intense process of revision. Georgina Leighton has been
extraordinarily supportive and patient as we have worked to bring
the collection into its final form. We would also like to thank
Tim Beck for his energetic work in copyediting, and Chandrakala
Chandrasekaran for managing production. In addition, we are grate-
ful to Claudia Kassner and Aparna Sundaram of Columbia Law
School for preparing the index. Needless to say, the greatest share
of efforts has been exerted by the authors of the chapters in this book;
we are extremely grateful to them all.
Giuseppe Dari-Mattiacci
Dennis P. Kehoe
List of Figures
18.1. Cesena, Italy (44°N, 12°E), showing RS demarcation

persistence. 213
Map data: Google, DigitalGlobe. Image © 2013 DigitalGlobe.
18.2. Ancient Rome at its greatest extent under Trajan (c.117 ). 216
18.3. Roman system of rectangular demarcation. 220
18.4. Roman (left, a) and US (right, b) rectangular
demarcation systems. 222
18.5. Depiction of a groma and mensor (surveyor). 225
18.6. Documented location of centuriae in ancient Rome. 235
Source: Museo della Centuriazione.
18.7. Roman centuriation in Italia. 236
18.8. Hypothetical Roman centuriation along a river. 238
18.9. Carthage, Tunisia (36°N, 10°E), showing RS demarcation
persistence. 242
Source: Google Earth.
19.1. Choice between personal and impersonal exchange
to maximize social value of productive resources. 253
Source: Adapted from Arruñada and Garoupa (2005: 717, fig. 2).
19.2. Hypothesized evolution of titling when, after distant trade
becomes more common, mancipatio becomes costlier and less
effective and is replaced by traditio. 272
Source: Adapted from Arruñada and Garoupa (2005: 717–18,
figs 2 and 3).
22.1. Deterring wrongdoing. 358
Whilst every effort has been made to secure permissions, we may have failed in
a few cases to trace the copyright holders. If contacted, the publisher will be
pleased to rectify any omissions at the earliest opportunity.
List of Tables
14.1. Manumission terms in the Digest (# of fragments). 52

14.2. Some Roman-Egyptian slave prices (in silver dr.). 62
14.3. Patron rights and the effects of manumission. 67
Adapted from Loreti-Lorini 1925.
14.4. Claims to freedman inheritance (Gai., Inst. 3.40–3.44
and 3.56; Ulp., Epit. 29.1–29.3). 70
17.1. Probable birthplaces of forty-one prominent Roman writers
and teachers born 200 – 200. 177
18.1. History of ancient Rome. 217
18.2. Roman measurements with US equivalents. 219
Source: Dilke 1971; Campbell 2000.
18.3. Characteristics of selected Roman centuriae in Italia. 234
Source: see text.
24.1. Aedilician and praetorian remedies. 409
24.2. Buyers’ and seller’s valuations of the goods. 412
Whilst every effort has been made to secure permissions, we may have failed in
a few cases to trace the copyright holders. If contacted, the publisher will be
pleased to rectify any omissions at the earliest opportunity.
List of Contributors
Barbara Abatino, Research Fellow in the Faculty of Law at the

University of Amsterdam, the Netherlands.
Jean Andreau, Director of Studies at the School for Advanced Studies
in the Social Sciences (EHESS), France.
Benito Arruñada, Professor of Business Organization at Pompeu
Fabra University, Spain.
Giuseppe Dari-Mattiacci, Alfred W. Bressler Professor of Law at
Columbia Law School, USA.
Robert C. Ellickson, Walter E. Meyer Professor Emeritus of Property
and Urban Law and Professorial Lecturer in Law at Yale Law
School, USA.
Richard A. Epstein, Laurence A. Tisch Professor of Law at NYU
School of Law, the Peter and Kirsten Bedford Senior Fellow at the
Hoover Institution, and the James Parker Hall Distinguished Service
Professor Emeritus of Law at the University of Chicago Law
School, USA.
Iole Fargnoli, Professor of Roman Law at the University of Bern,
Switzerland, and the University of Milan, Italy.
Robert K. Fleck, Professor, John E. Walker Department of Economics,
Clemson University, USA.
Andreas Martin Fleckner, Professor of Private Law, Roman Law, and
Commercial Law at Humboldt University of Berlin, Germany.
David Friedman, Professor of Law at the Santa Clara University
School of Law, USA.
Henry Hansmann, Oscar M. Ruebhausen Professor of Law at Yale
Law School, USA.
F. Andrew Hanssen, Professor, John E. Walker Department of
Economics, Clemson University, USA.
xviii List of Contributors

Ron Harris, Kalman Lubowsky Professor of Law and History at Tel
Aviv University, Israel.
Dennis P. Kehoe, Professor, Department of Classical Studies and
Andrew W. Mellon Professor in the Humanities (2010–13) at Tulane
University, USA.
Egbert Koops, Professor of Legal History at Leiden University, the
Netherlands.
Reinier H. Kraakman, Ezra Ripley Thayer Professor of Law at Harvard
University, USA.
Gary D. Libecap, Distinguished Emeritus Professor of Economics at
the University of California, Santa Barbara and a Research Associate
at the National Bureau of Economic Research, USA.
Luuk de Ligt, Professor of Ancient History at Leiden University, the
Netherlands.
Elio Lo Cascio, Professor Emeritus of Roman History at the Sapienza
University of Rome, Italy.
Dean Lueck, Director of the Program on Natural Resource
Governance at the Ostrom Workshop, Professor of Economics and
Affiliate Professor of Law at Indiana University Bloomington, USA.
Barbara Luppi, Assistant Professor of Economics at the University of
Modena and Reggio Emilia and Adjunct Professor of International
Economics at the Bologna Institute for Policy Research, Italy.
Thomas J. Miceli, Professor of Economics at the University of
Connecticut, USA.
Geoffrey Parsons Miller, Stuyvesant P. Comfort Professor of Law at
NYU School of Law, USA.
Francesco Parisi, Oppenheimer Wolff and Donnelly Professor of
Law at University of Minnesota Law School, USA, and Professor
of Economics at the University of Bologna, Italy.
Daniel Pi, Visiting Professor of Law at Mitchell Hamline School of
Law, USA.
Eric A. Posner, Kirkland and Ellis Distinguished Service Professor of
Law at the University of Chicago Law School, USA.
List of Contributors xix

Aldo Schiavone, Professor of Roman Law at the Università di Roma
La Sapienza, Italy.
Richard Squire, Alpin J. Cameron Professor of Law at Fordham
University School of Law, USA.
Peter Temin, Elisha Gray II Professor Emeritus of Economics at
MIT, USA.
Hendrik L. E. Verhagen, Professor of Private International Law,
Comparative Law and Roman Law and Society at Radboud University,
the Netherlands, and attorney-at-law at Clifford Chance (Financial
Markets, Amsterdam).
List of Abbreviations
C. Codex of Justinian
CIL Corpus Inscriptionum Latinarum
C.Th. Theodosian Code
D. Digest of Justinian
FIRA Fontes Iuris Romani Antejustiniani
Gai., Inst. Gaius, Institutes
Inst. Justinian, Institutes
Standard abbreviations are used for inscriptions, papyri, and other legal texts
are used, as well as for ancient literary sources.
12
Rome and the Economics

of Ancient Law II
Geoffrey Parsons Miller
The first volume of this collection enlists economic theory and

methodology to explore the legal and economic system of ancient
Rome—the evolution of its constitution, the organization of its mar-
kets and trading relationships, and the legal form of its business
enterprises. The contributions to that volume illustrate the fruitful
potential of legal-economic theory for shedding light on the institu-
tions of the ancient world, and in particular for enhancing our
understanding of the legal and economic arrangements found in the
Roman Republic and Roman Empire.
The Introduction to the first volume discusses the essential prem-
ises of economic analysis of ancient law, which need only be sum-
marized here. The essential points are twofold. First, legal-economic
analysis provides a useful approach to understanding economic activ-
ity in the ancient world. Such analysis starts with the premise that,
while we must take into account basic social factors affecting eco-
nomic choices, people in ancient societies tended to use rational
means to pursue the most important things that they valued, such
as status, prestige, wealth, security, and the chance for a better life for
their children and loved ones. This basic assumption provides a
starting point for analyzing how people in the Roman world
responded to various incentives as they pursued their life goals.
Second, at a sufficient level of abstraction, the problems of social
organization facing jurists, law-makers, and business people in
Geoffrey Parsons Miller, Rome and the Economics of Ancient Law II In: Roman Law and Economics: Exchange,
Ownership, and Disputes, Volume II. Edited by: Giuseppe Dari-Mattiacci and Dennis P. Kehoe, Oxford University Press
(2020). © Oxford University Press.
DOI: 10.1093/oso/9780198787211.003.0012
2 Geoffrey Parsons Miller

ancient Rome were similar to those facing their counterparts of today.
The society of ancient Rome, like societies today, faced the challenge
of facilitating cooperative behavior while deterring exploitation and
opportunism. Roman jurists and law-makers, like jurists and law-
makers today, sought to facilitate transactions that enhance value,
limit risk, and allow people the means to live a decent life under
conditions of reasonable economic security. In developing norms and
rules to achieve these ends, legal systems both ancient and modern
create incentives for people to act in value-enhancing ways and to
refrain from acting in harmful ways.
This second volume carries forward the economic analysis of
Roman legal and economic institutions. A fundamental institution
in the Roman economy was slavery. Aldo Schiavone’s chapter, “Law,
Slaves, and Markets in the Roman Imperial System,” observes that the
Roman imperial economy was fundamentally dependent on slavery.
“[W]ithout the presence of slaves in agriculture, in manufacturing, in
service functions,” Schiavone declares, “economic and administrative
life of the empire would never have reached the levels we know.”
Roman law dealt comprehensively with the rights, duties, and obli-
gations pertinent to slaves: their servitude and manumission; their
purchase and sale; their obligations to their masters; their rights to
own property and to make contracts; their disabilities before the law;
and much else besides. The author observes that Roman jurists did
not succeed in conferring legal recognition on slaves but only man-
aged to rationalize and improve the functional efficiency of slavery as
a means for enriching the slaveholding class.
Manumission forms the topic of Egbert Koops’s chapter, “The
Practice of Manumission through Negotiated Conditions in Imperial
Rome.” Koops observes that contracts for the manumission of slaves
were common in Roman law. Slaves were able in many cases to
acquire sufficient wealth to purchase their freedom and negotiate
contracts for manumission. Koops infers that the price of manumis-
sion was not overwhelming in many cases: a slave might be able to
purchase freedom within a period of years rather than decades.
Manumission in the author’s account was a transaction that advanced
the economic interest of slave owners: the prospect of manumission
induced diligent and trustworthy efforts by slaves who hoped for their
freedom, fostered cooperation and subservience from other slaves,
and promised the master continued benefits flowing from the acqui-
sition of a freedman.
Rome and the Economics of Ancient Law II 3

Credit was a crucially important feature of the Roman economy.
Jean Andreau, “Banking, Money-Lending, and Elite Financial Life in
Rome,” examines the development of credit laws and markets from
the last centuries of the Republic until the fourth century of the
Common Era. Andreau notes that lending transactions were com-
monplace throughout the history of ancient Rome. Particularly
enlightening is the author’s description of Roman banks, which in
some respects appear to have operated in a manner not so dissimilar
from banks of today.
Hendrick L. E. Verhagen’s “Secured Transactions in Classical
Roman Law” offers an important and controversial reinterpretation
of the law applicable to secured lending transactions. The conven-
tional view is that the Roman rules on credit were awkward and
cumbersome, and that, for this reason among others, the most
important source of credit enhancement was personal security such
as guarantees. Drawing in part on an archive of professional Roman
money-lenders, the author argues that this view is mistaken. Far from
being legally disfavored, loans on real security were flexible financing
arrangements that played an important role in the Roman economy.
The author’s analysis is informed by a sophisticated understanding
of the economic principles underlying secured lending, including
concepts such as monitoring, bonding, adverse selection, and moral
hazard.
Robert C. Ellickson’s chapter, “Ancient Rome: Legal Foundations
of the Growth of an Indispensable City,” turns to Roman urbaniza-
tion, and in particular to the role of Rome as the capital of a great
empire. Ellickson takes issue with the view of economic historian
Moses Finley, who claimed that the enormous growth in population
and prosperity of ancient Rome occurred at the expense of subject
populations elsewhere. Finley’s depiction of Rome as a parasite of the
Empire is fundamentally mistaken, in Ellickson’s view. Far from being
parasitic, Rome exported many benefits to areas under its dominion,
including government, law, literature, access to financial and capital
markets, and the security and peace that came with being part of a
great empire. Meanwhile, Rome’s enormous growth can be explained
as a function of wise economic policy and sound institutional foun-
dations. In particular, the government refrained from engaging in
populist growth-limiting policies, but instead allowed the private
market to provide housing for the population, supported by
government-provided public services such as aqueducts. Ellickson’s

analysis of Rome’s policies towards land development suggest that
cities of today might have much to learn from their ancient forebear.
Economic historians Gary D. Libecap and Dean Lueck carry for-
ward the economic analysis of land law in their chapter, “Land
Demarcation in Ancient Rome.” As in the case of Ellickson’s chapter,
these authors find much to admire in Roman administration of land
policy. In particular, they study centuriation, a system of land demar-
cation with decidedly modern elements based on systematic survey-
ing and mapping of territory. Libecap and Lueck model a decision
facing Roman administrators of new territories: whether to maintain
the existing demarcation systems inherited from the previous rulers,
or rather to implement the new “rectangular” system. They argue that
despite its higher up-front costs, the rectangular system offered
superior benefits by virtue of its efficacy at reducing boundary con-
flicts, facilitating the development of roads and canals, fostering of
new settlement, lowering the costs of transactions, and improving the
identification of land parcels.
Benito Arruñada’s “The Institutions of Roman Markets” highlights
the role of Roman law in facilitating market transactions. Impersonal
markets work best when participants are protected against idiosyn-
cratic risks due to features of counterparties that are difficult to
observe or control. On the one hand, to afford such protections, the
law must dispense with the strict enforcement of rights in property
(otherwise all transactions would be subject to potentially unknown
claims of prior owners). On the other hand, if property rights are not
well protected, people will not invest optimally in productive activ-
ities. The law must seek to find the right balance between enforce-
ment of property rights and the facilitation of impersonal
transactions. Arruñada argues, based on a subtle and nuanced ana-
lysis, that the Roman state and Roman society developed market-
enabling institutions that, while not impersonal in the modern sense,
at least allowed for a wider scope of transactions than would have
been possible under a regime of strict enforcement of property rights.
No legal system can be effectively administrated if the rules are not
enforced. Richard A. Epstein’s contribution, “One Step at a Time in
Roman Law: How Roman Pleading Rules Shape the Substantive
Structure of Private Law,” offers a favorable view of how the Roman
legal system dealt with problems of enforcement in the courts.
Extending on his impressive prior work in this field, Professor Epstein
cogently argues that the clear classifications of Roman pleading

provided a mechanism for improving the precision of doctrines of
delict, contract, property, and restitution, and by this pathway facili-
tated the creation of efficient substantive rules. Epstein is alert to
the implications of his analysis for contemporary public policy: mod-
ern systems of pleading, he suggests, may look to Roman models
when searching for ways to improve the efficiency of adjudicatory
proceedings today.
David Friedman’s chapter, “Private Prosecution and Enforcement
in Roman Law,” underscores the private element in Roman adjudi-
catory contestation. Friedman points to evidence that Roman pro-
cedures evolved out decentralized systems of privately enforced
law. Drawing on the analogy to the feud system in saga-period
Iceland—a subject Friedman has analyzed in prior work—the author
here points to holdovers of an earlier regime of private enforcement
in the rules of Roman law. These include the use of damage payments
as sanctions for crimes; the delegation of the enforcement of judg-
ments; the use of sureties to enforce legal rights; and the rule of
requiring disputing parties to deposit the equivalent of bonds for
the benefit of the winning party. Even in the imperial period, when
the task of enforcement shifted increasingly to state actors, private
citizens remained principally responsible for the prosecution of
both civil and criminal offenses. Friedman’s analysis of the historical
antecedents of Roman procedure is necessarily conjectural, given
the paucity of historical evidence, but it does appear to be a
plausible reconstruction based on the residues and traces found in
later law.
Modern economic analysis recognizes that the principal value of
legal rules is not their efficacy in rectifying wrongs, but rather their
effect at deterring people from engaging in wrongdoing in the first
place. Francesco Parisi, Daniel Pi, Barbara Luppi, and Iole Fargnoli’s
“Deterrence of Wrongdoing in Ancient Law” addresses this issue.
Like Friedman’s contribution, which traces Roman procedures to a
hypothesized origin in private enforcement regimes, these authors
offer a dynamic account of the growth of Roman (and other) law out
of earlier antecedents. In their case, the methodology of choice is a
formal economic model exploring the growth of individual liability
principles out of an earlier framework of lex talionis grounded in
biologically based desires for revenge.
Thomas J. Miceli’s chapter, “Collective Responsibility,” investigates
the topic of punishment of a group for misconduct by one of its

members. Collective responsibility was exemplified in the Roman
context by the doctrines of quasi-delicts and noxal liability. The
author models the choice between individual and collective liability
as reflecting a tradeoff between the costs of failing to punish the true
offender and the costs of erroneously punishing an innocent person.
The general (although not total) elimination of collective responsi-
bility from modern legal systems, the author suggests, is not so much
due to the inherent superiority of individual responsibility under all
circumstances, but rather to the fact that under conditions prevailing
in the modern world, the benefits of collective responsibility are
almost never sufficient to overcome its costs. This interesting article
provides a useful corrective to conventional value judgments, and
illustrates the potential of the economic analysis of ancient law to
place contemporary beliefs and attitudes in a more comprehensive
perspective.
Barbara Abatino and Giuseppe Dari-Mattiacci’s study, “The Dual
Origin of the Duty to Disclose in Roman Law,” takes on a topic with a
decidedly modern resonance: the legal obligations of parties to dis-
close private information to their counterparties in contracts of sale.
The issue is salient for contemporary commercial and business law—
consider the disclosure obligations of merchants under state unfair
trade practices laws; rules against trading on confidential information
in stock markets; and SEC requirements for public company reporting.
It is a testament to the sophistication of Roman law that disclosure
obligations in commercial transactions were also regulated under that
body of rules. Abatino and Dari-Mattiacci note that Roman law
imposed such obligations under two institutions: the aediles, pertain-
ing to the context of private market transactions effected through
auctions; and the praetor, covering private transactions in general.
The remedies differed in important respects: for example, the aedili-
cian remedies included both restitution and enforcement of the
contract plus price reductions, while the praetorian remedy required
the buyer to affirm the contract but allowed offsetting claims of
damages. Abatino and Dari-Mattiacci argue that both remedies
were efficient in their sphere of application: the aedilician remedies
maximized the value of market transactions between professional
sellers and private buyers in organized markets; whereas the praetor-
ian remedy maximized the value of transactions between private
parties outside of organized markets. Like other contributions to
this volume, this chapter by Abatino and Dari-Mattiacci illustrates

the potential of economic analysis to shed light on institutions of
ancient law that resist analysis through traditional doctrinal
approaches.
This second volume of Roman Law and Economics carries forward
and further validates the promise of its predecessor. In its various
chapters, we observe the lived reality of an ancient legal system,
sensitively analyzed and illuminated by principles of economic ana-
lysis. Taken as a whole, these studies demonstrate that economic
analysis has enormous potential to enhance our understanding and
appreciation of Roman legal institutions.
13
Law, Slaves, and Markets in the

Roman Imperial System
Aldo Schiavone
The Roman Mediterranean Sea is the historical and natural

background of the most advanced western economy before the age
of capitalism—a system in which slavery is both a factor of strength
and a limit to any further growth. This contradiction is also inherent
in Roman law. On the one hand, Roman law anticipates some of the
fundamental characteristics of the private and commercial law of
modern Europe: the centrality of consent, freedom of contract, reci-
procity, fairness, and good faith. On the other hand, Roman law
unsuccessfully tries to overcome the classical form of chattel-slavery
and ends up reinforcing its presence in Roman culture and the
broader economy. The chapter supports this point of view through
the study of Roman legal texts from first century BC to first century AD.
13.1. LAW, ECONOMY, AND POWER
The Roman Mediterranean between the end of the second century BC

and the beginning of the third century AD formed the geographic
backdrop for the most vast and complex economic system ever to
have existed before the Industrial Revolution and the definitive cap-
italistic lift-off in the second half of the eighteenth century. Notwith-
standing the qualitative changes that had already taken place in the
Aldo Schiavone, Law, Slaves, and Markets in the Roman Imperial System In: Roman Law and Economics: Exchange,
Ownership, and Disputes, Volume II. Edited by: Giuseppe Dari-Mattiacci and Dennis P. Kehoe, Oxford University Press
(2020). © Oxford University Press.
DOI: 10.1093/oso/9780198787211.003.0013
12 Aldo Schiavone
fabric of the European economies since the late medieval period, it
was only in seventeenth-century Holland or, possibly, about a century
before, the Spanish empire of Philip II, that we find levels of produc-
tion performance and wealth accumulation no longer comparable to
those of the Roman world.¹
The sea was the chief protagonist of ancient commerce, and the
rivers and winds along with it. Water was the only means of convey-
ing large quantities of bulk goods over long distances. “On ships with
swift sails he often traveled the great sea”: thus describes an epitaph
from Brundisium one merchant’s life.² The soft technology of mari-
time transport (hulls, sails, rudders, knowledge of winds and cur-
rents) won out over the hard one of overland freight (wheels, carts,
yokes and beasts of burden, and difficult and costly road-building).
Once again, the ancient economy seems to have been more
dependent on geography and anthropology than on history. Its mer-
cantile side involved almost exclusively the coastal areas. When,
beginning at the end of the second century AD, the axes of the empire
in the west shifted toward the interior of Europe, the commercial
networks were thrown into disarray: the fragile threads of large-scale
exchanges tended to fray and break (though we must not underesti-
mate the persistence of Mediterranean trade in late antiquity, as in the
high Middle Ages), and the traditionally dominant agrarian economy
began to take on overtones of feudalism.
Roman law would not have been what it is if its course had not
become entwined first with the development and then with the
maintenance of this Mediterranean and tricontinental network of
production and markets: in a certain sense, the first “world economy”
in our history. Historians of law have almost always forgotten this
very elementary truth, led as they are by a long tradition to recon-
struct legal forms as if they developed in a world apart, far removed
from society, economics, and politics; but such an obstinate conceal-
ment in no way lessens the absolute evidence of the connection.³
It is less easy, however, to identify the nature of this relationship.
One should not in fact imagine any unilinear causality, in one
¹ The economic history scenarios outlined in the text presuppose my work in

Schiavone (2000).
² Corpus Inscriptionum Latinarum IX 60: see Giardina (1993: 261). See also the
book of Temin (2013).
³ See my work in Schiavone (2012).
Law, Slaves, and Markets in the Roman Imperial System 13

direction or another: from economics toward law, or from law toward
economics. It is necessary instead to hypothesize a complex reci-
procity of interactions, where every innovation—in the field of law
or economics—was at once cause and effect of further developments
occurring on both fronts.
Without fear of exaggeration, we can safely say that law was an
invention of the Romans: in the sense that they were the first to isolate
the legal function as a specific technique of social regulation, detach-
ing it from every other kind of cultural production or institutional
apparatus—from religion, from morals, from politics itself—and
identifying it instead with an entirely autonomous field, with its
own principles and structure: the world of ius.
At the center of this universe were the jurists (a word unknown to
any ancient language, except for Latin), who, at least in the centuries
of Rome’s imperial heyday, were not just the adepts and interpreters
of ius, but, in large part, its creators as well: a body of experts who
labored for dozens of generations, moving along a previously
untrod path.
Admittedly, besides the work of the jurists there was in Rome, in
that period, also a significant presence of legislation—from the
Twelve Tables, still in the archaic age, to the laws passed by the
people’s assemblies in republican and Augustan times, through to
the normative provisions decided by the emperors during the Prin-
cipate. And a no less important role was played, in the same epochs,
by the praetor’s edict: a text formed through an uninterrupted strati-
fication that had involved, year after year, hundreds of magistrates
responsible for civil jurisdiction (the praetors).
Yet this many-stranded whole was never concretely operative
outside the thick weave of prescriptions, interpretations, innovations,
and amendments that would be built around it by jurisprudence.
“Law cannot exist unless there is some jurist who can improve it
from day to day,” wrote Pomponius in the Enchiridion (in D.1.2.2.13)
around the middle of the second century, when that model had
already begun to slip into crisis.
What we generally think as “Roman law” is therefore first and
foremost a construction of the jurists. A “living law of custom,”⁴ case
based and guided by experts: much closer in its structure, in its
⁴ The expression comes from Carl Schmitt (1950, who epitomized Savigny), and it
is used in reference to English law. See also Buckland and McNair (1952).
14 Aldo Schiavone
mature configuration, to modern English and (in a number of ways)
American law, than to French law after the codification under
Napoleon, or to Italian law subsequent to the codification of 1865.
In the epoch of the great conquests—between the second century
BC and the age of Augustus—the collaboration between jurists and
praetors gave rise to a kind of “Mediterranean trading law,” which
immediately became a decisive instrument of Roman hegemony. The
imperial economy never built a unitary structure: nothing like the
capitalistic unification of contemporary economies. Rather, it
involved a constellation of diverse systems, determined much more
by the history and geography of the different places than by the force
of the impact and presence of Rome. But the empire, through the
lever of tax levies, and due to the initiatives of avid, unscrupulous, and
enterprising groups of merchants, managed to create an unprece-
dented network of trade exchanges and wealth redistribution mech-
anisms, which contributed in a decisive way to stabilizing and
maintaining Roman power. This network required a framework of
relations that it would have been impossible to build and consolidate
without the experience of Roman law, which gave that context, albeit
through many mediations, the certainty of a juridical foundation
capable of protecting and regulating, in a rigorous and specific man-
ner, property expectations, single and group interests, dealings,
markets. In turn, the presence of such an extensive and developed
economic fabric offered an inestimable stimulus for legal thought—
an almost endless supply of different cases and situations—and,
above all, limitless proof of an essential lesson: that to govern so
very diverse spaces and peoples, consensus was much more necessary
than armies, and law, then, more than force.
But the picture was in reality even more complex. A third element
present on the scene—besides law and economy—should also be
borne in mind: politics. In fact, the functioning of a Mediterranean
economy presupposed, in turn, imperial political unification: in the
ancient context, it was almost always politics that underpinned eco-
nomics, and not vice versa, as we are accustomed to seeing in the
modern world. And what has just been said of economics holds true
for politics as well. The institutional construction of the empire would
have been inconceivable if a decisive role had not been played by
Roman law. The latter, in turn, would never have reached the degree
of refinement and formalization that we know, and that enabled it to
perform its essential function, if its development had not been

constantly accompanied by an unprecedented accumulation of
political power, concentrated in the circle of the same elites that
were elaborating the law—if the great imperial legal culture had not
taken root in the shadow of a world power. The growth of political
strength, economic successes, and the disciplinary capacity of law—
power and order—were inextricable strands of the same hegemonic
construction. The birth of a new paradigm within Roman legal
thinking, which transformed law into a science, changing the quality
of its precepts from acts of authority into acts of knowledge—an
epistemological revolution destined to mark the history of the West
from then on—would have been impossible if it had not occurred
within this intertwined web: a picture about which historians still
have a lot to discover, and a lot to recount.⁵
Roman jurisprudential law was an order with a fluid, changing
surface, constantly moving in the attempt to grasp as closely as
possible the social reality it faced and had to regulate. The shared
core of a scientific and preceptive paradigm that had been preco-
ciously developed was constantly riven by polemics between different
jurists, disagreements and divergent solutions.
Partly due to this intrinsic instability, and for many other reasons
besides—both political and cultural, as well as more strictly technical
and practical—the Roman case law shaped by the jurists was never
really the law of the whole empire. It was not the law of the whole
empire in a “horizontal,” geopolitical sense, because wherever pos-
sible Roman imperial praxis envisaged the regulatory independence
of the local communities, which were to govern themselves “accord-
ing to their own laws and their own traditions”—suis moribus legi-
busque suis uti.⁶ Nor was it in what might be termed a “vertical,”
social sense, because Roman law almost always remained the law if
not just of the elites, then at least of the property-owning orders—in
Rome itself, in Italy and in those provincial contexts where Roman-
ization had penetrated most strongly. It never became, in other
words, if not following heavy mediation, the law of the vast, outlying
territories under Roman domination, from the Red Sea to the Atlantic
Ocean; nor, by the same token, was it the law of all layers of the
population, of the entire social body.
⁵ See Schiavone (2012: 245).

⁶ Gellius, Noctes Atticae 16.13.4, referring to Hadrian’s oratio ad Italicenses.
16 Aldo Schiavone
The law of the Roman jurists always remained an extremely
circumscribed model, in terms of its effective range of application,
but carried great prestige and a potent charge of exemplarity: a bright,
luminous center surrounded by a myriad of minor and local legal
systems, sometimes lacking any formal recognition, which assumed
Roman law as a distant and unattainable point of reference that still
had to be taken into account, even if only in a rough and ready
manner.
The activity of Roman jurisprudence had developed entirely within
the heart of this economic and political supremacy, and it would be
incomprehensible now if we did not trace it back to the everyday
exercising of a hegemony that it helped so greatly to construct. And
the jurists themselves frequently became embroiled in the manage-
ment and the problems of a world government. In the republican age,
insofar as they all belonged to the nobility, they were also at the head
of the Republic. During the Principate, first as advisors to the prin-
ceps, then directly as high-ranking functionaries of the imperial
administration (the entire history of Roman legal thought can be
drawn together around three pictures, each with the same protagon-
ist, though with different dress: that of an archaic priest; of a repub-
lican nobleman; and finally, of a great specialist at work in the milieu
of the princeps and the court). But the science of the jurists seems
never to have left the slightest trace of such involvement in the
governing of the empire, and they always managed to avoid becoming
shackled by too close a bond between legal learning and political
action. On the contrary, they were capable of defending the isolation
of their knowledge, rebalancing it as a technique that was self-
legitimating well beyond the underpinnings of political power, even
though entirely constructed right next to it. And in turn, politics
would always find itself in need of alliances and compromises with
jurisprudence, in which the relative equilibrium of powers was deter-
mined by the conditions prevailing at any given time.
13.2. THE SLAVE SYSTEM
The Roman imperial economy had a trait that makes it quite dis-
tinctive and totally recognizable: it was a slave economy, like the one
of classical Athens between the fifth and fourth centuries BC, or that of

the Southern U.S. states before the Civil War. It is hard now to doubt
the evidence of this, though often in the past “the ideologies of the
moderns” (to use an expression of Finley 1980)⁷ have tried to mask
this incontestable truth, and many have worked to soften history and
to prevent the image of the classical world—partially invented by
themselves in libraries—from being sullied by such a stain.
But even just from a quantitative point of view, the slightest hesita-
tion would be unthinkable. There are grounds to sustain that in
the years of Augustus, at the end of the first century BC, perhaps
over 30 percent of the entire population of Italy was made up of
slaves—around three million in total, a massive figure and a very high
proportion, maintained for over a century. And this is without mention-
ing the incidence of slave labor on the structure and performance of the
production fabric: without the presence of slaves in agriculture, in
manufacturing, in service functions, the economic and administrative
life of the empire would never have reached the levels we know.
For sure the situation outside Italy was different, and we cannot
expect such figures to be matched in the peripheral areas as well. But
the presence of a considerable slave population can be clearly traced
in Sicily, Cisalpine Gaul, Africa, in the Spanish and Greek mining
regions, in Gaul, especially in the southern areas, and almost all of the
eastern provinces. This was especially true of the large coastal cities:
no less than one hundred thousand slaves lived in Alexandria
(around 25 percent of the local population). And in Antioch in the
fourth century AD some wealthy citizens still possessed between one
and two thousand slaves each. Nor have we any reason to think that
the picture was very different in Carthage or Marseilles.⁸
Slavery was not just an ancient phenomenon. But modern slavery
was entirely colonial in origin—in Brazil, in the Caribbean, in
the American South—and it became established for particular
reasons (the scarcity of labor in the New World, accompanied by
the European penetration along the coasts of West Africa) only in
rural environments that were relatively peripheral with respect to the
increasingly manufacturing-based and industrial center of the new
European and Atlantic economy. New-World slavery always had
⁷ See also Andreau and Descat (2006: especially 65ff., 107ff.). A vast and accurate
bibliography (until the beginnings of the 2000s) in Bellen and Heinen (2003).
⁸ Schiavone (2000: 112–13). See also Bradley (1994: esp. 10ff., 57ff.); Turley (2000);
and Thompson (2003).
18 Aldo Schiavone
problems of compatibility in relation to a different and much more
expansive and dominant mode of production—both agricultural and
industrial—founded on free labor. It was, in short, a recessive figure
in comparison to the heart of modernity.
In the ancient civilizations, by contrast, in classical Athens as in
imperial Rome, the slave system was by far the most developed and
dominant economic form in terms of results and organization: the
real propulsive core of the whole Mediterranean economy. And it
never basically had any alternative, either theoretical or practical.
From Plato to Aristotle and through to the Roman jurists of the
Severan age, it was constantly viewed as a social necessity, a universal
principle of the civilizing of humanity, as something that could not at
any rate be questioned.
Indeed in this respect Roman thinking—both philosophical and
legal—represented an undoubted novelty in comparison to the Greek
tradition. Rather than insisting on the truth of an ontological differ-
ence that, “by nature” (kata physin), divided the human species into
those who were born to command and those to obey, as sustained in a
celebrated passage from Aristotle’s Politics,⁹ Roman thought, at least
from Cicero onwards,¹⁰ preferred to abandon the rigidity of this
model and turn to a justification of slavery that we might call socio-
logical and historical rather than fully naturalistic: at any rate better
suited to the reality of the slave system it was faced with, distinguished
not only by the annihilation of personal identity and harsh repres-
sion, but also (as we shall see) by the valuing of the abilities and
talents of slaves, and the regular concessions of freedom.
It involved, in other words, an adjustment dictated by imperial
reality: many slaves arrived from regions, like those of the eastern
Mediterranean, that were culturally more advanced than the society
of the conquerors, and speaking of an ontological diversity destining
them to slavery would have been embarrassing. Unlike the Greek
model of slavery, the Roman one never, or almost never, had specif-
ically racial overtones.
This fully explains how jurists between the second and third
century AD could affirm that men should be classified as either free
or slaves (and from the legal point of view, it was as if slaves did not
exist, being closer to things than to human beings), but at the same
⁹ Arist., Politics 1253b–1255a. ¹⁰ Cic., De rep. 3.25.37.


time that slavery did not rest on natural law (in relation to which they
even went so far as to say that “all men are equal”), but only on that
“of the peoples”:¹¹ it was, that is, the result of a unanimously accepted
convention, not the reflection of a natural datum.
In any case, the idea of a society without servile labor formed no
part of the ancient Mediterranean cultures. Nor did the great slave
revolts of the second and first centuries BC (in Sicily and in Italy,
including that of Spartacus) set themselves such an aim.¹² They just
sought to overthrow local power arrangements, and to exact ven-
geance on inhuman masters, not to uproot an overall system. Neither
philosophical thought, nor political or legal experience offered any
points of reference. And, what is more, in all the criticism raised
about the uncontrolled spread of chattel slavery and the harsh con-
ditions it entailed, the suggested alternative looked to an archaic and
idealized past, rather than to the present or to the future. It just
envisaged the return to a world of domestic autarchies, with a softer
form of slavery, numerically limited and “according to nature,”
incorporated within the family structures of former times—visions
fed solely by literary commonplaces concerning the revival of the
virtuous poverty of the ancestors.¹³ Such models were unable to do
anything other than create a yearning for archaizing restorations, in
which the rejection of the excesses of slavery inevitably carried with it
an aversion towards large-scale trade, money, and mercantile net-
works. Basically, towards everything that was now an inerasable
feature of Roman civilization, delivering the economic performance
that underpinned the whole social structure of the empire: urban life,
mass consumption, military and administrative bodies, intellectual
creativity.
And so, after brushing up against a dangerous critical point with
the bloody insurrections of the previous one hundred years, the neo-
aristocratic stabilization that characterized Roman society with the
advent of the Augustan regime was accompanied by a no less impos-
ing stabilization of the imperial slave system. It thus consolidated its
dual and contradictory aspect as an indispensable requirement for
¹¹ D. 1.1.4 (Ulp. 1 Inst.) and D. 50.17.32 (Ulp. 43 ad Sab.); D.1.5.4.1 (Florentinus 9

Inst.); D.12.6.64 (Triphoninus 7 disp.).
¹² Schiavone (2013).
¹³ I am referring in particular to the thought of Posidonius, as inferred from Athen.
6.84.263c–d = FGrHist 87 F8 = Edelstein-Kidd F 60.
20 Aldo Schiavone
maintaining the levels of intellectual and material life achieved by that
society, and at the same time as an insuperable barrier to any further
development, which would instead have presupposed a massive
spread of free labor, as would occur in the modern economies. Slavery
underpinned that world and yet thwarted it.
The terrible coercion that relentlessly accompanied the existence of
millions of slaves was never substantively eased, nor was the ferocious
discipline to which they were subjected. A measure introduced at the
behest of Augustus himself reaffirmed, for example, that in the event
of the homicide of a master, all the slaves living “under the same roof ”
as him were to be tortured and put to death, because, as a great jurist
Ulpian would explain two centuries later, “no house could otherwise
ever have been safe, if not by obliging slaves, on pain of death, to
defend their masters both from dangers arising within his home and
from without.”¹⁴ In AD 61, the intransigent application of this norm,
despite the disguised opposition of the emperor himself—Nero, prob-
ably influenced by Seneca—would lead to the summary execution, in
the heart of Rome, of four hundred slaves, including many women
and children.¹⁵
But such ruthlessness did not prevent a great many slaves, espe-
cially in the cities—hundreds of thousands perhaps, in the course of
the whole of Roman history—from becoming integrated into imperial
society, and not all just at the lowest levels, and from being accepted
not as an extraneous body, but as an active and vital element. Such a
tendency reveals the extraordinary Roman talent—as yet unequalled
in world history—for the unprejudiced assimilation of different cul-
tures and peoples, provided they participated in some way in the great
process of unifying the empire.
Terror and integration, then: these were the only apparently
contradictory pillars that underpinned the Roman slave system.
And both were analytically regulated by law.
The most significant form of integration consisted of manumissions
(manumissiones). Through them, every single master, without any
intervention from the political authorities, could give a slave his
liberty, making him a freedman (libertus): and this guaranteed him
¹⁴ D. 29.5.1 pr. and 26–7 (Ulp. 50 ad ed.). See also Tac., Ann. 14.40.1–44.4. About
the sc. Silanianum see Dalla (1994); the recent contribution of Harries (2013: 51ff.);
and Miceli in this volume.
¹⁵ Schiavone (2000: 108ff.).

at the same time Roman citizenship (it seems incredible, but has a
precise historical explanation). In the final decades of the Republic, the
practice became extremely widespread. Many masters preferred to
have free citizens bound to them by legal and economic obligations,
as freedmen were, rather than elderly slaves who had to be maintained.
And it was Augustus, once again, who had to establish that no more
than one hundred slaves could be manumitted in a will, though he left
it to the master’s discretion as to how many to free while he was alive.¹⁶
In the Rome of these years, tens of thousands of citizen had servile
origins. Later, in the age of Nero, there was talk in the senate, accord-
ing to Tacitus, that the majority of knights, not to mention many of the
senators themselves, had some blood deriving from a freedman.¹⁷
So there took shape an elastic and multiform space of cultural and
social intertwining, of relations and exchanges, shared by the world of
the free and by that of slaves. And it was precisely in this circuit that
the most lively part of the imperial economic system developed.
13.3. LEGAL THOUGHT
In Roman jurisprudential texts from the time of Servius Sulpicius

Rufus, at the end of the Republic, through to the Severan authors at
the beginning of the third century AD, slavery occupies a position of
great prominence. This centrality appears entirely evident to us, even
though two important elements have long worked together to par-
tially obscure it.
The first concerns the tradition of the available documents. We
owe much of our knowledge about Roman legal thought to what has
been transmitted to us from Justinian’s Digest, compiled in the heart
of the Byzantine age, in the first half of the sixth century AD. In the
time and context in which the compilers were laboring on this great
collection—at once a code and an anthology, or rather an anthology
in the form of a code—slavery had become a much less crucial
phenomenon, socially and economically, than it had been in the age
of the ancient jurists at the height of the Roman Empire. It is therefore
¹⁶ With the lex Fufia Caninia of 2 BC, which covered all the various cases possible
with great precision (one hundred was the maximum): Gai., Inst. 1.42–3.
¹⁷ Tac., Ann. 13.27. See also Hopkins (1978: 115ff.).
22 Aldo Schiavone
reasonable to think that the Byzantine compilers, in their choice of
materials, tended to attenuate rather than emphasize the original
preponderance of references to slavery: a point of view that by now
appeared to them to be decontextualized and lacking in topical
concern. If, notwithstanding this filter, slaves still stand out so power-
fully in those ancient writings, it means that their invasive presence
had in any case become an integral and insuppressible part of juris-
prudential thinking: preserving essential nuclei from those works also
meant rendering their slave-based orientation inerasable.
The second element relates instead to our perspective as moderns.
For a long time European legal culture approached Roman law
with practical much more than strictly historiographical intent, the
objective being to extract from that legal order—updated where abso-
lutely necessary—rules for the age. In this massive effort of actualizing
interpretation, which reached a peak in nineteenth-century bourgeois
Europe, the original slave-oriented imprint no longer had any utility;
indeed, as far as possible it was worth concealing and passing it off as an
accidental and negligible aspect, highlighting instead what was sup-
posed to be the supratemporal dimension of that law, its capacity to
regulate very different and distant societies in the best possible way.
Yet despite this dual conditioning—first in the Justinianic selec-
tion, then in modern interpretation—the reality of things came to the
fore: every strand of Roman law was shot through by the presence of
slavery, which indeed constituted the chief condition of existence of
its most advanced component, both practically and conceptually.
Slavery disciplined mercantile dealings and the circulation (and accu-
mulation) of commercial capital. What is more, this observation
had already been made at the beginning of the twentieth century,
opening what can still be considered the most important book on the
theme in modern historiography: The Roman Law of Slavery by
W. W. Buckland, published in Cambridge in 1908 (and then
reprinted in 1970). In the Preface we can read: “There is scarcely a
problem which can present itself, in any branch of the [Roman] law,
the solution of which may not be affected by the fact that one of the
parties to the transaction is a slave, and, outside the region of pro-
cedure, there are few branches of law in which the slave does not
prominently appear.”¹⁸ We would not say anything different today.
¹⁸ Buckland (1908: v). See also Watson (1987).


13.4. PERSONAL DEPENDENCY
AND ECONOMIC AUTONOMY
Servius was the first of the great Roman jurists to cast an expert and
attentive eye on the economic reality of his time. We are in the middle
years of the first century BC, and imperial society now appeared to be
undergoing a radical transformation. We are in the heart of the
Roman slave-holding and mercantile boom, at the culmination of a
wave of expansion without precedent in the ancient economies: a
process of growth still tied to limits and contradictions that kept it
from creating the foundations for an authentic capitalist take-off, but
sufficiently powerful to achieve a level of performance that, as we
have said, was long unrivaled in the West: new farming estates, with
impressive standards of productivity; factories created exclusively for
the production of commodities; large quantities of slave labor; a
spectacular accumulation of public and private money; urban centers
supported by magnificent building programs; trade networks that had
formed an interdependent grid of markets extending across the entire
Mediterranean; opulent levels of consumption, though concentrated
among tiny elites.
The slave trade, in particular, was one of the most successful
commercial activities in the Roman Mediterranean, which was also,
no less than the Atlantic of the slave trade, a sea of slaves. In the
second and first century BC—the golden age of imperial slavery—a
high level of supply was maintained by the virtually uninterrupted
sequence of victorious wars of conquest (thousands upon thousands
of prisoners were sold to merchants following in the wake of the
troops, then resold at higher prices) and the indefatigable activities of
organized bands of pirate slave traders—the same ones who, having
become too powerful and a danger to navigation, were eradicated by
Pompey. The port of Delos (on the island sacred to Apollo) was
perhaps the largest marketplace; according to Strabo, up to ten
thousand prisoners could be sold there in a single day.¹⁹
The writings of Servius—who had been quaestor in Ostia at the
beginning of his political career, and was thus able to gain a very good
idea of what the trading reality of the empire had become—are the
mirror of this world. Unfortunately, Justinian’s compilers did not use
¹⁹ Strabo 14.5.2 (Meineke).

24 Aldo Schiavone
any work of his in assembling their collection (perhaps they no longer
even had a copy), and we cannot read anything by him through the
Digest. But the latter does include seventy-six texts taken from the
Digest of Alfenus Varus, the first jurist to write a work with that title,
and the most important, along with Aulus Ofilius, of Servius’ pupils.²⁰
Alfenus reproduced a great deal of Servius’ thought, enabling us to
regard it as a full-blown edition, with commentary, of the responses
of the master himself. And even if we have some difficulty in exactly
distinguishing each time between the commentary of the pupil and
the thought of the master, we can consider these documents to be
reliable testimony of Servius’ ideas.
At any rate, in at least twenty-six of these texts,²¹ at the center of
the economic issue underlying the legal problem, we find the presence
of a slave. This is a very high proportion: it is as if Servius had
discovered the pervasiveness and essentiality of slavery in the reality
of his own world.
There is no doubt that in Servius’ eyes a slave was above all else a
living machine, a thing, a commodity—a “speaking tool”²²—in keep-
ing with a tradition that dated back at least as far as Aristotle.²³ But in
the eyes of the jurists, this state of pure thingness, though accepted as
unquestionable, did not fully describe the condition of a Roman slave.
Alongside the institutional annihilation of their humanity (Quod
attinet ad ius civile, servi pro nullis habentur—“as far as the civil law
is concerned, slaves are regarded as not existing”—Ulpian would still
write in a passage of his Libri ad Sabinum),²⁴ for jurisprudence there
was also a need, dictated by the empirical rationality of the slave
system which the Romans had developed with particular attention, to
valorize the abilities and talents of the subjected, exploiting in the best
possible way potential and aptitudes that did not fit into the scheme
of their reduction to simple things.
In this way there opened up what seems to us to be a glaring
contradiction, but which legal thought, with a ductility that did not
fear incongruence but pursued only the optimization of results, did
not regard as such, but rather as a kind of opportune coexistence of a
dual slave regime: one centering on the retention of the bond of
personal dependence, which was never denied; and the other founded
²⁰ Lenel, Pal. I. 38–54. ²¹ Schiavone (2012: 522 n. 22).

²² As Varro put it in Res. rust. 1.17.1. ²³ Pol. 1253b; NE 1161b.
²⁴ See n. 11.

on the valorization of the labor force and the productive capacity of
slaves. In this respect, they were granted such a degree of autonomy in
their economic activities that they could almost be mistaken for free
men: a sort of permanent state of exception, valid as long as one
remained within the sphere of production and markets. It was an
entirely Roman peculiarity—nothing like it had been seen in Greek
slavery.
In the systematic scheme of Gaius (a jurist from the second century
AD, and the author of a manual destined to enjoy extraordinary
posthumous success), which was probably influenced by a scholastic
approach that may also have been marginal with respect to the
chief lines of development of the legal thought of his time, but is
in case revelatory for us of a widespread attitude and tendency, the
slave appeared to lie at an ambiguous point of intersection between
personae and res, between “persons” and “things”: an intersection that
marked the critical point of a whole civilization.²⁵ An intersection that
fixed the atypical and exceptional status of a “human thing” where two
different ontologies were bound up to the point of being indistin-
guishable, in a sphere of reciprocal influences without equal in ancient
civilizations. It was as if, in the jurists’ doctrines, slavery, with all its
power to annihilate identity, appeared not only to be an advantage—
according to the common notion of any slave system—that permitted
the total exploitation of the labor of the subjected at a very low cost, but
also ended up becoming, in a completely unexpected way, a brake as
well, a kind of obstacle to be overcome in order to permit that
combination of human resources and trade circuits which the classical
form of chattel slavery would never have made possible.
Servius was the first to conceptualize this phenomenon—
subjection in terms of “status,” but autonomy with regard to com-
mercial transactions—and to employ a formal device to account for
this particular condition. In a response, published by Alfenus Varus,
we read: “A person leased land to a slave of his for cultivation, and
provided him with some oxen. . . . ”²⁶ If we were to take as the sole
point of reference the relationship of dependency between master and
slave, this sentence would make no sense. The slave was totally subject
to the master’s will and so could not enter into any form of economic
²⁵ Gai., Inst. 1.8–9; 48–49; 52; 2.13. See Esposito (2014: 9ff.).
²⁶ In D.15.3.16 (Alf. 2 dig.): Quidam fundum colendum servo suo locavit et boves ei
dederat.
26 Aldo Schiavone
or legal exchange with him. And yet, both Servius and Alfenus took
the existence of such a relationship for granted, and in defining it,
they did not hesitate to use a technical legal verb, locare, that placed it
in the juridical scheme of the res locata. This was a conceptually
subversive qualification, in which slave and master figured as distinct
subjects, facing off against one another on a plane of formal equality.
Though restricted to well-defined limits, and only with regards to
transactions in progress, the formal determinations induced by the
legal models of exchange overlapped with and prevailed over those of
the relationship of dependency: “a person leased land to a slave of
his”—servo suo locavit. The contrast could not be expressed more
effectively; in these few brief words two opposite worlds really come
face to face: the discriminating one of status, and the inclusive one of
the contract. Servius’s idea was to construct in an analogical fashion,
as far as seemed possible, the economic relations between slave and
master as if they were between free men, on every occasion in which
the developments of the new economy, and the need to enhance the
value of the slave’s abilities, intervened to modify the typical and
traditional structure of the tie of dependency.
13.5. A DUAL REGIME
Servius’ account did not only speak about leasing. In the course of the
event described in the same text, we see the slave selling and buying,
receiving sums and failing to pay them, and becoming insolvent.²⁷
None of this was rare in that time. In the context of a broad and
detailed series of cases—carefully analyzed by jurisprudence—we find
slaves commanding ships, running shops, administering estates, look-
ing after libraries, and directing commercial enterprises just like
managers.²⁸
²⁷ “ . . . but as the oxen proved unsuitable, he told the slave to sell them and to buy
replacements with the money he got for them. The slave sold the oxen and bought
replacements but became insolvent without paying the person he bought them from”
( . . . cum hi boves non essent idonei, iusserat eos venire et his nummis qui recepti essent
alios reparari: servus boves vendiderat, alios redemerat, nummos venditori non sol-
verat, postea conturbaverat . . . ).
²⁸ Di Porto (1984) and Carandini (1988), besides, naturally, the previously cited
work of Buckland (1908).

Strata of privileged slaves thus formed in the servile condition,
especially in the big cities, who lived in circumstances of often
considerable well-being, free from physical coercion and with full
mastery of their bodies. For many of them, the availability of a
patrimony was accompanied by effective recognition of the family
units that had formed, and respect for their inviolability. At the same
time, a minutely detailed regulation of manumissions opened up
further perspectives, and in many cases came close to the goal of a
liberty that could mean the attainment, even in the space of a single
generation, of a social level that would provide a shield from the most
serious dangers.
The dual regime that governed, at least potentially, the servile
condition, personal dependency and economic autonomy, ended up
modifying, as we have just seen, not just the relations between the
slave and a third party, but also those between slave and master,
reaching the very core of the bonds of dependency.
And that was not all. There was, in fact, a habit, already fully
consolidated at the end of the republican age, whereby a slave was
effectively permitted to possess a patrimony (peculium) independ-
ently from the carrying out of a specific commercial activity. This
might consist of money or various goods—even other slaves—that in
principle belonged to the master, but was actually managed freely by
the slave, so much so that it could even be used to buy his freedom.
The exact determination of the peculium could prove to be important
in various circumstances, and indeed Servius devoted himself to that
task with great care, as we learn from Ulpian: “Everything owed by
the slave to the master is deducted before the peculium is valued,” he
wrote, before going on to say that, to this “definition,” Servius “adds
that one must also deduct anything owed by the slave to people in the
master’s power, as this too is admittedly due to the master.”²⁹ We
cannot say with certainty who was the first to formulate the definition
mentioned by Ulpian, to which Servius “adds” a second part. But
certainly it drew almost literally on a rule already formulated by
Servius in the text mentioned earlier (“it did not seem to form part
of the peculium, if not the balance remaining after deducting what the
²⁹ D. 15.1.9.2–3 (Ulpian, 29 ad ed.): Peculium autem deducto quod domino debetur

computandum esse. . . . Huic definitioni Servius adiecit et si quid his debetur qui sunt in
eius potestate, quoniam hoc quoque domino deberi nemo ambigit.
28 Aldo Schiavone
slave owed the master”).³⁰ Evidently, then, the jurist had worked to
achieve a genuine doctrinal organization of the patrimonial relations
between slave and master, designed to obtain as great a flexibility as
possible, in the light of the new requirements taking shape in the
Roman economy.
There were, however, some insurmountable barriers. Actual cir-
cumstances, acknowledged on the fringes of law in order to permit
slaves to operate in the commercial reality to the best of their abilities,
could not translate into formally recognized juridical obligations of
the master towards the slave. Let us look at a further response of
Servius, this time transcribed by Aufidius Namusa, another of his
pupils: “a master left his slave five gold pieces thus: ‘let my heir give to
my slave, Stichus, whose freedom I have directed in my will, the five
gold pieces which I owe him by my accounts.’ Namusa reports Servius
to have been of the opinion that no legacy was due to the slave,
because a master could owe nothing to his slave.”³¹ The legal recog-
nition of economic relationships within the bond of slavery ran up
against an impassable asymmetry. It is the other side of the coin to
Roman flexibility that we can see operating here: that of slaveholding
intransigency. The admission of a debt on the part of the master
counted for nothing, even if it was certified in his accounting books:
to admit its existence in terms of law would have meant pushing the
analogy to the point of accepting the possibility of a legal obligation
on the master’s part toward the slave. If such a step had been taken,
one of the fundamental presuppositions of the whole slaveholding
mechanism would have been swept aside. The fact that the binding
force of dependency appeared, in the context of Servius’ reasoning,
only as a limit and obstacle to the legal qualification of economic
relationships completely integrated into everyday life is an indication
of the degree to which the jurist’s thinking was capable of capturing
(“which I owe him . . . because a master could owe nothing . . . ”) the
sharpest contradiction that the growth of trade had introduced into
the fabric of a society based on slavery: the contrast between the
³⁰ Again in D. 15.3.16: . . . non videri peculii quicquam esse, nisi si quid deducto eo,
quod servus domino debuisset, reliquuum fieret.
³¹ This text can be read through a citation by Iavolenus Priscus, 2 ex post. Lab., in
D. 35.1.40.3: Dominus servo aureos quinque eius legaverat: “heres meus Sticho servo
meo, quem in testamento liberum esse iussi, aureos quinque, quos in tabulis debeo,
dato.” Nihil servo legatum esse Namusa Servium respondisse scribit, quia dominus
servo nihil debere potuisset.

desirability of valorizing productive forces, which would have
required the setting aside of personal relationships of dependency,
and the presence of a generalized form, without any practicable
alternatives in that world, of chattel slavery.
13.6. PRAETORS AND SLAVES
Another insuperable obstacle imposed by the persistence of the

relationship of dependency was the absolute impossibility for
the slave to seek legal redress or to be summoned himself before the
law—and little did it matter if the trial concerned his own master or a
third party.
But the praetors responsible for the administration of civil justice,
in close collaboration with the jurists, tried to get round this difficulty
as well. To deal with the problem they invented a trial device that
permitted the assignment to the masters of legal responsibility for
business affairs conducted by slaves, without, however, canceling the
role played in them by the latter. The technical expedient consisted of
indicating the name of the slave in the part of the judicial “formula” in
which the event was described (it involved a simple statement of the
circumstances that had led to the dispute, and mention of the slave
and his activities did not entail any recognition of his fitness to plead),
and the name of the master in that part of the “formula” envisaging
the conviction or acquittal of the defendant—which could not regard
a slave without denying his condition, but only a free man. It was a
manipulation carried out at the very margins of ius, utilizing all of its
ancient and shrewd formalistic wisdom, a sort of exchange within the
ceremony of the trial, that in some way made it possible, albeit
obliquely, to incorporate personal dependency and participation in
commercial networks into the slave’s activity.
In the space of a few decades the edict issued by the praetor at the
beginning of his mandate and modified year by year, which estab-
lished the formulas of the actions permitted to those who, with the
appropriate prerequisites, had requested them, reached the point of
making provision for a series of cases that ended up offering legal
protection for the most significant economic enterprises of slaves.
These ranged from agros colere (farm management) to pecuniam
faenerare (credit lending), to mercaturas redempturasque facere
30 Aldo Schiavone
(mercantile activities), according to the significant list proposed
by Labeo (the greatest jurist of the Augustan age) and recalled by
Ulpian.³²
The essential (and probably original) nucleus was made up of
actions de peculio (“on patrimony”) and de in rem verso (“on enrich-
ment”)—which envisaged a single formula with two possible variants,
already the subject of reflection by Servius and Labeo³³—whereby the
master was called to answer for commitments made by his slaves
within the limits of size of the peculium granted to the latter, or,
alternatively, the enrichment gained as a result of the activities
undertaken by the slave. To these, more or less in the same period
of time, were added the more specific formulas of the actions exerci-
toria (for maritime trade), institoria, quod iussu, and tributoria (for
other cases of non-maritime or speculative trading activity).³⁴ The
result led to a kind of commercial law of slavery that was unparalleled
in any other slave-holding society, ancient or modern.
That was, moreover, the age in which the creative force of the
praetorian edicts had reached its peak, in a constant effort by the
aristocratic elites holding legal knowledge to adapt to and anticipate
the new needs of a world dominion. A handful of solid principles,
evident but never explicitly formulated, guided their work. First,
consensualism (“consensus”): that is to say, acknowledgment of the
agreement between the parties, however manifested, provided it was
demonstrable, and expressed within the context of a typology of
transactions rigorously envisioned in the edicts. Then, reciprocity
(ultro citroque obligatio in the lexis of Labeo),³⁵ according to which,
in a transaction, any economic performance by one of the parties
should be matched by a symmetrical performance of the other party.
Next, the idea of good faith (bona fides), that is to say the commit-
ment to a reliable and trustworthy form of behavior, based on the
³² D.14.3.5.2 (Ulp. 28 ad ed.).

³³ As far as Servius is concerned, these actions are dealt with in the final part of the
already-mentioned text in D. 15.3.16 (qui boves vendiderat nummos a domino petebat
actione de peculio aut quod in rem domini versum esset . . . —“that person now claimed
the price of the oxen from the master by means of an action ‘on the peculium’ or ‘on
enrichment’ . . . ”). For Labeo, see for example D.15.3.1.1 (Ulp. 29 ad ed.), who reports
the thinking of the Augustan jurist.
³⁴ See the reconstruction in Di Porto (1984: 31ff.). On the so called actiones
adiecticiae qualitatis, see de Ligt (1999: 205ff.) and, more important, Miceli (2001).
³⁵ Cited by Ulp. 11 ad ed. in D. 50.16.19.

Roman tradition, as a way of limiting Mediterranean mercantile
cunning. Finally, and in many ways the most important of all, fairness
(aequum): a criterion whereby, in the assessment of a case for trial
purposes, the literal application of the traditional law (ius civile) could
be set aside in favor of a more flexible measure that was better suited
to the interests and values at stake. Roman law thus took on its
consolidated guise: a law made up of cases and judicial actions,
constructed day by day through the very long collaboration between
magistrates and jurists.³⁶ It was, in short, imperial power that turned
into the effectivity of a more mature legal order, which in turn favored
the accumulation of new power, in a spiral that seemed to have
no end.
13.7. CONCLUSI ON
In appearance, it seemed that not so much was required for the

decisive step to be taken, and for slavery to dissolve, as it were, from
within, destroyed by a sort of precapitalist maturing of economic
networks, guided and supported by a precursory legal reflection: and
on certain occasions the jurisprudence between Servius and Labeo
really does give the impression of being on the brink of that threshold.
But that jump was never made, and Roman slavery did not slip into
a gentle transition toward dominant forms of free wage labor. It was
also a question of numbers: the new developments introduced by the
praetors and re-elaborated by the jurists concerned only a minority—
albeit not small, as we have seen, and the focus of great attention as
well—of the millions of slaves who lived in Italy between the age of
the Gracchi and the time of Augustus. The outlook of the majority of
them, divided between the prison dormitories of the villas, the dam-
nation of the mines, or the grinding drudgery of the most humiliating
domestic duties in the city, would never extend beyond pure survival.
And, what is more, never in history has the emergence from a slave
system taken place from within the system itself, without the necessity
of traumas or at any rate interventions induced from outside; nor
have slave revolts ever sufficed. Great systemic changes have always
³⁶ Schiavone (2012: 145–53); see also Cardilli (2014).

32 Aldo Schiavone
been required to get out of it: the end of the ancient world, in one
case, and a bloody civil war with a formidable industrial transform-
ation behind it, in another. Slavery has always been a total institution,
causing such a profound split in the consciousness both of those
suffering and those imposing it as to render impossible, from within,
a collective change with the specific goal of abolishing these practices.
And so, over a few generations the creative and brilliant excogita-
tions of the Roman jurists would end up in a blind alley. In order to
continue, what was needed were social, economic, and cultural foun-
dations that it was impossible for that world to provide. Later juris-
prudence would not cancel out the results that had been achieved;
instead, it would perfect and systematize them. But there would be no
further progress of any significance along the path to further legal
recognition for the slaves. And in fact the goal of legal thought had
never been to arrive at the point of questioning the existence of the
slave system but only to attain its maximum functional efficiency with
a view to the profits of the masters, and to pursue an economic
rationality that, while not rejecting slaveholding, reconfigured its
structures to achieve a valorization of the productive forces compat-
ible with the preservation of the given structure: a result that the law
made possible with its empirical flexibility, and which no doubt
contained within it the seeds of a culture already in some way proto-
capitalist, but without there emerging a class capable of taking pos-
session of and developing it. It is a lack that marked the whole of
ancient history, defining its form and destiny.
REFERENCES
Andreau, Jean, and Raymond Descat. 2006. Esclave en Grèce et à Rome. Paris:
Hachette.
Bellen, Heinz, and Heinz Heinen. 2003. Bibliographie zur antiken Skaverei,
vols I–II. Stuttgart: Franz Steiner Verlag.
Bradley, Keith. 1994. Slavery and Society at Rome. Cambridge: Cambridge
University Press.
Buckland, William W. 1908. The Roman Law of Slavery. Cambridge: Cam-
bridge University Press.
Buckland, William W., and Arnold D. McNair. 1952. Roman Law and
Common Law, ed. Frederick H. Lawson, 2nd edn. Cambridge: Cambridge
University Press.

Carandini, Andrea. 1988. Schiavi in Italia: gli strumenti pensanti dei Romani
fra tarda repubblica e medio Impero. Rome: Nuova Italia Scientifica.
Cardilli, Riccardo. 2014. “Bona fides”: tra storia e sistema, 3rd edn. Turin:
Giappichelli.
Dalla, Danilo. 1994. Senatus consultum Silanianum. Milan: Giuffré.
de Ligt, Luuk. 1999. Legal history and economic history: the case of the
actiones adiecticiae qualitatis. 67 The Legal History Review 205–26.
Di Porto, Andrea. 1984. Impresa collettiva e schiavo “manager” in Roma
antica (II sec.a.C–II sec.d.C.). Milan: Giuffré.
Esposito, Roberto. 2014. Le persone e le cose. Turin: Einaudi.
Finley, Moses I. 1980 [new edn 1998]. Ancient Slavery and Modern Ideology,
ed. Brent D. Shaw. Princeton: Princeton University Press.
Giardina, Andrea. 1993. “The merchant,” in Andrea Giardina, ed., The
Romans, trans. Lydia G. Cochrane. Chicago: The University of Chicago
Press, 245–71.
Harries, Jill Diana. 2013. “The Senatus Consultum Silanianum; court deci-
sions and judicial severity in the early Roman Empire,” in Paul du Plessis
(ed.), New Frontiers: Law and Society in the Roman World. Edinburgh:
Edinburgh University Press, 51–72.
Hopkins, Keith. 1978. Conquerors and Slaves: Sociological Studies in Roman
History. Cambridge: Cambridge University Press.
Miceli, Maria. 2001. Sulla struttura formulare delle “actiones adiecticiae
qualitatis.” Turin: Giappichelli.
Schiavone, Aldo. 2000 [repr. 2002]. The End of the Past: Ancient Rome and
Modern West. Cambridge, MA: Harvard University Press (orig. edn: La
storia spezzata. Roma antica e Occidente moderno. Rome and Bari 1996,
several reprints).
Schiavone, Aldo. 2012. The Invention of the Law in the West. Cambridge,
MA: Harvard University Press (orig. edn: Ius. L’invenzione del diritto in
Occidente. Turin 2005, new ed. Turin 2017).
Schiavone, Aldo. 2013. Spartacus. Cambridge, MA: Harvard University Press
(orig. edn: Spartaco. Le armi e l’uomo. Turin 2011).
Schmitt, Carl. 1950. Die Lage der europäischen Rechtswissenschaft. Tübingen:
Internationaler Universitäts-Verlag.
Temin, Peter. 2013. The Roman Market Economy. Princeton: Princeton
University Press.
Thompson, Frederick H. 2003. The Archaeology of Greek and Roman Slavery.
London: Duckworth.
Turley, David. 2000. Slavery. London: Blackwell.
Watson, Alan. 1987. Roman Slave Law. Baltimore: Johns Hopkins University
Press.
14
The Practice of Manumission

through Negotiated Conditions
in Imperial Rome
Egbert Koops
14.1. INTRODUCTION
Roman slaves often had to meet expressly negotiated conditions to

obtain their freedom (§14.2). The use of such conditions helps
to explain why the Romans freed so many slaves (§14.3). They are
an expression of the economic considerations that underlie the
extraction and manumission model of Roman slavery (§14.4).
Agreements between masters and slaves occurred in practice and
were recognized at law (§14.5). Conditions could be set among the
living or by testament and could consist of settling accounts, money
payments, or services in kind; some followed the slave and were
actionable (§14.6). The money to pay for freedom often came from
the slave’s patrimony or peculium (§14.7). Although evidence is
scarce, conditions and the corresponding manumission prices
seem to have been of a type that could be met within years rather
than decades (§14.8). Extracting a price from slaves for their free-
dom lessened the future claims of patrons (§14.9). For a certain type
of slave, negotiated manumission conditions may have been the
norm (§14.10).
Egbert Koops, The Practice of Manumission through Negotiated Conditions in Imperial Rome In: Roman Law and
Economics: Exchange, Ownership, and Disputes, Volume II. Edited by: Giuseppe Dari-Mattiacci and Dennis P. Kehoe,
Oxford University Press (2020). © Oxford University Press.
DOI: 10.1093/oso/9780198787211.003.0014
36 Egbert Koops
14.2. TWO CASES: CRONION AND CALENUS
In  142, the cavalry trooper Antonius Silvanus drew up his will at

his legion’s winter camp at Alexandria.¹ Silvanus instituted his son as
heir, made out several legacies, and provided for the manumission of
a slave: “I wish my slave Cronion to be free after my death if he has
handled everything properly and transfers it to my aforementioned
heir or agent, and I wish the manumission tax for him to be paid from
my estate.”² As a libertus orcinus, someone freed from beyond the
grave, Cronion’s obligations to his patron’s son would have been
quite limited. An orcinus was not required to provide labor or
sustenance, or to show particular deference to his patron under
classical Roman law (Loreti-Lorini 1925: 36–46; Champlin 1991:
139; Mouritsen 2011: 51), making him largely independent from his
manumittor’s family circle (Garnsey 1981: 362–3).³ In this case the
heir was even burdened with the 5 percent manumission tax of the
slave’s market value (Liebs 2000: 126; Günther 2008: 118–24).⁴ So
why was Cronion freed? Referring to the clause si recte tractaverit, “if
he has handled everything properly,” Wiedemann (1985: 164), rec-
ognizing that “manumission is a just reward for faithful service,”
argued that Silvanus wanted to fashion himself as a benign master.
Yet the clause has a different and more legal meaning. Si recte
tractaverit is a condition that should be read in conjunction with
the obligation to transfer something (et tradiderit). The best explan-
ation, it seems, is that Cronion had to account for and hand over the
assets of the inheritance.
¹ FIRA III 47. The will surfaced from the Egyptian sands in 1938 and remains the
completest specimen of a Roman testament to date. Liebs 2000: 114–19.
² FIRA III 47, lines 31–7: Cronionem / seruom meum pos(t) mortem meam / si
omnia recte tractaverit et / tra(di)derit heredi meo s(upra) s(cripto) uel / procuratori
tunc liberum uolo / esse uicesima<n>{m}que pro eo ex / bonis meis dari uolo.
³ Justinian made extensive changes to the Roman law of patronage by his consti-
tution of  531, found in C. 6.4.4 (partially restored from the Basilica and referred to
in Inst. 3.7.3). The position of the orcinus was completely revised by C. 6.4.4.27
(preserved in the Veronese Codex). On the interpolation of other texts in consequence
of this law, see Loreti-Lorini 1925: 47–60; Harada 1939a, 1939b. Further discussion
in §14.9.
⁴ A similar clause appears in line 53 of the so-called testament of ‘Dasumius’ (AE
2005, 191), though the clause in lines 117–18 refers to the vice(n)sima hereditatis. For
an example of manumission inter vivos with payment of the manumission tax by the
master, see P.Tebt. II 407 ( 199); for manumission testamento with payment by the
heir, see P.Sijp. 44 ( c.130).
The Practice of Manumission 37

The legacies of 1,350 denarii total represent the value of two to six
slaves,⁵ quite apart from the gift of freedom to Cronion. The inher-
itance will have been bigger still, totaling perhaps 4,000 denarii
(Champlin 1991: 54, 184–6), and so the cavalry soldier Antonius
Silvanus appears to have been a moderately wealthy man. Perhaps
Cronion was a trusted slave who looked after Silvanus’ accounts; or
perhaps he had to render an accounting for the business he conducted
out of his slave patrimony (peculium). Even though a slave had little
legal personality and could technically hold no patrimony, his master
could allow the use of certain assets as if they belonged to the slave,
giving him a measure of autonomy and an incentive to increase his
holdings (§14.7). In any case, it is fairly common in legal sources to
find slaves accounting for their administration to their masters’ heirs
in fulfillment of a condition for their release (Buckland 1908: 494–6).
The Digest contains roughly seventy instances of a clause stipulating
that a slave shall be free if he has rendered an accounting (si rationem
dederit/reddidit),⁶ many of which are drawn from actual practice
(Morabito 1981: 170–2).
In an example that follows the wording of Silvanus’ will quite
closely, probably also drawn from an actual testament, the jurist
Pomponius cites the writings of the late Republican jurist Labeo:
“my accountant Calenus is to be free and to have all his goods and
a legacy of a hundred, if it appears that he has kept the accounts
diligently” (si rationes diligenter tractasse).⁷ Pomponius goes on to
explain that a diligence is required that profits the master rather than
the slave, not only in calculating accounts but also in paying over the
balance (reliqua). This second part is revealing. What Pomponius and
Labeo have in mind is a case in which a slave receives his freedom
together with his peculium, if he calculates and pays over in good faith
the remainder of the accounts in his care. The slave-accountant
⁵ Compare the slave prices mentioned in Ruffing and Drexhage 2008: 321–36. The
total comes to over four years’ of legionary wages at 300 denarii a year, not including
donatives (Liebs 2000: 121). Difficulties in comparing slave prices: Harper 2010: 212.
Influence of exonerations and guarantees: Arzt-Grabner 2010: 30–1. Market integra-
tion: Temin 2001, 2004; Scheidel 2005b.
⁶ Similar clauses appear in lines 41 and 51 of the ‘Dasumius’ testament (AE
2005, 191).
⁷ D. 40.7.21 pr. (Pomp. 7 Plaut.; Lab., Post.): Calenus dispensator meus, si rationes
diligenter tractasse videbitur, liber esto suaque omnia et centum habeto. The same case
is referred to in D. 40.4.8 (Pomp. 5 Sab.). For similar cases, see D. 33.8.23.1 (Scaev. 15
dig.) and D. 40.7.40.3 (Scaev. 24 dig.).
38 Egbert Koops
Calenus had personal money as well as access to his master’s money,
and inevitably the two would get mixed up. The less of reliqua there
would be for the master, the more peculium there would be for
Calenus: hence the need to explain that the slave’s interest should
not prevail. Elsewhere, Pomponius refers to a statement by the late
first-century jurists Neratius and Aristo that a slave can give security
for the unclear (obscurius) part of the balance. Otherwise hardly any
slave would ever become free, “owing to the uncertainty of the
account and this type of affair.”⁸ Celsus, writing in the first half of
the second century, allows a similar glimpse. If a slave has been freed
on condition of providing an account, but the heir will not allow
him to sell off his peculium to raise money to pay the balance, then the
slave will be free as if the condition were fulfilled.⁹ The Roman jurists
consider it near-normal that the accounts of master and slave become
entangled. They also consider it acceptable that a slave supply from
his personal money what went missing from his master’s accounts.
And they consider this sufficient to fulfill the condition for manu-
mission.¹⁰ Africanus, from the middle of the second century, is clear
on the point: “to give an accounting means no more than to pay
the balance” and as long as this happens in good faith, the slave will
be free.¹¹
Set against the legal sources, Cronion’s manumission appears in a
different light than Wiedemann thought. “Faithful service” is not the
condition for freedom, but Cronion’s truthful accounting and hand-
ing over of the balance. Much like Calenus, Cronion had access to his
master’s accounts and by implication probably held a peculium. If so,
it was not explicitly granted to him as a legacy and was therefore
considered withdrawn in as far as anything would be left after settling
the accounts.¹² Seen in this light, his manumission appears less as
an act of benevolence and more as the winding-up of a partnership.
⁸ D. 40.7.5 pr. (Pomp. 8 Sab.): incerta causa rationis et genere negotii huiusmodi.
⁹ D. 40.7.23.1 (Cels. 22 dig.). This follows from the legal principle that a condition
is automatically met if someone who stands to gain from its non-fulfillment hinders or
prevents its fulfillment (Kaser 1971: 257).
¹⁰ e.g. D. 40.5.41.7 and 9 (Scaev. 4 resp.); D. 40.7.40 pr. (Scaev. 24 dig.).
¹¹ D. 35.1.32 (Afr. 9 quaest.): rationes reddere nihil aliud sit quam reliqua solvere.
But also see D. 35.1.82 (Call. 2 quaest.) for an additional obligation to surrender the
books. Columella (RR 1.8.13) advises masters not to allow farm overseers to trade for
their own profit at all, because it increases the difficulty of settling accounts.
¹² Fr. Vat. 261 (Pap. 12 resp.); D. 33.8.8.7 (Ulp. 25 Sab.); C. 7.23.1 (Diocl./Max., 
294); Inst. 2.20.20.

The slave’s help is indispensable to establish the contents of the
inheritance, and manumission is a strong incentive to produce a
truthful account.¹³ Of course Silvanus could have revoked Cronion’s
privileges and peculium at any time, or could have ordered him to
render an accounting (§14.5).¹⁴ But death is a natural moment to
extricate the affairs of master and trusted slave, precisely because the
heirs lack sufficient information about the inheritance if the slave
bookkeeper does not cooperate (Dari-Mattiacci 2013: 84–7). The
legacies that are often coupled to this type of manumission, as in
the case of the accountant Calenus, only reinforce this view. They put
a premium on a truthful account, leading to manumission, while the
penalty for a detected lie is continued slavery, thereby tipping the
scales in favor of truth.
14.3. THE FREQUENCY OF MANUMISSION
It would be amiss to suggest that Cronion or the dispensator Calenus

were typical Roman slaves. Not every employee in a modern firm has
access to the board of directors or the corporate bank account and not
every Roman slave had access to his master, let alone his master’s
accounts. Slaves occupied a spectrum of social positions (Finley 1985:
67–8; Alföldy 2011: 196). Even so, Cronion and Calenus represent a
type that is frequently encountered in literary sources, that predom-
inates in the epigraphy, and that receives most of the jurists’ attention.
Selection bias cannot be avoided, since too little is known about the
vast bulk of undocumented Roman slaves, including almost all the
agricultural slaves, to say whether, when, and how they were manu-
mitted, or in what numbers. For this reason, most of what follows
applies only to a certain class of privileged or skilled slaves (Fenoaltea
1984), or to those employed in positions of trust (Dari-Mattiacci
¹³ This is not to say that the account will necessarily be truthful. The slave may take
the dangerous gamble that his fraud will go undetected, netting him both freedom and
(more) money. D. 27.3.1.3 (Ulp. 36 Ed.); D. 40.4.22 (Afr. 9 quaest.); D. 40.5.23 pr.
(Pap. 9 resp.); D. 40.5.41.11 (Scaev. 4 resp.); D. 40.7.40 pr. (Scaev. 24 dig.).
¹⁴ D. 33.8.19 pr. (Pap. 7 resp.): “when a master wanted to manumit a slave, he
ordered him to provide an inventory of his peculium, and thereby the slave received
his freedom” (cum dominus servum vellet manumittere, professionem edi sibi peculii
iussit atque ita servus libertatem accepit). Also see D. 18.1.7 pr. (Ulp. 28 Sab.).
40 Egbert Koops
2013: 105–6), and not to Apuleius’ slave millers or the Spanish miners
mentioned by Diodorus Siculus, “who would rather be dead than
alive.”¹⁵ Certain claims can be made, however, with regard to this
select group of documented Roman slaves.
One of these claims takes the form of a question famously posed by
Keith Hopkins (1978: 115): “Why did the Romans free so many
slaves?” The number of references to freed slaves is staggering
indeed. Rare cases of mass manumission aside (Westermann 1955:
135–6; Hopkins 1978: 115), many active, prosperous, or high-ranking
freedmen appear in literary sources (Garnsey 1981: 359–61)¹⁶ and
the economy of Roman Italy has been rightfully characterized as
a “freedmen economy” (Verboven 2012). Three-quarters of the
funerary inscriptions in Rome concern former slaves and the propor-
tion is high as well for other Italian cities (Mouritsen 2005: 38–9,
Verboven 2012: 90–1). Of the more than 506,000 Latin inscriptions
catalogued in the Clauss-Slaby online database, over 27,000 refer to
liberti. Freedmen figure in more than 1,000 legal fragments and their
position was subject to a constant stream of legislation.¹⁷ Unfortu-
nately, although such figures provide the notion that freedmen were
common enough to matter, they reveal little about their relative
number. The interests of writers and their readership shape literary
sources,¹⁸ inscriptions reflect the strong epigraphical habit of freed-
men (Mouritsen 2005: 55–63), and not every legal text presents a case
drawn from actual practice. Some support can be found in the
importance literary sources accord the vice(n)sima manumissionis,
the manumission tax, but no reliable figures can be drawn from it
(Bradley 1987: 149–50; Günther 2008: 98–9), nor from the number of
public grain recipients (Mouritsen 2011: 120–3). Be this as it may, few
scholars would dispute that the Romans freed ‘many’ slaves.
Quantifying ‘many’ remains difficult, both in terms of the number
of slaves at any given time and in terms of the ratio of slaves to free
persons. Two figures are generally adduced: the census returns for
¹⁵ Apul., Met. 9.12; Diod. 5.38.1: αἱρετώτερος γὰρ αὐτοῖς ὁ θάνατός ἐστι τοῦ ζῆν.
¹⁶ e.g. Cass. Dio 50.10.4–5 and 51.3.3; Cic., Balb. 24; Philo, Leg. ad Gai. 155; Tac.,
Ann. 13.27.
¹⁷ Lopez Barja de Quiroga (1998: 161–3) provides a long list, only covering the
Junian Latins.
¹⁸ Even freedmen authors such as Epictetus, Phaedrus, and Terentius provide little
information on the prospects of the slave population. On Phaedrus, see Bradley (1987:
150–3). On the anonymous Life of Aesop, see Hopkins (1993: 10–27).

Roman Egypt, which show a ratio close to 11 percent (Bagnall
and Frier 1994: 48), and Galen’s cryptic reference to the roughly
20–30 percent slave population of Pergamum (Harris 1999: 65).¹⁹ It
is difficult to proceed from these figures to an estimate for Roman
Italy, let alone the rest of the Empire. As Scheidel (2005a: 64–6) has
pointed out, the prevalent notion that slaves made up a third of the
Roman population is nothing more than a shaky top-down guess. By
stocking the Roman elite with a number of slaves according to their
status, applying the data from Roman Egypt to the Italian cities, and
extrapolating from the production levels and demand for a slave
workforce in agriculture, Scheidel (2005a: 66–71) arrived at a
bottom-up estimate of anywhere between 500,000–1,000,000 urban
slaves and 250,000–750,000 rural slaves for Roman Italy. Depending
on the size of the free population, a figure which is also problematic,
slaves may have made up between 13–28 percent of the Italian
population if one assumes the ‘low count’ for the total population
(Scheidel 2008: 19–30). Such a margin of uncertainty precludes
quantitative claims about the percentage of slaves manumitted, par-
ticularly since the weight of various sources in the slave supply is still
heavily debated (Harris 1980; Scheidel 1997; Harris 1999; Scheidel
2005a). Because births seem to have been a major source of new
slaves, female manumission rates cannot have been overly high at
child-bearing age or the slave population would plummet, which is
not supported by other evidence. But the incidence of male manu-
mission and late-life female manumission has less of a demographic
effect and may have been high.²⁰ Scheidel (1997: 160–6, 2005: 76) has
employed a purely hypothetical model of 10–20 percent per five-year
age cohort, starting at age 25, which means roughly half to three-
quarters of the slave population would have been released by age 50.
This hypothetical model has met with general silent acceptance
(Verboven 2012: 92), which means no better guess is currently
available.
In a seminal study of the age at death given in over 1,800 funerary
inscriptions, Alföldy (1972: 346–54) has shown that a large majority
of the epigraphically commemorated ex-slaves had been released
¹⁹ Gal., Propr. an. 9.13 (Corpus medicorum Graecorum V 4.1.1).

²⁰ Manumission was also used as an incentive for childbirth. Colum., RR 1.8.19; Fr.
de iure fisci 13; D. 1.5.15 (Tryph. 10 disp.); D. 40.7.3.16 (Ulp. 27 Sab.).
42 Egbert Koops
before the age of 30, or at most between 30 and 40.²¹ The age of 30
was not chosen randomly: it was the legal age for manumission
leading to Roman citizenship under the Augustan lex Aelia Sentia
of  4.²² Alföldy claimed that manumission shortly after 30 consti-
tuted the “Normalfall” with which every slave could reckon. This
assertion met with severe criticism (Hopkins 1978: 127; Garnsey
1981: 361–3; Wiedemann 1985: 162–3). It was argued that slaves
were only commemorated at an early age; that only trusted slaves
were commemorated at all; and that epigraphically commemorated
freedmen form an atypical group of “successful” ex-slaves by their
very nature. In addition, Alföldy’s figure could not generally apply to
female slaves, or the slave population would never come close to
reproducing itself (Scheidel 1997: 165–6). Yet the simple fact remains
that the majority of epigraphically commemorated (ex-)slaves were
manumitted early, when a reasonable part of their productive lives
still lay ahead. True enough, the epigraphical data cannot be used as a
demographic sample (Mouritsen 2011: 132–6), but it does show that
for a certain class of slaves, early manumission was the norm.
Employing similar methods as Alföldy with a non-overlapping
data set, Weaver (1972: 97–104) concluded that manumission
between the ages of 30 and 40 was normal for imperial slaves (familia
Caesaris). A different, more random sample is offered by the census
returns from Roman Egypt.²³ Bagnall and Frier (1994: 71, 94, 156–8,
342–3) indicate that male slaves in Roman Egypt were generally freed
between ages 30 and 40, though women were as a rule not manumit-
ted while they could still bear children. A similar trend appears in
Romano-Egyptian documents that provide the age of slaves or freed-
men (Wiedemann 1985: 163). Finally, a Greek census inscription
from the late fourth century lists 152 named slaves and 87 ages on a
single agricultural estate on the island of Thera (Harper 2008:
106–16). Males may be under-reported for tax reasons, but even so
the same pattern emerges as from the Romano-Egyptian data, that is
to say a high level of probable male manumission after 30. Perhaps it
²¹ As much as 67 percent in Rome itself, though regional differences should be kept

in mind. As Alföldy notes, this percentage even understates the case, as freedmen who
died at a later age may have been released earlier.
²² Gai., Inst. 1.18; Ulp., Epit. 1.12.
²³ The point can always be made that the published census returns for Roman
Egypt cover three centuries, only mention 118 slaves, and may reflect local circum-
stances such as a lesser demand for slave labor.

is a localized and fairly late snapshot. But at some point the question
becomes how many “special cases” it takes to establish a general
case, particularly since no study has provided any evidence to the
contrary.²⁴ The incidence of male and female manumission varies
remarkably between sources,²⁵ but for men, at least, manumission
between the ages of 30 and 40 appears to have been frequent. The
jurists suggest that the requirements for formal manumission were
interpreted very leniently after 30, and quite strictly beforehand.²⁶
Even so, it goes without saying that some slaves were never freed.
14.4. ECONOMIC CONSIDERATIONS

FOR MANUMISSION
All slaves were expensive, both measured against wheat equivalent

and against daily wages (Scheidel 2005b; Ruffing and Drexhage 2008:
341–7), and outside of the elite most households will have owned only
a few (Bagnall and Frier 1994: 48–9), either as a status symbol and for
domestic work, or to work beside their owner and provide income in
the “family economy” (Groen-Vallinga 2017: 98–9). Slaves needed
sustenance²⁷ and slave children at least had little earning power
(Morabito 1981: 64–5).²⁸ Some were taught trades or skills at considerable
²⁴ The over 1,200 Delphic manumission records from the second century 
through the first century  do not provide firm age indications (Hopkins 1978:
133–71). The same is true of the epitaphs of the urban slaves and freedmen of the elite
Roman households (Treggiari 1975a: 58, 1975b: 395–400).
²⁵ Alföldy (1972: 351) noted a higher incidence of female than male manumission
before 30 in the funerary inscriptions. The census returns from Roman Egypt and the
Thera inscription show quite the opposite pattern, which is probably closer to reality.
The difference may be due to the practice of releasing and then marrying young slave
women (Wacke 2001), who stood a much better chance of being commemorated.
²⁶ Gai., Inst. 1.20: maiores vero triginta annorum servi semper manumitti solent,
“slaves over 30 are wont to be manumitted at any time.” Also see D. 40.2.7 (Gai. 1
cott.); D. 40.2.8 (Ulp. 5 Ed.).
²⁷ Juvenal and Seneca complain of the cost (Sat. 3.166–67; Tranq. 8.8). A wet nurse
for an infant slave foundling cost 10 drachmae and half a liter of oil per month (BGU
IV 1107, 13 ); a slave wet nurse for a free infant cost 400 drachmae for two years
(P. Oxy. I 91,  187).
²⁸ e.g. a legacy of the use of a slave child only takes effect after infancy, and no value
is placed on his labor until the age of five: D. 7.1.55 (Pomp. 26 Q. Muc.); D. 7.7.6.1
(Ulp. 55 Ed.).
44 Egbert Koops
cost (Verboven 2012: 94).²⁹ Even the residual value of slaves was high.
Petronius has a freedman claim that he paid 1,000 denarii for his
freedom after forty years of slavery.³⁰ The highest actual age that can
be related to a specific price is in the manumission certificate of
Antonius Hermes, aged 40, for whom 20 denarii in taxes were paid,
putting his value at a minimum of 400 denarii.³¹ To release any slave
meant a sacrifice both in status and earning capacity for most house-
holds, not to mention the difficulty of recouping the investment under
a system of early manumission. So why were they freed in such
numbers? As Hopkins noted (1978: 117), “Roman society was not
marked by altruism.”
Of course, some manumissions were prompted by non-pecuniary
considerations. Slaves and masters lived in close proximity and rela-
tionships were doubtlessly formed. Some slaves were released in
recognition of their accomplishments³² or in gratitude for exceptional
service,³³ and to a large extent such practice represents the literary
ideal (Wiedemann 1985: 163–5; Mouritsen 2011: 30–5).³⁴ Slave girls
²⁹ P. Oxy. IV 724 = Chr. 140 (apprenticeship to stenographer, 120 drachmae, 

155); but compare P. Oxy. XIV 1647 ( c.185, paid apprenticeship to weaver, wages
rising from 8 drachmae per month in the first year to 20 drachmae per month in the
fourth year, with eighteen days off per year). Value added by training: Plut., Cat. 21.7;
Cic., Q. Rosc. 10; Hor., Ep. 2.2.5–8; Mart. 10.62. Also see the famous case reported by
Paulus, D. 19.1.43 (5 quaest.). Conversely, Habinnas counts himself lucky that
he spent no money to teach what his slave could also pick up on the streets (Petr.,
Sat. 68.7).
³⁰ Petr., Sat. 57.6 and 57.9.
³¹ FIRA III 10bis (Egypt, second century). The true price may have been higher still
because of tax evasion. If both examples are explained away as cases of self-purchase
(which is not evident in the case of Antonius Hermes), the next highest age is a slave
woman aged 30–39 who sold for 200–25 denarii (800–900 drachmae, P. Mich. XV
707, after  185, damaged); or a male slave aged 32 who sold for 225 denarii (900
drachmae, P. Oxy. XXXVIII 2856,  91/92). In P. Hamb. I 63 ( 125/126) two
prisoners of war, one aged 38 and the other’s age not given, are sold for a combined
price of 350 denarii (1,400 drachmae).
³² e.g. after offering money the grammarian Lenaeus was freed gratis by Pompey in
recognition of his learning (Suet., Gramm. 15), and when Cicero freed his secretary
Tiro in 53 , his brother Quintus famously wrote that Tiro would rather be their
friend than their slave (Cic., Fam. 16.16.1).
³³ Fending off robbers, uncovering a plot, defending or healing their master, etc.:
D. 40.2.15.1 (Paul. 1 Ael. Sent.). Denouncing the murder of their master: Sc. Silania-
num, D. 29.5 rubr. (Buckland 1908: 600–2). The grant of freedom for informing
against a criminal master to the benefit of the state seems to be a later development
(Buckland 1908: 598–9); but see App., Bell. civ. 4.11 for special circumstances (pro-
scriptions).
³⁴ Dion. Hal., Ant. Rom. 4.24.4.

were bought and set free at an early age for the purpose of marriage,
which brought material benefit to their masters under the Augustan
matrimony laws (Wacke 2001). The lex Aelia Sentia contained an
exception for next of kin and future wives, who could be formally
manumitted and acquire Roman citizenship before the prescribed age
of 30.³⁵ If a union with a slave girl proved fruitful, the resulting
children might be freed as well, though the relation between a master
and his natural slave children should not be sentimentalized by
assuming this as a necessary outcome.³⁶ Dying slaves were sometimes
manumitted. The practice had little value in the eyes of a jurist and
was the butt of one of Petronius’ jokes, since the bereaved master still
owed manumission taxes (Kleijwegt 2002).³⁷ Sometimes, dying slaves
were simply abandoned and left to fend for themselves.³⁸ Slave
abandonment may either have been a common practice or, under
the emperor Claudius, an outrageous but isolated incident that
sparked imperial intervention. The same is true for the manumission
of actors or gladiators following popular acclaim, which was criticized
or forbidden by several emperors.³⁹ Equally frowned upon was the
large-scale manumission to gain clients or fill out funeral processions,
which Dionysius of Halicarnassus claims to have witnessed.⁴⁰ Such
cases seem rare (Champlin 1991: 141). Finally, a slave could be forced
to take up a debt-ridden inheritance as a necessary heir (heres neces-
sarius), a quaint institution that merited considerable legal attention
because a necessary heir was often included in the order of succession
³⁵ Gai., Inst. 1.19; D. 40.2.11 (Ulp. 6 procons.); D. 40.2.12 (Ulp. 2 Ael. Sent.);
D. 40.2.13 (Ulp. 6 procons.); D. 40.2.20.2–3 (Ulp. 2 off. cons.). The slave’s age
requirement was abolished by Justinian (C. 7.15.2,  530).
³⁶ Some testaments provide for the manumission of likely slave children, which
implies that the master and father had no intention of manumitting immediately:
FIRA III 50 = Chr. 316 (will of C. Longinus Castor,  194); FIRA III 10 (Dameis aged
13 freed with peculium and liberated from patronage, third century).
³⁷ D. 40.4.17 pr. (Iul. 42 dig.); Petr., Sat. 65.10. For deathbed manumission, see
Mart. 1.101; Pliny, Ep. 8.16; CIL VI 9317; CIL X 2381 = ILS 7842; SB XXII 15345.
³⁸ Claudius decreed that abandoned slaves need not return to their masters if they
recovered, but were free with the status of Junian Latins, giving them at least some
recognition: Suet., Claud. 25.2; Cass. Dio 60.29.7; D. 40.8.2 (Mod. 6 reg.); C. 7.6.1.3
(Just.,  531).
³⁹ Tiberius: Cass. Dio 57.11.6; Suet., Tib. 47; Hadrian: Cass. Dio 69.16.3; Marcus
Aurelius: D. 40.9.17 pr. (Paul. 3 Ael. Sent.); C. 7.11.3 (Alex.,  223–4); Cass. Dio
72.29.
⁴⁰ Dion. Hal., Ant. Rom. 4.24.8. Also see Pers., Sat. 3.105–6; Petr., Sat. 42.6.
46 Egbert Koops
to ensure that the inheritance was taken (Buckland 1908: 505–12).⁴¹
Again, the number of cases in which this failsafe was activated and the
slave was freed may have been limited (Champlin 1991: 137).
The problem with such manumission motives is that blood ties,
faithful service, deathbed grants, abandonment, popular success in
the arena or on stage, debt-ridden inheritances, etc. can hardly
account for systematic male manumission at an early age. It is
difficult to reconcile that manumission was “both very common
and very selective,” as Mouritsen claims (2011: 140), who admits
that it was “essentially a financial transaction” (2011: 146) while
simultaneously downplaying economic considerations (2011:
159–80). Instead of the fanciful notion of “common selection,”
the best explanation why the Romans freed so many slaves remains
the one offered by Hopkins (1978: 118) and Alföldy (1972: 360–3):
slaves paid considerable sums of money for their freedom. The
Roman slave system exploited the desire for freedom to great effect,
offering manumission as an incentive for the ideal type of hard-
working and frugal slave who would merit freedom (Hopkins 1978:
128, 147–8). This meant that extraction of the full value of a slave’s
labor was postponed for some time. Even so, giving slaves a stake in
their productivity by allowing them to save part of their earnings
ensured cooperation and harder work at lower supervision costs
(see already Cohen 1951: 222).⁴² The risk of undisclosed funds
following the slave into freedom was partially countered by giving
patrons a legally protected stake in any future earnings of a freed-
man, through an expectancy to inherit (§14.9) as his closest agnatic
kin (Kaser 1971: 697; Gardner 1993: 19–20).⁴³ Thus, slavery
and manumission were marked by a stepped extraction process.
First, labor was extracted for which the slave received less than full
compensation; then a manumission price would be paid, enabling
the master to reinvest while obtaining a freedman; and upon
the freedman’s death, the patron would often take part of the
inheritance.
⁴¹ Gai., Inst. 2.153; Ulp., Epit. 1.14. See line 13 of the testament of ‘Dasumius’ (AE
2005, 191).
⁴² Varro, RR 1.17.5–7 and 1.19.3. But note the quite different opinion of Colum.,
RR 1.8.13 and 11.1.24. Aristotle hints at the problem as well (Oecon. 1.1344b).
⁴³ Gai., Inst. 3.39 and following. Gaius dedicates a full thirty-eight paragraphs to
the convoluted law of freedman inheritance.

The mechanism described here was not invented by Augustus.
Statuliberi, slaves freed by testament on condition of payment of a
sum of money, already appear in the Law of the Twelve Tables.⁴⁴ But
the Augustan manumission laws surely fostered this form of manu-
mission to serve state interests.⁴⁵ On the one hand, a high manumis-
sion rate was desirable as it meant that enough freedmen were visible
and successful for manumission to present a believable scenario,
fostering obedience on the part of the slave population (Bradley
1994: 163).⁴⁶ Moreover, freed slaves were taken into the citizen
body, adding new citizens in a practice that some considered an
important contribution to the Roman success.⁴⁷ On the other hand,
Augustus aimed to add only the “best” slaves as citizens,⁴⁸ that is to
say those who deserved freedom in the eye of a discerning master by
their frugality or other virtues. These conflicting aims were both
addressed by the lex Iunia (Norbana) of 17 ,⁴⁹ which created a
waiting room class of second-rate free persons, consisting of freed-
men who had been granted their liberty through informal manumis-
sion and lacked full citizenship in consequence.⁵⁰ Such Junian Latins
do not stand out in the epigraphy (Weaver 1990) and are difficult to
track in other sources.⁵¹ With some justice Weaver (1997: 55) has
called this status group “a black hole of large but unknown
⁴⁴ Tab. VII,12 (Bruns) = Tab. VI.2 (Flach): Ulp., Epit. 2.4; D. 40.7.25 (Mod. 9 diff.);
D. 40.7.29.1 (Pomp. 18 Q. Muc.). Also see Fest., s.v. statu liber.
⁴⁵ Augustus’ program included the lex Iunia (17 ), lex Fufia Caninia (2 ), lex
Aelia Sentia ( 4) and the lex Papia Poppaea ( 9). With minor modifications these
laws were in place for over five hundred years, until most were abolished by Justinian.
⁴⁶ Hinted at by Dion. Hal., Ant. Rom. 4.24.7, perhaps in response to the slave
revolts of the late Republic.
⁴⁷ Cf. the famous (but slightly erroneous) letter of Philip V of Macedon to the city
of Larissa, SIG³ 543 = ILS 8763 (214 ). Also see Cass. Dio 56.7.6.
⁴⁸ Cass. Dio 56.33.1–3 (but perhaps mistaken, see Suet., Aug. 101); Suet., Aug.
40.3–4.
⁴⁹ The date is in dispute (the other possibility being  19, under Tiberius), but
since the status group is named after the Junian law, it seems likely this law preceded
the lex Aelia Sentia that also regulated their position. Buckland 1908: 534–7; Sirks
1981: 250–1; Weaver 1997: 58–60; Lopez Barja de Quiroga 1998: 137–8.
⁵⁰ That is to say not by iusta manumissio in front of a magistrate or by testament,
but by the mere will of their master. Gai., Inst. 1.17 and 3.56; Fr. Dos. 4–5; Quint.,
Decl. min. 340 and 342. Of course proof was essential: see Mart. 9.87 and the unsavory
tale in Cic., Att. 7.2.8.
⁵¹ Isolated cases: Suet., Vesp. 3.1 (later declared freeborn); Plin., Ep. 7.16.4 (offer of
magistrate’s assistance); Plin., Ep. 10.5.2, 10.11.2 and 10.104 (requests for citizenship
grants); Mart. 1.101 (deathbed manumission); AE 1959, 297 (probatio anniculi). See
now Emmerson (2011).
48 Egbert Koops
proportions.”⁵² It seems likely that informal manumission was
frequently employed. Patrons of Junian Latins retained a claim to
their entire inheritance, which would fall to them as if the deceased
freedman had remained a slave.⁵³ This gave Junian Latins a strong
incentive to obtain full Roman citizenship by a second, formal manu-
mission (iteratio, Buckland 1908: 714–18; Sirks 1983). The require-
ment of further cooperation from the patron for formal manumission
ensured continued dependence and further opportunities for extrac-
tion.⁵⁴ In any case, the overall effect of the Augustan laws was that
informally manumitting slaves as Junian Latins was relatively easy,
maintaining a high manumission rate, while formally turning slaves
into Roman citizens was more difficult, but not impossible, main-
taining the desirability of citizenship. From a systematic point of
view, manumission reinforced slavery (Hopkins 1978: 118).
14.5. IMPLICATIONS OF FREEDOM PURCHASE:

AGREEMENTS AND TERMS
The extraction process described here has two implications. The first
is that manumission was often the outcome of negotiations between
master and slave. This may be a difficult proposition, considering the
power disparity, but there is enough evidence to believe agreements
or promises occurred frequently. In the Roman view, slaves were
capable of forming relations governed by bona fides or good faith
(Horsmann 1986: 317), which in turn meant that to break one’s word,
even against a slave, was to act against good morals (Kaser 1971: 200).
Pliny allowed slaves to bequeath part of “their” property, as long as
they made it over within the household, and Columella advised to
offer exemption from work to slave women who bore more than two
⁵² As a status group, Junian Latins survived the mass grant of Roman citizenship by
the Constitutio Antoniniana of  212, and the status was only abolished by Justinian
(C. 7.6.1.12–12a,  531). Koops 2012.
⁵³ Gnomon 19 = BGU 1210; Gai., Inst. 3.56; Inst. 3.7.4; C. 7.6.1.1b (Just.,  531).
⁵⁴ Fr. Dos. 14; Gai., Inst. 1.35; Fr. Vat. 221; Ulp., Epit. 3.4. Also see Tac., Ann. 13.27.
The state tapped this source of income as well, as successive emperors granted
citizenship in return for any number of expensive civic services. Gai., Inst. 1.32c–34;
Ulp., Epit. 3.6.

children, and freedom to those who bore more than three.⁵⁵ The
novelist Achilles Tatius describes a scene in which the slave girl
Lakaina negotiates with Melite for her purchase for 2,000 drachmae
from an abusive master, offering to repay as soon as possible.⁵⁶ In an
inscription at the Rijksmuseum of Oudheden at Leiden, a certain
Ericthonius is mentioned, “who died age 28, when he ought to be
manumitted at age 30.”⁵⁷ The use of the verb debere may imply a
social norm, but also an agreement between a master and his slave
or a third party concerning the term of servitude; in any case
Ericthonius’ master was felt to be morally or legally required to
free him at age 30.⁵⁸ Manumission agreements, pactiones pro libertate,
are common in legal sources. As early as the first century , Alfenus
discusses a case in which a slave has contracted with his master to
purchase his freedom; when the master dies, the slave has not yet paid
the full amount but is freed by testament.⁵⁹ Whether he can reclaim
what he has already paid depends on whether the master entered the
payment into his own accounts, or kept it apart as belonging to the
slave’s peculium. The implication that legal obligations could arise
between masters and slaves as debtors and creditors is borne out by a
multitude of texts (Buckland 1908: 220–5; Morabito 1981: 105–10).
As Ulpian notes, slaves were bound and could bind others naturaliter
by their contracts.⁶⁰
True, a master was under no legally enforceable obligation and
merely under a moral obligation to uphold agreements made with his
slaves. Masters might withdraw the peculium (Johnston 2002:
11–12),⁶¹ sell the slave or assign him to a different station, or renege
on their promise of manumission (Buckland 1908: 205, 640–3).
⁵⁵ Plin., Ep. 8.16.1–2 (note the inversion in Petr., Sat. 53.8, where Trimalchio is
disinherited by his slaves); Colum., RR 1.8.19. In the same letter, Pliny suggests that he
manumits immaturos easily, that is those under 30.
⁵⁶ Ach. Tat. 5.17.
⁵⁷ RMO Leiden (unpublished, EDCS-58700011): Dis manibus/Ericthonio animae/
sanctissimae hic/cum deberet ann(is) XXX/manumitti ann(os) XXIIX/decessit/
C(aius) Cilnius Philetus/filio carissimo/fec(it) (Smyrna, second to third century).
Compare CIL X 4917.
⁵⁸ Especially in cases of self-sale into slavery, such agreements will have been
common. Morabito 1981: 70–4; Harris 1999: 73–4. For discussion of the clause ut
manumittatur, see §14.6.
⁵⁹ D. 40.1.6 (Alf. 4 dig.). Also see D. 33.8.8.5 (Ulp. 25 Sab., referring to Labeo).
⁶⁰ D. 44.7.14 (Ulp. 7 disp.). Also see D. 12.6.13 pr. (Paul. 10 Sab.); D. 12.6.64
(Tryph. 7 disp.).
⁶¹ D. 15.1.7.6 (Ulp. 29 Ed.); D. 15.1.8 (Paul. 4 Sab.). Compare Plaut., Aul. 820–32.
50 Egbert Koops
A rescript from Diocletian states that a mistress who agreed that her
slave would be free after a certain period of servitude was not required
to uphold the bargain; but even so, a different rescript from the same
emperor states that a master who has received a slave’s payment
should at least be urged by the provincial governor to keep his
promise of manumission.⁶² The ability to renege raises the question
whether some masters would freeload on a social structure in which
many kept their word (Watson 1987: 65, Kleijwegt 2012: 113–15).
After the murder of the urban prefect Pedanius Secundus by his slave
in  61, the senate debated whether to apply the law that all slaves
under the same roof were to be killed. Tacitus notes that the murderer
was spurred to violence either “because he had been refused the
freedom for which he had paid the price” or out of a lover’s jealousy;
but later, in the speech he attributes to the jurist Gaius Cassius
Longinus, it is stated that the unnamed slave was avenging wrongs
“because he had bargained with paternal money or because an ances-
tral slave was taken away”; this suggests that Tacitus rather believed a
failed manumission deal to be the cause of violence.⁶³ Roman masters
were well aware of this potential threat: “as many enemies as slaves,”
as the proverb went.⁶⁴ Not to keep one’s word on matters as import-
ant as manumission could be dangerous in a society with a propensity
toward violence (Morabito 1981: 135). Considerable legal ingenuity
was expended on ways to neutralize this social danger by making the
master’s obligation to manumit enforceable at law. Examples are the
use of a trusted third party or the allowance of certain claims against
the master if an agreement had been confirmed by testament (§14.6).
Such devices are far more common in the legal sources than simple
pactiones between master and slave.
The second implication of agreements to purchase freedom would
appear to be that Roman slaves were manumitted more often inter
vivos, during the lifetime of their masters, than by testament. This runs
counter to a tradition going back to Buckland (1908: 546), that
testamentary manumission was more common by far because it
guaranteed the use of the slave until death (Duff 1928: 25; Garnsey
⁶² C. 7.16.36 (Diocl./Max.,  294); C. 4.6.9 (Diocl./Max.,  294). Also see

C. 7.16.20 (Diocl./Max.,  293).
⁶³ Tac., Ann. 14.42: seu negata libertate, cui pretium pepigerat, sive amore exoleti
incensus et dominum aemulum non tolerans; Tac., Ann. 14.43: quia de paterna
pecunia transegerat aut avitum mancipium detrahebatur.
⁶⁴ Sen., Ep. 47.5; Fest. s.v. quot servi (314L); Macr., Sat. 1.11.13. Also see Curt. 7.8.28.

1981: 362; Patterson 1982: 223–4).⁶⁵ Recently, scholarly opinion has
swung the other way, claiming that manumission inter vivos repre-
sents the standard (Roth 2010: 98–9; Mouritsen 2011: 180–3). In truth,
there is little evidence for either claim. Testamentary manumission is
referenced four times more often in the Digest (Morabito 1981: 166),
but that may be due to the greater complexity of testaments and
legacies, giving the lawyers and Byzantine lawmakers cause for atten-
tion. In literary and epigraphic sources both are encountered in
roughly equal proportion, but accidents of preservation forbid any
conclusion. Little importance can be attached to the rule that the slave
peculium was considered granted unless withheld for manumission
inter vivos, and was considered revoked unless granted for testament-
ary manumission. As Roth notes (2010: 95–104), because slaves paid
for freedom out of their peculia, it would have been the subject of
negotiation on the one hand, while on the other hand an implicit grant
was often construed, for instance in cases of testamentary manumis-
sion on condition of payment of a certain sum.
Since it touches on slave expectations as well as manumission
motives and numbers,⁶⁶ it would seem important whether manumis-
sion inter vivos or by testament was more common. However, once
one accepts the existence of a social practice of manumission pay-
ments, this opposition loses its relevance. A master would negotiate
with a certain class of slaves and set certain terms. If the terms were
met during his lifetime, he would usually free the slave; but it made
sense to enshrine the terms in a testament or codicil as well, leading to
a legally recognized position of the slave as a statuliber in case the
master died before freeing his slave. Testamentary manumission put
the agreement between master and slave on a securer footing for the
slave: not because he expected to be released only at his master’s
death, but because he knew that the bargain would hold even if his
master died. The heirs were not bound by any agreements between
the deceased and his slave, but they were bound by the testator’s will.
A testament such as that of Silvanus (§14.2) cannot discount the
possibility that Cronion may have been manumitted before the will
came into effect; but if Cronion was still a slave when Silvanus died,
⁶⁵ The self-purchase model vitiates this argument since it allows for the purchase of
a replacement slave and the addition of a freedman.
⁶⁶ Because of the lex Fufia Caninia of 2 , which limited the number of slaves who
could be freed by testament: Gai., Inst. 1.42; Ulp., Epit. 1.24; Paul., Sent. 4.14.4.
52 Egbert Koops
his position as a statuliber would have been unassailable, not only by
the heirs but also by any new owner to whom the heirs might sell him.
Such conditions followed the slave (Buckland 1908: 286–7). Of
course, slaves were powerless to prevent the testament from being
modified by the testator during his lifetime,⁶⁷ which happened fre-
quently (Champlin 1991: 64–70); but that was no different than their
position regarding any other agreement with their master. Moreover,
there are traces of a practice of reading wills in public, asserting the
social order of the household (Champlin 1991: 23–4) by communi-
cating what lay in wait.⁶⁸
Seen in this light, the important question is not how the Romans
freed so many slaves—by testament or inter vivos—but under which
terms. A testament is merely a way to enshrine a negotiated agree-
ment and make it actionable. Over 90 percent of the fragments in the
Digest dealing with manumission place a condition on attaining
liberty (Morabito 1981: 174). The condition may consist of an
amount to be paid, at once or over time, or of years of servitude,
but it may also take the form of future service as a freedman; and it
may be set by agreement or by testament, either directly or through
an intermediary. The possibilities are endless (Champlin 1991:
139–40). And though their frequent occurrence should be taken
with a grain of salt, since conditions give rise to greater legal problems
than gratis manumission, it does show that manumission terms
were at the forefront of jurists’ minds (Table 14.1). Put differently,
Table 14.1. Manumission terms in the Digest

(# of fragments).
payment of price 190
- price set inter vivos (71)

- price set by testament (119)
render accounting 66
fixed term or fixed age 87
imposition of operae 101
no condition (testament) 31
no condition (inter vivos) 14
⁶⁷ If (part of) a testament containing a manumission clause was suppressed, the

slave could sue the heirs under a rescript from Marcus Aurelius and Commodus:
D. 5.1.53 (Herm. 1 epit.); D. 48.10.7 (Marc. 2 Inst.).
⁶⁸ Petr., Sat. 71.1–3.

such extensive discussion only makes sense if the purchase of
manumission was common, and freedom had its price.
14.6. FREEDOM PURCHASE AGREEMENTS

AND THEIR ENFORCEABILITY
Many slave systems in antiquity developed mechanisms whereby

slaves purchased their freedom (Patterson 1982: 282–4). The
best example is provided by the 1,200+ inscriptions from Delphi
(Hopkins 1978: 133–71), though manumission payments are encoun-
tered throughout the Hellenistic world (Zelnick-Abramovitz 2005:
207–47). Even so, a similar Roman practice is harder to substantiate
outside of the legal texts. Few inscriptions mention whether a freed-
man paid for his liberty or received it gratis.⁶⁹ Even when family
relationships make paid-for manumission unlikely, a gift is hardly
ever mentioned, which warrants the assumption that the reason for
manumission was simply not considered worth inscribing. Unlike
Greek manumission inscriptions, which were publicized on durable
temple walls, Roman deeds and state records were kept on perishable
materials. Freedom purchase appears quite often in the far better
preserved documentary evidence from Roman Egypt (Scholl
2001),⁷⁰ but no conclusions can be drawn about its prevalence, save
that informal manumission is more common than the iusta manu-
missio leading to Roman citizenship. Literary sources provide more
references to the purchase of freedom, starting as early as Plautus
(d. 184 ). When the slaves Stichus and Sagarinus encounter the
courtesan Stephanium, they exclaim that “the peculium is wafted
away, it is done; liberty has fled this slave.”⁷¹ Many Plautine slaves
have a peculium. It may be sizable through thrift or for comic effect,
or merely limited to some livestock, but the Plautine slaves generally
⁶⁹ CIL XI 5400 = ILS 7812 (doctor, 50,000 sesterces); CIL XII 5026 ([e]t pretio
[obtin]uit, quod prec[e]/non valuit); CIL VI 2211 = FIRA III 80g (free); CIL VI 20905
(free, but curse tablet). CIL II² 7.432 can be read either way.
⁷⁰ e.g. P. Oxy. IV 202 = Chr. 361 (per epistulam); FIRA III 11 = Chr. 362; P. Lips. II
151; P. Oxy. IX 1205; (inter amicos); P. Oxy. IV 722 = Chr. 358; P. Turn. 19 (in front of
agoranomoi); P. Flor. I 4 = Chr. 206; P. Oxy. XLIII 3117 (registration through court).
⁷¹ Plaut., Stich. 751: [Stich.] Vapulat peculium, actum est. [Sag.] Fugit hoc libertas
caput.
54 Egbert Koops
attempt to buy their liberty or that of loved ones once they acquire
some money.⁷² Other authors assume that slaves will work hard and
save, or rob and swindle, even cheat their own bellies to find the
necessary funds.⁷³ The promise of freedom is considered an effective
incentive for better work,⁷⁴ and although literary authors can con-
ceive of slaves who prefer sheltered servitude to insecure life on their
own,⁷⁵ such cases must have been rare in practice, considering
the grave disadvantages of “social death” (Patterson 1982: 86–94;
Hopkins 1993: 14). Martial refers to the advantages of “mastering”
oneself⁷⁶ and Trimalchio’s conlibertus Hermeros, who bought his
freedom for the high price of 1,000 denarii after forty years of
servitude (§14.4), also claims to have purchased the freedom of
his slave partner (contubernalis) to save her from groping hands.⁷⁷
“Liberty is favored above all things,” according to Gaius,⁷⁸ and though
the worth of freedom cannot be measured, its value can be priced.
Skilled slaves may have cost a fantastic sum in the past, Pliny the
Elder notes, but if that figure is topped in the present, it is because
these slaves are purchasing their freedom.⁷⁹ And when Vespasian
became emperor, his old herdsman complained that the fox had
changed his fur, but not his nature, so that he still needed to pay for
manumission; conversely, Virgil’s old shepherd Tityrus never had a
hope of freedom and hence never a thought of saving.⁸⁰
The purchase of freedom is ubiquitous in legal sources. In the
Digest alone, close to two hundred fragments deal with the compli-
cations arising from its various legal forms. Aside from simple pac-
tiones and the more elaborate obligation to render an accounting
⁷² Plaut., Asin. 497–8 (frugality); Plaut., Asin. 539–40 (flock); Plaut., Trinum. 727–8
(talent in the bank); Plaut., Trinum. 433–4 (peculium generally assumed). Purchase of
freedom is mentioned inter alia in Plaut., Asin. 650–2 and 673; Aul. 309–10; Most.
300; Pers. 34–8; Rud. 1408–10.
⁷³ Cic., Par. stoic. 5.39; Sen., Ep. 80.4; Pers., Sat. 5.73–99 and 5.174–5. Also see Ter.
Phor. 1.35–46 (saving); Dion. Hal., Ant. Rom. 4.24.4 (saving, theft); Plin., NH 33.6.26
(theft of food); Apul., Met. 10.14 (theft of food); Juv., Sat. 3.188–9 (bribes).
⁷⁴ Cic., Rab. Perd. 15. Also see Sen., Ep. 80.4.
⁷⁵ Plaut., Capt. 119–20 and 270–3; Epict. 4.1.33–5; Mart. 9.92. ⁷⁶ Mart. 2.68.
⁷⁷ Petr., Sat. 57.6: “I have purchased the freedom of my slave bedmate, so that no
one might wipe his hands in her hair” (contubernalem meam redemi, ne quis in
<capillis> illius manus tergeret).
⁷⁸ D. 50.17.122 (Gai. 5 Ed. prov.): Libertas omnibus rebus favorabilior est.
⁷⁹ Plin., NH 7.39.128–9. Also see Suet., Gramm. 13 (Staberius Eros) and 15
(Lenaeus); D. 12.4.3.5 (Ulp. 26 Ed.); Tac., Ann. 13.27 (Paris).
⁸⁰ Suet., Vesp. 16.3; Virg., Ecl. 1.26–35.

(§14.5 and 2), slaves could be sold under the obligation to manumit
(ut manumittatur); sold in trust against a payment from the slave’s
peculium or money borrowed to the purpose (emptio suis nummis);
manumitted by testament under a condition of payment (statuliber);
or left under a testamentary trust (fideicommissum) under the taker’s
obligation to manumit against payment. Many of these arrangements
were put on firmer legal footing by the emperors,⁸¹ though it should
be recalled that statuliberi already appear in the Law of the Twelve
Tables. Under the Trajanic SC Rubrianum ( 103),⁸² the praetor
could declare a slave free if the fiduciary taker under a legacy would
not free him and did not appear in court to explain himself (Buckland
1908: 611–20; Finkenauer 2010: 26–34). If they were not freed by
their fiduciary buyer, slaves who bought their freedom with their own
money (suis nummis) could claim their freedom in court under a
rescript of Marcus Aurelius and Lucius Verus ( 161–9, Buckland
1908: 636–40; Horsmann 1986; Finkenauer 2010: 44–50).⁸³ And by
a constitution of Marcus Aurelius and Commodus ( c.178), a slave
who was sold or, less often, donated to be manumitted by the new
master, was automatically free if the master did not comply (Buckland
1908: 628–36; Finkenauer 2010: 34–44).⁸⁴ This construction of a trans-
fer ut manumittatur was particularly suited to cases of self-sale into
slavery, which required the use of a proxy seller. The self-seller could
negotiate the conditions of servitude and future manumission
(including the grant of a peculium) by having the proxy seller transfer
him under the condition ut manumittatur, for instance after a certain
period of servitude or when a certain sum had been paid. Such a
construction gave the self-seller legal redress, though still a slave, if
the conditions were met but manumission did not follow. A similar
but far older rule applied to statuliberi, who became free immediately
once they either fulfilled the testamentary condition of paying money,
rendering an accounting, serving for a certain period, etc., or were
prevented from fulfilling that condition (Buckland 1908: 487–8,
492–505). In other words, slaves could actively proceed against their
⁸¹ The causes for this development are not clear. A lessened credibility of mere
moral commitments, favor libertatis, the greater importance accorded to will theories
and by extension to testaments, or even the desire to raise the proceeds of manumis-
sion taxes may all have played a part.
⁸² D. 40.5.26.7 (Ulp. 5 fideic.). ⁸³ D. 40.1.4 pr. (Ulp. 6 disp.) and following.
⁸⁴ D. 40.1.20 pr. (Pap. 10 resp.); D. 40.8.3 (Call. 3 cogn.); C. 4.57.2 (Alex.,  222);
C. 4.57.3 pr. (Alex.,  224).
56 Egbert Koops
masters in court on the strength of many manumission agreements,
though not on a mere pactio. This forms an important departure from
the rule that slaves had no capacity to appear in civil suits (Buckland
1908: 83–4; Garnsey 1981: 363).⁸⁵ Such varied manumission arrange-
ments imply negotiation between masters and slaves, but even more
tellingly a concern for the possibility of enforcing the bargain.
Two final arrangements need mentioning because they clarify the
limits of this system. The Roman law of civil procedure dictated that all
judgments were expressed in terms of money, which might suggest the
trick of colluding with a free person to bring a suit concerning
the status of a slave (Buckland 1908: 653, 665). If the master proved
slave status, the suit was lost and the damages would simply be paid.
Yet in opposition to the general principle of pecuniary condemnation,
Papinian explicitly states that a victorious master cannot be forced to
accept money if he wishes to take away the slave.⁸⁶ Negotiated manu-
mission always depended on the will and assent of the master (Gardner
1993: 14). A second trick only strengthens the point. A slave could give
someone a mandate to buy him and subsequently set him free.⁸⁷ Such a
mandate has essentially no legal force, although it may give rise to
obligations if the slave has indeed been sold and transferred, that is to
say if the master assents (Buckland 1908: 639–40; Heinemeyer 2013:
238–80). But since a slave has no authority to sell or transfer himself,⁸⁸
and cannot irrevocably bind the master to a mandate to sell (which can
always be withdrawn as long as there is no performance), it falls within
the master’s power to comply with the slave’s scheme or not (Heine-
meyer 2013: 270–1). For this reason, according to Ulpian, “In this case
I should be no more liable on the mandate than I am forced to sell
him.”⁸⁹ In the final analysis, manumission required the master’s assent.
⁸⁵ D. 5.1.53 (Herm. 1 epit.): slaves may inter alia appear in court against their
masters in cases of (conditional) testamentary or fideicommissary manumission,
fiduciary purchase with their own money (suis nummis), purchase with money loaned
to the purpose of manumission, and to have an arbiter appointed to supervise the
conditions of accounting.
⁸⁶ D. 40.12.36 (Pap. 12 resp.): Dominus qui optinuit, si velit servum suum abducere,
litis aestimationem pro eo accipere non cogetur. If he accepted the pecuniary condem-
nation however, the slave became a Junian Latin: C. 7.6.1.8 (Just.,  531).
⁸⁷ D. 17.1.19 (Ulp. 43 Sab.); D. 17.1.54 pr. (Pap. 27 quaest.); C. 4.36 l (Diocl./Max.,
 293).
⁸⁸ A slave is never part of his own peculium: D. 33.8.16.1 (Afr. 5 quaest.).
⁸⁹ D. 17.1.19 (Ulp. 43 Sab.): nec magis in hunc casum debeo mandati teneri, quam
ut eum tibi venderem.

14.7. THE PECULIUM AS A FREEDOM FUND
The linchpin of the Roman system of self-purchase by slaves is the

peculium, the slave patrimony. That certain slaves were allowed the
use of property that formally belonged to their master is again not
uniquely Roman. The Delphic manumission inscriptions imply
that the phenomenon was also known in the Greek world (Hopkins
1978: 147) and traces of a similar practice appear in Jewish law
(Cohen 1951: 150–86). What is characteristically Roman, however,
is the amount of legal attention devoted to the institution. In theory it
could not exist as the slave had no relevant legal capacity; but in
practice the jurists treated it in over 1,000 fragments of the Digest with
their eyes wide shut.⁹⁰ For the present purpose, only a few aspects
matter. It has been noted that the peculium could be granted and
taken away at any time, though this was undoubtedly circumscribed
by social norms. Whether masters always knew what their slaves
controlled is another matter entirely (Johnston 2002). Some placed
their money in a bank, lent it to debtors, or sent assets abroad, and
many will have had secret funds hidden away.⁹¹ The implication of
rendering an accounting as a condition for manumission (§14.2),
with the insistence on accounting in good faith, is that “squaring
up” was more important than the provenance of the funds.⁹² As
noted, some peculation was assumed, but evidently outweighed by
lower supervision costs.
Slave peculia could be extensive, containing “chattels, land, sub-
slaves and their peculium, and even claims against debtors,” or very
small, consisting only of clothes or some livestock.⁹³ Sizable holdings
⁹⁰ D. 40.1.4.1 (Ulp. 6 disp.): “First, it seems that it cannot be properly said that
someone is bought with his own money, since a slave cannot have his own money: but
with closed eyes it should be thought that he has been purchased with his own money,
as long as he is not bought with the money of the one who buys him” (et primo quidem
nummis suis non proprie videtur emptus dici, cum suos nummos servus habere non
possit: verum coniventibus oculis credendum est suis nummis eum redemptum, cum
non nummis eius, qui eum redemit, comparator).
⁹¹ D. 16.3.1.33 (Ulp. 30 Ed.); D. 40.7.40.6 (Scaev. 24 dig.); D. 44.7.14 (Ulp. 1 disp.).
⁹² D. 40.7.23.1 (Cels. 22 dig.).
⁹³ D. 15.1.7.4 (Ulp. 29 Ed.): in peculio autem res esse possunt omnes et mobiles et
soli: vicarios quoque in peculium potest habere et vicariorum peculium: hoc amplius et
nomina debitorum. Clothes: D. 15.1.25 (Pomp. 32 Sab.). Livestock: Varro, RR 1.17.7
and 1.19.3.
58 Egbert Koops
are attested for slaves in the imperial household,⁹⁴ though the legal
sources also contain references to private slaves owning money, wine,
buildings, cattle, gold and silver, and vicarii or sub-slaves.⁹⁵ Many
slaves erected costly inscriptions from their own money (de suo) for
themselves or for others: the masseur Arphocras, for instance, bought
space in a columbarium for 80 denarii and an ossuary for 175
denarii.⁹⁶ Slaves received some sort of allowance⁹⁷ and there are
glimpses of wage-earning or tenant slaves who made payments to
their absent masters,⁹⁸ presumably in exchange for greater independ-
ence. Alfenus reports that he was consulted on the case of a statuliber
who paid off his duty to perform day labor in money. These money
payments did not count toward fulfilling the manumission condition,
Alfenus explained, just as a money payment from the slave tenant of a
farm instead of a payment in kind would not count.⁹⁹ Even Cato, who
is not otherwise known as a considerate master, ran a paid brothel for
his slaves and lent them money to invest in other slaves and share in
the profit, both of which imply the grant of a peculium.¹⁰⁰
Several authors have argued that Roman businesses were structured
through the peculium to benefit from the legal advantages offered by the
master’s limited liability (Di Porto 1984; Cerami and Petrucci 2010:
61–7; Abatino et al. 2011). But in spite of occasional references,¹⁰¹ there
is preciously little evidence outside of the legal texts¹⁰² that economic
⁹⁴ e.g. CIL VI 5197 = ILS 1514 (a dispensator Scurranus with 16+ vicarii, 
14–37); Suet., Otho 5.2; Plin., NH 33.52.145.
⁹⁵ Money: D. 19.1.38 pr. (Cels. 8 Dig.). Wine: D. 33.6.9.3 (Ulp. 23 Sab.). Buildings:
D. 15.1.22 (Pomp. 7 Sab.); D. 33.8.6 pr. (Ulp. 25 Sab.). Cattle: D. 15.3.16 (Alf. 2 dig.).
Gold and silver: D. 14.4.5.13 (Ulp. 29 Ed.). Vicarii figure in sixty-four fragments from
the Digests (Morabito 1981: 111 n. 605).
⁹⁶ A direct link between the peculium and paying for an inscription de suo is found
in several inscriptions: CIL II 6338ff; CIL II² 7.981; CIL XI 6314 = ILS 3581; AE 1903,
140. Arphocras: AE 1980, 150 ( c.50).
⁹⁷ Sen., Ep. 80.8; Hor., Ep. 1.14.40; Petr., Sat. 75.4.
⁹⁸ Wage-earning: D. 12.6.55 (Pap. 6 quaest.); D. 19.2.60.7 (Lab. 5 post.); D. 40.7.3.8
(Ulp. 27 Sab.) and possibly D. 33.7.19.1 (Paul. 13 resp.). Tenancy: D. 15.3.16 (Alf. 2
dig.); D. 18.1.40.5 (Paul. 4 epit.); D. 20.1.32 (Scaev. 5 resp.); D. 26.7.32.3 (Mod. 6 resp.);
D. 33.7.12.3 (Ulp. 20 Sab.); D. 33.7.18.4 (Paul. 2 Vit.); D. 33.7.20.1 (Scaev. 3 resp.);
C. 4.14.5 (Gord.,  243).
⁹⁹ D. 40.7.14 pr. (Alf. 4 dig.). ¹⁰⁰ Plut., Cato 21.2 and 21.7.
¹⁰¹ Banking: Hipp., Refut. omn. haer. 9.7, but also see D. 14.5.8 (Paul. 1 Decr.).
Pasturage: Varro, RR 2.10.5.
¹⁰² The Digest contains various examples of economic activity structured through
the peculium. Shops: D. 14.4.5.13 (Ulp. 29 Ed.); D. 14.4.5.16–17 (Ulp. 29 Ed.) and
possibly D. 15.1.47 pr. (Paul. 4 Plaut.). Banks: D. 2.13.4.3 (Ulp. 4 Ed.). Inns and

activity was purposely structured in this manner, with the institor model
of direct agency appearing dominant instead (Aubert 1994; Fleckner
2010: 235; Mouritsen 2011: 175). Peculium-run businesses may have
been rather small (Abatino and Dari-Mattiacci in this volume) and
more often than not an extension of the family economy. Even so,
peculia undoubtedly formed an important aspect of the Roman slave
household, and it does not seem far-fetched to claim that most slaves
held one, especially if they were involved in tasks that were difficult to
monitor (Dari-Mattiacci 2013). Rather than an organizational principle
for Roman businesses, the peculium appears to have been a reward or
incentive for better work, both serving to ameliorate a slave’s condition
and to allow the prospect of eventual self-purchase. It was a recognized
legal principle that the peculium served as a “freedom fund,” and so a
statuliber became free if he was blocked from the peculium which he
needed to fulfill the condition of his release (Buckland 1908: 502–3).
The use of sub-slaves is of particular interest in this regard. Horace
explicitly connects the term vicarius to slave speech by opposing him to
the co-slave (conservus) that such a slave actually is.¹⁰³ Sub-slaves could
be employed to perform their slave master’s work, netting time for other
activities, or to enhance the slave master’s peculium, for instance by
being hired out to others. Aside from vicarii working as farmers, sailors,
captains, and shopkeepers (Morabito 1981: 112), the Digest also con-
tains the sinister example of sub-slaves put to work as prostitutes.¹⁰⁴
Perhaps the most revealing use of vicarii is that of delivering them pro
libertate, in exchange for freedom, either by the slave who is seeking
freedom or by a third party.¹⁰⁵ Thus, it is tacitly assumed for such
stables: D. 4.9.7.6 (Ulp. 18 Ed.). Mule driving: D. 19.2.60.7 (Lab. 5 post.). Shipping:
D. 4.9.3.3 (Ulp. 14 Ed.); D. 4.9.7.6 (Ulp. 18 Ed.); D. 9.4.19.2 (Paul. 22 Ed.); D. 14.1.1.22
(Ulp. 28 Ed.); D. 14.1.6 pr. (Paul. 6 brev.); D. 47.2.42 pr. (Paul. 9 Sab.). Dye factory:
D. 32.91.2 (Pap. 7 resp.). Clothing factory: D. 14.4.5.15 (Ulp. 29 Ed.); D. 15.1.27 pr.
(Gai. 9 Ed. prov.). Bakery: D. 33.7.18.1 (Paul. 2 Vit.). Managing conveyances:
D. 11.6.3.6 (Ulp. 24 Ed.). Prostitution: D. 3.2.4.3 (Ulp. 6 Ed.).
¹⁰³ Hor., Sat. 2.7.79–80. For the jurists, the term has a precise technical meaning: a
slave belonging to the peculium of another slave. Vicarii appear in many inscriptions
but the meaning is less clear there, since the term was also used as a functional
description (“assistant,” “underling”) for slaves belonging to the imperial household,
cities, or tax associations (Weaver 1972: 200–6).
¹⁰⁴ Hire: D. 14.3.11.8 (Ulp. 28 Ed.); D. 14.3.12 (Iul. 11 dig.). Prostitution: D. 3.2.4.3
(Ulp. 6 Ed.), though it should be noted that the master slave is considered infamis after
manumission.
¹⁰⁵ Slave: D. 41.3.4.16 (Paul. 54 Ed.); D. 41.4.9 (Iul. 3 Urs. Fer.). Third party:
D. 19.5.5.2 (Paul. 5 quaest.).
60 Egbert Koops
exchanges that a slave of equal market value is as good as any other
slave, and the only question that concerns the jurists is what happens if
title to the replacement slave is disturbed.
14.8. ESTABLISHING THE PRICE OF FREEDOM
It remains an open question whether Roman slaves paid a premium, a

reduced fee (Morabito 1981: 57), or market price (Garnsey 1981: 364)
for their freedom. All options are theoretically possible: a slave may
have valued his freedom higher than a potential buyer, but the price
may also have been lower out of liberality, or because the slave would
not have required guarantees. Much will have depended on specific
negotiations. The jurists assume an objective market price when
discussing the replacement value, although they are certainly aware
of subjective considerations. If a slave is killed, emotional attachment
plays no part in establishing the damages, though it might have raised
the cost of purchasing this specific slave, and the same rule applies if
the value of a slave’s labor is to be appraised.¹⁰⁶ Freeing a slave in
return for money does not lead to enrichment, as the money com-
pletely covers the loss of the slave, but if the slave pays more than his
value, presumably the market value, then the master is enriched to the
amount of the surplus.¹⁰⁷ A captured slave who has been ransomed
from the enemy by someone else and set free need only pay his true
master his estimated market price to acquire freedom.¹⁰⁸ If parents
sold a child into slavery, Constantine allowed them to reclaim the
child against payment of its value or delivery of a replacement
slave,¹⁰⁹ again stressing the connection between market price and
freedom. For calculations concerning the Falcidian fourth, release
under a fideicommissum, or manumission of someone else’s slave by
mistake, the market value is again generally used. In all, the over-
whelming tendency among the jurists is to value slaves at their market
price for replacement purposes.
¹⁰⁶ D. 9.2.33 (Paul. 2 Plaut.); D. 7.7.6.2 (Ulp. 55 Ed.).

¹⁰⁷ D. 15.3.2 (Iav. 12 Cass.); D. 15.3.3 pr. (Ulp. 29 Ed.). By implication, market
value and slave are interchangeable while the surplus constitutes a gift from the
peculium to the master.
¹⁰⁸ D. 29.2.71 pr. (Ulp. 61 Ed.).
¹⁰⁹ Fr. Vat. 34 (Const.,  313); C.Th. 5.10.1 (Const.,  329).

The scarcity of credible and comparable manumission prices
makes it difficult to establish a relation between the market price and
the price of freedom in practice.¹¹⁰ Almost all slave prices mentioned in
the Digest are unusable as evidence (Buckland 1908: 8)¹¹¹ and many
of the literary prices are suspect. In Delphi, Hopkins (1978: 158–63)
found evidence for a rising manumission premium for full freedom
due to increased prices as a result of market integration. If anything,
manumission prices seem to be on the high side as well in the docu-
ments from Roman Egypt, suggesting a premium, but the evidence
does not allow for any conclusions. The following prices are merely
given as an indication,¹¹² but they do show that manumission and
replacement at market value operate on the same plane, or put differ-
ently, that the price of freedom is not an order of magnitude higher
than the market value of slaves.
Circumstantial support comes from relating slave prices to prices
for other commodities. Ruffing and Drexhage (2008: 340–5) found
that the somewhat believable documented slave prices could be
expressed in terms of 250–1,000 day wages for the first century 
and 200–1,000 day wages for the second century. For Roman Egypt,
where better data are always available, prices lay between 1,200–2,000
day wages in the first century, between 250–2,800 day wages in the
second century, and between 400–1,125 day wages in the third
century, before the massive inflation of the last quarter of that century
(Table 14.2). Converting slave prices to wheat equivalent, Scheidel
(2005b: 2–6) arrived at prices that represented 2.7–3.2 years of
income for a rural laborer in Roman Egypt. Diocletian’s Edict on
maximum prices of  301 contains a schedule of slave prices, as
well as ample figures on wages and daily necessities (Lauffer 1971;
¹¹⁰ Interestingly, the Sumerian Code of Lipit-Ishtar from the nineteenth century 
contains a provision (§14) that a slave shall be freed if he compensates his master
twofold.
¹¹¹ Slave prices are mentioned in 133 fragments of the Digest. The price is almost
always 10, with 5 being half a slave or a pledged slave, and 20 being two slaves, a skilled
slave, or a case of mistake or deception. Justinianic price schedules aside, the only
actual price seems to be the fixed sum of 20 aurei that was to be paid by slaves who had
been set free under an invalid will yet wanted their freedom: D. 4.4.31 (Pap. 9 resp.);
D. 5.2.8.17 (Ulp. 14 Ed.); D. 5.2.9 (Mod., inoff. test.); D. 40.4.47 pr. (Pap. 6 quaest.).
Contra Morabito 1981: 54–9; Scheidel 2005b: 6.
¹¹² See Ruffing and Drexhage 2008: 321–36 for the completest list to date. On the
sale of slaves in Roman Egypt, see Straus 2004.
62 Egbert Koops
Table 14.2. Some Roman-Egyptian slave prices (in silver dr.).
Manumission prices have been bolded.¹¹³
Year () Price Age Source
c.80–100 1,161 23 P. Oxy. LXXV 5051

86 1,125 – P. Oxy. I 48 + XXXVIII 2843
91/92 900 32 P. Oxy. XXXVIII 2856
101 1,000 17 P. Turn. 19
107–15 500 0–19 P. Strasb. VI 505
c.100–25 700 4 BGU XI 2111
...
212 1,600 19 P. Oxy. XXXVI 2777
221 2,200 34 FIRA III 11
221 2,200 14 P. Vind. Bos. 7
225 1,600 9 SB XIV 11277
234 2,200 25 PSI III 182
c.235 1,600 – P. Oxy. XLIII 3117
Salway 2010).¹¹⁴ It suggests that a male unskilled slave between 16

and 40 (maximum price: 30,000 denarii) cost the equivalent of 4.8–6
years of unskilled day wages (maximum price: 20–25 denarii), at 250
working days per year. It is under debate whether slave labor and
wage labor stood in direct competition, and if so for which occupa-
tions (Harper 2010: 212–14, Groen-Vallinga 2017: 280–2). But even
so it is interesting to note that a slave working as a day laborer—
which included limited meals—would potentially earn back his pur-
chase price within a period of years rather than decades. The Price
Edict supplies enough information to fill out this amortization period
by using a ‘consumption basket’ of daily necessities. Adapting the
figures supplied by Allen (2009: 327–43) and assuming no rent
payments, with lodgings provided by the master, the investment in
an unskilled male slave of 18 years might be recouped between 4.7
years (“bare bones diet”) and 7.7 years (“respectable diet”), assuming
¹¹³ Prices in talents and copper drachmae have been converted using the scale: 1
talent = 6,000 copper drachmae = 100 silver drachmae. For tax purposes, 300 copper
dr. converted to 1 silver dr., but a conversion rate of 56:1 seems to have been used in
practice (Johnson 1936: 282–3). Note that P. Oxy. XXXVIII 2856 mentions both
prices, at a conversion rate closer to 69:1.
¹¹⁴ Since the Price Edict offers maximum prices, without explaining their proven-
ance or their internal relation, the schedule should be used with care. Scheidel 1996;
Harper 2010: 219–20.

250 working days per year. A late legal source implies an awareness of
this range: if a captive is ransomed from enemies, he has to repay the
ransom price or work off the debt in five years’ time.¹¹⁵ Skilled slaves
may have earned more, but their value was higher as well, since skill
level is an important determinant of market price.¹¹⁶ Diocletian’s
Edict expressly provides that the price for skilled slaves is at most
double that of unskilled slaves, which corresponds to the differenti-
ation between skilled (50–75 denarii per day) and unskilled labor
(20–25 denarii per day), and so the range stays roughly the same.
Interestingly enough, the high prices that are mentioned for certain
exceptional slaves (doctors, grammarians, actors), fall in the same
range when set against the comparably high figures for their wages:
the ratio is roughly 0.2 to 7 years of wages.¹¹⁷ Needless to say, these
figures need to be taken with a large grain of salt. Their purpose is
merely to draw attention to the fact that the overall ratio is one to
eight years of wages, tending to five to seven years, rather than twenty
to thirty years.
How long, then, would a slave have to labor to purchase his
freedom at a close to market price? Unfortunately, it is almost
impossible to relate manumission prices to specific terms of servitude.
Not a single source refers to the amount that a master would allow
slaves to keep for themselves in the form of wages. Seneca states that a
slave actor received 5 denarii and 5 modii of wheat for a show, but he
does not mention the master’s cut; Martial refers to a gratuity of
25 asses.¹¹⁸ Such figures are hardly enough to hazard a guess, and
in any case the amount of wages will have differed from case to
case. Other legal systems from antiquity such as Babylonian and
Mosaic law mention relatively short terms of servitude, three to six
years, although these relate mostly to debt bondage and legislation
¹¹⁵ C.Th. 5.7.2 = Const. Sirm. 16 (Hon.,  408) = C. 8.50(51).20.2 (Hon.,  409).
¹¹⁶ The jurists valued the use of a skilled slave against his market price, but the use
of an unskilled slave against the daily wages earned by his usual work: D. 7.7.6 pr.
(Ulp. 55 Ed.).
¹¹⁷ Doctor: 250,000 HS per year (Plin., NH 29.5.7, imperial physician), manumission
50,000 HS (CIL XI 5400 = ILS 7812). Actor: 200,000 HS per year (Cic., Q. Rosc. 23),
manumission 700,000 HS (Plin., NH 7.39.128). Grammarian: 100,000 HS per year (Suet.,
Gramm. 17), sale 700,000 HS (Plin., NH 7.39.128).
¹¹⁸ Sen. Ep. 80.5: Ille qui in scaena latus incedit et haec resupinus dicit [ . . . ] servus
est, quinque modios accipit et quinque denarios; Mart. 10.75: Sportula [ . . . ] quad-
rantibus arida centum [ . . . ] puero diximus esse datam.
64 Egbert Koops
on periodic debt remission (Wiedemann 1985: 166; Chirichigno
1993).¹¹⁹ Roman law is less forthcoming. As mentioned, the manu-
mission prices from the Digest yield no information; similarly, the
terms of servitude provide little insight (Wiedemann 1985: 169–74).
Testamentary provisions for terms of one, two, three, five, seven, ten,
and even fifteen years are mentioned, as well as provisions for pay-
ment of certain sums during or within three, five, or ten years. But
without any particulars of a slave’s age and status, such clauses mean
very little. Neither provisions that a slave is to be released after a
certain period of time (examples are one, two, three, five, eight, ten, or
twelve years), nor clauses that he is to be released when the heir or the
taker under a legacy dies or comes of age, can be related to a specific
period of servitude. If a manumission clause is coupled to a slave’s
age, it is almost always the 30 years of age that corresponds to the
lex Aelia Sentia.¹²⁰ The one exception occurs in a rescript from
Alexander Severus considering a slave girl named Firmia who was
sold at age 7, to be released at age 25.¹²¹
The single explicit reference to a common term of servitude comes
from Cicero, who informs the senate that it entertains some hope of
freedom after enduring six years of slavery, which is longer than
diligent captives and frugal slaves usually suffer.¹²² Cicero is referring
to the time that has passed since Caesar crossed the Rubicon, and so
the period is of limited value (Mouritsen 2011: 137). Nevertheless, the
reference would fall flat if slaves only had a chance of freedom after,
say, thirty productive years. Considering the social advantage of
gaining freedmen while recapitalizing, relatively short terms of servi-
tude seem likely for those who became slaves later in life, while the
practice of manumitting certain slaves at an early age (§14.3) may
have been the lot of those born into slavery or enslaved at a very
young age. Again, it should be remembered that some slaves were
never freed.
¹¹⁹ Codex Hammurabi 117 (in the fourth year); Exod. 21.2; Deut. 15.12–18; Philo,
Spec. leg. 2.84 (in the seventh year).
¹²⁰ D. 10.2.39.2 (Scaev. 1 resp.); D. 34.5.29(30) (Scaev. 18 dig.); D. 40.4.46 (Pomp. 7
var. lect.); D. 40.7.13.5 (Iul. 43 dig.).
¹²¹ C. 4.57.3 pr. (Alex.,  224).
¹²² Cic., Phil. 8.32: Etenim, patres conscripti, cum in spem libertatis sexennio post
sumus ingressi diutiusque servitutem perpessi, quam captivi servi frugi et diligentes
solent.

14.9. PATRONS AND FREEDMEN
Hopkins’s question, “why did the Romans free so many slaves?,” is

best answered by a model in which the benefits of allowing slaves
autonomy by the undisturbed use of their peculia, while factoring in
the cost of peculation, outweigh the gains of brute exploitation. It
implies that value extraction and recapitalization are delayed until an
agreed-upon price has been met, because a slave cum peculio will
eventually outperform a slave sine peculio (Alföldy 1972: 360–2). This
scenario is not so unlikely, particularly when supervision costs are
taken into account (Dari-Mattiacci 2013), as successful manumission
deals inspire confidence in other slaves, leading to better work and
subservience. But aside from the material benefit of better work at
lesser cost, and the immaterial benefit of strengthened social peace,
furthered by making some manumission agreements actionable, there
is another mechanism at work to ensure that masters keep their word.
The loss of a slave is sweetened by the acquisition of a freedman.¹²³
Few parts of Roman law are as convoluted as the law of patronage.
The rights of patrons take a number of forms, and these rights also
vary considerably according to the type of manumission. The differ-
ence between slaves manumitted as citizens and as Junian Latins has
already been mentioned (§14.4), but even for formally manumitted
slaves, the legal relationship to their patron was contingent on the
mode and conditions of manumission. Freedmen manumitted dir-
ectly and among the living out of kindness or generosity remained
firmly under their patron’s control and supervision. All family rela-
tions were destroyed by slavery, and so the manumittor was legally
closest to a parent, to whom a duty of obedience and deference was
owed (Mouritsen 2011: 36–51).¹²⁴ But as a general rule, a patron who
¹²³ The patron’s patrimony is enriched by acquiring a freedman. Thus, a manu-

mittor in good faith of a slave that was transferred to him in error is enriched to the
value of the operae and his claim to the freedman’s inheritance: D. 12.6.65.8 (Paul. 17
Plaut.). From the palingenesis, it is clear that the far too general statement in
D. 50.17.126.1 (Ulp. 15 Ed.), that no one is enriched who acquires a freedman, initially
only referred to the Sc. Iuventianum of  129 concerning the vindicatio caducorum.
See D. 5.3.20.6c (Ulp. 15 Ed.); D. 5.3.23 pr. (Ulp. 15 Ed.). A counter argument is not
provided by D. 19.5.5.5 (Paul. 5 quaest.) either, since the reason why the freedman’s
value is not calculated in that particular case is that the action is limited to quanti
interest mea servum habere in the first place.
¹²⁴ The duty of obsequium covers any number of rules, running from an inability to
sue the patron in certain suits or raise certain defenses, to an obligation enforced by
66 Egbert Koops
had received compensation for manumission, be it in payment or as
taker under a fideicommissum, had fewer rights against his freedman
(Garnsey 1981: 363–4). By extension, a trustee owner such as the
transferee ut manumittatur or the buyer in an emptio suis nummis
had even fewer rights, since he was merely lending his name to the
manumission and not making any personal sacrifice at all.¹²⁵ Finally,
the slave who was freed by testament, the orcinus, was under little
obligation at all (Loreti-Lorini 1925; Champlin 1991: 139) since his
true patron was dead and the patron’s heir, being next in line, had
made no sacrifice nor willed the manumission. All the more reason
existed to apply this rule if a slave was freed as a statuliber by
testament, for instance under the condition of rendering an account-
ing, paying a sum of money, or working a number of years, since the
heir even gained by the process when compared to direct manumis-
sion. If and when the condition was met, the former statuliber became
a freedman of the deceased testator and not of the testator’s heir.
Within this continuum, two categories of rights stand out for the
present purpose since they carry a direct monetary value. These are
claims on day labor or services (operae) and claims on the estate of a
freedman upon his death, either in the form of a right to intestate
succession or in the form of a claim for possession in contravention of
the will (bonorum possessio contra tabulas). An overview is provided
in the following, rudimentary table (Table 14.3).
The claim to labor duties may seem paramount among a patron’s
rights since it ensures the continued services of a freedman after
manumission. But this is not the case. Operae could only be imposed
upon freedmen who had been manumitted for free, and then only if
they had promised or sworn to perform such duties (Garnsey 1981:
364; Gardner 1993: 20, 29–32).¹²⁶ They could not be imposed if the
master had accepted any money in return for manumission or had
only served as a trustee manumittor, at least since the time of Hadrian
(Morabito 1981: 168–9). In other words, the obligation to perform
operae was mutually exclusive with paid-for manumission. It appears
public law to treat patrons with respect (Gardner 1993: 23–5). Similar considerations
of the “parental” nature of the relationship underlie a freedwoman’s need for permis-
sion to marry, reciprocal obligations of tutelage, and reciprocal claims for mainten-
ance (alimenta) between patron and freedman.
¹²⁵ D. 40.1.4.7 (Ulp. 6 disp.).
¹²⁶ D. 38.1.7.2 (Ulp. 28 Sab.); D. 38.1.31 (Mod. 1 reg.); D. 40.4.36 (Paul. 7 Plaut.).
Table 14.3. Patron rights and the effects of manumission. Adapted from Loreti-
Lorini 1925.
Manumission The patron acquires a right to . . .
occurred . . .
Inter vivos, and Labor Intestate Overrule a Be maintained Deference

(operae) succession testament (alimenta) (obsequium)
- for free Yes¹²⁷ Yes¹²⁸ Yes¹²⁹ Yes¹³⁰ Yes¹³¹

- by pactio No¹³² Yes¹³³ Yes¹³⁴ Yes?¹³⁵ Yes¹³⁶
- by transfer to No¹³⁷ Yes?¹³⁸ Yes¹³⁹ Yes?¹⁴⁰ Yes¹⁴¹
be freed
(ut manumittatur)
- by purchase suis No¹⁴² No¹⁴³ No¹⁴⁴ No¹⁴⁵ No¹⁴⁶
nummis
(cont. )
to have been the functional equivalent of borrowing the manumission

price from the master and paying it off in instalments. Freedmen
operae are accounted in labor days while the obligation to perform a
certain number of labor days as a slave is conversely mentioned as a
possible condition for manumission.¹⁴⁷ The analogy between paying
before and paying after manumission through labor is strengthened
by the fact that freedmen could agree to pay a sum of money instead
of performing labor, though patrons could not force them to do so
¹²⁷ C. 6.3.1 (Sev./Ant.,  204).

¹²⁸ Gai., Inst. 3.41; D. 26.4.3 pr. (Ulp. 38 Sab.); D. 34.5.9(10).2 (Tryph. 21 disp.).
¹²⁹ Gai., Inst. 3.41.
¹³⁰ D. 25.3.5.18 (Ulp. 2 cons.); C. 6.3.1 (Sev./Ant.,  204).
¹³¹ D. 37.15.7.4 (Ulp. 10 Ed.); D. 38.2.1 pr. (Ulp. 42 Ed.); D. 44.4.4.16 (Ulp. 76 Ed.);
D. 47.2.92(91) (Lab./Paul. 2 pith.); C. 2.41(42).2 (Just.,  531); C. 6.7.1 (Ant.,
 214).
¹³² C. 6.3.3 (Sev./Ant.,  206). ¹³³ C. 6.4.1 pr. (Sev./Ant.,  210).
¹³⁴ C. 6.4.1 pr. (Sev./Ant.,  210) contra D. 38.2.3.4 (Ulp. 41 Ed.), which is
generally amended.
¹³⁵ By inference: no direct source available.
¹³⁶ D. 37.15.3 (Marc., resp.); C. 6.6.3 (Alex.,  223).
¹³⁷ C. 6.3.2 (Sev./Ant.,  205); D. 38.1.13 pr. (Ulp. 38 Ed.).
¹³⁸ By inference: no direct source available. ¹³⁹ D. 38.2.3.3 (Ulp. 41 Ed.).
¹⁴⁰ By inference: no direct source available. ¹⁴¹ D. 2.4.10 pr. (Ulp. 5 Ed.).
¹⁴² C. 6.3.8 (Alex.,  224). ¹⁴³ C. 6.4.4.4 (Just.,  531).
¹⁴⁴ C. 6.4.1 pr. (Sev./Ant.,  210); C. 6.4.4.4 (Just.,  531).
¹⁴⁵ D. 25.3.5.22 (Ulp. 2 cons.).
¹⁴⁶ D. 2.4.10 pr. (Ulp. 5 Ed.); D. 37.15.3 (Marc., resp.); C. 6.3.8 (Alex.,  224).
¹⁴⁷ D. 38.1.3.1 (Pomp. 6 Sab.); D. 38.1.15.1 (Ulp. 38 Ed.); D. 40.7.4.4 (Paul. 5 Sab.).
Also see D. 40.7.14.1 (Alf. 4 dig.).
68 Egbert Koops
Table 14.3. Continued
Manumission The patron acquires a right to . . .
occurred . . .
Inter vivos, and Labor Intestate Overrule a Be maintained Deference

(operae) succession testament (alimenta) (obsequium)
By testament, and Labor Intestate Overrule a Be Deference

(operae) succession testament maintained (obsequium)
(alimenta)
- directly (orcinus) No¹⁴⁸ Yes¹⁴⁹ Yes¹⁵⁰ No¹⁵¹ Some¹⁵²

- by fideicommissum No¹⁵³ Yes¹⁵⁴ Yes¹⁵⁵ No¹⁵⁶ Some¹⁵⁷
(Morabito 1981: 91–2).¹⁵⁸ Again, the number of operae owed would

have been a matter of negotiation (Verboven 2012: 97). Seen in this
light, there is little economic difference between manumission among
the living under the imposition of operae, and manumission by
testament under the condition to perform a number of services or
pay a sum of money. The legal difference, however, is that manumis-
sion is already complete under the first option, while slavery
¹⁴⁸ D. 38.1.13.1 (Ulp. 38 Ed.). Changed by Justinian so that a patron’s son could
impose operae (Loreti-Lorini 1925: 53). See D. 38.2.29.1. (Marc. 9 inst.); D. 40.5.33 pr.
(Paul. 3 fideic.).
¹⁴⁹ Gai., Inst. 3.41. ¹⁵⁰ Gai., Inst. 3.41.
¹⁵¹ The argument is that a slave who was to be released under a fideicommissum that
the taker neglected to fulfill was declared free under the Sc. Rubrianum of  103 and
given an equal position to that of an orcinus: D. 40.5.26.7 (Ulp. 5 fideic.); D. 40.5.5
(Paul. 57 Ed.). By extension, the same held true for a slave bought suis nummis whom
his fiduciary master would not free: D. 5.1.67 (Ulp. 6 disp.). The true orcinus cannot
have been worse off than these orcini by analogy. Therefore, there probably was no
obligation to provide alimenta or act with full deference. See Loreti-Lorini 1925: 35–41.
¹⁵² See the previous footnote.
¹⁵³ D. 38.1.7.4 (Ulp. 28 Sab.); D. 38.1.13.1 (Ulp. 38 Ed.); D. 38.1.42 (Pap. 9 resp.);
D. 38.1.47 (Val. 6 fideic.); D. 38.2.29 pr. (Marc. 9 inst.); C. 6.3.5 (Ant.,  212);
C. 6.4.4,7 (Just.,  531). Changed by Justinian so that a patron’s son could impose
operae (Loreti-Lorini 1925: 53). See D. 38.2.29.1. (Marc. 9 inst.); D. 40.5.33 pr. (Paul. 3
fideic.).
¹⁵⁴ Fr. Vat. 225 (Pap. 11 quaest.); D. 38.2.29 pr. (Marc. 9 inst.); D. 38.16.3.1 (Ulp. 14
Sab.).
¹⁵⁵ Fr. Vat. 225 (Pap. 11 quaest.); D. 38.2.29 pr. (Marc. 9 inst.); C. 6.13.1 (Gord.,
 239).
¹⁵⁶ D. 25.3.5.22 (Ulp. 2 cons.); C. 6.4.4.7 (Just.,  531).
¹⁵⁷ D. 2.4.9 (Paul. 4 Ed.); C. 6.3.5 (Ant.,  212); C. 6.7.1 (Ant.,  214).
¹⁵⁸ D. 38.1.32 (Mod. 1 reg.); D. 38.1.39 pr. (Paul. 7 Plaut.).

continues under the second option, albeit under the unassailable
position of a statuliber.
The paramount right of the patron, then, lies in the claim to inherit
part of the estate of a freedman (Gardner 1993: 21–3) in addition to
either receiving a manumission price or imposing operae. This claim
travelled along the male line,¹⁵⁹ ensuring that testamentary manu-
mission, for example, did not overly diminish the family fortune. A
freedman could only exclude his patron from inheriting by leaving the
inheritance to his freeborn natural children. Even then, strengthening a
patron’s right to sizable estates (over 100,000 sesterces), the Augustan
lex Papia Poppaea of  9 provided that a patron was entitled to a share
equal to that of the freedman’s children, while it took three surviving
freeborn children to exclude the patron completely. In light of the
regular age at manumission, plausibly between 30 and 40, complete
exclusion was possible but perhaps not likely. This is even more true for
women, who tended to be manumitted at a later age (Gardner 1993: 22;
Scheidel 1997: 167–8).¹⁶⁰ In addition, it was made legally difficult to
resort to subterfuge to escape the patronal claim. Donating assets to
children prior to death would be of no avail, since two actions were in
place by the first century to rescind alienations by a freedman that
could defraud the patron, further protecting his interest (Lenel 1927:
352–3; Gardner 1993: 23).¹⁶¹
This well-protected expectation of inheritance further strength-
ened the predilection of masters to delay extraction from their slaves
until the payment of a manumission price, when they would receive
replacement value and gain a freedman, instead of settling for full
exploitation of a slave’s labor (Table 14.4). Manumission did not
amount to a full loss of the slave’s earning capacity. In fact, a patron
might benefit if a freedman did well for himself, which in turn
ensured a continuing supervision that was desirable to the state.
¹⁵⁹ The claim to a Junian Latin’s inheritance was a simple patrimonial right that did
not follow the rules of agnatic succession, but rather the normal rules of inheritance.
Gai., Inst. 3.58 and following.
¹⁶⁰ Freedwomen would have had to acquire the ius quattuor liberorum first (Kübler
1909), freeing them from their patron’s tutelage by giving birth to four freeborn
children, before they could make their own testament and even think of bypassing
their patron. Considering the pattern of female manumission past childbearing age,
acquisition of this right must have been limited to exceptional cases of early manu-
mission. An example is perhaps found in CIL VI 10246 (Septimia Dionisias).
¹⁶¹ The Fabian and Calvisian actions: D. 38.5 rubr. and C. 6.5 rubr.
70 Egbert Koops
Table 14.4. Claims to freedman inheritance (Gai., Inst.

3.40–3.44 and 3.56; Ulp., Epit. 29.1–29.3).
Freedman has . . . Patron receives . . .
Status of Junian Latin All

No heres suus All
Extraneus as heir (not a child) Half
1 Child as heir and under 100,000 HS Nothing
1 Child as heir and 100,000+ HS Half
2 Children as heir and 100,000+ HS One-third
3 Children as heir and 100,000+ HS Nothing
Freedwoman has . . . Patron receives . . .
No testament All
Testament, less than 4 children Patron decides
Testament, 4 living children One-fifth
Testament, 3 living children, 1 dead One-quarter
Testament, 4 dead children All
This legal structure also explains why the remainder of the peculium
could be handed to a freedman if anything was left after paying the
freedom price. Patrons might confidently expect that after serving as
an incentive and a freedom fund, the money would serve as venture
capital (Garnsey 1981: 367–70; Mouritsen 2011: 176–80), to be
returned to the family fortune with interest in good time. Since
many skilled freedmen went on to work in the family business, capital
accumulation could be monitored by the patron and was safeguarded
by economic family ties (Verboven 2012: 98, Groen-Vallinga 2017:
146–52).
The one modality that escapes this framework is the emptio suis
nummis (Heinemeyer 2013). Recognized by Marcus Aurelius, in this
legal construction the slave was purchased by a fiduciary buyer for the
express purpose of manumission, with money coming either from the
slave’s peculium or borrowed to the purpose. The seller was not
manumitting the slave and so he did not become patron, while the
buyer merely lent his name to the arrangement and made no sacrifice,
leading to a near-complete loss of patron’s rights including the right
to inherit. This particular arrangement would seem to defeat the
entire purpose of manumission for the original owner. Why not
revoke the peculium, or even contract with the slave through a

manumission pact, retaining the patron rights, instead of selling the
slave to a fiduciary buyer? As with slaves freed under the condition of
rendering an accounting—necessary to truthfully establish the size of
an inheritance—some slaves will have had enough power and influ-
ence that the realization of replacement value from their no doubt
hidden peculia required the surrender of patron rights. In this view,
emptio suis nummis is not the starting point of legal development
toward actionable manumission agreements, but rather the second-
century endpoint. Instead of an unattractive proposition that was to
be avoided in favor of other modalities (Gardner 1993: 37), or an
anomaly (Mouritsen 2011: 173–4), its existence is testament to the
bargaining power of certain well-placed slaves (Morabito 1981:
166–7). It put the enforceability of manumission arrangements
made inter vivos on an equal level to that of testamentary arrange-
ments and effectively neutralized almost all of a patron’s claims. As
such, it formed a logical pinnacle to the Roman system of negotiated
freedom payments.
14.10. CONCLUSION
Conditional manumissions are known from the Hellenistic world and

Roman Egypt (Zelnick-Abramovitz 2005). This chapter has suggested
that the Roman lawyers, at least, considered the fulfillment of certain
conditions the normal precursor to Roman manumission as well.
Important differences exist between Greece, Egypt, and Rome, par-
ticularly with regard to the residual power over conditionally freed
slaves, but the general point stands that relations between masters
and slaves were formalized by negotiations and agreements concern-
ing manumission. By enshrining an otherwise merely morally bind-
ing agreement in a testament, it became enforceable by the slave after
his master’s death. Gradually, as a result of imperial intervention,
most manumission agreements made inter vivos became enforceable
as well. The general effect of setting conditions prior to manumission
was that a patron’s hold over his freedman was lessened; or put
differently, that a patron who accepted a price for freedom lost part
of his bundle of rights. Aside from the emptio suis nummis, however,
the most important patron right remained intact, in the form of an
expectancy to inherit.
72 Egbert Koops
Agreements between master and slave could concern the period of
servitude, but more often referred to a manumission price. This price
was paid from a slave’s peculium, assets which masters allowed slaves
to hold as their own (Garnsey 1981: 364). Serving both as an incentive
and as a freedom fund, the slave peculium and the concomitant social
practice of purchasing freedom explain why the Romans freed so
many slaves. Though it is difficult to relate the size of the peculium to
the proceeds of a slave’s labor or the duration of slavery, its social
importance can hardly be underestimated. It gave slaves an interest in
their labor and the hope of advancement on the one hand, while
reducing supervision costs, increasing security and raising product-
ivity for their masters on the other. Though there was no legal
guarantee of tenure, the peculium and the negotiations and agree-
ments surrounding it show that at least a certain class of Roman
slaves, that is to say those encountered in literary and epigraphic
sources, rose well above the status of socially dead instrumentum
vocale. The practice of frequent and early manumission in exchange
for a negotiated and agreed upon price does not make the Roman
slave system any less brutal or exploitative during the period of
enslavement. But it does help to understand why masters, slaves,
and freedmen, living in close proximity, largely succeeded in main-
taining social order under the Roman empire.¹⁶²
REFERENCES
Abatino, Barbara, Giuseppe Dari-Mattiacci, and Enrico C. Perotti. 2011.

“Depersonalization of business in ancient Rome.” 31 Oxford Journal of
Legal Studies 365–89.
Alföldy, Géza. 1972. “Die Freilassung von Sklaven und die Struktur der
Sklaverei in der römischen Kaiserzeit.” 2 Rivista Storica dell’Antichita
97–129; repr. with add. in Helmuth Schneider, ed., Sozial- und
Wirtschaftsgeschichte der römischen Kaiserzeit, Darmstadt: WBG 1981,
336–71.
Alföldy, Géza. 2011. Römische Sozialgeschichte, 4th edn. Stuttgart: Steiner.
¹⁶² Research for this chapter was conducted under grant 400-09-009 from the
Netherlands Organisation for Scientific Research (NWO).

Allen, Robert C. 2009. “How prosperous were the Romans? Evidence from
Diocletian’s Price Edict (AD 301),” in Alan K. Bowman and Andrew
Wilson, eds, Quantifying the Roman Economy: Methods and Problems.
Oxford: Oxford University Press, 327–45.
Arzt-Grabner, Peter. 2010. “ ‘Neither a truant nor a fugitive’: some remarks
on the sale of slaves in Roman Egypt and other provinces,” in Traianos
Gagos, ed., Proceedings of the 25th International Congress of Papyrology.
Ann Arbor: Scholarly Publishing Office, 21–32.
Aubert, Jean-Jacques. 1994. Business Managers in Ancient Rome: A Social and
Economic Study of Institores, 200 ..–.. 250. Leiden/New York: Brill.
Bagnall, Roger S., and Bruce W. Frier. 1994. The Demography of Roman
Egypt. Cambridge: Cambridge University Press.
Bradley, Keith R. 1987. Slaves and Masters in the Roman Empire: A Study in
Social Control. Oxford: Oxford University Press.
Bradley, Keith R. 1994. Slavery and Society at Rome. Cambridge: Cambridge
University Press.
Buckland, William W. 1908 [repr. 1970]. The Roman Law of Slavery. Cam-
bridge: Cambridge University Press.
Cerami, Pietro, and Aldo Petrucci. 2010. Diritto commerciale romano: profilo
storico, 3rd edn. Turin: Giappichelli.
Champlin, Edward. 1991. Final Judgments: Duty and Emotion in Roman
Wills, 200 ..–.. 250. Berkeley: University of California Press.
Chirichigno, Gregory C. 1993. Debt Slavery in Israel and the Ancient Near
East. Sheffield: Sheffield Academic Press.
Cohen, Boaz. 1951. “Peculium in Jewish and Roman law.” 20 Proceedings of
the American Academy for Jewish Research 135–234.
Dari-Mattiacci, Giuseppe. 2013. “Slavery and information.” 73 Journal of
Economic History 79–116.
Di Porto, Andrea. 1984. Impresa collettiva e schiavo ‘manager’ in Roma
antica (II sec. a.C.–II sec. d.C.). Milan: Giuffrè.
Duff, Arnold M. 1928. Freedmen in the Early Roman Empire. Oxford:
Clarendon Press.
Emmerson, Allison A. C. 2011. “Evidence for Junian Latins in the tombs of
Pompeii?” 24 Journal of Roman Archaeology 161–90.
Fenoaltea, Stefano. 1984. “Slavery and supervision in comparative perspec-
tive: a model.” 44 Journal of Economic History 635–68.
Finkenauer, Thomas. 2010. Die Rechtsetzung Mark Aurels zur Sklaverei.
Stuttgart: Franz Steiner.
Finley, Moses I. 1985 [repr. 1999]. The Ancient Economy, 2nd edn. Berkeley:
University of California Press.
Fleckner, Andreas M. 2010. Antike Kapitalvereinigungen. Ein Beitrag zu den
konzeptionellen und historischen Grundlagen der Aktiengesellschaft.
Cologne: Böhlau.
74 Egbert Koops
Gardner, Jane F. 1993. Being a Roman Citizen. London/New York: Routledge.
Garnsey, Peter. 1981. “Independent freedmen and the economy of Roman
Italy under the Principate.” 63 Klio 359–71.
Groen-Vallinga, Miriam J. 2017. The Roman World of Work, (diss. Leiden).
Rotterdam: Optima Grafische Communicatie.
Günther, Sven. 2008. “Vectigalia nervos esse rei publicae”: Die indirekten Steuern
in der Römischen Kaiserzeit von Augustus bis Diokletian. Wiesbaden:
Harrassowitz.
Harada, Keikichi. 1939a. “Textkritische Studien zur adsignatio liberti orcini.”
59 Zeitschrift der Savigny-Stiftung 498–508.
Harada, Keikichi. 1939b. “Zwei Quellenstellen zum Patronatsrecht,” in Fest-
schrift Paul Koschaker I. Weimar: Böhlau, 401–3.
Harper, Kyle. 2008. “The Greek census inscriptions of late Antiquity.”
98 Journal of Roman Studies 83–119.
Harper, Kyle. 2010. “Slave prices in late antiquity (and in the very long
term).” 59 Historia 206–38.
Harris, William V. 1980. “Toward a study of the Roman slave trade.”
36 Memoirs of the American Academy in Rome 117–40.
Harris, William V. 1999. “Demography, geography and the sources of
Roman slaves.” 89 Journal of Roman Studies 62–75.
Heinemeyer, Susanne. 2013. Der Freikauf des Sklaven mit eigenem Geld—
Redemptio suis nummis. Berlin: Duncker & Humblot.
Hopkins, Keith. 1978. Conquerors and Slaves. Cambridge: Cambridge Uni-
versity Press.
Hopkins, Keith. 1993. “Novel evidence for Roman slavery.” 138 Past &
Present 3–27.
Horsmann, Gerhard. 1986. “Die ‘divi fratres’ und die ‘redemptio servi suis
nummis’ (zu den Motiven der epistula ad Urbium Maximum, Dig.
40,1,4).” 35 Historia 308–21.
Johnson, Allan C. 1936. Roman Egypt to the reign of Diocletian, in Frank
Tenney, ed., An Economic Survey of Ancient Rome, vol. II. Baltimore:
Johns Hopkins University Press.
Johnston, David. 2002. “Peculiar questions,” in Paul McKechnie, ed., Think-
ing Like a Lawyer: Essays on Legal History and General History for John
Crook on his Eightieth Birthday. Mnemosyne suppl. 231. Brill: Leiden,
5–13.
Kaser, Max. 1971. Das römische Privatrecht. Vol. I: Das altrömische, das
vorklassische und klassische Recht, 2nd edn. Munich: Beck.
Kleijwegt, Marc. 2002. “Cum vicensimariis magnam mantissam habet
(Petronius Satyricon 65.10).” 123 American Journal of Philology 275–86.
Kleijwegt, Marc. 2012. “Deciphering freedwomen in the Roman Empire,” in
Sinclair Bell and Teresa Ramsby, eds, Free at Last! The Impact of Freed
Slaves on the Roman Empire. London: Bloomsbury, 110–29.

Koops, Egbert. 2012. “Second-rate citizens: Junian Latins and the constitutio
Antoniniana.” 19 Maastricht Journal of European and Comparative Law
223–39.
Kübler, Bernhard. 1909. “Über das Jus liberorum der Frauen und die
Vormundschaft der Mutter. Ein Beitrag zur Geschichte der Rezeption des
römischen Rechts in Ägypten.” 30 Zeitschrift der Savigny-Stiftung 154–82.
Lauffer, Siegfried. 1971. Diokletians Preisedikt. Berlin: De Gruyter.
Lenel, Otto. 1927 [repr. 2010]. Das edictum perpetuum. Ein Versuch zu seiner
Wiederherstellung, 3rd edn. Leipzig: Tauchnitz.
Liebs, Detlef. 2000. “Das testament des Antonius Silvanus, römischer
Kavallerist in Alexandria bei Ägypten, aus dem Jahr 142 n. Chr.,” in
Klaus Märker, ed., Festschrift für Weddig Fricke zum 70. Geburtstag.
Freiburg: Alber, 113–28.
Lopez Barja de Quiroga, Pedro. 1998. “Junian Latin: status and number.” 86
Athenaeum 133–64.
Loreti-Lorini, Bradamante. 1925. “La condizione del liberto orcino nella
compilazione giustinianea.” 34 Bullettino dell’Istituto di Diritto Romano
29–66.
Morabito, Marcel. 1981. Les réalités de l’esclavage d’après le Digeste. Paris:
Belles Lettres 1981.
Mouritsen, Henrik. 2005. “Freedmen and decurions: epitaphs and social
history in imperial Italy.” 95 Journal of Roman Studies 38–63.
Mouritsen, Henrik. 2011. The Freedman in the Roman World. Cambridge:
Cambridge University Press.
Patterson, Orlando. 1982. Slavery and Social Death: A Comparative Study.
Cambridge, MA: Harvard University Press.
Roth, Ulrike. 2010. “Peculium, freedom, citizenship: golden triangle or
vicious circle? An act in two parts,” in Ulrike Roth, ed., By the Sweat of
your Brow: Roman Slavery in its Socio-Economic Setting (BICS Suppl. 109).
London: Institute of Classical Studies, 91–120.
Ruffing, Kai, and Hans-Joachim Drexhage. 2008. “Antike Sklavenpreise,”
in Peter Mauritsch, ed., Antike Lebenswelten: Konstanz—Wandel—
Wirkungsmacht. Festschrift für Ingomar Weiler zum 70. Geburtstag
(Philippika 25). Wiesbaden: Harrassowitz, 321–51.
Salway, Benet. 2010. “Mancipium rusticum sive urbanum: the slave chapter
of Diocletian’s edict on maximum prices,” in Ulrike Roth, ed., By the Sweat
of your Brow: Roman Slavery in its Socio-Economic Setting (BICS Suppl. 109).
London: Institute of Classical Studies, 1–20.
Scheidel, Walter. 1996. “Reflections on the differential valuation of slaves in
Diocletian’s Price Edict and in the United States.” 15 Münstersche Beiträge
zur Antiken Handelsgeschichte 67–79.
Scheidel, Walter. 1997. “Quantifying the sources of slaves in the early Roman
Empire.” 87 Journal of Roman Studies 156–69.
76 Egbert Koops
Scheidel, Walter. 2005a. “Human mobility in Italy II: the slave population.”
Scheidel, Walter. 2005b. “Real slave prices and the relative cost of slave labor
in the Greco-Roman world.” 35 Ancient Society 1–17.
Scheidel, Walter. 2008. “Roman population size: the logic of the debate,” in
Luuk de Ligt and Simon J. Northwood, eds, People, Land, and Politics:
Demographic Developments and the Transformation of Roman Italy, 300
— 14. Leiden: Brill, 17–70.
Scholl, Reinhold. 2001. “ ‘Freilassung unter Freunden’ im römischen Ägyp-
ten,” in Heinz Bellen and Heinz Heinen, eds, Fünfzig Jahre Forschungen
zur antiken Sklaverei an der Mainzer Akademie, 1950–2000. Stuttgart:
Steiner, 159–69.
Sirks, A. J. Boudewijn. 1981. “Informal manumission and the lex Junia.” 28
Revue Internationale des Droits de l’Antiquité 247–76.
Sirks, A. J. Boudewijn. 1983. “The lex Iunia and the effects of informal
manumission and iteration.” 30 Revue Internationale des Droits de l’Anti-
quité 211–92.
Straus, Jean A. 2004. L’achat et la vente des esclaves dans l’Egypte romaine:
contribution papyrologique à l’étude de l’esclavage dans une province
orientale de l’Empire romain. Munich: Saur.
Temin, Peter. 2001. “A market economy in the early Roman Empire.” 91
Journal of Roman Studies 169–81.
Temin, Peter. 2004. “The labor market of the early Roman Empire.” 34
Journal of Interdisciplinary History 513–38.
Treggiari, Susan. 1975a. “Jobs in the household of Livia.” 43 Papers of the
British School at Rome 48–77.
Treggiari, Susan. 1975b. “Family life among the staff of the Volusii.” 105
Transactions of the American Philological Association 393–401.
Verboven, Koenraad. 2012. “The freedman economy of Roman Italy,” in
Sinclair Bell and Teresa Ramsby, eds, Free at Last! The Impact of Freed
Slaves on the Roman Empire. London: Bloomsbury, 88–109.
Wacke, Andreas. 2001. “Manumissio matrimonii causa. Die Freilassung
zwecks Heirat nach den Ehegesetzen des Augustus,” in Heinz Bellen and
Heinz Heinen, eds, Fünfzig Jahre Forschungen zur antiken Sklaverei an der
Mainzer Akademie, 1950–2000. Stuttgart: Steiner, 133–57.
Watson, Alan. 1987. Roman Slave Law. Baltimore: Johns Hopkins University
Press.
Weaver, Paul R. C. 1972. Familia Caesaris: A Social Study of the Emperor’s
Freedmen and Slaves. Cambridge: Cambridge University Press.
Weaver, Paul R. C. 1990. “Where have all the Junian Latins gone? Nomen-
clature and status in the early Empire.” 20 Chiron 275–305.

Weaver, Paul R. C. 1997. “Children of Junian Latins,” in Beryl Rawson and
Paul R. C. Weaver, eds, The Roman Family in Italy: Status, Sentiment,
Space. Oxford University Press: Oxford, 55–72.
Westermann, William L. 1955. The Slave Systems of Greek and Roman
Antiquity. Philadelphia: American Philosophical Society.
Wiedemann, Thomas E. J. 1985. “The regularity of manumission at Rome.”
35 The Classical Quarterly 162–75.
Zelnick-Abramovitz, Rachel. 2005. Not Wholly Free: The Concept of Manu-
mission and the Status of Manumitted Slaves in the Ancient Greek World
(Mnemosyne suppl. 266). Leiden: Brill.
15
Banking, Money-Lending, and Elite

Financial Life in Rome
Jean Andreau
In the Roman world the practice of making loans at interest was

widespread and quite ancient. “In truth, the practice of making loans
at interest was a perennial problem in Rome, and a frequent cause of
sedition and strife” (Sane vetus urbi faenebre malum et seditionum
discordiarumque creberrima causa), in the words of Tacitus (Tac.
Ann. 6.16.1). He adds that the Law of the Twelve Tables, in order to
remedy this evil, stipulated that interest should be limited to a twelfth
part (unciarium fenus). What is meant by a “twelfth”? The twelfth
part of the loaned capital, annually? It may rather have been 1/12 of
this capital, but repaid every month, as Zehnacker (1980) has sug-
gested, I think rightly. In any case allusion to the Twelve Tables attests
that among the Latins, charging interest for loans had already become
widespread in the mid-fifth century . Livy also mentions social and
political crises in the fifth and fourth centuries  caused by heavy
indebtedness affecting large numbers of people, especially the poor
(Liv. 6.11–20; 7.16.1; 7.21.3; 7.27.3–4). At the time of the Twelve
Tables, in the fifth century, the city of Rome had not yet begun to
strike discoid coins. Some Greek cities in southern Italy or Sicily
minted disc-shaped coins in silver, but it does not appear that these
coins circulated freely in Roman territory. Bronze bars were used
there, both as standards of value and as a medium of exchange. Some
of these bars bore no markings, and they were known as aes rude
(rough or raw bronze); others were scored with marks like the
branches of trees, so that they are now called “ramo secco” (Thomsen
Jean Andreau, Banking, Money-Lending, and Elite Financial Life in Rome In: Roman Law and Economics:
Exchange, Ownership, and Disputes, Volume II. Edited by: Giuseppe Dari-Mattiacci and Dennis P. Kehoe,
DOI: 10.1093/oso/9780198787211.003.0015
82 Jean Andreau
1978; Zehnacker 1990). Charging interest was possible, then, even in
the absence of disc-shaped coins. Even at that period, the activity of
lending was not limited to lending food, or payment in kind. There
were other kinds of loans.
Repeated crises brought things to the point where in 347 
interest rates were limited to 1/24th part, half the amount of the
previous rate. Then in 342, through the lex Genucia, charging interest
was completely prohibited in the city of Rome (Liv. 7.42.1). This is the
only time in the entire history of ancient Rome that such a prohib-
ition was enacted. Probably, the edict was never formally repealed (on
the situation at the beginning of the first century , see App., Bell.
civ. 1.54). But was it ever applied? We just do not know. And if it was
applied, for how long? Probably not for long. In fact between 318 
and 310, deposit bankers appeared in the Forum (the argentarii). At
the same time, Rome began to mint silver coins called Romano-
Campanian didrachmas. A century later (late third century, early
second century), at the time when Plautus and Cato the Elder lived,
such a prohibition was completely out of the question, and so things
would remain up to the time of the end of the Western Roman
Empire, in the fifth century . No such prohibition would be
considered in all that time. This does not mean that public author-
ities, in Italy and in the provinces, would not make attempts to
regulate or even limit interest rates, as we shall see in what follows.
During the last three centuries of the Republic, between the end of
the fourth century and Augustus’ reign, the pace of economic and
commercial life increased, driven by the conquests of Rome. Such
developments cannot be quantified, but there is no serious doubt
about the matter. The size of the estates of members of the elite of
Rome became significantly larger (on average), and the lifestyle of this
group was transformed. The Roman tradition may have greatly
exaggerated the supposed austerity of the lives of the Roman senators
of previous centuries, but there is no doubt that at the time of Cicero,
members of Italian elites (senators, equestrians, municipal aristo-
crats) were in general richer and lived lives of luxury in comparison
with the elites of a few centuries before. The practice of charging
interest continued, and it created problems as before. Indebtedness
probably increased as the size of estates increased. It is not possible to
measure, though, the speed with which this took place, in Italy or in
conquered territories. However, a fair number of documents show
that charging interest was a common practice.
Banking, Money-Lending, and Elite Financial Life in Rome 83

With regard to the first century , it can be supposed that all
wealthy or well-off men and women borrowed money or lent it, at
one time or another. Many of these loans were secured, for instance
by pignora. The same people both borrowed and lent, depending on
their needs, at different times in their lives or even at the same time,
depending on the case. Women of a certain social level did not
specialize in lending money as a profession, but they could find
themselves managing estates and engaging in either borrowing or
lending. The documentation we have concerning the poor (in cities as
well as in the countryside) is much more limited than what is
available for rich people; but a fair number of testimonies and
indicators show that the poor frequently got into debt, such that
many of them were always in debt; they also show that some poor
people practiced low-level money-lending themselves.
Shatzman has compiled a list of twenty-five senators known to
have engaged in money-lending (1975: 76), and Nicolet made a list of
the names of seventeen equestrians (1966: 372–3). In view of the
amount of documentation available and the nature of that documen-
tation, this is a very significant sample. Further, at the end of the
Republic, and under the early Principate, lending is mentioned in all
the general texts that concern wealth and the sources of income of the
rich. In addition to owning lands, livestock, personal residences, and
residential investment properties, slaves and precious objects such as
gold, a rich person of note (not necessarily a senator or equestrian) is
likely to be a creditor (Petr., Sat. 37.8–9; Sen., Ad Lucil. 2.6 and 4.41.7;
Tac., Hist. 1.20.3; Juv., Sat. 11.39–41; Tert., Cult. fem. 1.9.3; etc.).
As for this last century of the Republic, there are also more
references to the way such a source of income must be handled.
The wealthy man is normally helped out by his family, by friends, or
by those who depended on him, such as slaves or freedmen: he
could borrow from them with more convenient conditions and
receive information from them. But in any case, he must find out
about the character of those to whom he lends, and their solvency.
He must pay attention to repayment dates and other aspects of the
loan agreement, and know when it is time to sue a bad debtor,
although Saller (1982: 121) insists that creditors exhausted all
other possible means of settling debts before bringing their debtors
into courts of law. In Rome, creditors posted complaints about
deadbeats and their bad faith upon the columna Maenia; Cicero,
in the Pro Cluentio, speaks of one Quintus Manilius who lost his
84 Jean Andreau
reputation in this manner (Cic., Pro Cluent. 13.38–9; see Andreau
1987b: 163–4).
Obviously sums great and small were lent out, and financial activity
of this kind was not of the same importance for everyone. Loans did
not all carry the same rate of interest or the same conditions. Even
among members of the same social group, for example senators or
equestrians, there were differences with regard to loans made at
interest. Many senators needed to borrow at various times in their
political careers, or for reasons pertaining to their social lives, but
such situations did not mean they might not still lend money, for
example to credit intermediaries. One year, a senator might need to
borrow a great deal of money for a political campaign; another year,
he might be obliged to splurge on his daughter’s wedding. But at other
times, they might profit from certain very lucrative political functions
or from inheritances. Not all the members of the elite were trained
financiers, although some of them might handle very large sums.
Cicero himself is an excellent example. He borrowed often, and in
large amounts, but he also engaged in money-lending. One of his
letters shows that he and Atticus both did, Cicero through an inter-
mediary, Cluvius, and Atticus with the assistance of Vestorius (Cic.,
Ad Att. 6.2.3; see Früchtl 1912; Shatzman 1975: 416–22). According
to Rauh, the debts Cicero held in 45  amounted to an enormous
sum, possibly several million sesterces (Rauh 1989: 60–9). But Cicero
was not a trained financier.
It was fashionable to rail against money-lending, but well-known
Greeks and Latins do not appear to have tried very hard to hide the
fact that they were money-lending, except when there was something
illegal about it (for example if the interest was high enough to
constitute usury). If charging interest, which was not illegal, had really
been the subject of a moral and social taboo among members of the
elite, the number of general mentions of it, and the number of loans
we hear about, would not be so large. Charging interest was criticized
from a moral standpoint, but it was also accepted by Roman public
opinion.
Who were Cluvius and Vestorius, two men who are mentioned
several times in Cicero’s letters? What activity were they engaged in?
They lived and worked in Puteoli (Pozzuoli). They had a relationship
with the businesspeople of Puteoli, but what exactly was it? Cluvius
owned many things, including real estate (commercial buildings
and gardens). Vestorius must have had interests in manufacturing,

because in a passage in Vitruvius and another in Pliny the Elder, he is
said to have been the first in Pozzuoli to produce the dye or pigment
Egyptian blue, and to have created a new variety of that blue (Vitr.
7.11.1; Plin., N.H. 33.162). From the way Cicero talks about him, he
knew a great deal about accounting and financial practices. At that
time, in Italy, there were professional bankers (argentarii), as well as a
variety of types of financiers and money-lenders, at various levels of
the social hierarchy. Were Cluvius and Vestorius, who appear not to
have belonged to the equestrian order, really bankers, or something
different, financiers of another sort? On the basis of the available
information, I am convinced that they were not professional bankers.
But this cannot be established with complete certainty; if someone
wishes to think of them as bankers, I cannot prove this wrong.
Both interpretations can be defended (D’Arms 1981: 49–55; Andreau
1997: 99–118).
Aristocrats who borrowed and lent money at interest, according
to the requirements of the moment, were concerned above all with
their estates, the dowries of their wives, and perhaps with the
estates of their close friends. Thus, they were obliged to know how
to maintain buildings and how to manage their lands and their
money. Some of them, all while leading political and social lives in
accordance with their rank, showed themselves to be particularly
greedy, even rapacious.
Such was the case for the senator Marcus Junius Brutus, with
whom Cicero had dealings in his capacity as governor of the province
of Cilicia and Cyprus, in the year 51–50 . The city of Salamis in
Cyprus, in the year 56 , borrowed a large sum of money (the exact
amount is not known) from two Italians based in Cilicia, Marcus
Scaptius and Publius Matinius. When Cicero arrived in Cilicia in 51
, the debt had not been repaid. The people of Salamis thought that
the two businessmen were the actual creditors, and at first Cicero also
believed this; but later he learned that the capital had come from
M. Junius Brutus, and the risk in the lending operation was his
problem. Brutus had taken advantage of the official mission that
had taken him to Cyprus in 58, in the company of the senator who
would later be known as Cato of Utica, to make the deal to loan the
money. His two intermediaries and he demanded a usurious interest
rate; Brutus contacted a succession of provincial governors in order to
put pressure on the citizens of Salamis (Cic., Ad Att. 5.21.10–13;
6.1.3–8; 6.2.7–9; 6.3.5–6; see Andreau 1997: 109). Brutus was not a
86 Jean Andreau
professional financier, but he was obviously more avid for profit
than Cicero.
Titus Pomponius Atticus, Cicero’s friend and brother-in-law, was
better acquainted with the business of finance than either Cicero or
M. Junius Brutus; still, he could not be counted among the greediest
members of the elite. There were varying degrees separating the most
disinterested members of the elite and those who were the least
knowledgeable about finance on one hand, and those who were the
greediest or the most experienced on the other. Among the greedy, we
find professional money-lenders (the faeneratores this word means
lender at interest), a certain number of equestrians who controlled the
companies of publicans—companies specializing in tax collection,
among other things, and a number of Italians who had set up their
businesses in different provinces.
Even if the imperial regime changed the political context, and even
if the importance and influence of companies of tax collectors grad-
ually decreased, such financial activities undoubtedly continued
under the Principate. But with regard to the Principate, we do not
have discourses such as those of Cicero, nor any correspondence
comparable to his. The available evidence is less weighty.
15.1. SOCIAL PRACTICES OF GRATUITY

AND GENEROSITY
Aristocrats of the first century  also made loans without charging

interest or at very low rates, by reason of friendship, or in order to
help their relatives or clients; they might themselves stand surety for
their relatives, friends, or clients, but they also made outright gifts.
These financial operations having to do with some degree of gener-
osity have been studied in depth and in a very subtle way, in recent
decades, by Ioannatou (2006: 229–307) and Verboven (1993, 2002).
Members of the elite were pleased to express admiration for disinter-
ested actions and generous gestures, and they prided themselves on
acting this way themselves, as we see in treatises by Cicero (De officiis)
and Seneca (De beneficiis). In truth, in most cases the practice of
giving gifts or no-interest (or low-interest) loans, even if it was part of
the ideology of the elites, was not completely disinterested, because

the lenders in these cases were most often hoping to motivate
symbolic or political results that would benefit them. As Ioannatou
wrote (2006: 244), in such actions “the social aspect was far more
important than the financial aspect.” The gifts and loans made with-
out interest (or at low interest rates) were part of social habits, and
were related to a social hierarchy (even within the aristocratic milieu).
They were often involved in exchanges between family members
or political allies, involving a series of gifts and counter-gifts (see
Verboven 2002; Ioannatou 2006).
Three remarks must be made with regard to the interest-free loans.
First, it appears that they were rarely completely obligatory. Most
of the time, the father of an aristocratic family had some latitude
in choosing, even if, after the fact, he presented his choice as an
absolute duty which could not be avoided. Secondly, such gestures
are primarily known to us as occurring in the first century , because
of the existence of Cicero’s works. But they were already made before
the first century , and continued to exist under the Principate.
Thirdly, some historians might be tempted to insist on the reality
of these gifts from generosity, as opposed to financial practices
imbued with greed and self-interest. Others might give more credence
to the inverse tendency, and emphasize the predominance of greed.
But neither of these positions is satisfying, because the desire for gain
and riches existed alongside altruistic social tendencies. These two
ways of doing things were always present in the city of Rome and in
the Roman Empire, although beginning with Augustus, the emperor
often tried to limit the effects of the most extreme examples.
As for the law, loans at interest and loans without interest gave rise
to contracts that were only partly identical. A stipulation could be
made in either case, and the mutuum as well (even if in principle, the
mutuum was gratuitous), because it could be accompanied by a
stipulation of interest. But certain contracts, whose gratuitous status
was better respected, were better adapted to interest-free loans. This is
the case with the mandatum and the commodatum. The latter was a
gratuitous contract that designated movable and immovable goods to
be loaned for use, and which belonged to the category of good offices
and services rendered as between friends or neighbors (Michel 1962:
95–127 and 168–97). One may say with Ioannatou (2006: 230–1) that
interest-free loans “fitted the legal pattern of the mutuum and the
commodatum,” without forgetting that the mutuum was also used in
the case of loans for which interest was charged.
88 Jean Andreau
15.2. THE BEGINNINGS OF BANKING IN
ROME AND ITS DEVELOPMENT (FOURTH
TO FIRST CENTURIES )
Let’s return to the matter of loans at interest and financial affairs.

Between the end of the fourth century  and the beginning of the
common era, not only did the volume of financial operations in Rome
and Italy certainly increase by a very large amount, (although it is
difficult to prove this), but the nature of these operations became
more complex, as the world of business became more diversified.
On one hand, deposit banks, which had existed in the Greek world
since the second half of the fifth century, began to operate in Rome at
the end of the fourth century, and were soon found in other cities in
Italy and the western Mediterranean. In the last decades of the
Roman Republic, we know that several deposit bankers, called argen-
tarii, operated not only in Rome itself, but in western Mediterranean
cities such as Pompeii, Regium, Lepcis Magna, and Syracuse
(Andreau 1987a: 313–29). Later, under the Principate, inscriptions
attest the presence of some banks in a larger number of cities, but, as
regards the activity of these professions, the situation that can be
observed in the Ciceronian period remains more or less the same
until the second part of the third century .
When they appeared at Rome, these bankers undoubtedly verified
the value of coins and operated as money-changers, but without more
documentation it is impossible to know what other services they
might have provided for their clients at that time. A century later,
during the time of Plautus, we can see that they received deposits,
made loans, and provided routine banking services for their clients
(cashing drafts, making payments to and on behalf of their
clients, allowing withdrawals, etc.).
Further, during the second half of the second century , bankers
in Rome and Italy began to provide a service that Greek bankers had
not provided: they furnished credit for purchases made at auction.
Auctions were held in Greece as well as in Rome, but in the Greek
world bankers did not play the same role in auctions as in Rome (see
Thielmann 1961; Andreau 1974; Andreau 1987: 583–97; Petrucci
1991, 2002; Garcia Morcillo 2005, 2008, 2014). In the Roman case,
at an auction, or not long afterward (a few days), the banker, the
argentarius, would pay the seller the auction price for objects, slaves,

or supplies purchased at auction, loaning the money to the buyer,
who would reimburse the banker several months later (not more than
a year, in the examples we know) and pay interest. Several texts from
Cicero attest the existence of this auction credit during the times
when he lived (see for example Cic., Pro Caec. 6.16–17 and 10.27).
And for the Julio-Claudian period, 153 tablets were found at Pompeii,
tablets that had belonged to the argentarius or to the coactor argen-
tarius Lucius Caecilius Jucundus, of which 137 were receipts from
auctions, given by the seller to the banker, attesting that he had
received from the banker the price of sale. The other sixteen tablets
had to do with business transactions (rentals, tax collection) the
banker had concluded, during the period  50–60, with the city of
Pompeii (Andreau 1974; Jongman 1988).
In the De agricultura of Cato, mention is made of a coactor, that is,
a private person who handles sums of money on behalf of his clients,
collecting payments and debts, without however acting as a banker
(Cato, De agr. 150.2). The father of Horace was a coactor in Venosa
(Hor., Sat. 1.6.86), and we know of a few other coactores, at the end of
the Republic and under the Principate. Finally, two other occupations
appear at the end of the second century , or during the first century
: the nummularii used to verify the value of coins and change
money, but they were not deposit bankers (the nummularii would
not become deposit bankers until the first half of the second century
). And there were also coactores argentarii, who were at one and
the same time deposit bankers and coactores.
Professional bankers, argentarii and coactores argentarii, all loaned
money, as is shown by the documentation on credit provided for
auctions, but these people were never confused with money-lenders
who charged interest yet were not bankers (Barlow 1978: 67–80;
Andreau 1999: 9–49). Plautus borrows his characters who are pro-
fessional bankers from Greek models, and he calls them tarpezitai or
trapezitai, after their Greek names, or just argentarii, the Latin term.
He never speaks as if they were the same as money-lenders who were
not bankers, a status he indicates by using the Greek word daneistai
or by the Latin expression faeneratores (on banking in Plautus and
Terence, see Andreau 1968).
Plautus calls his comedies palliatae, and in them he shows himself
to be the inheritor of Greek comic traditions; he in fact claims this.
His plays take place in Greek cities and feature characters with Greek
names. But he also describes and satirizes Roman mores. What he
90 Jean Andreau
says about bankers and credit is thus part of the Greek reality of his
times, but is also part of Roman reality—although he distorts all these
situations in accordance with the comic effect of his language. His
plays are valuable sources of information about Roman life. The
distinction he established between professional bankers (argentarii)
and money-lenders (faeneratores) is part of the information we can
glean from his writing, and what he says in this regard is largely
confirmed by other sources.
That which constituted the essence of the idea of a bank (argen-
taria) for the ancients and particularly for legal experts (jurisconsults)
had two aspects. On one hand, this is what constitutes in our eyes as
well the specific function of a deposit bank—the double service of
receiving deposits and extending credit. The banker, acting as such,
did not loan out his own money, but rather a part of the money that
had been deposited with him by his clients (Plaut., Curc. 71–9). On
another hand, this was the connection between the banker and the
client. Clients deposited money, and they could leave it on deposit or
withdraw it, or ask the banker to make payments or send money. This
connection is thus exhibited in a series of operations, and by the
records of transactions that were kept. The result of such operations
involved deposit accounts held by clients, and such accounts were
called rationes.
Transfers could be made between clients of the same bank, as is
attested by two passages, one in Plautus and the other in Terence
(Plaut., Asin. 436–40 and Ter., Phorm. 921–3). There was no system
of institutionalized compensation between the banks of the same city,
but that did not prevent transfers from taking place, as we may see in
a number of papyri. To facilitate such transfers, the banker sometimes
held an account in the bank of one or several of his colleagues
(Bogaert 1983a: 218–21; 1983b: 34). Between banks in different cities,
such an arrangement, although not impossible, was very probably
rarer, and must have depended on the personal relations of the
individual bankers concerned (on such categories of financial oper-
ations, see Harris 2006, 2008; Hollander 2007). Professional bankers,
unlike elite financiers, were local businessmen, who specialized in
operations on the spot.
A cheque is a written note by which a client of a bank orders it to
pay a precise sum of money to the beneficiary of the cheque. But the
cheque is given to the beneficiary, whereas other payment orders are
given to the banker. In antiquity, there is no trace of cheques in the

western Mediterranean. However, in the eastern Mediterranean, the
use of cheques is attested in at least two regions: in Canaan, following
rulings made by Mosaic law on the payment of wages (Bogaert 1968:
340), and in Egypt, where cheques have been found from the end of
the Hellenistic period. Such cheques continued to be used in Egypt
after the Roman conquest (Bogaert 1983a, 1987: 73–7). Yet, the
existence of these non-transmissible cheques was not as important
historically as that of the transmisible cheques which came into use
later, in modern Europe.
At the end of the Republic and during the classical period of
Roman law, jurisconsults had a point of view regarding deposit
banks (argentaria) that was at one and the same time very rigid and
very flexible. It was rigid with regard to the definition at its center:
they only thought of these banks in terms of deposit accounts, and
nothing that was external to accounts had anything, strictly speaking,
to do with banks. It was very flexible as concerned the details of the
operations that the bank account permitted: a bank account is noth-
ing but this series of operations. For the jurisconsult and for the client,
what the banker was supposed to do was to lend out part of the
money deposited. But it is also assumed that the banker will make
payments in the name of the depositor, and allow him to make
withdrawals at will (D. 2.13.9.2, Paul. 3 ad ed.).
Before examining more closely the orientation of Roman law in
relation to banking and financial activity, I would like to make a few
remarks about the norms applied to non-Romans, that is, to Latins
and to foreigners (who were called socii, or “Allies” in Italy during the
republican period, and peregrini in the provinces).
15.3. THE NON-ROMANS AND FINANCIAL LIFE
Roman citizens were subject to the norms of Roman civil law. But in
territories controlled by Rome, there were people who were not
Roman citizens. There were non-citizens in Italy prior to the end of
the Social War (91–89 ), and there were also non-citizens in the
provinces, at least up to the time of the famous Constitution of
Caracalla or the Antonine Constitution (Constitutio Antoniniana)
of  212. What consequences did such a situation have on banking
92 Jean Andreau
and credit? An episode from the history of the Roman Republic
allows us to understand a little better the elements of the problem.
It concerns the “crisis” of 193 , concerning which we have only a
short passage from Livy (Liv. 35.7). Some have called this a banking
crisis, but it would be more precise to speak of a financial crisis, since
nothing indicates that the professional bankers had anything to do
with it. That year, many Romans were deep in debt, even though the
greed of financiers (Livy relates) had been limited by a number of laws
governing interest-bearing loans—laws whose substance we do not
know. How was this possible? We can see that some Roman citizens,
in order to avoid the effects of that legislation, transferred the debts
they owned to Italian Allies, who were not subject to the same norms
because they were not Roman citizens. Livy says the transfer was
fraudulent, because the money came from Romans who transferred it
to Allies and then allowed the Allies to lend it to Romans (Barlow
1978: 78–80). How did the Roman creditors manage this? We just do
not know.
The first attempt to straighten this out involved requiring Allies to
declare all the loans they had made to Roman citizens, going back to
the most recent Feralia. Then the extent of the practice was dis-
covered, Livy says. So a second measure was tried: a plebiscite pro-
posed by the tribune M. Sempronius Tuditanus required the Latins
and Allies to conform to the same norms as Romans with regard to
loaning money (pecuniae creditae ius). We do not know the amount
of the debts declared, and this is a pity, because this is the only time
we see the Roman state inquire as to the amounts of debts contracted
by Romans, or in any case a part of these debts. The episode shows the
consequences to which a multiplicity of statuses of persons before the
public law can give rise, and how the state managed to palliate these
consequences by making all persons subject to the same legal norms.
Roman jurisprudence (as for example in the manual of Gaius)
distinguishes between ius civile, only applicable to Roman citizens,
and ius gentium, which was applicable to Roman citizens as well as to
other free inhabitants of the Empire. But Livy, in this passage, does
not speak of ius gentium. As Schulz has emphasized, when legal
advisors or legislators wanted to apply a norm already in force for
Romans to a non-Roman, they did not seek to create a system that
would apply equally to both and that would be a part of the ius
gentium. They simply applied the civil law to the non-Roman (Schulz
1946: 73, 137, and 163). This is why the norms pertaining to the civil

law are very often present in contexts that involve foreigners. The ius
gentium only had practical significance in international law.
After the beginning of the second century , the number of
citizens slowly grew. After the Social War (91–89 ), all the Italians
of the Peninsula became Roman citizens, and later the residents of
northern Italy were granted the same status.
But at the beginning of Augustus’ reign, the Italians residing in
provinces and provincial citizens who had acquired Roman citizen-
ship represented only a tiny minority of the population of those
provinces. In the second century , the situation had obviously
changed. At the beginning of the third century , not long before
the Constitution of Caracalla, how many foreigners were left? For
some historians a very large majority of the free inhabitants of the
Empire had already received this citizenship. For others, Roman
citizens continued to be a minority in the eastern part of the Empire,
and in North Africa, communities of foreigners were still numerous
(Jacques-Scheid 1990: 281–7). Otherwise, following the establishment
of the Constitution, legal norms of Roman law were not applied
everywhere, nor were they applied to all populations.
It is important to know in which cases judgment remained in the
hands of Roman magistrates (meaning by this primarily the gover-
nors of provinces), and which cases were handled by other magis-
trates (usually municipal magistrates). But it is just as important to
know the norms to which foreigners were subject. The first question
is a little less difficult than the second, but both are delicate because
the documentation for them is extremely limited.
Geographically, the evidence for financial and banking-related
matters is not divided up in the same way as the documentation for
law and the legal system. As regards the late Republic, the cases of
Sicily and of Cilicia are the least obscure, simply because of the
speeches of Cicero against Verres and of his letters. In imperial
times, we have from Egypt (via the papyri) a copious enough amount
of documentation for legal matters and for the professional part of
financial life. Rather than making general remarks, I prefer to base
myself on two examples, about which we are somewhat less ignorant:
Sicily at the time of Cicero, and Egypt during the Principate.
In the Sicily of the time of Gaius Verres, a single category of litigation
was judged by Sicilian judges in accordance with Sicilian norms: litiga-
tion opposing two peregrini from the same city (Maganzani 2007).
Conversely, if the two parties engaged in litigation were Roman
94 Jean Andreau
citizens, the procedure followed and the norms applied would obvi-
ously be based on Roman law. When the case involved two peregrini
from different cities, it was necessary to appear at a session of the
conventus, where the governor of the province would be acting in a
judicial capacity.
Banking and money-lending in Roman Egypt have been very well
studied by Bogaert (1994) and Lerouxel (2008, 2012, 2016). As
regards the norms applied to non-Romans in Egypt, one has to
refer to the research of Mélèze-Modrzejewski (1970, 1990) and to a
recent article by Yiftach-Firanko (2009). For instance, Mélèze studied
the mandatum cases described in Egyptian papyri, some of which had
a financial aspect. They were gratuitous mandates of the kind known
under Roman law, but had to do with Greco-Egyptian customs, and
not with Roman law (Mélèze-Modrzejewski 1990: II, esp. 473).
In the second century , we find allusions to a “Law of the
Egyptians” that was the result of a fusion between Greek elements
and properly Egyptian ones. This was a sort of manual that aimed at
explaining local legal customs to Roman judges, and the norms con-
tained in it applied to all the peregrini of Egypt, whatever their origin.
Conclusion: in a number of cases, when the opposing parties were
not Roman, the applied norms were more or less divergent with
regard to Roman law, and we know far less about them than we
know about Roman law. But we should not exaggerate the size of the
unknown area of the law for foreigners, because the influence of
Roman law was probably always very strong.
15.4. LAW AND J URISPRUDENCE

AND THE WORLD OF FINANCE
Even if Roman law seems systematic and specializes in definitions

and classifications, its role can be undeniably characterized as flexible.
The fact that, in the Republican period, corporate and banking law
came mainly under the ius honorarium, that is under the praetor’s
edict, and was not mainly a matter of statute, is a symptom and an
expression of this flexibility. For, in his edict, the praetor, at least in
theory, might take new social and economic situations into account,
whereas it would have been more difficult to change statutory law so
frequently.

Under the Principate, corporate law was not mainly shaped by
imperial constitutions (which were equivalent to Republican leges
publicae). As the praetor’s edict was in some way fossilized, in
practice, and then officially (edictum perpetuum), it did not ensure
the same flexibility. But this evolution of the praetor’s edict was
balanced by the leading role played by jurisconsults, to whom, from
Augustus’ reign onwards, the emperor granted the ius respondendi. In
the second and third centuries , the major part of the available texts
about banking and financial life were written by jurisconsults. It
needs to be stressed how precise and concrete these fragments,
included in the Digest, were. But there were a number of jurisconsults,
and they did not always agree on any particular question, so some of
these questions were subjects of debate. It seems that the jurists’
opinion bound the judges only when a consensus had been reached
(Magdelain 1990: 103–52).
Compared to the literary evidence for financial life and banking
from the Ciceronian period, the evidence from the Principate is poor.
We know a fair number of funerary inscriptions concerning bankers
or money-changers, but the literary texts concerning banking and
financial life are rather few during the Principate. For this reason, the
historian who is not a specialist in law may gain the impression that
there was much less innovation in the Principate. But this certainly is
not correct as concerns law and juridical thought.
In financial matters, a first token of the flexibility of Roman law is
that jurisconsults paid a great deal of attention to the concrete
application of technical procedures. The authors of the fragments
collected in the Digest, as well as Gaius in his Institutes, were much
concerned with the material conditions of the various occupations
and activities, and they based their legal opinions thereon.
The jurists often expressed their opinions on one occupation, or
one form of activity, after another. So, in writing about the actio
institoria, Ulpian, not hesitating to go into detail, enumerates an
impressive list of functions that a slave might perform: there are
managers of buildings and monuments, bankers, money-lenders,
weavers and vilici, and also baker’s apprentices, spice vendors, cloth-
ing merchants, travelling salesmen, and wholesalers in oil or grain
(D. 14.3.5, Ulp. 28 ad ed.).
The definition of bank-account which Labeo produced and which
is handed down by Ulpian is very descriptive: it is a list of counter
transactions rather than a general statement about the whole banking
96 Jean Andreau
practice: “Labeo says that the bank-account is composed of mutual
transactions which consist in paying out, collecting payments, lend-
ing, obliging, and paying off one’s debts” (D. 2.13.6.3, Ulp. 4 ad ed.).
The jurists attached much significance to the bank-account, which
they called either ratio (“an account”), or ratio accepti et expensi (“an
account of deposits and payments”), or ratio implicita propter accepta
et data, “a complex account including both deposits and payments”
(D. 2.14.47.1, Scaev. 1 dig.). The fact that this notion of bank-account
could be called in different ways suggests that it arose out of profes-
sional practice and was not a creation of the jurists themselves.
The jurists were concerned to keep in touch with the development
of business transactions and professional life, in order to handle
the juridical problems caused by financial activity as efficiently as
possible.
But at the same time, as Thomas remarked, the art of the law
provided for human activities “a means for their formal elaboration”
(Thomas 2004: 212). Thus we are concerned first of all with concrete
and precise observations on the conditions of an activity; and then
based on these observations, we have the beginnings of conceptual-
ization of a kind we could characterize as economic. In the financial
field, this beginning of conceptualization led the jurists to insist on
the specificity of banking. One might have thought that all financial
activity would be handled according to the modalities of private
law applicable to every citizen; but, on the contrary, one category
was isolated, that of people who had the right to open deposit
accounts and who did business on the basis of such deposit accounts
(D. 2.14.47.1, Scaev. 1 dig.). Only the deposit banker had the right to
open an account for each of his clients. His transactions were ordered
around the notion of the bank-account, which created a relationship,
considered as long-lasting, between his clients and himself.
Besides this empirical description and its formal elaboration, the
jurists found another way of handling such problems. It consisted in
appealing to juridical instruments that already existed. For instance,
they resorted several times to stipulation. This formal contract was
very rigid, for it was made up of a question and an answer which had
to be completely identical. But, despite that rigidity, it was very
flexible at the same time, since the stipulation could be applied to
nearly every requirement (Johnston 1999: 77–8). So, in connection
with a loan contract that was in principle interest-free, such as a
mutuum, a stipulation could be added to allow the payment of

interest (Johnston 1999: 84–5). The stipulations was also used to
manage the credit that the bankers granted the buyers in the context
of auctions (stipulatio argentaria).
Another route taken by ius honorarium and jurisprudence was the
creation of new instruments, for instance new actions. Conspicuous
here are the five actiones adiecticiae qualitatis and the actio tributoria,
which were instituted by the praetor in the late Republican period
(see for instance Aubert 1994: 40–116). Their remit consisted in
protecting third parties who had professional relations with slaves
or with dependent sons (filiifamilias). Through these actions, third
parties were allowed to take legal action against the slave’s owner or
the son’s father. For our present purposes the most significant is the
actio institoria. It concerned a slave or a filiusfamilias who ran a non-
maritime business, that is a shop, a workshop, a storehouse, or a bank,
on behalf of his owner or his father. The actio institoria was instru-
mental in enabling slave owners to take advantage of their slaves’
commercial and financial activity. These actiones adiecticiae qualitatis
were one of the most striking and original features of the Roman slave
system. Of course, and without a doubt, this system had all the defects
and flaws of slavery and slave systems, but at least it allowed a modest
number of slaves and freedmen to become technicians and small or
medium-sized managers.
Finally, with regard to other aspects of banking and credit, debates
of long duration seem to have taken place among the jurists. The best
example of such debates concerns the so-called “irregular deposit,”
which nevertheless is at the center of banking activity. In the juristic
texts on this contract one encounters no consensus, and possible
juristic interpolations of the Justinianic period does not provide
sufficient explanation.
In Roman law, under the ordinary contract of deposit, the deposi-
tee was not entitled to use the deposited object (Andreau 1987a:
529–44). Even if this object was an amount of money, it had to be
kept as it was, until it was given back to the depositor, and it could not
bear interest. But, of course, such a deposit, which was retained
without being touched by the depositee, is not the defining charac-
teristic of banking activity. The banker generally wished to be entitled
to make use of the money and to return not the precise coins
deposited, but the equivalent value. In practice, the argentarii, the
coactores argentarii, and, later on, the nummularii received deposits
of money that they were entitled to use. But, at the end of the
98 Jean Andreau
Republican period and under the Principate, did the jurists acknow-
ledge such a kind of deposits (which, later on, was called an “irregular
deposit”)? Or did they consider such transactions as loans received by
the banker from the customer? Three elements were singled out for
attention by the jurists: did the depositee use the money he had
received from the client? Did he lend it at interest? And did he pay
the depositors interest? The most probable hypothesis is that, in the
second and third centuries , the interest-bearing deposits, i.e. the
deposits for investment, were considered as loans, whereas the others,
the deposits which did not provide interest to the depositors, were
considered as true deposits.
These observations point up some of the ways in which Roman law
succeeded in coping with the problems caused by financial life and
credit: it paid much attention to the practical details of financial life; it
called up juridical instruments that already existed; it created new
tools, and particularly new actions; and, moreover, long debates took
place among the jurisconsults, especially about the depositum which,
later on, after the classical period, was called the “irregular deposit.”
15.5. PRODUCTIVE LOANS AND

COMMERCIAL CREDIT
In Greece and in Rome, did productive loans exist, that is loans which
played a part in production and trade? If such productive loans
existed, to what extent did professional bankers and businessmen
who used to lend money play a role in such loans? And what can
be said of the role of law and jurist in connection with them?
On the first of these three questions, the evidence is very scarce, but
the answer is positive, beyond all reasonable doubt. Seneca, for
instance, wrote in a letter to Lucilius (Sen., Ad Lucil. 119.1): “yet,
you will need a creditor; if you wish to be in business, you have to
contract debts” (opus erit tamen tibi creditore; ut negotiari possis, aes
alienum facias oportet). Let us note that, four centuries before,
Demosthenes had written a very similar sentence on the role of
loans in trade (Dem. 34.51). Plutarch, when he composed a treatise
condemning debts, treated in the same manner those who borrowed

to buy estates or productive slaves and those who borrowed because
they had expensive tastes or for purposes of evergetism (Plut., Mor.
830 E). In another passage, he speaks of a loan contracted in order to
buy grain, vineyards, or olive groves (Plut., Mor. 523 F). Moreover,
the scraps of information provided by jurists are consistent with those
derived from literary texts: several fragments of the Digest, written by
the jurists Africanus and Ulpian, deal with cases in which the loan is
contracted in order to buy an estate, repair a ship, feed sailors, or buy
goods (D. 12.1.4 pr., Ulp. 34 ad Sab.; D. 14.1.1.8–11, Ulp. 28 ad ed.;
D. 14.1.7, Afric. 8 quaest.).
Such texts show that the notion of productive loan was absent in
antiquity. Greeks and Romans were not accustomed to distinguish
between “productive” loans and loans for consumption. However, in
practice, this conceptual gap did not prevent such productive loans
from being contracted. They were interest-bearing loans. From the
juridical point of view, they were treated in the same way as the other
kinds of loans, except the maritime loans. What proportion of all
loans were productive? We do not know. As regards Athens in the
fifth and the fourth century , both Millett (1991: 188) and Cohen
(1992: 151), who very often disagree, are as one in holding that
productive loans constituted a rather substantial minority. For the
Roman world of the end of the republican period and of the Princi-
pate, such a guess would also be appropriate.
To what extent did professional bankers and the rest of the finan-
ciers play a role in these loans the object of which was to invest in
economic activities? This second question leads us inevitably to a
consideration of the three Sulpicii. The Sulpicii were three first-
century  businessmen, of whom we know through the tablets
found in the area called Murecine (or Moragine), in the territory of
Pompeii. The tablets concern transactions concluded in Puteoli under
the Julio-Claudians. They have been published and studied by Camo-
deca, who has written two outstanding books and many papers about
them (Camodeca 1992, 1999, 2000, and 2003).
What was the business of the three Caii Sulpicii, Faustus, Cinnamus,
and Onirus? Two different hypotheses have been proposed: they were
either professional bankers, or money-lenders without being bankers.
Relying mainly on the tablet TPSulp. 82 (mainly, but not exclusively),
Camodeca concluded that they were professional bankers (argentarii,
cf. Jones 2006). On the contrary, Verboven (2003a, 2003b) disagrees,
100 Jean Andreau

holding that they were not bankers, but faeneratores, that is, special-
ized financiers and money-lenders. I share Verboven’s opinion
(Andreau 1999: 71–9). At the moment none of the two assumptions
may be considered as surely true.
Every category of businessmen could be brought to lend money for
a productive purpose, and so to have an influence on economic life,
but not all of them in the same manner, nor to the same extent.
Howgego (1992) has listed all the categories of persons and institu-
tions that could lend money in Roman antiquity. In brief, let us
say that short-term commercial loans were made by several categor-
ies of people, and particularly by professional bankers and wholesale
merchants, who lent money at the same time as they sold goods.
As for long-term credit, which was certainly less frequent, it was
mainly practiced by financiers belonging to the elite and by big
businessmen.
I have already said that, in auctions, professional bankers of the
Latin world used to advance short-term loans to the buyers. These
auctions often took place in places of commerce such as ports,
markets for retailing or for wholesale trade. For intance, in Rome,
there were auctions in the Forum Boarium and the Macellum Mag-
num. Outside Rome, receivers and deposit-bankers are known in
towns where fairs took place, such as Cremona, or in towns with
periodic markets (nundinae). These nundinae constituted regional
networks. I have shown elsewhere that in Pompeii and in Puteoli the
auctions took place on the day of the nundinae (Andreau 1974, 1999:
149–50, 2000; De Ligt-De Neeve 1988; De Ligt 1993).
Besides serving as country markets for peasants and small land-
owners of the surrounding area, the nundinae had two other func-
tions. They were one of the places where patrimonial transactions
could take place, where the landowners, even the important ones,
could buy and sell land, buildings, cattle, slaves, and where they could
sell some of the products of their estates. At the same time, merchants
were present at the nundinae, and they played a significant role in the
marketing of agricultural products, which were conveyed to the coast
and then to other ports and cities, for instance to Rome. When the
landowners and the farmers sold agricultural products by auction,
the buyers included merchants and tradesmen. A letter of Pliny
provides an example of such a situation: Pliny sells his grape-harvest
by auction, and the buyers are negotiatores (Cat., De agr. 146–8; Plin.,
Epist. 8.2).

In other cases, argentarii and coactores argentarii were involved in
auctions which were more to do with craftwork or mining; however,
in such cases too the buyers might be merchants. For instance, in
Southern Portugal, at Aljustrel, a Roman mining site, the bronze
tables of Vipasca show that mining pits could be sold by auction,
besides slaves, mules, donkeys, horses, i.e. men and animals playing a
role in mining, and that argentarii customarily took part in such
auctions (CIL II.5181; Domergue 1983).
Whatever views one holds of the activities of the Sulpicii, the
professional bankers provided a short-term commercial credit, and
they were not the only ones to do it. For instance, in a fragment
written by Paul, a slave entrusted by his master with money-lending
at interest was active in commercial credit, on his own initiative,
paying barley traders on behalf of buyers (D. 14.5.8, Paul. 1 decret.).
I am convinced that neither Cicero’s Cluvius nor the Sulpicii were
professional bankers; but, if they were bankers, these Murecine tablets
would show that, in a port such as Puteoli, the professional bankers,
besides short-term credit, also made long-term loans, and that, in that
way, they created more currency than I previously believed to be
the case (though, in fact, the volume of their transactions was not
sufficient to create a lot of currency). If the Sulpicii were bankers,
the Murecine tablets would show professional bankers of greater
significance than any others known from the first and the second
centuries . But, in any case, the outline of professional law which
I will present in what follows would be the same. And moreover, in
the available evidence—including the Murecine tablets—no argentar-
ius concludes a maritime loan. They may play a role as intermediaries
in maritime loans, but do not seem to have concluded such loans
themselves. In the Murecine archive, two tablets, TPSulp. 31 and
TPSulp. 78, have first been considered as regarding maritime loans;
but Camodeca has shown that it is not the case (Camodeca 1999:
97–9 and 177–80). Admittedly, the available evidence is scanty. But in
my view maritime loans were mainly transacted by big businessmen,
faeneratores or merchants who were financiers at the same time.
Among the known examples of production loans (that is, of loans
contributing to production, transportation, or distribution of goods),
medium-term and long-term loans are very rare. This may seem
surprising, but such surprise is probably misplaced, for the following
three reasons. First, merchants and manufacturers could have relied
on short-term loans that were repetitively refinanced. Then, even in
102 Jean Andreau

eighteenth-century England, such long-term loans were not as fre-
quent or as massive as it was previously thought. Verley (1985: 48–51)
has offered a compelling picture of the ways in which capital was
raised for commercial ventures in pre-industrial economies. Com-
mercial ventures often required only modest levels of start-up capital.
But when they were successful, they could grow rapidly through the
massive reinvestment of profits. Thus pre-industrial commercial ven-
tures tended to rely on internal over external funding, and a low
demand for fixed as opposed to liquid capital. Such conclusions are a
fortiori valid as regards Roman antiquity. Self-financing was also
present in antiquity.
In addition, Rome had institutions that offered financing choices
other than long-term loans. The first of the institutions is what we
would call a limited partnership: one of the associates (socii) of this
societas provided the capital to one or a few associates who were in
charge of the actual work and asset management. This was well
suited to the needs of members of the elite interested in increasing
their patrimony: they had the opportunity to extract profits from
commercial, industrial, or even financial enterprises without having
to live the life of an entrepreneur. Crassus participated in such
societates, as in all probability did Vespasian (Cic., Par. Stoic. 6.46;
Suet., Vesp. 4.6). Some municipal notables were involved in such
limited partnerships as well. Such was the case with P. Alfenus Pollio,
identified as such in Jucundus’ tablet no. 45, found at Pompeii
(CIL IV, Suppl. 1.3340.45).
The second institution is the peculium. To get his slave to work
with third parties, the master had available two juridical procedures:
the preposition (praepositio) and the peculium (Andreau 2004). It is
important to make a clear distinction between them. In the prepos-
ition, the master maintains his role as the entrepreneur and the slave
is a “manager,” to use a term proposed by Aubert. This is not the case
with the peculium. Aubert (1994: 4) wrote: “even though a slave with
peculium was legally dependent, his economic activities were in
practice kept separate from his master’s. . . . A slave with peculium
was not acting as business manager on behalf of a principal.” He was
an entrepreneur.
To summarize: Roman banking and finance did not function, first
and foremost, to facilitate economic investment. Nevertheless, their
economic role should not be entirely dismissed.

15.6. STANDARDS IMPOSED ON
PROFESSIONAL BANKERS
Did commercial law exist in the Roman world as a distinct branch of

law? Did Roman jurisprudence deal with commercial transactions in
a specific way, in accordance with a set of standard rules and proced-
ures? The matter has long been considered problematic, as is indi-
cated by the books of Goldschmidt (1891) and Fadda (1903,
published again in 1987). The conclusion that has been generally
accepted since the beginning of the twentieth century is that a specific
and distinct commercial law did not exist in Rome, but only appeared
with the ius mercatorum, in the Middle Ages (twelfth and thirteenth
centuries).
But, since the eighties, contrary opinions have surfaced in Italy,
within the framework of the controversy between “primitivists” and
“modernists.” In fact, the absence of commercial law as such has been
perceived by some historians as an indication that economic life and
thought were at a lower level of development in antiquity than in
early modern history and even in the late Middle Ages. Serrao, leader
of a “modernizing” tendency in legal history, rebelled against such a
vision, and his disciples, Cerami, Petrucci, and Di Porto, have
returned to this problem of the existence of a Roman commercial
law. Readers may refer to a book published by Cerami and Petrucci
(2010; cf. Petrucci 1991, 2002).
Of course, how we define commercial law is crucial. This matter,
and the general problem of commercial law, will not be treated here.
Our present concern is to provide a rough outline of the professional
law that Roman jurisprudence created around banking. I say profes-
sional law rather than commercial law, because it only concerned the
members of the banking professions, and not all those involved in
finance and banking.
The constituent parts of this professional law are fourfold: produc-
tion of rationes; solidarity in a society of bankers; the receptum
argentarii; compensation.
Did these four sets of regulations give preferential treatment to the
bankers, as compared with other categories of financiers, or were they
disadvantageous to them? And did they have a positive influence on
economic life? In my book about the bankers (Andreau 1987a), I have
stressed that some of these rules might have damaged their interests,
104 Jean Andreau

even if this was not the intention of the authorities and jurists. I am
rather less convinced of this than I was twenty-five years ago. As the
authorities gave a lot of attention to securing the good working order
of monetary circulation and of juridical contracts, it was not in their
interest that the banks should go bankrupt! On the other hand, the
financial situation of the bankers’ enterprises was certainly precar-
ious, if one considers their size and how risky their activity was. As
regards the Greek banks, Bogaert (1968) has very much stressed this
fragility. If the bankers had been faced with the authorities’ hostility
or with disadvantageous regulations, the fact that they carried on with
their activity for several centuries would not be plausible. So, in order
to reassess this conclusion, we need to look again at the relevant
regulations.
First of all is the “production of the bank-accounts, of the register”
(editio rationum). All the operations effected on an account were
entered in a register kept by the banker, the rationes, which consti-
tuted the tangible reality of his clients’ accounts. We do not know
precisely how such a register of accounts was organized (Andreau
1987a: 615–26). Anyway, it had to be produced by the banker when-
ever a client was involved in a lawsuit, even if the banker was not
personally concerned. The banker was required to produce (edere)
everything relating to the particular account, for, in some manner,
these records were considered to be the property of the client. This
standard did not make the bankers’ activity easier, and it could have
been an indirect way to control their professional behavior. However, it
simplified the financial practice, for it allowed the relationship between
bankers and clients to be exclusively based on accounts kept by the
banker without adding other formalities to each operation.
Second norm: compensation. If the client Titius had deposited 30
and owed 40 to the banker, and if Titius took the banker to court, he
was not expected to calculate the compensation between all the
operations of his account. The banker, however, did have to calculate
this compensation, taking into consideration all the debts and
deposits of the client. He could claim only the outstanding difference,
10, and moreover had to be careful to make no errors in his calcula-
tion of that difference. This rule may be regarded as harmful to the
bankers; but it was favorable to the clients; a rule sensibly favorable to
clients, if known by clients, would reduce transaction costs and help
generate business for bankers. Moreover, in some way, like the
previous rule, it may have been intended to minimize the distance

between the banker and his client. Banking is a commercial business,
and the banker runs the financial risk of his activity. But at the same
time he is a credit intermediary, and these regulations made him a
kind of direct financial assistant of his client.
A third element of the professional law relates to the partnership of
bankers. Generally speaking, in Roman law, the private partnerships
of merchants or craftsmen did not have legal personality. They did
not possess a patrimony distinct from the patrimonies of their socii.
From that point of view, the bankers’ partnerships (argentarii socii)
were no different from other private partnerships. Nevertheless, as
regards their partners’ joint liabiblity, they did have a particular legal
status. If one of the partners was creditor of a third party, the sum
could be claimed and perceived by another partner (which is called
active solidarity). In the same way, if one of the socii bankers con-
tracted a debt, the creditor could claim for the entire sum from his
partner or partners (which is called passive solidarity). These two
solidarities were not applied to other commercial or manufacturing
partnerships (Arangio Ruiz 1965: 78–83 and 144–5; Andreau 1987a:
626–31). In the bankers’ case, they seem to have become more sub-
stantial with the passing of time. In fact, the Rhetorica ad Herennium,
at the beginning of the first century , considers it as a simple custom
which was not a matter of legal standards, whereas, for the jurist Paul,
at the beginning of the third century , it was a well-established
juridical rule (Rhet. Her. 2.13.19; D. 2.14.25 pr., Paul. 3 ad ed.).
This rule concerning solidarity (in case of a societas of bankers)
gave more authority to the banking profession (compared to other
businesses) and to the bank-accounts. Like the norm concerning
editio rationum, it shows that banking was considered a prestigious
activity, even if the bankers themselves were socially less important
than some money-lenders belonging to the senate or the equestrian
order. The specificity of banking was understood by the ancients, as
the jurists’ texts stress. Thus, some of the texts speak of the publica
causa and of the utilitas publica of the banks, or of the fides publica of
the bankers. These complex notions deserve fuller discussion than is
possible here. They show that the Romans regarded banking specif-
ically as having some standing, even dignity—in connection with the
notion of bank-account (ratio)—and this had consequences for the
rights and duties of bankers.
The last element, but not the least: the receptum argentarii. The
receptum was reserved for argentarii (until the second century ,
106 Jean Andreau

when it was extended to the nummularii). It was the undertaking that
the banker gave to a third party, when he promised to pay him the
money that his client owed to a third party (Lenel 1881: 62–71;
Andreau 1987a: 597–602). The receptum involved three people (the
banker, his client, and the third party), but it was binding upon only
two of them, the banker and the third party. Legally, according to the
receptum, the client did not need either to express his agreement or to
be present. Clearly, however, in one form or another, an agreement
also existed between the client and his banker on the payment of the
debt in question. A praetorian action was instituted as regards
the receptum, at the end of the second century or at the beginning
of the first century , and it was used down to the third century 
(after which it lapsed).
The receptum is probably the most important among these profes-
sional regulations, but also the least known, because it had disap-
peared in the post-classical period: nearly nothing is said of it in
Justinian’s Corpus iuris. It simplified the client’s role and the relations
between him and the banker: the client only needed to worry about
paying the debt to the banker and not about the arrangements
that the banker had to make with the third party. Moreover, it is
one of the indirect ways in which Roman law created forms of agency
relationships.
G. Dari Mattiacci draws my attention to the fact that the four
standards I have just explained might reduce transaction costs
among bankers, clients, and third parties, and I think it is a very
useful idea, from the point of view of a neo-institutionalist analysis
(on such analysis in Roman economic history, see for instance
Lerouxel 2006, 2012, 2015, 2016; Terpstra 2008; Kehoe, Ratzan, and
Yiftach 2015). In fact, even if all of them were not favorable to the
bankers themselves, they could protect their clients’ interests, that is,
those of artisans, wholesale merchants, and retailers, and play a
stimulating influence on commercial and financial life. The bankers’
professions were linked with commerce and manufacture. With their
presence in the fora, in markets, ports, and auctions, they were part of
the fabric of commercial and manufacturing life. This is particularly
relevant to Peninsular Italy, where they played a role in the food-
supply of the city of Rome, through the ports of Puteoli, Ostia,
and Portus.
Anyway, in the first century  and in the first century , there is
no indication that banks suffered particularly because of these rules.

But the situation was transformed in the third century , because of
monetary, military, and political developments, particularly during
the last third of the third century, when prices soared. In that period,
the banking professions virtually disappear from sight, for several
generations. For financial life and the banking professions, it is a
period of major crisis. The institutional role of bankers in auctions
definitively disappears and will never reappear in history, in its earlier
form. The professions of argentarii and coactores argentarii, which
were directly linked to the provision of credit in auctions, also
disappeared completely for almost a century; from the first half of
the fourth century , the word argentarii designates silversmiths. It
is only in the last part of the fourth century that money-changers and
bankers are attested again (Andreau 1997: 133–55).¹
REFERENCES
Andreau, Jean. 1968. “Banque grecque et banque romaine dans le théâtre de

Plaute et de Térence.” 80 Mélanges de l’École Française de Rome 461–526.
Andreau, Jean. 1974. Les Affaires de Monsieur Jucundus. Rome: Éditions de
l’École Française de Rome.
Andreau, Jean. 1987a. Vie financière dans le monde romain, Les métiers de
manieurs d’argent (IVe siècle av. J.-C.-IIIe siècle ap. J.-C.). Rome: Éditions
de l’École Française de Rome.
Andreau, Jean. 1987b. “L’Espace de la vie financière à Rome,” in L’Urbs,
Espace urbain et histoire (Ier siècle av. J.-C.-IIIe siècle ap. J.-C.). Rome:
Éditions de l’École Française de Rome, 157–74.
Andreau, Jean. 1997. Patrimoines, échanges et prêts d’argent: l’économie
romaine. Rome: “L’Erma” di Bretschneider.
Andreau, Jean. 1999. Banking and Business in the Roman World. Cambridge:
Andreau, Jean. 2000. “Les marchés hebdomadaires de Latium et de Campa-
nie au 1er siècle ap. J.-C.,” in Elio Lo Cascio, ed., Mercati permanenti e
mercati periodici nel mondo romano. Bari: Edipuglia, 69–91.
Andreau, Jean. 2004. “Les Esclaves ‘hommes d’affaires’ et la gestion des
ateliers et commerces,” in Andreau, France, and Pittia, eds, 111–26.
Andreau, Jean, Jérôme France, and Sylvie Pittia, eds. 2004. Mentalités et
choix économiques des Romains. Bordeaux-Pessac: Ausonius.
¹ I am grateful to Giuseppe Dari-Mattiacci and Robert Ellickson for their useful

comments on a first version of this chapter.
108 Jean Andreau

Arangio Ruiz, Vincenzo. 1965. La Società in diritto romano. Naples: Eugenio
Jovene.
Aubert, Jean-Jacques. 1994. Business Managers in Ancient Rome: A Social
and Economic Study of Institores, 200 – 250. Leiden/NewYork/
Cologne: Brill.
Barlow, Charles T. 1978. Bankers, Moneylenders and Interest Rates in the
Roman Republic. Ann Arbor/London: University Microfilms International.
Bogaert, Raymond. 1968. Banques et banquiers dans les cités grecques.
Leiden: Sijthoff.
Bogaert, Raymond. 1983a. “Note sur l’emploi du chèque en Egypte ptolé-
maïque.” 58 Chronique d’Egypte 212–21.
Bogaert, Raymond. 1983b. “Les kollybistikai trapezai dans l’Egypte gréco-
romaine.” 3 Anagennesis 21–64.
Bogaert, Raymond. 1987. “Recherches sur la banque en Egypte gréco-
romaine,” in Tony Hackens and Patrick Marchetti, eds, Histoire écono-
mique de l’Antiquité. Louvain-la-Neuve: Séminaire de Numismatique
Marcel Hoc, 49–77.
Bogaert, Raymond. 1994. Trapezitica Aegyptiaca: recueil de recherches sur la
banque en Egypte gréco-romaine. Florence: Gonnelli.
Camodeca, Giuseppe. 1992. L’Archivio puteolano dei Sulpicii, vol. I. Naples:
Eugenio Jovene.
Camodeca, Giuseppe. 1999. Tabulae Pompeianae Sulpiciorum: edizione cri-
tica dell’Archivio puteolano dei Sulpicii. Rome: Quasar.
Camodeca, Giuseppe. 2000. “Per un primo aggiornamento all’edizione
dell’archivio dei Sulpicii (TPSulp.).” 11 Cahiers du Centre Gustave Glotz
173–91.
Camodeca, Giuseppe. 2003. “Il Credito negli archivi campani: il caso di
Puteoli e di Herculaneum,” in Elio Lo Cascio, ed., Credito e moneta nel
mondo romano (Atti degli Incontri Capresi di Storia dell’Economia antica
(Capri, 12–14 ottobre 2000). Bari: Edipuglia, 69–98.
Cerami, Pietro, and Petrucci, Aldo. 2010. Lezioni di diritto commerciale
romano, 3rd edn. Turin: Giappichelli.
Cohen, Edward E. 1992. Athenian Economy and Society: A Banking Perspec-
tive. Princeton: Princeton University Press.
D’Arms, John H. 1981. Commerce and Social Standing in Ancient Rome.
de Ligt, Luuk, and Piter-Willem W. de Neeve. 1988. “Ancient periodic
markets: festivals and fairs.” 66 Athenaeum 391–416.
de Ligt, Luuk. 1993. Fairs and Markets in the Roman Empire. Amsterdam:
J. C. Gieben.
Domergue, Claude. 1983. La Mine antique d’Aljustrel (Portugal) et les tables de
bronze de Vipasca. Bordeaux-Talence: Publications du Centre Pierre Paris.

Fadda, Carlo. 1903 [repr. 1987]. Istituti commerciali del diritto romano.
Naples: Jovene.
Früchtl, Alois. 1912. Die Geldgeschäfte bei Cicero. Erlangen: Junge & Sohn.
Garcia Morcillo, Marta. 2005. Las Ventas por subasta en el mundo romano,
La Esfera privada. Barcelona: Universitat de Barcelona.
Garcia Morcillo, Marta. 2008. “Auctions, bankers and public finances in the
Roman world,” in Verboven, Vandorpe, and Chankowski, eds, 257–75.
Garcia Morcillo, Marta. 2014. “Publicidad, transparencia y legitimidad: sub-
astas en la obra de Cicerón,” in Catherine Apicella, Marie-Laurence
Haack, and François Lerouxel, eds, Les Affaires de Monsieur Andreau,
Économie et société du monde romain. Bordeaux-Pessac: Ausonius,
209–22.
Goldschmidt, Lewin. 1891. Universalgeschichte des Handelsrechts. Stuttgart:
F. Enke.
Harris, William V. 2006. “A revisionist view of Roman money.” 96 Journal of
Roman Studies 1–24.
Harris, William V. 2008. “The nature of Roman money,” in William
V. Harris, ed., The Monetary Systems of the Greeks and Romans. Oxford:
Oxford University Press, 174–207.
Hollander, David B. 2007. Money in the Late Roman Republic. Leiden/
Boston: Brill.
Howgego, Christopher. 1992. “The supply and use of money in the Roman
world, 200  to  300.” 82 Journal of Roman Studies 1–31.
Ioannatou, Marina. 2006. Affaires d’argent dans la correspondance de
Cicéron, L’aristocratie sénatoriale face à ses dettes. Paris: De Boccard.
Jacques, François, and Scheid, John. 1990. Rome et l’intégration de l’Empire,
44 av. J.-C.–260 ap. J.-C., vol. 1, Les Structures de l’Empire romain. Paris:
Presses Universitaires de France.
Jakab, Eva. 2011. “Financial transactions by women in Puteoli,” in Paul J. du
Plessis, ed., New Frontiers: Law and Society in the Roman World. Edin-
burgh: Edinburgh University Press, 123–50.
Johnston, David. 1999. Roman Law in Context. Cambridge: Cambridge
University Press.
Jones, David Francis. 2006. The Bankers of Puteoli: Finance, Trade and
Industry in the Roman World. London: Tempus.
Jongman, Wim. 1988. The Economy and Society of Pompeii. Amsterdam:
J. C. Gieben.
Kehoe, Dennis P., David M. Ratzan, and Uri Yiftach, eds. 2015. Law and
Transaction Costs in the Roman Economy. Ann Arbor: University of
Michigan Press.
Lenel, Otto. 1881. “Beiträge zur Kunde des Edicts und der Editcommentare.”
2 Zeitschrift der Savigny-Stiftung für Rechtsgeschichte 14–83.
110 Jean Andreau

Lerouxel, François. 2006. “Les femmes sur le marché du crédit en Égypte
romaine (30 av. J.-C.–284 apr. J.-C.). Une approche néo-institutionnaliste.”
37 Cahiers du Centre de Recherches Historiques 121–36.
Lerouxel, François. 2008. “La banque privée romaine et le marché du crédit
dans les tablettes de Murecine et les papyrus d’Égypte romaine,” in
Verboven, Vandorpe, and Chankowski, eds, 169–98.
Lerouxel, François. 2012. “Le marché du crédit privé, la bibliothèque des
acquêts et les tâches publiques en Égypte romaine.” 67 Annales, Histoire
Sciences Sociales 943–76.
Lerouxel, François. 2015. “The βιβλιοθήκη ἐγκτήσεων and transaction costs
in the credit market of Roman Egypt (30 ... to ca. 170 ..),” in Kehoe,
Ratzan, and Yiftach, eds, 162–84.
Lerouxel, François. 2016. Le Marché du crédit dans le monde romain, Rome:
Éditions de l’École Française de Rome.
Magdelain, André. 1990. Ius imperium auctoritas: Études de droit romain.
Rome: Éditions de l’École Française de Rome.
Maganzani, Lauretta. 2007. “L’Editto provinciale alla luce delle Verrine:
profili strutturali, criteri applicative,” in Julien Dubouloz and Sylvie Pittia,
eds, La Sicile de Cicéron: Lectures des Verrines. Besançon: Presses Uni-
versitaires de Franche-Comté, 127–46.
Mélèze-Modrzejewski, Joseph. 1970. “La Règle de droit dans l’Egypte
romaine,” in D. H. Samuel, ed., Proceedings of the Twelfth International
Congress of Papyrology. Toronto: A. M. Hakkert, 317–77.
Mélèze-Modrzejewski, Joseph. 1990. Droit impérial et traditions locales dans
l’Egypte romaine. Aldershot: Ashgate.
Michel, Jacques-Henri. 1962. La Gratuité en droit romain. Bruxelles: Uni-
versité Libre de Bruxelles.
Millett, Paul. 1991. Lending and Borrowing in Ancient Athens. Cambridge:
Nicolet, Claude. 1966. L’Ordre équestre à l’époque républicaine (312–43 av.
J.-C.), vol. I: Définitions juridiques et structures sociales. Paris: De Boccard.
Petrucci, A. 1991. Mensam exercere: studi sull’impresa finanzaria romana: II
secolo a.C–metà del III secolo d.C. Naples: Jovene.
Petrucci, A. 2002. Profili giuridici delle attività e dell’organizzazione delle
banche romane. Turin: Giappichelli.
Rauh, Nicholas. 1989. “Finances and estate sales in republican Rome.” 63
Aevum 45–76.
Saller, Richard P. 1982. Personal Patronage under the Early Empire. Cam-
Schulz, Fritz. 1946. History of Roman Legal Science. Oxford: Clarendon Press.
Shatzman, Israel. 1975. Senatorial Wealth and Roman Politics. Bruxelles:
Latomus.

Terpstra, Taco. 2008. “Roman Law, Transactions Costs and the Roman
Economy: Evidence from the Sulpicii Archive,” in Verboven, Vandorpe,
and Chankowski, eds, 345–69.
Thielmann, Georg. 1961. Die römische Privatauktion, zugleich ein Beitrag
zum römischen Bankierrecht. Berlin: Duncker & Humblot.
Thomas, Yan. 2004. “Travail incorporé dans une matière première, travail
d’usage et travail comme marchandise. Le droit comme matrice des
catégories économiques à Rome,” in Andreau, France, and Pittia, eds,
201–25.
Thomsen, Rudi. 1978. “From libral aes grave to uncial aes reduction, the
literary tradition and the numismatic evidence,” in Les Dévaluations à
Rome: Epoque républicaine et impériale, vol. 1. Rome: Éditions de l’École
Française de Rome, 9–30.
Verboven, Koen. 1993. “Le système financier à la fin de la République
romaine.” 24 Ancient Society 69–98.
Verboven, Koen. 2002. The Economy of Friends: Economic Aspects of Amicitia
and Patronage in the Late Republic. Bruxelles: Latomus.
Verboven, Koen. 2003a. “The Sulpicii from Puteoli, argentarii or faenera-
tores?,” in Pol Defosse, ed., Hommages à Carl Deroux, vol. III, Histoire et
Epigraphie, Droit. Bruxelles: Latomus, 429–45.
Verboven, Koen. 2003b. “The Sulpicii from Puteoli and Usury in the Early
Roman Empire.” 71 Tijdschrift voor Rechtsgechiedenis 7–24.
Verboven, Koen, Katelijn Vandorpe, and Véronique Chankowski, eds, Pistoi
dia tèn technèn: Bankers, Loans and Archives in the Ancient World, Studies
in Honour of Raymond Bogaert. Leuven: Peters.
Verley, Patrick. 1985. La révolution industrielle. Paris: Editions MA.
Yiftach-Firanko, Uri. 2009. “Law in Graeco-Roman Egypt: Hellenization,
fusion, Romanization,” in Roger S. Bagnall, ed., The Oxford Handbook of
Papyrology. Oxford: Oxford University Press, 541–60.
Zehnacker, Hubert. 1980. “Unciarium fenus (Tacite, Annales, VI, 16),” in
Mélanges de littérature et d’épigraphie latines, d’histoire ancienne et d’arch-
éologie (= Mélanges Pierre Wuilleumier). Paris: Belles Lettres, 353–62.
Zehnacker, Hubert. 1990. “Rome, une société archaïque au contact de la
monnaie (Vie–Ive siècles av. J.-C.),” in Crise et transformation des sociétés
archaïques de l’Italie antique au Ve siècle av. J.-C. Rome: Éditions de
l’École Française de Rome, 307–26.
16
Secured Transactions in Classical

Roman Law
Hendrik L. E. Verhagen
“For historians and lawyers interested in the connection of law

with economics the Roman law of real security will forever
remain an instructive phenomenon.”
(Schulz 1951: 405)
16.1. INTRODUCTION
When the rules for the taking and enforcement of security are
“cumbersome, inefficient and awkward” (Fleisig 2008: 90) security
may lose its essential economic function of mitigating risk for lenders.
This is the condition of the Roman law of real security according to
the prevailing view in modern literature on Roman law. It is often
added, however, that flaws in real security were not a major problem,
as personal security (guarantees, suretyships) was far more important
in ancient Rome than real security.¹ This may have been true for the
“economy of friends,” where loans were provided within the frame-
work of amicitia (“friendship”) relationships of members of the elite
¹ Schulz (1951: 405)—see §16.5. In a similar sense, Zimmermann (1996: 115–16);

Johnston (1999: 94). More nuanced Kaser (1971: 457); Kaser, Knütel, and Lohsse
(2017: 180). For positive assessments of the Roman law of real security, see Krämer
(2007: 1–6); Harris (2006); Terpstra (2008: 356–60).
Hendrik L. E. Verhagen, Secured Transactions in Classical Roman Law In: Roman Law and Economics:
DOI: 10.1093/oso/9780198787211.003.0016
114 Hendrik L. E. Verhagen

and friends were expected to enhance each other’s borrowing capacity
by standing surety and to discourage one another’s threatening
insolvency.² But in the archive of the Sulpicii, a collection of legal
documents involving a group of professional Roman (Puteoli)
money-lenders, we see that in ordinary commercial relationships
personal security was used for small loans and real security for larger
ones.³ In the papyri of Roman Egypt personal security is rare
throughout the Principate (Lerouxel 2016: 245–91). It is also hard
to believe that the Roman law of real security was relatively insignifi-
cant when so many opinions of the jurists on pignus have survived.⁴
The Roman law of security was highly sophisticated and versatile,
allowing multiple charges, non-possessory security, and even floating
charges. As legal systems often adapt in reaction to impulses from
their economic environment, the complexity of the Roman law of real
security would suggest that pignus and fiducia did play an important
role in the Roman economy.⁵ In this contribution it will be argued
that the prevailing view is wrong and that the Roman law of real
security was capable of facilitating credit in a similar manner as its
descendants in modern economies. The economic functions of
real security will be reviewed (§16.2). The requirements for an effect-
ive law of security, which are to a large extent a reflection of these
functions, will subsequently be applied to Roman law (§16.3). The
² Kaser (1982: 216–7); Krämer (2007: 3); Verboven (2002: 151).

³ Gröschler (2008: 303–5); Krämer (2007: 182). For the loans secured by real
security no personal security seems to have been granted. In Kaser, Knütel, and
Lohsse (2017: 180) it is noted that, although at the beginning of the Principate
personal security was still preferred, in the second century  real security had
developed into the most significant form of security. The archive of the Sulpicii
would suggest that this development already took place in the first century . This
is confirmed by Chemain (2015): 377–386. There are two excellent editions of the
Sulpicii archive: Camodeca (1999) (TPSulp) and Wolf (2012) (TPN).
⁴ In the Digest a whole book (D. 20) and a large title (D. 13.7) are devoted to real
security (pignus and hypotheca), while the Codex also contains a large number of
rescripts and other imperial constitutions dealing with pignus (C. 4.24; C. 8.13–8.34).
Also elsewhere in the Digest there are many fragments on pignus and hypotheca. In the
sources, the term pignus (“pledge”) is often used as the generic term for possessory
and non-possessory pledges, while the term hypotheca usually (but by no means
always) denotes non-possessory pledges specifically.
⁵ On the co-evolution of law and economy, see in particular Roe (1996) and
Luhmann (2004: 230–73). See also the contributions to Zumbansen and Callies
(2011) (with many further references).
Secured Transactions in Classical Roman Law 115

(lack of) publicity in the Roman law of real security will be given
separate attention (§16.4), as it is the most controversial aspect of the
Roman law of real security.
16.2. ROMAN LAW AND ECONOMICS
16.2.1. Economic Analysis of Roman Security Interests

In this chapter it will be examined what law and economics has to say
about the role of the law of secured credit in modern economies and it
will be indicated whether there is enough similarity with the role of
the law of secured credit in the Roman empire in order to apply the
findings of law and economics to it.
The laws of real security of modern civil law jurisdictions are built
on Roman foundations and are strikingly similar to them. In particu-
lar the Roman legal rules on pignus—as contained in the Corpus iuris
civilis—were received in the ius commune of medieval Europe and
were eventually (usually in only slightly modified form) adopted in
the great codifications of modern Europe. There is therefore a con-
tinuity of congruent legal solutions, which allows the Roman law of
real security to be analyzed in a similar fashion as modern laws
(Wieacker 1988: 29–30). It is true that the formal rules in the codifi-
cations may have acquired different meanings in the course of his-
tory. Moreover, the functioning of the formal rules may be subject to
different informal constraints. More generally, the socio-economic
environment of modern legal rules may be so different from that of
their Roman ancestors that a one-to-one transposition cannot be
made.⁶ One of the most important historians of the Roman economy,
Moses Finley, speaks of “the inapplicability to the ancient world of a
market-centred analysis” (Finley 1999 (1973): 26). If Finley is right
that ancient society did not have a market economy and that the
ancient economy is better analyzed with sociological methods exam-
ining the status of persons and their relationships, then the economic
analysis of ancient law might also be doomed to fail.
⁶ As Schiavone (2000: 177) warns us: “one should not be misled by the apparent
persistence of the juridical forms that cover the economic relationships, in cases in
which the modern usage of Roman law seems to be evidence of uninterrupted contact
between the ancient and the modern.”

Finley has been enormously influential in the Anglo-American
world. The school of thought which is considered by the Oxford
Handbook of Roman Studies (Scheidel) as the most influential one,
still has clear “primitivist” connotations. It considers the Roman
economy as “a highly localized, fragmented, and largely agrarian
economy that sustained a thin veneer of coerced transfers and trade
in luxuries and a network of towns that were dominated by land-
owning elites” (Scheidel 2010: 593). The purpose of this chapter is not
to take sides in the “Bücher/Meyer controversy” between “primitiv-
ists” and “modernists.” This chapter merely seeks to demonstrate that
Roman law did offer a legal framework sophisticated enough for a
market economy in which finance plays an important role. If that is
so, however, one wonders how likely it is that a state with a “primi-
tive” economy, even as described by moderate “neo-primitivists” as
having only a “thin veneer of coerced transfers and trade in luxuries,”
would develop a highly versatile and differentiated law of secured
credit?⁷ In any case, although for other areas of the law this may well
be different, I would claim that the interests at stake when a grain
merchant in first-century  Puteoli pledges his stocks in favor of his
financier are not much different from those when his colleague in
twenty-first-century Rotterdam does the same.⁸ As will be elaborated
below, the economic functions that are attributed to real security in
studies on secured credit in modern economies can, without diffi-
culty, also be applied to the Roman law of real security.
16.2.2. Economic Functions of Security

16.2.2.1. Primary Function of Security: Risk Reduction
In their groundbreaking Yale Law Journal article Jackson and
Kronman (1979: 1143) observed that “[t]o a considerable extent,
the value of a security interest depends on the degree to which it
insulates the secured party from the claims of the debtor’s other
creditors.” The primary economic function of security is therefore
⁷ In a similar sense, with reference to the Roman law of sale, Crook (1996: 35–6).
So also for commercial law generally Johnston (1999: 110–1), quoted with approval by
Frier (2000: 447).
⁸ See also Arruñada (in this volume), who assumes that “the basic problems that
market participants faced in Roman times were essentially the same as they face
now . . . ” (p. 256).

risk reduction for the creditor (Röver 1999: 108). In case of secured
loans the risk concerned is that of non-payment of principal and
interest at the agreed time. The more this risk is reduced by taking
security, the greater its value will be to the creditor. The risk manifests
itself in particular when the debtor is insolvent. In that event creditors
are likely to receive only a small portion of their claims, since the
liquidation proceeds of the debtor’s assets are usually significantly less
than the aggregate amount of his or her debts. Under the principle of
pari passu or paritas creditorum (“equality of creditors”) all creditors
receive distributions proportionate to their claims. Real security,
however, creates a separate class of creditors, with preferential rights
to the proceeds of specific assets (the charged assets). Thus, a secured
lender who has obtained a security interest on one or more of the
debtor’s assets is not affected by this pari passu principle. In other
words, where the law allows a creditor to obtain effective security, this
will considerably reduce the risk that he or she will suffer losses when
the loan is not fully repaid at the agreed time. The corresponding
benefit for debtors granting security is that they may obtain credit in a
manner (amount, term) which otherwise would not be available to
them or only at higher costs (interest). Also, in ancient Rome the
primary reason for taking security cannot have been anything else
than maximizing the lender’s prospects for recovery, and so it is likely
to have facilitated the borrower’s access to credit.⁹
16.2.2.2. Monitoring, Bonding, and Moral Hazard

Another economic function of real security has to do with monitoring
the debtor. Where security is taken in order to secure recovery, the
relatively simple process of monitoring the market value of the
charged assets can replace the far more complex process of monitor-
ing the borrower’s business as a whole. By focusing on the continued
availability of the charged assets, the secured creditor can achieve a
substantial reduction of his or her monitoring costs (Jackson and
⁹ Harris (2006: 6) mentions Cicero, Ad Att. 16.6.3 as an example of a borrower’s

differential treatment of secured and unsecured debt. The Sulpicii archive unfortu-
nately does not contain documents mentioning interest rates (although they must
have calculated interest). In particular for Roman Egypt there is relatively much
evidence for the enhancement of borrowing capacity by real security. See Lerouxel
(2015), who does, however, not discuss whether the interest rate for secured loans was
substantially lower than for unsecured loans.

Kronman 1979: 1153).¹⁰ A related function is that where the security
interest is attached to the charged property, this will restrict debtor
misbehavior, consisting of transferring assets to third parties (Jackson
and Kronman 1979: 1153). The same purpose (i.e. bonding) could be
achieved by contractual covenants in loan agreements in which the
borrower undertakes not to sell certain assets without the lender’s
permission. However, when the borrower is declared bankrupt, a
contractual action for breach of covenant will normally only be
marginally recoverable. A proprietary security interest will allow the
lender to enforce the security, even when the charged assets have been
transferred to third parties. Taking security may also prevent the
debtor from shifting risks to the detriment of his or her creditors by
adopting a more risky investment strategy after the credit has been
granted (moral hazard). The advantages of such a new strategy would
normally accrue for the debtor, while the disadvantages (insolvency)
would belong to the creditors. Security will entail that the debtors will
lose the charged assets if they default on their borrowings, which may
discourage them from adopting risky investment strategies (Röver
1999: 14). One economic function of security therefore is that it can
effectively restrict the debtor’s ability to transfer certain assets of
stable value to fund more risky transactions (Armour 2008: 7–8).
Again there is no reason why these functions should not also be
performed by real security in Rome. Let us look at one example
derived from the Sulpicii archive. In TPN 43 (TPSulp 51) the debtor
declares that the wheat, pledged in order to secure the initial loan of
HS 10,000, is stored “with me” (penes me) in the public granaries of
Puteoli. When some time later an additional loan of HS 3,000 was
granted, the same wheat was pledged to secure the total debt of HS
13,000. This time the debtor no longer declares that the wheat is
stored “with me” (TPN 44/TPSulp 52). From another document
(TPN 86/TPSulp 45) it becomes clear why: the creditor rented the
space in the granaries where the pledged wheat was stored (for a
nominal amount of HS 1 per month), thus effectively bringing the
pledged property under his control. What must have happened here
is that the pledge was originally created as a non-possessory pledge
and was later converted into a possessory one (Krämer 2007: 300–38;
Verhagen 2011: 28). The motivation behind this conversion is not
¹⁰ See also my discussion of impersonal exchange below, §16.2.2.3.


expressed, but is likely to have been that the creditor wished to
strengthen his position. Bringing the pledged wheat under his control
would entail that the creditor could be sure that as long as there was
sufficient wheat in the storeroom rented by him, it would be less
urgent to keep an eye on the debtor (monitoring). Moreover, the
borrower’s loss of control over the charged wheat may have been an
important stimulus for the proper repayment of the secured loan and
would prevent him from using the assets for more risky investment
strategies (bonding and moral hazard).
16.2.2.3. Depersonalizing Credit Relationships:

Impersonal Exchange and Adverse Selection
Economic development is facilitated by legal and other social institu-
tions that support impersonal exchange (Arruñada 2012: 17). In
particular Verboven, however, has demonstrated that highly person-
alized relationships based on amicitia played an important role in the
Roman economy. Amicitia was not merely a relationship based on
mutual affection and altruism; it could also involve the exchange of
money, goods, and services. In many cases, the exchange aspect was
even the raison d’être of the amicitia relationship. In financial transac-
tions amicitia entailed that one friend helped out another, by providing
interest-free loans or at low(er) interest (Verboven 2002: 116). Also
productive loans at higher interest (but still on more favorable terms)
were preferably obtained from friends. Amicitia would provide the
lender with an extra guarantee that repayment would take place, while
it would protect the debtor from excessive interest rates and allow him
to count on more leniency when unable to repay the loan in time
(Verboven 2002: 170–4). However, if amicitia was as pervasive in the
financial world as Verboven claims it to be, one wonders how the
substantial evidence for real security can be explained.¹¹ Precisely real
security enables a lender to provide credit outside the circle of his or her
relatives, friends, and their dependents. The lender’s lack of information
as to the borrower’s creditworthiness is much less a concern when the
¹¹ Terpstra (2013: 29): “if indeed amicitia and clientela were all-important in
business, it is simply not visible in the Murecine documents.” The Murecine docu-
ments are those of the Sulpicii archive, discovered at the locality of Murecine near
Pompeii (but originally stemming from Puteoli).

value of the collateral will be sufficient to cover the debt.¹² The law of
real security may have provided a mechanism of bringing lenders and
borrowers together who otherwise would not have engaged in credit
transactions. The Roman law of real security may have facilitated
impersonal exchange in financial transactions and may thus have
contributed to economic growth in the Roman empire. To put it
differently, the existence of a large body of fragments in the Digest
and the Code on real security at least suggests that many credit rela-
tionships were depersonalized.¹³
In these depersonalized relationships Roman real security may also
have had an information function, distinguishing good borrowers
from bad ones (adverse selection). Adverse selection may occur in
markets with asymmetrical information: information available to
one party (often about himself or herself) may remain hidden for
the other party. This could lead to adverse selection: parties who have
most to hide are most eager to enter into certain contracts (Stiglitz
and Weiss 1981: 393–410). The availability of security may, however,
reduce this process of adverse selection. Professional lenders generally
anticipate that only serious borrowers will be prepared (or be able) to
grant security. Thus security not only has a risk reducing function but
also an information function, distinguishing good borrowers from
bad ones (Röver 1999: 116–17).
16.2.2.4. Zero-Sum Game or Redistribution

of Wealth to Secured Creditors?
From a macro-economic perspective, the most important advantage
of the microeconomic function of risk reduction is that it increases
available credit, which leads to higher investments, increased produc-
tion, and ultimately a higher gross domestic product (Röver 1999:
¹² For Roman Egypt, see Lerouxel (2012: 960–3). According to Kay there are two
ways around the problem of the lender’s lack of information as to the borrower’s
creditworthiness: using a financial broker (with information) introducing the bor-
rower to the lender and deposit banking. In the latter case the lender takes a credit risk
on the banker rather than on the ultimate borrower, while the banker’s expertise is in
evaluating the creditworthiness of persons. Kay (2014: 110) does not mention,
however, real security as another way of remedying the lender’s lack of information.
¹³ On the other hand, rights of pledge were also created in highly personalized
contexts, such as between husband and wife (e.g. D. 20.6.11, Paul. 4 resp.) and between
brothers (e.g. D. 20.4.3.2, Pap. 11 resp.).

109–10). A fundamental objection, however, against the efficiency of
proprietary security is that secured credit involves a zero-sum game:
interest rate savings for debtors granting security are wiped out by
corresponding interest rate increases for debtors unable to offer secur-
ity (Schwartz 1984). So-called “redistributive” theories even hold that
the benefit of lower interest rates is acquired at the expense of non-
secured creditors, who will receive less in the insolvency of a debtor
who has granted security rights to one or more other creditors. The
growing empirical literature on the use of secured credit, however,
refutes these redistributive theories and demonstrates that, as Armour
(2008: 11–2) states, “secured credit is, on the whole, socially beneficial,
and . . . such benefits are highly likely to outweigh the social costs of
any transaction motivated by redistribution.” Armour argues that
granting security, by facilitating monitoring and bonding, reduces
the probability of the debtor engaging in risky wealth-reducing trans-
actions and thus increases the value of all creditors’ claims (2008: 7–8).
Some years ago it was argued that, in the more than thirty years of
debate since the Jackson/Kronman article, “a comprehensive justifi-
cation of secured commercial credit on efficiency grounds is unproven
and perhaps not provable” (McCall 2009: 11). However, legal history
seems to provide strong indications that security is efficient. As
McCormack (2004: 26) has put it: “an inefficiency conclusion would
go against the grain of history. Security devices are widespread and
pervasive not only in the modern industrialized world but also in
ancient societies, and one might ask the rhetorical question: why
does secured credit persist for so long if it is inefficient?”
16.3. EFFECTIVENESS OF THE ROMAN

LAW OF REAL SECURITY
16.3.1. Priority and Certainty in Ranking

The law of real security will, as we have seen, only be able to perform
its economic function of facilitating access to credit if the secured
lender’s claims are insulated from the claims of other creditors. An
effective law of security therefore requires, first and foremost, that the
creditor’s claims can be satisfied with priority over the other creditors
from the proceeds of the charged assets. In addition, the basic

requirement of priority should be supplemented with rules creating
certainty in ranking. It should—at the outset of a secured finance
transaction—be perfectly clear to the lender what the ranking of his
or her security right will be: whether the tax authorities will have a
higher-ranking privilege, whether other secured or privileged credit-
ors could prevail, et cetera.
Roman law does to a large extent meet these requirements for an
effective law of security. There could be no doubt for secured creditors
that the proceeds of the charged property were exclusively available in
order to discharge the secured debt. Also when insolvency proceed-
ings were conducted in respect of the debtor, the secured creditor
could take recourse against the charged property with priority over
other creditors. If the creditor was in possession of the charged
property, he or she could sell it to third parties and would only be
required to pay the surplus to the person conducting the insolvency
proceedings (in particular: the bonorum emptor). If the pledge was
non-possessory, the creditor could institute the actio Serviana to
recover the pledged property from the person in charge of the
relevant stage of the insolvency proceedings (Kaser and Hackl 1996:
402). There could, however, sometimes be creditors (in particular: the
imperial treasury and local authorities) with superior rights of
recourse against the pledged property, but this was rare.¹⁴ Only
from the second century onwards there was an increasing number
of cases in which all the debtor’s assets would be subject to a right of
pledge arising by operation of law, in particular in favor of the tax
authorities. These statutory general hypothecs may not have been as
devastating for secured credit as is often assumed in modern litera-
ture. They would generally also be subject to the prior tempore
principle, so that they could not adversely affect existing security
rights. Besides, when a creditor satisfied the tax debts of a debtor,
¹⁴ In D. 49.14.28 (Ulp. 3 disp.) the imperial treasury’s preferential right of recourse

over an after-acquired asset was granted preference over an anterior general pledge
created by the debtor in favor of a private individual. On this interesting text, in which
Ulpian restates an opinion by Papinian, see Wagner (1974: 180–92). Sometimes the
ranking was reversed by operation of law, in particular when the money provided by a
later secured creditor was used for the benefit of the pledged property, such as the
equipment, crew, or repair of ships or the storage of merchandise. See D. 20.4.5 (Ulp. 3
disp.) and D. 20.4.6 (Ulp. 73 ad ed.). In a rescript from 293 by Diocletian and
Maximian (C. 8.17.7) the pledge over real estate granted to the lender who financed
the purchase of the estate is considered to have priority over any other pledge (e.g. a
general pledge created earlier).

such creditor would be granted the same preferential position as the
tax authorities (Wagner 1974: 191). From Diocletian onwards, how-
ever, higher tax pressures may have increased the cases where assets
were previously charged with a fiscal pledge.
16.3.2. Creating Security

One benefit of taking security is that it provides a lower-cost method
for achieving what otherwise should be achieved by more complex
contractual arrangements. Jackson and Kronman (1979: 1157) observe
that if the law prohibited debtors from preferring one or more creditors
over others by granting security, a similar “network of priority rela-
tionships could be expected to emerge by consensual arrangement
between creditors.” However, particularly where the number of general
creditors is large, such arrangements will be more complicated to
achieve and involve more transaction costs. This would be true for
ancient Rome as well. However, security will only have this advantage
of saving on transaction costs for more complex priority relationships
when the law of secured credit offers efficiency in creation. In order
to reduce transaction costs, there should be as few formalities for
creating security as possible. Ideally there would be a single method
for creating security rights rather than a multiplicity of methods for
different types of charged assets (Armour 2008: 16). Also the creation
of security over after-acquired property should not be cumbersome.
It seems that originally in Rome res mancipi were always provided
as collateral by way of fiducia (cum creditore) and res nec mancipi by
way of pignus. In case of fiducia the debtor would—through manci-
patio—transfer land, slaves, horses, or cattle as collateral to a creditor.
A transfer in court by way of in iure cessio would also be possible,
though is likely to have been rare. Gaius explains why: “But generally,
in fact more or less always, we mancipate. For there is no point and no
need to do with greater difficulty in the presence of a praetor or the
governor of a province what we can do ourselves in the presence of
friends” (Gai., Inst. 2.25).¹⁵ Gaius can be regarded as offering second-
century advice to save on transaction costs! The epigraphic sources all
¹⁵ Plerumque tamen et fere semper mancipationibus utimur: quod enim ipsi per nos
praesentibus amicis agere possumus, hoc non interest nec necesse cum maiore difficul-
tate apud praetorem aut apud praesidem prouinciae agere. Translation taken from
Gordon and Robinson (1997: 134–5).

concern a fiducia cum creditore with a mancipatio (Noordraven 1999:
154–5). In the classical period it was possible that res mancipi could
also be charged by way of pignus (Verhagen 2011: 8–9).¹⁶ Thus a single
method for creating security over all types of asset became available,
which, moreover, could also be used for property subsequently. For the
creation of a right of pignus no formalities existed. Initially it was
probably required that the charged property was brought under the
factual control of the creditor, but at a later stage non-possessory
pledges could be created “by mere agreement” (nuda conventione).
16.3.3. Comprehensiveness
Another requirement for an effective law of real security is that of
comprehensiveness. First, granting and taking security should be
available to a wide range of debtors and creditors. For instance,
granting security should not be confined to persons acting in the
exercise of a trade or profession, and taking security should not be
confined to licensed credit institutions. No such limitations existed in
Roman law. Moreover, women could grant or accept security, and
security was often granted by or to slaves acting for their masters,
the latter category often in connection with a banking business.¹⁷
Security could also be granted for someone else’s debts (D. 20.1.5 pr.,
Marcian., ad form. hyp.), although for women this was prohibited by
the senatus consultum Vellaeanum ( 46?). Furthermore, security
rights should be possible with respect to a wide range of assets: not
only movable and immovable property, but also receivables (the
pledgor’s claims against his or her debtors). This was the case in
Roman law: all types of assets (including receivables) could be
encumbered with a right of pignus (Verhagen 2013a: 60–2). More-
over, it should be possible to create generic charges, for instance over
a herd or a shop’s inventory, without it being necessary that each
individual item be identified and that frequent updates be executed
for after-acquired items. This would increase transaction costs and
¹⁶ Ultimately pignus completely supplanted fiducia (although not yet in the late
classical period); Justinian’s compilers deleted fiducia from the jurists’ fragments: one
of the most important examples of systematic interpolation.
¹⁷ See e.g. TPN 73 (TPSulp 90) and the Mancipatio Pompeiana, FIRA III, nr. 91
(security (fiducia) by women); TPN 43, 44, and 69 (TPSulp 55, 51, 52, and 79) (slaves).

would be likely to affect the availability or cost (interest rate) of credit.
Under classical Roman law generic pledges could be created over
present and future assets, without the need to specify each individual
item.¹⁸ From the second century  it even became possible to create
a “floating charge” over all the debtor’s present and future assets
(Verhagen 2013a: 63–6).¹⁹ Finally, it should be possible to create
security for a wide range of debts. Roman law also complied with
this condition for an effective law of security. Security could be
granted not only for debts assumed at the time of its creation, but
also for previously existing debts, contingent debts, and even future
debts. The priority of the right of pledge could then be determined by
the time at which the pledge was granted rather than by the time the
secured debt arose. For example, in D. 20.4.9 pr. (Afr. 8 quaest.) the
slave Eros was first pledged to a landlord, in order to secure debts
under a lease that had not yet entered into effect, and then to a lender
in order to secure a loan, which (as was often the case in Rome) was
immediately due and payable. Although the debt owed to the lender
came into being first, the lessor’s right of pledge was first-ranking, as
it had been created earlier. Where, on the other hand, the contract
pursuant to which the secured debt would arise had not yet been
entered into at the time of creation, the pledge’s priority would be
determined by the time at which the secured debt arose. Thus, where
a pledge had been granted in order to secure a loan that had yet to
be entered into, the priority of the pledge would be determined by
the time at which the borrowed money was actually paid out to the
borrower (D. 20.4.1.1, Pap. 8 quaest.).
16.3.4. Rights of Use

16.3.4.1. Hypotheca and Debtor’s Right to Dispose
Empirical studies for contemporary economies demonstrate that
non-possessory security leads to a greater availability of credit
(Armour 2008: 19). Charged assets that are essential for the debtor’s
¹⁸ D. 20.1.13 pr. (Marcian., ad form. hyp.): herd; D. 20.1.34 pr. (Scaev. 27 dig.): shop.
¹⁹ e.g. D. 20.1.1. pr. (Pap. 11 resp.). See Wagner (1968). For a complete overview of
all Digest fragments on generic and general pledges, see Mentxaka (1986: 279–350).
For a comparison of the Roman general pledge with the floating charge of the
common law, see Verhagen (2013b: 135–42).

business should still be able to generate income with which the
secured debt can be discharged. By allowing the debtor to remain in
possession of the charged assets, he or she will be able to continue
using the assets for the exercise of a trade or profession. Also, the
debtor should be able (with the creditor’s permission) to dispose of
the charged assets in the ordinary course of his or her business.
In all likelihood the Roman right of pignus originated in the early
ius civile as a form of possessory security.²⁰ It is often held in modern
writing that the non-possessory pledge was only first recognized at
the time of the second-century jurist Julian, who is the author of the
earliest legal opinions on this form of security.²¹ It seems, however,
that pignus already could exist as a non-possessory security right
much earlier than Julian. In the Digest there are several fragments
from before Julian that seem to concern non-possessory security
(Krämer 2007: 249–75).²² The archive of the Sulpicii provides epi-
graphic evidence of a transactional practice of creating non-
possessory pledges already in the first century .²³ In any case, it is
clear that in the classical period pignus could be created as a non-
possessory security interest and that it was widely used. Moreover,
although originally the debtor could not transfer pledged property,
this later changed. With the secured creditor’s (express or implied)
consent the debtor could even transfer the pledged property without
the pledge continuing to attach to it.²⁴
16.3.4.2. Multiple Pledges

In particular when the (expected) liquidation proceeds of a charged
asset exceed the amount of the secured debt, the debtor can optimize
²⁰ The great majority of modern authors accepts this: see Krämer (2007: 121), with
further references.
²¹ Krämer (2007: 37), whose own view is that the non-possessory pledge was
recognized much earlier. See also Kaser (1971: 457).
²² One example is D. 13.7.30 Paul. 5 Epit. Alf. dig.), in which Paul restates the
opinion of the jurist Alfenus Varus (first century ) on what may have been a non-
possessory pledge of a boat. Krämer (2007: 211–48). Another (in my opinion more
problematic) example is D. 13.7.18.3 (Paul. 29 ad ed.), referring to the opinion of
Cassius (first century ).
²³ See §16.2.2.2 above. Fiducia cum creditore had always been possible as a form of
non-possessory security. The mancipatio did not require an actual transfer of posses-
sion. Often the debtor would rent the property from the creditor or hold it as bailee
(precarium): Gai., Inst. 2.60.
²⁴ See §16.4.4.4 below.

the collateral value of the asset if the law enables him or her to create
several layers of security in favor of multiple creditors. In classical
Roman law this was for some time problematic. For, after the first
fiduciary transfer, a debtor would no longer be the owner of the
property and would therefore not have the power to transfer it validly
to another creditor. When the security granted to the creditors was a
right of pignus the position originally was largely the same. Although
there is good reason to assume that the non-possessory pledge was
recognized relatively early, the recognition of its “logical sequel”
(Kunkel 1973: 155)—the multiple pledge—took much longer. For a
long time, the granting of the pledge would deprive the debtor of his
or her power to dispose of the pledged asset. This may have been a
relic from the past, when pignus—like fiducia—may also have been
regarded as a form of ownership (Kaser 1982: 72–3). The functional
similarity and analogous treatment of fiducia and pignus by the jurists
may also have contributed to this (Krämer 2007: 370). During the
course of classical law the inability to create multiple pledges over the
same assets disappeared, under the influence of transactional prac-
tices (Kaser 1976: 201–5). In fact, one of the reasons why fiducia
ultimately lost the struggle for life against pignus and disappeared
from practice may have been that it never did enable a debtor and
creditor deliberately to create lower-ranking security.
16.3.4.3. Rights of Use for Creditor: Antichresis,

Rehypothecation, and Assignment of the Secured Debt
Sometimes, the availability of credit may be enhanced by giving the
creditor rights to use the charged assets. In Roman law such rights of
use could take several forms. The creditor could use the (natural or
civil) fruits of the charged property to discharge the secured debt or
to cover the interest (antichresis). The creditor could also charge the
collateral for his or her own debts (“rehypothecation”).²⁵ Also, when a
creditor assigns a secured claim to another person, that other person
(the assignee) should be able to enforce the security when the
assigned debt is not properly discharged. Although the assignment
of debts was problematic in Roman law, devices were used to achieve
²⁵ e.g. D. 36.4.5.21 (Ulp. 52 ad ed.): fruits; D. 44.3.14.3 (Scaev. quaest. publ. tract.):
rehypothecation.

the same results, and in those situations the “assignee” was entitled to
enforce the security.²⁶
16.3.5. Enforcing Security

In many if not most legal systems, going to court is a relatively costly
and time-consuming way for creditors to get what is due to them.
However, precisely an effective law of real security enables a secured
creditor to take enforcement measures without the need to go to court
first (Jackson and Kronman 1979: 1143). Empirical studies indicate
that efficient enforcement mechanisms lead to greater availability of
credit, lower interest rates, and less collateral for equivalent levels of
borrowing (Armour 2008: 19). Also, the opening of insolvency pro-
ceedings should not adversely affect the secured creditor’s ability to
take recourse against the charged assets. At the same time, it should
be born in mind that certain formalistic enforcement procedures, in
particular public auctions, may have the positive effect of generating
the highest possible price for the charged property. This may benefit
not only the secured creditor, but also (in case of a surplus) lower-
ranking creditors or the debtor.
Under Roman law the collateral would originally be forfeited to
the creditor, who would become its owner. From that point on the
creditor could treat the object as his or her own, even if its value
exceeded the amount of the secured debt (Verhagen 2013a: 69–78).²⁷
Already for this early stage, we have epigraphic evidence of contract-
ing parties expressly agreeing that the creditor would be authorized to
sell the object of security in case of a payment default, in which case
the creditor should turn over any surplus to the debtor, while the
debtor would remain liable for any deficit.²⁸ In practice the creditor
will usually have auctioned the charged property. In the first century
 a well-organized auction practice existed, both for “ordinary”
commercial sales and for enforcement sales (Noordraven 1999: 247;
Lerouxel 2008: 189).²⁹ In the Sulpicii archive we find many
²⁶ e.g. D. 18.4.6 (Paul. 5 quaest.).

²⁷ The only difference was that in case of fiducia ownership would transfer
immediately to the creditor, whereas in case of pignus this would take place upon
the debtor’s default.
²⁸ e.g. TPN 69 (TPSulp 79): pignus; Formula Baetica, FIRA III, nr. 92: fiducia.
²⁹ The documents in the Sulpcii archive concerning auctions are reviewed by
Noordraven (1999: 247–59). See also Romeo (2006), with further references.

documents relating to the auction of assets (slaves, cloth, immovable
property) that had been pledged or fiduciarily transferred.³⁰ We also
know that there were many bankers (argentarii) in the Roman empire
specializing (inter alia) in providing credit to bidders at auction
(Andreau 1999: 38–9). The Sulpicii may have belonged to this cat-
egory, which would mean that enforcing security by way of auction
would be very familiar to them.
For a long time, it was purely at the creditor’s discretion whether or
not recourse would be taken against the charged assets by selling
them (at auction or otherwise). Only at the end of the classical period
had the creditor become obliged by operation of law to sell the object
of security and turn over the surplus to the debtor. Court permission,
however, was still not required for selling the charged assets and
taking recourse against the proceeds of the collateral. Only forfeiture
required the permission of the imperial chancery: impetratio dominii.
This evolution from taking recourse by forfeiture to mandatory
enforcement by way of sale can also be observed in other legal systems
(Verhagen 2011: 11–13, 2013b: 152–9). As Moses Finley observed
(1953: 266): “[i]n its earliest form, security is always substitution, a
forfeit.” According to Finley the cause of this “profound economic
transformation” lies outside the law of security and has everything to
do with the growing importance of credit. Credit providers are not
interested in obtaining ownership of the object of security as such:
they want to liquidate it as soon as possible, so as to be able to take
recourse against the proceeds in order to satisfy their claims. At the
same time, debtors wish to lay their hands on the excess value of
the collateral they provided. Thus, rules providing for a compulsory
sale of collateral and the return of the superfluum (excess proceeds) to
the debtor, as had developed in Rome by the end of the classical
period (but have their origin in transactional practices from the early
Principate), result in an equilibrium in which the interests of creditors
and debtors are more or less balanced.
Interestingly, forfeiture clauses, in the form of conditional sales of
the charged property to the secured creditor, were still used at the end
of the classical period, so apparently there was a need for them. They
were legally valid, provided that the purchase price payable by the
secured creditor to the debtor reflected the fair market value (iustum
³⁰ TPN 70, 71, 73, 74, 75, 76, 78, 79, and 80 (TPSulp 83, 84, 90, 91, 92, 85, 87,
and 89).

pretium) of the property, so that when its value exceeded the amount
of the secured debt the surplus would go to the debtor (D. 20.1.16.9,
Marcian., ad form. hyp.; Verhagen 2013a: 73–7). Conditional sales of
this type can indeed be an efficient means of taking recourse against
charged assets, as they dispense with the need to auction the property
soon after the debtor defaults, so that in a depressed market the
creditor has the option to wait for better times. What at the beginning
of the twenty-first century is regarded as “state of the art” in collateral
enforcement was already possible in the Roman law of the second
century !³¹
16.4. PUBLICITY
16.4.1. Publicity: Introduction

The main criticism endorsed by modern Romanists on the Roman
law of real security concerns its lack of publicity. Non-possessory
security could be created over movable and immovable property,
whether or not by way of a “floating charge,” without registration in
a public register. In contrast to most developed modern countries,
Rome did not have public mortgage registries for immovable property
that could be consulted by prospective secured creditors in order to
determine whether the debtor was the owner of the property and
whether “limited” rights in rem (e.g. pledges, usufructs) had previ-
ously been created over the property. Moreover, again in contrast to
modern legal systems, Roman law did not recognize a general prin-
ciple allowing third parties in good faith to rely on possession of
movable property.³² In other words, the adverse consequences of lack
of publicity were not compensated for by rules protecting bona fide
third parties. The oldest right in rem (nearly) always prevailed and
could not be adversely affected by dispositions made without the
consent of the oldest title holder (Verhagen 2014: 982). “Not ‘security
of intercourse’ but security of the vested right was dear to the Roman
heart” (Schulz 1936: 252).
³¹ See e.g. Directive 2002/47/EC of the European Parliament and of the Council of
June 6, 2002, on financial collateral arrangements and revised (2006) article 2078 of
the French civil code.
³² See however usucapio, discussed in §16.4.4.2 below.

16.4.2. Registries Support Impersonal Exchange
In a monograph on the economic and legal functions of property
registries, Benito Arruñada (2012) sketches a dilemma concerning
security interests that can be summarized as follows. The point of
departure is that impersonal exchange expands the number of trans-
actions that can be entered into and thus enhances economic growth
(Arruñada 2012: 17). In secured lending transactions impersonal
exchange means that the characteristics (reliability, solvency, etc.) of
the debtor are not so much relevant as those of the charged assets.
Granting the creditor a pledge, mortgage, or other form of security
right in rem enables him or her to rely effectively on the charged assets.
A proprietary security interest has a strong enforcement advantage, in
that it allows the creditor to take recourse against the charged assets
with preference over other creditors. The downside for commercial
and financial intercourse of rights in rem, however, is that the other
creditors run the risk of getting something of much less value than they
expected. This may happen, for instance, when previously charged
property is purchased by, or charged to, someone who is not aware
of the charge. Should, on the other hand, the purchaser or the subse-
quent secured creditor be protected against a previously created right
in rem, this would reduce the enforcement advantage for the original
secured creditor. Property registries can solve this dilemma by provid-
ing verifiable information, so that the law can allow rights in rem
(benefiting the original secured creditor) while at the same time enab-
ling subsequent buyers or secured creditors to verify what the legal
status of the property is.³³ However, in a particular (contemporary or
ancient) society this will only be true when registries are technically
feasible and reliable and can be used at relatively low costs.
16.4.3. Public Registers in the Roman Empire

16.4.3.1. Registration in the Roman Empire
Douglass North (2010: 15) describes efficiency as “a condition in
which given the state of technology and information costs, the market
has the lowest production and transaction costs attainable.”
³³ For a distinct summary of the advantages of registration, see Arruñada

(2012: 65).

According to North the “term is almost always used in relative rather
than absolute terms.” Thus whereas in a digitalized world a public
registration system for security rights may be efficient, it may very
well be an inefficient system for the ancient Roman world. For
immovable property operating a public registration system may gen-
erally be less problematic (although it may still be difficult) than for
movable assets (e.g. stocks) and intangible assets (e.g. trade receiv-
ables). This may already have been the case in the ancient world:
Roman Egypt did have registration systems for real estate (Taubens-
chlag 1955: 222–30; Wolff 1978: 184–262; Lerouxel 2012). In Egypt
the Romans even introduced a new registration system for the most
valuable assets (real estate and slaves): the “archive of acquisitions”
(βιβλιοθήκη ἐγκτήσεων). Although the prime motivation for these
changes may have been to control documentary practice with a view
to taxes and liturgies, the archive clearly also had a publicity function
for private parties. In particular, one concern of the Romans was
to prevent making dispositions of land, houses, and slaves by
persons without title (Von Woess 1924: 29). Moreover, the archive
also recorded security interests and even fiscal preferences (proto-
praxia), so that potential purchasers or secured creditors could see
whether the property had been previously charged (Von Woess
1924: 3–4).³⁴ Von Woess and more recently Lerouxel have stressed
the high quality and accessibility of the archive (Von Woess 1924:
31; Lerouxel 2012: 967–8). Lerouxel (2015: 178) even considers the
creation of the archive a “pivotal moment” in the history of the
private credit market in Roman Egypt, causing a “tremendous
growth” in the number of secured loans and the amounts loaned
out by individuals.
a. Registration in Roman Italy

The registration of charges of real estate was not a phenomenon
completely unknown to Roman Italy. When citizens entered into
certain contracts with the Roman state or with local municipalities,
³⁴ A papyrus discovered in Oxyrhynchos records an ordinance dating from 89 by

Mettius Rufus, prefect of Egypt. One of the provisions of this ordinance reads as
follows: “ . . . Therefore I command that all owners shall register their property at the
record-office within six months, and all lenders the mortgages that they hold, and
other persons the claims which they possess.” FIRA I, nr. 60 (P.Oxy. 2, 237):
translation taken from Bowman and Wilson (2009: 40).

they might be required to grant a special security interest over real
estate in favor of these bodies: the cautio praedibus praediisque. This
form of security was created by a special declaration of the owner of
the estate (subsignatio), which was registered in the records of the
state (tabulae publicae) or of the municipality (tabulae communes).
This form of (non-possessory) security was not a private law institu-
tion but rather a charge arising under public law (Kaser 1971:
459–60).³⁵ As Schulz (1951: 412–13) observed, “whenever we cross
the boundary of private law we enter a new world: we find documents,
registration and publicity, phenomena which were entirely foreign to
classical private law.” The Roman practice of recording security
interests primarily served to enforce the state’s claims rather than to
publicize these interests to potential future acquirers or secured
creditors. These observations also apply to another form of security
granted over real estate to public bodies, recorded in the so-called
tabulae alimentariae (Criniti 1991). In the second century the imper-
ial treasury granted loans secured by a pledge over real estate, with the
intention that the interest generated by these loans be for the benefit
of children in need. The secured loans were recorded in a document,
containing the names of the borrowers, descriptions of the real estate,
the amounts of the loans, and the rates of interest, all of which was
stored in public archives (Wenger 1953: 761–6). Moreover, there is
evidence that already during the Principate Roman cities did have a
cadaster, a registry whose purpose it was to demarcate individual
plots of (often: public) land (Moatti 1993; Arruñada in this volume).
For taxation purposes, registries of land in the census existed from the
time of Augustus.³⁶ So it would have been technically possible for the
Romans to create a registration system for immovable property,
similar to the Egyptian archive of acquisitions. But apparently the
Romans saw no need for it in Rome and elsewhere in the western part
of the empire.
³⁵ This is undisputed for praedia granted to the Roman people, but less so for
praedia granted to municipia: van Gessel (2003: 97).
³⁶ Arruñada (in this volume), with further references. That such registers could
also be relevant in property disputes between private parties is demonstrated by
D. 10.1.11 (Pap. 2 resp.), where it is observed that in boundary disputes, when there
are no “old records” (vetera monumenta), one should follow the most recent regis-
tration by the tax authorities (census).
b. Why No Property Registries in Rome?

Why did the Romans introduce a property registry in Egypt, but fail to
do so in their home territory?³⁷ By way of conjecture I submit that the
following factors could have played a role. In Egypt the Romans built
upon existing registration systems set up and operated under the
Ptolemies (Von Woess 1924: 28). The costs of building upon existing
registration systems, including a body of professionals (notaries)
experienced in executing and registering transaction documents,
may be considerably lower than setting up an entirely new system.³⁸
Also the Romans’ desire to get a grip on property and commercial
relationships in their new province may have played a role, in par-
ticular with a view to taxation.³⁹ Another reason is more of a psycho-
logical nature and may be that Roman citizens would consider it an
unacceptable infringement upon their privacy if third parties could
see whether their property had been charged for indebtedness (Von
Woess 1924: 30).⁴⁰ A lesser degree of social cohesion in Roman Egypt
than in Italy may also provide (part of ) the explanation. As will be
discussed below, an important reason why the Roman law of property
did not develop a principle protecting third parties in good faith is
likely to have been that in the Republic and in the Principate social
norms (amicitia, infamia) generally provided safeguards against a
defrauding debtor who charged someone else’s property or failed to
disclose an earlier charge. These social norms may not have been as
stringently observed in Roman Egypt, where the population was
largely non-Roman.⁴¹
³⁷ According to Arruñada (in this volume), the costs of archiving should not have
been a serious obstacle for creating property registries.
³⁸ Lerouxel (2012: 958): “À l’époque romaine, l’Égypte est donc une grande terre de
notariat et elle a atteint un haut degré de culture pratique de l’écrit.”
³⁹ It is not uncommon that institutions are introduced in colonial territories that
are considered unnecessary in the dominating country. For instance, a land registra-
tion system was introduced in Ireland and the colonial possessions of the United
Kingdom long before it was introduced in England and Wales: Watson (2001: 56).
⁴⁰ Arruñada (2012: 108–9) mentions that the failure of the Crown’s attempt to
introduce property registers for land in England in 1581 and 1673 has been attributed
to the opposition of the nobility, who did not want their debts to be known to the
public. Other European states have similar experiences.
⁴¹ Lerouxel (2012: 946, 967–76) argues that the interaction between the private
credit market and the finance of state expenditures may explain why the archive was
created. These funding techniques were, however, not unique for Roman Egypt, as
also Lerouxel notices.

16.4.4. Informal Publicity through Possession
16.4.4.1. No Protection of Bona Fide Third Parties
Two fundamental principles in the modern continental laws of
property—which are of Roman origin—are the nemo plus-principle
and the prior tempore-principle. The nemo plus-principle entails that
a non-owner cannot transfer ownership or grant other rights in rem
(e.g. pignus) to someone else.⁴² The prior tempore-principle entails
that when two or more competing real rights exist over the same asset
the oldest of these rights prevails.⁴³ In modern civil law jurisdictions
the nemo plus-principle and the prior tempore-principle can—in
respect of movable property—be overridden by a third principle of
property law, what one could call the “legitimizing effect” of posses-
sion. A person who in good faith relied on the semblance of owner-
ship reflected by his or her predecessor’s possession will be protected:
possession legitimizes the other party’s belief that the possessor was
entitled to dispose of the property. Applied to security interests, this
principle provides that when a non-owning possessor creates a pos-
sessory security interest over movable assets in favor of a creditor who
had no reason to believe that the debtor did not own the property, the
creditor will acquire a valid and enforceable security interest.⁴⁴ Like-
wise, when the owner in possession grants a possessory security
interest in favor of a creditor, that security interest will prevail in
ranking over a previously created non-possessory security interest of
which the creditor could not have been aware.⁴⁵ In Roman law all this
is different. Here the nemo plus-principle and the prior tempore-
principle were not counterbalanced by a general principle allowing
third parties in good faith to rely on their counterparty’s possession.
There was therefore no general protection for the creditor who in
good faith accepted security on property already charged in favor
⁴² The maxim nemo plus iuris ad alium transferre potest quam ipse haberet (“no
one can transfer more in law to another than he himself has”) is as such articulated as
a general principle of property law from the twelfth century onwards by the Glossators
(see e.g. Glossa Ordinaria, gl. Nemo plus ad D. 50.17.54). For pignus, see e.g.
D. 20.1.3.1 (Pap. 20 quaest.).
⁴³ Prior tempore potior iure (“earlier in time, stronger in law”). See Ant. C. 8.17.3
( 212, on pignus of land). Another principle of property law is that rights in rem
follow the property: “droit de suite.” For pignus see D. 13.7.18.2 (Paul. 29 ad ed.).
⁴⁴ e.g. §1207 German civil code (BGB); art. 3:238(1) Dutch civil code (BW).
⁴⁵ e.g. §1208 BGB; art. 3:238(2) BW. In both cases bona fide creditors will only be
protected when their security interest is possessory.

of someone else or on property owned by someone else.⁴⁶ The
oldest right in rem prevailed and could not be adversely affected
by dispositions made without the consent of the oldest title holder
(Verhagen 2014).
16.4.4.2. Usucapio and Possessory Interdicts

Roman law did have an institution which could have evolved into
such general principle protecting third parties in good faith: acquisi-
tive prescription by way of usucapio. Already at a very early stage,
however, the practically important category of stolen or embezzled
property (res furtivae) was excluded from usucapio (Kaser 1971: 37).
Moreover, usucapio could neither lead to the acquisition of a pledge
created by a non-owner, nor could it protect a secured creditor
against an earlier pledge unknown to him or her. Nevertheless,
usucapio did fulfill a useful role in Roman society. Not only did it
dispense with the need to transfer res mancipi by way of mancipatio,
it also relieved the purchaser (or other acquirer) from the probatio
diabolica (“diabolical proof”) of an unbroken chain of rightful
acquisitions.⁴⁷ As such it could also benefit the creditor when he or
she had to demonstrate that his or her right of pledge had been
granted by a debtor owning the pledged property. To this it may be
added that any possessor, even of stolen property, would have the
possessory interdicts at his or her disposal, also against the owner, in
order to act against threatening or actual interferences with his or her
possession. The possessory interdicts were also available to the
secured creditor whose factual possession of the pledged object was
(or threatened to be) interfered with. The rules on possessio and
usucapio thus reached a “compromise between the Roman dislike of
interference with vested rights and the practical need to give some
recognition to established facts” (Nicholas 1962: 124–5).
⁴⁶ Similarly a purchaser of previously charged property would not be protected.

⁴⁷ Already in preclassical law when the civilian (Quiritary) owner would transfer a
res mancipi by way of traditio (informal transfer of possession) the transferee would
after one (movable property) or two (immovable property) years acquire civilian
ownership through usucapio. In the meantime the transferee could reclaim the
property from every possessor with the actio Publiciana and ward off a rei vindicatio
by the civilian owner with the exceptio rei venditae et traditae. This strong legal
position is in modern literature (e.g. Jolowicz and Nicholas (1972: 99–100)) often
referred to as “bonitary” ownership. See in particular Ankum and Pool (1989).
16.4.4.3. Occasional Protection against Lack of Ownership

In a limited number of cases Roman law did protect a creditor against
the invalidity of a pledge created by a non-owner (Verhagen 2014).
To this category belong, in the first place, texts dealing with a
“relative” pledge. Wubbe (1960: 58–62, 92) has argued that whenever
a non-owner in factual possession of the pledged asset would grant a
pledge, the creditor could use the actio Serviana to recover the
pledged assets from anyone possessing them, except from the full
(pleno iure) owner, the bonitary owner, and their successors. A less
radical position is that only a bona fide possessor, who was in the
process of acquiring civilian ownership through usucapio, could
create a relative right of pledge (Verhagen 2014: 986–8).⁴⁸ Secondly,
a rescript by Septimius Severus and Caracalla from  205 (C. 8.15.2)
holds that, although a pledge created without the permission of the
owner is invalid, the situation is different when the owner knowingly
allowed the pledge to be created. Fraudulent behavior in respect of an
innocent creditor by the owner has the consequence that the inval-
idity of the pledge cannot be invoked against a pledgee in good faith
(Ankum 2010: 37). Here Roman law comes close to results reached by
modern rules on protecting bona fide secured creditors.
16.4.4.4. Purchasers of Pledged Property

The position of the purchaser of pledged property in classical Roman
law is that the secured creditor could always invoke the pledge even
when the purchaser was in possession of the property and could
reasonably have been unaware of the pledge. Obviously this legal
rule could have seriously affected commercial intercourse in Rome.
However, we have a large number of texts indicating that the secured
creditor’s permission to the debtor to sell the pledged assets would
enable the debtor to transfer the pledged assets unencumbered.⁴⁹
In other words, where the debtor was allowed by the secured creditor
to sell goods in the ordinary course of business, commercial
⁴⁸ According to Ankum and Pool (1989: 32) it was even required that the debtor
be a bonitary owner, i.e. someone to whom res mancipi had been conveyed by an
owner.
⁴⁹ See in particular the fragments in D. 20.5, D. 20.6, and C. 8.25.

intercourse would not be disrupted. Moreover, Digest fragments on
general pledges indicate that, although a practice existed of creating
security over all one’s assets, such a security interest did not prejudice
the interests of third parties dealing with the debtor in the course of
business (Verhagen 2013a: 63–6). In other words, the general pledge
may have created an equilibrium between, on the one hand, the
secured creditor’s interest of having a preferential right of recourse
against the assets comprised from time to time in the debtor’s estate,
and, on the other hand, the debtor’s interest of being able to sell assets
in the ordinary course of business and the debtor’s customers’ interest
of acquiring them free of a security interest. This equilibrium is
essentially the same as that reached by modern rules allowing third
parties in good faith to rely on their counterparty’s possession.
16.4.5. Transactional Practices Entailing Publicity

16.4.5.1. Introduction
In ancient Athens there was a practice of placing “mortgage stones”
(horoi) on mortgaged real estate, indicating the name of the creditor
and the amount secured by the mortgage (Fine 1951; Finley 1952,
1953: 249–68). Even this (simple) practice is not attested for Rome.
But this is not to say that Roman legal practice did not have other
practices informing third parties or otherwise reducing the risk that
the secured creditor would be confronted with adverse claims by the
real owner or another secured creditor. There were certain transac-
tional practices which, in combination with the social and criminal
sanctions discussed below, must have offered some compensation for
lack of publicity and protection of good faith.
16.4.5.2. Mancipatio
Under Roman law res mancipi were transferred as security to a
creditor by way of fiducia cum creditore. The debtor would, through
mancipatio, transfer land, slaves, or cattle as collateral to a creditor.
As a consequence of the mancipatio, the creditor acquired dominium
ex iure Quiritium (civilian ownership) of the object of fiducia. In the
accompanying pactum fiduciae it would normally be expressly agreed
that the creditor would retransfer the collateral if and when the
secured debt was discharged. The mancipatio can be regarded as a

symbolic sale, even when it took place by way of security.⁵⁰ It required
the presence of six Roman citizens: five witnesses and a sixth person,
the libripens, who held a pair of scales. In the presence of these
persons the transferee would (in case a slave was transferred) declare:
“I assert that this man is mine by Quiritary right, and let him have
been acquired to me with this piece of bronze and a bronze scale.”⁵¹
Subsequently, the transferee would strike the scales with the piece of
bronze (aes) and give it to the transferor in payment of the “purchase
price.” Although the original function of the witnesses may have been
that of members of the “clan” or family expressing their consent to
the transfer of property to which they were co-entitled, by the time of
the Twelve Tables its main function must have been to create cer-
tainty as to the validity of the transfer of ownership of important
capital goods (Kaser 1956: 120–1; Amunátegui Perelló 2012: 343).
The large number of witnesses not only served to facilitate proof that
the mancipatio had actually taken place, it also enhanced publicity
in the preliminary stage (Kaser 1971: 42). Through the formal invi-
tation of witnesses (rogatio) to the mancipatio, the intended transfer
was made public, so that persons with adverse claims could come
forward and object to the transfer, for example, by refusing to act as
witnesses, by asking other members of the local community to do so,
or by initiating legal proceedings (Kaser 1956: 120–2). Thus the
publicity was also aimed at preventing transfers of property over
which third parties exercised better rights. In other words, the pres-
ence of witnesses publicized the intended and the completed manci-
patio. When the res mancipi was a slave, horse, or cattle the acquirer
could put up an uncontested mancipatio as a defense against the
accusation of theft (Kaser 1956: 134). Moreover, an uncontested
mancipatio would make it easier for the acquirer to prove that he
was the true owner of the property, especially when he could also
⁵⁰ Also in the case of fiducia cum creditore the “sale” was for a nominal purchase
price of HS 1. See the Mancipatio Pompeiana, FIRA III, nr. 91 (singula sestertis nu-
[mmis sin]gulis) and Formula Baetica, FIRA III, nr. 92 (sestertio n(ummo) I). The
transferor’s liability for eviction under the actio auctoritatis for double the purchase
price was based on the mancipatio itself, so that a purchase price of HS I effectively
exempted the transferor from liability.
⁵¹ Hunc ego hominem ex iure Quiritium meum esse aio, isque mihi emptus esto hoc
aere aeneaque libra. Translation taken from Jolowicz and Nicholas (1972: 144).

prove a chain of earlier transfers by way of mancipatio (Kaser 1956:
134–6). As the mancipatio was, in contrast to the traditio, “abstract,”
it was not necessary to prove that the earlier transfers had all been
made pursuant to valid contacts or other legal transactions.
The mancipatio may have served this publicity function well when
Rome was a small agrarian community around the Palatine, but must
have lost much of its effectiveness when Rome developed into a
much larger community with hundreds of thousands of inhabitants.⁵²
This may have been one of the reasons why—in the Principate—
mancipatio was in practice often replaced by informal traditio and
res mancipi came to be charged by way of the (informal) pignus rather
than by way of fiducia cum creditore. Nevertheless, in the Principate,
the presence of witnesses probably continued to have a publicity
function, particularly when agricultural property (farms, arable land)
was transferred in rural areas.⁵³
The formalities for the mancipatio itself were relaxed during the
course of the classical period. According to Gaius the parties and
witnesses did not actually need to be on the land when the mancipatio
took place.⁵⁴ In the late classical period the oral formalities and the
use of the scale were replaced with a document in which the manci-
patio was recorded. There is, however, difference of opinion in
modern literature as to whether the presence of witnesses was still
required in the late classical period (Kaser 1975: 274). In any case, it is
likely that in practice witnesses were used both in mancipatio and
in an informal traditio whenever there were no other means of
⁵² Also the fact that the mancipatio could only be used by Roman citizens or
citizens benefiting from commercium may have contributed to this: Sturm (1993: 349).
Another factor contributing to the demise of mancipatio was that simpler legal
arrangements, in particular the stipulatio duplae (undertaking to pay twice the
amount of the purchase price), were available to make a seller liable in case of eviction:
Sturm (1993: 354).
⁵³ Documents recording the conveyance of land usually contained the name of the
plot of land, its extent, the municipality (and sometimes also: pagus and vicus) in
which it was situated, the names of neighbors, natural features of the land, buildings
on the land. See for instance Tabula Veleia (Criniti 1991) and Formula Baetica (FIRA
III, nr. 92).
⁵⁴ Gai., Inst. 1.121. In postclassical law this changed. At the beginning of the fourth
century  Emperor Constantine issued a long constitution which required that the
sale and transfer of land should take place on site (reversing Gai., Inst. 1.121) and in
the presence of neighbors. See Fragmenta Vaticana 35: Honoré (1989: 142–9); Voss
(1982: 135–77).

establishing the boundaries of the transferred land.⁵⁵ In case
of provincial land (which was res nec mancipi) this was common
practice and this is likely to also have been the case with land
in Italy.⁵⁶
16.4.5.3. Documents
The fact that in classical Roman law writing was rarely prescribed
as a legal formality does not mean that in practice many transac-
tions were not recorded in written documents and witnessed. As
the archive of the Sulpicii and other epigraphic and papyrological
evidence demonstrate, the creation of security (fiducia or pignus)
was often evidenced in writing in the style of chirographa or
testationes.⁵⁷ Writing served as evidence for the granting of a
pledge (D. 20.1.4, Gai., ad form. hyp.). This was particularly rele-
vant for non-possessory pledges (hypothecae), which were other-
wise more difficult to prove than possessory pledges. Rather than
using the relatively inexpensive medium of papyrus, the parties
even chose to record their transactions in wooden writing tablets
(tabulae ceratae), to which the seals of witnesses were attached.
The sealed tablets included the date on which the pledge was
created and sometimes an express warranty that the security was
validly created and first-ranking.⁵⁸ Obviously the use of documents
did not guarantee that the creditor could enforce his or her pledge
as first-ranking creditor. There are Digest fragments discussing
cases in which documents were lost (D. 44.2.30.1, Paul. 14 quaest.),
not dated (D. 20.1.34.1, Scaev. 27 dig.), wrongly dated (D. 48.10.28,
Mod. 4 resp.), or deliberately antedated by the parties in order to
defraud other creditors (D. 22.4.3, Paul. 3 resp.; Wacke 1969:
399–400).
⁵⁵ The main function of witnesses in the classical period may been to demarcate the
plot of land that was actually transferred (e.g. Jav., D. 18.1.63.1, Iav. 7 ex Cassio): Voss
(1982: 158–64).
⁵⁶ D. 18.1.35.8 (Gai. 10 ad ed. prov.); D. 18.1.63.1 (Iav. 7 ex Cassio). Honoré (1989:
139) observes that the traditio of provincial land was influenced by the mancipatio.
⁵⁷ On the use of chirographa, testationes, and tabulae ceratae, see Verhagen (2018:
251–62).
⁵⁸ See TPN 40, 43, 44, and 69 (TPSulp 55, 51, 52, and 79). For an express warranty
see e.g. FIRA III, nr. 91 (Mancipatio Pompeiana) and D. 20.1.34.1 (Scaev. 27 dig.).
16.4.5.4. Handing-Over Deed of Sale

An interesting fragment from the second-century jurist Scaevola
refers to a debtor’s handing over the deed of sale (pursuant to
which he had acquired the real estate) to the secured creditor
(D. 13.7.43 pr., Scaev. 5 dig.).⁵⁹ Several explanations have been offered
for this practice. It has been argued that the practice demonstrates the
difficulties in departing from the principle that only possessory
pledges could be created (Wagner 1968: 59–60). This is not very
likely, as by the time of Scaevola the pledge created by mere agree-
ment (nuda conventione) was firmly established. A more convincing
explanation is that handing over the deed of sale had always been
customary, but had never been mentioned by the jurists as they were
not particularly interested in these practical matters (Wagner 1968:
60). Scaevola’s writings, however, are an exception and show a great
interest in actual transaction practices, so that may explain why the
practice is first attested precisely in one of his opinions. There is also a
rescript by Septimius Severus and Caracalla (C. 8.16.2,  207, when
the concept of a non-possessory pledge was at least a century old)
on a case in which the deed of sale of land had been handed over to
the secured creditor. The rescript states that when someone has
pledged the deed of sale of his land he must have intended to pledge
the land itself.
The creditor’s possession of the deed of sale certainly could have
practical advantages. As the Severan rescript demonstrates it could
serve as evidence of the pledge. It could make a prospective secured
creditor or purchaser suspicious if the seller would be unable to
produce the deed of sale. Moreover, possession of the deed of sale
could enable the creditor to sell the property more easily when the
debtor was in default. Also, as D. 19.1.48 (Scaev. 2 resp.) demon-
strates, documents concerning real estate were used in order to
determine which rights the seller’s predecessor had over the property
sold: this could also be relevant for the creditor who took security
over real estate.⁶⁰ The disadvantages of this practice are that the
⁵⁹ See also Sev./Ant. C. 8.16.2 ( 207) and D. 19.1.48 (Scaev. 2 resp.). On this
practice in other times and places, see Arruñada (2012: 46–7).
⁶⁰ Titius had inherited the fundus Sempronianus from Sempronius and sold it to
Septicius. The contract of sale included the following clause: “Whatever rights Sem-
pronius had in the Sempronian farm are yours by purchase for such and such a
number of coins.” The question was asked to the jurist Scaevola whether purchaser

secured creditor could (without good reason) impede a sale and that it
could prevent multiple security interests being created over the same
property when there was sufficient “equity” in the property to do so
(Arruñada 2012: 46).
16.4.6. Reputation-Based Mechanisms

and Criminal Liability
As D. 20.1.15.2 (Gai., de form. hyp.) suggests (“the danger run by
those who create multiple charges’”), it can reasonably be assumed
that loss of credibility, civil liability, and infamia must have had a
strong deterrent effect against pledging someone else’s property or
the failure to disclose already existing charges.⁶¹ Particularly for
Romans belonging to the higher social classes, infamia could have
serious repercussions: it (inter alia) entailed the exclusion from
important offices (Krämer 2007: 367–8). For a long time the com-
bination of social norms, legal rules, and transactional practices must
have worked reasonably well to remedy the lack of publicity and of
protection of bona fide third parties. The Romans must have been
aware—in the first century —that the registry they created for
Egypt had been very beneficial for the availability of (secured) credit.
Is it too speculative to conclude from this that they would surely have
set up similar registries in the city of Rome if their own law of real
security was, because of lack of registries, so poorly adapted to the
Septicius could institute the action on purchase (actio empti) against seller Titius, in
order to force him to show what rights Sempronius had on the basis of records
relating to the inheritance and to point out the boundaries. Scaevola answered that
decisive is what the parties are considered to have intended with the clause cited
above. But if this is unclear, good faith entails that seller Titius must produce the
farm’s records (instrumenta fundi) as well as point out its boundaries. Interestingly,
Papinian observes in D. 10.1.11 (2 resp.) that in boundary disputes, when there are no
“old records” (vetera monumenta), one should follow the most recent registration by
the tax authorities (census).
⁶¹ When one reads D. 20.1.15.2 (periculum, quod solent pati qui saepius easdem res
obligant) one would think that the “danger” (periculum) referred to would have been
criminal liability (stellionatus). However, the prevailing view seems to be that stellio-
natus only became a crime at the end of the second century . Before that time the
“danger” must have been that of being sued with the actio de dolo (Kaser 1976: 183).
A condemnation on the basis of the actio de dolo would lead to infamia. See also
D. 20.1.34.1 (Scaev. 27 dig.), in which the debtor declares that the creditor must have
given credence to the debtor’s assurance “as man of honor” that the pledge was first-
ranking.

needs of the financial world? It was only by the end of the second
century  that imperial rescripts stated that the fraudulent non-
disclosure of earlier pledges should be criminally prosecuted.⁶² Know-
ingly pledging someone else’s property or property that had already
been charged constituted a special crime (crimen stellionatus).⁶³ The
maximum punishment was severe: forced labor for plebeians and
banishment or exclusion from their order for members of the higher
classes.⁶⁴ It must be assumed that this criminal liability served to
enhance security of transactions and that it took place in response to
abuses that threatened commercial intercourse (Wacke 1969: 400;
Bürge 1979: 128).
It must be doubted, however, whether these criminal sanctions
were always an effective remedy to ensure a properly functioning
law of secured credit.⁶⁵ The uncertain legal status of pledged assets,
which despite the newly introduced criminal sanction, apparently
remained a problem, is likely to have contributed in the late classical
period to the increasing practice of excluding liability for eviction in
contracts entered into by creditors enforcing their rights of pledge by
selling the pledged assets to third parties.⁶⁶ In that case the purchaser
could sue the debtor with an adapted version of the action which a
purchaser would normally have against the seller (actio empti utilis;
D. 20.5.12.1., Tryph. 8 disp.). A creditor liable for eviction could
assign to the purchaser his or her contractual action arising against
the debtor under the pledge agreement. In both scenarios the
⁶² See D. 13.7.36 pr. (Ulp. 11 ad ed.), referring to the imperial rescripts; D. 13.7.16.1
(Paul. 29 ad ed.), expressly mentioning pledging someone else’s property and pledging
property already pledged to someone else; Philip. C. 9.34.4 pr. ( 244). When a
valuable asset had been pledged for a small debt, granting a second pledge without
disclosing the first one was not a crimen stellionatus (D. 13.7.36.1, Ulp. 11 ad ed.).
⁶³ The pledgor would not be criminally liable if he or she had disclosed the earlier
pledge to the second pledgee: D. 20.1.15.2 (Gai., de form. hyp., discussed above,
p. 143).
⁶⁴ Philip. C. 9.34.4 pr. ( 244); D. 13.7.36.1 (Ulp. 11 ad ed.); D. 47.20.3.2 (Ulp. 8
de off. proc.).
⁶⁵ As Arruñada (2012: 51) observes with respect to the need to rely on criminal law:
“the need to rely on such harsh mechanisms for enforcement provides a glimpse of the
opportunity costs caused by lack of proper institutions for conveyancing land and
securing credit.”
⁶⁶ D. 21.2.68 pr. (Pap. 11 resp.); D. 17.1.59.4 (Paul. 4 resp.). As with so many
transactional practices, these clauses may have evolved into a rule of law. From Alex.
C. 8.45.1 ( 223) it appears that the secured creditor is not liable for eviction, unless
he or she has expressly agreed otherwise.

purchaser could directly sue the debtor who had pledged someone
else’s property. This would, however, be of little benefit when the
debtor was insolvent, and for this reason the possibility of eviction
may have had a depressing effect on prices paid for pledged assets
(Bürge 1979: 130–1).⁶⁷ From D. 13.7.22.4 (Ulp. 30 ad ed.) it appears
that in the late classical period it became a customary practice for
creditors selling pledged property to enter into a stipulatio duplae
with purchasers, this promising twice the amount of the purchase
price in case of eviction. This may have had a positive effect on sales
prices realized by creditors enforcing their pledge. At the same time
this transactional practice increased risks for secured creditors, which
may have resulted in increased interest rates for borrowers.⁶⁸
Criminal liability, stipulationes duplae for eviction, granting rem-
edies for eviction to the creditor or the purchaser of pledged property,
these are all symptoms of serious disruptions of the credit markets in
the late classical period resulting from the lack of publicity and of
other forms of protection against pledges created over someone else’s
property or previously charged property. More generally, it seems
that the uncertain legal status of land was becoming more and more
problematic. It is therefore not surprising that in late antiquity meas-
ures were introduced to remedy this.
16.4.7. Late Antiquity: Publicly Executed

Documents and gesta municipalia
In classical law many juridical acts (e.g. mancipatio, stipulatio)
required that certain words were to be used by the parties executing
them. These juridical acts derived their legal validity from these
formalities. Already during the course of classical law this perception
⁶⁷ Liability for eviction could be a reliable alternative for the secured creditor when
creditworthy third parties are liable, such as title insurers (Arruñada 2012: 51). I have
not carried out extensive research on this, but, although via a stipulatio or fideiussio
(or other form of personal security) such liability was technically possible, it seems
unlikely that there was an established Roman practice of title insurance.
⁶⁸ In the classical period the stipulatio duplae was broadly used in the context
of ordinary sale. In a number of fragments additional remedies are granted to the
executing pledge-creditor who entered into a stipulatio duplae: D. 13.7.8.1 (Pomp. 35
ad Sab.) (right of retention); D. 13.7.22.4 (Ulp. 30 ad ed.) and D. 13.7.23 (Tryph. 8
disp.) (right of recourse against debtor). See also D. 13.7.24 (Ulp. 30 ad ed.): the
creditor has an actio emptio utilis when evicted after impetratio dominii.

changed and the will of the parties came to be regarded as the cause of
the legal consequences of a juridical act. The consequence of this was
that the old oral formalities were replaced with new forms, which now
only served to enhance legal certainty. The new forms were mainly
(witnessed) documents and the cooperation of public authorities
(Kaser 1975: 73–4). We have already briefly looked at the first form
(written mancipatio in late classical law), and will now focus on the
second form, which eventually resulted in an imperial constitution on
the ranking of pledges.
In the late Roman period private documents were often deposited
with local archives (gesta municipalia), particularly in the western
Roman Empire. A protocol was prepared describing the essentials of
the transaction recorded in the deposited document or repeating the
text of the registered document verbatim. The deposited document
and the protocol were stored in an archive, while the parties received
copies (Hirschfeld 1904: 63–4). Direct evidence of gesta records is to
be found in the Ravenna papyri, dating from after the fall of the
western Roman Empire.⁶⁹ An extremely interesting example, which
appears to follow a procedure of (late) Roman origin (Tjäder 1955:
286), is a protocol of  489, concerning a donation by the (first)
Germanic king of Italy, Odoacer. This document gives us a detailed
picture of how title deeds were registered in late antiquity and the early
Middle Ages. The donation by Odoacer was drawn up by a notary of
the royal chancery (notarius regiae sedis) in Ravenna. The text of the
donation was recited and copied into the gesta protocol executed in
Ravenna and deposited with the gesta muncipalia of that city. Both the
notarial donation deed and the Ravenna gesta protocol were also
registered with the gesta municipalia of Syracuse (Sicily), where the
land was situated and where possession of it was transferred to the
donee.⁷⁰ Subsequently the name of the former owner was replaced
with that of the new owner (the donee) in the tax registry (a polypthicis
⁶⁹ The authoritative edition (with translations in German) of the Ravenna papyri

(P.Ital.) is Tjäder 1955 (P.Ital. 1–28) and Tjäder 1982 (P.Ital. 29–59). In one docu-
ment, a letter, reference is made to a mortgage (fiducia) of land. P.Ital. 1, Tjäder (1955:
168–78).
⁷⁰ P.Ital. 10–11, Tjäder (1955: 279–93). The donation was to a Roman supporter of
Odoacer, Pierius, who had acted as a military commander under the king. Pierius died
a year later at the battle of Adda ( 490) in which Odoacer was defeated by
Theoderic the Great. Tjäder (1955: 283–4). Another interesting example of a protocol
accompanied with a “letter of transfer” (epistula traditionis), executed in Ravenna (
540), is P.Ital. 31, Tjäder (1982: 62–71).

publicis).⁷¹ Although the core function of the gesta municipalia was
fiscal (i.e. to record the real estate and transactions relating thereto
with a view to taxation), they were also voluntarily used by private
individuals to authenticate and give some form of publicity to their
transactions.⁷² In particular, registering a deed with public authorities
would enhance the evidentiary force of the document.⁷³
From Constantinople we even have an enactment by Emperor Leo
I from  472 (C. 8.17.11) specifically dealing with the ranking of
pledges created by “publicly executed instruments” (instrumentis
publice confectis). According to this constitution “ . . . if anyone
seeks to claim for himself the right of pledge or hypothec under
such documents, the one who uses publicly executed instruments is
preferred even if there is a later date on them . . . ”⁷⁴ In other words, a
publicly executed deed of pledge takes precedence over a privately
executed one. This constitution is very important for our subject: for
the first time in the history of Roman law the “public” execution of a
document fundamentally affects the ranking of a right of pledge.
16.4.8. Comparative Perspective

16.4.8.1. Registration
From a comparative perspective it is important to note that classical
Roman law was not unique in not having title registers for immovable
⁷¹ According to Hirschfeld (1904: 21) the registration was in a property registry,

which not only listed the actual owners of the property, but also the holders of limited
rights in rem (hypothecs, usufructs, etc.). According to Tjäder (1955: 442, with further
references) the registry was a tax register, registering for each plot of land the name of
the owner, the extent of the property and the amount of taxes to be paid.
⁷² Brown (2013: 102). The functioning of the gesta municipalia in the late Roman
period is discussed at p. 95–102. See also Hirschfeld (1904: 33–49). See also Everett
(2013: 63–94), who criticizes the view that the primary purpose of the gesta was the
creation and storage of authoritative documents (2013: 72), but later discusses several
instances where the parties clearly had that purpose in mind (2013: 79, 80, and 82).
Moreover, as far as I am aware, nobody has suggested that this was the primary
purpose of Constantine and other Roman emperors for organizing gesta muncipalia.
⁷³ Kaser (1975: 80–1). This (secondary) function of the gesta muncipalia was rather
that of “recordation of documents” than that of “registration of rights.” For this
distinction, see Arruñada (2012: 55–67).
⁷⁴ Sin autem ius pignoris vel hypothecae ex huiusmodi instrumentis vindicare quis
sibi contenderit, eum qui instrumentis publice confectis nititur praeponi, etiamsi
posterior dies his contineatur. Translation Frier (2016).

property. In respect of land, houses, and other immovable property,
registration as a general institution is a relatively recent phenomenon,
even for modern countries with developed economies. The function
of informing prospective secured lenders was carried out by other
mechanisms, such as a network of local notarial archives in sixteenth-
century and seventeenth-century Paris (Hoffman, Postel-Vinay, and
Rosenthal 2000: 20–1). French, English and other modern experi-
ences show that developed credit markets can exist without registra-
tion systems and that setting-up and operating such registries is not
without difficulties.⁷⁵ As far as movable property (e.g. stocks, inven-
tory, machinery) and receivables are concerned, there are widely
divergent views as to the desirability of having public registration
systems in contemporary jurisdictions.⁷⁶ In the United States in
particular, it is regarded essential to establish a public general security
rights registry for recording notices about the possible existence of a
security right.⁷⁷ In the great majority of civil law jurisdictions (e.g.
Germany, France, Italy, Netherlands) there are no public filing sys-
tems for assets other than immovable property.⁷⁸ In any case, where
in contemporary jurisdictions a registration system for movable
property and receivables is at least technically feasible, in ancient
Rome this must have been—given the state of technology—much
more difficult to achieve. The exception was slaves, the ownership
and charge of whom could be registered in the archive of acquisitions
of Roman Egypt. However, slaves must generally have been easier to
identify with a certain estate or household than other movable assets.
All this demonstrates that it would be too simplistic to discredit the
⁷⁵ According to Arruñada (2012: 233), “governments have struggled for almost ten
centuries to organise reliable registries . . . .” Even today few countries have succeeded
in making property registries fully functional.
⁷⁶ Arruñada (2012: 84–6) does not appear to recommend registration of property
rights on movable property.
⁷⁷ Under Article 9 of the Uniform Commercial Code a security interest can be
“perfected,” so that it can be invoked against competing claimants and insolvency
officials, in several ways, such as by possession or control or by public filing.
⁷⁸ The original draft for the 1992 Dutch Civil Code did provide for a non-
possessory pledge which needed to be registered in a public register. The banking
and trade community opposed such publicly registered security interest. Under the
present Civil Code, a non-possessory pledge can be created either by notarial (or other
“authentic”) deed or by registering the deed of pledge with the Dutch tax authorities in
a non-public register (article 3:237 Civil Code). In 2018 Belgium introduced the
registration in a public register of non-possessory pledges over movable assets (title
XVII, Book III, Belgian Civil Code). See Dirix (2013).

Roman law of real security simply because it lacked registries for
security interests.
16.4.8.2. Protection of Bona Fide Secured Creditors

More remarkable is that in respect of movable property Roman law
did not develop a general principle protecting bona fide persons
against lack of title or already existing charges. However, even this
feature of Roman law can be put in perspective by a comparison with
modern law. In contemporary market economies contractual claims
are valuable assets and therefore an important category of collateral:
receivables-based financing is of immense importance for many
companies.⁷⁹ Nevertheless, as far as the protection against multiple
dispositions or unauthorized dispositions of claims is concerned, in
many (if not most) civilian jurisdictions, the situation is exactly the
same as with movable property in classical Roman law. There is
neither a public filing system for receivables nor are there rules on
the protection of bona fide purchasers or chargees of receivables.⁸⁰ In
the credit market this is no obstacle for taking receivables as collat-
eral, as there are sufficient non-legal incentives to prevent a debtor
from creating security over claims that have already been charged
(Flessner 2008: 336–49). These non-legal incentives may, at least for
many centuries, even have been stronger in ancient Rome.
16.5. CONCLUDING OBSERVATIONS
The Roman law of security satisfied nearly all the requirements for an
effective law of secured transactions. Secured creditors would in
principle be able to take recourse against the charged assets with
preference over the other creditors. Roman law offered a single and
⁷⁹ “In developed economies the bulk of corporate wealth is locked up in debts,”

Oditah (1991: vii).
⁸⁰ See, however, art. 1690 section 3 and art. 2075 of the Belgian Civil Code. For
English law, see Dearle vs. Hall (1828) 3 Russ 1. In these jurisdictions notification
fulfills a function for the protection of bona fide assignees against earlier assignments:
the assignee who has first given notice of the assignment to the debtor has priority.
Oditah considers it a very poor form of publicity to make the debtor of the assigned
claim “a kind of public register,” Oditah (1991: 140).

informal method for creating security rights rather than a multiplicity
of methods for different types of charged asset. Security could be
granted by a wide range of debtors to a wide range of creditors and
could cover a wide range of debts (including future debts). Rights of
use could be granted to the debtor (non-possessory and multiple
pledges) and to the creditor (antichresis, rehypothecation, and assign-
ment), thus preventing the collateral from remaining “dead capital.”
The rules on enforcement allowed creditors to take recourse against
the charged assets by selling them to third parties without the need to
obtain a court’s permission.
The Achilles’ heel of the Roman law of security was its lack of
publicity for security created over real estate and its lack of protection
of bona fide third parties in respect of more liquid assets. Schulz
(1936: 251) observed that “under the narrower and more easily
controllable condition of life in ancient days there was no urgent
need for the principle of publicity and the protection of a person
acquiring a thing in good faith.” In respect of security interests in
particular, Schulz writes that “hypothecary law did not, in those
times, play the same part in economic life as it does today and that
therefore the absence of protection for a creditor accepting security in
good faith was less noticeable.” Although Schulz was usually highly
perceptive of the interactions between law and economy in Rome,
his view on real security may need revision. Already in the first
century  the city of Rome itself had several hundred thousand
inhabitants. In other words, here the “narrower and more easily
controllable condition of life” no longer existed. Second, the increas-
ing sophistication of the law of real security in the classical period of
Roman law during the Principate coincided with a time of economic
growth. The law seems to have responded to the demands of the
economy, by allowing non-possessory pledges, pledges of claims, gen-
eral pledges, multiple pledges, et cetera. Roman law thus did offer a
legal framework for secured credit which would be versatile enough for
a market economy in which secured finance plays an important role.
If that is so, one wonders how likely it is that real security was as
insignificant as Schulz (and others) profess it to have been.
For a long time the Roman economy seems to have done fairly well
without registries for security interests on immovable property and
protection of bona fide creditors taking security over movable prop-
erty. The smooth operation of the Roman (secured) credit market
may only have become seriously disrupted by the end of the classical

period. It may not be a coincidence that only by the end of the
classical period the fraudulent non-disclosure of earlier pledges
came to be criminally prosecuted and that contractual practices
emerged dealing with the creditor’s liability when selling the charged
property. In other words, (only) by the end of the classical period the
system of secured credit may have become considerably less
effective.⁸¹
REFERENCES
Amunátegui Perelló, Carlos F. 2012. “Problems concerning mancipatio.” 80

Tijdschrift voor Rechtsgeschiedenis 329–52.
Ankum, Hans, and Eric Pool. 1989. “Rem in bonis meis esse and rem in bonis
meam esse: Traces of the Development of Roman Double Ownership,” in
Peter Birks, ed., New Perspectives in the Roman Law of Property. Oxford:
Oxford University Press, 5–41.
Ankum, Hans. 2010. “Spätklassische Problemfälle bezüglich der Verpfän-
dung einer res aliena,” in H. Altemeppen et al., eds, Festschrift für Rolf
Knütel zum 70. Geburtstag. Heidelberg: Müller, 35–44.
Armour, John. 2008. “The Law and Economics Debate about Secured Lending:
Lessons for European Lawmaking?,” in Horst Eidenmüller and Eva-Maria
Kieninger, eds, The Future of Secured Credit in Europe. European Company
and Financial Law Review, Special Volume 2. Berlin: De Gruyter, 3–29.
Arruñada, Benito. 2012. Institutional Foundations of Impersonal Exchange:
The Theory and Policy of Contractual Registries. Chicago: University of
Chicago Press.
Brown, Warren C. 2013. The gesta municipalia and the public validation of
documents in Frankish Europe,” in Warren C. Brown et al., eds, Docu-
mentary Culture and the Laity in the Early Middle Ages. Cambridge:
Cambridge University Press, 95–151.
Buckland, W. W. 1963. A Text-Book of Roman Law: From Augustus to
Justinian, 3rd edn, revised by Peter Stein. Cambridge: Cambridge Univer-
sity Press.
Bürge, Alfons. 1979. Retentio im römischen Sachen- und Obligationenrecht.
Zürich: Schulthess.
⁸¹ The author thanks Benito Arruñada and Giuseppe Dari-Mattiacci for their
helpful comments. Some of the descriptions of the law of pignus and fiducia in this
chapter are adaptations from parts of Verhagen (2013a, 2013b, and 2014).

Camodeca, Giuseppe. 1999. Tabulae Pompeianae Sulpiciorum, Edizione cri-
tica dell’archivio puteolano dei Sulpicii. Rome: Quasar.
Chemain, J.-F. 2015. L’argent des autres: le cautionnement dans le monde
romain du IIe siècle avant J.-C. au Ier siècle après J.-C. Versailles: Via
Romana.
Criniti, Nicola. 1991. La tabula alimentaria di Veleia: introduzione storica,
edizione critica, traduzione, indici onomastici e toponimici, bibliografia
veleiate. Parma: Deputazione di Storia Patria per le Province Parmensi.
Crook, J. A. 1996. “Legal history and general history.” 41 Bulletin of the
Institute of Classical Studies 31–6.
Deakin, Simon, and Fabio Carvalho. 2011. “System and evolution in
corporate governance,” in Peer Zumbansen and Gralf-Peter Callies, eds,
Law, Economics and Evolutionary Theory. Cheltenham/Northhampton:
Elgar, 111–30.
Dirix, Eric. 2013. De hervorming van de roerende zakelijke zekerheden.
Mechelen: Kluwer.
Everett, Nicholas. 2013. “Lay documents and archives in early medieval
Spain and Italy, c.400–700,” in Warren C. Brown et al., eds, Documentary
Culture and the Laity in the Early Middle Ages. Cambridge: Cambridge
University Press, 63–94.
Fine, John V. A. 1951. Horoi: Studies in Mortgage, Real Security, and Land
Tenure in Ancient Athens. Athens: American School of Classical Studies at
Athens.
Finley, Moses I. 1952. Studies in Land and Credit in Ancient Athens, 500–200
: The Horos-Inscriptions. New Brunswick, NJ: Rutgers University Press.
Finley, Moses I. 1953. “Land, debt, and the man of property in classical
Athens.” 68 Political Science Quarterly 249–68.
Finley, Moses I. 1999 (1973). The Ancient Economy, updated with a foreword
by Ian Morris. Berkeley: University of California Press.
Fleisig, Heywood. 2008. “The Economics of Collateral and of Collateral
Reform,” in Frederique Dahan and John Simpson, eds, Secured Transactions
Reform and Access to Credit. Northampton, MA: Edward Elgar, 81–109.
Flessner, Axel. 2008. “Security interests in receivables—A European perspec-
tive,” in Horst Eidenmüller and Eva-Maria Kieninger, eds, The Future of
Secured Credit in Europe, European Company and Financial Law Review,
Special Volume 2. Berlin: De Gruyter, 336–49.
Frier, Bruce W. 2000. “Roman law’s descent into history.” 13 Journal of
Roman Archaeology 446–8.
Frier. Bruce W. 2016. The Codex of Justinian: A New Annotated Translation,
with Parallel Latin and Greek Text, 3 vols. Cambridge: Cambridge
University Press.
Gordon, W. M., and O. F. Robinson. 1997. The Institutes of Gaius, 2nd edn.
London: Duckworth.

Gröschler, Peter. 2008. “Die Mittel der Kreditsicherung in den tabulae
ceratae,” in K. Verboven et al., eds, Pistoi dia tèn technèn: Bankers,
Loans and Archives in the Ancient World, Studies in Honour of Raymond
Bogaert. Louvain: Peeters, 301–19.
Harris, Ron. 2003. “The encounters of economic history and legal history.”
21 Law and History Review 297–346.
Hirschfeld, Bruno. 1904. Die Gesta municipalia in römischer und frühgerma-
nischer Zeit, Marburg: [no publisher] 33–49.
Hoffman, Philip T., Gilles Postel-Vinay, and Jean-Laurent Rosenthal.
2000. Priceless Markets: The Political Economy of Credit in Paris,
1660–1870. Chicago: University of Chicago Press.
Honoré. Tony. 1989. “Conveyances of land and professional standards in the
later Empire,” in Peter Birks, ed., New Perspectives in the Roman Law of
Property. Oxford: Oxford University Press, 137–52.
Jackson, Thomas H., and Anthony T. Kronman. 1979. “Secured financing
and priorities.” 88 Yale Law Review 1143–82.
University Press.
Jolowicz, H. F., and Barry Nicholas. 1972. Historical Introduction to the Study
of Roman Law, 3rd edn. Cambridge: Cambridge University Press.
Kaser, Max. 1971. Das römische Privatrecht, Erster Abschnitt: Das
altrömische, das vorklassische und klassische Recht, 2nd edn. Munich:
Beck.
Kaser, Max. 1975. Das römische Privatrecht, Zweiter Abschnitt: Die nachk-
lassischen Entwicklungen, 2nd edn. Munich: Beck.
Kaser, Max. 1976. Ausgewählte Schriften, vol. II. Napels: Jovene.
Kaser, Max. 1982. Studien zum römischen Pfandrecht. Napels: Jovene.
Kaser, Max, and Karl Hackl. 1996. Das römische Zivilprozessrecht, 2nd edn.
Munich: Beck.
Kaser, Max, Rolf Knütel, and Sebastian Lohsse. 2017. Römisches Privatrecht:
Ein Studienbuch, 2nd edn. Munich: Beck.
Kay, Philip. 2014. Rome’s Economic Revolution. Oxford: Oxford University
Press.
Krämer, Gerd. 2007. Das besitzlose Pfandrecht: Entwicklungen in der
römischen Republik und im frühen Prinzipat. Cologne/Vienna: Böhlau.
Kunkel, Wolfgang. 1973. “Hypothesen zur Geschichte des römischen Pfan-
drechts.” 90 Zeitschrift der Savigny-Stiftung für Rechtsgeschichte: Roma-
nistische Abteilung 15–170.
Lerouxel, François. 2012. “Le marché du crédit privé, la bibliothèque des
acquêts et les tâches publiques en Égypte romaine.” 67 Annales HSS
943–76.

Lerouxel, François. 2015. “The βιβλιοθήκη ἐγκτήσεων and transaction costs
in the credit market of Roman Egypt (30 ... to ca. 170 ..),” in Dennis P.
Kehoe, David M. Ratzan, and Uri Yiftach, eds, Law and Transaction Costs
in the Ancient Economy. Ann Arbor: University of Michigan Press,
162–84.
Luhmann, Niklas. 2004 (1993). Law as a Social System. Oxford: Oxford
University Press (orig.: Das Recht der Gesellschaft, Frankurt am Main:
Suhrkamp).
McCall, Brian. 2009. “It’s just secured credit: a natural law case in support of
some forms of secured credit.” 43 Indiana Law Review 7–44.
McCormack, Gerard. 2004. Secured Credit under English and American Law.
Cambridge: Cambridge University Press.
Mentxaka, Rosa. 1986. La pigneracion de colectividades en el derecho romano
clasico. Bilbao: Servicio Editorial de la Universidad del País Vasco.
Moatti, Claude. 1993. Archives et partage de la terre dans le monde Romain
(IIe siècle avant—Ier siècle après J.-C.). Rome: École française de Rome.
Noordraven, Bert. 1999. Die Fiduzia im römischen Recht. Amsterdam:
Gieben.
Nicholas, Barry. 1962 [repr. 2010]. An Introduction to Roman Law. Oxford:
Oxford University Press.
North, Douglass C. 2010. Understanding the Process of Economic Change.
Princeton/Oxford: Princeton University Press.
Oditah, Fidelis. 1991. Legal Aspects of Receivables Financing. London: Sweet
& Maxwell.
Roe, Mark J. 1996. “Chaos and evolution in law and economics.” 109
Harvard Law Review 641–68.
Romeo, Stefania. 2006. “Fiducia auctionibus vendunda nelle Tabelle Pom-
peiane. Procedure e modalità di redazione delle testationes nelle avctiones
puteolane del 61 d.C.” Polis. Studi interdisciplinari sul mondo antico 207–58.
Röver, Jan-Hendrik. 1999. Vergleichende Prinzipien dinglicher Sicherheiten.
Munich: Beck.
Röver, Jan-Hendrik. 2007. Secured Lending in Eastern Europe: Comparative
Law of Secured Transactions and the EBRD Model Law. Oxford: Oxford
University Press.
Schäfer, Hans-Bernd. 2008. “The law and economics debate about secured
lending: lessons for European lawmaking. Commentary?,” in Horst Eiden-
müller and Eva-Maria Kieninger, eds, The Future of Secured Credit in
Europe. European Company and Financial Law Review, Special Volume 2.
Berlin: De Gruyter, 30–5.
Scheidel, Walter. 2010. “Economy and quality of life,” in Alessandro Barch-
iesi and Walter Scheidel, eds, The Oxford Handbook of Roman Studies.
Oxford: Oxford University Press, 593–609.
Schulz, Fritz. 1936. Principles of Roman Law. Oxford: Clarendon Press.
Schulz, Fritz. 1951 [repr. 1969]. Classical Roman Law. Oxford: Clarendon Press.

Schwartz, Alan. 1984. “The continuing puzzle of secured debt.” 37 Vanderbilt
Law Review, 1051–69.
Schiavone, Aldo. 2000 (1996). The End of the Past: Ancient Rome and the
Modern West. Cambridge, MA/London: Harvard University Press (orig.:
La storia spezzata: Rome antica e Occidente modern. Rome/Bari: Laterza).
Stiglitz, Joseph E., and Andrew Weiss. 1981. “Credit rationing in markets
with imperfect information.” 71 American Economic Review 393–410.
Sturm, Fritz. 1993. “Das Absterben der mancipatio,” in Stephan Buchholz
et al., eds, Überlieferung, Bewahrung und Gestaltung in der rechtsgeschich-
tlichen Forschung. Paderborn (etc.): Schöningh.
Taubenschlag, Raphael. 1955. The Law of Greco-Roman Egypt in the Light of
the Papyri. 332 –640 , 2nd edn. Warsaw: PNW.
Terpstra, Taco. 2008. “Roman law, transaction costs and the Roman econ-
omy,” in Koenraad Verboven et al., eds, Pistoi dia tèn technèn: Bankers,
Loans and Archives in the Ancient World, Studies in Honour of Raymond
Bogaert. Louvain: Peeters, 345–69.
Terpstra, Taco. 2013. Trading Communities in the Roman World, A Micro-
Economic and Institutional Perspective. Leiden/Boston.
Tjäder, Jan Olof. 1955. Die nichtliterarischen lateinischen Papyri Italiens aus
der Zeit 445–700, I. Papyri 1–28. Acta Instituti Romani Regni Sueciae,
Series in 4o, XIX:1. Lund: Gleerup.
Tjäder, Jan Olof. 1982. Die nichtliterarischen lateinischen Papyri Italiens aus
der Zeit 445–700, II. Papyri 29–59. Acta Instituti Romani Regni Sueciae,
Series in 4o, XIX:2. Stockholm: Paul Åström.
Van Gessel, Christian. 2003. “Praedes, praedia, cognitores: les suretés
réelles et personelles de l’adjudicataire du contrat public en droit romain
(textes et réflections),” in Jean-Jacques Aubert, ed., Tâches publiques et
entreprise privée dans le monde romain. Actes du Diplôme d’Etudes
Avancées, Universités de Neuchâtel et de Lausanne, 2000–2002. Geneve:
Droz, 95–122.
Verboven, Koen. 2002. The Economy of Friends: Economic Aspects of Amici-
tia and Patronage in the Late Republic. Brussels: Latomus.
Verhagen, Hendrik L. E. 2011. “Das Verfallpfand im frühklassischen
römischen Recht.” 79 Tijdschrift voor Rechtsgeschiedenis 1–46.
Verhagen, Hendrik L. E. 2013a. “The evolution of pignus in classical Roman
law.” 81 Tijdschrift voor rechtsgeschiedenis 51–79.
Verhagen, Hendrik L. E. 2013b. “Ius honorarium, equity and real security:
parallel lines of legal development,” in W. J. Zwalve and E. Koops, eds,
Law and Equity: Approaches in Roman Law and Common Law. Den Haag/
Boston: Martinus Nijhoff, 129–60.
Verhagen, Hendrik L. E. 2014. “The (lack of) protection of bona fide pledgees
in classical Roman law,” in Rena van den Bergh et al., eds, Meditationes de
iure et historia: Essays in Honour of Laurens Winkel. Fundamina. Editio
specialis 2014 (20–2). Pretoria: Unisa Press, 982–92.

Verhagen, Hendrik L. E. 2014. “Chirographs in classical Roman law: consti-
tutive or probative?” 65 Revue Internationale des Droits de l’Antiquité
251–306.
Voss, Wulf Eckart. 1982. Recht und Rhetorik in den Kaisergesetzen der
Spätantike: Eine Untersuchung zum nachklassischen Kauf- und Über-
eignungsrecht. Forschungen zur Byzantinischen Rechtsgeschichte 9.
Frankfurt am Main: Löwenklau.
Wacke, Andreas. 1969. “Prozessformel und Beweislast im Pfandrechtsprä-
tendentenstreit.” 37 Tijdschrift voor Rechtsgeschiedenis 369–414.
Wagner, Herbert. 1968. Voraussetzungen, Vorstufen und Anfänge der
römischen Generalverpfändung. Marburg: Elwert.
Wagner, Herbert. 1974. Die Entwicklung der Legalhypotheken am Schuldner-
vermögen im römischen Zeit (bis zur Zeit Diokletians). Cologne/Vienna:
Böhlau.
Watson, Alan. 2001. Society and Legal Change, 2nd edn. Philadelphia:
Temple University Press.
Wenger, Leopold. 1953. Die Quellen des römischen Rechts. Vienna:
Holzhausen.
Wieacker, Franz. 1988. Römische Rechtsgeschichte. Quellenkunde, Rechtsbil-
dung, Jurisprudenz und Rechtsliteratur, Erster Abschnitt. Munich: Beck.
Woess, Friedrich von. 1924. Untersuchungen über das Urkundenwesen und
den Publizitätsschutz im römischen Ägypten. Munich: Beck.
Wolf, Joseph Georg. 2012. Neue Rechtsurkunden aus Pompeji, Tabulae Pom-
peianae Novae, 2nd edn. Darmstadt: Wissenschaftlichte Buchgesellschaft.
Wolff, Hans Julius. 1978. Das Recht der griechischen Papyri Ägyptens in der
Zeit der Ptolemaeer und des Prinzipat, Bd. 2: Organisation und Kontrolle
des privaten Rechtsverkehrs. Munich: Beck.
Wubbe, Felix B. J. 1960. Res aliena pignori data: de verpanding van ander-
mans zaak in het klassieke Romeinse recht. Leiden: Universitaire Pers.
Zimmermann, Reinhard. 1996 [1990]. The Law of Obligations: Roman Foun-
dations of the Civilian Tradition. Oxford: Oxford University Press.
Zumbansen, Peer, and Gralf-Peter Callies, eds. 2011. Law, Economics and
Evolutionary Theory. Cheltenham/Northhampton: Elgar.
17
Ancient Rome
Legal Foundations of the Growth
of an Indispensable City
Robert C. Ellickson
“The ancients themselves were firm in their view that civilized

life was thinkable only in and because of cities.”
M. I. Finley (1977: 305)
“For the upper-class Roman, it would seem, the provincial towns
did not exist: Rome’s prestige held them, as London and Paris hold
similar groups today. To live well, he must dwell in Rome . . . .”
Lewis Mumford (1961: 210)
In 200 , the city of Rome had a population of about 200,000

(Morley 1996: 39). Two hundred years later, the city’s population
had leapt to roughly 1,000,000, the largest urban agglomeration the
world had yet seen.¹ Almost two millennia would elapse before
Europe would again witness equivalent urban expansions, those of
London and Paris around 1800.
¹ Temin (2013: 31) summarizes the varying estimates of ancient Rome’s popula-
tion. At the time of the early Empire, about one-seventh of the peninsula of Italy’s
population resided in Rome (ibid.: 259). During civil wars and other periods of socio-
political instability, the rise in the city’s population may have paused or reversed
(Turchin and Scheidel 2009). In  1500, Rome’s population numbered roughly
50,000 (Jongman 1988: 73). At that time, the combined populations of the twelve
largest cities of Christian Europe totaled about 1,000,000, the equivalent of ancient
Rome’s population at its peak (Scheidel 2007: 78).
Robert C. Ellickson, Ancient Rome: Legal Foundations of the Growth of an Indispensable City In: Roman Law and
DOI: 10.1093/oso/9780198787211.003.0017
160 Robert C. Ellickson

The explosive growth of Rome during and after the late Republic
poses two major questions. The first is whether the city’s emergence
as a major metropolis benefited other residents of the Empire, or
instead, as Moses Finley (1999) and others have contended, was
parasitic on the resources of outsiders. I contribute to this long-
standing debate by applying the theory of cities that Edward Glaeser
(1998) and other urban economists have developed. This body of
theory helps identify the many sorts of agglomeration benefits that
ancient Rome provided to persons residing outside the city. I include
an analysis of the origins and career destinations of forty-one
renowned Romans, among them poets, historians, and jurists. This
small dataset strongly supports the thesis that many of Rome’s most
talented inhabitants were born elsewhere and migrated to the city.
They saw the immensity of Rome as advantageous because it enabled
them to specialize and to exchange ideas with peers. Interactions
among these talented people generated information spillovers that
benefited not only themselves, but also outsiders who consumed the
many services that the capital exported. Far from being parasitic, the
city of Rome in its prime contributed mightily to the relative pros-
perity of the Mediterranean region.²
My second central undertaking is to identify the institutional
underpinnings that enabled the city of Rome to grow as rapidly as
it did. Land was the principal form of ancient wealth (Crook 1967:
147; Finley 1999: 188). I contend that lawmakers in both the Republic
and early Empire generally resolved key issues affecting urban devel-
opment more sagaciously than had lawmakers of other notable
regimes in the ancient world. Like most of the prominent states
that predated it, Roman lawmakers authorized private property in
land. The Romans, however, were exceptionally willing to permit
a landowner to alienate real estate by sale, lease, or testamentary
disposition.³ For example, the Twelve Tables, legislation adopted
² Support for the geographic breadth of prosperity in the Roman world can be
found in Engels (1990: 125–6); Ward-Perkins (2005: 87–100); Lo Cascio (2007:
619–22); and Temin (2013: 220–61). Benefits hardly were universal. During their
conquests, the Romans killed and enslaved many.
³ The phrase real estate challenges Finley’s (1999) vision that transactions in the
ancient world differed in kind from those in a mature commercial nation. Ancient
Rome unquestionably lacked many of the institutions present in a twenty-first-
century market economy, including specialized real estate brokers and large-scale
mortgage lenders. See n. 63. But market transactions were pervasive (Temin 2013).
Legal Foundations of the Growth of an Indispensable City 161

soon after the founding of the Roman Republic, affirmed unfettered
freedom of testation, an extraordinary legal principle in a near-
archaic society. The Romans’ willingness to put land tenure in motion
helped unleash the energies of the host of private entrepreneurs who
developed and managed the apartment blocks of the expanding city.
The decisions that determined the scope of the public sector in the
city of Rome also were relatively sound. Public officials provided
public works such as forums and aqueducts that were essential to
Rome’s growth. But these officials also limited the scope of govern-
ment in ways that enhanced the city’s prosperity. They declined to
provide private goods, such as housing, that the city’s entrepreneurs
could more efficiently supply, and learned to resist political pressures
for growth-inhibiting populist enactments, such as rent controls and
edicts cancelling debts.
I do not contend that the legal policies just identified were suffi-
cient to assure the rise of the city of Rome. They surely were not. But
in the absence of this cluster of enabling policies, Rome would have
remained a backwater.
17.1. THE CITYSCAPE OF ANCIENT ROME
Especially for the benefit of non-classicists, I begin with a brief

description of Rome’s demography and landscape during its prime,
defined here as the conveniently symmetric period between 200 
and  200. During that stretch, Rome was, for better or worse, a city
in a class by itself. Zipf ’s Law, an intriguing nugget of urban theory,
predicts that the population of a nation’s largest city will be roughly
twice the population of its second largest city, three times the popu-
lation of its third largest, and so on (Gabaix 1999). Around  150,
the populations of the cities throughout the Roman empire com-
ported well with Zipf ’s Law. After Rome, a city with about 1,000,000
inhabitants, came Alexandria, with about 500,000, and then Carthage,
with 300,000 (Wilson 2011: 181–6). But within the boundaries of
present-day mainland Italy, the second most populous city in  150
A “capitalistic” economy can be defined as one in which decentralized actors routinely

exchange goods and services for money. By this definition, Rome’s land and housing
markets were capitalistic.

is thought to have been Ostia, Rome’s seaport. Ostia’s population
then was a mere 40,000, a twenty-fifth of Rome’s (Wilson 2011: 182).⁴
On the peninsula of Italy itself, the city of Rome dwarfed all rivals.⁵
17.1.1. Private Buildings and the Entrepreneurs

Who Built and Owned Them
During the late Republic and early Empire, curule and plebeian
aediles and other public officials supervised the provision of public
works such as the streets and aqueducts essential to Rome’s growth.
There is no indication, however, that officials of either the late
Republic or early Empire ever sponsored the building of housing
projects (Frier 1980: 25; Robinson 1992: 4; see also Frier 1977: 36).⁶
Private entrepreneurs, whose names are mostly lost in history, pro-
duced and managed the apartment blocks, lodging houses, and shops
that made up the bulk of the city’s buildings.
The best evidence about Roman housing comes from archaeo-
logical remains in Ostia and the Vesuvian cities (Pirson 1997). Apart-
ment blocks built in Ostia during the first century , in particular,
are thought to have been patterned after structures in Rome (Packer
1971: 63; Garnsey 1976: 129; Frier 1980: 3–6, 14). During the early
Empire, the typical private housing development in Rome, at least for
upscale tenants,⁷ was an apartment block (insula)⁸ on the order of
four to six stories in height.⁹ An apartment block commonly included
shops (tabernae) on the ground floor fronting on the street, and
apartments (cenacula) on the floors above.
⁴ Morley (1996: 182) provides a somewhat different set of estimates of the popu-
lations of ancient cities on the Italian peninsula.
⁵ When analyzing the growth of urban giants, Ades and Glaeser (1995: 216–18)
selected ancient Rome as one of their case studies.
⁶ By forfeit or escheat, however, a domus or insula might fall into government
ownership (Frier 1980: 25).
⁷ On the existence and size of a Roman “middle class,” see Mayer (2012). Perhaps 6
to 12 percent of the Roman Empire’s population could be so categorized (ibid.: 21).
For more discussion, see Scheidel and Friesen (2009).
⁸ The proper translation of insula is disputed (Storey 2002). I adopt the interpret-
ation in Frier (1980). Dubouloz (2011), reviewed by Machado (2012), unconvention-
ally defines an insula not as a type of building but as a property that was owned
indirectly.
⁹ Frier (1980: 14–16) describes the Casa di Via Giulio Romano, the insula in Rome
that has best survived.

Over the centuries, improvements in building technology raised
the average quality of Rome’s apartment blocks. Among these
advances were the refinements of Roman concrete (opus caementi-
cium) during the late Republic, and fired brick during the first two
centuries of the Empire (Schneider 2007: 162–3). These materials
enabled construction of taller, stronger, and more fire-resistant apart-
ment blocks, commonly ones graced with arches and vaults.¹⁰ Within
an apartment, however, rafters and other structural elements com-
monly were of timber (Robinson 1992: 34–5). Textual sources indi-
cate that tenants in Rome’s apartment blocks were frequently
endangered by fire and collapse (Crook 1967: 154).
An apartment on an intermediate floor of an upscale insula might
have an area of 150 to 300 square meters (Frier 1980: 6). This typically
would have accommodated a household of six to eight occupants,
counting slaves¹¹ and other hangers-on (Frier 1980: 16). Upper
apartments were reached by way of staircases, commonly external
(Pirson 1997: 177–8). An apartment on a lower floor might be
serviced by gravity-fed piped water (Robinson 1992: 99, 119–20; see
also Beard (2008: 64), on Pompeii), and perhaps a toilet seat situated
over a chamber pot or pipe into a wall (Beard 2008: 107).
Most of Rome’s relatively poor households—perhaps as much as
90 percent of a city’s free population (Frier 1980: 5)—found housing
in one of three ways.¹² As noted, the ground floor of an insula
fronting a well-traveled street included specialized shops.¹³ The arti-
sans and traders—commonly slaves or ex-slaves—who worked in
these tiny shops commonly slept in a back room or on an elevated
mezzanine (Pirson 1997: 173).
¹⁰ Vitruvius estimated that a concrete tenement house in Cicero’s day would have
had a lifespan of eighty years, an expected duration shorter than that of a brick
structure (Garnsey 1976: 129).
¹¹ The percentage of the population enslaved likely was higher in Rome than
elsewhere, and perhaps peaked at the outset of the Empire at between 25 percent
and 40 percent. For estimates, see Morley (1996: 39); Scheidel (2005); and Temin
(2013: 135–6). See generally Koops in this volume.
¹² A sub-middle-class household was far less likely to include slaves and ex-slaves,
and might have had five to six occupants on average (Bowman and Wilson 2011b: 12).
On the residences of the poor of Pompeii, see Beard (2008: 105–6). Wallace-Hadrill
(1994: 75–9) reports that the distribution of dwelling sizes in Pompeii was broad and
not bimodal.
¹³ Holleran (2012: 99–158) describes the various services that these shops offered.

Second, a landlord of an apartment block whose lower floors were
designed for prosperous tenants might subdivide the topmost stories
into small rooms or cubicles that poor households could afford.¹⁴
These quarters, although perhaps enjoying better light and view,
lacked piped water and required a long climb. Residents of upper
floors also typically bore the greatest risks of both fumes wafting and
fires spreading from the apartments below. Because landlords tended
to market apartments on upper floors to the less wealthy, in Rome
there commonly was a vertical sorting of tenants by income.¹⁵
Households in Rome also were horizontally sorted (Yavetz 1958:
505–6).¹⁶ The poorest neighborhoods contained disproportionate
concentrations of tenements and lodging houses (known as dever-
soria, among other terms). The owner of a lodging house offered
short-term rentals of squalid quarters and perhaps also meals in a
common room (Frier 1977: 31–5; Frier 1980: 27–9, 51; Scobie 1986:
402). Although archaeological evidence is sparse, some lodging
houses may have been made with cheaper materials such as mud
brick and wood (Pirson 1997: 95). Public latrines provided an essen-
tial service.
The wealthiest Romans, less than 1 percent of the population,
sought the distinction of owning and residing in a domus, a single-
household abode that typically was spacious, luxuriously appointed,
and situated on a large lot.¹⁷ A domus household might include
dozens, or even hundreds, of slaves, ex-slaves, and other hangers-on
(Saller 2007: 107). During the late Republic, some members of the
upper crust clustered their houses on Palatine Hill, close to the
jewelers, goldsmiths, and pearl dealers on Via Sacra (Engels 1990:
243; Holleran 2012: 55).
Rome’s neighborhoods likely were less stratified by social class than
those of a twenty-first-century city. Because the richest households
¹⁴ An apartment builder indeed might use wooden materials, not masonry, to

frame the top floor or two (Robinson 1992: 37).
¹⁵ Compare the vertical stratification characteristic of the luxurious Hausmannian
apartment houses of late nineteenth-century Paris, in which tiny rooms for chamber-
maids were provided on the top floor.
¹⁶ Frier (1980: 35) refers to Argiletum and Aventine as “lower-class districts”
of Rome.
¹⁷ Treggiari (1999) depicts Cicero’s emotional attachments to his domus. A census
of Rome at the end of the third century  counted 1,790 domus structures and 44,300
insulae (Benevolo 1980: 176).

included numerous slaves and ex-slaves, pedestrians on Via Sacra
likely would have been socially diverse. And residents of upscale
apartments commonly lived in close proximity to the ground-floor
shops that housed low-status artisans and traders.¹⁸
Little is known about the entrepreneurs, designers, contractors,
financiers, and construction workers who together erected Rome’s
private buildings.¹⁹ Because senators tended to regard conspicuous
involvement in real estate development as undignified (Finley 1999:
41–2, 60–1), most subdividers and apartment builders likely were
drawn from a somewhat lower social rank, perhaps relatively wealthy
freedmen. Cicero’s letters mention two real estate operators,
C. Sergius Orata and Damasippus, the latter a subdivider of land
into lots (Rawson 1976: 100–2).²⁰ Plutarch’s Parallel Lives includes a
depiction of M. Licinius Crassus, an exceptionally rich Roman famed
for his military, political, and real estate exploits. Crassus’ adulthood
spanned the first half of the first century . Writing in Greece almost
two centuries later, Plutarch asserted that Crassus built nothing
except his own house, but eventually acquired “the largest part of
Rome” from sellers panicked by fires breaking out in their apartment
blocks.²¹ Plutarch surely exaggerated. Even the wealthiest Roman
lacked the resources to buy a major fraction of the private real estate
in the city, and had no feasible means of managing thousands of
apartment blocks. There were no large business enterprises in
Rome (Crook 1967: 229), partly because Roman law provided no
workable mechanism for the creation of a private business entity
that would have been attractive to capital investors (Fleckner in
volume I; Hansmann et al. in volume I). Ownership of Rome’s
many thousands of apartment blocks likely was distributed widely
among individuals and small enterprises. The wealthiest residents of
Rome tended to invest not only in rural property, but also in urban
real estate (Frier 1980: 21–6; Parkins 1997; Rosenstein 2008). Cicero,
¹⁸ On the paucity of horizontal stratification in Pompeii, a much smaller city, see

Wallace-Hadrill (1991: 261–4); Beard (2008: 62); and Mayer (2012: 51–8).
¹⁹ Kehoe (2012: 123–4) provides a brief discussion of builders and construction
workers, and Hawkins (2012: 180–1), of specialty subcontractors. A well-established
profession of land surveyors existed; see Campbell 2000; and Libecap and Lueck in
this volume, but most of the textual evidence depicts their activities in rural areas, not
cities.
²⁰ Rawson draws on Cicero, Ad Atticum 12.33.
²¹ Plut., Crass. 2.4–2.5; for discussion, see Frier (1980: 32–4).

whom Finley (1999: 41–2) cites as contemptuous of urban commerce,
himself owned several apartment blocks (Garnsey 1976: 126–7).
Many housing complexes would have been in the hands of either
individuals who had inherited their ownership interests, or guardians
managing inheritances for wards.
Patronage networks supported by gift exchange pervaded many
aspects of Roman society.²² These helped foster trust among long-
distance traders, for example. The owner of an apartment block
similarly could take advantage of a patron-client relationship when
managing the building. The owner might rely on a former slave, for
instance, to handle relations with tenants.²³ A patron-builder simi-
larly might prefer to choose a client, rather than a more socially
distant provider, as the architect or contractor. In a dense city such
as Rome, however, the reputational concerns of a repeat player would
have partially allayed a builder’s worries about the trustworthiness of
an agent who was not a client.
17.1.2. Density
During the reign of Augustus, Rome is estimated to have had about
five hundred residents per hectare (Morley 1996: 38). This was
roughly twice the density of Manhattan’s population in 2010. In
Rome, both people and buildings were packed together. Streets were
narrow. By the early Empire, most of Rome’s streets had a width of
about 4.5–5.0 meters between building lines (Chevallier 1976: 67).²⁴
An apartment block typically abutted its sidewalk and covered most
of its lot. In most neighborhoods, narrow streets lined with shop-
fronts threaded through masses of hulking apartment blocks bursting
with humanity. Pedestrians had to share a street right-of-way with,
among others, peddlers of goods, guiders of pack-animals, riders on
horseback, and drovers of carts drawn by oxen (Finley 1999: 126).
A staple of urban economic theory holds that land values tend to
decline with distance from a city’s center (Alonso 1964).²⁵ As this
²² See nn. 35–7 and accompanying text.

²³ See text accompanying nn. 99–100.
²⁴ Beard (2008: 58) states that street pavements in Pompeii typically were less than
3 m wide.
²⁵ Frier (1980: 23 n. 7) marshals some of the few extant sources on land prices in
ancient Rome.

theory predicts, rents in central Rome seem to have been astro-
nomic.²⁶ The price of agricultural land also was roughly ten times
higher near the city than elsewhere in Italy (Engels 1990: 25–6),
further evidence of the value of proximity to the capital.
Rome’s epicenter was the venerable Forum, which evolved into a
collection of adjoining forums. Storey’s (2002: 420–1) study of
fourth-century  census data suggests that residential densities
were highest near the Forum and the adjoining Palatine Hill. During
the period 200  to  200, Rome had a single defensive perimeter,
the Servian Wall, which dated from the fourth century . A Roman
living inside that wall would have resided no more than two kilo-
meters from the Forum, a half-hour’s walk at most. The territory
within roughly a day’s walk (30–7 km) of the Forum comprised
Rome’s suburban district (Goodman 2007: 20, 50). Partly because
the Servian Wall, despite its sixteen gates, impeded access, gradients
of both land value and population density likely dropped markedly
beyond that wall. Nonetheless, the decision to build the Aurelean
Wall in  270–5 at a farther-outlying location suggests that the area
just beyond the Servian wall had already become heavily urbanized
with apartment structures.²⁷
17.1.3. Unpleasant Aspects of Life in Rome

Life in the capital city was exasperating and unhealthful in many
respects. Scobie (1986) provides a particularly dystopian assess-
ment.²⁸ Roman literary figures, most notably Juvenal in Satura III
(The Sixteen Satires 1974: 87–98), stress the downsides of city life and
the superiority of rustic environments. Juvenal’s mouthpiece is
Umbricius, a man who had failed to prosper in Rome and who
rants about negative aspects of the city. Some current-day interpret-
ers, overlooking Juvenal’s satirical approach, have been too quick to
take Umbricius’ laments literally (Braund 1989).
Even allowing for Umbricius’ exaggerations, the environment in
Rome was indeed challenging in many respects. For starters, it was a
²⁶ Frier (1980: 43) refers to the “enormously high rents at Rome,” an assertion
consistent with Juvenal (1974: 93).
²⁷ Angel (2012: 93) provides a map that demarcates both walls.
²⁸ Laurence (1997) argues that Scobie exaggerates.

noisy place.²⁹ Late into the evening, from the streets echoed shouts of
draymen (Juvenal 1974: 95) and the clatter of iron-clad wheels of
carts on stone pavements (Beard 2008: 80). Because acoustical insu-
lation within and between apartment blocks was mediocre, a would-
be sleeper might have to put up with the thumps of footsteps, the cries
of babies, and the barking of the ubiquitous pet dogs.³⁰
Rome also was a smelly city. Latrines and other systems for
disposal of excrement and urine, both human and animal, were
primitive at best (Scobie 1986: 407–22; Robinson 1992: 119–22).
A pedestrian risked being splattered by a dump of household waste,
perhaps from a chamber pot, from a window above (Scobie 1986: 417;
Beard 2008: 55).³¹ These putrid materials, when added to the wastes
of draft animals, could make a pedestrian’s walk along a street an
unpleasant trek (Beard 2008: 56). The remains of Pompeii suggest
that Rome’s aediles might have arranged for installation of stepping
stones in streets to provide a path above the slop (Beard 2008: 54).
Braziers, oil lamps, and kitchen fires, all commonplace in apartments
(ibid. 92), emitted fumes and posed risks of fire. Rome suffered as
many as one hundred building fires per day, twenty of them large
(Robinson 1992: 108)
Worst of all, Rome was a sickly environment, especially for
young children. Because infectious diseases were rampant, life expect-
ancies were shorter in Rome and other cities than in the countryside
(Scheidel 2007: 39). Given all these distasteful aspects of life in Rome,
how could the city’s population growth have been so explosive after
200 ? Why would the head of a free household have ever chosen to
migrate to the capital?
²⁹ Martial, Epigrammata, Book 12, poem 57, proclaims the impossibility of a good
night’s sleep in Rome. In Epistles 56, Seneca complains of noise from a bathing
establishment.
³⁰ Even poor urban households commonly kept dogs as pets (Scobie 1986:
418–19).
³¹ A dump of wastes could give rise to legal liability. “The praetor says . . . ‘If
anything should be thrown out or poured out from a building onto a place where
people commonly pass and repass or stand about, I will grant an action to be brought
against whoever lives there for double the damage caused or done as a result’ ” D. 9.3.1
pr. (Ulp. 23 ad ed.). See also D. 9.3.1.10 (Ulp. 23 ad ed.): “If a number of people occupy
a lodging house and something is thrown down from it, action may be brought against
any one of them.” (Throughout this chapter, references to the Justinian Digest are to
the Watson ed. (1985) version.)

17.2. WHY TALENTED PEOPLE MOVED TO ROME,
AND HOW THEIR PRESENCE THERE
BENEFITED OUTSIDERS
To individuals and households who were free to decide where to

live—as slaves obviously were not—Rome offered countervailing
benefits that no other city in the eventual western empire could
match. This section identifies these benefits, and, more controver-
sially, argues that the city of Rome also provided valuable services to
people throughout the empire. In spirit, my analysis echoes both
Donald Engels’s (1990) assessment of ancient Corinth’s role as a
service city, and many of the essays in Bowman and Wilson (2011).
But, unlike the authors of those works, I explicitly draw on the theory
of cities that Glaeser and other urban economists have refined since
the 1990s. This theory seeks to explain why people are willing to put
up with the high costs and hassles of city life, and why, throughout
history, cities have tended to be centers of innovation, as ancient
Rome surely was.
17.2.1. The Once Fashionable Conception

of Rome as a Consumer City
In several works first published in the 1970s, the deservedly eminent
classicist Moses Finley (1977, 1999) depicted Rome as a “consumer
city”—a “parasite” of residents of rural areas.³² Finley’s conception of
the Roman economy is commonly characterized as “primitivist,” a
position contrasted with the “modernist” views of his critics. Three of
the premises underlying Finley’s analysis are uncontroversial. The
first is the truism that, on balance, a city inhabited by people owning a
disproportionate share of their polity’s financial assets and real estate
will be a net importer of foodstuffs and manufactured goods. The
phrase consumer city captures this truism.³³ Second, no one disputes
the fact that rich Romans disproportionately lived in Rome. The
members of the senatorial class, the wealthiest of all, were centered
³² Max Weber and Karl Polanyi influenced Finley’s thinking. Finley applied his
thesis to ancient cities generally, not just to Rome.
³³ Glaeser (2011: 259–60), in a book that cites neither Finley nor Weber, describes
contemporary London, New York and Paris as “consumer cities”—that is, places that
offer services and entertainments that attract the wealthy.

in the capital, and indeed the senate was required to convene there.
Third, it is uncontested that the city of Rome lacked large-scale
manufacturing facilities (Finley 1999: 23). The city thus could hardly
have counterbalanced its massive imports of food solely with exports
of manufactured wares.³⁴
Much more controversial, however, has been Finley’s (1999:
35–61) assertion of a profound difference between the motivations
of ancient and modern traders. He admitted that transfers of land and
houses were “relatively frequent” during the time of Cicero (Finley
1976b: 4).³⁵ But he contended that parties negotiating these transac-
tions, especially when both were members of the social elite, would
have been more concerned about attaining honor and status than
material wealth. A land sale or apartment lease between two wealthy
parties thus would not have been an arm’s length bargain about the
division of the monetary gains from trade, but rather an episode in an
ongoing process of gift exchange embedded within a larger social
network.³⁶ To Finley, the ancient economy was qualitatively different,
not just quantitatively different, from a modern economy (Morris
1999: introduction).
Many of Finley’s critics (e.g. Wallace-Hadrill 1991; Temin 2013:
4–11) reject this vision.³⁷ Finley himself was well aware that members
of Rome’s upper crust frequently acted materialistically (Jongman
1988: 34–5). Land typically was sold for cash and at arm’s length
(Frederiksen 1975: 168). In all known urban leases, the rent due was
stated in money (Frier 1980: 60). Conversely, gift exchange is hardly
absent in a current-day market economy. On the order of one-third of
US GDP currently is produced within households, much of it through
non-market processes (Ellickson 2006: 231–2). And inhabitants of a
³⁴ There is some evidence of the mass manufacture of pottery at sites other than
Rome (Ward-Perkins 2005: 96–100; Peña 2007: 32). Classicists agree, however, that
most goods were manufactured either in households or small shops (Hawkins 2012).
³⁵ Rawson (1976: 86) cites a handful of examples of land parcels, urban and rural,
that were sold four or more times within a fifty-year period.
³⁶ Kirschenbaum (1987: 171–96) provides examples of transactions among intim-
ates, some involving friends of Cicero and Pliny the Younger. On the operation of
reputational constraints within these sorts of social networks, see Temin (2013:
110–13).
³⁷ On both Finley and his critics, see Morris (1999). Erdkamp (2001), who gener-
ally supports the consumer-city thesis, has also summarized the debate. Duffy and
Puzzello (2014) invoke theory and laboratory evidence to evaluate, as a general
matter, the relative merits of gift exchange and monetary exchange.

modern market economy remain acutely status-conscious (McAdams
1997: 355–8). Thorstein Veblen would have recognized Cicero.
Even if members of Rome’s upper crust had been qualitatively
different from contemporary traders—to a modernist, a doubtful
proposition—they constituted no more than one percent of the
city’s population (cf. Engels 1990: 123–4). Would honor and status
have been central when a person of average means was making a
quotidian arrangement with a barber, butcher, or fuller? Might not
the parties in these instances have generally sought to buy low and sell
high? And would not market forces have tended to weed out suppliers
indifferent to covering their costs?
17.2.2. The Agglomeration Efficiencies

That Rome Generated
Even if ancient plutocrats were to have been primarily motivated by
considerations of honor, the conclusion does not follow that Rome
was a “complete parasite-city” that sucked resources out the rest of
the Empire (Finley 1999: 130).³⁸ Imagine that after 300  the city of
Rome had never grown, and also that no other major city had
emerged on the Italian peninsula. How would those counterfactual
conditions have affected the prosperity of a tenant farmer in central
Italy?³⁹ Adam Smith foreshadows, In the Wealth of Nations (1937:
356–7), a vision contrary to Finley’s:
We must not . . . imagine that the gain of the town is the loss of the
country. The gains of both are mutual and reciprocal, and the division
of labor in this, as in all other cases, advantageous to all the different
persons employed in the various occupations into which it is
subdivided.
17.2.2.1. The Economic Theory of Cities

Over the course of history, cities have been engines of economic
growth and technological and artistic innovation. Urban economists
³⁸ See also Finley (1999: 125, 194).

³⁹ In this thought experiment, the tenant farmer would owe rent not to a landlord
residing in a city such as Rome, but to a landlord who was a member of the local
gentry.

such as Edward Glaeser (1998), building on the ideas of Alfred
Marshall and others, have developed a theory of why this is so.⁴⁰
The central point is that people in close proximity can more
cheaply initiate mutually valuable economic and social interactions.
The “agglomeration benefits” of cities can be grouped into three
categories: (1) reductions in the transportation and information
costs of arranging interactions; (2) enhanced specialization of labor
and capital; and (3) spillover benefits arising out of the exchange of
ideas. Although some scholars in classical studies have occasionally
referred to one or more of these benefits, none appears to have made
systematic use of the economic theory of cities.
An example featuring a hypothetical sandal-maker will serve to
illustrate each type of agglomeration benefit. The sandal-maker is an
ex-slave living in a shop on a well-trafficked street in ancient Rome.
Assisted by others, including several slaves of his former master, he
fashions sandals out of leather, wood, and cord, and sells them out of
his shop. How might this sandal-maker benefit from operating a shop
in Rome, as opposed to a rural village?
First, and most mundanely, the sandal-maker would profit from
being in close proximity to numerous suppliers, workers, and custom-
ers. Within a human beehive such as Rome, a sandal-maker could find
within easy walking distance numerous different suppliers of finished
leather and other inputs necessary in the trade, including workers.
A large number of potential buyers of sandals also would live in close
proximity to his shop. The transportation and search costs these poten-
tial customers would incur when engaging in comparative shopping for
sandals would be far lower in a big city than in a rural setting. Moreover,
specialty retailers in a large metropolis commonly cluster their shops
together, thereby reducing the transaction costs of both customers and
suppliers. There is evidence that specialty retailers indeed did cluster
their outlets in Rome (Holleran 2012: 53–60).⁴¹ A big-city location thus
would likely benefit both sandal-makers and sandal-buyers by enhan-
cing their mutual gains from footwear transactions.
Second, a sandal-maker located in Rome, unlike one in a rural
village, could readily specialize in crafting a particular style of sandal.
Opportunities for specialization benefit both suppliers with distinct
⁴⁰ Like Glaeser, Schleicher (2010) presents the theory of agglomeration efficiencies

in accessible fashion. Rosenthal and Strange (2004) summarize the technical literature.
⁴¹ Beard (2008: 60–1) describes the bunching of specialty retailers in Pompeii.

talents and consumers who value product differentiation. As a result,
in a big city, jobs are more varied, social opportunities richer, and
entertainments more diverse. As the satirist Juvenal (1974: 95) put it,
“If you can face the prospect of no more public games, purchase a
freehold house in the country.” Juvenal himself appears to have been
unwilling to face that prospect.⁴²
Third, and commonly of greatest regional value, interactions within
a great city such as Rome generate positive information spillovers. In
the sandal-making context, these would have entailed the invention
and diffusion of new ideas about how to make and market sandals.
Specialists in a big metropolis are likely to fraternize, exchange infor-
mation, and stimulate, through status competition, innovative think-
ing. Paradigm instances include the interactions of Impressionist
painters in Paris and of computer scientists in the Silicon Valley.
The high point of the Golden Age of Latin literature in ancient
Rome is analogous. During Augustan times, two figures centrally
involved were Maecenas and Messalla Corvinus, wealthy patrons of
the arts. Their social and patronage circles came to include a broad
array of Rome’s most distinguished poets, among them Horace, Ovid,
and Virgil. Had Virgil remained in his birthplace, by legend the small
village of Andes in northern Italy, it is a safe bet that he would never
have written the Aeneid. The works of these various poets diffused
throughout the Empire, benefiting outsiders who had never set foot in
the capital. New ideas about the design and manufacture of sandals
would have spread from Rome in similar fashion.⁴³
Although urban residents obtain benefits from city living, they
also must bear a variety of congestion costs. These include the
expense of big-city real estate (or, in the alternative, high commuting
costs), enhanced environmental disamenities, and higher prices for
most goods and services.⁴⁴ The economic theory of cities posits,
⁴² Robinson (1992: 173) asserts that Juvenal, a denizen of Rome, was “someone
who clearly had no intention of living elsewhere.”
⁴³ Perhaps in implicit recognition of the agglomeration benefits that cities generate,
Roman law encouraged migration to Rome. A migrating citizen had full political
rights in the capital, but not, at least as a formal matter, in any alternative destination
(Morley 1996: 174–6).
⁴⁴ The market price of wheat was higher in Rome than elsewhere (Temin 2013:
40–52, 254–5; Temin in volume I;). The city of Rome depended on imports of wheat,
and its grain dealers would have borne relatively high expenses for both space
and labor.

unsurprisingly, that people voluntarily move to a city when they
perceive that their benefits from city life would outweigh their costs.⁴⁵
17.2.2.2. The Many Services That Rome Exported

The consumer-city model is overly narrow because it focuses exclu-
sively on flows of manufactured goods and agricultural products.
Rome, on balance, was not a parasite because it exported a wide
variety of valuable services to residents of its hinterland and prov-
inces. After the founding of the Empire, Rome served as the capital
city until  330, when Constantine moved the capital to Byzantium.
On account of Rome’s special status, members of the city’s elite
assumed central responsibilities in imperial governance. As Jane
Jacobs (1969: 143) succinctly put it, “ultimately government became
Rome’s chief export work.”
During the first several centuries of the Empire, the primary service
that the city of Rome exported was the Pax Romana (Lo Cascio 2007:
621–6; Temin 2013: 222–3). In this interlude the Romans were
relatively successful at deterring internal unrest, invasions across the
Empire’s frontiers, and piracies at sea. Mediterranean trade flour-
ished. Although the legions and navies that were the instruments of
Rome’s hegemonic control were commonly based far from the capital
city, the imperial government in Rome coordinated these expensive
military forces and designed and supervised the taxation system that
paid for them.⁴⁶ During both the late Republic and early Empire, the
governors and other top officials of the provinces (proconsuls, pro-
praetors, legates, and others) were mostly drawn, as part of their
regular career path, from Rome’s senatorial class.
In addition, residents of the city of Rome were disproportionately
involved in establishing the legal and financial institutions that pro-
moted trade throughout Italy and the empire. The praetors, jurists,
and others who refined the principles of Roman law and judicial
administration were disproportionately based in Rome, and the city
also appears to have been a favored venue of litigants.⁴⁷ Officials
⁴⁵ Because a decision to move may give rise to externalities, the population of a city
may not be optimal (Tolley 1974).
⁴⁶ Hopkins (1980, 2002) analyzes the Roman system of taxation.
⁴⁷ D. 4.6.1.1 (Ulp. 12 ad ed.) provides that a party “absent on state business” should
not suffer from any resulting delay in the initiation of an action for restitution of lost
property. But D. 4.6.5.1 (Ulp. 12 ad ed.) adds that “those who serve the state in Rome

in the capital likely coordinated efforts to standardize coinages
(Lo Cascio 2007: 626–30), and weights and measures (Beard 2008:
164–5, 195). From Rome came ideas, funding, and perhaps even
designs for sundry economic development projects throughout the
empire. These included roads between population centers, aqueducts,
harbors, and the planting of cities in the various provinces.
Rome also likely was an important center of private finance, such as
that was. The wealthiest Romans congregated in Rome and would have
sought help in investing their capital. Specialists in banking and credit
emerged (Andreau 1999; Andreau in this volume; Harris 2006; Temin
2013: 157–89), and many of them would have regarded the capital city
as an ideal base for operations. The Roman plutocrats who invested in
rural and urban real estate would have appreciated having at hand
specialists, such as lenders, surveyors, and lawyers, who were capable of
assisting both in the effectuation of land transfers (Crook 1967: 78–9,
89) and the management of rental property. Guardians of the estates of
wards controlled a significant fraction of wealth (Saller 2007: 99–100;
Nicholas 2008: 90–7). They also would have been potential consumers
of financial advice, legal services, and accounting assistance.
As its role in the Golden Age of Latin literature illustrates, Rome was
a center of innovation in the arts, sciences, and engineering, all poten-
tially of benefit to outsiders. Although other cultural centers, such as
Alexandria and Athens, may well have been more vibrant by some
measures, Rome was unmatched in the central portion of the eventual
empire. Many buildings and artistic works in Pompeii, for example,
reflected architectural styles in the capital (Beard 2008: 51–2). Two
libraries, one for works in Latin and the other in Greek, were completed
on Palatine Hill in 28 , and many more followed (Robinson 1992:
57–8). These were a boon to writers of literary, historical, and technical
works. Rome became a center of schooling, tutoring, and apprentice-
ships (cf. Saller 2012: 83). The father of a talented son might either send
him, or accompany him, to Rome to advance his education. As youths,
Cicero, Horace, and Ovid all were sent to the big city.
Perhaps Rome’s most important “export” was its success in attract-
ing to the city hordes of traders, tourists, envoys, and others who paid
short-term visits. These itinerants consumed services and acquired
goods while in the city, and in effect exported those goods and
are not absent on state business,” implying that the city of Rome was seen as an
especially convenient venue for litigation.

services when they returned home.⁴⁸ Rome was an inland city, over 20
kilometers up the Tiber from the port of Ostia. Fifteen roads radiated
out in all directions from Rome into the surrounding hinterland
(Chevallier 1976: 133). The capital offered specialized marketplaces
that drew in outsiders seeking to peddle or acquire goods (Holleran
2012: 159–93). Cattle, for example, were auctioned at the Forum
Boarium.
But the city’s attractions were far more varied. In the mid-first
century , Seneca described Rome’s appeal to both permanent and
temporary migrants:
They have flocked here from their townships and colonies, in short,
from every part of the world: some have been drawn by ambition, some
by the obligation of a public service, some by the office of envoy
entrusted to them, some by luxury seeking a suitable and rich field for
vice, some by desire for higher studies, some by public shows. . . . ⁴⁹
Ah, public shows. Besides hosting religious festivals and theater
performances, Rome was famous for its games. These typically were
open to all, including women and slaves (Robinson 1992: 170). What
spectator could forget an event at the Colosseum featuring lions,
elephants, and rhinoceroses (Beard 2008: 270)? Rome surely
imported bread, but circuses it in effect exported to well-to-do tour-
ists drawn to the city. Visitors attracted to a show or festival might
stay in Rome for several days to shop (Holleran 2012: 191), a practice
that would have enhanced revenues at the city’s many inns, bars, and
houses of adult entertainment.
17.2.3. Evidence That Rome Was a Magnet for Talent

The economic theory of cities posits that the ancient city of Rome would
have been able to attract, and retain, a disproportionate share of the
empire’s most talented individuals. Residence in a city helps enable
people of genius to specialize in tasks that they perform best. They
also can gain from integrating themselves into a social network of people
who share their interests and talents. A specialized network of this sort
provides opportunities for picking up ideas, making social connections,
and motivating oneself through status competition.
⁴⁸ Finley (1999: 139) himself mentions a city’s “invisible exports of trade and tourism.”
⁴⁹ Seneca, “Consolation to Helvia” 6.2, in Seneca (2007: 167).

The life histories of forty-one leading writers and teachers of
the late Roman Republic and early Empire indicate that the city of
Rome indeed served as a magnet for talent. The names of the forty-
one appear in Table 17.1. I selected these individuals prior to
researching their birthplaces and career paths. Each was eminent
enough to have been included in John Hazel’s Who’s Who in the
Roman World (2001).⁵⁰ The forty-one can be loosely grouped into
five categories: historians (n = 9), literary figures (10), jurists and
advocates (13), educators (6), and writers on technical subjects (3).
With the exception of the jurist Modestinus, the lives of each
Table 17.1. Probable birthplaces of forty-one prominent Roman writers and

teachers born 200 – 200.
A. Born prior to 1  (n = 22)
Capito (j), Hortensius Hortalus (j), Lucretius (l),

City of Rome
Quintus Mucius Scaevola (called “Pontifex”) (j)
Within one day’s walk of Rome Cato the Elder (h), Aelius Stilo (e)
Cicero (j), Verrius Flaccus (e), Labeo (j), Ovid (l),
Elsewhere in Central Italy
Sallust (h), Marcus Terentius Varro (h), Vitruvius (t)
Northern Italy Gaius Valerius Catullus (l), Livy (h), Virgil (l)
Southern Italy Ennius (l), Horace (l), Orbilius (e)
North Africa Terence (l)
Spain Porcius Latro (e)
Turkey Alexander Polyhistor (e)
B. Born after  1 (n = 19)
In Rome or within one day’s walk of it [None]

Elsewhere in Central Italy Juvenal (l)
Northern Italy Celsus Publius (j), Pliny the Elder (h), Tacitus (h)
Julianus (j), Papinian (j), Quintus Cervidius
North Africa
Scaevola (j), Suetonius (h)
Spain Columella (t), Lucan (l), Martial (l), Quintilian (e)
France Frontinus (t)
Greece Plutarch (h)
Turkey Modestinus (j)
Levant Josephus (h), Ulpian (j)
Unknown Gaius (j), Paul (Julius Paulus) (j)
Notations: (e) denotes an educator, (h) a historian, (j) a jurist or advocate, (l) a literary figure, and (t) a
writer on a technical subject.
⁵⁰ Although the data in Table 17.1 are primarily drawn from Hazel (2001), in some
instances Magill (1998) and Hornblower and Spawforth (2000) also were consulted.
The latter two sources contain fewer entries than Hazel does.

overlapped at least part of the period 200  to  200.⁵¹ Twenty-two
of the forty-one were born prior to 1 . For the sake of non-
classicists, Table 17.1 employs present-day geographical designations
to identify the region of a birthplace.
Some caveats about the data must be noted. In many cases, classi-
cists have been forced to conjecture about birthplaces and career
paths, and they have not always agreed. The selections are contestable
and forty-one is a small number. Perchance this preliminary venture
will provoke others to compile a larger dataset.
Despite their limitations, the data robustly support the proposition
that a city attracts people of talent. Of the thirty-seven writers and
teachers whose career histories are sufficiently documented, thirty-
four spent much, indeed likely most, of the primes of their careers in
the city of Rome.⁵² These brainy individuals, of course, were not
always in the capital. A person with talent and wealth at times
might travel (perhaps as a member of the emperor’s entourage (Millar
1977)), spend time in a rural villa, and perhaps even serve out a
multiyear tour of duty in a province. Nonetheless, the city of Rome
appears to have been the main locus of the professional lives of thirty-
four out of the thirty-seven.⁵³ Each of the remaining three of the
thirty-seven, moreover, had first-hand knowledge of the capital. Plu-
tarch was born in Greece, but lectured in Rome and “had influential
friends there” (Hazel 2001).⁵⁴ Gaius, who is thought to have received
his legal education in Rome, apparently moved to Berytus (Beirut) as
a young man. If so, Gaius was the only one of the thirty-seven who
affirmatively spurned choosing Rome as his career base. Columella,
the foremost Roman writer on agriculture, unsurprisingly was the
third exception. Columella owned several farms not far from Rome
and likely spent most of his prime years on them. His treatise, De re
⁵¹ Modestinus apparently was born shortly after  200. The Law of Citations of
 426 gave special weight to his opinions and those of four other jurists. Modestinus
was added to the group of forty-one so that it would include all five of these notables.
On the role of the jurists in the Roman legal system, see Nicholas (2008: 28–38).
⁵² For four—Lucretius, Alexander Polyhistor, Verrius Flaccus, and Vitruvius—
there is insufficient information about the venue of the prime career.
⁵³ Two close cases are included in the thirty-four: Virgil, who appears to have split
the prime of his career between Rome and Naples, and Orbilius, who moved to Rome
at age 50.
⁵⁴ Plutarch had predecessors. Ando (2003: 359) identifies numerous Greek writers
who visited Rome during the late Republic and then circulated back to cities in greater
Greece.

rustica, written late in life, includes an extensive bibliography. To
compile this bibliography, Columella would have needed a good
library, and it is easy to guess to what city he went to find one.
Roman leaders, like their counterparts in many other ancient
societies, used banishment as a sanction for perceived misconduct.
During the late Republic and early Empire, banishment invariably,
and tellingly, included a prohibition against residing both in the city
of Rome (Braginton 1944: 392) and in a much broader territory
around it. Three individuals in the group of forty-one—Cicero,
Ovid, and the jurist Paul (Julius Paulus)—unquestionably were ban-
ished, and Juvenal likely was as well. Of these four, only Ovid failed to
receive eventual permission to return to Rome. Exiled to a site in
present-day Romania, Ovid vainly begged for dispensation in letters
to Augustus and, later, Tiberius (Braginton 1944: 404).
The distribution of the birthplaces of these forty-one writers and
teachers powerfully demonstrates Rome’s widespread appeal. The data
in Table 17.1 indicate that no more than four were born in the city of
Rome itself, and only an additional two within a one-day walk of the
city. At least thirty-three of these forty-one talented contributors to
Roman culture were born beyond the immediate hinterland of the city.
This small data-set suggests that, as the Empire matured, the
talented individuals migrating to Rome increasingly came from
distant parts of the greater Mediterranean. In the dataset, three
(14 percent) of the twenty-two talents born prior to 1  were born
beyond the boundaries of modern Italy. By contrast, of the seventeen
who were born after  1 and whose birthplaces are known, thirteen
(76 percent) were born outside Italy, including four in North Africa
and four in Spain.⁵⁵ Birthplaces have been identified for only three of
the five jurists whose eminence was recognized in the Law of Cit-
ations. The three are Modestinus, born in Turkey; Papinian, born in
North Africa;⁵⁶ and Ulpian, born in Lebanon. Each migrated to Rome
and spent the prime of his juridical career there.
Remarkably, none of the nineteen talented individuals born after
 1 is known to have been a native of either the city of Rome or its
⁵⁵ Grounded conjectures about the birthplaces of jurists Gaius and Paul could not
be found.
⁵⁶ Hypotheses about Papinian’s birthplace differ. Tony Honoré’s guess, reported in
Hornblower and Spawforth (2000), is North Africa. Others have surmised that
Papinian was from Syria.

immediate hinterland. This suggests, but, because the numbers are
small, hardly proves, that the city of Rome for some reason came to
lack capacity to produce home-grown talent. The relative shortness of
life expectancies in Rome may have been part of the story (Scheidel
1999: 280).⁵⁷
The life of Josephus, a leading Jewish historian of the first century
, illustrates the varied sources of Rome’s appeal. Josephus was born
in Jerusalem in  37. After participating in the Jewish war against
the Romans of  66, he surrendered to the Romans and soon came
to Rome. The emperor Titus eventually granted Roman citizenship to
Josephus, and provided him a pension and house. Josephus appears
to have spent most of the following three decades in Rome, at work on
his major historical works.
The data in Table 17.1, limited as they are, counter Finley’s con-
ception that Rome was a consumer city parasitic on outsiders. From
Italy and beyond, Rome attracted exceptionally talented people.
By clustering in the capital, these individuals spurred one another
to compose histories, poems, legal treatises, and technical works.
These cultural achievements, like the Pax Romana, were services
that the city of Rome exported empire-wide.
17.3. LEGAL RULES REVEALING ROME’S

COMMITMENT TO “COMMERCIALIZATION
OF THE SOIL”
The Romans adopted two general policies that were necessary, but
not sufficient, conditions for the phenomenal growth of the ancient
city of Rome. This section depicts Rome’s relative willingness, com-
pared to that of other ancient civilizations, to make land alienable by
sale and lease. The next section stresses the relative sagacity of Rome’s
leaders in avoiding the twin perils of too little, and too much, gov-
ernment. In the absence of these two foundational policies, Rome’s
entrepreneurs would not have been able to build and maintain the
apartment blocks that accommodated the 800,000 people added to
⁵⁷ The theory that lead-poisoning is the explanation has fallen out of fashion
(Scheidel 1999: 274).

the city’s population between 200  and  50. Rome doubtless
benefited from other advantages. These arguably included, among
many others, a favorable location, other supportive cultural and legal
traditions, and luck. But Rome’s relatively sound real-estate institu-
tions played an essential role.
17.3.1. Norms and Laws Fostering the

Alienability of Land
The bedrock institution of private land ownership promotes prosper-
ity by incentivizing smallholders to conserve and improve their
holdings with scant supervision from on high (Aristotle, 1946:
49–51; Demsetz 1965; Ellickson 1993). Evidence from early Mesopo-
tamia, Egypt, and China indicates that, in each location, individuals
and families were allowed to own, in perpetuity, much of the land best
allocated to agricultural and residential use.⁵⁸ Roman law, for its part,
firmly protected the rights of private owners of both land and mov-
ables, so much so that nineteenth-century German critics chafed at
the Romans’ “property absolutism” (Whitman 1996).⁵⁹
The placement of land into commerce is a hallmark of an archaic
society’s transition to a market economy. Karl Polanyi (1944: 179)
refers to this event as “commercialization of the soil.” Polanyi con-
tended that land was not bought and sold until after the demise of
feudalism in the late Middle Ages. Polanyi was wrong (Ellickson 1993:
1375–8). In Mesopotamia, evidence of land sales and land leases dates
⁵⁸ On practices in the Near East, where the solidest evidence is from Mesopotamia,
see Ellickson and Thorland (1995) and Westbrook (2003a). On China, see Scogin
(1990). An ancient regime, while permitting private ownership in land, at times might
grant some lands in feudal tenure, and manage others through a palace, temple, or
other hierarchical organization.
⁵⁹ For the regulation of land in Roman law, see Kehoe (2016). According to both
Mommsen and Weber, just after Rome’s founding, arable land in Latium, with the
exception of small plots assigned to families (heredia), was owned communally,
purportedly following traditional clan or tribal practices (Love 1991: 20). Buckland
(1963: 239) concurs: “There was no separate property in land in early Rome except for
the heredium, or houseplace, which was not alienable.” Historically, communal
ownership has been common for pastures, but has been exceedingly rare for arable
lands during the growing season (Ellickson 1993: 1331–2, 1346–8, 1399). The notion
that the founding Romans owned their arable land communally thus warrants
skepticism. If they did, there can be no doubt that their descendants emphatically
rejected that form of land tenure.

at least as far back as 2000  (Westbrook 2003b), and in Egypt, to
the New Kingdom (Jasnow 2003). In China, policies favoring the
alienability of private land were evident in the fourth century  or
earlier (Scogin 1990: 1338–46, 1354). Legal authority to use land to
secure a loan requires acceptance of the possibility that the land
ultimately will end up being owned by the lender. Ancient Mesopota-
mian regimes were willing to accept that possibility (Ellickson and
Thorland 1995: 393–9), and, as early as 500 , Athenians were placing
stones to signify that a land parcel had been pledged to secure a debt
(Finley 1951).
In most ancient societies, loyalties to family and tribe inhibited
unconstrained land transfers, as feudal entails later would in England.
The alienability of land can have both beneficial and harmful conse-
quences. On the positive side, a land transfer is likely to generate gains
from trade, potentially benefiting both the buyer and seller. Other
affected parties also may gain. In Rome, the voluntary sale of an
apartment block, on average, would have shifted its stewardship to
an owner relatively more competent in arranging for the building’s
management—a possible boon for tenants. More generally, land
alienability enhances the mobility of workers, households, and cap-
ital, fueling the dynamism of an economy.
On the negative side, unconstrained land transfers can disrupt
grounded social networks. If a community insider were to sell land to
an outsider, the buyer likely would have fewer loyalties to local resi-
dents than the seller had had. An influx of strangers thus may vitiate a
community’s endeavors, for example, to defend and police itself and
provide its members with informal social insurance. The norms and
laws of archaic civilizations therefore tended to deter an owner of
ancestral land from alienating it to persons other than the members
of the owner’s extended-family or tribe (Ellickson and Thorland 1995:
354–8, 387–93; Westbrook 2003a: 56–62; cf. Finley 1976b: 4–5).⁶⁰
Although the inhabitants of many other ancient civilizations
permitted transactions in land, the Romans were particularly eager
to do so.⁶¹ The city of Rome became, by the standards of ancient
⁶⁰ Romans symbolically recognized their precommercial roots by granting “tribal”

assemblies (comitia tributa) a central role in republican governance. In practice, these
groups were defined by geographic boundaries, not bloodlines.
⁶¹ On the commitment of Roman authorities, from the late Republic to late
antiquity, to the workings of free markets, see Lo Cascio in volume I.

times, a real estate bazaar (Rawson 1976: 86–9; Temin 2013:
139–48).⁶² Ferdinand Tönnies, a prominent nineteenth-century
German critic of Roman society, regretted this commercialization of
the soil. Tönnies (1887) admired institutional arrangements that
foster the intimate and enduring local social networks characteristic
of a precommercial society (Gemeinschaft).⁶³ Like Otto von Gierke
and other fellow German critics, Tönnies thought that Rome’s com-
mercialism had led to a loss of trust and devotion to social duty
(Whitman 1996: 1845–9).
Although the soundness of the German critics’ normative critique
is debatable, they unquestionably were right to conclude that the
residents of Latium, Rome’s region, had been uncommonly willing,
at an early date, to substitute commercial relationships for kinship
relationships. The Romans’ openness to change would later be
reflected in their eagerness to borrow from other cultures, particularly
Greek ones, as opposed to hewing to their own traditions.
The Twelve Tables, a collection of tenets of private law codified in
451–450 , is a primary source of evidence on the predilections
of the Roman elite during the early Republic.⁶⁴ The Twelve
Tables staunchly embraced principles of both private property and
freedom of contract. Most pertinently, this ancient source included
two truly exceptional legal policies that expedited land alienation.
17.3.1.1. Broad Freedom of Testation

The Twelve Tables articulated a principle of complete freedom of
testation (Crook 1967: 122; Watson 1995: 174–5).⁶⁵ This declaration,
⁶² Nonetheless, compared to a present-day real estate market, Rome’s was sluggish.

There were relatively fewer transfers by sale, and more by either testamentary
disposition or conferral of a dowry (Temin 2013: 144). No evidence has been found
of brokers of either land sales or leases (Finley 1999: 117–18). Markets for loans
secured by land appear to have been thin, in part because the wealthy preferred to
borrow from family members or friends (Crook 1967: 170–8; Temin 2013: 168–71).
⁶³ Sociologists debate whether life in a large city is as socially alienating as Tönnies
conceived. Wirth’s analysis (1938), supportive of Tönnies’s view, has been challenged
by Fischer (1982).
⁶⁴ On the Twelve Tables, only fragments of which survive, see Westbrook (2009a).
Livy asserted that, prior to the enactment of the Twelve Tables, the Roman Republic
sent a mission to Athens to learn about that city’s legal traditions. Many scholars are
skeptical of Livy’s account (Watson 1995: 111–12).
⁶⁵ On Roman inheritance law, see generally Crook (1967: 98–138).

presumably not at violence with the norms of the early republicans,
overtly spurned the archaic impulse to discourage the transfer of
property outside the family. On its face, the Twelve Tables authorized
a paterfamilias to transfer patrimonial land by testamentary devise to
a patron, a client, an influential politician, a mistress, or whomever.
As a matter of comparative law, complete freedom of testation is an
extreme legal outlier. In the contemporary United States, virtually all
states confer on a surviving spouse an elective share (Kelly 2013: 1128
n. 9), typically at least one-third of the decedent’s estate. Louisiana,
influenced by French practice, entitles surviving children to shares
of their parents’ estates as a matter of right (ibid.: 1181). In pre-
industrial England, the norms of the aristocracy favored the voluntary
use of entails to keep patrimonial real estate within the family. Roman
law never permitted that option (Buckland 1963: 189, 276; Crook
1967: 121–2; but cf. Champlin 1991: 177–80).
Champlin (1991) has cogently analyzed actual testamentary prac-
tices after 200 , when textual evidence first becomes available.
A wealthy Roman, seeking to avoid the “horror of intestacy,” typically
chose to execute a will (ibid.: 41–63). A large majority of testators,
however, honored informal duties of family support and included
both children and spouses among the named heirs (ibid.: 103–30).
Nonetheless, many also made some use of their broad freedom of
testation. They commonly chose to confer smaller shares on daugh-
ters than on sons (ibid.: 112–20), and not uncommonly provided
legacies to servants and friends (ibid.: 131–54).⁶⁶ Crook (1976: 83)
asserts that, in practice, testators “frequently” favored persons outside
their families, “another spur to the brisk movement of land.”⁶⁷
During the late Republic, formal Roman inheritance law was
amended to soften the Twelve Tables’ unbridled endorsement of
freedom of testation. After these amendments, for example, a testator
had to expressly disinherit a son to bar his claim (Champlin 1991: 14).
And beginning no later than 52 ,⁶⁸ Roman law provided a proced-
ure, querela inofficiosi testamenti, through which a descendent or
other claimant could assert that a testator had violated moral duties
by completely disinheriting the claimant (Crook 1967: 120–1; Cham-
plin 1991: 15). This remedy seems to have been relatively frequently
⁶⁶ Because heirs were responsible for the debts of the decedent, a testator might
have reason to select ones whose assets were meager.
⁶⁷ See also Rawson (1976: 85–9). ⁶⁸ Val. Max. 7.7.2.

invoked, despite the burdens of litigation it placed on a challenger.
But Roman law continued to be relatively protective of freedom of
testation. The lex Falcidia of 40 , for example, limited a claimant
who had successfully invoked querela inofficiosi testamenti to no
more than one-quarter of the testator’s estate. This law also restricted
the portion of the estate that the testator could give away in legacies,
requiring that the heirs receive at least one-fourth of the estate.
17.3.1.2. Legal Practices That Provided

Title Assurance to Land Buyers
A robust real estate market depends on an effective system of title
assurance. In its absence, risk-averse land buyers either may get cold
feet, or have to bear the additional costs of acquiring title informa-
tion.⁶⁹ Legal systems far more ancient than Rome’s struggled to
provide procedures that would help assure a land buyer of the quality
of a seller’s title (Ellickson and Thorland 1995: 382–7).⁷⁰ Some
developed a method of publicizing an upcoming sale. Others pro-
vided a buyer a means of examining documents evidencing prior
sales. The city of Rome, however, appears to have never developed a
system of land records on which land buyers could confidently rely
(Arruñada 2012: 12–16, and his chapter in this volume; Nicholas
2008: 104).⁷¹
From an early date, however, Roman law did provide several
procedures that helped reassure wary land buyers.⁷² Two central
⁶⁹ Frier (1980: 35–6, 122–3) cites instances of the fraudulent leasing of residential
property by a person purporting to be its owner.
⁷⁰ On the general problem, see Baird and Jackson (1984: 304–6) and Arruñada in
this volume.
⁷¹ Rome’s officials may have used records from census surveys for a variety of
purposes, such as determining individuals’ eligibility for the grain dole and vulner-
ability to the military draft. But the jurists seem not to have regarded evidence in land
records to be pertinent to the resolution of title disputes between private parties. Land
records appear to have been better developed beyond the city of Rome. The Romans’
system of centuriation in the colonies gave rise to both land maps and a version of
land records (Duncan-Jones 1976: 12–20; Libecap and Lueck in this volume). In
Roman Egypt a land sale was supposed to be registered as soon as it took place
(Crook 1967: 147–8). By  478, the Eastern Empire may have instituted a formal
system of land records (Macnair 1999: 557–8).
⁷² After their refinement by jurists, Roman law conceptions of property were
significantly more sophisticated than those of prior legal systems in the Near East.
Among Roman law’s achievements were the distinctions between immovable

ones were usucapio and mancipatio, institutions that dated back at
least to the early Republic. Specialists in Roman law have devoted
great attention to both.
a. Usucapio
The Twelve Tables included a second provision stunningly support-
ive of land alienation. Virtually all legal systems entitle a long-time
land possessor who has satisfied specified requirements to obtain title
to another’s land without having to pay for it. In Anglo-American
law, the pertinent doctrine is adverse possession. The closest Roman
law analogue was usucapio (see generally Buckland 1963: 241–9;
Nicholas 2008: 122–5; Arruñada in this volume). To usucapt success-
fully, an adverse possessor had to have acquired the property on a
lawful basis and in good faith—that is, to have thought himself
the rightful owner—and to have remained in overt and continuous
possession for a sufficiently long period of time.⁷³
How long? The Twelve Tables provided that two years would
suffice. Compared to other legal systems, a two-year period is extra-
ordinarily short. England during its Industrial Revolution required
twenty years for successful adverse possession. A solid majority of US
states currently require a period of ten years or more, and none
require as few as two (Foster and McKinney 2011: 201). Remarkably,
the two-year period for usucapio announced in the Twelve
Tables remained a rule of Roman law for at least the ensuing seven
hundred years (Honoré 1989: 141). The shortness of this statute of
limitations, on balance, promoted the alienability of land by limiting
how long a good-faith purchaser had to worry about the quality of a
seller’s title.⁷⁴
property (real estate) and movable property, and between rights in personam (against
single individuals) and in rem (against the world) (Nicholas 2008: 99–100). By
clarifying landowners’ legal rights and remedies, these conceptual advances might
have contributed to Rome’s growth (cf. ibid.: 1; Westbrook 2009a: 69). I myself
suspect that the size of the economic boost from these abstractions would have
been trivial at most.
⁷³ Watson (1971: 64) thinks that the requirement of good faith likely was added
after 200 .
⁷⁴ Offsetting some of this gain would be a buyer’s increased risk of later losing
purchased land to a third-party claimant invoking usucapio.
b. Mancipatio
Even prior to the Twelve Tables, Roman citizens regarded mancipatio
as the favored procedure for an inter vivos transfer of Italic land
(Buckland 1963: 235–41). A mancipatio ceremony required the pres-
ence of not only the transferor and transferee, but also (in effect) six
witnesses, each required to be an adult male citizen. The Digest and
other textual sources reveal nothing about the procedures governing
the selection of these witnesses,⁷⁵ and commentators have seldom
focused on how they were chosen.⁷⁶
If negotiating in good faith, both a prospective seller and a pro-
spective buyer of land would have gained from the presence at a
mancipatio ceremony of witnesses likely to be knowledgeable about
prior transactions involving the same property. Fitting this descrip-
tion might be a prior owner of the land being transferred, or an owner
or long-time occupant of adjacent property. A custom of choosing
knowledgeable witnesses would have made it more difficult for an
unscrupulous seller to sell the same property twice. A buyer might
also have worried about a hidden encumbrance, such as an undis-
closed land-security interest⁷⁷ or servitude.⁷⁸ If able to reduce the
⁷⁵ In Berger (1953), an unusually exhaustive source, neither the entry under

Mancipatio nor that under Testis indicates the party or parties responsible for
selecting these witnesses.
⁷⁶ Arruñada in this volume cogently stresses how a mancipatio ceremony could
serve to publicize a transaction; see also Nicholas (2008: 142, 152). But the usually
insightful Raymond Westbrook (1989: 209) failed to discern mancipatio’s potential
title assurance function. Cf. Champlin (1991: 75–81), presenting evidence on the
selection of witnesses for the execution of a will.
⁷⁷ During the Republic, fiducia denoted a loan secured by property that passed into
the ownership of the creditor. Roman law later recognized a number of alternative
arrangements (pignus, hypotheca, and antichresis) (Buckland 1963: 431–3, 473–81;
Crook 1967: 243–9). Because parties typically used mancipatio to execute a fiducia
transaction (Wigmore 1941: 384–5; Watson 1971: 84–8), a subsequent buyer’s adroit
selection of mancipatio witnesses potentially could have reduced his risks of undis-
closed security interests (Buckland 1963: 231). On land security generally, see Verha-
gen in this volume.
⁷⁸ An urban servitude was not res mancipi, and thus not creatable by a mancipatio
ceremony, but rather only by an in iure cessio proceeding before a magistrate (Watson
1971: 82). On account of this difference in mode of creation, neighbors, assumed here
to be a buyer’s favored mancipatio witnesses, might be less likely to know about
servitudes than about other title defects. But Roman law did provide a land buyer
several protections against a hidden servitude. A mancipatio ceremony created a
warranty of title that a buyer could enforce against a seller by means of an actio
auctoritatis action. This warranty appears to have included protection against an
undisclosed servitude, especially if the mancipation had included the locution optimus

risks of title flaws through the adroit selection of mancipatio
witnesses, a buyer would have been willing to pay a higher purchase
price. If so, the seller and buyer would have had more surplus
to share.
If the parties involved in a land sale had wished the witnesses at a
mancipatio ceremony to be locals, they likely would have preferred to
hold the ceremony at or near the land being transferred. But, in his
Institutes published in  161, Gaius declares that “Lands, however,
are usually mancipated at a distance” (Gai., Inst. 1.121). The customs
for siting these ceremonies doubtless varied across the many centuries
and territories under Roman influence. An off-site location, the
routine choice according to Gaius, would have been sensible if the
buyer and seller were to have been trusting intimates. Suppose, for
instance, that a rich patron residing in the city of Rome had wished to
sell a rural estate in Northern Italy, farmed by tenants, to a client who
resided close to the estate. These trusting parties might have decided
to conduct the mancipatio event in Rome to save the patron the
burden of making the long journey to the rural site. By contrast,
suppose the seller and buyer were to have been strangers, and the
property in question, an apartment block in Rome itself. Under those
circumstances, the parties likely would have preferred to hold the
mancipatio ceremony on-site, where it would have been easier to
assemble witnesses knowledgeable about prior real estate transactions
in the immediate vicinity.⁷⁹ In a society where land titles were uncer-
tain, the mancipatio ritual thus was structured to serve as a potential
source of buyer comfort.
17.3.2. Legal Practices That Fostered the Leasing

of Houses and Shops
Ancient Rome was a city of renters. A third-century census
found that apartment blocks outnumbered houses by more than
maximus (Crook 1976: 72–3). Roman law also refused to enforce a servitude that
either imposed an affirmative duty or whose beneficiaries did not own appurtenant
land (Nicholas 2008: 142–3). These policies would have lessened a land buyer’s risks
of surprise.
⁷⁹ Constantine the Great, emperor from 306 to 337 , eventually decreed that a
land-sale ceremony had to be conducted on site and in the presence of neighbors
(Macnair 1999: 556–7).

twenty-to-one.⁸⁰ The condominium form, an ownership arrangement
that might have enhanced the value of some apartment blocks, was
not recognized (Crook 1967: 143–4; Natelson 1987). These facts
imply that 90 percent of imperial Rome’s residents, if not more,
lived in rental units.⁸¹ For a relatively prosperous household that
could not afford the luxury of a house, the best option was to lease
an apartment in an upscale apartment block for a multiyear term.
Members of a poorer household either resided in the shops where
they worked, were short-term tenants in a squalid rooming house, or
rented a cubicle apartment on an upper floor of a tall apartment
building.⁸²
Frier (1980) provides a definitive analysis of landlord-tenant rela-
tions in Roman cities during the early Empire.⁸³ He persuasively
argues that, in a city such as Rome, the formal landlord-tenant law
depicted in the Digest would have governed at most the upscale
portion of the rental housing market (Frier 1980: 39–52). In the
downscale sector, a legal proceeding typically would have been far
too costly for either a landlord or tenant to consider. And even in the
upscale sector, in most residential-lease disputes, the stakes would
have been small, and the landlord and tenant both would have been
anticipating continuing relations. Under these circumstances, the
disputants commonly would have decided to forgo the expense and
hassle of legal proceedings. They instead would have tended to resolve
their disputes informally, applying social norms that might not have
corresponded to the strictures of formal law (Ellickson 1991; cf.
Kehoe 2007: 13–15, 100–5).
If Roman landlord-tenant law were to have severely impeded the
ready consummation of long-term and short-term residential leases,
the city could never have grown as it did. Dozens of passages in
Justinian’s Digest refer to urban landlord-tenant disputes. Many of
⁸⁰ See n. 17.
⁸¹ Although the members of a household dwelling in a domus typically were far
more numerous than those of a household in an insula, an insula might accommodate
several dozen households, while a domus accommodated only one.
⁸² See nn. 11–16 and accompanying text.
⁸³ The practice of leasing real estate long predated the rise of Rome. In Mesopo-
tamia, the leasing of both houses and fields was common by the middle of the third
millennium  (Ellickson and Thorland 1995: 369). But, as in other areas of law, the
Romans fleshed out, and in this instance helped urbanize, a legal institution that
others had pioneered.

the classicists who have analyzed these entries have assessed the
various doctrines separately. This section offers a holistic, if
stripped-down, account of applicable Roman law and custom.
I discuss the interplay among three norms that were central in
Roman urban leasing practice. Each of the three one-sidedly favored
either the landlord or tenant, and each is entirely at odds with
contemporary US landlord-tenant practice. Frier has concluded that
Roman residential landlord-tenant law, in the limited domain where
it mattered, achieved a reasonable “balance” between the interests of
landlords and tenants, thereby enabling the flourishing of an active
market for urban apartments (Frier 1980: 192–3, 209).⁸⁴ Because the
three one-sided norms crudely offset each other, on balance, I agree
with Frier’s assessment.
First, Roman law protected a lessee’s entitlements against a land-
lord solely by means of liability rules, as opposed to property rules
(Calabresi and Melamed 1972; Frier 1980: 64). In the eyes of Roman
jurists, a landlord retained dominium (Buckland 1963: 187–9). The
upshot, in their view, was that a landlord could oust a tenant from
leased premises at any time, albeit at the risk of being liable to the
tenant for damages incurred.⁸⁵ These legal consequences of domin-
ium were regarded as immutable, that is, not modifiable by the
parties. Most US states, and doubtless many other jurisdictions,
rightly reject this rule and instead, as a default, provide property-
rule protection to a tenant not in breach. A residential tenant who has
robust security of tenure has fewer worries about suddenly losing a
home, and is more likely to invest in both improving the premises and
developing local social connections.
Roman law included a second rule that also was decidedly pro-
landlord. If a lease so provided, and perhaps even if it did not, a
landlord was entitled to unilaterally seize most of the tenant’s fur-
nishings to satisfy the tenant’s arrears on rent payments, and possibly
other debts as well (Frier 1980: 105–35).⁸⁶ In most contexts, a legal
system should disfavor this remedy, known as distraint in Anglo-
⁸⁴ Kehoe (2007: 193–7) similarly concludes that Roman jurists sought to balance
the interests of landlords and tenants when developing the imperial law that governed
the leasing of farms.
⁸⁵ By extension, the buyer of a landlord’s interest was immutably entitled to oust a
sitting tenant. If ousted in this fashion, the ex-tenant’s sole possible remedy would
have been an in personam action against the former landlord (Crook 1976: 73–4).
⁸⁶ Beard (2008: 89–93) describes furnishings in Pompeii apartments.

American law. A tenant may be provoked to resist violently a land-
lord’s agents who have peremptorily entered the tenant’s apartment
and begun to empty it. Most US states prohibit this self-help action
(Rabin 1984: 538), and Roman jurists were aware of its risks.⁸⁷
But there was a third, and countervailing, practice. During the
period of Rome’s rapid growth, the Roman law of apartment leases
embodied an overarching, although hardly blanket, commitment to
freedom of contract (Frier 1980: 61–3).⁸⁸ Many residential tenants
appear to have succeeded in wielding their contracting power to
restore a rough balance of power in their relationships with landlords.
At least during the early Empire, most urban leases either explicitly or
implicitly provided that a tenant’s rent did not become due until after
a period of occupancy (Frier 1977: 29; 1980: 37).⁸⁹ This remarkable
practice, coupled with the related custom that a tenant did not have to
provide a security deposit in advance (Frier 1980: 61 n. 14), vastly
increased the leverage of urban tenants. The term of a residential lease
in Rome typically extended for several years, began on the standard
moving day, July 1, and ended on June 30. The multiyear term was
divided into payment periods, seldom less than six months each. In
the middle of a payment period, a landlord therefore would have been
apprehensive that the tenant would fail to remit the significant sum
due at the period’s end. The tenant’s control over this contingency
would have deterred the landlord from unexpectedly ousting or
otherwise abusing the tenant in mid-period. An ousted tenant likely
would have become enraged, and less likely to perform.
Twenty-first-century landlord-tenant law rejects each of these
three features of Roman landlord-tenant practice. Considered in
isolation, each is suspect on the merits. In most instances, legal
rules should protect, perhaps immutably, both a tenant’s security of
tenure and immunity from a landlord’s unilateral seizure of
⁸⁷ Kehoe (2007: 155–7) reviews Roman law’s limits on the use of distraint in a rural
context.
⁸⁸ See e.g. D. 19.2.15.1 (Ulp. 32 ad ed.): “ . . . [O]r if they agree on something else in
a clause of the hire and this duty is not carried out, there will be an action on hire.” In
some contexts, bargaining to modify a default rule resulted in a pro-landlord clause
(Frier 1980: 61–3, 141–2). See e.g. D. 19.2.11.1 (Ulp. 32 ad ed.), in which a tenant is
deemed liable for damages from a fire that the lease forbade. Conversely, a tenant
could bargain for a waiver of the landlord’s remedy of distraint (Frier 1980: 115).
⁸⁹ D. 19.2.19.6 (Ulp. 32 ad ed.) mentions a rent that was payable in advance. Frier
(1980: 37) asserts that this provision’s detailed discussion implies the rarity of advance
payments, an interpretation contrary to Crook’s (1967: 154).

furnishings. And a landlord has good reasons for insisting that a
tenant both pay rent in advance and provide a security deposit.
The Romans’ trio of one-sided landlord-tenant practices, however,
did crudely offset, resulting in the “balance” that Frier has identified.
17.4. THE ACTIVE, BUT CABINED,

ROLE OF GOVERNMENT
During Rome’s heyday, the praetors, aediles, emperors, and others

who had a hand in the capital’s governance generally made sensible, if
hardly flawless, decisions about the scope of governmental undertak-
ings. Either overly passive governance, or overly intrusive govern-
ance, would have held the city back (Morris et al. 2007: 11–12).
Despite their occasional confiscations of landed property, Rome’s
officials largely respected Roman law’s staunch commitments to
protection of property rights and freedom of contract. They also
were cleverly selective in identifying additional roles that the city
government might affirmatively play. This prudence was another
necessary, although not sufficient, condition for the city’s emergence
as the titan of the Mediterranean.
17.4.1. Policies Addressing Negative

Land Use Externalities
As noted, most of Rome’s neighborhoods were noisy, smelly, and
unsanitary. The city’s governors nonetheless applied a light hand
to problems of negative land use externalities. Although Pirson
(1997: 175) refers to work by German scholars on “building codes,”
there is no evidence that a builder needed a public permit in
advance (Robinson 1992: 33).⁹⁰ Nor is there any evidence of zoning
policies.⁹¹ Augustus’ decree limiting building heights along Rome’s
public streets to 70 feet, if actually enforced, seldom would have been
⁹⁰ Scobie (1986: 405) cites a regulation limiting the width of a party wall to 1.5
inches, a puzzling requirement.
⁹¹ Holleran (2012: 58–9) refers to clusters of tanners in Rome, and Goodman
(2007: 47) to a concentration of industry on the right bank of the Tiber. A cluster of
enterprises of course may arise in the absence of a government directive.

both a binding and burdensome constraint on the developer of an
apartment block.⁹² In some periods, city authorities did require a
landowner who wished to demolish a structure both to obtain per-
mission in advance and immediately build a replacement structure
(Crook 1967: 261–2; Garnsey 1976: 133–6). This policy, like the limit
on heights, may have been motivated in part by aesthetics. It also
helped assure lateral support of adjoining buildings.
To achieve land-use coordination at the block level, Rome’s public
authorities wisely tended to rely on decentralized mechanisms more
than on inflexible command-and-control regulations.⁹³ Roman law
included the rudiments of a law of nuisance.⁹⁴ A neighbor damaged
by another’s fire or smoke, for example, might be entitled to pursue a
legal action for damages.⁹⁵
More important seems to have been the well-developed Roman law
of servitudes (Buckland 1963: 259–68; Crook 1967: 149–52; Rodger
1972; Nicholas 2008: 140–8; Bannon 2009).⁹⁶ Neighbors could resolve
a lurking externality problem, such as a potential blockage of light or
view from an apartment block, by negotiating a servitude that bound
and benefited not only themselves but also their successors in own-
ership (Buckland 1963: 264).
Perhaps most important, Roman customs of real estate develop-
ment produced a set of agents who were well-positioned to negotiate
servitudes and other decentralized solutions to externality problems.
Coase (1961) famously observed that, regardless of the state of the
law, neighboring landowners are incentivized to resolve potential
⁹² The best-preserved insula in Rome, mentioned in n. 9, slightly exceeds Augus-

tus’ height limit (Scobie 1986: 406). Trajan may later have shortened the limit to 60
feet (Robinson 1992: 35–6).
⁹³ On land use issues, see generally Crook (1967: 151, 165–7); Watson (1971:
75–83); and Rodger (1972).
⁹⁴ Anglo-American law students learn that nuisance law derives from the principle
sic utere tuo ut alienum non laedas (use your property in such a way that you do not
damage another’s). This Latin phrase appears to be a post-Roman invention. In
English reports, the first sighting is Edwards vs Halinder, 74 Eng. Rep. 385 (1594).
Lord Coke later quoted the maxim in his famous opinion in William Aldred’s Case, 77
Eng. Rep. 816, 821 (K.B. 1611).
⁹⁵ See e.g. D. 9.2.27.9 (Ulp. 18 ad ed.) (applying negligence standard to determine
liability for spread of a fire); D. 8.5.8.5 (Ulp. 17 ad ed.) (a much-discussed passage
authorizing relief against smoke from a ground-level cheese shop); but cf. D. 8.5.8.6
(Ulp. 17 ad ed.) (a moderate amount of smoke from a hearth would not be actionable).
⁹⁶ Servitudes predate Roman law. They were recognized in both ancient Mesopo-
tamia (Westbrook 2003b) and New Kingdom Egypt (Jasnow 2003).

nuisance conflicts in mutually advantageous fashion. He also stressed
that the transaction costs of reaching an agreement might foil their
bargaining efforts.
In ancient Rome, Coasean bargaining to resolve a potential land-
use conflict typically would have involved relatively few agents, a
condition conducive to decentralized coordination. Most houses
and apartment blocks were situated on spacious lots. City blocks in
ancient Rome, as in modern Rome, generally were small in area.
Because city officials apparently did not attempt to regulate the size
and shape of lots, entrepreneurs could consolidate and subdivide lots
within a city block at will. Most apartment blocks in Ostia, thought to
be similar to those in Rome, sat on lots that had varying shapes and
commonly included an area that exceeded 1,000 square meters
(Packer 1971: 93–110).⁹⁷ A small city block in Rome thus might
have included only handful of apartment blocks, each teeming with
tenants.
A high-status owner of an apartment block or lodging house in
Rome had several means of avoiding the hassle, and potential social
stigma, of involvement in the building’s day-to-day management.
An owner commonly entrusted supervision of the building either
to a middleman who had taken out a master lease on the entire
premises (Frier 1980: 30–1, 36; du Plessis 2006), or to a caretaker
(insularius) (Frier 1980: 24, 29–30; Wallace-Hadrill 1994: 132; Frier
and Kehoe 2007: 130–4). Given the limitations of Roman agency law,
promising candidates for the caretaker role would have included the
owner’s son, slave, or ex-slave.⁹⁸ Roman law recognized a particular
business form, the peculium, that was especially adapted to a slave-
run enterprise (Kirschenbaum 1987: 31–47; Hansmann et al. 2006:
1360–4; Abatino and Dari-Mattiacci in volume I). A hard-working
slave had a good prospect of eventually buying freedom from a
master.⁹⁹ In a much-quoted advertisement painted on a wall of an
⁹⁷ Compare Frier (1980: 9), describing as “typical” an Ostian apartment block that
was situated on a slightly rhomboid lot whose sides averaged 21 meters.
⁹⁸ Roman law generally did not permit an agent other than a son, slave, or ex-slave
to bind a principal (Kirschenbaum 1987: 1–7, 13–14).
⁹⁹ Love (1991: 136) estimates “that probably over three-fifths of slaves were freed
before they were thirty.” This is probably incorrect. Koops in this volume implies that
manumission after age 30 might have been more typical. On slavery in the Roman
Empire, including the incentives of slaves to work, see Scheidel (2012b).

insula in Pompeii, prospective tenants were advised to “consult
Primus, slave of Cnaius Alleius Nigidius Maius” (Beard 2008: 109).
On each lot, in short, there typically would have been an individual
with presumptive authority to bargain over smoke, noise, and other
externality issues. In the case of an apartment block, the key figure
likely would have been either the caretaker whom the owner had
selected, or the middleman to whom the owner had leased the entire
complex. Roman law incentivized an owner of an apartment block to
prompt the owner’s agent to consider the welfare of current tenants.
The partial blockage of a tenant’s light by a neighbor, for example,
could give rise to a tenant claim against his own landlord (Frier
1980: 102).
For a domus, the agent for Coasean bargaining normally would
have been the paterfamilias of the household (Saller 2007; Nicholas
2008: 65–9).¹⁰⁰ The paterfamilias would have been well-positioned
not only to expeditiously resolve land-use conflicts among members
of his often-large household, but also to represent the household in
negotiations with neighbors over block-level environmental issues.
During the late Republic and early Empire, Rome attracted
continuing waves of immigrants. Most of these newcomers must
have been aware of the noise, fires, diseases, and other risks of life
in the big city. Narrowly targeted building regulations possibly
would have made Rome an even more attractive place of residence.
But crude building regulations, such as a draconian limit on the
heights of the city’s buildings, would have stifled the supply of
private housing. The Romans likely were wise to relegate the
resolution of most nuisance and building-quality issues to block-
level actors.
17.4.2. The Supply and Financing of Public Goods

Rulers in ancient Mesopotamia, Egypt, and China attained legitimacy
in part by providing dams, canals, and other water projects that
would have been beyond the capabilities of non-state actors
¹⁰⁰ A Roman household did not invariably include a paterfamilias. Probably over
20 percent of land was owned by women, and almost as large a fraction by orphans (a
male orphan would also be a paterfamilias) whose affairs were being managed by a
guardian (Saller 2007: 97, 100). But in these situations as well, there likely would have
been an agent authorized to speak for the household.

(Wittfogel 1957). The Romans similarly developed a public sector
that bore responsibility for providing aqueducts and other public
goods. Roman jurists made a sharp distinction between public
property (res publica) and private property, and favored the
staunch protection of both (Buckland 1963: 183). In the absence of
government improvements and services, Rome could never have
grown as it did.
The theory of public finance, pioneered by Paul Samuelson (1954)
and Richard Musgrave (1959), seeks to identify the goods and services
that a government can provide more capably than can the decentral-
ized actors in the private sector. In brief, a public good is one that
either is nonrivalrously consumed, such as national defense, or one,
such as a street, from which a provider cannot practically exclude
entrants. Rome’s officials, most prominently the curule and plebeian
aediles, had no overarching theory of the proper scope of govern-
ment, and determined its ambit by muddling through. During the
period of the city’s ascendance, the decisions of these officials, com-
pared to those of many other ancient polities, were relatively sound.
Ancient travelers commonly regarded Rome’s public works to be
among its greatest achievements (Robinson 1992: 59, 99).
Management of the city streets was a key public undertaking
(Robinson 1992: 59–82). Archaeological findings indicate that the
city of Rome evolved from a settlement that dated back to the early
second millennium . On account of path dependence, the locations
and widths of the settlement’s earliest street rights-of-way would
have tended to endure (Ellickson 2013). Rome’s officials began to
pave its streets in 238  (Chevallier 1976: 71) with polygon-shaped
stones, likely of basalt. Street congestion eventually became a
serious problem. In 45 , Julius Caesar famously banned, with
some exceptions, traffic by carts between sunrise and late afternoon
(Beard 2008: 80). Evidence from Pompeii (Beard 2008: 65–70)
implies that Rome’s officials may have experimented with one-way
streets and pedestrian-only streets. City officials also made efforts to
beautify streetscapes. By the early Empire, there were an estimated
500,000 statues in Rome (Beard 2008: 77–8), many of them placed in
streets, forums, and other public spaces.
Rome’s public sector also saw to the provision of public buildings,
many of world-class design. These included temples to various gods,
public baths, theaters, libraries, and the famous venues for games
(Robinson 1992: 47–58). Censors, the officials mainly responsible for

public contracts during the Republic, typically hired specialized con-
tractors (publicani) to execute these building projects (du Plessis
2004; Hansmann et al. 2006: 1360–4).
City officials provided essential utility services as well. By 200 ,
Rome already had two aqueducts. The fast-growing city added seven
more by  52, quadrupling Rome’s public water supply (Martini
1976: 563). The aqueducts supplied the city’s public fountains, where
most households drew their water. Partly to help clear this incoming
water, Rome also became a pioneer in sewer technology (Robinson
1992: 117–19).
Besides attending to these physical improvements, public officials
provided services that arguably can be characterized as natural mon-
opolies. These included the coordination of street cleaning (Beard
2008: 56) and, beginning in the early Empire, the provision of mobile
crews of fire-fighters.¹⁰¹
Although theorists of public finance might quibble over the wis-
dom of some of Rome’s public undertakings, most of them likely
would give the city’s high marks overall. As important as what Rome’s
governors did, however, was what they declined to do. During the
period of the city’s rapid growth, the production and management of
houses, apartment blocks, shops, and many other goods and services
were relegated to the private sector. Most theorists of public finance
similarly would endorse the decision to leave this portion of the
economy privatized. A private individual who supplies a good or
service personally bears the cost of not making a profit, and therefore
is likely to be more cost-conscious than a public supplier. More
generally, economic forces, in a competitive market, tend to be far
better than political forces at weeding out inefficient suppliers. Con-
sistent with this theory, the governments that initiated major efforts
to socialize housing supply during the twentieth century typically
performed poorly.¹⁰² The ancient Romans avoided this stumble.
¹⁰¹ See Robinson (1992: 105–8), describing the roles of the hundreds of vigiles who
patrolled Rome nightly.
¹⁰² During the mid-twentieth century, the governments of France, Netherlands, the
United Kingdom, and many other nations assumed responsibility for providing a
majority of their nation’s rental housing stock. These policies, like the public housing
program in the United States, have generally been more wasteful than government
housing assistance programs that would have relied more on the private providers of
rental housing (Ellickson 2010: 986, 995–1003).

In an economy growing on the order of 0.1 percent per year, only a
few basic policy mistakes would have been sufficient to push Rome off
the path toward prosperity.¹⁰³
The Romans also were relatively adroit at choosing methods of
financing their various public goods. Although localities collected
general taxes in certain periods (Lo Cascio 2007: 625), these seldom
were a major source of revenue. Rome defrayed more costs through
the imposition of user fees on benefited landowners. An owner of
land abutting a street, for example, might be charged with responsi-
bility for maintaining that stretch of street, and, on failure to perform,
billed for the costs of a contractor hired to do the job (Crook 1967:
257–60; Robinson 1992: 59–73). Some consumers of water also were
charged user fees (Robinson 1992: 104).
Far more important in the system of public finance was the custom
of euergetism, a refinement of a tradition previously developed in
some Greek city-states (Engels 1990: 128; Finley 1999: 150–4).
Rome’s richest residents were under great pressure from both gov-
ernment officials and social peers to finance voluntarily public
improvements and services.¹⁰⁴ During the early Empire, for example,
the emperor might set a good example by “giving” a public building
or group of gladiatorial games to the people of Rome (Millar 1977:
195, 365). Victorious generals and ambitious plutocrats contributed
in similar fashion (Robinson 1992: 160–72). Supplementing the sys-
tem of euergetism were informal requirements that men appointed to
powerful positions make out-of-pocket donations (summae) to
finance public endeavors. After Augustus appointed his friend
M. Agrippa curator aquarum, Agrippa paid for a new aqueduct out
of personal funds, a precedent soon followed by the emperor Claudius
(Rodgers 2004: 15–16).¹⁰⁵ Rome’s sumptuary laws (Harris 2007: 530)
may have been partly designed to redirect the expenditures of the
wealthy away from conspicuous consumption (a zero-sum game of
¹⁰³ The 0.1 percent figure is Richard Saller’s rough estimate of the average annual
rate of economic growth in the western Roman empire during the three centuries
following 200  (Morris et al. 2007: 5). Temin (2013: 197) stresses the roughness of
this figure.
¹⁰⁴ In Pompeii, local notables similarly financed the erection of statues in public
places (Beard 2008: 197).
¹⁰⁵ Champlin (1991: 158–9) describes bequests to fund public buildings and
services.

individual status enhancement), and toward the financing of public
goods (potentially a positive-sum game).
17.4.3. Rome’s Relative Resistance to Populist Policies

Ancient regimes in the Near East and elsewhere commonly adopted
measures that were likely to generate immediate popular acclaim, but
impair long-term economic growth.¹⁰⁶ Three examples of these popu-
list policies were edicts cancelling debts, rent control measures, and
conferrals on land sellers of immutable rights to repurchase (redeem)
transferred land. Rome’s officials certainly succumbed at times to
populist political pressures (Lévy 1967: 55–8, 68–9).¹⁰⁷ But, particu-
larly after the ascension of Augustus, Roman officials were relatively
resistant to them.¹⁰⁸
On assuming office, a king of an ancient Near Eastern empire
commonly proclaimed a partial or complete cancelation of debts
owed to private creditors (Ellickson and Thorland 1995: 401–4;
Westbrook 2009b: 151–60). Although applauded by debtors, a debt-
cancelation decree (misharum) typically would have counterproduct-
ive effects on financial markets in the long term.¹⁰⁹ A potential lender
aware of the risk of cancelation might either refuse to lend, or charge
a higher interest rate. At minimum, when negotiating loan terms, a
lender and borrower uncertain about the timing of the next debt
cancelation would incur higher transaction costs.
In 86 , in response to shortages of coinage, the Roman Republic
approved a law cancelling 75 percent of outstanding debts, and, in 49
, another law cancelling roughly 25 percent (Harris 2007: 519).
Julius Caesar ordered a partial remission of rents in 48  (Frier 1977:
34), prompting Finley (1999: 143) to speculate that Caesar’s oppon-
ents worried that he would cancel yet more debts. In 41 , three
¹⁰⁶ On the prevalence of these policies in even more ancient regimes, see Ellickson
and Thorland (1995: 400–8).
¹⁰⁷ In the late Republic there were efforts to limit interest rates through usury laws
(Harris 2007: 520). And Tiberius and Nero made sporadic efforts to control the price
of grain (Temin 2013: 33–5).
¹⁰⁸ In 58 , the tribune Clodius, the late Republic’s most notable populist, helped
institute what became a long-lived, but increasingly restricted, policy of doles of free
wheat to adult male citizens resident in Rome (Lo Cascio 2007: 639–41).
¹⁰⁹ A debt-cancelation measure that cooled popular wrath conceivably might have
aided an ancient economy by preventing an even more destructive event, such as a
bloody revolution or civil war (cf. Roe 1998).

years after Caesar’s assassination, the Second Triumvirate ordered
yet another rent remission (Frier 1977: 36). After Augustus
assumed the emperorship, however, for several centuries there appear
to have been no debt-cancelation decrees. By enabling developers
and buyers of apartment blocks to obtain loans on more favorable
terms, this abstention would have contributed to the growth of the
city of Rome.
Rent controls are embraced by many twenty-first-century states.
The details of these measures vary, and so do, accordingly, the
inefficiencies that they cause. Crude controls on urban rents severely
crimp both the construction of new private housing and the main-
tenance of existing housing. Rent ceilings also can reduce household
mobility and produce mismatches between households and dwellings
(Glaeser and Luttmer 2003). Other than the aforementioned remis-
sions of rents in 48 and 41 , the city of Rome appears to have been
free of legislated rent controls (Frier 1977: 36).¹¹⁰ The notion of a “just
price” was not formally recognized in Roman law until the late
Empire (Gordley 1981: 1601–2). During the period of Rome’s growth,
market prices generally were regarded as fair prices (Lo Cascio in
volume I). A judge therefore would have rebuffed a tenant’s com-
plaint that a negotiated rent had been pegged too high.
Prior to 200 , states in China and the Near East commonly
provided a land seller, and his heirs, an unwaivable entitlement to
repurchase sold land at the original purchase price (Westbrook
2009b: 148–51; Ellickson 2012).¹¹¹ This redemption right would
have helped keep ancestral land, perhaps mortgaged during a period
of financial distress, within a family. In some contexts, these rights
might have generated net economic benefits by preserving local
social networks and providing informal social insurance (Ellickson
2012: 294). In a large city such as Rome, however, redemption rights
on balance would have fouled land markets. They would have made,
for example, a potential buyer of a lot suited for a new apartment
block more wary of purchasing, and after purchase, of building.
¹¹⁰ Centuries later, emperors Constantine and Justinian attempted to peg rents of
agricultural lands to customary levels (Kehoe 2007: 134–5).
¹¹¹ Although Roman law did not create implicit redemption rights, it permitted
parties to create a repurchase right by express agreement. See D. 18.1.75 (Hermogen-
ian 2 iuris epit.); D. 19.1.21.5 (Paul. 33 ad ed.); D. 19.5.12 (Proculus 11 epist.).

Roman law, in contrast to the law of ancient China, Israel, and
several Near Eastern states, appears never to have conferred immut-
able redemption rights.¹¹² Yet another populist bullet thus
was dodged.¹¹³
17.5. CONCLUSI ON
The tools of law and economics generate insights into the causes and
effects of the unprecedented growth of the ancient city of Rome.
Cities historically have been centers of creativity, and urban econom-
ics helps reveal why this has been so. Rome was the indispensable
nerve center not only of the empire’s predations, but also of its
accomplishments. The economic perspective helps elucidate the
many services that Rome exported, and why talented individuals
from around the Mediterranean flocked to it. The theory of public
goods similarly illuminates the city of Rome’s division of responsi-
bilities between its public and private sectors.
A law-and-economics perspective provides a useful prism,
although hardly the only one, on the merits of specific institutions.
By present-day standards, many of Rome’s core policies were deeply
flawed. The city’s governance system was clumsy. Legal barriers kept
women from fully developing their talents. Slavery, and the many
other formal status distinctions that pervaded Roman culture, tended
to impair social and economic mobility.¹¹⁴ The institutions governing
real estate transactions—the focus of much of this chapter—also had
¹¹² Westbrook (1989: 208–9) asserts that mancipatio was a procedure designed to
extinguish redemption rights, an entitlement that had “entirely disappeared” from
Roman law by classical times.
¹¹³ Populist policies were more in vogue during the late Empire. For example, in 
301, Diocletian’s Edict on Maximum Prices sought to regulate the prices of over a
thousand commodities. Bartlett (1994) asserts that unbridled statism of this stripe
contributed significantly to the decline and fall of the Western Empire.
¹¹⁴ As a general matter, Roman slavery was less harsh than the subsequent slavery
systems of the post-conquest Americas (Temin 2013: 114–38). In exceptional
instances, some freeborn individuals in Rome voluntarily entered slavery to obtain
managerial positions that otherwise would have been unavailable (Dari-Mattiacci
2013: 98).

shortcomings. Rome lacked a reliable system of land records and a
condominium form of ownership. The state’s power of compulsory
purchase (eminent domain) was ambiguous.¹¹⁵ And Rome’s officials
at times baldly confiscated real estate, such as Cicero’s house on the
Palatine (Hales 2000: 45–6).¹¹⁶ A growing city requires the construc-
tion of buildings, and confiscations hardly foster their production.
Measured against the practices of other ancient civilizations, how-
ever, Rome’s land institutions were relatively growth-promoting.
Provisions of the Twelve Tables signaled that the people of the
early Republic, despite their roots in a tribe-based culture, were
uncommonly committed to putting land into commerce. Embracing
the foundations of a market economy, the Romans recognized
private property in land, labor, and capital, and also conferred on
an owner broad discretion to transfer an asset by sale, lease, or will.
In addition, the Romans came to be relatively resistant to populist
enactments. Their embrace of this cluster of policies was a necessary,
although hardly a sufficient, condition for the transformation of a
polity in the backwater of Latium into a city that coordinated the
conquest of the entire Mediterranean. An intriguing question is why
the descendants of the members of tribes residing along the left bank
of the Tiber had initially opted for this relatively radical set of
policies. On that issue, classicists steeped in Roman culture will
have ideas far better than mine.¹¹⁷
¹¹⁵ In 1625 in De jure belli et pacis, Hugo Grotius coined the phrase dominium
eminens (supreme lordship) to describe this power. Roman lawyers appear never to
have crisply conceptualized the issue. Crook (1967: 262–4) asserts that public author-
ities only “sparingly” forced land transfers, and, when they did, typically provided at
least partial compensation; see also Reynolds (2010). Taylor (2000: 93–127), by
contrast, argues that Roman governments lacked the power of compulsory purchase,
but were willing to confiscate an owner’s assets to punish a perceived misdeed.
¹¹⁶ Land could be confiscated piecemeal by court order or imperial fiat, or en masse
as a result of military action. In the latter case, the land became ager publicus, a spoil
available for grant or subsidized lease to soldiers or others (Finley 1999: 119). On land
confiscations in societies more ancient than Rome’s, see Ellickson and Thorland
(1995: 345–6).
¹¹⁷ For comments and other help, I thank Clifford Ando, Jean Andreau, Emma-
nuelle Chevreau, Cyril Courrier, Giuseppe Dari-Mattiacci, Adriaan Lanni, Joseph
Manning, David Schleicher, Steven Shavell, Pierre Vesperini, James Whitman, and
the two reviewers. Responsibility for errors is entirely mine. Zachary Herz provided
essential research assistance, and Sarah Kraus was indispensable in locating sources.

REFERENCES
Ades, Alberto F., and Edward L. Glaeser. 1995. “Trade and circuses: explain-
ing urban giants.” 110 Quarterly Journal of Economics 195–227.
Alonso, William. 1964. Location and Land Use: Toward a General Theory of
Land Rent. Cambridge, MA: Harvard University Press.
Ando, Clifford. 2003. “Review of Being Greek under Rome by Simon Gold-
hill.” 57 Phoenix 355–60.
Angel, Shlomo. 2012. Planet of Cities. Cambridge, MA: Lincoln Institute of
Land Policy.
Aristotle. 1946. Politics, trans. Ernest Barker. Oxford: Clarendon Press.
Theory and Practice of Contractual Registries. Chicago and London: Uni-
versity of Chicago Press.
Baird, Douglas, and Thomas Jackson. 1984. “Information, uncertainty, and
the transfer of property.” 13 Journal of Legal Studies 299–320.
Bannon, Cynthia. 2009. Gardens and Neighbors: Private Water Rights in
Roman Italy. Ann Arbor: University of Michigan Press.
Bartlett, Bruce. 1994. “How excessive government killed ancient Rome.” 14
Cato Journal 287–303.
Beard, Mary. 2008. The Fires of Vesuvius: Pompeii Lost and Found. Cam-
bridge, MA: Harvard University Press.
Benevolo, Leonardo. 1980. The History of the City, trans. Geoffrey Culver-
well. Cambridge, MA: MIT Press.
Berger, Adolf. 1953. “Encyclopedic dictionary of Roman law.” 43 (New
Series) Transactions of the American Philosophical Society 333–809.
Bowman, Alan, and Andrew Wilson, eds. 2011a. Settlement, Urbanization,
and Population. Oxford: Oxford University Press.
Bowman, Alan, and Andrew Wilson, eds. 2011b. “Introduction,” in Alan
Bowman and Andrew Wilson eds, Settlement, Urbanization, and Popula-
tion. Oxford: Oxford University Press, 1–14.
Braginton, Mary V. 1944. “Exile under the Roman emperor.” 39 Classical
Journal 391–407.
Braund, Susan H. 1989. “City and Country in Roman Satire,” in Susan
H. Braund, ed., Satire and Society in Ancient Rome. Exeter: Exeter Uni-
versity Publications, 23–47.
Buckland, William Warwick. 1963. A Text-Book of Roman Law, ed. Peter
Stein, 3rd rev. edn. Cambridge: Cambridge University Press.
Calabresi, Guido, and A. Douglas Melamed. 1972. “Property rules, liability
rules, and inalienability: one view of the cathedral.” 85 Harvard Law
Review 1089.

Campbell, Brian. 2000. The Writings of the Roman Land Surveyors. London:
Society for the Promotion of Roman Studies.
Champlin, Edward. 1991. Final Judgments: Duty and Emotion in Roman
Wills, 200 – 250. Berkeley: University of California Press.
Chevallier, Raymond. 1976. Roman Roads, trans. N. H. Field. London:
B. T. Batsford.
Coase, Ronald H. 1960. “The problem of social cost.” 3 Journal of Law and
Economics 1–44.
Crook, John A. 1967. Law and Life in Rome. Ithaca: Cornell University Press.
Crook, John A. 1976. “Classical Roman law and the sale of land,” in Finley,
ed., 71–83.
Dari-Mattiacci, Giuseppe. 2013. “Slavery and information.” 73 Journal of
Dubouloz, Julien. 2011. La propriété immoblière à Rome et en Italie (1er–Ve
siècles). Rome: École Française de Rome.
Duffy, John, and Daniela Puzzello. 2014. “Gift exchange versus monetary
exchange: theory and evidence.” 104 American Economic Review 1735–76.
Duncan-Jones, Richard P. 1976. “Some configurations of landholding in the
Roman Empire,” in Finley, ed., 7–33.
du Plessis, Paul. 2004. “The protection of the contractor in public works
contracts in the Roman Republic and early Empire.” 25 Journal of Legal
History 287–314.
du Plessis, Paul. 2006. “Janus in the Roman law of urban lease.” 55 Historia
48–63.
Ellickson, Robert C. 1991. Order without Law: How Neighbors Settle Disputes.
Ellickson, Robert C. 1993. “Property in land.” 102 Yale Law Journal
1315–1400.
Ellickson, Robert C. 2006. “Unpacking the household: informal property
rights around the hearth.” 116 Yale Law Journal 226–328.
Ellickson, Robert C. 2010. “The false promise of the mixed-income housing
project.” 57 UCLA Law Review 983–1021.
Ellickson, Robert C. 2012. “The costs of complex land titles: two examples
from China,” 1 Brigham-Kanner Property Rights Conference Journal,
281–302.
Ellickson, Robert C. 2013. “The law and economics of street layouts: how a
grid pattern benefits a downtown.” 64 Alabama Law Review 463–510.
Ellickson, Robert C., and Charles Dia. Thorland. 1995. “Ancient land law:
Mesopotamia, Egypt, Israel.” 71 Chicago-Kent Law Review 321–411.
Engels, Donald. 1990. Roman Corinth: An Alternative Model for the Classical
City. Chicago: University of Chicago Press.
Erdkamp, Paul P. M. 2001. “Beyond the limits of the ‘consumer city’: a model
of the urban and rural economy in the Roman world.” 50 Historia 332–56.

Finley, Moses I. 1951. Studies in Land and Credit in Ancient Athens, 500–200
: The Horos Inscriptions. New Brunswick, NJ: Rutgers University Press.
Finley, Moses I., ed. 1976a. Studies in Roman Property. Cambridge: Cam-
Finley, Moses I. 1976b. “Introduction,” in Finley, ed., 1–6.
Finley, Moses I. 1977. “The ancient city: from Fustel de Coulanges to Max
Weber and beyond.” 19 Comparative Studies in Society and History
305–27.
Finley, Moses I. 1999. The Ancient Economy, ed. Ian Morris. Berkeley:
University of California Press.
Fischer, Claude S. 1982. To Dwell among Friends: Personal Networks in Town
and City. Chicago: University of Chicago Press.
Foster, Lynn, and J. Cliff McKinney, Jr. 2011. “Adverse possession and
boundary by acquiescence in Arkansas.” 33 University of Arkansas at
Little Rock Law Review 199–262.
Frier, Bruce W. 1977. “The rental market in early imperial Rome.” 67 Journal
of Roman Studies 27–37.
Frier, Bruce W. 1980. Landlords and Tenants in Imperial Rome. Princeton:
Princeton University Press.
Frier, Bruce W., and Dennis P. Kehoe. 2007. “Law and economic institu-
tions,” in Scheidel, Morris, and Saller, eds, 113–43.
Frederiksen, Martin W. 1975. “Theory, evidence and the ancient economy.”
Gabaix, Xavier. 1999. “Zipf ’s law for cities: an explanation.” 114 Quarterly
Journal of Economics 739–67.
Garnsey, Peter. 1976. “Urban property investment,” in Finley, ed., 123–36.
Glaeser, Edward. 1998. “Are cities dying?” 12 Journal of Economic Perspec-
tives 139–60.
Glaeser, Edward. 2011. Triumph of the City. New York: Penguin Press.
Glaeser, Edward L., and Erzo F. P. Luttmer. 2003. “The misallocation of
housing under rent control.” 93 American Economic Review 1027–46.
Goodman, Penelope J. 2007. The Roman City and its Periphery. London:
Routledge.
Gordley, James. 1981. “Equality in exchange.” 69 California Law Review
1587–1656.
Hales, Shelley. 2000. “At home with Cicero.” 47 Greece & Rome (2nd Series)
44–55.
Hansmann, Henry B., Reinier Kraakman, and Richard Squire. 2006. “Law
and the rise of the firm.” 119 Harvard Law Review 1333–1403.
Harris, William V. 2007. “The late Republic,” in Scheidel, Morris, and Saller,
eds, 511–39.

Hawkins, Cameron. 2012. “Manufacturing,” in Scheidel, ed., 175–94.
Hazel, John. 2001. Who’s Who in the Roman World. London: Routledge.
Holleran, Claire. 2012. Shopping in Ancient Rome. Oxford: Oxford Univer-
sity Press.
Honoré, Tony. 1989. “Conveyances of land and professional standards in the
later empire,” in Peter Birks, ed., New Perspectives in the Roman Law of
Property. Oxford: Clarendon Press, 137–52.
Hopkins, Keith. 1980. “Taxes and trade in the Roman empire (200 ..–..
400).” 70 Journal of Roman Studies 101–25.
Hopkins, Keith. 2002. “Rome, taxes, rents and trade,” in Walter Scheidel and
Sitta von Reden, eds, The Roman Economy. New York: Routledge,
190–230.
Hornblower, Simon, and Tony Spawforth, eds. 2000. Who’s Who in the
Classical World. Oxford: Oxford University Press.
Jacobs, Jane. 1969. The Economy of Cities. New York: Random House.
Jasnow, Richard. 2003. “Egypt: New Kingdom,” in Raymond Westbrook, ed.,
A History of Ancient Near Eastern Law, vol. 1. Leiden: Brill, 289–359.
Jongman, Willem. 1988. The Economy and Society of Pompeii. Amsterdam:
J. C. Gieben.
Juvenal. 1974. The Sixteen Satires, trans. Peter Green, rev. edn. Harmonds-
worth: Penguin Books.
Kehoe, Dennis P. 2007. Law and Rural Economy in the Roman Empire. Ann
Arbor: University of Michigan Press.
Kehoe, Dennis P. 2012. “Contract labor,” in Scheidel, ed., 114–30.
Kehoe, Dennis P. 2016. “Tenure of land and agricultural regulation,” in
P. J. du Plessis, C. Ando, and K. Tuori, eds, Oxford Handbook of Roman
Law and Society. Oxford: Oxford University Press, 646–59.
Kelly, Daniel B. 2013. “Restricting testamentary freedom: ex ante versus ex
post justifications.” 82 Fordham Law Review 1125–85.
Kirschenbaum, Aaron. 1987. Sons, Slaves and Freedmen in Roman Com-
merce. Jerusalem: Magnes Press.
Laurence, Ray. 1997. “Writing the Roman metropolis,” in Helen M. Parkins,
ed., Roman Urbanism: Beyond the Consumer City. London: Routledge,
1–20.
Lévy, Jean-Philippe. 1967. The Economic Life of the Ancient World, trans.
John G. Biram. Chicago: University of Chicago Press.
Lo Cascio, Elio. 2007. “The early Roman empire: the state and the economy,”
in Scheidel, Morris, and Saller, eds, 619–47.
Love, John R. 1991. Antiquity and Capitalism: Max Weber and the Socio-
logical Foundations of Roman Civilization. London: Routledge.
Machado, Carlos. 2012. “Review of La propriété immobilière à Rome et en
Italie (Ier–Ve siècles) by Julien Dubouloz.” Bryn Mawr Classical Review
2012.07.36.

Macnair, Mike. 1999. “Vicinage and the antecedents of the jury.” 17 Law and
History Review 537–90.
Magill, Frank N., ed. 1998. Dictionary of World Biography, vol. 1: The
Ancient World. Chicago: Fitzroy Dearborn Publishers.
Martini, Pierluigi. 1976. “A review of the history of the aqueducts of Rome.”
68 Journal of the American Water Works Association 560–7.
Mayer, Ernst Emanuel. 2012. The Ancient Middle Classes: Urban Life and
Aesthetics in the Roman Empire, 100 –250 . Cambridge, MA: Har-
vard University Press.
McAdams, Richard H. 1997. “The origin, development, and regulation of
norms.” 96 Michigan Law Review 338–433.
Millar, Fergus. 1977 [repr. 1991]. The Emperor in the Roman World (31 –
337). London: Duckworth.
Morley, Neville. 1996. Metropolis and Hinterland: The City of Rome and the
Italian Economy 200 – 200. Cambridge: Cambridge University Press.
Morris, Ian. 1999. “Foreword to updated edition,” in Moses I. Finley, The
Ancient Economy, updated edn. Berkeley: University of California Press,
ix–xxxvi.
Morris, Ian, Richard P. Saller, and Walter Scheidel. 2007. “Introduction,” in
Scheidel, Morris, and Saller, eds, 1–12.
Mumford, Lewis. 1961. The City in History. New York: Harcourt, Brace &
World.
Musgrave, Richard A. 1959. Theory of Public Finance: A Study in Public
Economy. New York: McGraw-Hill.
Natelson, Robert G. 1987. “Comments on the historiography of condomin-
ium: the myth of Roman origin.” 12 Oklahoma City Law Review 17–58.
Nicholas, Barry. 2008. An Introduction to Roman Law. Oxford: Oxford
University Press.
Packer, James E. 1971. The Insulae of Imperial Ostia. Memoirs of the
American Academy in Rome, vol. 31. Ann Arbor: University of Michigan
Press.
Parkins, Helen. 1997. “The ‘consumer city’ domesticated? The Roman city in
élite economic strategies,” in Helen M. Parkins, ed., Roman Urbanism:
Beyond the Consumer City. London: Routledge, 83–111.
Peña, J. Theodore. 2007. Roman Pottery in the Archaeological Record. Cam-
Pirson, Felix. 1997. “Rented accommodation at Pompeii,” in Ray Laurence
and Andrew Wallace-Hadrill, eds, Domestic Space in the Roman World:
Pompeii and Beyond. Journal of Roman Archaeology Suppl. 22. Ports-
mouth: Journal of Roman Archaeology, 165–81.
Polanyi, Karl. 1944. The Great Transformation. New York: Rinehart.
Rabin, Edward H. 1984. “The revolution in residential landlord-tenant law:
causes and consequences.” 69 Cornell Law Review 517–84.

Rawson, Elizabeth. 1976. “The Ciceronian aristocracy and its properties,” in
Finley, ed., 85–102.
Reynolds, Susan. 2010. Before Eminent Domain: Toward a History of Expro-
priation of Land for the Common Good. Chapel Hill: University of North
Carolina Press.
Robinson, Olivia F. 1992. Ancient Rome: City Planning and Administration.
London: Routledge.
Rodger, Alan. 1972. Owners and Neighbors in Roman Law. Oxford:
Clarendon Press.
Rodgers, Robert H., ed., 2004. Frontinus, De Aquaeductu Urbis Romae.
Roe, Mark J. 1998. “Backlash.” 98 Columbia Law Review 217–41.
Rosenstein, Nathan 2008. “Agriculture and aristocrats in the middle and late
Republic.” 98 Journal of Roman Studies 1–26.
Rosenthal, Stuart S., and William C. Strange. 2004. “Evidence on the nature
and sources of agglomeration economies,” in J. Vernon Henderson and
Jacques-François Thisse, eds, Handbook of Regional and Urban Econom-
ics, vol. 4. Amsterdam: Elsevier, 2119–71.
Saller, Richard. 2012. “Human capital and economic growth,” in Scheidel,
ed., 71–86.
Saller, Richard P. 2007. “Household and gender,” in Scheidel, Morris, and
Saller, eds, 87–112.
Samuelson, Paul A. 1954. “The pure theory of public expenditure.” 36 Review
of Economics and Statistics 387–9.
Scheidel, Walter. 1999. “Emperors, aristocrats, and the grim reaper: towards
a demographic profile of the Roman elite.” 49 Classical Quarterly 254–81.
Scheidel, Walter. 2005. “Human mobility in Roman Italy, II: the slave
population.” 95 Journal of Roman Studies 64–79.
Scheidel, Walter. 2007. “Demography,” in Scheidel, Morris, and Saller, eds,
38–86.
Scheidel, Walter, ed. 2012a. The Cambridge Companion to the Roman Econ-
omy. Cambridge: Cambridge University Press.
Scheidel, Walter. 2012b. “Slavery,” in Scheidel, ed., 89–113.
Scheidel, Walter, and Steven J. Friesen. 2009. “The size of the economy and
the distribution of income in the Roman empire.” 99 Journal of Roman
Studies 61–91.
Scheidel, Walter, Ian Morris, and Richard Saller, eds. 2007. The Cambridge
Economic History of the Greco-Roman World. Cambridge: Cambridge
University Press.
Schleicher, David. 2010. “The city as a law and economic subject.” 2010
University of Illinois Law Review 1507–63.
Schneider, Helmut. 2007. “Technology,” in Scheidel, Morris, and Saller, eds,
144–71.

Scobie, Alex. 1986. “Slums, sanitation and mortality in the Roman world.” 68
Klio 399–433.
Scogin, Hugh T., Jr. 1990. “Between heaven and man: contract and state in
Han Dynasty China.” 63 Southern California Law Review 1325–1404.
Seneca. 2007. Dialogues and Essays, trans. John Davie. Oxford: Oxford
University Press.
Smith, Adam. 1937. The Wealth of Nations. New York: Modern Library.
Storey, Glenn R. 2002. “Regionaries-type insulae 2: architectural/residential
units at Rome.” 106 American Journal of Archaeology 411–34.
Taylor, Rabun. 2000. Public Needs and Private Pleasures: Water Distribution,
the Tiber River, and the Urban Development of Ancient Rome. Rome:
“L’Erma” di Bretschneider.
University Press.
Tolley, George S. 1974. “The welfare economics of city bigness.” 1 Journal of
Urban Economics 324–45.
Treggiari, Susan. 1999. “The upper-class house as symbol and focus of
emotion in Cicero.” 12 Journal of Roman Archaeology 33–56.
Turchin, Peter, and Walter Scheidel. 2009. “Coin hoards speak of population
declines in ancient Rome.” 106 PNAS 17276–9.
Wallace-Hadrill, Andrew. 1991. “Elites and trade in the Roman town,” in
John Rich and Andrew Wallace-Hadrill, eds, City and Country in the
Ancient World. London: Routledge, 241–72.
Wallace-Hadrill, Andrew. 1994. Houses and Society in Pompeii and Hercu-
laneum. Princeton: Princeton University Press.
Ward-Perkins, Bryan. 2005. The Fall of Rome and the End of Civilization.
New York: Oxford University Press.
Watson, Alan. 1971. Roman Private Law around 200 . Edinburgh: Edin-
burgh University Press.
Watson, Alan, ed. 1985. The Digest of Justinian. Philadelphia: University of
Pennsylvania Press (available for download at Project Muse, http://muse.
jhu.edu/).
Watson, Alan, ed. 1995. The Spirit of Roman Law. Athens: University of
Georgia Press.
Westbrook, Raymond. 1989. “Restrictions on alienation of property in early
Roman law,” in Peter Birks, ed., New Perspectives in the Roman Law of
Property. Oxford: Clarendon Press 207–13.
Westbrook, Raymond. 2003a. “Introduction: the character of ancient Near-
Eastern law,” in Raymond Westbrook, ed., A History of Ancient Near
Eastern Law, vol. 1. Leiden: Brill, 1–90.
Westbrook, Raymond. 2003b. “Mesopotamia: Old Babylonian period,” in
Raymond Westbrook, ed., A History of Ancient Near Eastern Law, vol. 1.
Leiden: Brill, 361–430.

Westbrook, Raymond. 2009a. “The nature and origins of the Twelve Tables,”
in Bruce Wells and Rachel Magdalene, eds, Law from the Tigris to the
Tiber: The Writings of Raymond Westbrook, vol. 1. Winona Lake: Eisen-
brauns, 21–71.
Westbrook, Raymond. 2009b. “Social justice in the ancient Near East,” in
Bruce Wells and Rachel Magdalene, eds, Law from the Tigris to the Tiber:
The Writings of Raymond Westbrook, vol. 1. Winona Lake: Eisenbrauns
143–60.
Whitman, James Q. 1996. “The moral menace of Roman law and the making
of commerce: some Dutch evidence.” 105 Yale Law Journal 1841–89.
Wigmore, John H. 1941. “The pledge-mortgage idea in Roman law: a
revolutionary interpretation.” 36 Illinois Law Review 371–93.
Wilson, Andrew. 2011. “City sizes and urbanization in the Roman empire,”
in Bowman and Wilson, eds, 161–95.
Wirth, Louis. 1938. “Urbanism as a way of life.” 44 American Journal of
Sociology 1–24.
Wittfogel, Karl. 1957. Oriental Despotism: A Comparative Study of Total
Power. New Haven: Yale University Press.
Yavetz, Zvi. 1958. “The living conditions of the urban plebs in republican
Rome.” 17 Latomus 500–17.
18
Land Demarcation in Ancient Rome

Gary D. Libecap and Dean Lueck
18.1. INTRODUCTION
Ancient Rome was a vast social network—lasting nearly one thousand

years—that created wealth by facilitating specialization and trade at
the level not seen in the world before or for many centuries afterward.¹
At its height the Empire comprised over 5 million square kilometers,
or roughly three-fourths of the land area of the continental US and
half of land area of the European continent. For centuries, the popu-
lation of the city of Rome was about 1,000,000.² After the (Western)
Empire fell in the fifth century CE Rome’s population plummeted to
20,000 and did not again reach 1,000,000 until the twentieth century.³
The networks that facilitated wealth generation included common
languages and a common currency, a large standing army (the
legions), a large navy, and a system of law (and public administra-
tion). The Roman Empire also included a system of roads (and
¹ Temin (2006, 2013) provides evidence of wealth and economic growth during the
Empire period and also provides evidence of a wide market for wheat across the
Empire. Arruñada (2012) discusses the importance of contractual registries in foster-
ing networks of impersonal trade, although the evidence for Roman registries is
not clear.
² Alexandria and Carthage were the second and third largest cities in the Roman
world and had populations perhaps as high as 500,000 and 300,000, respectively.
³ We follow the modern scholarly convention of using CE for “common era” rather
than AD, and BCE for “before common era” for BC.
Gary D. Libecap and Dean Lueck, Land Demarcation in Ancient Rome In: Roman Law and Economics:
DOI: 10.1093/oso/9780198787211.003.0018
212 Gary D. Libecap and Dean Lueck

shipping routes) and property rights in land that promoted a vast
market in agricultural products that not only sustained a wealthy
capital city but also a population that, on a conservative estimate,
reached a maximum of about 65 million in the second century CE,
roughly 15 percent of the entire world population (Temin 2013).
Indeed much of the Empire’s agricultural production came not
from the area near Rome, but from North Africa, the Po Valley of
northern Italy, and from Spain and France.
A key component of Rome’s agricultural productivity was its
sophisticated system of land demarcation known as centuriation.
Centuriation was the demarcation of large tracts of land into a
collection of uniform squares (rectangles in some cases) with precise
borders surveyed by highly trained professionals. Vast tracts were
demarcated by Roman surveyors (called agrimensores, literally mean-
ing land measurers) in North Africa and the Po Valley. This demar-
cation typically followed the conquest or annexation of new territory.
Once the land was demarcated it was often colonized by Roman
settlers, who might also serve as soldiers, and put to agricultural use.
Land demarcation has been dominated by indiscriminate or unsys-
tematic patterns of irregularly shaped plots, such as metes and
bounds (Thrower 1966; McEntyre 1978; Linklater 2002; Libecap
and Lueck 2011a, 2011b; Libecap, Lueck, and O’Grady 2011). This
type of demarcation includes nearly all of Europe and the eastern
United States (Thrower 1966; Estopinal 1998; Linklater 2002). Metes
and bounds (MB) systems demarcate boundaries in terms of natural
features of the land (e.g. rivers, cliffs, beaches) and even some rela-
tively permanent human structures (e.g. bridges, walls). The most
famous deviation from metes and bounds is the American rectangu-
lar system (RS) established in 1785, now governing the vast majority
of land in the U.S. (Estopinal 1998; Libecap and Lueck 2011a, 2011b;
Linklater 2002). This was part of a general effort to implement the RS
in the British Empire, the largest empire to follow that of Rome
(Libecap, Lueck, and O’Grady 2011).
The Roman system of land demarcation was a purposeful, highly
centralized and coordinated effort to standardize land into well-
defined and enforced parcels. It was the first widespread system of
rectangular demarcation. Under centuriation land was typically
demarcated land into square units called centuriae quadratae, with
sides of 710 meters (Bradford 1957). Each of these squares was
comprised of 100 square heredia (Johnson 1976). Typically, at the
Land Demarcation in Ancient Rome 213

center of the centuria a north–south axis intersected an east–west
line, making four quarters.⁴
Despite having been established two millennia ago centuriation
still persists in the landscape throughout the Roman world. It is
found, for example, in northern Italy, especially in the Po River
Valley, at Braga in Portugal, Chester in England, Tarragona and
Merida in Spain, Cologne and Trier in Germany (Stanislawski
1946), and Carthage in Tunisia. Figure 18.1 shows a satellite image
of the rectangular centuriation system still present and dominant
near the town of Cesena, Italy (northeast Italy). The image is dom-
inated by squares but the left-hand side of the figure and the far
upper-right corner also shows land demarcated under a non-
rectangular system.
This chapter examines how the centuriation system of land demar-
cation was used and implemented as part of the great Roman
Figure 18.1. Cesena, Italy (44 N, 12 E), showing RS demarcation

persistence.
Map data: Google, DigitalGlobe. Image © 2013 DigitalGlobe.
⁴ Unlike the modern US system, however, the Roman system did not cover
contiguous stretches of land, but was established at various new cross-points and
thus varied somewhat with natural land features. And as we note below there was
considerable variation over time and across space as well.

network. We merge the economics of property rights with the
economics of networks to examine the establishment and evolution
of the Roman centuriation.⁵ Our approach predicts that centuriation
will be adopted in those areas where the potential value from settled
agriculture and colonization will be largest. This suggests that
Romans established the centuriation system in areas where potential
land values were high and where the terrain was relatively flat so that
the costs of implementing the system were relatively low. These are
also the areas where the gains from coordinated demarcation to
promote tillage, drainage, irrigation, land trading, and addressing
for taxation and other purposes were high. We also examine how
changes in political and bureaucratic institutions and markets for
land and agricultural commodities, as well as improvements in survey
techniques influenced Roman land demarcation and its effects.
Until now historians and archaeologists have been the primary
scholars of Roman centuriation and have by far the most well-
developed theories to explain the practice. For example, Larsen
(1938) argued that the practice was derived from political goals to
increase tax revenue and establish colonies. Alternatively Stambaugh
(1988) discussed squares as religious artifacts. Our approach is purely
economic, although consideration of political forces will be important
in the empirical analysis because of their impact on implementation
costs. In this regard we differ from Roman legal scholars in that we do
not rely on verification of motivation from primary sources, such as
the writings of Roman jurists. Instead, we draw functional parallels
between ancient and modern solutions to similar problems, and we
use modern economic theory to inform the evidence on land demar-
cation that comes from a variety of sources.
The chapter is organized as follows. We begin in §18.2 with a brief
history of ancient Rome and its land demarcation systems. In §18.3
we develop an economic framework for analyzing the demarcation of
land under the Roman system. §18.4 is our empirical analysis of the
adoption of the system of centuriation in ancient Rome. In §18.5 we
summarize our findings and discuss their implications.
⁵ The works of Acemoglu and Robinson (2012) and others in the literature on
economic institutions and economic development are related though our approach is
distinct.

18.2. HISTORY OF LAND DEMARCATION
IN ANCIENT ROME
Classical scholars typically divide the history into three periods that
coincide with three distinct political regimes: the Kingdom or Mon-
archy (753–509 BCE), the Republic (509–27 BCE), and the Empire
(27 BCE–476 CE).⁶ The Republic was marked by a political system in
which authority was given to elected magistrates and unelected sen-
ators. During this early period territorial expansion was moderate
with Roman territory consisting of Italy, Spain, Sicily, and parts of
Gaul, North Africa, and the Near East (Cornell and Matthews 1982).
The start of the Imperial period marked an era of more aggressive
territorial expansion with the Empire ultimately spanning about
6.5 million square kilometers (2.5 million square miles) at its peak
under Emperor Trajan.
In the later Empire the focus shifted from expansion to defense
against invading enemies (Adkins and Adkins 2004). Diocletian first
divided the Empire into an eastern and western half in 293 CE and it
was permanently divided in 395 with the establishment of the Byzan-
tine Empire and the Western Roman Empire. In 476 CE, Germanic
invasions caused the fall of the Western Roman Empire,⁷ but the
Byzantine Empire did not fall until 1453 when Constantinople was
conquered by the Ottoman Turks.
Figure 18.2 shows a map of the Roman Empire in the first century
CE when the Empire was at it largest extent. The map reveals the
division of the Empire into provinces such as Africa and the Hispa-
niae (Spain), in addition to Italia (Italy). Table 18.1 contains a sum-
mary of the history of Rome relevant for our study. For each century
it shows the system of government, land area and colonization,
population, and major events.
⁶ We rely on various general sources including Adkins and Adkins (2004). Temin
(2013) also provides a concise economic history of ancient Rome. There is of course
debate over the precise ending of the Empire and this date is specific to the Western
Empire and does not include the Eastern Roman (Byzantine) Empire.
⁷ In the fall of 476 CE the German military office general Odoacer deposed the
emperor Romulus Augustus and became the king of Italy.
Figure 18.2. Ancient Rome at its greatest extent under Trajan (c.117 CE).
Table 18.1. History of ancient Rome.
Century Government Size of area Total Population of Important Events
Population city of Rome
7 BCE Kingdom 2,750,000 sq km 20,000 NA Foundation of Rome

(1,061,781 sq mi)
6 BCE Kingdom NA NA
5 BCE Republic
4 BCE Republic 10,000,000 sq km 165,000* 30,000
(3,861,000 sq mi)
3 BCE Republic 360,000,000 sq km 262,322 150,000 Rome colonizes and conquers Italy
(138,997,000 sq mi)
2 BCE Republic 800,000,000 sq km 214,000 ** NA Annexation of Corsica and Sardinia, First Punic War between
(308,882,000 sq mi) Rome and Carthage, Second Punic War
1 BCE Republic 1,200,000 sq km 394,336 1,000,000 Third Punic War, Conquest of Macedonia, Africa, and Greece
and Empire (463,323 sq mi)
0 Empire 1,950,000 sq km 4,063,000*** NA Octavian named Augustus, first emperor of Rome, annexation
(752,899 sq mi) of Syria
1 CE Empire 2,750,000 sq km 6,944,000 1,000,000 Roman invasion of Britain
(1,061,781 sq mi)
2 CE Empire 4,200,000 sq km NA NA Conquest of Dacia
(1,621,629 sq mi)
3 CE Empire 5,000,000 sq km 56,000,000# 800,000 Abandonment of Dacia
(1,930,511 sq mi)
4 CE Empire 4,400,000 sq km NA 700,000–800,000 Theodosius the Great declares Christianity to be the sole
(1,698,849 sq mi) religion of the empire
Split in West and East Empire
5 CE Empire NA 90,000–150,000 Vandals invade Gaul Fall of (Western) Roman Empire
* These figures do not include the inhabitants of the Latin colonies nor of the allied states.
** The falling off from the number of the preceding census of 220 BCE was a result of the Hannibalic war.
*** These figures and those of the enumerations for 8 BCE and 13 CE are from the Monumentum Ancyranum. The increased number are given by the census of 70 BCE over that of 115
BCE registers the result of the admission to the city of the Italians at the end of the Social war. Available on website http://www.tulane.edu.
18.2.1. The Roman System of Centuriation

The main source of understanding for Roman demarcation comes
from the Corpus Agrimensorum Romanorum (Corpus), a collection of
Roman surveyor manuals copied later into manuscripts by monks.⁸
It is the most important source of information on centuriation.⁹ The
manuscripts that make up the Corpus were written by different
authors through several historical periods. The most important
manuscripts date from the sixth and ninth centuries CE (Dilke 1971:
17). The earliest technical writer of the Corpus is Sextus Julius Fron-
tinus, governor of Britain from 74–8 CE.¹⁰ Below we discuss the key
features of centuriation and Roman land demarcation.
18.2.1.1. The Basic System

The process of centuriation is a system of land division that is built
around two wider central lines or avenues called limites, the cardo
maximus and decumanus maximus, and subsequent narrower limites,
creating a grid pattern.¹¹ There is a sophisticated process of measur-
ing and extensive boundary marking, with a system of measurements
used for land and other assets.
The basic unit of measurement for length was the Roman foot and
the main unit of measurement for area was the actus, with a length of
120 Roman feet on each side. The length that constituted a Roman
foot, however, changed over the course of Roman civilization.¹²
Because the actus was based on the foot, inconsistencies are observed
⁸ We rely on Bradford (1957), Campbell (2000), and Dilke (1971) for our primary
understanding of centuriation and Roman land institutions.
⁹ However, monks who copied the manuals had little understanding of the
technical Latin used to describe Roman surveying procedures (Dilke 1971: 17). As a
result, the contents of the Corpus are not always consistent.
¹⁰ Some of the manuscripts are accompanied by well-preserved and colored mini-
atures. These illustrations served as a tool for teaching Roman surveying students.
Within the manuals, land disputes are addressed, as well as the technicalities and
proper methods associated with measuring, marking, mapping, and allocating land.
¹¹ In Latin, cardo generally referred to a hinge (as in a door) or a pivot (as with a
clamshell), so the application to demarcation and mapping as a north–south or main
axis is intuitive. Among English writers cardo is sometimes spelled “kardo” (e.g. Dilke
1971).
¹² There were three different standards that defined a Roman foot: The early foot,
the normal foot, and the late Roman foot, which came into use in the third century CE
(Dilke 1971: 84). Each new foot was shorter than the previous one.

in the size of the actus at different locations that were centuriated in
different time periods (Dilke 1971: 85). The length of a standard 20 × 20
actus centuria measures 709.68 meters on a side using the normal
foot. The average length of 20 actus in Italy has been calculated to
be about 707 meters, with a range from 704 to 714 meters (ibid.).¹³
Table 18.2 defines the units used in the Roman measurement system.
The actus is the basic unit of land area for the Roman land system,
measuring 14,440 square Roman feet or approximately 0.312 acres.
Two actus made up one iugerum and two iugera made up a heredium,
which was commonly thought of as a reasonable parcel size for an
individual. As Table 18.2 shows a heredium is equivalent to 1.246
acres. The 100 heredia grouped together gave the centuria its name.
A standard centuria consisted of 20 × 20 actus, or 200 iugera. Sides of
centuriae were always multiples of actus, which helps determine if
traces of squares are a part of Roman centuriation; however, there
were non-standard centuriae indicated from the Corpus, and from
observation in the field. Figure 18.3 shows an actus and iugera within
a heredium.
Table 18.2. Roman measurements with US equivalents.

Roman Definition US Equivalent
Measurement
Pes Early Roman foot 11.70 in/29.73 cm

Normal Roman foot 11.64 in/29.57 cm
Late Roman foot 11.58 in/29.42 cm
Passus Paces, 5 Roman feet 58 in/147.85 cm
Mille passus The mile, 100 paces 4833.33 ft
Actus 14,400 sq Roman feet 13,920 sq ft/0.312 acres
Iugerum 2 sq actus, 28,800 sq Roman feet 27,840 sq ft/0.624 acres
Heredium 2 iugera 55,680 sq ft/1.246 acres
Centuria* 200 iugera, 100 heredia 124.6 acres
Saltus Area of woods or pasture, or a continuous 500 acres or 3,115 acres
tract of land, 2 2 centuria or 5 5 centuria
Source: Dilke 1971; Campbell 2000.
¹³ In Tunisia the average is about 708 m, but some areas have been measured as low
as 703 m. The low measurements in Africa reflect the use of the later Roman foot.
Some of the variations seen in Italy, however, may simply be due to errors in
measuring (Dilke 1971: 85).
ACTUS HEREDIUM
IUGERUM
(2 sq. actus)
Panel A: Illustration of an actus Panel B: Depiction of two iugeria

within a heredium
Figure 18.3. Roman system of rectangular demarcation.
18.2.1.2. Orientation and Division

The main lines of survey, the decumanus maximus (DM) and the
cardo maximus (CM), were typically north–south and east–west,
respectively, as calculated by a sundial (Dilke 1971: 86). In the
minority of cases, centuriation was laid out on an alignment other
than the standard N–S orientation. In such cases, alignment was
oriented around existing main roads or geographical features. In a
minority of cases, orientation of demarcated land did not follow these
rules. Some surveyors were influenced by the length of the land,
picking the direction where it was the longest to mark the DM.¹⁴
In the division of the measured land, there were limites (boundary
lines) of different sizes. The DM was the widest, with the CM slightly
narrower. At first, there were no set widths for the limites until
Augustus standardized the system and established that the DM
would be 40 feet in width and the CM 20 feet. Outer limites would
be sighted at a distance of 2,400 Roman feet, or the length of 20 actus,
in each direction from the main intersection. The inner limites can be
referred to as limites intercisivi, and intersect a centuria into multiple
plots. Every fifth main limes was made slightly wider so to ensure that
it would become a usable road. These limites were called quintarii.
Each of these limites was sighted at a right angle and cross-checked to
ensure accuracy.
¹⁴ There is evidence that the origins of such orientations take their root from
Etruscan practices (Dilke 1971: 87, 89), including the origin of the word cardo and
many surveying practices.
18.2.1.3. Allocation, Numbering, and Mapping

Allocating centuriated land was one of the surveyor’s most important
responsibilities. They were required to direct the settlers to their plots
of land to avoid confusion and mistakes. To record the details of the
allocation, surveyors were responsible for creating maps. Some maps
were inscribed on wood tablets, other were inscribed on parchment or
bronze. The government kept the maps of all divided and allocated
lands. If there was ever a dispute, the sanctuarium (where the maps
were kept) could be referenced. Figure 18.4 shows the centuriation
system and also shows the US rectangular system developed more
than two millennia later.¹⁵
In order to keep track of the allocations, it was important to have
a numbering system. The system uses the following abbreviations:
(1) DD, to the right of decumanus; (2) SD, to the left of decumanus;
(3) UC, beyond cardo; and (4) CC, on the near side of cardo.
Following each of the inscriptions would be a Roman numeral indi-
cating the distance from either the CM or DM, depending on what
boundary it refers to. As Figure 18.4 shows, intersection of the decu-
manus and cardo maximus (called the umbilicus) is not at the center
of the settlement. In the Corpus, Hyginus Gromaticus prescribes that
in centuriated colonies, the main intersection of the DM and the CM
should be in the center. In practice, however, this layout was not
common, except in North Africa (Dilke 1971: 88). If an existing road
ran through the planned settlement, then this would often become
the place of intersection (ibid.). In cases of rough terrain, or if
¹⁵ The American RS also begins with the establishment of an Initial Point with a
precise latitude and longitude. Next, a Principal Meridian (a true north–south line)
and a Baseline (an east–west line perpendicular to the meridian) are run through the
Initial Point. On each side of the Principal Meridian, land is divided into square (six
miles by six miles) units called townships. A tier of townships running north and
south is called a “range.” Each township is divided into thirty-six sections; each
section is one mile square and contains 640 acres. These sections are numbered 1 to
36 beginning in the northeast corner of the township. Each section can be subdivided
into halves and quarters (or aliquot parts). Each quarter section (160 acres) is
identified by a compass direction (NE, SE, SW, NW). Each township is identified by
its location relative to the Principal Meridian and Baseline. For example, the Seventh
Township north of the baseline and Third Township west of the First Principal
Meridian would be T7N, R3W, First Principal Meridian. In this manner, properties
are positioned relative to one another in a standardized way. Dilke (1971) discusses
the likelihood that Thomas Jefferson, who was the prime architect of the US system,
based it on the Roman system.
EXAMPLE 3
SECTION DETAIL
D 6 5 4 3 2 1
E
7 8 9 10 11 12
U quintarius
C 18 17 16 15 14 13
TOWNSHIP 1 NORTH. RANGE
M 1 WEST. OILA & SALT RIVER
19 20 21 22 23 24 BASELINE & MCHIDIAN
A
N 30 29 28 27 26 25
U 36
S 31 32 33 34 35
M
A
X
SD I DD I DD II
I
VK II VK II VK II
M
U
SD I DD I DD II
VK I S
VK I VK I
K AR D O MA XI MU S
SD I DD I DD II
CK I CK I CK I Direction of
Sighting
SD I DD I DD II
CK II CK II CK II
quintarius
Figure 18.4. Roman (left, a) and US (right, b) rectangular demarcation systems.


centuriation was being implemented on an existing settlement, the
surveyor would often prefer to make a fresh start at a point not far
outside the existing or planned settlement (ibid. 88–9). In cases where
mountains, rivers or some other major obstruction impeded centur-
iation, the intersection could be at a distant point (ibid. 89).
Roman survey techniques and the administrative bureaucracy that
made such sophisticated demarcation possible also gradually improved
as the empire expanded. These allowed for more accurate measurement
and extension of the RS. Roman survey and administration were unpre-
cedented in their precision and scope in the Empire. With the fall of
Rome many of these practices were lost until redeveloped and aug-
mented in the late eighteenth and nineteenth centuries in Britain and
the United States (Libecap, Lueck, O’Grady 2011).
18.2.1.4. Agrimensores and Surveying

In the first century BCE, a surveyor was also referred to as metator or a
mensor, both of which mean measurer (Dilke 1971: 37). Colonization
under Julius Caesar led to early growth of surveying. The need for
more land for citizens and veteran soldiers, especially in Augustus’
time, encouraged the natural growth and development of centuria-
tion. It was under Augustus’ rule that the first steps leading to the
creation of a large bureaucracy were taken (ibid.). Conquered land
was resurveyed and measurements standardized; 40 feet for DM;
20 feet for CM; for other decumani and cardines 12 feet; and for
subsidiary roads 8 feet. Augustus’ successors continued to settle
veteran soldiers on newly acquired lands. After the Empire’s expan-
sion halted less new land was added and fewer new soldiers required
settlement. Accordingly, the need for land reallocation and the role of
surveyors and administrators changed.
With the development of centuriation and distribution procedures,
there were also advancements in the surveying. At first, agrimensores
were not hired, but instead offered their services and expertise as a
favor. Any compensation that they received was optional. Later, as
the quantity of new land grew, Roman influence spread, and the
number of soldiers and settlers seeking parcels for settlement
increased, the demand for surveyors rose. The Roman state began
to view surveyors as professionals with certain expertise that could be
held accountable. If a surveyor made an incorrect measurement, one
could take legal action against him.

During the imperial period, the establishment of new colonies was
less common, so that the quantity of land needing to be surveyed
began to decrease while the frequency of legal disputes increased. The
rise in land disputes corresponds with population increase and grow-
ing land scarcity. As a result, the role of surveyors shifted to serving as
judges or arbitrators in land law. When there was a land dispute, it
was the provincial governor’s responsibility to send out surveyors, or
they could be employed privately. Surveyors were expected to know
the extensive details of land law, to analyze land disputes, and to give
advice to judges.
The two instruments commonly used and improved over time by
surveyors were the groma and the perticae. The groma was mainly
constructed out of wood, so there are limited remains of this tool that
have survived to modern times. The Corpus does not give a picture of
the groma, consequently, scholars have had limited resources to aid
them in the creation of a model. The only example of what might be a
true groma is one whose metal parts were located in Pompeii in 1912
(ibid. 15).
The groma was a staff, around two meters long with a cross
mounted at the top and, likely, a tripod or sharp point at its base
(ibid. 69–70). The arms of the cross were wooden with metal sheeting.
To prevent inaccuracy due to the wearing of wood, the arms were
reinforced near the center of the cross by bronze angle brackets.
A plumb line hung through holes at the end of each arm. Sighting
was done by lining up opposite plumb bobs. Sighting using the plumb
lines required that the cross be off-centered from the staff to avoid
obstruction. To achieve this task, the cross was placed on a bracket
which projected out from the top of the staff. The distance from the
center of the cross to the staff was 23.5 cm (9.25 in). Once plumb bobs
were aligned, sights were set on a second groma, positioned a speci-
fied distance away (such as one actus). Perticae were measuring rods
that allowed surveyors to measure standardized distances (ibid. 73).
Sights were set on a third groma the same distance away, making a
right angle with the other two gromae, creating a square.
The groma allowed for the measurement of straight lines, squares
and rectangles as required by Roman surveying for RS. A surveyor
with a groma, plumb-line level, sundial, perticae, and writing and
drawing materials could trace out rectangular plots for land demar-
cation. Figure 18.5 shows how Roman scholars believe the groma was
designed and used.
Figure 18.5. Depiction of a groma and mensor (surveyor).
A necessary skill for the agrimensor was that of sighting and

leveling through use of the groma. There were multiple cross-checks
in order to ensure accuracy. Cross-checks were scheduled at every
fifth intersection, or quintarius. If there was a defect in the groma, or
the sighting, a wrong observation would be recognized and could be
corrected by resurveying. Uneven land and rivers, however, made
measuring rectangles more complex and the RS more costly.
18.2.1.5. Boundaries
Boundaries were an important part of centuriation and were marked
by stones. Land boundaries (termini) had religious connections for the
Romans (Dilke 1971: 98). They were named for the god of boundary
stones, Terminus (Adkins and Adkins 2004: 304). To move a bound-
ary stone without permission was considered not only a civil offense
but also a religious one. As noted above when boundary disputes
occurred, the surveyor acted as a judge or arbitrator.

18.3. ECONOMICS OF LAND
DEMARCATION SYSTEMS
The discussion of the history of Roman land demarcation suggests a

number of questions, including what determines when centuriation
will be chosen and implemented; what determines the characteristics
of a centuria; and what are the effects of centuriation on land use and
land markets. We describe an economic framework to examine these
and other related questions and derive implications for empirical
analysis. Our approach is to distinguish the Roman centuriation
from non-systematic or metes and bounds demarcation. We use RS
to represent a centuriated system (RS for Roman system or rectangu-
lar system) and MB for non-Roman demarcation (or metes and
bounds). In the context of Roman land, MB refers generally to a
pre-existing, often uncoordinated local demarcation practice.¹⁶ We
briefly examine the choice between RS and MB as well as the optimal
configuration of land parcels, land disputes, land markets, land val-
ues, and long-term land value and use under both MB and RS.¹⁷
18.3.1. Demarcation Basics: Parcel Shapes, Sizes,

Alignment, and Topography
We begin by examining a simple decentralized demarcation problem.
In our framework, the individual parcel demarcation decision will
depend on both the expected value of plot productivity, demarcation
costs, and the demarcation regime. The innate characteristics of the
land will partly determine its productivity and the cost of demarcat-
ing each parcel. In particular, topography or ruggedness plays a
critical role. More rugged terrain will have lower productivity and
higher surveying and policing costs. The demarcation regime will
constrain the size, shape, and alignment (i.e. spatial or directional
orientation) of the parcels.¹⁸
¹⁶ See Dilke (1971) for a discussion of pre-Roman demarcation systems.

¹⁷ We draw on the analysis in Libecap and Lueck (2011b) and Libecap, Lueck,
and O’Grady (2011). We later consider the choice of leaving a tract of land
un-demarcated.
¹⁸ It is also likely the RS affected irrigation systems by lowering the costs of moving
water and coordinating landowners, though we do not collect evidence on irrigation.

We can thus define the net value of a parcel of land as a function of
area, shape, perimeter, topographic features, and the demarcation
regime. Similarly, demarcation costs for land parcels are a function
of area, parcel perimeter, topography, and demarcation regime (RS or
MB). Topography and the demarcation system will affect parcel size
and perimeter choices, and parcel alignment. They will also directly
affect both value and demarcation costs. Under each regime initial
landholders pick a parcel size and shape to maximize land value.
Under MB land claimants make these selections independently, based
on their own net value calculations, while under RS parcel demarca-
tion is predetermined and generally rectangular parcels are presented
to claimants. Accordingly, the question facing Roman administrators
was which demarcation system provided for the highest value of the
land for claimants, and hence, highest productive value for the
Empire. Once chosen, the demarcation system was implanted on
lands acquired by the expanding Empire.
The next section examines the implications of these demarcation
choices as well as the implications for land markets and values
generated by the different demarcation regimes under varying top-
ography. As we describe, a Roman administrator may choose to
impose RS or utilize the existing MB arrangement.
18.3.2. Initial Parcel Demarcation

To consider the comparative implications of the two land demarcation
systems, we make the following assumptions. First, we assume that
under each exogenously imposed (by Rome) demarcation regime there
is a large tract of land whose external boundary is enforced collectively
or otherwise by the sovereign. Second, within this large tract, there is a
group of non-cooperative agents who claim and enforce separate plots
in order to maximize the value of their land, net of demarcation and
enforcement costs. Each claimant can only choose and demarcate a
single parcel. Within the external borders, there is no coordination or
contracting among claimants.¹⁹
¹⁹ We assume simultaneous claiming and we ignore the optimal time to claim

under first possession rules (Lueck 1995).
18.3.2.1. Demarcation under Non-Roman Systems (MB)

Under non-Roman local demarcation, or MB, each claimant chooses
parcel size and the length of parcel boundaries to enforce in order
maximize land value less demarcation costs. All parcels fill the new
area acquired by the Empire. Under MB there are few central con-
straints on the size and shape of individual claims, and no mechanism
to coordinate alignment of individual parcels. The solution to the
parcel demarcation problem facing each claimant will depend on the
structure of value and cost functions, and importantly, on how
topography affects them. For example, rugged terrain will affect
agricultural potential and boundary marking and observation costs.
In the simple case in which claimants have the same productivity and
the same enforcement costs, and value does not depend on topog-
raphy or shape, the claimant’s problem is simply to minimize border
demarcation and enforcement costs, constrained by the productivity
of the land.
Because we assume a total land constraint and that all land is to be
included in tangent parcels, and we further note that rectilinear plots
have important production advantages for agriculture and urban use
(Lee and Sallee 1974; Amiama et al. 2008), we predict square parcels
in flat areas for MB. Squares fill the interstitial space or gaps between
parcels.²⁰ They have relatively low perimeter-to-area ratios and thus
have low demarcation and enforcement costs. Survey and fencing
(enclosure) costs are lower for plots with fewer angles and longer
straight boundaries (Johnson 1976).
Deviating from flat topography, however, will alter the net value
function by affecting both land value and demarcation costs. If
demarcation and enforcement costs (surveying-fencing-policing)
and land value depend on terrain, borders will roughly follow topog-
raphy under non-Roman demarcation (e.g. follow streams and ridge
tops rather than cross them). We expect a pattern of parcel sizes and
shapes that depends on the character of the land and of the potential
of claimants (farming productivity, violence and monitoring
²⁰ Regular polygons maximize the area enclosed by a given perimeter and thus
have the lowest p/a ratio for any n-sided polygon but only three regular polygons—
triangles, rectangles (squares), and hexagons—can create patterns with a common
vertex and have no interstitial space. Eliminating interstitial space means less open
access waste and fewer conflicts in the future.

productivity). Non-Roman or MB demarcation is flexible and allows
parcels to be adapted to terrain in order to maximize land values.
In addition, because MB does not coordinate or otherwise con-
strain the spatial structure of individual parcels, alignment will be
solely the choice of the individual claimant. Under MB, the lack of
coordination among claimants means that individual parcels will
not have a common alignment. Even in the case where the land is
flat and the optimal shapes are squares, there is no reason to expect a
grid of squares. The coordinated grid is a public good that will not
be provided individually, unless there is an authority coordinating the
claimants’ actions, perhaps in the form of a developer or, in the case
of Rome, a sovereign.
18.3.2.2. Demarcation under the Roman System (RS)

Initial demarcation under RS is distinct from that under
MB. The land is surveyed by the sovereign into squares prior to
original allocation. Unlike MB, claimants under RS do not explicitly
choose area and perimeter, but are constrained by the system so that
the choice under RS is to pick a (square) parcel that maximizes
difference between land value and acquisition costs. Demarcation
costs are born initially by the sovereign and hence are zero for the
claimant. Because, however, RS demarcation does not allow for
deviation from squares, the value of parcels will decrease and demar-
cation costs will rise with more rugged topography.
18.3.3. Aggregation: Total Value and Land Markets

The previous analysis considered initial demarcation of individual
parcels under MB and RS. In this section we examine the net value of
aggregated land under both systems and also examine how each
system affects the market for land after initial demarcation.
18.3.3.1. Total Value of the Land in a System

As above, we assume the region governed by either a pre-Roman MB
or RS is a fixed area, totally partitioned into parcels. Some areas
acquired by Rome might have had zero net values after consideration
of agricultural productivity and demarcation costs, and hence

remained unclaimed and not demarcated. We incorporate a temporal
dimension to account for difference in demarcation setup and con-
tinuation costs for the two regimes. Under MB the net value of the
land is the sum of individual values and costs, less the continuing
costs associated with individual parcel adjustments resulting from the
lack of coordination. These include individual enforcement efforts,
responding to parcel border disputes, search, and other costs associ-
ated with misaligned plots. Under MB there is a one-time demarca-
tion cost born by each claimant. In addition, land demarcation and
production begin immediately, and the continuing costs associated
with MB are assumed to be increasing in proportion the size of the
region, parcel size, topography, and to rise over time as these prob-
lems accumulate.
The total value under RS has a different structure from MB
because it is a coordinated, imposed regime. We assume the network
effects of RS are such that any person’s or group’s use of the system
also benefits others and that it further increases the incentive to
participate (Baird, Gertner, and Picker 1994; Farrell and Klemperer
2007). The network benefits are the public goods of common
addresses, survey coordination, and standardized, aligned and fixed
parcel boundaries that reduce border disputes and expand the land
market because parcels are of uniform size and shape and can be
more easily located than with MB. These network and coordination
benefits come, however, at the cost of a necessarily extensive system.²¹
Under RS there are upfront costs of design, survey, and controlling
land access until demarcation is completed.
Under these assumptions the total present value of the land in the
RS region is a function of parcel square shapes, sizes, and perimeters;
topography; and the discount rate. The RS system setup costs occur
prior to claiming and use. Network effects are incorporated into the
parcel value function, which is increasing in the number of parcels
governed by the RS. RS system costs are increasing in proportion to
the size of the area to be surveyed and demarcated, but at a decreasing
rate, revealing network economies. RS costs are increasing because of
topography and become greater than MB because of the square-
parcel constraint. It is possible to extend the network effects by
²¹ Our MB–RS cost distinction is similar to Dixit’s (2003) distinction between local
(informal) and large (formal-legal) trading systems, where the latter have greater
setup costs like RS.

assuming that system costs fall over time as RS is implemented in
many other areas, while there are also improvements in survey
technique and system administration. Under RS surveying occurs
prior to parcel selection, so the time horizon for generating value
from the land begins after land has been allocated.
18.3.3.2. Net Value of the Roman System

This framework implies that it is efficient to implement RS when it
provides greater economic value to Rome than does the existing local
(MB) system. This will be the case in some settings but not in others.
Here we focus on the determinants of the potential net gains of the RS
relative to MB in order to derive testable implications for analysis.
Total value of land under each system in a particular area is the sum
of the values of the individual parcels. If the land is flat, then central-
ized RS parcel shapes and sizes are the same as would be chosen
under decentralized MB, although alignment would differ and there
would be no coordinated addressing under MB. The net gain of RS
over MB for a particular region is determined by four factors: (1) the
increase in land value under RS compared to MB from the time that
the RS is implemented, arising from the productivity advantages of
squares, addressing, alignment, and reduced parcel border disputes;
(2) the foregone land rent during the setup period and setup costs
during that period assumed by the sovereign; (3) the one time MB
demarcation cost assumed by individuals; and (4) the continuing MB
costs as described earlier.
These four factors constitute the tradeoffs between RS and MB.²²
The first focuses on the network gains from RS over MB; the second
on the gains from MB that would be sacrificed during the period the
RS is being implemented, in terms of output under MB and RS setup
costs; the third is the foregone individual demarcation costs under
MB not required under RS; and the fourth is the avoided continued
costs of MB over the time horizon of Roman land control. Compara-
tives statics emerge: the net value of RS will increase in the size of the
governed land area; increase in the expected time horizon for captur-
ing net rents; decrease in the time of RS implementation facilitated by
flat land; and increase with improvements in survey technology and
²² A similar analysis can be undertaken for the pre-Roman choice to demarcate

using MB or to leave the land un-demarcated.

administration that occurred over time. A number of predictions
follow:
Prediction 1: A rectangular system is more likely to be adopted
when (a) Rome controls large tracts of land for RS implementation;
(b) when the time horizon is longer; (c) when implementation can
be rapid; and (d) when network effects are larger.
Considering forces likely to change the model parameters can illu-
minate these predictions. For instance, more rugged topography
would reduce net gains from RS by increasing the costs and time of
RS implementation and perhaps even by reducing the losses of sub-
optimal parcel shape. Similarly, we expect that implementing RS in a
region where no incumbent demarcation system existed would lower
costs. There would be no resistance from incumbent land holders to
the demarcation change and no need to move or replace such
demarcation-specific investments as road, fences, buildings, and
canals. Where the topography dictates benefits from coordinated
drainage of land (e.g. a large flat swampy area), then RS is more likely
to be implemented. Finally, political authority and stability will
increase the expected time horizon and make RS adoption more likely.
18.3.3.3. Land Markets: Roman (RS) versus

Non-Roman Demarcation (MB)
RS has economic implications for land markets once it has been
adopted. Because parcel boundaries are standardized and aligned,
there are fewer overlapping borders and unclaimed gaps outside
property descriptions. These factors imply another prediction:
Prediction 2: There will be fewer legal disputes (and litigation)
over boundaries under the RS than under non-Roman demarcation.
Centralized RS demarcation with uniformly defined parcels and
common addressing lowers the cost of using the market and allows
for reorganization of plots as conditions change (Barzel 1982). This
should be observed as a greater number of market transactions under
RS than MB and an increase in the value of land.²³ The following
predictions are implied by this analysis.
²³ Over time the Romans used different size distributions. Subsequent market
activity, however, would lead to subdivision or aggregation of rectilinear plots to

Prediction 3: There will be higher land values under RS than MB.²⁴
Prediction 4: There will be more land transactions under RS than
under MB and parcel size will have a lower variance under RS
and MB.
Prediction 5: There will be higher portion of land improved
under RS than MB.
The coordinated clarity of RS is also expected to have an impact on
public infrastructure, such as roads, drainage ditches, and irrigation
canals, that require long rights-of-way and coordination among
owners. Contiguous linear borders should lower the cost of assem-
bling rights of way and coordination along parcel boundaries.²⁵ This
implies another prediction:
Prediction 6: There will be more roads and canals for drainage
and irrigation per unit of land under the rectangular system than
under metes and bounds.
18.4. EMPIRICAL ANALYSIS
In this section we use historical information and data on specific

centuriation systems to test some of the predictions about the use and
characteristics of centuriae and their effects on economic activity in
ancient Rome. As we noted in the introduction, we interpret the
evidence in light of our economic framework rather than look for
verification in primary Roman legal sources.
18.4.1. Detailed Data on Centuria

Table 18.3 shows currently available data from Roman Italia (Italy).
The information includes the name and location of the centuriae. It
achieve optimal production sizes and we argue that the RS would smooth these
market transactions.
²⁴ As indicate above, we also predict that under RS land values will be decreasing in
the ruggedness of the topography of the land and that there would be a break-even
point beyond which the gains of the RS would be offset by the costs of rigidness.
²⁵ The value of straight roads and their benefits for surveyors and civil engineers is
discussed by Johnson (1976: 167).
Table 18.3. Characteristics of selected Roman centuriae in Italia.

Roman name date of Roman Pre-Roman Size of centuria Alignment
foundation settlement
Aecae 214 BC yes Via Traiana

Alba Fucens 303 BC yes mountains
Allifae 241 BC
Aquileia 181 BC 12 12 actus road
Aquinum 100 BC yes 21 20 actus
Arimimium 268 BC yes
(Ariminum)
Asculum 269 BC
Aug. Bagiennorum 1 century BC
Augusta Praetoria 25 BC yes 20 16 actus
Salassorum
Augusta Taurinorum 28 BC 21 20 actus
Belunum 181 BC 20 20 actus
Beneventum 268 BC yes 25 16 actus
Bononia 189 BC yes 20 20 actus Via Aemilia
Brixia 1 century BC
Caesena 3 century BC yes 20 20 actus North-South
Capua 312 BC yes 20 20 actus
Cosa 273 BC 16 32 actus mountains
Cremona 218 BC yes 21 20 actus
Fanum 1 century BC coast line
Ferentinum 2 century BC
Florentia 59 BC 20 20 actus North-South
Formiae 3 century BC
Grumentum 3 century BC
lulia Concordia 42 BC 20 18 actus
Lazio 326 BC
Luca 180 BC North-South
Luceria 321 BC 80 16 actus
Luni 177 BC
Minturnae 295 BC yes river
Norba 2 century BC mountains
Ostia 338 BC 193.94 125.70 river
meters
Parma 183 BC 20 20 actus Via Aemilia
Patavium 45 BC yes 20 20 actus North-South
Pisarum 184 BC coast line
Pompeii 4 century BC yes 18 23 actus
Praeneste 90 BC
Puteoli 194 BC yes
Sena Gallica 283 BC
Surrentum 1 century BC yes
Tarracina 509 BC Via Appia
Venaf rum 2 century BC 16 16 actus
Venetia Via Postumia
Verona 75–50 BC
Vicetia 135 BC 20 20 actus
Source: see text.

Figure 18.6. Documented location of centuriae in ancient Rome.

Source: Museo della Centuriazione.
also includes information about the size of the centuriae, the date of
establishment, prior settlement, and the type of centuriation (town,
colony, military).
As noted above, prediction 3 states that large landowners or sov-
ereigns are more likely to adopt a centralized rectangular system
because it provides the public goods of systematic location of prop-
erties, coordinated survey, reduced title conflict, and greater infra-
structure investment. Figure 18.6 shows documented centuriae based
on information from Museo della Centuriazione Romana (Museum
of Roman Centuriation) in Borgoricco, Italy.²⁶ Figure 18.7 shows the
locations of centuriae in the Roman Italy.
As Table 18.3 and Figures 18.6 and 18.7 suggest, the locations of
centuriae can be determined and linked to data as described above.
²⁶ See http://www.euromuse.net/en/museums/museum/view-m/museo-della-
centuriazione-romana/ accessed February 5, 2014.
Figure 18.7. Roman centuriation in Italia.
Flatter land with higher quality soils is more likely to be centuriated

than other lands. A simple examination of the location of the
centuriae in Figure 18.7 provides evidence consistent with this pre-
diction as the highest density of centuriae is in the flat fertile Po River
Valley. An absence of centuriae in the Alps and Apennines is evident
as well.
18.4.2. Historical Information on Centuriation

Many of the implications from the model can be tested against
information from the scholarly literature on ancient Rome and
scholars’ understanding of centuriation and how it was linked to
the market for agricultural products. As Temin (2006, 2013) notes,
there is considerable information on ancient Rome but relatively
little that can be used in standard econometric analysis. In this
section we provide details of centuriation that are consistent with the
prediction.
18.4.2.1. Agrimensor Responsibilities and Skills

For centuriation to create an efficient network the system of survey-
ing must have been well designed and consistently applied. Surveyors
measured the land, decided boundaries, distributed individual allo-
cations, directed settlers to their parcels, made maps, recorded land
transactions, and dealt with disputes. As mentioned before, they were
at the center of the most important activities in the Roman Empire.
In order to fulfill these responsibilities significant education was
required. The skills that surveyors needed to master included: orien-
tation, sighting and plotting lines, accurate measurement, boundary
marking and recognition, and legal knowledge. Many of these skills
were best learned by observing other skilled surveyors, so in many
cases those who wanted to become surveyors contacted an established
surveyor in the community to become fully trained in geometry,
measurements, and tax calculations, all of which a surveyor needed
on a regular basis (Dilke 1971: 51–6, 87). A surveyor was required to
be able to determine orientation using the equipment described
above. Usually in centuriation, sundials and compass points were
employed to establish north–south orientation. This practice was
most likely to be followed for military sites and remote settlements.
In some cases, orientations of centuriae were fixed by reference to
existing roads or geographical features. Two examples include Dal-
matia (modern day Croatia), where most centuriation conforms to
the coastline, and the Po valley, where the decumanus was the Via
Aemilia, the major Roman trunk road in that region.
As predicted, over time the operations of the Empire became
increasingly bureaucratic, including those of the agrimensor. Agri-
mensores were given the responsibility of handling property disputes
requiring considerable legal knowledge. Surveyors needed to be
familiar with both the body of law governing the classification of
land and the body of law concerned with boundaries and boundary
disputes. Even though legal definitions were important in property
lawsuits, surveyors had to apply reason and make allowances for
custom. These surveying and bureaucratic improvements suggest
that RS would be adopted more over time.
18.4.2.2. Optimal Alignment

Parcel alignment was an important feature of the RS. In the American RS
all land was demarcated and all land was aligned along a north–south

axis. The Romans did not implement a contiguous RS over their
entire empire, presumably because it was not efficient to have a
uniform alignment rule. Instead each centuriation system was an
independent demarcated region surrounded by lands that were either
un-demarcated or demarcated in a local MB system. As noted above,
the default alignment was north–south, but depending on the
topography a north–south alignment might be more costly than an
alternative orientation. In particular, if there was an existing trans-
portation corridor (road or river), it was cheaper to align the centur-
iae perpendicularly to the corridor rather than north–south. Such an
alignment preserves optimal square parcels along the river or the road
system (or along a coast). Because such a centuriation system is
generally isolated, it will not create alignment conflicts with other
lands. Figure 18.8 shows possible alignment outcomes and illustrates
how alignment can depend on the topography of the land.
Although we have no predictions about the alignment, the issue is a
good example of how important coordination was for the RS and how
land demarcation followed a maximization pattern. The historical
record offers evidence consistent with the hypothesis that alignment
depended on the topography of the land. Dilke (1971: 87) lists
A: oriented N/S C River
B: oriented along river parallel
C: oriented N/S
D: oriented along river parallel
Figure 18.8. Hypothetical Roman centuriation along a river.


settlements that conformed to a non-standard orientation:²⁷ “at Tar-
racina (now Terracina) the decumanus was the Via Appia; in much of
the Po valley it was the Via Aemilia; at Acelum (Asolo) it was the Via
Postumaia. In the area round Florence the topography, particularly as
regards rivers, dictated the alignment of the centuriation. . . . In some
coastal areas the alignment of the coast affected the centuriation; thus
much of the Dalmatian coast runs roughly NW–SE, and at Iader
(Zara) the centuriation was made to conform to this coastline, even
spanning the coastal indentations.”
18.4.2.3. Types of Land

There are different categories of land in ancient Rome, which allows
us to examine the determinants of these categories that were gov-
erned by different legal rules and by different demarcation systems.
The discussion here that follows Campbell (2000) indicates that more
valuable land tends to be governed by the RS. In particular Campbell
(ibid.) discusses several types of land:
Land that was divided and allocated. There are two categories, land
contained within limites and land allocated by means of straight-line
boundaries, rigores. Land that received limites is centuriated land,
which Frontinus calls land within decumani and cardines (ibid. 3).
Land that was divided by rigores lengthwise is said to be separated by
strigae, while land divided breadthwise is called by scamna. The
choice of demarcating land by strigae or scamna was based on long-
established custom in the way in which public arable land in Italy was
cultivated. RS was usual in these settings via centuriation.
Land that was contained in a survey throughout its extent. This land
was allocated to a community and in some provinces a tax was
imposed as a lump sum on the community as a whole (ibid. 319),
which explains the importance of accuracy and thoroughness in
demarcating this type of land. The land of private individuals was
also surveyed on the same principle (ibid. 3). It was common for a
surveyor to enter land to map it according to how it had been divided
and allocated, even though only its boundaries had been surveyed. RS
was often used through centuriation.
²⁷ The Via Appia, Via Aemilia, and Via Postumaia were major trunk roads in the
Roman system of roads.

Land of uncertain boundary not contained in any survey (arcifinius).
The term ager arcifinius (also called as ager occupatorius, ibid. 3, 472)
refers to land that was not surveyed but was bound by rivers, ditches,
roads, or mountains. This land and these borders were given bound-
ary markers over time because of the frequency of legal disputes in the
places where the territory ended. Ager arcifinius was often land that
had been newly conquered by the Romans, but which had not been
properly annexed and allocated. This may have resulted from a lack
of resources to do so, or because the land was very remote as well as
potentially of low value (ibid. 472). The first settlers of these remote
areas were not organized by the central government in Rome, but,
instead, were individuals or small groups of Roman farmers who
migrated to the region. These early settlers were frontiersmen who
were constantly faced with the possibility of violent confrontation
from the recently expelled occupants. To mitigate the risk of attack
the early settlers often demarcated their land holdings along
natural defenses such as streams and hills. Even so, there was little
formal demarcation because of the lack of central survey. MB was
common when there was no official survey or presence of the legion
to enforce RS.
When an area was surveyed and the land within it divided by limites
into rectilinear plots, there was often land left over that was not
divided into centuriae. This type of land was classified under the
general term subsecivum. The word subsecivum is derived from the
word subseco, meaning the line that cuts it away. Campbell (ibid.)
notes two types of subseciva that occurred in centuriated settlements.
The first type was the area on the outer boundaries of lands allocated
for centuriation, but where square centuriae could not be completed.
The second type was land that was in the middle of a centuriated area
but that is not allocated and was marked off by a line. In non-
centuriated settlements, land that lay between the outer boundaries
of a settlement and rectilinear divisions was also called subsecivum
(ibid. 3). Land designated as subsecivum was at the disposal of the
founder of the colony (ibid. 321). The surveyor also marked some
land as public areas such as woods and pastures (Dilke 1971: 107).
These lands were not available for private ownership, but rather were
left for firewood for the public bathhouses or cemeteries for the poor
(ibid.). As a result, in these areas there was a mix of MB and RS as well
as RS centuriation that did not involve rectangular parcels.
18.4.2.4. Long-Term Effects and Path Dependence

One of the most striking features of Roman centuriation is its aston-
ishing persistence. Perhaps no other institution shows the path
dependence of land demarcation. Governments come and go. Cities
rise and fall. Twenty centuries after the Roman agrimensores created a
grid of square parcels, the land is still farmed in rectilinear fields and
the roads run straight with right angles as they did under Roman rule.
As John Bradford (1957) describes the modern traces of ancient
systems of centuriation:
The forceful imprint of the elaborate gridded road-systems which
betoken it can be still traced across some thousands of square miles
on both sides of the Mediterranean. In origin, most of these systems
were carved out of territories raw from conquest, and even now, in
retrospect, their appearance deeply stirs the imagination,—so boldly
artificial was the conception and drastic the creation as compared with
any earlier man-made landscape in this region. Such a purposeful
regular dissection of land for cultivation was not again matched in
formal precision until comparatively recent times: e.g. in the settlement
of North America or in the systematization of the Great Alfold after the
Ottoman tide had receded from Hungary in the 18th century.
Bradford (1957: 145, 154)
The satellite photograph of the area around Cesena, Italy
(Figure 18.1) is an example. In fact, in the modern era, the vast extent
of centuriation was first noticed by pilots during World War II
(Bradford 1957) and ultimately led to the field of aerial archaeology.²⁸
Another example is Carthage, Tunisia. Figure 18.9 shows a modern
satellite photograph of the city and surrounding countryside and its
unmistakable Roman demarcation. Carthage, according to tradition,
was founded in 814 BCE and was the capital of an extensive empire
during the early and middle years of the Roman Republic. The regime
was often in conflict with Rome and the city was ultimately destroyed
by Romans in 146 BCE after the Third Punic War. Not only did the
Romans burn the city and destroy surrounding farmland, they also
re-demarcated the land using the centuriation system.²⁹ Soon
²⁸ Libecap and Lueck (2011b) find dramatic persistence of demarcation in Ohio

(USA) and in the differential value of the land under RS and MB systems.
²⁹ The Romans also extensively centuriated farmland in North Africa and this area
was an important source of wheat over the centuries.
Figure 18.9. Carthage, Tunisia (36 N, 10 E), showing RS demarcation

persistence.
Source: Google Earth.
thereafter Carthage became one of the largest cities in the Roman

world, with a population of hundreds of thousands. As Figure 18.9
shows, the Roman grid system is still intact over two thousand years
later. It goes without saying that since the Romans implemented their
centuriation system much has changed in Carthage, but not the
demarcation of the land.
18.5. CONCLUDING REMARKS
We study the demarcation of land in Ancient Rome in order to

examine the choice between centralized and decentralized coordin-
ating property institutions.³⁰
This chapter is the first systematic economic analysis of land
demarcation outside the United States, examining the two dominant
systems in ancient Rome. Our findings increase our understanding of
a fundamental feature of property institutions. In particular, we find
³⁰ As we have stressed, we do not claim that Roman jurists conceptualized such

problems in the same terms as we do.

that the implementation of the rectangular system of land demarca-
tion known as centuriation is consistent with a wealth-maximizing
model; and that such demarcation institutions have a remarkable
persistence and impact on land use and land markets. Both findings
suggest the importance of initial institutional arrangements. Still this
chapter leaves many questions unanswered and available for further
study, including the role of political institutions on the use and
structure of centuriation, the impact of centuriation on the develop-
ment of the Roman economy, and the effects of the decline in the
Empire on the use of centuriation in peripheral regions.
This study of land demarcation is connected to a growing literature
on the nature of property rights in contributing to different patterns
of economic growth across countries (de Soto 1989; North 1990;
Acemoglu and Robinson 2012). It is also linked to an empirical
literature on property rights with application to natural resources
(Bohn and Deacon 2000; Libecap and Smith 2002), American Indian
reservation agriculture (Anderson and Lueck 1992), and urban resi-
dential land development (Miceli et al. 2002). It thus informs the
larger issues of institutions and economic development.³¹
REFERENCES
Acemoglu, Daron, and James A. Robinson. 2012. Why Nations Fail: The
Origins of Power, Prosperity, and Poverty. New York: Crown Publishers.
Adkins, Lesley, and Roy A. Adkins. 2004. Handbook to Life in Ancient Rome,
updated edn. New York: Facts on File.
Amiama, C., J. Bueno, and C. J. Alvarez. 2008. “Influence of the physical
parameters of fields and of crop yield on the effective field capacity of a
self-propelled forage harvester.” 100 Biosystems Engineering 198–205.
Anderson, Terry L., and Dean Lueck. 1992. “Land tenure and agricultural
productivity on Indian reservations.” 35 The Journal of Law and Econom-
ics 427–54.
The Theory and Policy of Contractual Registries. Chicago: University of
Chicago Press.
³¹ We received helpful comments from the volume editors Paul du Plessis, Joe
Manning, Julio Ramos, and two anonymous reviewers. We have also benefitted from
the assistance of Bruno Pegorin, Francesco Mazzucato, and Director Silvia Cipriano at
the Museum of Roman Centuriation in Borgoricco (Padova) Italy.

Baird, Douglas C., Robert H. Gertner, and Randal C. Picker. 1994. Game
Theory and the Law. Cambridge: Harvard University Press.
Barzel, Yoram. 1982. “Measurement cost and the organization of markets.”
25 Journal of Law and Economics 27–48.
Bohn, Henning, and Robert T. Deacon. 2000. “Ownership risk, investment,
and the use of natural resources.” 90.3 American Economic Review 526–49.
Bradford, John. 1957. Ancient Landscapes: Studies in Field Archaeology.
London: G. Bell and Sons.
Campbell, Brian. 2000. The Writings of the Roman Land Surveyors: Introduc-
tion, Text, Translation and Commentary. London: Society for the Promo-
tion of Roman Studies, Journal of Roman Studies Monograph No. 9.
Cornell, Tim, and John Matthews. 1982. Atlas of the Roman World. Oxford:
Phaidon.
De Soto, Hernando. 1989. The Other Path. New York: Harper and Row.
Dilke, Oswald A. W. 1971. The Roman Land Surveyors: An Introduction to
the Agrimensores. New York: Newton Abbot.
Dilke, Oswald A. W. 1985. Greek and Roman Maps. Ithaca: Cornell
University Press.
Dixit, Avinash. 2003. “Trade expansion and contract enforcement.”
111 Journal of Political Economy 1293–1317.
Estopinal, Stephen V. 1998. A Guide to Understanding Land Surveys,
2nd edn. New York: John Wiley and Sons.
Farrell, Joseph, and Paul Klemperer. 2007. “Coordination and lock-in: com-
petition with switching costs and network effects,” in Handbook of Indus-
trial Organization. Amsterdam: Elsevier, 3, Chapter 31.
Johnson, Hildegard Binder. 1976. Order Upon the Land: The US Rectangular
Land Survey and the Upper Mississippi Country. New York: Oxford Uni-
versity Press.
Larsen, J. A. O. 1938. “Roman Greece,” in A. H. M. Jones, ed., An Economic
Survey of Ancient Rome. Baltimore: The Johns Hopkins Press; London:
Milford.
Lee, D. R., and G. T. Sallee. 1974. “Theoretical patterns of farm shape and
central place location.” 14 Journal of Regional Science 423–30.
Libecap, Gary D., and Dean Lueck. 2011a. “Land demarcation systems,” in
Kenneth Ayotte and Henry E. Smith, eds, Research Handbook on the
Economics of Property Law. Research Handbooks in Law and Economics.
Northampton, MA: Edward Elgar.
Libecap, Gary D., and Dean Lueck. 2011b. “The demarcation of land and the
role of coordinating property institutions.” 119 Journal of Political Econ-
omy 426–67.
Libecap, Gary D., Dean Lueck, and Trevor O’Grady. 2011. “Large scale
institutional changes: land demarcation within the British Empire.”
54 Journal of Law and Economics S295–S327.

Libecap, Gary D., and James L. Smith. 2002. “The economic evolution of
petroleum property rights in the United States.” 31:S2 Journal of Legal
Studies S589–S608.
Linklater, Andro. 2002. Measuring America. London: Harper Collins.
Linklater, Andro. 2013. Owning the Earth. London: Harper Collins.
Lueck, Dean. 1995. “The rule of first possession and the design of the law.” 38
The Journal of Law and Economics 393–436.
McEntyre, John G. 1978. Land Survey Systems. New York: John Wiley and
Sons.
Miceli, Thomas J., Henry J. Munneke, C. F. Sirmans, and Geoffrey
K. Turnbull. 2002. “Title systems and land values.” 45 The Journal of
Law and Economics 565–82.
North, Douglass C. 1990. Institutions, Institutional Change and Economic
Performance. Cambridge: Cambridge University Press.
Stambaugh, John. 1988. The Ancient Roman City. Baltimore: Johns Hopkins
University Press.
Stanislawski, Dan. 1946. “The origin and spread of the grid-pattern town.”
36 Geographical Review 105–20.
Temin, Peter. 2006. “The economy of the early Roman Empire.” 20 Journal
of Economic Perspectives 133–51.
University Press.
Thrower, Norman W. 1966. Original Survey and Land Subdivision. Mono-
graph Series of the American Association of Geographers. Chicago: Rand
McNally and Company.
19
The Institutions of Roman Markets

Benito Arruñada
19.1. INTRODUCTION: THE INSTITUTIONS

OF IMPERSONAL EXCHANGE
Opportunities for economic growth are greater when trade, instead of

being limited to known people, is impersonal.¹ This is so especially
when “impersonal” means that contractual performance is totally
independent of hard-to-observe parties’ characteristics, including
not only their reputation and wealth but also their legal authority to
commit strangers to the contract.² However, such fully impersonal
trade requires contractual enforcement to be based on assets (i.e. to
convey in rem rights or, in general, priority to acquirers) and this
poses a conflict between those holding and those acquiring property
rights. This conflict between owners and buyers is present in both
property and business contexts in which owners explicitly act through
contractual agents. This chapter analyzes the sophisticated institu-
tions used in classical Rome to overcome such conflict, reducing
information asymmetry without endangering property rights.
Given that property rights are the foundation of economic incen-
tives and prosperity, one might think that, in case of conflict with
acquirers, goods should always be returned to their owners unless
¹ This section summarizes the argument in Arruñada (2012: 15–42) and Arruñada,
Zanarone, and Garoupa (2019).
² Other concepts of impersonal exchange use less stringent requirements, such as,
mainly, trade in the shadow of an independent court (North 1990: 34–5), but also
trade with strangers, equal treatment of market participants, presence of assurance
intermediaries, or posted prices available to any buyers (Arruñada 2012: 15–18).
Benito Arruñada, The Institutions of Roman Markets In: Roman Law and Economics: Exchange, Ownership, and
Disputes Volume II. Edited by: Giuseppe Dari-Mattiacci and Dennis P. Kehoe, Oxford University Press (2020).
© Oxford University Press.
DOI: 10.1093/oso/9780198787211.003.0019
248 Benito Arruñada

they had granted their consent. However, such strict enforcement of
property rights as rights in rem would increase transaction costs by
worsening the information asymmetry suffered by acquirers of all
sorts of rights, who would always have to gather the consent of the
previous owners without even knowing who they are. Strictly enfor-
cing property rights would therefore endanger trade. Moreover, it
would also endanger specialization, because specialization is often
based on having agents acting as owners’ representatives, and, with
universal strict enforcement, acquirers would have reasons to doubt
the legal authority of sellers.
Economic growth therefore requires this conflict between property
enforcement and transaction costs to be minimized, so that both
current owners and acquirers are efficiently protected. Protecting
owners’ property rights encourages investment, and reducing the
transaction costs faced by acquirers encourages them to trade imper-
sonally and thus improves the allocation and specialization of
resources. Owners’ consent must be preserved but enforced in a
way that makes trade possible. And current owners have an interest
in tackling the conflict, not only because they are potential sellers but
also because, being acquirers with respect to the previous owners,
they could eventually lose their title.
The nature of the problem can be clarified by considering that most
economic transactions are interrelated sequentially, as most transac-
tions legally interact with previous transactions. In the simplest
sequence, with only two transactions, one or several “economic prin-
cipals”—such as owners, employers, shareholders, creditors, and the
like—first voluntarily contract with one or several economic “agents”—
possessors, employees, company directors, and managers—in an “ori-
ginative” transaction. Second, the agent then contracts “subsequent”
transactions with third parties.³ Understandably, it is necessary to
optimize the total costs of transacting, considering both originative
and subsequent transactions.
Sequential exchange is necessary to specialize the tasks of princi-
pals and agents—between landowners and farmers, employers and
employees, shareholders and managers, and so on in the originative
contract. But it also gives rise to substantial transaction costs, because,
³ This use of “agency” language for describing property cases may puzzle readers
familiar with the legal concept of agency. However, it is convenient for generalizing
the argument, encompassing all types of transactions.
The Institutions of Roman Markets 249

when third parties contract with the agent in the subsequent contract,
they suffer information asymmetry regarding not only the material
quality of the goods or services being transacted but also the legal
effects of the previous originative contract. In particular, third parties
are often unaware if they are dealing with a principal or an agent, or if
the agent has sufficient title or legal power to commit the principal.
Moreover, principals face a commitment problem when trying to
avoid this asymmetry because their incentives change after the third
party has entered the subsequent contract. In an agency setup, before
contracting, principals have an interest in third parties being con-
vinced that agents have proper authority. However, if the business
turns out badly and there are no further incentives in place, principals
will be inclined to deny such authority.⁴ The typical dispute triggered
by the sequential nature of transactions is one in which the principal
tries to elude obligations assumed by the agent in the principal’s
name, whether the agent had legal authority or not.
In principle, judges may adjudicate in such disputes in favor of the
principal or the third party. I will refer to favoring the third party as
enforcing “contract” rules, as opposed to the seemingly more natural
“property” rules that favor the principal.⁵ The effects of these rules are
⁴ This is easy to grasp in a legal agency setup as e.g. when contemplating the
relationship between a shareholder (principal), a corporate representative (agent), and
a corporate lender (third party). A typical property conflict is that between a first
buyer (principal), a seller (agent), and a second buyer (third party). In such a case, the
commitment refers to the seller’s promise not to sell twice. Ex ante, she is interested in
committing herself, in order to encourage the buyer to buy; but her incentives change
after the sale. The text could therefore read: before the originative sale contract, sellers
(i.e. agents) have an interest in buyers (i.e. principals) being convinced that they will
not cheat through a subsequent transaction (e.g. a second sale), but their incentives
change after the first sale.
⁵ These labels parallel the legal origin of the dichotomy and should prevent
confusion with related but drastically different concepts. In particular, the rules are
similar but distinct from the “property” and “liability” rules defined in an influential
work by Calabresi and Melamed (1972) because, instead of a taking that affects only
two parties, here the rules are defined in the context of a three-party sequence of two
transactions. Moreover, my analysis focuses on the role played by the parties in each
transaction, disregarding that current third parties will often act as principals in a
future sequence of transactions. Consequently, when good-faith third parties win a
dispute over their acquisitive transaction (i.e. when they are given a property right),
they do not win by virtue of a property rule, as this—by definition—would have given
the good to the original owner. In such a case, the third party does not pay any
monetary damages to the original owner, as in Calabresi and Melamed’s liability rule.
A final difference is that Calabresi and Melamed’s property rule is weaker, referring
only to the ability to force a would-be taker to bargain for a consensual transfer similar

clear. Take the simple case in which an agent exceeds his legal powers
when selling a good to an innocent third party (i.e. a good-faith party
who is uninformed about the matter in question). When applying the
“property rule” that no one can transfer what he does not have, judges
have the sold good returned to the previous owner, who is therefore
granted a right in rem, and give the innocent third party a mere
claim in personam against the agent. Conversely, judges may apply
an indemnity or contract rule so that the sold good stays with the
acquiring third party and the owner-principal only wins a personal
claim against the agent.
The difference in the value of these legal remedies—real or personal—
is substantial because the enforceability of these two types of right is
often markedly different. While rights in personam are only valid
against specific persons, inter partes, rights in rem are valid against all
individuals, erga omnes. The latter, therefore, provide the strongest
possible enforcement: without the consent of the rightholder, rights
in rem remain unaffected. When the thing is a parcel of land, the
difference in value often ranges from full value for the party being
adjudicated the land, to zero value for the party being given a claim
to be indemnified by an insolvent person. (Similar differences arise
in business and corporate contexts, where a parallel distinction is
often made, but framed in more general terms: not in terms of rights
but in terms of legal priority.)
If judges apply a property rule, maximizing property enforcement,
owners will feel secure with respect to future acquisitions but all
potential acquirers will suffer greater information asymmetry with
respect to legal title, endangering trade. Conversely, if judges apply a
contract rule, minimizing information asymmetry for potential
acquirers, they weaken property enforcement, making owners feel
insecure, endangering investment and specialization. The choice of
rule therefore involves a conflict between property enforcement and
transaction costs—more generally, a conflict between the transaction
costs of originative and subsequent contracts.
To overcome the conflict, expanding the set of viable contractual
opportunities with minimal damage to property rights, different
solutions will be appropriate, depending on the circumstances of
each type of right and transaction. The legal system may directly
to specific performance, which thus arguably has little to do with a right in rem
(Merrill and Smith 2001).

choose to enforce some rights in a certain way or, more generally, give
freedom to rightholders about which rule to apply. However, when
rightholders are free to choose, the legal system must guarantee their
commitment, which in some cases is automatic as a byproduct of
market interactions but in others requires costly organizations such
as public registers.
In general, for judges to apply property rules, which favor owners,
owners must have publicized their claims or their rights, which should
protect acquirers. That is, owners can opt for a property rule to make
their rights stronger, but, thanks to publicity, acquirers suffer little
information asymmetry. Conversely, for judges to apply contract
rules, which favor acquirers, owners must have granted their consent,
which should protect them. That is, when it is in the owners’ interest to
reduce transaction costs, they choose a contract rule, so that acquirers’
rights are stronger, whereas owners’ rights are weaker. This weakening
of property is safeguarded by the fact that it is owners—in general
terms, principals—who choose e.g. the agent to whom they entrust
possession or appoint as their representative, while implicitly, and
simultaneously, opting for a contract rule. However, commitment to
their choice is also necessary.
Smooth operation of this conditional application of rules poses
varying degrees of difficulty for different transactions. The difficulty is
relatively minor when the originative transaction produces verifiable
facts, such as the physical possession of movable goods by a merchant
or a house by a lessee, or the ordinary activity of an employee. For
these cases, judges can base their decisions on this public information,
which is produced without any explicit formal intervention. What
judges or legislatures have to do is only to define clearly the rules to be
applied.
The difficulty is greater when the originative transaction produces
fewer verifiable facts, making private (i.e. contractual, as opposed to
institutional) solutions harder to apply. Such private solutions may
even be impossible if all the information on the transaction remains
hidden and its consequences are not verifiable. Consider, for example,
the difficulties of using land as collateral when hidden mortgages are
enforced on the basis of contractual documents (Arruñada 2003).
Given the possibility of antedating mortgage deeds, judges would be
basing their decisions on unreliable information, and lenders would
be reluctant to contract for fear of previous mortgages emerging. In
such a context, rules alone are not enough because applying them

requires verifiable information on the titling-relevant elements of
originative contracts. To produce such information, it is necessary
to enforce only those mortgages that have been made public, usually
by entering them into a public register. Something similar happens
when establishing by purely private contract the existence of a cor-
poration, distinguishing the corporation’s assets from the personal
assets of its shareholders (Arruñada 2010).
This analytical perspective, which focuses on contract interaction
and sequential exchange, in which there are at least two transactions
and three relevant parties, is necessary for improving our understand-
ing of transaction costs in impersonal markets, and the role that
institutions play to contain them. In particular, it prevents contractual
problems from being considered as referring to only two parties, even
for situations in which the dominant conflict may well be multilateral.
For instance, when considering the institutions of Roman markets, a
prominent survey (Frier and Kehoe 2007: 134–7) discusses the tragedy
of the commons and the bilateral conflict between the owner and the
potential trespasser or usurper, but sidelines the multilateral conflict,
which is central for any property market, between alternative claimants
and acquirers (e.g. between owner, seller, and acquirer or between an
owner in debt and successive creditors).
19.2. INSTITUTIONAL CHOICE
Environmental circumstances, such as the cost and benefits of registry

services and their palliatives, as well as the distribution of trading
opportunities in the economy, will also affect the relative value of
personal and impersonal solutions. To structure the analysis, and
taking land transactions as the baseline, Figure 19.1 represents the
hypothetical social decision on whether or not to build the institutions
to support impersonal exchange. In essence, it assumes that more
impersonal exchange requires a fixed cost F for putting in place the
necessary institution, such as a land registry.⁶ If such an institution is
⁶ Using discrete categories of personal and impersonal exchange is a simplification.

Moreover, personal exchange also often requires institutional support, which incurs
fixed and variable costs, as the historical examples below will make clear. However, for
the sake of simplicity, I will omit them, assuming that they are lower than those for

Social value of
resources subject to 45°
titling conflict
Public
titling
Private titling
G (i.e., “privacy”)
Indifference point for Value of resources

L individual decisions assuming no
titling conflict
Figure 19.1. Choice between personal and impersonal exchange to maxi-

mize social value of productive resources.
Note: The figure represents how the social value of resources such as e.g. land (vertical axis)
is a function of their theoretical value assuming no conflicting claims (horizontal axis) and
the availability of institutions (e.g. land registries) for public titling. Social choice of
institutions is driven by these costs and the statistical distribution of the value of resources
in the economy along the horizontal axis. Area G represents the potential gain from using
institutions suitable for impersonal exchange for high-value resources, and area L the
potential loss from using them for low-value resources (by e.g. overtitling low-value land).
Source: Adapted from Arruñada and Garoupa (2005: 717, fig. 2).
created, individuals can then choose to trade personally or imperson-

ally. To trade impersonally, they incur an additional unit cost but their
assets become more valuable. Following Arruñada and Garoupa
(2005), the social value of an asset, such as e.g. a parcel of land
(represented on the vertical axis), is assumed to depend on the
impersonal exchange, so that the costs considered in the analysis must be understood
as the cost differences between institutions enabling relatively more or less impersonal
exchange.

probability that claimants with better legal rights may appear and is
thus a fraction of its value in an ideal world with no conflicting claims,
represented on the horizontal axis. This fraction depends on the
availability of institutions and the individual decisions to use them or
not, and it is lower if the asset is traded in a more personal way—in
land, it remains under private titling, or “privacy,” for short. To
simplify matters, the figure represents all the assets available in the
area served by a certain institution, such as a registry, that costs an
amount F to put in place. Publicly titling a particular parcel of land
means that, by incurring an additional unit cost (given by the intercept
of the public titling line), the probability of a conflicting claim on that
land being successful falls with respect to privacy. In other words, by
publicly titling, the value of the asset increases by a certain percentage:
with respect to the privacy line, the public titling line has a negative
intercept but is steeper.
The social choice of institutions is therefore driven by the fixed and
variable costs of the different institutions and the value of resources in
the economy, represented in the figure by their distribution along the
horizontal axis. When public titling is available, owners incur a unit
per-parcel cost for publicly titling their land, given by the intercept of
the public titling line. They will choose not to publicly title the less
valuable parcels: those below the indifference point in the figure. Area
G therefore represents the potential gain from using public titling
institutions, which are only used for higher-value resources.⁷
In a given geographical or historical context, the optimality of
introducing public titling hinges on the difference between the titling
gains G and the fixed costs, if any, of establishing the titling system, F;
and the main determinants of these gains and costs are: (1) the
statistical distribution of the resource in the economy, represented
by its distribution along the horizontal axis; (2) the costs and effect-
iveness of public titling, given respectively by the differential fixed
cost of establishing the titling system (F), the differential variable or
unit costs represented by the intercept of the public titling function
and its slope (which for land is equal to one, minus the probability of
⁷ For a more detailed analysis, see Arruñada and Garoupa (2005: 713–25). Area L
is the potential loss when public titling is made mandatory, so that even low-value
resources are publicly titled, resulting in overtitling of low-value land. This possibility
is more theoretical than real, because, in practice, the effectiveness of mandatory
titling is limited to initial allocation of titles, as shown by the recurrent failure to
register second transactions after land titling efforts (Arruñada 2012: 147).

eviction); and (3) the cost and effectiveness of privacy, given by the
variable costs represented by its intercept (assumed in the figure to
be zero) and the slope of the privacy function.
The situation in the classical period of Roman law (including the
last two centuries of the Republic and up to the second half of the
third century CE) can be hypothesized as: (1) offering considerable
gains as a consequence of relatively extensive markets, so that a
substantial proportion of resources lies to the right of the horizontal
axis, capturing the effect of greater demand and opportunities for
impersonal exchange; (2) maybe suffering high fixed and variable
costs of creating (i.e. cost F) and operating (the line’s intercept) what
would be relatively ineffective (i.e. with a relatively flat line) registries;
and (3) enjoying refined institutions for personal exchange (privacy),
which made palliatives relatively effective by enacting rules and
maximizing the use of available evidence (i.e. obtaining a steeper
slope for the personal exchange line).
The rest of the chapter will address these three issues in the Roman
context. §19.3 explores to what extent Roman markets would have
benefited from institutions making impersonal exchange easier. §19.4
examines the difficulties that the Romans of the West may have faced
for archiving information, making registries more costly or less
effective. Then, the bulk of the chapter studies the main sets of
palliative solutions used for channeling different types of exchange
in the areas of property (§19.5), business (§19.6) and the enforcement
of personal obligations (§19.7), making personal exchange (privacy)
regimes more effective. Sections 19.7.3 and 19.7.4 explore the factors
behind the institutional choice while sections 19.7.5 to 19.7.7 review
their content. §19.8 concludes by examining the overall role of the
Roman state with respect to the market.
Before moving ahead, some caveats are in order to clarify the
boundaries of the analysis. First, the above conjectures assume that
privacy was the predominant solution for the conveyancing of land
during the classical period. Even though some public titling solutions
were also available—in particular, public conveyances through cere-
monies (mancipatio) and judicial intervention (in iure cessio)—these
were not based on registries and therefore presented different costs and
effectiveness (see, for a complementary discussion, Arruñada 2015).⁸
⁸ Especially in the pre-classical period, the Romans used a form of conveyance

known as mancipatio that I will analyze in section. It was public but did not rely on

Second, when the analysis is applied not to land but to other types
of goods, it must be kept in mind that for originative transactions that
produce some verifiable consequences, such as the possession of an
asset or the activity of an employee, these consequences can be relied
upon to ground judicial decisions on subsequent contracts with third
parties. In terms of figure 19.1, this means that the personal-exchange
line will be—ceteris paribus—steeper than for those other transactions
which lack verifiable consequences, such as, for instance, mortgages.
Third, from this perspective, solutions for personal and impersonal
exchange are seen as substitutes: if personal exchange is made easier,
impersonal exchange becomes more expensive.⁹ However, the state
also builds institutions that make personal exchange more easily
enforceable: e.g. if the state opens jails for putting debtors in prison,
these will be a complement to personal exchange. The same holds for
civil registries and other systems that facilitate the identification of
individuals and therefore the enforcement of their personal obligations.
Fourth, there has been much discussion about to what extent it
makes sense to use modern economic concepts when analyzing
ancient economies. The above analysis assumes, and the next sections
aim to show, that the basic problems that market participants faced in
Roman times were essentially the same as those they face now: the
need to overcome information asymmetries efficiently with respect to
title and legal authority in order to enforce property rights and reach
specialization advantages in contracting. Institutions of course adapt
to the specific technological and economic circumstances at any time,
but this adaptation is always governed by the same basic forces. The
analysis may be new but the problems are not. The relative import-
ance of the problems is probably different, and, in the Roman case,
this is applicable, in particular, to the relative prevalence and the
consequent interactions between personal and impersonal exchange.
registries, therefore did not incur a fixed cost F, but its effectiveness was likely limited
to the local market. Roman Egypt did rely on a sophisticated system of registers, but it
predated Roman conquest.
⁹ Note that my argument is about substitution between personal and impersonal
exchange, both happening in the market, not between market and non-market forms
of exchange, as seemingly discussed by e.g. Temin when arguing that the Roman
“economy of friends” was a complement of the formal market (Temin 2013: 110–11,
148); i.e. to the extent that reputation, friendship, and other social ties acted as
safeguards of market exchange, they were complements of the market, but making
personal exchange possible, which was a substitute for impersonal exchange.

However, one should not ascribe personal exchange to nonmarket
relations. Even today, most market exchange is personal to the extent
it relies on information about the parties’ attributes. In particular,
exchange can be personal without being based on domination or
power, an aspect that might be relevant for the ongoing discussion
between “primitivist and/or substantivist” and “modernist and/or
formalist” positions (using the labels in Scheidel 2012: 7–10).
Lastly, I will not analyze factors making all types of market
exchange—both personal and impersonal—and, in particular, long-
distance trade viable, such as, mainly, the effectiveness of the Roman
state in providing: (1) a stable currency; (2) a tolerable tax burden and
little government intervention; (3) security of property in a wide
geographical area, with e.g. little piracy in the Mediterranean; and (4)
a hybrid but comprehensive judicial system which, without destroying
local legal systems, covered the whole of the Empire through potential
appeals to Rome. My objective is to analyze the main features of a few
salient institutions from the perspective of personal-versus-impersonal
exchange, in the hope of suggesting some alternative interpretations.
With this in mind, I will examine some classic controversies, aiming to
clarify some of their main elements.
19.3. INSTITUTIONAL DEMAND: OPPORTUNITIES

FOR IMPERSONAL EXCHANGE IN THE ROMAN
ECONOMY
With the expansion and consolidation of the Roman state during

the Republic, the Roman economy developed in the direction of
providing greater opportunities for impersonal exchange. In terms
of figure 19.1, it is therefore likely that the distribution of resources
moved to the right along the horizontal axis. It is debatable, however,
to what extent such movements and opportunities were limited to a
few activities or even geographical areas. Let us now review and
discuss some empirical indicators in this regard.
19.3.1. An Extensive Market Economy

Several indicators point out that the economy of the early Roman
Empire was an exchange economy to an unprecedented level, albeit

with some redistribution by the state and the presence of customary
systems in rural areas. It has been estimated that 7.4–19.5 percent of
Italy was urbanized at that time (Wilson 2011: 191). If these numbers
are taken as an index of economic performance, the Roman economy
performed similarly to those of the most advanced European econ-
omies in 1600–1750. This was achieved “through the combined
operation of moderately stable political conditions and markets for
goods, labor and capital, which allowed specialization and efficiency”
(Temin 2006: 134).
It has even been argued that the Roman economy was not merely a
collection of independent markets but an integrated market economy:
“markets for goods, labor and capital were relatively well-developed
in ancient Rome, which in turn encouraged specialization and effi-
ciency. These markets were able to work well in the environment
created by public authorities who provided local public services in
cities and a functional rule of law across most of the Empire” (Temin
2006: 137).
In particular, “the market for land in the Roman Empire worked
approximately like the land market today” (Temin 2013: 139) because
land was typically held in fee simple with few impediments to land
usage and sale. The market in urban real estate enjoyed not only a
substantial provision of public goods but, compared with other
ancient societies, much less interference from public authorities in
terms of debt moratoria, rent control, mandatory redemption, and
tenant protectionism (Ellickson in this volume). In addition, prices
varied substantially and transactions were sophisticated, including
mortgages and subdivisions.
Moreover, the Roman economy was not only well developed in
terms of size and specialization but in the predominant role of market
exchange in the allocation of all types of resources. In addition to a
market for slaves that had little in common with “closed” slavery
(Scheidel 2008; Temin 2013: 114–38), there was an active market for
the rights to sue or to inherit (Riggsby 2010: 127), and monetary
compensation was acceptable even for criminal acts (Nicholas
1962: 209).
The environment for market exchange only ceased to be favorable
in the later centuries of the Empire, when trade collapsed at least in
some periods and areas as a consequence of political instability. The
market economy was seriously damaged in some regions in the third
century CE, with the collapse of frontiers and other events: the

doubling in size and cost of the army; the increase in taxation; the
emergence of a new more militaristic and local ruling class; and the
rise of Christianity (Brown 1971). Both the market and the common
culture retreated earlier in the West: e.g. market institutions remained
in place in the Eastern empire in the seventh century (Brown
1971: 145).
19.3.2. Discussion
However, the extent, integration and nature of market exchange in
ancient economies, and in Rome in particular, remain controversial
among historians.¹⁰ For our purposes, even if the Roman economy
was a market economy, it is debatable to what extent it opened up
opportunities for impersonal exchange: in terms of figure 19.1, even if
productive resources moved to the right along the horizontal axis, it is
unclear to what extent or in which areas they did move.
Certainly, in addition to being extensive, the Roman market econ-
omy also showed signs that it might have benefited from less personal
exchange, such as substantial productive specialization and long-
distance trade. There was specialization mainly in agricultural pro-
duction and processing industries, with technological change and
diffusion (Greene 2000). Since the second century BCE, provinces
had specialized in producing different types of goods: e.g. Italy
exported wine and oil; Egypt, grain; North Africa, oil and grain;
Spain, wine, oil, and minerals, etc. Long-distance trade in bulk com-
modities was prevalent (Temin 2006), and the degree of specialization
achieved for bulk commodities suggests that the difficulties for imper-
sonal exchange were overcome at least in those markets. Documen-
tary evidence shows that at least the large estates were organized with
the aim of reaching economies of scale and producing marketable
surpluses (Rathbone 1991). Moreover, the city of Rome, far from
¹⁰ Most ancient historians would subject claims about the integration of markets in
the Roman Empire to numerous caveats. See e.g. Erdkamp (2005) and Bransbourg
(2012). Moreover, judgments on ancient economies must always be taken cautiously,
given the lack of adequate evidence. See, for instance, Scheidel (2012: 2–5), for a
summary of the limitations for unambiguously interpreting the evidence on the
Roman economy, and Scheidel and Friesen (2009) for a comparative assessment of
works obtaining substantially different estimates of the mere size of the Roman
economy.

being an extractive parasite, as Finley (1999) contended, specialized
in producing and exporting all sorts of services (Ellickson in this
volume). After the First Punic War (264–241 BCE), the increase in
foreign population in Rome made it necessary to create a specialized
magistracy, the praetor peregrinus, for dealing with non-Roman citi-
zens by applying the ius gentium to them, with a focus on conflicts
related to interregional trade (Nicholas 1962: 7; Brennan 2001: 87–9;
Winkel 2013: 3553).
Financial intermediaries and transactions were sophisticated, and
bankers provided such an array of services (Andreau 1999) that it has
even been suggested that “financial institutions in the early Roman
Empire were better than those of eighteenth-century France and
Holland” (Temin 2013: 189). Moreover, much borrowing was for
commercial purposes,¹¹ and financial markets in different regions of
the Empire were linked, most likely through financial intermediaries
(Temin 2013: 178–9). Credit was also abundant, so much so that
indebtedness was often considered a problem (Crook 1967: 171–2).
Lastly, if interest rates, “perhaps the most evident quantitative dimen-
sion of the efficiency of the institutional framework” (North 1990:
69), were indeed not much higher than those of today (Crook 1967:
211), they would suggest low default risk and therefore effective
guarantees. Despite usury regulations, it seems that market partici-
pants were able to adjust the interest rate to the risk of the loan.¹²
However, opportunities for impersonal exchange might have been
limited to some markets and transactions. With respect to trade, it
has been debated to what extent most production did remain local,
with most long-distance trade taking place only by water, given the
¹¹ Or even most borrowing, according to some authors (e.g. Johnston 1999: 84).
¹² Rates were often regulated via usury ceilings, but they were often evaded by e.g.
lending to peregrini (Livy 35.7; Andreau in this volume), not stating the rate in the
contract, and possibly granting hidden discounts on the principals (Verboven 2003).
In fact, “repeated statutes were made, for obviating all elusions, which by whatever
frequent expedients repressed, were yet through wonderful devices still springing up
afresh” (Tacitus, Ann. 6.16.2). In the light of usury-avoidance practices (Koyama
2010), the prevalence in the ancient evidence of loan contracts stating the regulated
and legally enforceable maximum rate of 12 percent or no rate at all (e.g. Lerouxel
2012b), irrespective of the borrower’s default risk, suggests that the effective implicit
rate—easily disguised by e.g. discounting the loaned amount below the stated
principal—was probably higher. Similarly, in the opposite direction, the Sulpicii
archive also provides some basis for believing that variable interest rates below the
legal limit were charged (Bransbourg 2014).

higher costs of transporting goods over land.¹³ Similarly, there are
doubts on the nature of the exchanges. For instance, for rural land in
Roman Egypt, Rowlandson (1981) considers that many of the trans-
actions she analyzes took place between relatives. Certainly, for a land
market to exist it is not necessary for most land to be transferred by
purchase, as Temin points out (2013: 144); however, transactions
between relatives are more likely to be personal. Similarly, in the
area of sureties, the effectiveness of guarantees raises the question
about to what extent exchange was personal or impersonal. Some
authors even give somehow contradictory accounts on this. For
example, Crook claims that the abundant supply of credit to the
dominant class was structured as a web of loans and personal sureties,
in a circular fashion, relying on officium, the “mutual serviceableness
between status-equals” (Crook 1967: 94) but based, deep down, on
the supposedly secure market for land (ibid. 170–2). This description
suggests that personal credit was grounded on secured and therefore
impersonal credit.
19.4. INSTITUTIONAL SUPPLY:

PUBLIC REGISTRIES IN ROME
In a context of substantial opportunities for impersonal exchange, it

may come as a surprise that land registration “was quite unknown to
Roman law” (Nicholas 1962: 104), especially considering that Rome
revitalized the land registers of Egypt and built numerous cadastres.
In accordance with the analysis in §19.1, there are two possible
reasons: the costs of registries and the effectiveness of hybrid pallia-
tives and purely personal-exchange solutions. First, as argued above,
the supply of institutions is costly and, therefore, for a given demand,
it might be inefficient to provide them even if they make some
additional exchange opportunities possible.¹⁴ Second, the efficiency
¹³ It is estimated that “the envisaged price ratio for moving a given unit of cargo
over a given unit of distance is 1 (sea) to 5 (downriver)/10 (upriver) to 52 (wagon).”
(Scheidel 2013: 4).
¹⁴ This complements the views in some of the chapters in volume I. For example,
Abatino and Dari-Mattiacci argue that lack of proper agency institutions constrained
growth; while de Ligt seems to hold that institutions followed market demands, and
Fleckner suggests that there was not enough demand for more stable business

of registries and, in general, of all solutions for impersonal exchange
depends on those for personal exchange. This section will explore
some elements of the cost hypothesis while the following sections will
examine Roman palliatives and, in particular, §19.7 will focus on the
personal-exchange hypothesis.
19.4.1. Constraints on Property Registries

in the Roman West
The cost of archiving and, especially, using written documents as
judicial evidence is difficult to ascertain but should not have been a
serious impediment to developing property registries. It is true that
Romans lacked paper and printing, and relied for recordkeeping on
perishable media, such as whitewashed boards and wax tablets
(Rhodes 2013). However, their reliance on wax tablets for all sorts
of important documents and archiving seems to have been a choice
(Meyer 2004), as they also made frequent use of papyrus imported
from Egypt, especially in the East (Bagnall 2010), and more generally
for the administration of the Empire (Innis 1950). The use of papyrus
could have reduced the costs of registries, but sealed wooden tablets
seemingly offered greater protection against fraud by tampering
(Meyer 2004: 2), so were harder to refute in court than other docu-
ments (Meyer 2004: 219, 226). Moreover, a basic facilitating factor
was probably a level of literacy that is believed to have been high not
only among the upper class (W. Harris 1989) but also among lower-
class participants in economic transactions (Haeussler 2013).
Roman archiving laws are often considered weak (Riggsby 2010:
89). Similarly, “the files of the government (except perhaps the
archives for census material) seem often to have been in disarray”
(Eich 2013). It is even claimed that Romans suffered archiving diffi-
culties for legislation, as the imperial office was sometimes unable to
find its own legislation, which “may surprise those accustomed to
think of the Romans as an administratively efficient people” (Crook
1967: 32). Moreover, forgery of public documents was prevalent at
least in certain periods (Moreau 1994). Even authors inclined to
believe that the Roman bureaucracy was well-developed recognize
associations. Other authors, however, hold a middle ground view (Hansmann et al. in
volume I.).

that the archives were at least sometimes in a state of abandonment
and negligence (Nicolet 1988: 179).
However, different levels of the Roman government did organize
extensive public archives for administrative purposes, where they
kept copies of different types of documents such as laws, contracts
for public works, international treaties, documents related to public
lands, minutes of political bodies, imperial rescripts, and mortgage
loans to provide subsidized assistance to children (Sánchez-Moreno
2013: 662‒4). Closer to property registries, they also operated public
registries on individuals, mainly the census, which was updated every
five years (Crook 1967: 46), and a birth registry, established by
Augustus in 4 CE, which served to provide evidence of status as well
as family or household size. Moreover, Romans kept records of
transactions in both the private and public spheres. Private record-
keeping was effective. For example, some banks preserved contracts
of third parties for which they had facilitated payment, for them or for
the benefit of their clients (Temin 2013: 184), and, more importantly,
lenders’ books were used as evidence in courts. The books of bankers
were seen as “unimpeachable legal evidence . . . . [And, with some
constraints to protect sensitive financial data,] the praetor required
bankers to disclose their entries as evidence on behalf of anyone to
whose case they were relevant” (Crook 1967: 233). This made their
records a sort of partial register, acting as notaries or specialized
witnesses. Similarly, the state also kept a record of transactions in the
grain trade (Temin 2013: 107). Therefore, registries seem to have
suffered neither insurmountable nor specific technological difficulties.
Nor did the Romans lack knowledge about property registries:
their conquest of Egypt (30 BCE) put them in touch with a long
tradition of property registration. In both Ptolemaic and Roman
Egypt, public registers played an effective role in enforcement and
trade of property rights in land.¹⁵ Furthermore, the Roman provincial
¹⁵ See e.g. Manning (2003); Monson (2012); and Muhs (2015). Very different
institutions for land contracting (deeds, notaries, eviction guarantees, witnesses,
checks on parties’ identity, formal consent of potentially affected rightholders,
repeated formal protestations of title, recordation of contractual transcripts in public
registries) were in place in different stages of Egyptian history and changed substan-
tially in line with increases in the volume and value of transactions (Muhs 2015). The
effectiveness of the register to protect good-faith purchasers against fraudulent vend-
ors (e.g. Daumas 1987: 162) and the other legal functions analyzed by Manning (2003)
and Muhs (2015) point towards a legal (i.e. transaction-cost-reducing) function of

administration of Egypt created between 68 and 72 CE a sophisticated
“registry of acquisitions,” the bibliotheke enkteseon, for property
rights in land and slaves. It was not merely a passive recording of
deeds but an active register of rights with certification of title: deeds
were void unless the grantor, to prove his title, presented a certifica-
tion issued by the bibliotheke to the scribe writing the deed, and the
scribe was then required to send the deed to the bibliotheke. Lerouxel
(2012a: 648‒51) shows that the introduction of the bibliotheke coin-
cided with an increase in the volume of transactions. Moreover, it was
created with the objective of protecting property rights and reducing
private transaction costs.¹⁶ This objective is also clear from the fact
that, from a merely fiscal perspective, tax authorities had an easier
alternative: to define land taxes and obligations as rights in rem, so
that they enjoy priority over all other rightholders, including innocent
third parties who thought they were acquiring free of charges. In fact,
this was the common solution in later centuries, at least in the West
(Nicholas 1962: 153; Crook 1967: 246).
19.4.2. Discussion
The situation is far from clear, however, and several aspects remain in
doubt. First, the registration of land in the census, begun with
Augustus,¹⁷ has sometimes been seen as “a system of land registra-
tion” (Crook 1967: 147). However, this part of the census was prob-
ably a mere administrative cadastre, instead of a property register,¹⁸ as
it was based on voluntary declaration, as cadastres usually are, given
Egyptian registries, as does the claim that they did not contain maps (Kain and
Baigent 1992: 1). Applying Demsetz’s (1967) argument, several factors could have
made land registries more useful in Egypt: the possibly higher value per unit of land
and recurrent annual flooding.
¹⁶ Lerouxel (2016: 145–92), and mainly Jördens (2010) and Yiftach-Firanko
(2010), discussed by Lerouxel (2012a: 652).
¹⁷ See, however, Nicolet (1988: 226‒32) for possible antecedents of cadastral
documents.
¹⁸ The confusion between cadastres and registries is a version of a wider confusion
in historiography between administrative (mainly tax) and contractual or “economic”
archives (Sánchez-Moreno 2013: 660). The necessarily public element in rights in rem
exacerbates the confusion by adding such a public element to what are functionally
private contractual registries (Arruñada 2017, 2018).

that they only create obligations, not rights (Arruñada 2012: 205).
The summary of the forma censualis reproduced in the Digest from
Ulpian, a jurist writing in the third century CE, makes clear that both
the properties and their “owners” were identified in the census, but its
purpose was fiscal, as most of the recorded information (on surface
areas, uses, products, and values) was irrelevant for titling purposes.
Consequently, the information could be based merely on what the
“owner” declared, without any reference to an independent review of
his title claim. Moreover, the Ulpian fragment makes clear that failing
to declare an ownership claim in the census did not damage the claim:
“a man who claimed ownership did not declare it, it is right that his
claim should remain” (D. 50.15.4.4; Ulp. 2 de cens.).
This declarative nature is a common feature of cadastres every-
where because tax authorities are happy to record a claimant as owner
to ensure tax collection, even if his claims are not legal. It also explains
why the census was not used as proof of title but only as one more
source of evidence: “the census-list, being based on individual dec-
larations, was only evidentiary, not automatically proof” (Crook 1967:
149). Of course, as in many other jurisdictions, ancient and new, after
volunteering to record their claim in the cadastre, Roman claimants
probably tried to use their position as taxpayers to perfect their title.¹⁹
But, given the lack of independent reviewing and purging of the
information contained in the cadastres, judges did not rely on them
(Ellickson in this volume, n. 65),²⁰ so property acquirers had no
reason to rely on them either. In particular, even if the vast majority
of the claimants recorded as owners in the cadastre were the true legal
owners, acquirers could not rely on such records to reduce the
information asymmetry that they suffer with respect to sellers. Therefore,
¹⁹ Indeed, payment of the land tax seems to have been considered proof of
ownership later, in the Byzantine empire (Laiou and Morrisson 2007: 50). This is
reminiscent of current practice in some jurisdictions: “In Illinois, for example, an
adverse possessor may establish his claim merely by paying taxes on the property, at
least against an owner who is familiar with real estate practice and records” (Rose 1985:
80). A mention in the Digest also refers to relying on the census for settling boundary
disputes in the absence of evidence on alterations (D. 10.1.11; Papin. 2 resp.).
²⁰ Considering the reluctance of judges to grant conclusive effects to evidence
produced by property registries, which has been observed in different countries and
historical circumstances (Arruñada 2012: 241, n. 39), the prevalence of fraud in public
Roman documents (Moreau 1994) may have added another difficulty.

it might well be the case that a variety of public registries did indeed
exist for all the purposes mentioned e.g. by Nicolet (1988: 213‒40), and
that the authorities did have lists that allowed them to identify “owners”
(in fact, claimants) to the level of legal certainty required by administra-
tive law in matters such as tax collection or street maintenance. However,
it is equally likely that such lists provided little valuable information for
ascertaining legal title to property.
Second, Rome also did extensive land surveying and recording
during its complex and repeated processes of land redistribution,
which relied on lists of beneficiaries and land parcels at different
levels (e.g. Moatti 1994). However, these were not prepared for
facilitating private transactions; therefore, labeling them as land
registers may be misleading (as in e.g. Duncan-Jones 1976). Even if a
double set of records with different information may have been used,
relying on bronze for the maps and more perishable media for more
detailed information, including the beneficiaries’ names and the
surface area assigned to each (Nicolet 1988: 219‒21), the information
was not updated for property transfers, so that “to clearly establish
property rights, it was necessary to perform all of a series of searches
to find their origin” (Nicolet 1988: 221). In addition to mapping
and administrative functions, the forma played a legal attestation
function “with the reserve, however, that not containing the transfers,
it must be, in case of title conflict, completed by other means” (Moatti
1994: 109).
Third, securities granted to public bodies over private land were
recorded by different means (Verhagen in this volume). However, the
purpose of this archiving might well have been merely that of enfor-
cing the state’s claims, instead of publicizing them to potential third
parties. From a land titling perspective, this archiving by the state
would therefore have served the same purpose as the safekeeping of
contracts and property deeds by landowners and lenders, and, even if
the rightholder was the state, it was in essence of a private nature.
When land charges were inscribed in stone or bronze, as in the
alimenta schemes of the Veleia and Ligures Baebiani tablets, by
which the state granted loans to landowners, the returns from
which were then used to maintain poor children, a publicity purpose
is more likely, but the function of such publicity is also unclear.
In addition to that of cautioning potential land acquirers that the
land was subject to the alimenta charge, publicity might also have
had euergetic purposes—i.e. to show that the emperor was doing

good, not only caring for children but also providing a stimulus to
agriculture (Lo Cascio 2000: 223–64).
Fourth, there seems to be growing consensus that the practices of
Egypt were more representative of the rest of the Roman Empire than
it was thought in the past and that available evidence may be seriously
biased by the type of media used (e.g. Ando 2010: 179). However, this
bias does not explain the lack of references to land registers in both
the preserved records of contracts documenting private land trans-
actions,²¹ and in the legal texts (until the last centuries of the
Empire).²² Both absences suggest that at least in the West and during
the classical period of Roman law, land registration did not play a
significant role in private contracting. Considering that jurists often
minimize the role of modern registries in property transactions
(Arruñada 2012: 114‒18), therefore exaggerating their own role,
which is diminished in the presence of effective registries, one may
wonder if Roman jurists simply omitted mentioning the titling role
played by property registries in Roman times. It seems unlikely,
however, that such bias would have left no trace of registries. Modern
treatises teach most property law as if private conveyancing (i.e.
without registries) were still the dominant paradigm. However, all
of them need to cover conveyancing “in registered land,” to use the
²¹ For instance Adams (2013); Camodeca (2013); and Rotman (2013). This is in
contrast to those found in Egyptian contracts (Yiftach-Firanko 2013). Private safe-
keeping of documents might have been prevalent even for publicly issued documents,
given that “it does appear that the safest way to preserve a decision that concerned you
was to take a copy of it to yourself when it was promulgated; and this is in fact how our
surviving documents have mostly come down to us” (Crook 1967: 33). This private
safekeeping of public documents led to forgeries (Moreau 1994).
²² The involvement of public authorities in private legal transactions increased in
the last centuries of the Empire, with some authorities recording private documents
and issuing copies of their records that enjoyed privileges as proof (Harter-Uibopuu
2013). In 313, Constantine enacted publicity for conveyances, with a clear fiscal
purpose (Honoré 1989: 142–6), and in 323 he made the public registration of
donations mandatory. A century later, Valentinian III (425‒55) required the registra-
tion of conveyances, allegedly with little visible effect with the exception of the papyri
of Ravenna, whose registration is attributed to this city being an important capital and
administrative center (Everett 2000: 73‒6). However, it is unclear how prevalent these
solutions were both before or even after Constantine. Other contractual evidence on
property, dated at the end of the fifth century, during the period of Vandal rule in
Africa, does not mention registers (Corcoran 2013). Several factors indicate that these
institutional changes had little, if any, impact on reducing transaction costs: their
fiscal objective, their coincidence with the decline in market activity, and their
apparent failure.

English terminology. They tend to present a distorted view of the
two modes of conveyancing but for obvious reasons they need to
present both. Similarly, considering practices in medieval European
cities, one may think that local courts kept written records of in iure
cessio transactions, which might then have been used as a sort of
land recordation system. But even this possibility remains unlikely
because such transactions were rare (Gai., Inst. 2.25) and reliance on
such records should have left some traces in the legal sources. Nor do
the municipal registration practices observed in later centuries
seem to be relevant with respect to classical Roman law, because
the earlier available evidence on municipal land recordation in the
West, the papyri of Ravenna, dates transactions starting in 445 CE
(Everett 2013).
Lastly, suggestions that lack of land registration in Rome was due
to an overriding preference for privacy seem unsupported. For
instance, Temin argues that “there was no registration of property
ownership on behalf of the state, as in Egypt and probably other
provinces too. Roman law and administration operated in a quite
different tradition of civil honesty compared to the prying bureau-
cratic registers of Hellenistic kingdoms” (2013: 184). However, reluc-
tance to publicize transactions seems to be absent in the legal texts
and commentaries (while it abounds e.g. in modern European dis-
cussions), and, perhaps most importantly, traditional Roman rules on
conveyancing did require publicity, as examined in the next section.
Therefore, the reason probably lay in other factors. As argued below,
more effective palliatives and stricter enforcement of personal obli-
gations possibly made registration less necessary in Italy and the West
than in Egypt and the Hellenistic kingdoms.
19.5. PALLIATIVE SOLUTIONS IN PROPERTY
The Roman law of property provided a variety of solutions for

contracting rights in rem, enabling the type of asset-based impersonal
exchange which is a main focus of this work. Allegedly, these solu-
tions changed substantially over time in parallel with changes
observed in the economy and, in particular, with the expansion of
long-distance trade.

19.5.1. Ceremonial Publicity: mancipatio
In principle, to transfer ownership of the most valuable goods, such as
land, slaves and cattle,²³ Roman law required a formal and public
conveyance, either through a collusive proceeding before the court (in
iure cessio) or through a public ceremony known as mancipatio. In
the latter, the seller and the acquirer appeared upon the property
before a person holding a scale and five witnesses. The acquirer then
declared the property to be his, struck a token coin against the scales
and handed it to the seller, who said nothing. This ceremony pro-
duced a public conveyance which was separate from the private
contract (the two-step contracting process characteristic of property,
as analyzed in Arruñada 2003). In this case, the public step served at
least two major purposes. First, it triggered the start of a prescription
period that can also be understood as a purging procedure. Contra-
dictory in rem rights were downgraded to in personam status unless
rightholders opposed the intended transaction either during the
ceremony or within the prescription period. Second, it provided the
basis for future legal procedures, as the law granted substantial
procedural advantages to possessors against other claimants.
The ceremony therefore had a titling function as it publicized the
conveyance, but also served to gather the consents of affected right-
holders and purge conflicting claims either at once or after a certain
period. In contrast to the creation of rights in personam by formal
stipulatio, which did not in principle require any witnesses (Nicholas
1962: 104), mancipatio required numerous witnesses, and this “was
probably needed not so much to prove that the act had taken place as
to give it publicity so that any defect of title (which in a small society
would be likely to be known to the witness) could be investigated
immediately” (Nicholas 1962: 256).²⁴ Indeed, in such a context,
witnesses often performed several functions, including, in addition
²³ The separation made economic and legal sense: given their value and durability,
res mancipi were assets suitable for enforcing multiple rights in rem. For this reason, it
made sense for the legal system to require publicity when transferring them. Views
considering that the separation was driven by the legal requirement probably reverse
the direction of causation. The inclusion of some types of movables in res mancipi
should not be seen as a “useless but ubiquitous complication” (Crook 1967: 141),
because those movables were high-value capital goods suitable for holding multiple
rights on them.
²⁴ It has even been argued, in the context of documents formalized in tabulae, that
“these testes or superstites, as they were called, were not witnesses in our sense,

to merely testifying, those of pointing out possible title clouds and
granting their consent. Acquirers achieved some title protection by
choosing the right witnesses: those with a good reputation (when
adjudicating a case, judges considered the reliability of witnesses)²⁵
and specially those who might know or even directly hold claims on
the land (including e.g. relatives and neighbors). These potential
claimants, by acting as witnesses, are implicitly consenting to the
transaction.²⁶ This also holds for boundary disputes, for which, in
Rome, “though not a legal requirement, it was standard practice to
call the neighbors to witness a conveyance by delivery. This allowed
the transferee to verify that the transferor and his neighbors agreed
on the boundaries, or, if they did not, to pinpoint the area/s of
dispute. If the dispute was not settled at the time, the conveyance
could be confined to the agreed area, or if necessary postponed”
(Honoré 1989: 139). Lastly, even if mancipatio did not in principle
require delivery of possession (Buckland 1912: 95–6; Gai., Inst.
4.131), the transferor had a duty to deliver it (Honoré 1989: 139).
In most cases, if the transferee did not obtain possession publicly, it
was not protected against dispossessors by possessory interdicts, this
being a speedy procedure by which the judge would adjudicate on
possession without looking at title, forcing possible owners to rely on
a more cumbersome action, that of vindicatio, in which the possessor
would be the defendant and would not therefore bear the burden of
proof (Nicholas 1962: 108–9).²⁷
expected later to testify to what they had seen or read, but judges, expected to stop an
act at the time of its making if the performance were flawed” (Meyer 2004: 118).
²⁵ Interestingly, witnesses sealing the Campanian tablets were ordered hierarchic-
ally, starting with those of higher status (Meyer 2004: 118).
²⁶ Referring to conveyances in Ptolemaic Egypt, it is said that “the seller identifies
those most likely to challenge the title of the new owner as his children and siblings”
and “children were present and consented to the transaction, or ceded their claim to
the property” (Muhs 2015).
²⁷ It must be kept in mind that, in Rome, possession was narrowly understood to
mean “not simply the holding of a thing but rather the holding of a thing in the
manner of an owner, the exclusive holding of a thing” (Nicholas 1962: 111). Those
holding without possessing had “detention”: they were physically occupying the land
but lacked relation to it (ibid. 112). This was mere possessio naturalis (Buckland
1912: 77).

19.5.2. Evolution towards Privacy
Conveyance by mancipatio therefore served for providing some
degree of publicity, gathering consents and purging property rights.
However, public ceremonial conveyances become less efficient with
economic development.²⁸ First, their cost increases if parties move
more and need to purchase from a distance. Second, they are more
effective in local markets, where transactions take place between
neighbors, as is common in rural societies, than in wider markets:
e.g. “the six participants in the mancipatio, though adequate enough
for a small community, could constitute no hindrance to secrecy in so
vast a society as imperial Rome” (Nicholas 1962: 104). Indeed, for
neighbors, it is easy to notice announcements and public deals,
especially for the kinds of rights common in rural societies, many of
which are linked to family and household matters. This is probably
less so for non-neighbors, as indicated by the longer prescription
periods usually granted to those absent, apparently balancing costlier
knowledge with longer time.
Figure 19.2 represents what might have been the historical evolu-
tion in Rome towards more private conveyancing. At the starting
point, most land is conveyed by mancipatio, with some public titling
for the highest-value land through the judicial procedure known as in
iure cessio. I hypothesize that the development of a wider market for
land had two effects: first, it increased the cost of holding public
ceremonies, which drove the mancipatio line downwards; and, sec-
ond, and probably more important, it made such ceremonies less
effective, reducing the line’s slope. With this move of the mancipatio
line, a more private titling procedure, using increasingly private
delivery and known as traditio (or even, in the late Empire, merely
granting a written deed),²⁹ a procedure which previously was
²⁸ Moreover, two other factors may be present, at least in the short term. On the
one hand, there is a permanent demand for privacy by economic agents, because of
reputational concerns and tax avoidance, a demand that could be uncorrelated to
economic growth. As argued by Nicholas, “there was a similar struggle to achieve
secret conveyancing in English law, culminating early in the seventeenth century in
the recognition of the device of a bargain and sale for a term followed by a release”
(1962: 104 n. 3). On the other hand, conveyancers—mainly lawyers—may have an
interest in privacy solutions to the extent that they increase the demand for their
professional services.
²⁹ Procedural law is also important here. In post-classical times, the more bureau-
cratic system of judicial procedure known as cognitio extra ordinem ends up
Value of property
right with titling Judicial titling
45°
conflict
Mancipatio (1)
Traditio
Mancipatio (2)
Value of property
right without
titling conflict
Indifference points between judicial

titling and mancipatio (1), right, and traditio, left
Figure 19.2. Hypothesized evolution of titling when, after distant trade

becomes more common, mancipatio becomes costlier and less effective and
is replaced by traditio.
Note: The figure aims to represent the historical evolution caused by the lesser efficiency of
mancipatio with the development of markets, which led to the predominance of convey-
ancing through traditio, characteristic of classical Roman law.
Source: Adapted from Arruñada and Garoupa (2005: 717–18, figs 2 and 3).
suboptimal, became individually optimal for land parcels whose value

lay to the left of the new indifference point.³⁰ A testable prediction of
the analysis is that the abandonment of mancipatio should lead to an
superseding the old formulary system. After Diocletian (284–305), documents issued
by a scribe started to constitute a proof superior to witness testimony (Thür 2013).
However, the role of written documents—mainly, on wooden tablets—may have been
underestimated in legal texts and scholarship (Meyer 2004: 112–20).
³⁰ Note that, in this context, relying on traditio could be individually optimal even
if mancipatio had remained more effective than deed titling in terms of title assurance.
(Imagine that in the figure the traditio line had simply moved downwards—and,
therefore, become more costly—while remaining parallel to the old one.)

increase in the number of judicial procedures, which would then also
be used for land values between the two indifference points.
19.5.3. Exercise of Possession as Access

to Ownership: usucapio
To the extent that the exercise of possession has observable conse-
quences, it is capable of providing the basis for titling, by having it
upgraded to ownership with the lapse of time—by acquisitive pre-
scription.³¹ In Rome, initially, undisturbed possession of land and
other capital goods for only two years (one year for other movables)
led to ownership by usucapio if acquired in good faith and based on
proper cause, even if conveyed by simple traditio (i.e. delivery of
possession) without mancipatio, and with the only exception of
goods stolen or taken by force. Furthermore, during the late Republic
the praetor enhanced the protection enjoyed by the good faith pos-
sessor on the way to usucapio by granting him an action that ensured
he succeeded even against the formal owner (if the possessor’s title
was formally defective: a recipient of land by informal delivery or
traditio instead of formal mancipatio) or against everyone but the
owner (if the title was substantively defective because e.g. he had
bought from a non-owner). As a consequence, “the recipient of a res
mancipi by traditio was for nearly all practical purposes in the
position of an owner” (Nicholas 1962: 127). It is believed that,
thereafter, informal delivery by traditio replaced formal mancipatio.³²
The key element for usucapio to work effectively as a titling
procedure is that the exercise of possession provides claimants with
the information they need to litigate and protect any property rights
infringed by the intended transaction (e.g. Rose 1985: 78–81). It
therefore implicitly provides the basis to start a selective purge and
³¹ For a more extensive treatment of the titling role of possession, see Arruñada
(2015).
³² In fact, it has been argued that mancipatio was retained longer for land, at least
until the fifth century, because it did not require the presence of the parties on the spot
(Buckland 1912: 73). However, it is likely that the mancipatio term appearing in later
Roman documents really means only conveyance (Nicholas 1962: 117). Meyer argues
that formal conservatism (syntax, style, rhythmic prose) in Roman documents con-
veyed authority (Meyer 2004: 59). Her argument might help explain why the word
mancipatio was still used even when the title was transferred by granting a
written deed.

acquirers need to check only if the seller has effectively been in
possession for the prescription period. In this context, doubts about
the information provided by possession may arise, especially if the
delivery of possession is unobservable. Indeed, traditio was increas-
ingly made without any visible transfer of the good. This, together
with the decreasing use and limited efficacy of mancipatio in a vast
economy, leads some authors to think that Roman publicity was
ineffective (e.g. Nicholas 1962: 104). However, this critique may be
unwarranted because, for existing rights to be protected under usu-
capio, possession only needs to inform affected rightholders with
sufficient time for them to take action. And owners can always
make their possession public if, foreseeing a possible sale, they want
to convince future acquirers that they will be protected by the usuca-
pio of the seller.
It is consistent with this titling view of possession that the time for
usucapio increased when the information provided by publicity might
have taken longer to process. Standard usucapio was part of the civil
law and therefore applied only to Roman ownership: i.e. Italic land
held by Roman citizens. For provincial land, longer periods were
required: for longi temporis praescriptio of land, ten or twenty years
when parties were or were not in the same district (Nicholas 1962:
128). This makes sense because of the greater distance and time
probably needed by adverse claimants to get notice. The time for
usucapio also became longer in the last centuries of the Empire. At the
time of Justinian, three years were required for usucapio of movables,
longi temporis praescriptio became the only usucapio for land, and a
longissimi temporis praescriptio of thirty years was introduced for
good faith acquisitions even for things that had been stolen (ibid.).
This lengthening of usucapio periods at the time of Justinian seems
consistent with greater civil unrest and consequent information and
purging difficulties.
19.5.4. Discussion
The public ceremonies of mancipatio, initially, and the sophisticated
use of possession and usucapio seemingly provided an effective pal-
liative for transfers of ownership. The limitations of such a palliative
are most visible in the inability of Roman law to implement efficient
real securities, which is understandable due to their abstract nature

and the consequent difficulties of making them publicly known by
such means. Given the lack of registers,³³ to provide lenders with real
security, mortgages were instrumented on the basis of the fiducia.³⁴
This therefore merely reduced the default risk of the lender by
creating a dispossession risk for the borrower, which helps explain
why they most often resorted to personal sureties.³⁵ Alternatively,
Romans could, of course, easily add real securities to their personal
obligations.³⁶ However, due to the possibility of hidden charges, the
³³ Efforts to make hypothecs public in the West came quite late: not until 472 CE
was priority granted to hypothecs created before a public authority or before three
witnesses (Nicholas 1962: 153), therefore with an element of publicity.
³⁴ A fiducia, “like the late medieval English mortgage, was a conveyance subject to a
covenant for reconveyance on payment of the debt” (Nicholas 1962: 151). In fiducia,
the owner gave ownership to the creditor by mancipatio or iure in cessio, usually
including pacts for reconveyance on payment of the debt, authorizing the creditor to
sell if the debt was not paid and allocating the possible surplus to the borrower (e.g.
Crook 1967: 244–5). There is also evidence in the Digest that at some point real
securities might have been reinforced, as in old Mesopotamia and modern England
(Arruñada 2012: 46–7), by pledging the deed or the whole chain of deeds with the
creditor (Scaevola D. 13.7.43 pr.; Scaevola 5 dig.). This pledging of deeds also subjects
the debtor to the creditor’s moral hazard, as he may fail to produce the deed when the
debtor needs it or may even fraudulently sell, and also makes it harder to subrogate
the debt and contract second mortgages. The Digest case makes this point clear,
as the lender fails to produce the deed to the borrower, who is therefore unable to
prove the extent of his land and cannot fully develop it.
³⁵ Personal security was very common for any substantial transaction on credit
(Nicholas 1962: 151). It has even been argued that Romans preferred personal rather
than real securities because of status: “a wealthy Roman’s word was his bond, and
security as potent as any pledge” (Johnston 1999: 88). Understandably, posting a real
security might even have been detrimental to personal reputation, as it might signal
that the debtor’s word was not valued by potential creditors and, worse still, that he
had no friends willing to back him. From this perspective, reluctance to accept
publicity would have been limited to the posting of real securities, not to mere
indebtedness; and such reluctance would be hard to reconcile with the above-
mentioned prevalence of circular credit networks among friends.
³⁶ Compare Verhagen (in this volume), for whom “the Roman law of real security
was capable of facilitating credit in a similar manner as its descendants in modern
economies” and, in particular, offered “efficiency in creation” because parties could
create a variety of securities very easily. However, ease of creation is, at most, a partial
and doubtful advantage. Partial, because such ease ex ante comes at the price of, first,
information asymmetry at the time of contracting (given the possibility of hidden
charges and title clouds, lenders know less than borrowers about the value of the
security); and, second, ineffectiveness of the guarantees ex post (if a prior hidden
charge appears or a title risk materializes) and costly enforcement to clear title before
repossession (even if hidden charges or title clouds do not surface). It is also a doubtful
advantage when considering that total transaction costs may well be lower when the
law reduces the variety of real rights by means of a numerus clausus of such rights
(Heller 1999; Merrill and Smith 2000; and Hansmann and Kraakman 2002).

enforcement of such securities was in fact most likely based on the
personal sureties offered by the borrower and his guarantors, not on
the collateral value of the borrower’s land.³⁷ Therefore, the presence and
sophistication of real securities in surviving contracts, or their treatment
in legal texts tells us little about their economic relevance.³⁸ Similarly,
and whatever its effectiveness, the legal decision to impose harsher or
criminal sanctions on those who hid such charges can also be seen as a
sign that secrecy and fraud were hindering these transactions.
Other controversial issues are also illuminated by the titling role
played by possession. For a start, analyses that consider that granting
ownership by prescription based on possession and the lapse of time
“makes possible the proof of ownership by depriving it of its ordinary
meaning” (Nicholas 1962: 156) disregard that the public nature
of possession allows claimants to vindicate ownership during the
prescription period. It therefore implicitly provides the basis for a
judicial purge of ownership claims: ownership does not exist as a
titling- (i.e. judicially) independent reality. Similarly, seeing posses-
sion as a main element of the titling process denies claims that the
importance Roman law attached to possession rather than ownership
“indicates that Roman law was seriously concerned with preservation
of the status quo and keeping the peace, and less so with questions of
formal entitlement” (Johnston 1999: 60). Possession was a tool to
ascertain which claimant would be the owner. Ownership claims
should not be confused with ownership rights. Lastly, when consid-
ering the role of possession in the titling process, views of a two-part
land tenure system in Roman land law also seem unwarranted. It is
said, for example, that “jurists and legal commentators described a
complex two-part law of land tenure. There were two sets of rules,
³⁷ Interestingly, the alimenta schemes mentioned above provide some indication

that (at least at the time for which most evidence on these schemes is available—early
second century CE) the collateral value of land was low, as owners had to provide a
security which has been estimated at an average of 12.5 times the amount of the loan
(Fernández 2013: 315). Low collateral value is consistent with the conjecture that real
securities were of little effect in reducing default risk.
³⁸ Verhagen (in this volume) bases its positive assessment of Roman real securities
on two facts: the Sulpicii bankers used real securities for large loans, and the extensive
treatment of real securities by jurists, as well as in the Digest and the Codex. However,
the Sulpicii relied on pignus—transferring possession to the creditor—an old form of
real security. And extensive legal treatment of a given topic may well be an indication
of conflict rather than low transaction costs, as shown, for instance, by the examples
on chirographa taken from the Digest by Verhagen in this volume.

one for what has been labeled ownership and one on possession. The
line between these categories is not clear, and one puzzle for modern
observers is to know which set of rules was applicable in any particu-
lar case” (Temin 2013: 147). From a titling perspective, this duality
makes sense not as a two-part tenure system for different land but as a
two-step titling process for the same land: possession was simply the
publicizing element in the second step within the titling process
ending with usucapio. Possession was the path to ownership, not its
alternative.
19.6. PALLIATIVE SOLUTIONS IN BUSINESS
Sequential exchange is prevalent in business, because using contrac-

tual agents capable of committing their principals often provides
substantial specialization advantages, expanding the scope and cover-
age of business operations. This explains why in many situations the
strict enforcement of property rights is superseded by “contract” rules
granting third parties rights in rem or giving them priority access to
the principals’ assets—i.e. making the principal liable for the agent’s
contractual obligations.
19.6.1. The Palliatives of Contractual Agency

However, early Roman law lacked a proper law of agency, as “the
strictly personal character of the Roman obligation . . . made agency
impossible [in the sense of creating a direct relationship between
principal and third party], and the mandatarius alone was both liable
and entitled on any contract which he made” (Nicholas 1962: 189).
Contracts therefore could only negatively affect their parties: what in
English law is called the “privity” of contract was rigidly applied (ibid.
199).³⁹ This strict enforcement of what I have labeled the “property”
rule made sense in a personal market with few trade opportunities but
³⁹ Initially, the only exception were rights acquired by slaves and filiusfamilias,
which vested in the paterfamilias (Nicholas 1962: 199). This asymmetry in the rule
(principals not liable but able to acquire rights through their sons’ and slaves’ actions)
made sense because the third parties granting a right to the slave owner were
themselves owners; therefore they had ultimate control of the situation and could

not after trade opportunities substantially increased in the second
century BCE.⁴⁰ Consistent with this trade-expansion hypothesis, prae-
torian edicts introduced important changes diluting such strictness,
by making principals liable and therefore protecting innocent third
parties. The key innovation was to enhance the role of slaves and sons
as economic agents, in two opposite directions: on the one hand,
making masters and fathers liable for the acts of their slaves and sons;
and, on the other hand, by limiting the masters’ and fathers’ liability
to the assets assigned to the sons or slaves. I will argue that these
solutions were close to modern solutions both in their function and
their operation, as both were based on appearance and publicity.
First, consistent with the analysis in §19.1, the principal was made
liable when it was clear he had granted his consent and therefore
authorized the agent. Since the second century BCE, by the remedies
later referred to as actiones adiecticiae qualitatis, praetorian law made
the paterfamilias liable for obligations entered by his sons or slaves,
with his liability increasing with the authority he had granted or with
his involvement in the businesses run by his sons and slaves (Nicholas
1962: 202; Abatino et al. 2011: 374–5). In particular, he was liable in
full when he had authorized the agent, either explicitly, implicitly or
by allowing similar dealings: not only when the master ordered the
transaction, but also when he put the agent in charge of a business or
a ship. In such cases, the scope of liability was limited to contracts
made in connection with the business, and could be narrowed by
giving proper notice (Crook 1967: 190; Johnston 1999: 103).⁴¹ More-
over, similar principles were extended outside the household when a
freeman was appointed as a manager or institor (Nicholas 1962:
203–4). And at least the actio institoria made it possible to commit
take good care of their interests. With little potential for conflict, this solution must
have facilitated some degree of exchange and specialization, as slave owners needed to
be present to commit themselves on any obligation but did not need to be present to
acquire rights. Indeed, we find similar asymmetries in today’s economy: banks are
often allowed to modify the land registry, but only against their own interest (e.g. by
filing a mortgage cancellation).
⁴⁰ According e.g. to archaeological evidence on the number of shipwrecks (Parker
1992).
⁴¹ “The area of competence of the agent was defined in a charter (lex praepositio-
nis) or by custom or common sense; it could be restricted by posting at the workplace
a written sign (proscriptio) in any language, or extended to specific activities by a
special invitation (iussus/-um) addressed by the principal to third contracting parties”
(Aubert 2013: 3466).

the principal on the basis of appearance (Petrucci 2003), so it played a
role similar to that of possession in property. The principal was
protected only by an explicit prohibition and the rules seems to
have evolved in such a way that the master was liable, and, therefore,
third parties were protected, when the agent exceeded his powers
through actions not explicitly forbidden and made adequately public
by the master (Petrucci 2003: 607–10, discussing D. 14.5.8; Paul. 1
decret.). Roman law was therefore considering the same basic elem-
ents of consent and publicity that we find today in business agency
(Arruñada 2010).
Second, the principal could limit his liability to the part of his
wealth managed by the agent (the peculium),⁴² therefore enjoying
some degree of “asset partitioning,”⁴³ but with a complex set of
legal actions that arguably resembled contemporary remedies aimed
to tackle standard conflicts of interests between shareholders and
company creditors, such as—in current terms (Johnson et al. 2000;
Djankov et al. 2008)—the abnormal distribution of dividends and self-
dealing,⁴⁴ or the “tunneling” of corporate resources to insiders.⁴⁵ In
general, the rules seemingly tried to link the liability of the principal to
the benefit he was obtaining from the business (Crook 1967: 189). In
terms of §19.1, this normative linkage to benefits avoided the typical
attempt by principals to get rid ex post of their own ex ante commit-
ments (i.e. to have a property rule applied), once the business entered
by their agent went against the principal’s interests.
⁴² The peculium was a separable set of resources which, even if by law it pertained
to the master, was controlled by the slave and was first in line in terms of satisfying the
contractual commitments made by the slave to third parties. Even if slaves’ peculium
was more important in economic terms, “it is overwhelmingly probable that married
sons living independently had such a fund” (Crook 1967: 110).
⁴³ See Hansmann and Kraakman (2000); and Hansmann et al. (2006). Having a
slave operating a business provided masters with the possibility of delegating the
management of business activities and benefiting from direct agency, limited the
master’s liability, provided some “entity shielding” against the master’s creditors,
and helped to ensure the continuity of business after changes in ownership and
management. See Abatino et al. (2011), who find evidence of entity shielding between
different businesses of the same owner.
⁴⁴ e.g. this could be behind the rule constraining the discretion of the master to pay
himself to the detriment of the slave’s creditors, according to Gaius (Inst. 4.72).
⁴⁵ e.g. the rule making the master liable for payments made to him from the
peculium (Gai., Inst. 4.73).

19.6.2. Discussion
Many aspects of the Roman law of business agency are subject to
alternative interpretations, however. A major question is why sons
and slaves played such a prominent role as contractual agents,
which suggests that they enjoyed a comparative advantage over
individuals sui iuris, those who were “in their own power” and not,
like sons and slaves, under the patria potestas of a paterfamilias. This
advantage seems to have lasted at least until the early third century
(Johnston 1999: 106) and is supported by the occasional practice of
free-born individuals becoming slaves to act as chief manager or
accountant of a big household (Crook 1967: 60; Ramin and
Veyne 1981).
It has often been argued that the advantage was linked to failures in
formal law—i.e. to the incompleteness of agency-independent labor
(Nicholas 1962: 203; Johnston 1999: 105–6)—but the survival of this
formal constraint seems unlikely, considering the adaptability shown
by praetorian law. Alternatively, from the perspective of §19.1, this
comparative advantage may have been rooted in lower transaction
costs in either their originative and/or their subsequent contracts: by
easier contracting of originative agency and/or subsequent trade—i.e.
by better incentives for agents and/or lower information asymmetry
for third parties.
First, the incentives provided by family (i.e. household) and slave
law lowered agency costs, not only by making harsher punishments
available, which may be suitable in conditions of information asym-
metry (Dari-Mattiacci 2013), but by judicial forbearance that implies
a radical asymmetry of decision rights, as I will argue in §19.7.2.1.
Second, using sons and slaves as contractual agents also reduced
information asymmetry in subsequent transactions if they were easier
to identify as such by third parties and, in particular, their status was
not easy to modify ex post, thereby preventing principals from freeing
themselves from their undesirable commitments. As argued in
§19.7.2.2, this explains the structure of the formalities which dealt
with household membership.
The advantage seems more nuanced in terms of subsequent trans-
actions with third parties: third parties may have been more reluctant
to contract with filiusfamilias and slaves to the extent that in some
circumstances they would have had access only to their peculium,
which made recovery harder (e.g. Gai., Inst. 4.74). However, for the

same reason, such contracting was safer for the paterfamilias, sug-
gesting the presence of a tradeoff which was arguably easy to optimize
by modifying the public involvement of the principal in the business
and, consequently, his liability.
There also seems to be substantial discrepancy on the prevalence of
corporate entities in the Roman Empire. On the one hand, through the
contract of societas (e.g. Riggsby 2010: 13), someone selling to the
partnership was selling to the individual partner with whom he was
dealing. If he was not paid, he had no recourse against other partners. In
general, “legal personality was not available for private businesses”
(Abatino et al. 2011: 368). On the other hand, big tax farming compan-
ies created by publicani enjoyed special rules: e.g. death of a partner or
litigation between partners did not dissolve the company (Crook 1967:
234). It has even been claimed that there were proper companies, at least
for tax farming and shipping, and that they were functionally similar to
the European joint-stock companies of the sixteenth and seventeenth
centuries, with their own legal identity and continuity after the death of
their investors, with separation of ownership and control and with
shares being traded at variable prices (Malmendier 2009).
Given that at least partnerships for tax collection and building
projects were governed by special rules that included incorporation
and entity shielding, the legal and judicial expertise for contracting
them was available: the basic “technology” for corporate contracting
had already been developed. However, in connection with the argu-
ment in §19.7, corporate contracting is exceedingly costly when, in
the absence of company registries, it is based purely on informal
publicity (Arruñada 2010: 556–8), as shown by the experience of
English unincorporated companies during the Industrial Revolution
(R. Harris 2000). Even if the company form was available to the
Romans, it must have been ineffective when used without some
form of public chartering. The historical record is consistent with
the interpretation that, as in the modern English case and given the
lack of company registries, large corporations were viable when based
on the public chartering process leading to the concession grant but
were hardly viable on a purely contractual basis.⁴⁶ However, in
⁴⁶ Compare Malmendier (2009), who omits any consideration of the difficulties

faced by companies created on a purely contractual basis and emphasizes political
considerations in the demise of the large societas publicanorum in the Roman Empire.

contrast with the English case, in which we observe for decades
unsatisfied demand for proper incorporation, leading to the solutions
analyzed by R. Harris (2000), there are no indications of such unsat-
isfied demand in Roman times.
The analysis in §19.2 therefore provides a more systematic explan-
ation as to why corporations were not used for other activities and
industries. The most basic reason might have been that there was little
need to use them for such activities. There was not enough demand
for more stable business associations: available technology did not
require the large amounts of capital and continuity that, for instance,
infrastructure projects and railroads were to require at the time of the
Industrial Revolution.⁴⁷ The situation would have corresponded to
the one represented by interpreting impersonal exchange as the
corporate form of business organization and personal exchange as
non-corporate organization. In the Roman context, few business
opportunities lie to the right of the horizontal axis, making—even
under voluntary registration, which makes the cost L irrelevant—the
fixed cost of creating a company register (F) bigger than the benefits
(G). Therefore, only a few large companies incorporated, relying on
the publicity provided by the public calls and auctions linked to tax
farming and government subcontracting.
They lost importance and ultimately disappeared because the state changed the way it
collected taxes and produced public services, moving away from pure tax farming into
more sophisticated forms of subcontracting, with greater control of subcontractors,
sharing of revenues and development of the state’s bureaucracy (Gibbs 2013). It is not
clear that these changes in public finance worsened the environment for the market
economy, which in fact flourished during at least a couple of centuries after them; but,
whatever their causes and merits, the most pressing question is why the corporate
form used by tax farming did not prosper in other activities and, in particular, why
this form was not used for organizing other private economic activities either before
or after such changes in fiscal policy. (The existence of corporate entities in shipping,
alleged by e.g. Malmendier (2009: 1089) and Temin (2013: 103) on the basis of a
statement attributed to Cato by Plutarch, who wrote almost two hundred years later
and might have been prone to exaggeration (see, for another likely example, Ellickson
in this volume), seems to be grounded on a weak, temporary, basis (Verboven 2002:
285), even if shipping is an activity that allegedly poses relatively discontinuous and
therefore simpler contractual problems.)
⁴⁷ Compare to the firm size constraint argument in Abatino and Dari-Mattiacci (in
volume I).

19.7. EFFECTIVE ENFORCEMENT OF PERSONAL
OBLIGATIONS
Both economic transactions and legal systems substitute between

impersonal and personal exchange. When impersonal exchange is
impossible, transactions are either made relying on personal safe-
guards or are not made at all and become lost opportunities. In
particular, the harder it is to enforce rights in rem, more rights will
be enforced in personam. For example, when real securities for credit
are weak, credit will tend to be more personal and transactors will rely
more on personal sureties. Informal norms and formal rules might
also be adapted to facilitate such substitution. If a society relies more on
personal exchange, one could expect individuals’ reputation to play a
greater role and personal sanctions to be harsher (Arruñada 2012:
50–2). This was seemingly the case in ancient Rome, where informal
social norms motivated people to fulfill their personal obligations.
Interestingly, both formal rules and organizationally supported insti-
tutions were also used to reinforce such informal social norms.
19.7.1. Informal Norms and Formal Rules

Roman society placed great importance on personal honor and repu-
tation, which was the basis for strong horizontal and vertical bonds
that provided safeguards for all sorts of exchanges. As a consequence,
fulfilling contractual obligations was a matter of personal honor
(Temin 2013: 12). Horizontal bonds meant that substantial mutual
help was customarily provided by people equal in status, mainly
under the informal institutions of amicitia and officium: respectively,
the informal bond and the obligation of friendship. “Friendship
(amicitia) gave rise to serious and substantial duties. Roman friends
made claims on each other which would cause a modern ‘friend’ to
break off the relationship without delay” (Schulz 1951: 555).⁴⁸
⁴⁸ There are signs that informal bonds became less effective over time, as indicated
by Seneca’s complaint in the first century CE that fides was no longer enough and debt
now needed to be formalized by means of written and sealed documents, with the seal
giving increased physical protection (Meyer 2004: 156).

Vertical bonds amounted to several layers of patronage by people of
higher status, who acted as patrons for their lower-status “clients,” the
clientela. In the early Republic the situation was close to serfdom, but
survived later mainly for the relation between patron and freedman
(Crook 1967: 93). Moreover, informal personal ties were also import-
ant among Roman merchants for gathering information relevant for
contractual enforcement. Traders often belonged to the same social
groups (Temin 2006: 139), were able to exchange information (Kess-
ler and Temin 2007) and, to reduce information asymmetries, relied
not only on personal ties, household members, and peer monitoring
but also on formal guilds and some state institutions (Temin
2013: 100).
These personal bonds were based on individuals’ informal reputa-
tion but were also supported in formal reputational mechanisms.
Thus, losing reputation was subject to informal sanctions, but these
soon acquired the formal version of ignominia and infamia, with a
judicially imposed exclusion from the legal protections enjoyed by a
Roman citizen (Crook 1967: 83–5). Interestingly, reputation itself was
also formally protected: defaming another person was itself harshly
punished with infamia. The role of reputation was reinforced with the
census, which established a formal classification of individuals defin-
ing their rights and duties and including a negative mark for those
who offended public morality. In this way, it provided a formal
register of reputation. Lastly, the formal legal system also relied on
and reinforced the informal system of social norms to the extent that
the value that Roman judges granted to evidence depended to some
extent on the social standing of those providing or witnessing it
(Meyer 2004).
Moreover, informal social norms were reinforced by a whole array
of formal rules ensuring enforcement of personal obligations by
several means, which included, to mention only a few, a self-help
version of debtors’ prison, strict punishments, and sophisticated
allocations of liability. Personal enforcement or self-help initially
consisted of execution of the wrongdoer by the creditor, including
victims of crimes, who were entitled to a payment from the wrong-
doer. Later, creditors could only privately imprison defaulting debtors
but were responsible with respect to other creditors (Nicholas 1962:
209–10). Judicial enforcement was also based on personal execution:
the plaintiff was authorized to imprison the judgment debtor and sell
his whole property (Crook 1967: 82–3; Johnston 1999: 108–10). After

326 BCE, a borrower could not pledge himself or a son as collateral for
a loan (nexum), to become the lender’s slave in case of default, but
debt bondage could still result from the debtor’s default (Crook 1967:
173). Lastly, the law can be understood to have tried to make personal
obligations effective by increasing punishments in line with asym-
metric information. In particular, an early legal rule imposed double
damages for eviction of res mancipi (Nicholas 1962: 161 n. 1), which
motivated sellers not only to disclose but also to discover and purge
title clouds.⁴⁹
19.7.2. Institutions Empowering Private Ordering

At a more general level, social norms protecting personal exchange
were also reinforced by two sets of arrangements enabling the extended
familia to act as a legal entity: first, those allocating most decision rights
to its head, the paterfamilias; and, second, those defining the familia’s
legal boundaries and, therefore, who could inherit and, crucially in the
context of impersonal trade, act as his contractual agent, committing
the familia’s assets to meeting his obligations.
19.7.2.1. The Role of Judicial Forbearance

in Enabling Self-Enforcement
First, concentration of property and decision rights in the paterfamil-
ias (including those on slaves and filiusfamilias) defined an ambit of
“forbearance” in which judges would not enter (using the term coined
by Williamson (1991) to designate the refusal of courts to hear
disputes internal to modern firms, for example, between their divi-
sions). The refusal of judges to enter into intra-familia disputes
allowed an asymmetric allocation to the paterfamilias of property
⁴⁹ It was also made a crime to mislead a creditor about prior charges (Johnston
1999: 93). Verhagen (in this volume) emphasizes that this happened only by virtue of
imperial rescript practice as from the end of the second century CE. This relatively late
criminalization could have been a response to greater incidence of fraud caused by the
slow erosion of personal bonds. This weakening of personal bonds might also have
been behind the earlier growing reliance on written and sealed legal documents and
the increasing physical protection of such seals, as claimed by contemporaries (Meyer
2004: 156).

rights on other familia members.⁵⁰ This should have made it possible
for familias to fully rely on self-enforcement in their internal trans-
actions and to act as quasi-firms with respect to third parties.⁵¹
Several factors increased the importance of this system of self-
enforcement. To begin with, it encompassed not only the familia
stricto sensu but the wider clientela: “it would have been impious
and unlawful for patrons and clients to litigate or testify against each
other or to support each other’s enemies” (Verboven 2013: 1578).
Second, judicial forbearance was enhanced by the paucity of manda-
tory rules about key aspects of family law, including, most promin-
ently, substantial freedom of testation (e.g. Crook 1967: 118–27),
which had obvious consequences in terms of stronger incentives for
sons and other possible heirs or beneficiaries, including slaves. Third,
the aforementioned reliance on self-help can also be understood as a
form of judicial forbearance, to the extent that it was parties them-
selves who were in charge of enforcement. Fourth, this concentration
of rights in the paterfamilias had consequences both inside and
outside the familia. Inside, it facilitated enforcement. Outside, it
helped achieving some of the “modularity” provided now by business
firms (Smith 2006, 2009). For instance, reputation was linked not
only to individuals but to the whole familia, alleviating possible
horizon problems. Lastly, it makes sense that Roman law carefully
protected the familia against possible mistakes when the young age of
the paterfamilias could lead him to exercise these decision rights
poorly. It did so not only with the institution of guardianship,
applied to children younger than 14 and in a different form to adult
women, but also with a “caretakership” which was common for
minors aged between 14 and 25, probably to avoid the difficulties
they would have faced for contracting, given that, otherwise, they
could easily have the transaction nullified later on by alleging fraud
(Crook 1967: 113–18).
⁵⁰ About agency, it is said that “these were relations that never reached the inside of
a courtroom. Their entire tone precludes contract and suit, action and liability; yet
they were most effective in fulfilling the roles and needs lawyers associate with agency”
(Kirschenbaum 1987: 180, cited by Temin 2013: 111). Indeed, they precluded contract
and suit, but only inside the familia, in a similar fashion to modern firms.
⁵¹ Asymmetric allocation of decision rights and judicial forbearance play an
important role in today’s economy (e.g. for franchising, Arruñada, Garicano, and
Vázquez 2001).

19.7.2.2. The Role of Formalities with Respect to
Household Membership
The importance of the familia as a legal entity also explains the
sophisticated formalities and cautions used by Roman law to
define its boundaries. All contractually relevant entries and exits of
the familia were subject to ceremonial formalities and, therefore, to
publicity: adoption, which was commonly used to achieve succession
by adopting adults who entered the patria potestas of their adoptive
fathers; emancipation, which led to dissolution of potestas, and was
often used as a sanction; as well as the purchase and manumission of
slaves. The fact that the dominant form of marriage under the Empire
was consensual and had no ceremonial requirement ties in with the
argument: public notice was unnecessary because the legal status of
the wife remained unchanged—that is, she remained under the patria
potestas of her father if she was alieni iuris and she remained inde-
pendent and retained her property if she was sui iuris.⁵² However,
publicity was required for manus marriage, in which the wife was in
the position of a daughter to her husband, as well as for any manci-
patio, adoption, emancipation, and manumission. All of them were
ceremonial and required the presence of witnesses or some other
form of publicity (Nicholas 1962: 71–83; Crook 1967: 103–13;
McGinn 2013; Scafuro 2013). Moreover, such witnesses often played
a purging function, as explained above. Some adoptions were even
subject to judicial supervision and popular vote: adrogatio, being the
adoption of a person not under the patria potestas of another man,
often produced the merger of two familiae, making public approval
advisable, not only because it extinguished a familia (Nicholas 1962:
77) but possibly because it may have had serious consequences for
third parties, mainly creditors. Lastly, the very prevalence of adoption
in Rome, by indicating the extraordinary importance of succession,
indirectly supports the argument as to the legal-entity nature of the
familia.
These requirements of publicity are understandable considering
that these legal acts could potentially affect third parties: in the
framework of §19.1, they are originative transactions which may
⁵² Even if form-free (Nicholas 1962: 80–1), marriage required a clearly manifested

intention, customarily evidenced by two-step contracting (a more private engagement
and a more public wedding) and elaborate public ceremonies (Crook 1967: 102–3;
Hersch 2010; Arruñada 2012: 238–9 n. 14).

have serious consequences for subsequent transactions. The origina-
tive transactions were those establishing or modifying individuals’
status, such as birth,⁵³ emancipation, adoption, marriage, manumis-
sion, etc. The affected subsequent transactions were those in which
the status of the individuals or their children was a key condition for
their legal effects. For example, marriage determined the status of
children, and adoption affected the right to inherit and the power of
the adopted to commit the paterfamilias. The consequences of slavery
and manumission were similarly important, as slavery also made a
form of contractual agency viable and manumission gave the right to
have born-free children: freedmen and freedwomen could marry
(Crook 1967: 52) and their children were freeborn (see a case in
ibid. 48–9).
19.7.3. Discussion
Based on this analysis of the enforcement of personal obligations, it is
worth considering a possible connection between the strength of
personal obligations and the convoluted development of the Roman
law of agency. When enforcement of all types (by first, second, or
third parties) is stricter, it becomes more necessary to ensure that
proper consent has been given when entering into a commitment,
which may be seen as an application of the principle in principal-
agent theory that establishes some degree of complementarity
between the intensity of both incentives and monitoring. From this
perspective, the strict enforcement of personal obligations, character-
istic of Roman law, thus ties in with, first, its initial reluctance to
commit an individual by the actions of others; and, second, its ulterior
development of a palliative of contractual agency. This palliative was
largely based on the role as agents of sons and slaves, who provided an
additional safeguard by being under the patria potestas of their father
⁵³ The birth registry was voluntary, but (as often happens with public registries)
people were motivated to record to enjoy legal effects (Rowlandson 2013). The
evidence can be interpreted as confirming the argument: recordation in the kalendar-
ium was declarative within thirty days of birth but in principle illegitimate children
(with fewer or no rights) could not be included in the album, which suggests some
form of purge, especially considering that the album was publicly displayed, with both
archives working respectively as a mere recording diary and as a proper register. It is
consistent with this interpretation that the copies used as proof of status were
produced from the album and were authenticated by up to seven witnesses.

and master, a situation that was carefully identified and often made
public by institutional solutions.⁵⁴ For a similar reason, protecting the
strict enforcement of personal obligations might also have been
behind the underdevelopment of the law of companies.
19.8. CONCLUSI ON
For decades, ancient history has been divided between divergent

“modernist” or “formalist” views, which commonly emphasize eco-
nomic growth and market exchange, and “primitivist” or “substanti-
vist” views, which tend to highlight constraints, power, and
domination.⁵⁵ My analysis of Roman institutions from the perspective
of personal vs impersonal exchange aims to transcend two aspects of
this recurrent discussion, relating to the nature and extent of the
market economy in Roman times, and the market-facilitating role of
the Roman state.
With respect to the nature and extent of the market, by focusing on
rights in rem, the paradigm of truly impersonal exchange, the chapter
sets a benchmark for the impersonality of market transactions, one of
the dimensions of transaction costs and, therefore, the “quality” of
market exchange. Using this metric, one may be tempted to frame
the discussion in ancient history in terms of personal versus imper-
sonal exchange, considering, from a primitivist standpoint, that
only impersonal markets deserve to be considered as markets. This,
however, would be mistaken on both empirical and theoretical
grounds. Empirically, because even today most markets and transac-
tions remain personal or at least retain personal dimensions, even if
other dimensions or attributes have been depersonalized.⁵⁶ More
⁵⁴ The fact that the ability to commit the principal was limited to dependents under
his potestas also refutes the argument by Verboven (2002: 260–3), according to which
the strong informal ties between Roman friends made agency unnecessary. In my
view, friends were less effective than sons and slaves as agents because, whatever the
strength of their ties with the principal and in contrast to sons and children, their
status as friends was totally informal and therefore more easily subject to misunder-
standing and information asymmetry on the part of and with respect to third parties.
⁵⁵ See e.g. for a summary account Scheidel (2012: 7–10).
⁵⁶ In reality as opposed to abstract models, impersonality is a more or less
continuous attribute of transactions (Arruñada 2012: 15–18) and even specific dimen-
sions of transactions may be situated at different levels of that continuum.

fundamentally, efficiency judgments must consider the fixed and
variable costs of introducing and operating alternative institutions;
and, even within the narrow framework of §19.2, impersonal
exchange is not intrinsically superior to personal exchange. There-
fore, the lack of land and company registries, and the consequent
reliance on palliative solutions ensuring strict enforcement of per-
sonal obligations should not necessarily be considered as a negative
indicator of the extent of Roman markets or of the support given to
the market by the Roman state.
With respect to this support, it has been hotly disputed if the
Roman state favored the development of the market or merely pur-
sued extractive objectives that were conducive, as an unintentional
byproduct, to the development of market exchange. In particular, it
has been argued that Roman markets were not supported by the state
and only worked in a limited manner within the informal networks of
merchants and households (e.g. Bang 2008). In essence, I proposed in
§19.2 a simple model of the state’s institutional decisions and then
confirmed in the following sections that the empirical evidence on
Roman institutions is broadly consistent with the model’s predic-
tions.⁵⁷ I did not discuss how the Roman state effectively provided
conventional public goods which were essential for market exchange,
such as security of property, a stable currency, standard weights and
measures, and an interconnected judicial system, but did analyze key
areas of the law in which the Roman state helped expand the market
by developing institutions that enabled a wider scope of transactions.
They were, if not impersonal, at least made possible by the effective
enforcement of personal obligations.
These market-enabling institutions included more or less centralized
solutions. Relatively decentralized solutions were based on superseding
allegedly inefficient legal rules by means of judicial decisions, in a
process which was to some extent similar to how modern common
law is often assumed to work. The most salient case is the evolution of
praetorian remedies, which filled gaps in traditional law. More central-
ized solutions, closer to the modern civil law tradition, were rarer and
are here represented by the bibliotheke enkteseon, the new land and
⁵⁷ The model assumes a benevolent decision maker, which is unreal but, under
proper constraints, individual maximization should be compatible with social welfare,
and judging the effectiveness of such constraints is not easier than judging overall
consequences.

mortgage register created by the provincial administration of Egypt.
This creation suggests that Roman administration was not inimical to
this type of institution when it was well suited but, more widely, was
interested in facilitating market transactions and limiting the market-
debasing effect of the in rem enforcement of tax duties through tacit
charges.⁵⁸
Lastly, the state contributed by supporting personal exchange
within and between familiae. In this area, the state helped the market
not only by enforcing strict rules on personal obligations and not
interfering with private contracting within familiae (forbearance), but
also by providing organized support for personal transactions.
Roman personal exchange was not mere private ordering, as the
state enacted publicity rules (e.g. on emancipation and manumission)
and implemented public personal registries (e.g. the birth registry and
the census) that reduced transaction costs and enhanced the enforce-
ment of personal obligations. From this perspective and to the extent
that informal social norms enforcing personal obligations were sup-
ported by such formal rules and registries, the separation between
formal and informal institutions, which seems to play a major role in
current discussions in ancient history (Lerouxel 2012a: 642), becomes
fuzzier, while the market-enabling role of the Roman state becomes
greater and more nuanced.
The picture that emerges when considering the whole set of solu-
tions used in the classical period of Roman law is one in which,
broadly speaking and relying on the framework developed in §19.1,
transactions are based on, first, automatic verifiability of originative
contracts (e.g. publicity of possession for property and public appear-
ance for business transactions); and, second, substantial public sup-
port for the effective enforcement of personal obligations. In
particular, institutional support provided practically no organized
verifiability (i.e. registries of originative contracts) but did provide
rules (e.g. penalties, family law), general organizational principles
(judicial forbearance), and even organizations (census) that facilitated
the enforcement of personal obligations.⁵⁹
⁵⁸ See Lerouxel (2012a: 656–7). Of course, this emphasis on enabling the market
was not permanent. Especially in later centuries, the state simply gave privileged status
to the treasury (Nicholas 1962: 153; Crook 1967: 246) without paying attention to the
increase in transaction costs caused by these tacit charges in the absence of registries.
⁵⁹ This work has benefited from advice and comments received from Maria
Brosius, Yun-chien Chang, Giuseppe Dari-Matiacci, Damián Fernández, Stephen

R E F E RE NC E S
Abatino, Barbara, Giuseppe Dari-Mattiacci, and Enrico C. Perotti. 2011.

“Depersonalization of business in ancient Rome.” 31 Oxford Journal of
Legal Studies 365–89.
Adams, Colin E. P. 2013. “Abinnaeus archive,” in Bagnall et al., eds, 5–6.
Ando, Clifford. 2010. “The administration of the provinces,” in David
S. Potter, ed., A Companion to the Roman Empire. Oxford: Wiley-
Blackwell, 177–92.
Andreau, Jean. 1999. Bankers and Business in the Roman World. Cambridge:
Arruñada, Benito, and Nuno Garoupa. 2005. “The choice of titling system in
land.” 48 Journal of Law and Economics 709–27.
Arruñada, Benito, Giorgio Zanarone, and Nuno Garoupa. 2019. “Property
rights in sequential exchange.” 35 Journal of Law, Economics, and Organ-
ization 127–53.
Arruñada, Benito, Luis Garicano, and Luis Vázquez. 2001. “Contractual
allocation of decision rights and incentives: the case of automobile distri-
bution.” 17 Journal of Law, Economics, and Organization 256–83.
Arruñada, Benito. 2003. “Property enforcement as organized consent.” 19
Journal of Law, Economics, and Organization 401–4.
Arruñada, Benito. 2010. “Institutional support of the firm: a theory of
business registries.” 2 The Journal of Legal Analysis 525–76.
Theory and Policy of Contractual Registries. Chicago and London: Univer-
sity of Chicago Press.
Arruñada, Benito. 2015. “The titling role of possession,” in Yun-chien
Chang, ed., The Law and Economics of Possession. Cambridge: Cambridge
Arruñada, Benito. 2016. “How Rome enabled impersonal markets.” 61
Explorations in Economic History 68–84.
Hansen, François Lerouxel, Joe Manning, Tom McGinn, Fernando P. Méndez

González, Pedro Pernas, Paul du Plessis, Dominic Rathbone, José Remesal Rodríguez,
Jane Rowlandson, Carlos Sánchez-Moreno Ellart, Peter Temin, Rick Verhagen, an
anonymous reviewer, and participants at the Economic History Seminar of Carlos III
University, the Bozen workshop on “The Institutions of Property Rights,” the Italian
Society of Law and Economics, the Spanish Association of Law and Economics, and
the 18th annual conference of the International Society for New Institutional Eco-
nomics (Duke University). Usual disclaimers apply. It received support from the
Spanish Ministry of the Economy and Competitiveness through grants ECO2014-
57131-R and ECO2017-85763-R, and the Severo Ochoa Program for Centers of
Excellence in R&D (SEV-2015-0563). This work is a considerably expanded and
revised version of Arruñada (2016).

Arruñada, Benito. 2017. “Property as sequential exchange: the forgotten
limits of private contract.” 13 Journal of Institutional Economics 753–83.
Arruñada, Benito. 2018. “Evolving practice in land demarcation.” 77 Land
Use Policy 661–75.
Aubert, Jean-Jacques. 2013. “Institor,” in Bagnall et al., eds, 3466–7.
Bagnall, Roger S. 2010. Everyday Writing in the Graeco-Roman East. Sather
Classical Lectures. Vol. 69. Berkeley: University of California Press.
Bagnall, Roger S., Kai Brodersen, Craige B. Champion, Andrew Erskine, and
Sabine R. Huebner, eds. 2013. The Encyclopedia of Ancient History, 1st
edn. Malden, MA/Oxford: Wiley-Blackwell.
Bang, Peter F. 2008. The Roman Bazaar: A Comparative Study of Trade and
Markets in a Tributary Empire. Cambridge: Cambridge University Press.
Bransbourg, Gilles. 2012. “Rome and the economic integration of Empire.”
Institute for the Study of the Ancient World Papers 3. http://dlib.nyu.edu/
awdl/isaw/isaw-papers/3/ (visited July 30, 2015).
Bransbourg, Gilles. 2014. “Les taux d’intérêt flottants des Sulpicii,” in Cath-
erine Apicella, Marie-Laurence Haack, and François Lerouxel, eds, Les
affaires de Monsieur Andreau: économie et société du monde romain.
Bordeaux: Ausonius, 123–35.
Brennan, T. Corey. 2001. The Praetorship in the Roman Republic, vol. 1:
Origins to 122 BC. Oxford: Oxford University Press.
Brown, Peter. 1971 [repr. 2006]. The World of Late Antiquity AD 150–750.
London: Thames & Hudson.
Buckland, William Warwick. 1912. Elementary Principles of the Roman
Private Law. Cambridge: Cambridge University Press.
Calabresi, Guido, and A. Douglas Melamed. 1972. “Property rules, liability
rules, and inalienability: one view of the cathedral.” 85 Harvard Law
Review 1089–1128.
Camodeca, Giuseppe. 2013. “Tabulae Herculanenses,” in Bagnall et al., eds,
6507–9.
Corcoran, Simon. 2013. “Tablettes Albertini,” in Bagnall et al., eds, 6499–500.
Crook, John A. 1967. Law and Life of Rome. Ithaca: Cornell University Press.
Dari-Mattiacci, Giuseppe. 2013. “Slavery and information.” 73 The Journal of
Daumas, François. 1987. La civilisation de l’Égypte pharaonique. Paris: Les
Editions Arthaud (Spanish translation used: Barcelona, Óptima, 2000).
Demsetz, Harold. 1967. “Toward a theory of property rights.” 57 American
Economic Review 347–59.
Djankov, Simeon, Rafael La Porta, Florencio Lopez-de-Silanes, and Andrei
Shleifer. 2008. “The law and economics of self-dealing.” 88 Journal of
Financial Economics 430–65.
Duncan-Jones, Richard P. 1976. “Some configurations of landholding in the
Roman Empire,” in Moses I. Finley, ed., Studies in Roman Property by the

Cambridge University Research Seminar in Ancient History. Cambridge:
Eich, Peter. 2013. “Administration, Roman,” in Bagnall et al., eds, 90–5.
Erdkamp, Paul. 2005. The Grain Market in the Roman Empire: A Social,
Political and Economic Study. Cambridge: Cambridge University Press.
Everett, Nicholas C. 2000. “Scribes and charters in Lombard Italy.” 41 Studi
Medievali 39–84.
Everett, Nicholas C. 2013. “Ravenna Papyri,” in Bagnall et al., eds, 5739–41.
Fernández, Damián. 2013. “Alimenta schemes,” in Bagnall et al., eds, 315–16.
Finley, Moses I. 1999. The Ancient Economy. Berkeley: University of Cali-
fornia Press.
tions,” in Walter Scheidel, Ian Morris, and Richard P. Saller, eds, The
Cambridge Economic History of the Greco-Roman World. Cambridge:
Gibbs, Matt. 2013. “Tax farming,” in Bagnall et al., eds, 6553–5.
Greene, Kevin. 2000. “Technical innovation and economic progress in the
ancient world: M. I. Finley re-considered.” 53 Economic History Review
29–59.
Haeussler, Ralph. 2013. “Literacy, Rome and provinces,” in Bagnall et al., eds,
4104–8.
Hansmann, Henry, and Reinier Kraakman. 2000. “The essential role of
organizational law.” 110 Yale Law Journal 387–440.
Hansmann, Henry, and Reinier Kraakman. 2002. “Property, contract, and
verification: the numerus clausus problem and the divisibility of rights.” 31
Journal of Legal Studies S373–S420.
Hansmann, Henry, Reinier Kraakman, and Richard Squire. 2006. “Law and
the rise of the firm.” 119 Harvard Law Review 1333–1403.
Harris, Ron. 2000. Industrializing English Law: Entrepreneurship and Busi-
ness Organization, 1720–1844. Cambridge: Cambridge University Press.
Harris, William V. 1989. Ancient Literacy. Cambridge: Harvard University
Press.
Harter-Uibopuu, Kaja. 2013. “Gesta,” in Bagnall et al., eds, 2908–9.
Heller, Michael A. 1999. “The boundaries of private property.” 108 Yale Law
Journal 1163–1223.
Hersch, Karen K. 2010. The Roman Wedding: Ritual and Meaning in
Antiquity. Cambridge: Cambridge University Press.
Honoré, Anthony Maurice. 1989. “Conveyances of land and professional
standards in the later Empire,” in Peter Goodwin Birks, ed., New Perspec-
tives in the Roman Law of Property: Essays for Barry Nicholas. Oxford:
Clarendon Press, 137–52.
Innis, Harold A. 1950 [repr. 2007]. Empire and Communications. Toronto:
Dundurn Press.

Johnson, Simon, Rafael La Porta, Florencio Lopez-de-Silanes, and Andrei
Shleifer. 2000. “Tunneling.” 90 American Economic Review Papers and
Proceedings 22–7.
University Press.
Jördens, Andrea. 2010. “Nochmals zur Bibliotheke Enkteseon,” in Gerhard
Thür, ed., Symposion 2009: Vorträge zur griechischen und hellenistischen
Rechtsgeschichte (Seggau, 25–30 August 2009). Vienna: Verlag der Öster-
reichischen Akademie der Wissenschaften, 277–90.
Kain, Roger J. P., and Elizabeth Baigent. 1992. The Cadastral Map in the
Service of the State: A History of Property Mapping. Chicago: University of
Chicago Press.
Kessler, David, and Peter Temin. 2007. “The organization of the grain trade
in the early Roman Empire.” 60 Economic History Review 313–32.
Kirschenbaum, Aaron. 1987. Sons, Slaves and Freedmen in Roman Com-
merce. Jerusalem: The Hebrew University Magnes Press.
Koyama, Mark. 2010. “Evading the ‘taint of usury’: the usury prohibition as a
barrier to entry.” 47 Explorations in Economic History 420–42.
Laiou, Angeliki E., and Cécile Morrisson. 2007. The Byzantine Economy.
Lerouxel, François. 2012a. “The private credit market, the Bibliotheke Enkte-
seon, and public services in Roman Egypt.” 67 Annales. Histoire, Sciences
Sociales 629–59.
Lerouxel, François. 2012b. “Des prêts sans intérêt? Le taux d’intérêt dans le
nome oxyrhynchite avant 79.” 181 Zeitschrift für Papyrologie und Epigra-
phik 161–72.
Lerouxel, François 2016. Le marché du crédit dans le monde romain (Égypte
et Campanie), Bibliothèque des Écoles françaises d’Athènes et de Rome
374. Rome: École française de Rome.
Lo Cascio, Elio. 2000. “Gli alimenta, l’agricoltura italica e l’approvvigio-
namento di Roma,” in Elio Lo Cascio, ed., Il princeps e il suo impero:
studi di storia amministrativa e finanziaria romana. Bari: Edipuglia,
223–64.
Malmendier, Ulrike. 2009. “Law and finance ‘at the origin.’ ” 47 Journal of
Economic Literature 1076–1108.
Manning, Joseph G. 2003. Land and Power in Ptolemaic Egypt: The Structure
of Land Tenure. Cambridge: Cambridge University Press.
McGinn, Thomas A. J. 2013. “Manus,” in Bagnall et al., eds, 4270–1.
Merrill, Thomas W., and Henry E. Smith. 2000. “Optimal standardization in
the law of property: the numerus clausus principle.” 110 Yale Law Journal
1–70.
Merrill, Thomas W., and Henry E. Smith. 2001. “What happened to property
in law and economics?” 111 Yale Law Journal 357–98.

Meyer, Elizabeth A. 2004. Legitimacy and Law in the Roman World: Tabulae
in Roman Belief and Practice. Cambridge: Cambridge University Press.
Moatti, Claude. 1994. “Les archives des terres publiques à Rome (1er siècle
av.–1er siècle ap. J.-C.): le cas des assignations,” in Ségolène Demougin,
ed., La mémoire perdue: à la recherche des archives oubliées, publiques
et privées, de la Rome antique. Paris: Publications de la Sorbonne,
103–19.
Monson, Andrew. 2012. From the Ptolemies to the Romans: Political and
Economic Change in Egypt. Cambridge: Cambridge University Press.
Moreau, Philippe. 1994. “La mémoire fragile: falsification et destruction des
documents publics au 1er siècle av. J.C.,” in Ségolène Demougin, ed., La
mémoire perdue: à la recherche des archives oubliées, publiques et privées,
de la Rome antique. Paris: Publications de la Sorbonne, 121–47.
Muhs, Brian. 2015. “Transaction costs and institutional change in Egypt in
the early first millennium B.C.E.,” in Dennis P. Kehoe, David M. Ratzan,
and Uri Yiftach, eds, Transaction Costs in the Ancient Economy. Ann
Arbor: Michigan University Press, 80–98.
Nicholas, Barry. [1962] 2010. An Introduction to Roman Law. Oxford:
Nicolet, Claude. 1988. L’inventaire du monde: géographie et politique aux
origines de l’Empire romain. Paris: Hachette.
North, Douglass C. 1990. Institutions, Institutional Change and Economic
Performance. Cambridge: Cambridge University Press.
Parker, Anthony John. 1992. Ancient Shipwrecks of the Mediterranean and
the Roman Provinces. British Archaeological Reports International Series
580. Oxford: Tempus Reparatum.
Petrucci, Aldo. 2003. “Orígenes romanísticos de la buena fe objetiva en la
contratación mercantil.” 15 Anuario Mexicano de Historia del Derecho
601–15.
Ramin, Jacques, and Paul Veyne. 1981. “Droit romain et société: les hommes
libres qui passent pour esclaves et l’esclavage volontaire.” 30 Historia
472–97.
Rathbone, Dominic. 1991. Economic Rationalism and Rural Society in Third-
Century A. D. Egypt: The Heroninos Archive and the Appianus Estate.
Rhodes, P. J. 2013. “Documents and archives,” in Bagnall et al., eds, 2181‒4.
Riggsby, Andrew M. 2010. Roman Law and the Legal World of the Romans.
Rose, Carol M. 1985. “Possession as the origin of property.” 52 University of
Chicago Law Review 73–88.
Rotman, Youval. 2013. “Babatha,” in Bagnall et al., eds, 1004–5.
Rowlandson, Jane. 1981. “Sales of land in their social context,” in Roger
S. Bagnall, Gerald Browne, Ann E. Hanson, and Ludwig Koenen, eds,

Proceedings of the Sixteenth International Congress of Papyrology. New
York: Chico Scholars Press, 371–8.
Rowlandson, Jane. 2013. “Birth, registrations of,” in Bagnall et al., eds,
1128–31.
Sánchez-Moreno Ellart, Carlos. 2013. “Archives,” in Bagnall et al., eds,
660–5.
Scafuro, Adele C. 2013. “Witnesses, Greek and Roman,” in Bagnall et al., eds,
7114–16.
Scheidel, Walter, and Steven J. Friesen. 2009. “The size of the economy and
the distribution of income in the Roman Empire.” 99 Journal of Roman
Studies 61–91.
Scheidel, Walter. 2008. “The comparative economics of slavery in the Greco-
Roman world,” in Enrico Dal Lago and Constantina Katsari, eds, Slave
Systems, Ancient and Modern. Cambridge: Cambridge University Press,
105–26.
Scheidel, Walter. 2012. “Approaching the Roman economy,” in Walter
Scheidel, ed., The Cambridge Companion to the Roman Economy. Cam-
bridge: Cambridge University Press, 1–21.
Scheidel, Walter. 2013. “The shape of the Roman world.” Princeton/Stanford
Working Papers in Classics. Version 1.0. April.
Schulz, Fritz. 1951. Classical Roman Law. Oxford: Clarendon Press.
Smith, Henry E. 2006. “Modularity in contracts: boilerplate and information
flow.” 104 University of Michigan Law Review 1175–1222.
Smith, Henry E. 2009. “Modularity in property, intellectual property, and
organizations.” Harvard Law School Working Paper, September 17.
Temin, Peter. 2006. “The economy of the early Roman Empire.” 20 Journal
of Economic Perspectives 133–51.
University Press.
Thür, Gerhard. 2013. “Proof, legal, Greek and Roman,” in Bagnall et al., eds,
5572–4.
Verboven, Koenraad. 2002. The Economy of Friends: Economic Aspects of
Amicitia and Patronage in the Late Republic. Brussels: Latomus.
Verboven, Koenraad. 2003. “The Sulpicii from Puteoli and usury in the early
Roman Empire.” 71 Tijdschrift voor Rechtsgeschiedenis 7–28.
Verboven, Koenraad. 2013. “Clientela, Roman Republic,” in Bagnall et al.,
eds, 1577–82.
Williamson, Oliver E. 1991. “Comparative economic organization: the ana-
lysis of discrete structural alternatives.” 36 Administrative Sciences Quar-
terly 269–96.
Wilson, Andrew I. 2011. “City sizes and urbanization in the Roman Empire,”
in Alan K. Bowman and Andrew I. Wilson, eds, Settlement, Urbanization,
and Population. Oxford: Oxford University Press, 161–95.

Winkel, Laurens. 2013. “Ius gentium,” in Bagnall et al., eds, 3553–4.
Yiftach-Firanko, Uri. 2010. “Comments on Andrea Jördens,” in Gerhard
Thür, ed., Symposion 2009. Vorträge zur griechischen und hellenistischen
Rechtsgeschichte (Seggau, 25–30 August 2009). Vienna: Verlag der Öster-
reichischen Akademie der Wissenschaften, 291–9.
Yiftach-Firanko, Uri. 2013. “Law, Greco-Roman Egypt,” in Bagnall et al., eds,
3941–45.
Ancient Sources
Digest. The Digest of Justinian, rev. Eng. edn Alan Watson. Philadelphia:
University of Pennsylvania Press.
Gaius. The Institutes of Gaius, trans. Edward Poste, rev. edn E. A. Whittuck.
Oxford: Clarendon Press, 1904.
Livius. The History of Rome, Books 27 to 36, trans. Cyrus R. Edmonds. http://
www.gutenberg.org/ebooks/12582 (visited December 1, 2013).
Tacitus. The Reign of Tiberius, Out of the First Six Annals of Tacitus; With his
Account of Germany, and the Life of Agricola, trans. Thomas Gordon, ed.
Arthur Galton. http://www.gutenberg.org/ebooks/7959 (visited December
1, 2013).
20
One Step at a Time in Roman Law

How Roman Pleading Rules Shape the
Substantive Structure of Private Law
Richard A. Epstein
One broad measure of the success of any legal system is its durability.
All legal systems must survive a world that is dominated by scarcity
and individual self-interest. Set the wrong set of rules and the self-
interest of the regulated parties will destroy the system. Set the right
rules and societies will, within the limits of their resource endow-
ments, thrive relative to their competitors. By this test, the extensive
body of private Roman law has done very well indeed. Its basic
precepts and organizational structure not only served as the founda-
tion for a great empire, but have migrated, largely intact, across a wide
range of different cultures, both in time and in space.
For the Romans, pleading rules played an integral part in their
procedural system (Metzger 2013: 13–28). The early works of
Justinian and Gaius contain an accurate exposition of Roman pleading
rules. Indeed, it has been observed more than once that many of the
pleading rules of English law are direct descendants of the Roman
pleading principles, down to the details of nomenclature, for instance
where the term replication is carried over from the Roman law into
the English to refer to the next responsive plea after the defendant has
Richard A. Epstein, One Step at a Time in Roman Law: How Roman Pleading Rules Shape the Substantive Structure of
Private Law In: Roman Law and Economics: Exchange, Ownership, and Disputes, Volume II. Edited by: Giuseppe
Dari-Mattiacci and Dennis P. Kehoe, Oxford University Press (2020). © Oxford University Press.
DOI: 10.1093/oso/9780198787211.003.0020
302 Richard A. Epstein

put forward a valid, but defeasible, defense.¹ Yet at the same time that
these rules have sparked admiration, little has been said about how the
mature Roman pleading rules established a procedural template that
lends greater precision to the private law doctrines of property, con-
tracts, and tort that work as well today as in Roman times. The purpose
of this paper is not to explain the historical development of Roman
procedural law, but to show why the Roman practices work as well
today as they did historically and why they tend to lead to the creation
of efficient legal rules.
In order to achieve this goal, I shall proceed as follows. §20.1
explains the operation of the Roman pleading system in order to
show how it facilitated the orderly development of substantive law.
§20.2 then offers concrete examples of how the basic distinctions in
Roman law facilitate its success in ways that carry over to modern
legal analysis applicable in a quite different historical period. For
these purposes, two bodies of law play a critical part. The first is the
distinction between actions in rem and actions in personam. The
second, whose use is set up by the first, is the Roman system of staged
pleadings, which tended toward efficient legal solutions by using a set
of successive approximations, each of which brought the legal system
closer to its ideal formulation. In §20.2, I shall then give a number of
concrete examples drawn from historical materials which help
explain the efficient results that the Romans achieved in a number
of difficult substantive areas of the law.
20.1. THE ROMAN PROCEDURAL SYSTEM
The great strength of the Roman procedural system was its clear
classifications, aided by the deliberate way in which the mature
Roman pleading system converted broad cases into precise issues
ready for an authoritative decision. On its face, this technical body
of rules appears to derive little from economic theory, given that its
¹ For the early invocation of replication in the Roman law, see The Institutes of
Gaius, Book 4, pt I, §§ tl 5–29 (ed. Francis de Zulueta, 1946) (noting the use of the
replication as a way to admit new matter that overcomes what would otherwise be a
valid exception to the prima case). The replication made its way into English law as
well (Sutton 1929: 78, 190, 198).
How Roman Pleading Rules Shape the Substantive Structure 303

basic framework stresses the organization of procedural elements of a
wide array of real and personal actions, where the substantive legal
framework emerges through the gradual application of the procedural
system. But a closer look at the rules does reveal something of the
implicit economic principles operating at every level. The emergence
of these categories was not just a fluke but rested on the happy
historical circumstance that the Romans institutionally separated
the formulation of their basic legal rules from the application of
those rules to particular disputes. The basic legal rules and principles,
applicable across the board, were carried over from each praetor, or
some other magistrate, in a continuous arc of legal development
(Gai., Inst. 4.111; Lenel 1927). A separate fact finder then applied
those principles to the given case (Metzger 2013: 8).
This strong degree of separation meant that the standard legal
formulas could be articulated with a rare precision uncluttered by
the factual disputes that encouraged the parties to make strategic
decisions in presenting their cases. In the course of litigation, triers
of fact are constrained by a host of practical and formal restraints.
They must dispose of litigation within a reasonable time. They also
must make sure that each party receives due process, so that he has a
right to be heard—audi alteram partem—before a neutral judge—
nemo judex in causa sua—and fair notice of the charges against him
(Danner 1912: 27, 33). Those requirements all speak to the simple
point that each party should have his day in court to present his case.
The institutional framework that the Romans used to formulate their
pleadings severed doctrinal developments from the short-term neces-
sities of practical litigation.
The central importance of this separation lay in the way in which
the Romans classified their causes of action and pleaded their cases to
joinder of issue. The want of any litigant pressure meant that the
Romans could develop rational classifications of causes of action. The
division between rights in rem and rights in personam set the stage for
the simple but critical line of actions ex contractu and ex delicto. This
two-stage design may appear intuitive, but even with the Roman
precedents, the English ad hoc system of substantive rights insecurely
organized within the forms of actions created an unfortunate disson-
ance between the form of action and the underlying cause of action,
which was evident in dealing with both actions for debt (which
covered consensual and nonconsensual claims) and actions on the
case (which covered both tort and contract actions; Maitland 1936).

Once their basic classifications were established, the Romans were
able to develop their substantive rules by adding in additional elem-
ents of the case through exceptions, replications, duplications, and so
on (Nicholas 1962: 24–5).
The modern framework always provides on grounds of natural law
(or the American due process of law) both litigants the right to
present their cases before the court (Schauer 1976: 47). This two-
part division may be justified on grounds of judicial economy insofar
as it lets each side speak exactly once and then uses some pretrial
meeting to set the stage for the trial.² But conceptually, this procedure
sets out the wrong template for pleading any case to joinder of issue.
The great advance in Roman pleading law was that it did not confine
the development of the underlying legal doctrine to two stages of a
claim and then a defense, but rather it allowed the parties’ back and
forth to continue so long as either party wanted to add some new
matter to the case that incorporated all allegations from the proceed-
ing stages of the complaint.³ At each stage, the party then opposing
the pleadings (in the second stage, this would be the defendant, to
whom the plaintiff would respond in stage three, and to whom the
defendant would subsequently respond in stage four, and so on) has
three options. The first of these is to deny the allegations. The second
is to admit that they are true, and then to insist that they do not state a
sufficient prima facie case or subsequent plea (in effect, to demur).
The third is to admit that the new material is a sufficient plea and
then to insist that some additional factor, by way of replication or
duplication, is consistent with a sense of justice and therefore should
be considered an exception to the previous plea. Gaius does not use
the English common-law classifications in which these three options
are termed: to deny, to demur (or object), or to add new material
through a plea of confession and avoidance (Sutton 1929; Epstein
² Federal Rule of Civil Procedure Rule 16(a) provides: “In any action, the court
may order the attorneys and any unrepresented parties to appear for one or more
pretrial conferences for such purposes as: (1) expediting disposition of the action;
(2) establishing early and continuing control so that the case will not be protracted
because of lack of management; (3) discouraging wasteful pretrial activities;
(4) improving the quality of the trial through more thorough preparation; and
(5) facilitating settlement.”
³ These rules were set out with great clarity in Book 4 of Gaius’ Institutes
(§§102–29); Just., Inst. 4.6, 12. The development of this theme has been a career
obsession (Epstein 1997, 2005).

1973a). Yet his analysis tracks these terms precisely, so I shall use
them here for purposes of clarity. Importantly, the demurrer intro-
duced new material into the case, and its legal significance (at that
stage) was determined by whether it was then accepted or rejected. It
is through this dialogue of pleadings that the Roman law narrowed
and refined the precise legal issues in dispute.
These procedural niceties are vital for institutional reasons, but
I shall for the moment put them aside to concentrate on the substan-
tive defenses, some good and some bad, that can be raised in these
Roman cases, such as assumption of risk, contributory negligence,
infancy, insanity, plaintiff ’s entry, and the like. Once the defendant
presents these substantive defenses, the correct procedure is for the
plaintiff at stage three of the case (the reply or replication) to have the
same set of choices that is made available to the defendant at stage
two. The denial again raises no important principle of fact. The
demurrer to the exception asks the same question about its sufficiency
that a demurrer to the complaint asks about the basic cause of action:
is the matter pleaded sufficient to switch the balance back in favor of
the defendant? Yet it is at this point that the third alternative, the
introduction of new material by a plea of confession and avoidance,
shows why the common law’s two-stage procedures, which may be
administratively convenient, in fact distort the overall development of
substantive law. As Gaius notes: “Sometimes an exception, which in
the absence of counter allegations seems prima facie to be just to the
defendant, is unjust to the plaintiff, and then, to protect the plaintiff,
the praetor adds to the instructions a clause called Replication,
because it is an undoing and counteraction of the force of the
exception” (Gai., Inst. 4.26). Once three stages are admitted into the
argument, there is no way in which it is possible to impose any
artificial limits on the process whereby new information is added.
Allowing the replicatio thus paves the way for allowing the duplicatio
or rejoinder, after which the cycle can continue indefinitely until a
particular party finds that it has nothing more that it would like to say
(Gai., Inst. 4.127–9).
A few observations are in order about the formation of this system.
The first of these is that the system proceeds by private initiative
followed by government approval. The state does not tell individuals
what issues they ought to raise or when they ought to raise them. That
part is left in private hands, subject of course to the oversight of the
court as to which of these pleas is allowed, and when. The choice of

this method has the sometimes unappreciated virtue of removing any
external constraints on the operation of the system as a whole. The
kinds of questions that can be raised are not those which arise solely
from the abstract reflections of bureaucrats, but from concrete dis-
putes as they percolate from the ground up, assuring that the doctri-
nal issues that matter will receive careful attention while those that do
not will be passed by in relative silence.
The second point is that, in dealing with these issues, Gaius does
not make the mistake of H. L. A. Hart in his famous lecture on
“Ascription of Rights and Responsibilities,” which sought to turn
defeasibility into a question of the meaning of key words like “con-
tract” (Hart 1949: 171–94).⁴ But in fact that is not the case. A perfectly
adequate definition of contract refers to an agreement between two or
more people calling for the performance or forbearance of various
actions by either or both parties. That definition makes it clear what
kinds of activities are governed but leaves open whether a particular
agreement is the kind that should be enforced, or whether, even if it
falls within the class of enforceable agreements and the defendant has
not performed, the promise still should not be enforced because of
some act of the plaintiff or extrinsic circumstances. None of these
matters are questions of the definition of contract. All are questions of
substantive judgment. The modern law and economics scholar might
raise his or her eyebrows about Gaius’ constant assertions that certain
forms of conduct are either “just” or “unjust,” which reflects the
modern inclination to distrust any pronouncement about natural
law. But I think that this charge is overstated. A close look at the
individual cases in which Gaius pronounces judgment shows a virtual
unerring sense for the right result that would please a thoughtful law
and economics scholar who subscribes to the overtly consequentialist
arguments that dominate modern economic thought.
Thus, to give Gaius’ example in Inst. 4.26, in which the prima facie
case is that the defendant has not paid a debt to the plaintiff, the
defendant can introduce as an exception that the plaintiff had agreed
not to sue on the debt. But then the debt could be reinstated if the
plaintiff, by way of the replication (in the next stage), could demon-
strate that he had renewed the debt by a fresh promise: what Gaius
⁴ For criticism of Hart’s position, see Geach (1960: 221–5); Pitcher (1960: 226–35).
Hart eventually abandoned that position (Hart 1983). However, none of these three
authors hit on the Roman formulation.

called a “contrary agreement.” Clearly, a duplication is in order if that
fresh promise was in turn induced by fraud, and so on. Indeed, as
I have argued elsewhere, the phrase “naturalis ratio”—literally the
natural reason—rests on implicit use of a modern economic argu-
ment that could not be articulated at the time because no one then
thought in terms of the economic principles on which that approach
rested (Epstein 1989: 713–51; 2003). The task of law and economics is
to explain the durability of these rules, and not to dismiss them as
some relic of outmoded thought, which decidedly they are not.
20.2. THE USES OF THE PLEADING SYSTEM
It is one thing to note the organization of any legal system. It is quite

another to show its economic utility. In this regard, the bottom-up
nature of the pleading system is congenial with the economic
approach that wants to gain a holistic look at the overall set of
rules. Modern law and economics scholars tend to develop formal
models that identify the optimal distribution of rights and duties. In
so doing they tend to rely on overbroad generalizations suggesting
that, for example, a negligence rule with a contributory negligence
defense will provide for the optimal allocation of resources (Brown
1973: 323–49). The claim is often made in both a positive and
normative guise, and is wrong in both those contexts. In this regard,
the constant Roman tendency is to subdivide causes of action into
discrete units, each with its own outcomes. That approach often leads
descriptively and normatively to the use of a wide range of potential
liability rules in different contexts. In stranger cases, the basic pos-
ition sets the prima facie case for all law suits as strict liability, so as to
flesh out that system. But in harms that arise out of consensual
situations, no uniform theory of liability can deal with all of the
actions and omissions that cause harm, so that a wider range of
solutions may prove necessary. The advantage of the pleading system,
rightly construed, is that it resists the kinds of overgeneralizations to
which modern law and economics often falls prey.
To see how this system functions, it is important to show in some
detail how the pleading rules work, when rigorously and consistently
applied. There is a tendency in Roman sources to use these rules
precisely in dealing with such matters as pleas in abatement and res

judicata. But by the same token, as noted long ago by Danner (1912),
there was less willingness to plead separately all of the substantive
elements of a given claim. Instead it was common to allow matters
that technically might be part of an exceptio to be introduced at trial
under a general denial of the plaintiff ’s case (Danner 1912: 36–7). But
the method loses much of its power by this relaxation of its formal
strictures.
In order to ensure that the doctrine is properly aligned, it is
necessary to make sure that there is a unity across the system,
which can only be obtained by making a couple of critical observa-
tions about the notion of “sufficiency” of a claim. First, in a negative
sense, the term sufficient in law never means what it does in formal
logic—if A, then B follows. Rather what it means for legal pleadings is
that if A’s claim is sufficient, then B cannot file a demurrer but must
either deny or plead over by confession and avoidance. Second, in
order for this process to remain harmonious, the prima facie case
should only require the minimum set of allegations that is necessary to
shift the burden of explanation, whether by excuse or justification, to
the other side, leaving all other potentially favorable facts for use, if
needed, at some later stage in their argument, should some valid plea
by the defendant then give rise to new opportunities for further
elaboration. The point here is intended to impose a major level of
self-restraint on the plaintiff who might wish to plead more. The
choice of action starts with the identification of the plaintiff ’s interest
that has been harmed, which is of course established, as will become
clear, in accordance with the rights embedded in the in rem/in
personam distinction. The next part of the prima facie case identifies
the act or omission that is said presumptively to interfere with either
of the two issues just identified.
Here we can examine some examples from common law torts and
contracts. By way of introduction, “tort,” in contrast to the Roman
“delict” (Buckland, McNair, and Lawson 1965: 352–66), is drawn
from English law. Delict has long been understood to be “the act of
a person, by fault or maliciously, which harms or otherwise causes
damage to another” (Pothier 1761 [1806]). It should be apparent
that this definition excludes strict-liability offenses in their entirety,
which is consistent with the Roman view that did not distinguish
sharply between public enforcement of criminal law and private
enforcement of personal wrongs. The law of furtum clearly has this
intentional element, but the Lex Aquilia reads differently. It starts out

in strict-liability fashion—if anyone kills—so that the addition of
negligence and malice is through gloss only. I regard that approach
as a serious intellectual error in both Roman and modern times, and
for the same reason. This gloss ignores the role that strict liability
(with defenses) has and should have in any system of private law. In
using the term “tort” to refer to private wrongs under both the
common-law and the Roman-law system, I am talking only about
civil enforcement and intend to stress the close parallels between the
two legal systems within this limited domain.
In actions for bodily harm of property damage, plaintiffs will often
choose to include in their pleadings that the defendant’s act intended
to cause harm, or that the defendant failed to take reasonable care in
dealing with matters that hurt the plaintiff. That view is in tension
with the explicit language of the Lex Aquilia, where the word occidere
(killing) is not coupled with either intention (dolus) or some form of
fault (culpa). Instead the word used is iniuria, which literally means
“unlawfully,” without linking it to the denial of negligence and inten-
tion (Nicholas 1962: 222), as is commonly done.⁵ That position is also
taken explicitly in Gaius, where it is said baldly, “Now a person is
understood to kill wrongfully (iniuria) when this death occurs
through his willful intent (dolus) or negligence (culpa)” (Inst.
3.211). Note that this approach precludes the possibility of all
strict liability cases, prior to any examination of the substantive
arguments in favor of that position. Under strict liability the
unadorned claim that the defendant killed the plaintiff is dismissed
just as if the complaint had said that the plaintiff had been killed by an
act of God, which is on both moral and philosophical grounds
indefensible.⁶ Nor does this ostensible definition of iniuria leave the
field open for the full range of affirmative defenses based on plaintiff ’s
conduct, let alone instruct on how to treat those cases in which some
excuse or justification is offered, when it is very unlikely that the
identical set of defenses applies alike to negligence and intentional
harm cases. As becomes clear later on, iniuria is best understood not
⁵ For a discussion of the common law of torts, see Williams (1951).

⁶ See Holmes (1881: 96): “Unless my act is of a nature to threaten others, unless
under the circumstances a prudent man would have foreseen the possibility of harm,
it is no more justifiable to make me indemnify my neighbor against the consequences,
than to make me do the same thing if I had fallen upon him in a fit, or to compel me to
insure him against lightning.” But what of the want of foresight by the plaintiff of the
defendant’s application of force?

as a description of any mental state or extrinsic circumstance, but as a
conclusion of law that could never be substituted for a particular
allegation, so that just as the earlier definition of iniuria was too
limited, so too is its common alternative, which has, of course, wide
acceptance in modern tort law as well (Holmes 1881: 88–96; Brown vs
Kendall 1850). A similar issue could arise with a breach of promise,
where, however, it is rarely held that no promise is actionable unless
the breach is either willful or negligent. And once again, allowing the
strict liability case at the first level still permits the introduction of
other mental states further down the road.
It is, to be sure, a wise pleading tactic on the part of plaintiffs in an
unregulated system to allege any negligence or harmful intention in
the prima facie case if only to quiet any moral qualms about the
strength of the claim, so that a trier of fact will be strongly inclined to
rule in the plaintiff ’s favor. Nonetheless, that approach must be
strictly forbidden as a systematic matter. The point here is not that
mental states and care levels are irrelevant to the law of tort; rather,
they have played a central role in fleshing out the system from the
beginning. The only question on this score is where to slot them into
the overall system, not whether they have a part to play in it.
The insistence that each case be only presumptive also structures
the terms that are permissible to introduce at any stage of a case.
Thus, the plaintiff cannot plead that the defendant trespassed on his
land, for that creates the conclusion that the defendant had commit-
ted a wrong, which may or may not have been the case. The correct
way of putting the point therefore is to say that the defendant entered
the plaintiff ’s land, which creates a prima facie case of trespass, which
in turn might be excused because of a mistake in the location of the
boundary or justified by showing that the plaintiff permitted the
defendant to enter the land. Likewise, the plaintiff cannot say that
the defendant killed the plaintiff ’s relative. The prima facie case is one
of (deliberate) killing, which is a prima facie case of murder, but not
necessarily murder given the possibility of affirmative defenses (how
intention to harm is introduced will be discussed later). Similarly, on
the contract side, the prima face case in the pleading can never say
that the defendant breached his contract. The correct form of argu-
ment is that the defendant did not keep his promise, which leaves
open the possibility that there was some excuse or justification for
that case of nonperformance, as the case may be. In effect the English
language (and I dare say all others) has these pairs—nonperformance,

breach; entry, trespass; deliberate killing, murder—where the first of
the pair sets the presumption which becomes a legal wrong only when
the various responsive pleas are inapplicable in the case.
On this score, therefore, systematic reliance on pleadings offers a
useful decision tree that allows the analyst to see where each discrete
allegation fits into the overall picture. The insistence that there be
only the minimum elements stated at a given point in the analysis
means that there is then an opportunity for additional elements to
be worked into the issue later on. Indeed, to have two or more
alternative theories of liability that start with the prima facie case
leads to the obvious question of how the two of them can fit simul-
taneously into the overall picture, and how, and why, to parcel out
different causes of action between them, which is the difficulty of the
standard account of iniuria. That problem is avoided by the system of
staged pleadings, which makes it possible in time to work all relevant
elements into the overall system.
The same result is reached when the issue is looked at from a more
explicitly economic view. The tapestry of potentially relevant facts is
always very large, and the correct approach for ordering them is to
ask of the plaintiff to select from the full range of facts those that
advance his case, that is, to give reason why the law will create a social
improvement if it imposes prima facie form of liability in the plain-
tiff ’s case. Once that is done, the plaintiff can rest as the defendant
comes forward with an argument to explain why first appearances
are, as Gaius said, deceiving and that further information may well
show why it is now unwise to allow that cause of action, subject to
further pleadings on the matter. It should be apparent whenever this
system is used that the vast bulk of the cases are governed in practice
by those issues that come up early in the pleading sequence. Most
cases of force used against strangers, for example, are not used in self-
defense, so that much is gained if the collection of accidental injuries
is handled before tackling any difficult questions arising from inten-
tional harms. The same kind of arguments can be made about various
issues in contract law, where there is again the same alternation. The
plaintiff has to plead the prima facie case in minimalist terms, after
which the defendant is allowed to excuse or justify his non-
performance subject to further sequential pleadings by both sides
until the issues are finally joined.
The success of the pleading system does not depend solely on the
fact of any one particular alternation. It is also critical to see that the

right substantive boxes are filled to make sure that the system does
not go wrong. In order to do this, the legal system has to solve two key
questions. The first is to determine the initial baseline that establishes
the relative position among all individuals, dispersed as they are
across the globe. The second is to develop a set of legal responses
for persons who make moves that are inconsistent with these critical
assignments of rights. The first task is often discussed in political
theory under the rubric of “state of nature” theory or, more recently,
under the more abstract Rawlsian formulation of the “original pos-
ition,” which is intended to make the question look less historical and
more analytical. The second task is to develop the rules that go into
place once the basic norms are established. The pleading issues are
key to this second development, but they can only be understood by
starting with how the key Roman distinction between actions in rem
and actions in personam sets the table for the subsequent pleading
analysis, by indicating the scope of the plaintiff ’s interest as protected
by law. Thereafter the pleadings set the roadmap for dealing with
cases in which individual defendants do not follow, as it were, the
rules of the road established in the initial legal position.
20.2.1. Actions in personam and in rem

That basic distinction between actions in rem and in personam is part
of every legal system, even if the Latin words “against the thing” or
“against the person” do not translate easily into English (Epstein
2011). In a procedural sense, an action is said to be in rem if it is
directed toward the thing, which the plaintiff claims to be his own
(Nicholas 1962: 99–103). That vindicatio rei (vindication of title in
the thing) does not name a defendant as such, but clearly anyone who
wishes to contest the claim of ownership is entitled to appear on his
own behalf. The hard procedural questions are the extent to which
notice has to be given, so that others with a claim to the property have
a chance to be heard before their ownership rights are blocked, and
that notice could be provided in principle either by personal service to
known rivals or publication to others. One way in which this is done
in English law, with chattels, is for the plaintiff to make a demand
for the return of the object, which when refused creates a prima facie
case of the tort of conversion against this particular defendant,
leaving all other parties (none of whom are in possession) out of

the picture. But the formal differences by which these suits begin
should not conceal their common object, which is to restore posses-
sion of the disputed thing to the plaintiff, who then has an interest in
the property that is good against the entire world. In that sense the
vindicatio rei is different from any action brought against a particular
person for some wrong that he has done to the plaintiff insofar as it
protects rights in rem (i.e. claims for forbearance against the rest of
the world) and not rights in personam, by which more exacting
obligations were imposed on a particular person or persons by agree-
ment. Conceptually, this distinction is as relevant as to modern law as
it is to Roman law (Chang and Smith 2012).
We are now in position to see the close connection between rights
in rem and actions in tort, and rights in personam and rights in
contract. In both cases, the basic method comports first with the
basic logic of ordinary language and second with a rational economic
approach that allows the overall system to be developed by a series of
successive approximations, each of which, if decided correctly, takes
us closer to some social optimum, much the way a mathematician
finds the area under a curve by approximating it with ever thinner
rectangular slices, which in the limit define the area under a curve.
20.2.2. Actions ex delicto

In the tort context, if mutual forbearance from entry is the require-
ment of an in rem action, the action that constitutes the prima facie
tort is a physical invasion that violates the plaintiff ’s space. The bare
entry is enough of course to allow the plaintiff to remove the defend-
ant. Similarly, the taking of a chattel offers a prima facie case for its
recovery. Actions for damages in the event of partial or complete
destruction of the property follow on closely, because no defendant
should be able to escape liability by the use of this special stratagem of
timing the sequence of taking or destroying. It would be grotesque to
allow a defendant to first destroy the property without legal conse-
quences, and then be responsible only for the return of the land or
object that he has destroyed. The constitutional law of the United
States treats takings and tort damages as a pair, precisely because the
private law has made that same fundamental structural move.
This narrow definition of the prima facie case prevents the disin-
tegration of the “harm principle” in the law of tort so that it covers

any externality, i.e. an action that has a negative impact on other
individuals. It is more than coincidental that the violation of these in
rem rights involves in the first instance some physical invasion of the
property of another—the boundary-crossings that Robert Nozick
made famous in Anarchy, State, and Utopia, without any knowledge
of its Roman antecedents or economic logic (Nozick 1974). Once that
invasion has taken place, the party responsible has singled himself out
of the mass of other individuals by upsetting the balance of mutual
forbearances found in the in rem world. At this point, traditional
notions of corrective justice, drawn in large measure from the work of
Aristotle, kick in and shape the remedy that should be awarded
(Nicomachean Ethics, Book 5). In those cases where the land or
chattel is taken, the remedy is its restoration to the true owner. That
restoration cannot, of course, restore the lost (rental) value of interim
use or compensate for any damage to the property while in the hands
of other individuals. Accordingly, the law supplies damages to fill in
the gap with the object of returning the system to the status quo ex
ante to the extent that the law is able.
It is critical to stress that corrective justice, so conceived, is thus not
diametrically opposed to some economic approach to the subject,
even if it is often treated in that fashion. To the contrary, a sound
understanding of corrective justice provides a platform that allows for
the achievement of particular economic objectives. In those instances
where the chattel or an improvement on land has been destroyed, the
restoration remedy is no longer feasible, so the system turns, for want
of a better solution to damages as a second-best substitute for
unwinding the transaction. Those damages are in general awarded
on the same ground as in the case of damage or delay. But as they
occupy the entire landscape instead of some small fraction of it, the
weaknesses in that system become more acute. In dealing with slaves,
for example, the Romans did not normally award pain and suffering.
In the analogous action allowed to a free person, some estimate of
those damages should normally be allowed notwithstanding the want
of any market standard to determine the harm in question. The basis
of liability under the principles of minimum sufficiency outlined
above should reflect (as the text of the Lex Aquilia suggests in ch. 1)
a strict liability system for anyone who kills (occidere = killing by
cutting) a slave of either sex or a herd animal (pecus). By the same
token the parallel passages in chapter three dealing with “burning,
breaking, or destroying” are subject to the same basic reading, Here,

as earlier, it is key to distinguish between a denial that the defendant
has acted and an effort by the defendant to excuse or justify his
actions. Conceptually, it is one thing for a defendant to be pushed,
and quite another for him to act under duress in order to escape some
greater harm.⁷ Thus it becomes important to ask what happens
whenever B enters A’s land because he, B, has been driven there
under threats of force by C. At this point, B cannot plead the general
denial, because acting under compulsion is not the same as not acting
at all. In principle, one could allow this narrow defense on substantive
grounds, by insisting that the only proper action should be brought
against C, who did not enter the land but who had maximum control
over the situation. The result is that if all persons were present and
solvent in a single proceeding, B would be exonerated even though
the direct cause of the harm, and C would be liable in full although he
did not enter the land.
Yet even this simple scenario poses serious procedural difficulties
under the Lex Aquilia, which does not supply an action of indemnity
by B against C—you made me kill him. Any direct action by
A against C forces these substantive ambiguities to the surface,
because it is now necessary to unpack longer causal chains with
multiple defendants. The correct set of answers requires not a yes/no
or on/off answer on causation, but instead demands an explicit
causal hierarchy among all relevant parties, such that A can recover
from either B or C, and B can recover from C, and C can recover
from no one at all (Epstein 1973b).
The English common-law cases raise exactly those same ambigu-
ities. Thus, in Weaver vs Ward (1616), the judges write:
no man shall be excused of a trespass (for this is the nature of an excuse,
and not of a justification, prout ei bene licuit [as it well appeared to
him]) except it may be judged utterly without his fault.
As if a man by force take my hand and strike you, or if here the
defendant had said, that the plaintiff ran cross his piece when it was
discharging, or had set forth the case with the circumstances, so as it
had appeared to the Court that it had been inevitable, and that the
defendant had committed no negligence to give occasion to the hurt.
⁷ For the variations, see for the Roman version (D. 9.2.7.3; Ulp. 18 ad ed.), and for
the English law the contrast between Smith vs Stone (1647) and Gilbert vs
Stone (1647).

Note that the evocative phrase “utterly without fault” conceals the
difference between the denial of voluntary action—“as if a man
take my hand by force and strike another”—and an affirmative
defense based on the plaintiff ’s conduct (running “cross his piece”).
The full passage closes the triad with the affirmative defense of
inevitable accident, which is overcome in the third stage by the
reply that the defendant must lose if he had done some prior act to
“occasion” the harm. The use of the weaker causal word “occasion”
strikes the correct balance, because the use of force has been pleaded
at stage one, and it cannot therefore be raised anew in the replication.
Since, moreover, no justification could be offered for an accidental
killing—in this case an early instance of friendly fire—the court
rightly said that only an excuse could be raised, without seeing, as
becomes clear later on, the distinctions between excuses that enter the
pleadings at stage two and stage six.
This discussion of both the Roman and English cases shows how
easy it is to jumble up different elements in a complex fact pattern
when the pleading rules are not scrupulously observed. The narrow
definition of causation involving the use of force, however, also sets
the stage for invoking a broader sense of causation which in Roman
law covers the case of “furnishing a cause of death” (causam mortis
praestare), which is set in opposition to kill (occidere), which gives
rise to an action in factum, or what the English call an action on the
case. The Digest makes this leap from direct force so that the protec-
tion of the plaintiff ’s individual autonomy cannot be circumvented
by taking simple evasive steps, as in the case where the defendant
literally sets poison in front of the woman who then drinks it herself
(D. 9.2.9 pr.–1; Ulp. 18 ad ed.). Yet this expansion of liability is not
open-ended because it will not cover cases in which the woman
drinks the poison knowing of its danger when not subject to any
external threat of force. As economic theory predicts, the key cases
are those of asymmetrical information when the defendant knows, as
the plaintiff does not, that the drink contains poison.
The situation takes on an added dimension because ordinary Latin
(and ordinary English) both speak of possible excuses and justifica-
tions for given courses of conduct. No one has to justify actions that
do not harm other people, or even those that harm them accidentally:
only excuses based on the plaintiff ’s conduct come into play in the
latter case. Blocking the plaintiff ’s right of way (which presupposes a
system of property rights, lest someone claim that the plaintiff ’s face

got in the way of the defendant’s fist) provides a good causal defense
against strict liability. People can debate whether that defense should
be total or partial, so long as they understand that proof of antecedent
negligence by the plaintiff (except in the odd sense of negligence per
se, a move back to strict liability in disguise), is technically speaking
not needed to raise the defense (Epstein 2012).
A parallel affirmative defense is that the plaintiff was injured while
on the defendant’s land. That entrance establishes plaintiff ’s prima
facie trespass—which can be answered in two ways: one for strangers,
and one for licensees and invitees, who enter consensually. By the
former, the plaintiff concedes the trespass but rightly insists that the
defendant deliberately struck the plaintiff who had crossed a field
where javelins were being thrown at “an inopportune time”
(D. 9.2.9.4; Ulp. 18 ad ed.). Lawson mistakenly treats deliberate
harm as a species of negligence, even though the word culpa means
culpability, i.e. a generalized form of blameworthiness, and not neg-
ligence. Clearly the word “noxious” lodges a more powerful charge
than negligence, which generally will not suffice, because the plain-
tiff ’s trespass does not impose a duty of care (Lawson 1950: 222–3).
Nonetheless, the exact content of deliberate harm still needs further
elaboration, especially in cases where the defendant acts in reckless
disregard of consequences. But what now becomes crystal clear is that
deliberate harms only make their way into the original strict liability
case at the third stage of the argument. Compressing the pleadings
into two stages thus conceals that the defendant’s mental element is
relevant if, and only if, it is necessary to overcome an initial excuse
based on plaintiff ’s conduct.
Negligence can also become operative in cases of consensual entry.
Thus, if the plaintiff alleges a want of reasonable care, the defendant
responds at stage four that the defendant as a social guest assumed
that risk as well, for example, of latent defects not known to the
defendant. But that defense will fail against a business guest with
respect to latent dangers discoverable by reasonable inspection.⁸
Similarly, when a defendant surgeon cuts the plaintiff patient,
assumption of risk is a defense against a strict liability case, but
⁸ For the old categories of duties owed to various kinds of entrants, see Robert
Addie & Sons (Collieries), Ltd. vs. Dumbeck (1929); for their rejection, see the
California Supreme Court’s decision Rowland vs. Christian (1968). For commentary,
see Epstein 1999: 329–31.

ordinarily not against negligence. The Lex Aquilia picks up that point
independently of the pleadings: “Proculus says, if a doctor operates
unskillfully on a slave, an action lies either ex locato (for hire of
services) or under the Lex Aquilia” (D. 9.2.7.8; Ulp. 18 ad ed.).
These contractual variations are great whenever two parties who
have singled out each other for special dealings. Accordingly, this
concurrence of actions only applies for “operations.” The defendant
who fails to treat a patient when there is a duty to do so is easily held
liable on the contract of hire, but not under the Lex, which has great
difficulty in dealing with cases of omissions. In principle, consensual
parties may allocate risk as they choose, for they are not bound by the
simple allocation rules that govern cases between strangers.
To return to these stranger cases, justifications for deliberate harms
only enter the case at the fourth stage. The first type, relevant at stage
two, involve excuses for accidental harms based on plaintiff ’s con-
duct. The second type of excuse enters at stage six as follows.
A justification for deliberate harm, e.g. self-defense, enters at stage
four, after an allegation of deliberate harm. At stage five, the plaintiff
argues that the defendant exceeds the scope of that justification, say
by the use of excessive force, or even disproportionate force (a much
tougher nut to crack). In both cases, the measure of damage is
reduced, to cover only the excessive force in the first case and the
disproportionate force in the second. Next at stage six the defendant
points to the weaknesses in his mental or physical ability that
prompted the use of excessive force. At this point an extended version
of the defense of “you take your victim as you find him” is incorpor-
ated in both Roman and English law (D. 9.2.7.5; Ulp. 18 ad ed.).⁹
The same logic applies to the excuse of non-consent to deliberate
harm. At stage five, the plaintiff could claim that he was induced to
accept the operation by fraud, or perhaps by the failure to disclose
known risks that a fiduciary physician should make clear. The injured
pugilist may allege that the defendant did not use proper equipment,
so that injuries inflicted went beyond the scope of the consent that
was given. Assumption of risk for accidental harms, at the second
⁹ “But if one gives a slight blow to a sick slave and he dies, Labeo rightly says that he
is liable under the Lex Aquilia, because different things are fatal to different people.”
The text is not quite right because it does not distinguish between the treatment case,
which is governed by the rules of medical malpractice, and the case involving the
stranger, which is not.

stage is thus distinguished from consent to deliberate ones at the
fourth.
By contrast any plea of insanity (as noted in Weaver vs Ward) is
quite inappropriate at the second stage. The tort system with its
strong views of individual autonomy does not allow any insane
person a direct claim of assistance from B, a stranger. If A cannot
force the obligation of care on B, a stranger, he cannot achieve
that result by raising the defense of insanity when his own actions
have injured or killed another. The same argument can be made
about infancy or any other forms of legally-recognized capacity
impairments.
The infant, the incompetent, or the novice cannot be asked, as the
victim of an attack, to perform beyond his capabilities. Nonetheless,
the law cannot allow any infant or insane person to hide behind his
status to use unnecessary force. To counter that risk, the usual way of
framing the stage six excuse asks whether the defendant’s use of either
excessive or disproportionate force was done reasonably and in good
faith. This balancing act helps address the irreducible uncertainties
that surround defensive actions under stress. The good faith element
allows a certain subjective inquiry; the reasonableness element tries to
cabin it within an objective standard. Once the initial balance has
been upset, the clear rules that work at stages one and two no longer
apply. But so long as the hardest cases arise in only the lowest
probability situations, the system will work even if perfection is
unattainable.
20.2.2.1 Summing up the tort side

These disparate pleading paths, then, explain how peaceful coexist-
ence among the three basic theories of liability should be achieved.
Strict liability sets the uniform prima facie case. Negligence comes in
as an offset to the assumption of risk cases in the various special
relationships of which physician and lifeguard, babysitter, landlord
and occupier are some common examples. These cases will be subject
to contractual variations, both express and implied, which explains
why divergence in pleas at stages three and four forms a necessary
and proper part of the overall system. By the same token, the use of
deliberate force at stage three raises squarely issues of consent and
self-defense, which show how closely entangled justifications, and the
limitations on those justifications, apply as well. None of these

arguments are particularly dependent on the fine points of Roman
culture. Indeed, the rules that were developed first with slaves cover
all sorts of situations involving personal injury and property damage
in a more or less timeless fashion. The move from Rome AD 160 to
New York AD 2018 is relatively seamless. That is indeed as it should
be, lest every change in social circumstance require a recalibration of
the fundamental rules of tort law. By the same token, the variations
that arise in cases of assumption of risk and consent allow for degrees
of individuation that are as valuable across different people working
in the same general social milieu, as it does for social changes across
milieus and millennia. The care expected of a babysitter is likely to
vary less significantly over generations than the care expected of an
experienced surgeon, and these rules accommodate all the relevant
distinctions. Start with the pleadings and the system can run its
course.
20.2.3. Actions ex contractu

The extended account of the role of pleadings in tort theory makes it
easier to see how the basic approach carries over to contract disputes.
The initial statement of the complaint should be bare bones, leaving
the complications for later pleas when they become relevant. It is
commonly said on moral grounds that people ought prima facie to
keep their promises.¹⁰ In this regard, the law deviates evidently from
the Kantian premise that “there are certain duties of perfect obliga-
tion, such as those of fulfilling promises, of paying debts, of telling the
truth, which admit of no exception whatsoever” (Ross 1930: 18). But
that rigid position goes so much against the accumulated weight of
every legal system that it is idle to take it as the final resting point.
What is needed is a way to systematize these insights as presumptions
instead of per se rules, which is what the pleading metric allows.
In this context, we start with the proposition that it is prima facie
wrong not to keep one’s promises. The point here is that it is only
through promising that people can gain the benefits of cooperation
¹⁰ See, for example, the discussion of prima facie obligations, in Ross 1930: 20–5.
The notion of the prima facie case “is an idea that simply had not occurred to critics of
deontological theories (Mill, Moore, Sedgwick, Rashdall) and therefore represented at
the time a major advance in the dialectic existing between utilitarians and non-
utilitarians” (Stanford Encyclopedia of Philosophy 2012).

over time. People make promises that they are willing to keep in order
to elicit a like response from others, so that both sides can mutually
improve their position. To allow the demurrer to this basic prima
facie case would upset every known fact of social organization. To
keep the plea sufficiently general, the basic case starts with a general
strict liability standard, so that presumptively contracts create war-
ranties of result, not promises for effort.
Yet this presumption can be overcome in a variety of ways. In most
cases, the moral force of a promise is not explicitly tied to the use of
legal remedies. It should therefore be a defense against the legal
enforcement of any particular promise for the defendant to demon-
strate that the parties had no intention to create legal relations
(Balfour vs Balfour 1919), which could be proved either explicitly or
implicitly from context. In addition, it could also be urged that the
promise in question was not supported by consideration so that it was
a gift promise, which should not ordinarily be enforced.¹¹ To that
position, it is then possible to create exceptions with respect to certain
forms of gift promises, including those for gifts to charities or on the
occasion of marriage, which might be made enforceable as a substan-
tive matter. It is also possible to argue that even though these prom-
ises may not be enforceable when fully executory, they are enforceable
after the plaintiff delivered some object to the defendant, for which
under the law of bailments the transferee has an obligation to return.
Or in some cases, the promise could simply be that the defendant
agreed to take care of a chattel that he borrowed from the defendant,
so that the issue of executory enforcement is nowhere to be found. It
is also possible in many cases to make the further argument that the
defendant who is prima facie obligated to repay a debt could plead
any release as binding, even if it is not enforced by consideration,
which is the Roman rule but not the English (Gai., Inst. 3.169–76).
The possible defenses to the enforcement of a promise are not
solely related to the issues of intent and consideration. In other cases,
an affirmative defense could be used in context-specific ways to
undercut the strictness of the basic contractual obligation. That
adjustment will in general not be needed for the provision of certain
kinds of standardized products. No one thinks that the implied
warranty of merchantable products like milk or canned goods is
¹¹ See, for example, Congregation Kadimah Toras-Moshe vs DeLeo (1989).


met by the seller’s good faith effort to deliver a safe product. But when
the analysis turns for contracts dealing with services, including those
for medical malpractice, the assumption of risk defense becomes
powerful given the intrinsic uncertainty in medical outcomes, so
that liability for negligence (usually based on customary practice)
becomes the norm. In rare cases that can be displaced, as when the
defendant is anxious to conduct experimental treatments of the
plaintiff for his own collateral benefit, at which point the standard
of care remains strict (Hawkins vs McGee 1929).
It is through this lens of assumption of risk, which is always
highly context-specific, that the multiple standards used, for
example, in the bailment cases are developed, each depending in
some measure on the distribution of benefits between the two
parties (Epstein 2009; Coggs vs Bernard 1704). The greater the
benefit to the bailee, the higher the standard of care should be, and
the similarly so in reverse. It is only in a limited subclass of cases that
the standards commonly chosen are the reasonable-care standard
associated with the formula Judge Learned Hand developed in
United States vs Carroll Towing (1947).
The same element of pleading is exceedingly important in connec-
tion with the law of conditions, which deals with issues related to the
sequence of performance of various promissory obligations in a
contract. Thus, the simplest agreement is one in which A promises
to sell goods for $10, with nothing said about the order of perform-
ance. The modern form of pleading generally says that the plaintiff
should plead the performance or the occurrence of all conditions in
order to maintain the action, where the former term refers to condi-
tions within the control of the defendant and occurrences refers to
natural events (Federal Rules of Civil Procedure 9(c)).¹² But that
general pronouncement has no heft, for all the work comes when
the defendant is required to deny that performance or occurrence
“specifically and with particularity.” The denial then puts the condi-
tion into play and sets up the possibility that the plaintiff may be
able to explain in the third stage why the nonperformance or
¹² “In pleading conditions precedent, it suffices to allege generally that all con-
ditions precedent have occurred or been performed. But when denying that a
condition precedent has occurred or been performed, a party must do so with
particularity.”

nonoccurrence of the condition should not be a defense. Thus, it
could be said that the defendant need not pay the money because he
has yet to receive the goods. But to that there is a further reply that the
plaintiff has tendered the goods so that this defense is not available, or
the defendant waived the condition so that replication undoes the
force of the defense. At this point, the system introduces excuses and
justifications for the nonperformance of conditions. There are still
other implied conditions, based on impossibility of performance or
fundamental mistake that are accepted in some case but not in others,
including the difficult Roman cases where the plea against the
enforcement of a promise is error in substantia, so that the thing
delivered is not the thing promised. The same approach that is
congenial to the law of tort carries over well here.
Other conditions are based on the sequence of performance that is
either stated in the agreement or which can be teased out from its
other terms or from a background set of default norms, the chief
purpose of which is to minimize the exposure that each side has over
the life of a complex contract, which happens when one party gets too
far ahead of the performance of the other side. The underlying
rationale is strongly economic. “The policy of the law, here as in the
tendency to construct concurrent conditions, is to minimize credit
risks” (Patterson 1942: 920). The simple rule that may demand either
payment or performance unless a party is prepared to tender his
obligations is one example of that basic position. Timing progress
payments on staged performance in construction contracts is yet
another way to minimize the risk that one party will get too far out
in front of the other.
There is, in the context of a short article on pleading, no possibility
of going through all these variations, each of which can be teased out
of the basic position of Roman and Anglo-American sales law. All
that can be done in this instance is to note that the well-organized
system of pleading tracks the usual strategy of a complete contingent
state contract, whereby, at each point in the sequence of performance,
each party knows what he must do to keep the contract alive, and
what consequences—damages, release of the other side, delay in
performance—will attend each type of nonperformance. Once
again, the right system of pleading does not make the substantive
judgments, but it does help set up the problem so as to allow doctrine
to be organized, and cases to be distinguished, with greater precision
that would otherwise be the case.

20.3. CONCLUSI ON
The strategy of this article has been two fold. It first develops the key
distinctions of Roman procedural law that helped facilitate the devel-
opment of the substantive law of property, contracts and torts. It then
applies those rules to a number of critical choices in the organization
of any legal system. To be sure, the historical development of these
rules reveals little or no conscious interest on the part of classical
Roman writers—or for that matter their English and American
peers—to find some explicit economic logic that guides the law of
pleading. But initial appearances can deceive, especially when it is
possible to examine these various doctrines with some degree of
closeness to demonstrate how they facilitate the articulation of a
substantive set of jurist-made or judge-made decisions that in large
measure have just that effect.
The Roman system proceeded in a bottom-up incremental fashion,
by breaking up complex cases and dividing them into their constitu-
ent parts that are then carefully reassembled into a large whole. This
effort to strip out one issue after another is quite different from the
modern approach in all substantive law, which invokes some over-
arching notion of reasonableness very early in the inquiry. There is no
doubt that every legal system ultimately involves an interest-
balancing inquiry that is implicit in these modern tests of reasonable-
ness. But the key point in this case is not whether these judgments
should be made, but how they should be made. I have indicated—and
this is the second main point of this article—how various maneuvers
dealing with the definition of terms such as iniuria can lead to an
oversimplification of tort and contract doctrine that conceal a set of
distinctions that, once laid bare, make the system more coherent than
if reasonableness were the starting point of analysis. The idea of
prima facie obligations has been around for a long time. The great
achievement of the Roman doctrine of pleadings is that it shows with
great force and clarity the way in which it can be extended from a
vague set of moral intuitions into a systematic body of law that
functions as well in modern times as it did in ancient times.¹³
¹³ My thanks to Mateo Aceves, University of Chicago Law School, Class of 2015,

and to Bijan Aboutorabi, Thomas Malloy, and John Tienken University of Chicago
Law School class of 2018—four accomplished Roman law students—for their invalu-
able research assistance on earlier drafts of this article.

REFERENCES
Brown, John Prather. 1973. “Toward an economic theory of liability,”

2 Journal of Legal Studies 323–49.
Buckland, William Warwick, Arnold Duncan McNair, and F. H. Lawson.
1965. Roman Law and Common Law: A Comparison in Outline.
Chang, Yun-Chien, and Henry E. Smith. 2012. “An economic analysis of civil
versus common law property.” 88 Notre Dame Law Review 1–55.
Danner, Henry R. 1912. Roman Law Pleading: An Outline of its Historical
Growth and General Principles. Indianapolis: H. R. Danner.
Epstein, Richard A. 1973a. “Pleadings and presumptions.” 40 University of
Chicago Law Review 556–82.
Epstein, Richard A. 1973b. “A theory of strict liability.” 2 Journal of Legal
Studies 151–204.
Epstein, Richard A. 1989. “The utilitarian foundations of natural law.”
12 Harvard Journal of Law and Public Policy 713–51.
Epstein, Richard A. 1997. “The modern uses of ancient law.” 48 South
Carolina Law Review 243.
Epstein, Richard A. 1999. Torts. Gaithersburg: Aspen Law & Business.
Epstein, Richard A. 2003. Skepticism and Freedom: A Modern Case for
Classical Liberalism. Chicago: University of Chicago Press.
Epstein, Richard A. 2005. “The not so minimum content of natural law.”
25 Oxford Journal of Legal Studies 219–55.
Epstein, Richard A. 2009. “The many faces of fault in contract law: or how to
do economics right, without really trying.” 107 Michigan Law Review
1461–77.
Epstein, Richard A. 2011. Design for Liberty: Private Property, Public
Administration, and the Rule of Law. Cambridge, MA: Harvard Univer-
sity Press.
Epstein, Richard A. 2012. “The irrelevance of the Hand formula: how
institutional arrangements structure tort liability,” in Jef De Mot, ed.,
Liber amicorum Boudewijn Bouckaert. Brugge: Die Keure, 65–76.
Geach, Peter T. 1960. “Ascriptivism.” 69 Philosophical Review 221–5.
Hart, Herbert L. A. 1949. ‘The ascription of responsibility and rights in
XLIX.” 49 Proceedings of the Aristotelian Society (New Series) 171–94.
Hart, Herbert L. A. 1983. Essays in Jurisprudence and Philosophy. Oxford:
Clarendon Press.
Holmes, Oliver Wendell. 1881. The Common Law. Boston: Little, Brown.
Lawson, Frederick H. 1950. Negligence in the Civil Law. Oxford: Clarendon
Press.
Lenel, Otto. 1927. Das Edictum Perpetuum: ein Versuch zu seiner Wiederher-
stellung. Leipzig: Tauchnitz.

Maitland, Frederick W. 1936. The Forms of Action at Common Law:
A Course of Lectures, ed. Alfred H. Chaytor and William J. Whittaker.
Metzger, Ernest. 2013. “An outline on Roman civil procedure.” 9 Roman
Legal Tradition 1–30. http://romanlegaltradition.org/contents/2013/RLT9-
METZGER.PDF.
Nicholas, Barry. 1962. An Introduction to Roman Law. Oxford: Clarendon
Press.
Nozick, Robert. 1974. Anarchy, State, and Utopia. Basic Books.
Patterson, Edwin W. 1942. “Constructive conditions in contracts.” 42 Columbia
Law Review 903–54.
Pitcher, George. 1960. “Hart on action and responsibility.” 69 Philosophical
Review 226–35.
Poithier, Robert Joseph 1761 [1806]. Traité du contrat de louage et traité des
cheptels, revised edition by M Hutteau. Paris: Letellier.
Ross, William David. 1930. The Right and the Good. Oxford: Clarendon
Press.
Stanford Encyclopedia of Philosophy, “William David Ross,” revised 19 June
2012, available at http://plato.stanford.edu/entries/william-david-ross/.
Schauer, Frederick F. 1976. “English natural justice and American due
process: an analytical comparison.” 18 William and Mary Law Review
47–72.
Sutton, Ralph. 1929. Personal Actions at Common Law. London: Butterworth
& Co. Ltd.
Williams, Glanville. 1951. “The aims of the law of tort.” 4 Current Legal
Probs. 137–76.
21
Private Prosecution and Enforcement

in Roman Law
David Friedman
21.1. INTRODUCTION
Modern legal systems use two mechanisms to impose costs on those

who injure others. Under tort law, the victim or his agents initiate and
prosecute the case; if the verdict is a monetary penalty, it is paid to the
victim. Under criminal law, the victim is replaced by the state. The
offense is treated as an offense against the state; it is the state that
initiates, controls, and pays for prosecution and, if the verdict is a fine,
collects it.
Anglo-American common law developed out of a legal system,
Anglo-Saxon law, in which offenses were for the most part treated
as torts; a killing created a claim for compensation by the kin of the
victim.¹ The state had a claim only if and to the extent that the offense
was also seen as a breach of the king’s peace. In a purer form of such a
system, exemplified by the legal system of saga-period Iceland, not
only the prosecution of the case but also the enforcement of the
verdict was entirely the responsibility of the victim.² The state’s
role was limited to authorizing the use of private force through its
system of laws and courts. Even that role is missing in a still purer
¹ For an extensive survey of wergeld in Anglo-Saxon and other early societies, see
Seebohm (1902).
² Friedman (1979); Friedman et al. (2019: 147–69).
David Friedman, Private Prosecution and Enforcement in Roman Law In: Roman Law and Economics:
DOI: 10.1093/oso/9780198787211.003.0021
328 David Friedman

feud system such as the traditional Somali legal system,³ based on
customary law and the private arbitration of disputes.
The logic of a feud system is simple. If someone wrongs me,
I demand compensation and threaten to harm him if I do not get it.
If someone kills me, someone else, typically my heir, inherits my
claim and the duty to enforce it. In order for this to work as a way of
discouraging wrongs, the threat must be more believable when I have
been wronged than when I have not—there must be some mechanism
such that right makes might. In both the Icelandic and Somali
systems, the relevant mechanism is a court trial, in the former case
in a pre-existing court system, in the latter in a court created for the
case under traditional procedures. In the feud system currently exist-
ing among the Romanichal gypsies in Britain, the mechanism is more
informal. When I accuse you of wronging me and threaten to beat
you up if I do not receive suitable compensation, both of us know that
if you did wrong me according to the norms of our community, my
friends will back me and your friends will not back you.⁴
The role of the avenger of blood, the heir of the victim of a killing in
Jewish law, suggests that that legal system developed out of a pre-
existing system of decentralized, privately enforced law. If the court
finds the killing capital, it is the avenger of blood who is supposed to
execute the sentence. If the court instead sentences the killer to
temporary exile,⁵ the avenger of blood has the right to kill him if he
can catch him on his way to the city of refuge.⁶ The law forbids the
³ A good account of the Somali system can be found in Lewis (1961). See also
Friedman et al. (2019: 170–82).
⁴ Weyrauch (2001: ch. 3).
⁵ Exile is in one of the cities of refuge and lasts until the death of the high priest. See
Neusner (1988: The Order of Damages, Makkot, 2:4–2:7, 614–15); Klein (1954:
V.1.195).
⁶ Another piece of evidence that Jewish law was built on top of a system of privately
enforced law is the treatment of situations where the amount of a debt is ambiguous—
where, for instance, a carelessly drafted document refers to the sum as 50 zuz in one
place and 100 zuz in another. The court can only compel the payment of 50 zuz, since
it does not know that the larger sum is owed. But if the creditor has seized 100 zuz
from the debtor, the court cannot compel him to return part of it, since it does not
know that 100 zuz was not owed. “Therefore, in the case of every instrument which
contains a recital with two different meanings, one referring to a greater sum and the
other referring to a lesser sum, the holder of the instrument collects only the lesser
sum. But if he seized the greater sum it may not be reclaimed from him without clear
proof ” (Rabinowitz 1949: 187). A similar pattern occurs with tort damages: “If an
animal stamps on the ground on the premises of the plaintiff and as a result pebbles
spring up and cause damage there, the defendant must pay for (only) one quarter of
Private Prosecution and Enforcement in Roman Law 329

avenger of blood from agreeing to a money payment instead,⁷ which
suggests that the equivalent of the Icelandic wergeld had at some
previous point been a common practice.
In Jinayat, the part of Muslim law that applies to one person killing
or injuring another, the offense is treated as an offense against the
victim or, if he is killed, his kin. The court determines guilt and sets a
penalty, either a payment in camels or retaliation against the offender.
The victim or kin can either enforce the penalty or agree to waive it,
gratis or on condition of some lesser penalty.⁸
Can a similar pattern can be found in Roman law? Is there evidence
that it too developed out of a system in which prosecution of offenses
and the enforcement of verdicts was primarily the responsibility of
the victim?
21.2. THE HISTORY
The traditional date for the foundation of Rome is 753 BC. The
political system was converted from monarchy to republic, again
according to the traditional dating, in 509 BC, and to an empire in
27 BC. The first written text of the law which can be substantially,
although not completely, reconstructed⁹ is the Twelve Tables, com-
posed according to the Roman accounts in 451–450 BC. While there
was extensive writing on the law over the next thousand years, the
next complete official account of the law was Justinian’s codification
project of AD 529–34.¹⁰
the damage . . . . But if the plaintiff seizes one-half of the damage, we may not take it
away from him” (Klein 1954: XI.1.2, part 6, p. 8). That implies a system where such
claims were at least sometimes privately enforced.
⁷ “and Scripture says, Moreover ye shall take no ransom for the life of a murderer
(Num. 35:31)” (Klein 1954: V.1.194).
⁸ Hallaq (2009: 310, 320–2); Friedman et al. (2019: 98–9). There is an additional
penalty due in the form of penance—freeing a Muslim slave or fasting during daylight
for two consecutive months.
⁹ Reconstruction is from multiple sources, not all consistent with each other. For a
discussion, see Crawford (1996).
¹⁰ Consisting of the Institutes, Digest, and Codex. There is, however, an almost
complete legal textbook by Gaius from the second century, Gordon and Robinson
(1988), which is a major source of information about the legal system at and before
his time.
330 David Friedman

Over that period of time there were major changes in the law, the
result of legislation by the assemblies of the people and edicts of the
magistrates, along with interpretation, first by the pontiffs and then
by jurists. By the late Republic, the ius honorarium, the law laid down
by the magistrates, had substantially transformed the existing legal
system. The process continued under the empire, supplemented by
imperial constitutions. The pattern of development in civil procedure
was one of an increasing role for the magistrate as judge, a reduced
role for the parties.¹¹ Roman law as codified under Justinian, the form
in which it was passed down to influence later legal systems, was, like
modern law, a system in which state judgment and enforcement
played a central role.
21.3. PRIVATE LAW: STRONG SENSE
Originally, the term ius (plural, iura) denoted that which is due in
human relations—the rightful power of a community member to act
in a certain manner vis-à-vis his fellow citizens. It referred to a course of
conduct that the community would take for granted and, in that sense,
endorse. Thus, a person who appropriated an object, entered upon land,
ejected or imprisoned another individual may in so doing be exercising
ius. The community had a general awareness of the circumstances when
such acts would be construed as iura and these were established by
custom. . . . At this stage, the exercise of ius had no connection with state
organization and thus ius was defined as any instance of approved self-
help. (Mousourakis 2007: 19–20)
To the legislators of the Twelve Tables, however, an interpretation of
criminal law in which the state played a part was still entirely foreign.
For them the natural and only result of a crime was the victim’s right
of vengeance, and they were merely concerned to restrict the right of
physical vengeance to the more serious crimes, to keep this right under
judicial control, to isolate the wrongdoer who had been found guilty,
and thus to protect the commonwealth from the effects of devastating
family vendettas. (Kunkel 1973: 29)
As these passages by two scholars of Roman law suggest, a central
characteristic of the legal process of the early Republic was its reliance
¹¹ du Plessis (2015: 79–83).


on private action by the plaintiff.¹² It was his job to bring the
defendant into court, if necessary by force.¹³ A debtor could bind
himself by pledging his body as security; if he defaulted, his creditor
could seize him without a prior court verdict for enslavement or
execution.¹⁴ If the court verdict created a debt by the defendant to
the plaintiff and the defendant failed to pay in thirty days, the plaintiff
was entitled to seize him and, if the debt was not paid, eventually to
sell him into slavery or execute him.¹⁵ As in the Icelandic system, the
court authorized the use of force but it was, in most or all cases,
¹² Mousourakis (2007: 64) describes litigation as essentially private arbitration

under the approval of the state, functioning as a substitute for self-help. For detailed
discussion of the early history of Roman courts, see Kaser and Hackl (1996: 25–34).
¹³ Twelve Tables, Table I.
“1. If the plaintiff summons the defendant to court the defendant shall go. If the
defendant does not go the plaintiff shall call a witness thereto. Only then the
plaintiff shall seize the defendant.
2. If the defendant attempts evasion or takes flight the plaintiff shall lay hand
on him. . . .
3. If one of the parties does not appear the magistrate shall adjudge the case, after
noon, in favor of the one present” (Johnson et. al. 1961).
¹⁴ “The Twelve tables contain a harsh form of debt-contract, in which the bor-
rower, through the receipt of the money weighed out to him before witnesses, literally
gave himself into the power of the creditor (whence the transaction’s name, nexum,
‘binding’). If he could not free himself punctually by repayment he fell into a
condition of debt-bondage without the necessity of the judgement of a court” Kunkel
(1973: 26). These options were abolished by the Lex Poetelia (326 BC).
¹⁵ Twelve Tables, Table III. Execution of Judgment:
“1. Thirty days shall be allowed by law for payment of confessed debt and for
settlement of matters adjudged in court.
2. After this time the creditor shall have the right of laying hand on the debtor.
The creditor shall hale the debtor into court.
3. Unless the debtor discharges the debt adjudged or unless someone offers surety
for him in court the creditor shall take the debtor with him. He shall bind him
either with a thong or with fetters of not less than fifteen pounds in weight, or if
he wishes he shall bind him with fetters of more than this weight.
4. If the debtor wishes he shall live on his own means. If he does not live on his
own means the creditor who holds him in bonds shall give him a pound of grits
daily. If he wishes he shall give him more.
5. . . . Meanwhile they shall have the right to compromise, and unless they make a
compromise the debtors shall be held in bonds for sixty days. During these days
they shall be brought to the praetor (possibly anachronistic for consul) into the
meeting place on three successive market days, and the amount for which they
have been judged liable shall be declared publicly. Moreover, on the third
market day they shall suffer capital punishment or shall be delivered for sale
abroad across the Tiber River” (Johnson et. al. 1961).
332 David Friedman

private force. And, as in that system, there was a schedule of
payments owed for injuries to the victim.¹⁶
All of these conclusions must be qualified by the limitations of the
available evidence. The earliest written text of the law described in
Roman tradition, the laws of the kings, the so-called ius Papirianum,
ascribed to a Papirius, a putative contemporary of Tarquin the Proud,
has not survived; some scholars doubt that it ever existed, although
purported fragments exist in later literary sources.¹⁷ While the law of
the Twelve Tables certainly existed and was for a thousand years
regarded by the Romans as the foundation of their legal system, no
complete text survives, leaving us with only incomplete reconstruc-
tions based on passages quoted in later texts.¹⁸
The information on the law of the Twelve Tables can, however, be
supplemented by descriptions of legal process as it existed in the early
Republic and by later enactments that, by abolishing rights of private
action, demonstrated that they once existed. Thus, for instance, as
Paul du Plessis (2015: 346) points out: “There had to be a dishonest
and forcible taking of property; if the taker acted in good faith, he was
not liable. This rule appears to have exonerated those who violently
enforced genuinely held claims. Not surprisingly, Marcus Aurelius
decreed that bona fide claims should be forfeited if enforced in a
violent manner.”¹⁹ And a decree in AD 389 (C. 8.4.7) provided that an
owner who forcibly seized his property should forfeit both ownership
and possession, suggesting that self-help was still common enough to
be seen as an issue at that late a date.²⁰
¹⁶ Gai., Inst. 3.223: “Under the Twelve Tables the penalty for this delict was, for a
damaged limb, retaliation (talio); for a broken or bruised bone, on the other hand it
was 300 ‘asses’ if a free man’s bone had been broken but 150 if it was a slave’s; for all
other contempts (inuriae), on the other hand, the penalty established was twenty-five
‘asses’ ” (trans. Gordon and Robinson 1988).
¹⁷ Mousourakis (2007: 23). “The collection attributed to Papirius is probably an
apocryphal publication of the end of the Republic, or of the time of Augustus” (Girard
1906: 32). On the so-called leges regiae, see Johnson et al. (1961: 3–6); Crawford et al.
(1996: 561–3); and du Plessis (2015: 28–9).
¹⁸ Girard (1906: 51–2). Watson (1992: 14–20) provides an account of the sources
for the reconstruction and an analysis of its reliability.
¹⁹ The context was the action for violent theft (actio vi bonorum raptorum).
²⁰ du Plessis (2015: 182). Gaius (Inst. 4:155) describes a similar result at an earlier
date. “Sometimes, however, even if the person whom I have forcibly dispossessed was
possessing from me by force, stealth or licence, I am compelled to restore possession
to him, for instance, if I expelled him by force of arms. For by such gross wrongdoing

Litigation in early law was “a private arbitration established under
the approval of the State, as a substitute for self-help, the business of
the State officials being only to see that this arbitration is conducted in
proper form” (Buckland and McNair 1952: 400). Mousourakis sug-
gests that the early procedure for trial (legis actio) “probably derived
from the practice established by custom where contested claims were
voluntarily submitted to arbitration, and must have been in habitual
use before its formal adoption.” In the form in which it was recorded,
it consisted of two stages, the first before a consul or pontiff, the
second “before a citizen appointed as the judge (iudex) by the magis-
trate and the parties concerned” (Mousourakis 2007: 32). In the
earliest and most important of the legal actions (legis actio sacra-
mento), each party deposited a sum of money as a wager that it was in
the right. The prevailing party recovered its deposit, that of the losing
party was forfeited to the state. The outcome of the trial demonstrated
to other parties which claimant was in the right, providing the
mechanism by which right made might in that society. At least one
scholar interprets the procedure of each party swearing to the truth of
his claim as a way of getting the priests, who were in charge of legal
procedures, to provide a verdict, since a false oath was a religious
offense.²¹
The law, as it later developed, distinguished between ius and lex.
The former corresponded roughly to the underlying principles defin-
ing the rights of individuals towards each other.²² The later was
legislation, specific rules created by an assembly to enforce ius. That
I must in all circumstances submit to the action restoring possession to him” (trans.
Gordon and Robinson 1988).
²¹ “The two parties took an oath of the truth of their pretensions, in such a way that
there was necessarily on the one side or the other perjury,—a sin and, consequently, a
delict, in this epoch when religion and law were not separated, and when, in order to
know who had incurred the penalty, it was necessary for the king, head of religion as
of criminal justice, to inquire who was right” Girard (1906: 28). And, on the same
point in the context of the Twelve Tables, Girard (1906: 59) writes: “But we see by
what roundabout means legal process is here grafted onto procedure under which one
takes justice for oneself without legal process.”
²² “In this broad normative sense ius is not the same as morality nor as positive law;
rather, it is right law, or positive law as it ought to exist in light of what morality and
justice ordain. Ius, as defined above, was distinguished from lex (plural leges). The
latter term signified a law created by a competent legislative organ of the state in
conformance with a prescribed procedure. During the Republic the term lex was used
to denote a statute enacted by a popular assembly and created in a form directed to all
citizens” Mousourakis (2007: 20).
334 David Friedman

suggests that ius may have represented the system of privately
enforced customary law that preceded explicit legislation.
Early Roman law also made a distinction between ius, private
rights, and fas, divine rights. Where state intervention in English
law was initially justified on the grounds that some acts that violated
individual rights also violated the king’s peace, state enforcement in
Roman law seems to have originated with enforcement of religious
rules.²³ Exactly where the line was drawn between prosecution by
magistrates and private action is unclear. “The distinction was not
between capital and pecuniary cases: private actions, e.g. against a
thief or a debtor, might give the prosecutor the right to kill or enslave
a citizen, while on the other hand tribunes often prosecute criminally
for a fine” (Jones 1972: 38). It sounds from Jones’s account as though
public prosecution was primarily for offenses actually against the
state—treason, military incompetence, peculation.
The legal status of murder in the early law is unclear. By the time of
the Twelve Tables, it was treated as a public rather than private
offense; the penalty was not a monetary payment to the heirs of the
victim, as in Icelandic and Somali law, but execution. There is,
however, a passage sometimes ascribed to the Twelve Tables that
suggests that at some point blood revenge was a legitimate response
to a killing.²⁴
The rules in the Twelve Tables, at least as we have them, are not
entirely consistent. One passage forbids putting anyone to death
without a trial while another specifies that it is legal to kill someone
who steals by night, provided that the killer raises a clamor to draw
²³ “Priestly officials were entrusted to enforce the religious duties as prescribed by

the norms of fas. These officials independently discharged their task by devising and
administering coercive practices” (Mousourakis 1007: 19). Note also that the earliest
form of action involved an exchange of oaths and a wager upon them placed in the
hands of the pontiffs.
²⁴ Table VIII, 24a (VIII 13 Crawford). “If a weapon has sped accidentally from
one’s hand, rather than if one has aimed and hurled it, to atone for the deed a ram is
substituted as a peace offering to prevent blood revenge.” Crawford et al. (1996: 693)
discuss a tradition that this rule originated with the laws of Numa. So it may reflect a
rule from a point before the punishment of killing shifted from privately to publicly
enforced. Alternatively, the phrase “to prevent blood revenge,” present in the recon-
struction by Johnson et al., absent in some others, might be a mistake and the ram
could be interpreted as an offering to the gods. Jolowicz (1932: 178) suggests that the
ram was “given to the agnates—a relic of the time when the agnates of the slain man
were entitled to take vengeance on the slayer.”

attention to the fact.²⁵ Similarly, one passage permits retaliation in
kind for a broken limb if compensation is not agreed on, while the
next specifies a fixed money payment.²⁶ This suggests that, at the time
of the composition of the Twelve Tables, the law was in the process of
moving from a system of private vengeance with the possibility of
private compensation to one of court authorized force with predeter-
mined monetary payments. That is consistent with Mousourakis’s
view that self-help was established in some royal decrees prior to the
republic, with retaliation (talio) for some personal injuries.²⁷
21.4. PRIVATE ENFORCEMENT OF CONTRACTS

AND JUDGMENTS: SURETIES, DISTRAINT,
AND HONOR
Our modern view of the surety and his functions is almost entirely
based on the fideiussor of classical Roman law. But he is at once
the product of a lengthy evolution and the symbol of a relatively
advanced social and economic organization. Behind him, it is true,
loom the dim figures of the uindex, the uas and the praes, but these
survive in the works of the great jurisconsults only as the ghosts of a
vanished order . . . . (Binchy 1970: 356)
Societies that lack a state apparatus willing and able to enforce
contracts and legal judgments develop other mechanisms for the
purpose, one of which is the institution of suretyship. Early Irish
law provides examples which, Binchy argues, may represent the
survival of institutions common to other early Indo-European legal
systems, including the Roman. To what extent is that conjecture
supported by what we know of suretyship in Roman law?
²⁵ IX 6. “For anyone whomsoever to be put to death without a trial and convicted . . . is

forbidden.” VIII 12. “If a thief commits a theft by night, if the owner kills the thief, the
thief shall be killed lawfully” (Johnson et al. 1961). One might interpret this as distin-
guishing between execution by a magistrate and legal killing by a private party.
²⁶ VIII 2: “If anyone has broken another’s limb there shall be retaliation in kind
unless he compounds for compensation with him.” VIII 3: “. . . If a person breaks a
bone of a freeman with hand or by club, he shall undergo a penalty of 300 asses; or of
150 asses, if of a slave . . .” (Johnson et al. 1961).
²⁷ Cf. du Plessis (2015: 29).
336 David Friedman

The Irish system recognized three sorts of sureties created in the
process of forming a contract.²⁸ The naidm, or enforcing surety, had
the job of guaranteeing that the party for whom he was a surety would
fulfill his obligations and was entitled to use force to make him do so.
The rath, or paying surety, was obligated to pay what his party owed if
the party failed to do so, and he was also entitled to collect from his
party what he had paid plus an additional sum. The aitire, or hostage
surety, guaranteed his party’s obligations not with his wealth but with
his body. If the party defaulted, the surety was obliged to surrender
himself to the claimant, eventually ransoming himself back—and,
again, obtaining a claim against his party for considerably more
than the sum originally owed.
The fideiussor in Roman law was to some degree the equivalent of
the Irish rath. He was obligated to pay what the party for whom he
was a surety owed if the party himself failed to do so. There might be
multiple sureties for a single contract. In the initial version of the
institution, the creditor could claim the entire debt from any of them.
A rule created under Hadrian allowed a surety to require a creditor to
divide his claim against any surety by proportion in the total number
of sureties. Redress of sureties among themselves was only firmly
fixed by Justinian in AD 535.²⁹ The praes seems to have been an earlier
form of the fideiussor, combining the roles of rath and aitire, guar-
anteeing a debt with both his estate and his person.³⁰ Another early
form of surety was the vas, who pledged to pay a fixed amount if the
third party he was guaranteeing failed to fulfill an obligation such as
paying a debt or appearing before a court.³¹
²⁸ For details, see Kelly (2009); Binchy (1970).

²⁹ For earlier redress among sponsores and fidepromissores (but not fideiussores) see
Gai., Inst. 3.122 (against other sureties) and 3.127 (against the creditor).
³⁰ “To guarantee performance of a duty by nominating praedes was the original
form the Romans had used in archaic times. It meant that if the duty was not duly
fulfilled by the contractor himself, the surety was bound in a personal way: he was
obligated through nexum or mancipium to the creditor. For the later period that
concerns us, this form of surety had been abandoned in private law and was replaced
by the principle that the surety was a second debtor who guaranteed the debt with his
means (i.e. his entire estate) but not with his person. The surety was no longer called
praes but was referred to as sponsor or fideipromissor or fideiussor, in accordance with
the formal legal act that established this special relationship between the creditor and
the person willing to guarantee for a debtor” (Rainier 2013: 178).
³¹ “In the oldest times the ‘vadimonium’ served the purposes of a suretyship.
Vadimonium was a solemn promise to pay a specified penalty, if a certain third
party failed to meet a particular obligation (e.g. to pay a debt or answer a summons

The vindex in early Roman law plays a different role from any of
the Irish sureties. The vindex could intervene at the point when a
debtor was about to be seized and led off in bonds by his creditor,
taking the obligation on himself and apparently freeing the debtor
from it. It was then up to him to show that the debt was not owed. If
he failed, he was liable for twice the original amount.³²
While the Roman forms of suretyship differ from the Irish, their
existence is consistent with Binchy’s suggestion that institutions for
enforcement in a decentralized system that survived in Irish law late
enough to be recorded in some detail may date back to a much earlier
point in the history of Indo-European institutions.
Another institution of which the same may be true is distraint.
Irish law provided elaborate mechanisms permitting a claimant to
seize cattle in satisfaction of his claim by a series of stages, taken in
presence of witnesses, which the defendant could interrupt by agree-
ing to arbitration of the claim.³³
The Roman equivalent, predating the Twelve Tables, was the
action-at-law of pignoris capio.³⁴ It is unclear what the plaintiff
could do with the property seized, other than holding it in order to
force the defendant to perform his obligation. Nor is it clear in what
sorts of cases it was permitted. But it is a self-help remedy of the sort
needed in a legal system where state enforcement is absent or weak,
one that may, as Binchy has suggested, have been more similar to the
Irish distraint in the period before the Twelve Tables.³⁵
before court). A ‘vas’ is not a surety in our sense of the word, because the liability he
undertakes is not the same as that of the principal, but is a new liability with different
contents . . .” Sohm (1892: 8 n. 2).
³² “In the process described by Gaius and Gellius the intervention of the vindex
apparently set the debtor free, and all liability was assumed by the vindex. This is
clearly a later stage in an institution whose origin has been the subject of considerable
speculation” (Schiller 1910: 206; Gai., Inst. 4.21, 25).
³³ Kelly (2009: 177–82). ³⁴ du Plessis (2015: 71–2).
³⁵ “Further, it seems to me reasonably certain that the legis actio in the older
Roman law called pignoris capio is also a vestigial relic of this archaic right of self-
help, even though by the time of Gaius it had become a rigidly circumscribed and
conventionalized exception. Indeed, this was the view held by most (if not all)
previous legal historians, though some modern Romanists reject it. But have not
these scholars lost sight of the fact that in the earliest stages of Roman society, well
before the Twelve Tables, when Rome was still, to use Ludwig Wenger’s word, a
Stammstaat, a remedy of this kind must have existed and that the old term for it was
retained in classical Roman law (as happened to many such terms in other legal
systems) for a much more specialized procedure?” (Binchy 1973: 24).
338 David Friedman

In Irish law every individual had an honor price determined by his
status, affecting both the amount owed for offenses against him and
the upper limit of the amount he could commit himself to in a
contract or as a surety. One consequence of some offenses was the
loss of honor price. The nearest Roman equivalent was the reduction
in status (but not in limits on suretyship) that made someone an
infamis or intestabilis or both as a result of some sort of discreditable
conduct.
As du Plessis (2015: 107–9) describes, infames could not “hold
offices or positions of honor, could not vote (at least in early law), or
bring criminal accusations, or appear as advocates, or act as repre-
sentatives (or be represented) in litigation.” Intestabiles “could not act
as witnesses, either in litigation or in formal transactions (such as the
making of wills or the conveyance of property).”
One further feature of early Roman legal procedure is how formal-
istic it was—a plaintiff could lose his case for putting it in the wrong
words.³⁶ The same seems to have been true of Icelandic law. In one
passage in Njal’s Saga, a step in a legal procedure requires a formula to
be recited three times. Two of the three recitations may be imprecise,
but one must be verbatim.³⁷ Such rigid rules may make sense in a
decentralized system of privately enforced law; it is important that
third parties can tell who is legally in the right, and in the Icelandic
context there is no superior authority to decide what variations in the
rules are or are not legitimate.
Given the limitations in our source materials, it is hard to be
certain of the nature of Roman law in the monarchy and early
Republic, but such evidence as we have is consistent with the view
that, like other early forms of law, it originated as a privately enforced
feud system.
³⁶ Mousourakis suggests that the formality of the law reflected the religious origin
and character of many legal rules. An example is a plaintiff who charged the defendant
with cutting his vines when the legal formalism required him to describe them as trees
in order to fit the wording of the legal rule in the Twelve Tables (Mousourakis 2007:
29). Gai., Inst. 4.11: “This is why the opinion was given that a man who raised an
action over the cutting down of vines in a way that used the word ‘vines’ in the action
had lost his case. He ought to have used the word ‘trees,’ because the Twelve Tables,
under which the action for cutting down vines was available, spoke in general terms
about cutting down trees” (trans. Gordon and Robinson 1988).
³⁷ Magnusson and Palsson (1960: 76–8).

21.5. PRIVATE LAW: WEAK SENSE
The system of procedure, or lack of a system, constituted a major

disincentive to involvement in litigation, especially in Republican
Rome, since the State’s involvement in civil procedure was minimal
(in both the legis actio and formulary procedure). This was because of
the notion that litigation was essentially “a private arbitration estab-
lished under the approval of the State, as a substitute for self-help, the
business of the State officials being only to see that this arbitration is
conducted in proper form.” (Buckland and McNair 1952: 400)³⁸
So far I have been offering evidence that Roman law developed from a
legal system that was private in the strong sense of a feud system,
privately prosecuted and enforced. To what extent, especially in
the later and better-documented period, was there private law in the
weaker sense, law that, like modern tort and contract law, is privately
prosecuted but publicly enforced?
The answer is that, through the republican period and into the
early Empire, Roman civil law was private in a sense intermediate
between the two. Under both the earlier legis actio and the later
formulary system, both getting the defendant to court and executing
the judgment were the responsibility of the plaintiff, not the state.
Under the earlier system, if the defendant, once summoned, failed to
either come to court with the plaintiff or find a guarantor (vindex) for
his later appearance, the plaintiff was entitled to call witnesses and
then drag the defendant to court by force. If he was unable to do so
and the defendant failed to appear, the defendant was considered
indefensus, “a status that entitled the plaintiff to seize his property on
the authorization of the praetor.”³⁹ Under the later formulary system,
the defendant “could make a promise (vadimonium) to appear on a
particular day, in which case he would promise to provide security
and to pay a penalty for failure to appear at the agreed time in the
proximity of the court. . . . Sanctions were imposed on an indefensus—
a defendant—who tried to avoid being summoned (e.g. by hiding) or
who refused to obey the summons or to provide a vindex. “For
example, the praetor could allow the plaintiff the possession of the
defendant’s estate, with a possible right of sale. The right to seize a
³⁸ du Plessis (2015: 64). ³⁹ du Plessis (2015: 66).

340 David Friedman

reluctant defendant and drag him to court still survived but was
probably used only as a last resort” (du Plessis 2015: 72–3).
In modern legal systems, we take for granted that judgments will be
effectively enforced (“executed”) with the full authority and backing of
the State. But for much of Rome’s history, the state took only an indirect
role in the execution of judgments. It authorized the successful plaintiff
to pressurize the defendant into complying with the judgment—a form
of regulated self-help, the onus being firmly on the plaintiff to obtain
satisfaction. (du Plessis 2015: 70)
Under the legis actio, if the case was in rem, for possession of a thing,
and the winner was already in possession he remained so. If the loser
was in possession but had granted temporary security at an earlier
stage in order to remain so, the security was forfeited to the plaintiff.
“If the security was unforthcoming or insufficient, further proceed-
ings could be brought to assess the compensation payable . . . . When
the amount had been assessed, execution of the judgment proceeded
along the lines of a case in personam . . .” (du Plessis 2015: 70).
If the case was in personam, for enforcing an obligation owed by
the defendant to the plaintiff, the successful plaintiff in the early
period could enforce the verdict either by manus iniectio, physical
seizure of the defendant, or by pignoris capio, seizure of his property,
the latter option being available only for certain categories of cases. In
the former case, the defendant had thirty days in which to pay the
judgment, after which the judgment creditor could seize him, take
him before a magistrate, and if he neither paid nor produced a vindex
to take his place imprison him for sixty days. If nobody came forward
to pay the debt, at the end of that time the judgment debtor could be
sold into foreign slavery or put to death. A rule instituted in 326 BC
eliminated those alternatives. “Instead, the creditor could keep the
debtor for as long beyond the 60 days’ limit as it took for the latter to
work off the debt” (du Plessis 2015: 71). Under the later formulary
system, the creditor was eventually given the additional option of
seizing and selling property of the debtor in satisfaction of the debt.⁴⁰
The formulary system was introduced in the third century BC,
initially for cases involving foreigners, and made available more
generally in the next century. The earlier legis actiones system “fell
largely into disuse in the late Republic and was formally abolished by
⁴⁰ du Plessis (2015: 77–8).


Augustus in 17 BC . . . ” (du Plessis 2015: 72). It was in turn
supplemented in the imperial period by the cognition system and
formally abolished in AD 342. That change reflected the conversion of
civil law from a system privately prosecuted and to a large extent
privately enforced to something more like modern civil law, with the
responsibility for enforcement shifting from the plaintiff to the state.
Under the type of procedure that depended on an “investigation”
(cognitio) by a single magistrate, the plaintiff lodged a statement of his
claim with the magistrate and the magistrate ordered the defendant to
undertake to appear in court and provide security that he would do
so. If not, he could be arrested by the magistrate. If the defendant lost
the case and failed to pay up, his property could be seized by court
bailiffs and sold.
Roman law took its final and most complete form under Justinian
in the Digest, a lengthy account of the law in the form of selected
passages from the writings of earlier authorities, the Codex, a com-
pilation of imperial enactments, and the Institutes, a condensed
version intended for students. One of the first things that strikes a
modern reader is how much of it is privately prosecuted. Theft, for
example, is prosecuted by the victim, either the owner of the stolen
property or someone else injured by its theft. The convicted thief
owes damages of twice or four times the value of what has been stolen,
the larger amount if caught in the act.⁴¹ The law distinguishes what
we would describe as compensatory damages from the additional
punitive sum. The former is an obligation that is inherited by the
thief ’s heirs, the latter is not.
The distinction between private and public prosecutions, in Roman
law as in Athenian law,⁴² is that public prosecutions “are called public
because as a general rule any citizen may come forward as prosecutor
in them” (Just., Inst. 4.18.1).
Finally, it should be observed that a person who has been outraged
always has his option between the civil remedy and a criminal indict-
ment. If he prefers the former, the penalty which is imposed depends, as
we have said, on the plaintiff’s own estimate of the wrong he has
suffered; if the latter, it is the judge’s duty to inflict an extraordinary
penalty on the offender. (Just., Inst. 4.4.10)
⁴¹ Just., Inst. 4.1; Gai., Inst. 3.189–90. The thief is also liable to return what was
stolen by an action in rem or its value by an action in personam.
⁴² MacDowell (1978: 57–8).
342 David Friedman

The Digest distinguishes between civil and criminal law, but most
criminal law, as in eighteenth-century England,⁴³ appears to have
been privately prosecuted.⁴⁴ That raises, for a modern reader, the
question of what distinguishes the two.⁴⁵ The answer is never made
explicit in the texts. At least part appears to be that criminal convic-
tion results in punishment of the criminal but not in a damage
payment or transfer of property to the private prosecutor. Another
difference is that in a criminal case the accuser was not free to drop
the case, as legislated in the Senatus Consultum Turpillianum of
AD 61.⁴⁶ Conceptually, from the beginning of the legal system, crim-
inal law had to do with offenses against the gods or the polity, civil
law with disputes among individuals.
As in a modern legal system, the same act might sometimes be
prosecuted as either a civil or a criminal wrong, as shown by the
passage quoted above. According to a constitution of the emperors
Valens, Gratian, and Valentinian, “whenever both a private law action
and a criminal proceeding lie concerning property, both may be
brought, without reference to which one is brought first . . . ” (C. 9.31.1
pr.; Trier, AD 378).
21.6. WHY?
Over a period of about a thousand years, Roman law developed from

a legal system in which most claims were privately prosecuted and
⁴³ Friedman (1995); Friedman et al. (2019: 223–58).

⁴⁴ This raises the interesting question of what the incentive was for a private party
to accuse another of a criminal offense. In Periclean Athens, the successful private
prosecutor typically received a sizable share of the fine paid by the convicted defend-
ant, but there seems to have been no equivalent mechanism in Roman law except
perhaps the delatores who notify the imperial Fiscus of certain events (D. 49.14.1 pr.;
Callistr. 1 de iure fisci) in exchange for a reward (lex Papia Poppaea, see Tacitus,
Annals 3.25 and 3.28). Where the accuser was a victim of the crime, the obvious
incentive would be revenge—or, to put it in something closer to the language of
economics, a commitment strategy of prosecuting those who committed crimes
against one in order to produce deterrence as a private good. Another incentive, as
in the Athenian case, might be to injure an enemy. And one cannot dismiss the
possibility that at least some Roman citizens would regard enforcement of the
criminal law as a civic obligation.
⁴⁵ du Plessis (2015: 325–6) discusses the civil/criminal distinction and the inter-
mediate status of the law of delicts.
⁴⁶ See D. 48.16 and C. 9.45. In eighteenth-century English law, it was an offense for
a private prosecutor to compound a felony—accept a payment to drop charges.

privately enforced to something more like a modern legal system,
although with a larger role for the victim as prosecutor than in
modern criminal law. A similar pattern can be observed in English
common law.⁴⁷ That raises the question of why, in both cases, the
change occurred.
Broadly speaking, there are three answers. One is that people
discovered that state enforcement of law was superior to private
enforcement—the Whig theory of history broadly defined.⁴⁸ One is
that circumstances changed so that, while private enforcement might
have been the best viable option in the early period, by the later period
it no longer was, and the legal system changed accordingly. Both of
those explanation depend on some mechanism to push legal institu-
tions towards optimality. The second also requires an explanation of
what changed in the environment of the legal system and why those
changes implied a change in how that system ought to function. An
example of such an explanation in the English case would be the shift
from eighteenth-century private prosecution of crime to nineteenth-
century public prosecution, arguably due to an increasing urbaniza-
tion that reduced the victim’s reputational incentive to prosecute.
The third answer is that the change occurred not because it was an
improvement but because making it was in the interest of those
controlling the legal system. The clearest example of such a change
for which that is at least a plausible conjecture is the shift from private
to public enforcement, from what was in effect a tort system to a
criminal system, in early English law. Under Anglo-Saxon law, as
under the contemporary and better recorded Icelandic system, the
chief legal effect of what we would consider a crime was to establish a
claim for damages by the victim or his heirs against the offender. Over
time, under late Anglo-Saxon and then Norman law, that was con-
verted to a claim by the crown for damages, forfeiture of property, in
capital cases a fine paid in exchange for a pardon. The result was that,
⁴⁷ Actually, it can be observed twice, once in the transition from early Anglo-Saxon
to medieval Norman, a second time in the transition from privately prosecuted (but
publicly enforced) criminal law in the eighteenth century to publicly prosecuted law
by the latter half of the nineteenth century.
⁴⁸ Narrowly defined, the Whig theory described in Butterfield 1930 is the view of
English history propounded by nineteenth-century historians, such as Macauley and
Stubbs, that interpreted past historical events in terms of their role in producing the
desirable institutions of the present. Broadly defined, it is the view of history as
trending towards a more nearly optimal set of institutions.
344 David Friedman

in medieval England, the right to run courts and collect fines was
viewed as a valuable part of the bundle of rights held by a feudal lord,
sometimes purchased from the Crown.⁴⁹
I do not know enough about the workings of the Roman legal
system as it developed to say if something similar is a plausible
explanation in that case. Perhaps someone who does can.
21.7. CONCLUSI ON
Roman law probably originated in the context of a decentralized

system of self-help, a feud system where enforcement was private
and the court system provided a mechanism for providing an
authoritative verdict on private disputes. In the earliest form of
which we have a good account, the private role both in bringing the
defendant to court and in enforcing the verdict made Roman civil law
more private than modern tort law. As the system developed, under
first the Republic and then the Empire, the state took an increasingly
active role. But even in its final form as codified by Justinian, pros-
ecution of both civil and criminal offenses was primarily the respon-
sibility of private citizens.⁵⁰
REFERENCES
Binchy, D. A. 1970. “Celtic suretyship: a fossilized Indo-European institu-

tion?,” in George Cardona, Henry M. Hoenigswald, and Alfred Senn, eds,
Indo European and Indo-Europeans: Papers Presented at the Third Indo-
European Conference at the University of Pennsylvania. Philadelphia:
University of Pennsylvania Press, 355–67.
Binchy, D. A. 1973. “Distraint in Irish law.” 10 Celtica 22–71.
Buckland, William Warwick. 1957. A Manual of Roman Private Law,
2nd edn. Cambridge: Cambridge University Press.
⁴⁹ I discuss some of these issues in Friedman (1996). For examples of revenue from
law enforcement under the Angevins, see Warren (1978): 177.
⁵⁰ I would like to thank Egbert Koops for extensive advice and the correction of
some of my errors on Roman law and Dennis Kehoe for updating my references. Also
Giuseppe Dari-Mattiacci for giving me a reason to learn some, if not enough, about
the subject.

Buckland, William Warwick, and McNair, Arnold D. 1952. Roman Law &
Common Law A Comparison in Outline, 2nd edn. Cambridge: Cambridge
University Press.
Crawford, Michael H. 1996. Roman Statutes, vol. II. London: Institute of
Classical Studies.
du Plessis, Paul. 2015. Borkowski’s Textbook on Roman Law, 5th edn. Oxford:
Friedman, David. 1979. “Private creation and enforcement of law: a historical
case.” 8 Journal of Legal Studies 399–415.
Friedman, David. 1995. “Making sense of English law enforcement in the
eighteenth century.” 2 The University of Chicago Law School Roundtable
477–85.
Friedman, David. 1996. “Beyond the tort/crime distinction.” 76 Boston
University Law Review 103–12.
Friedman, David, Peter Leeson, and David Skarbek. 2019. Legal Systems Very
Different from Ours. Self-published.
Girard, Paul Frédéric. 1906. A Short History of Roman Law, trans. Augustus
Henry Frazer Lefroy and John Home Cameron. Toronto: Canada Law
Book, Co.
Hallaq, Wael B. 2009. Sharia: Theory, Practice, Transformations. Cambridge:
Johnson, Allan Chester, Paul Robinson Coleman-Norton, and Frank Card
Bourne. 1961. Ancient Roman Statutes: A Translation, Commentary,
Glossary, and Index. Austin: University of Texas Press.
Jolowicz, H. F. 1932. Historical Introduction to the Study of Roman Law.
Jones, A. H. M. 1972. The Criminal Courts of the Roman Republic and
Principate. Totowa, NJ: Rowman and Littlefield.
Kaser, Max, and Karl Hackl. 1996. Das römische Zivilproxeßrecht, 2nd edn.
Munich: Beck/Hackl.
Kelly, Fergus. 2009. A Guide to Early Irish Law. Dublin: Institute for
Advanced Studies.
Klein, Hyman, trans. 1954. The Book of Torts, Book Eleven of the Code of
Maimonides (Mishnah Torah). New Haven and London: Yale University
Press.
Kunkel, Wolfgang. 1973. An Introduction to Roman Legal and Constitutional
History, 2nd edn. based on the 6th German edn. of Romische Rechts-
geschichte, trans. J. M. Kelly. Oxford: Clarendon Press.
Lewis, I. M. 1961. A Pastoral Democracy: A Study of Pastoralism and Politics
among the Northern Somali of the Horn of Africa. Oxford: Oxford Uni-
versity Press for the International Africa Institute.
MacDowell, Douglas M. 1978. The Law in Classical Athens. Ithaca: Cornell
University Press.
346 David Friedman

Magnusson, Magnus, and Hermann Palsson, trans. 1960. Njal’s Saga.
London: Penguin.
Mousourakis, George. 2007. A Legal History of Rome. London and
New York: Routledge.
Neusner, Jacob. 1988. The Mishnah: A New Translation. New Haven and
London: Yale University Press.
Pound, Roscoe. 1906. Readings in Roman Law. Lincoln: Jacob North & Co.
Prichard, A. M. 1961. League’s Roman Private Law, 3rd edn. London:
Macmillan.
Rabinowitz, Jacob J., trans. 1949. The Code of Maimonides Book Thirteen:
The Book of Civil Laws. New Haven: Yale University Press.
Rainer, J. Michael. 2013. “Public building contracts in the Roman Republic,”
in Thomas A. J. McGinn, ed., Obligations in Roman Law: Past, Present,
and Future. Ann Arbor: University of Michigan of Michigan Press,
174–88.
Schiller, A. Arthur. 1978. Roman Law: Mechanisms of Development. Berlin:
Mouton De Gruyter.
Seebohm, Frederic. 1902. Tribal Custom in Anglo-Saxon Law. London: Long-
mans, Green, and Co.
Sohm, Rudolf. 1892. The Institutes of Roman Law. Oxford: Clarendon Press.
Warren, W. L. 1978. King John. Berkeley: University of California Press.
Watson, Alan. 1992. The State, Law and Religion: Pagan Rome. Athens and
London: University of Georgia Press.
Weyrauch, Walter O., ed. 2001. Gypsy Law: Romani Legal Traditions and
Culture. Berkeley: University of California Press.
Ancient References
Moyle, J. B., trans. 1913. The Institutes of Justinian. Oxford: Clarendon Press.
http://gutenberg.readingroo.ms/5/9/8/5983/5983-h/5983-h.htm.
Frier, Bruce W., ed. 2016. The Codex of Justinian: A New Annotated Trans-
lation with Parallel Latin and Greek Text, 3 vols. Cambridge: Cambridge
University Press.
Gordon, W. M., and O. F. Robinson. 1988. The Institutes of Gaius. Ithaca:
Cornell University Press.
Johnson, Allan Chester, Paul Robinson Coleman-Norton, and Frank Card
Bourne. 1961. Ancient Roman Statutes: Translation, with Introduction,
Commentary, Glossary, and Index, gen. ed. Clyde Pharr. Austin: Univer-
sity of Texas Press. http://avalon.law.yale.edu/ancient/twelve_tables.asp.
22
Deterrence of Wrongdoing in
Ancient Law
Francesco Parisi, Daniel Pi,
Barbara Luppi, and Iole Fargnoli
Deterrence is among the principal objectives of modern legal systems

in a variety of disparate subject areas. For example, the criminal law
seeks to deter individuals from undertaking certain “wrongs”; tort
law seeks to deter individuals from exercising less than due care when
engaging in risky activities; environmental law seeks to deter pollution
of the environment, and contract law seeks to deter individuals from
reneging on promises. Indeed, broadly interpreted, the deterrence
objective may be transposed to nearly every realm of human activity
in which the law manipulates incentives to influence behavior.
So fundamental is the concept of deterrence to the law, it
may easily be imagined that the very first legal institutions were
created with this precise purpose in mind. Yet framing the function
of law in terms of deterrence has not historically been regarded as
intuitively obvious. Even in criminal law, arguably the most natural
application of the deterrence rationale, legal scholars are still today in
disagreement about whether deterrence matters,¹ or indeed whether
¹ See e.g. Farrington et al. (1994); Langan and Farrington (1998); Hirsch et al.
(1999); and Gendreau et al. (1999) (finding that increasing prison sentences actually
correlated with a 3 percent increase in recidivism). See generally Tonry (2008) and
Webster and Doob (2012) (surveying the empirical research contradicting deterrence
theories’ predictions).
Francesco Parisi, Daniel Pi, Barbara Luppi, and Iole Fargnoli, Deterrence of Wrongdoing in Ancient Law In: Roman
Law and Economics: Exchange, Ownership, and Disputes, Volume II. Edited by: Giuseppe Dari-Mattiacci and
Dennis P. Kehoe, Oxford University Press (2020). © Oxford University Press.
DOI: 10.1093/oso/9780198787211.003.0022
348 Francesco Parisi, Daniel Pi, Barbara Luppi, and Iole Fargnoli
deterrence should matter.² Thus, it is unlikely that protean legal
institutions consciously embraced deterrence as a policy objective.
Nevertheless, we find that in many cases modern legal systems have
evolved not only to effect deterrence, but indeed to incentivize effi-
cient levels of deterrence; this raises the question how the law evolved
to so precisely achieve the deterrence objective in such a wide variety
of contexts without necessarily recognizing it explicitly.
In this chapter, we explore the evolution of deterrence in ancient law.
We shall offer a combination of historical, comparative, and economic
analyses in identifying the salient forces at work in the emergence,
development, and refinement of the deterrence objective in early legal
systems. We begin in §22.1 with a description of the evolutionary and
cultural origins of “wrongdoing” and the desire for revenge in eco-
nomic terms. We trace the historical development from these inherited
biological traits to the intertribal norm of lex talionis. In §22.2, we
analyze the deterrent effects of the principles of lex talionis. In §22.3, we
describe the economic pressures that effected a shift from communal
liability to individual liability. In §22.4, we describe the evolution of
compensatory damages from the principle of lex talionis, which would
later develop into the modern law of torts. §22.5 traces a parallel branch
of legal evolution, describing the emergence of criminal law from lex
talionis. We conclude with some general remarks on the role of the
deterrence objective in the formation and maturation of law.
22.1. “WRONGDOING” AND RETRIBUTION
Let us postpone defining what “wrongs” are and provisionally assume

the naïve view that wrongs are offenses under some exogenously
determined standard. Instead, let us focus our attention on the effect
that wronged individuals tend to desire vengeance. We begin our
analysis by asking why victims of wrongs seek to “repay” the disutility
they have suffered by causing disutility to their injurers. One plausible
hypothesis is that victims of wrongs experience positive utility from
vengeance, and that “revenge” is simply the product of private utility
maximization.³ The utility they get from vengeance outweighs the
² See e.g. Lewis (1953); Kant (1965); Duff (1986); and Hirsch (1993).
³ Posner (1980: 27) seems to suggest something like this.
Deterrence of Wrongdoing in Ancient Law 349

cost of undertaking revenge activities. Recent psychological research
seems to indicate that this explanation is dubious, finding that
individuals tend not to experience greater happiness after exacting
revenge.⁴ However, the hypothesis may be partially recovered by
distinguishing between the pursuit of revenge and its result: we may
plausibly suppose that the act of exacting revenge itself provides
utility, though the end state of having achieved revenge may not.
The fact that individuals tend to seek revenge seems to reveal that
they perceive the pursuit of revenge at least to be a desirable activity.
To explain this preference, we find the argument from evolutionary
psychology persuasive: that individuals have a biologically inherited
predisposition toward retaliatory behavior, manifest by the sense that
offending parties deserve retaliation for perceived offenses. Naturally,
knowing that victims will have a predisposition for revenge creates an
expected cost for would-be injurers, which deters “wrongful” con-
duct, thereby reducing the incidence of wrongs and improving the
welfare of would-be victims.⁵ Parisi and Fon (2005) show that this
behavior can lead to equilibria with low rates of “wrongful” conduct.⁶
When an individual from one tribe harms an individual from
another tribe, either because the act of harm itself creates utility for
the injurer, or because injury results as the byproduct of some other
activity that creates utility for the injurer, an externality arises. If
members of the victim’s tribe desire revenge as a consequence of
the injury, then their utility-maximizing choice will often be to
retaliate, causing harm to the injurer, reducing the benefit extracted
from the activity.
We may describe the utility of the injurer by the function:
UI ðwÞ ¼ BðwÞ C πρðκÞ; ð1Þ
where B(w) is the benefit of the harmful activity, w is the activity level,
C is the opportunity cost, π is the probability of revenge being
⁴ See e.g. Carlsmith et al. (2008), finding that successful revenge tends to generate
disutility for the avenged.
⁵ It is relatively straightforward to understand how such a phenotype would be
selected. See e.g. Jacoby (1983); Chagnon (1988).
⁶ We think the intuition underlying this analysis to be reasonably straightforward.
Readers interested in the details of the formal analysis should consult Parisi and Fon
(2005). See also Posner (1988: 27).
triggered, and ρðκÞ is the magnitude of revenge, given the revenger’s
investment in revenge κ. Clearly, when πρ > B C, the rational
prospective injurer will choose to engage in some other activity.
That is, he will be deterred from the harmful activity.
We may describe the utility of the victim by the function:
UV ðκÞ ¼ HðwÞ κ; ð2Þ
where H(w) is the magnitude of harm, and κ is the cost of exacting

revenge. Let us assume that revenge is costly, but that it creates a
@ρ
greater cost for the injurer than the revenger. That is, @κ >1. However,
assuming that the game is played sequentially, such that injurers
move first and victims respond, it seems that the rational choice for
injurers is not to seek revenge. That is, maxκ UV ¼ 0. This result is
true for one-shot interactions, but does not hold for repeat inter-
actions, where an investment in κ >> 0 may leads to a reduction in
equilibrium of w over multiple rounds of play, resulting in better
long-run payoffs than a strategy of κ ¼ 0.
It is plausible that our prehistoric ancestors were incapable of
making such conclusions. The absence of abstract language in early
humans or limited cognitive faculties in pre-human hominids may
have rendered the information cost of rational deliberation in
repeated games prohibitive. Yet if we modified equation 2 so that
@r
the act of exacting revenge generated utility rðκÞ, such that @κ >1 for
an interval ð0; nÞ, then some level of revenge-seeking would maximize
the function:⁷
UIR ðκÞ ¼ HðwÞ κ þ rðκÞ: ð3Þ
Such behavior would improve the long-term prospects of tribes,

facing repeat interaction with neighboring tribes. It is therefore
plausible that such a revenge-preferring phenotype would tend to
succeed through a process of natural selection.⁸
⁷ The literature on “negative reciprocity” is very well developed. For more detailed
discussions of this topic, see e.g. Güth et al. (1982); Camerer and Thaler (1995); Roth
(1995); Falk and Fischbacher (2006); and Falk et al. (2008).
⁸ Parens (2014: 170).

We may observe this protean form of “retributive justice”
manifested in a variety of early societies.⁹ For example, the informal
intertribal law of Native American communities consisted, in many
cases, exclusively of the “law of revenge.”¹⁰ Biblical accounts
also describe revenge-motivated “justice” in early Judeo-Christian
society.¹¹ The honor value of “blood revenge” also figured promin-
ently in norms of ancient Athens,¹² Arab society,¹³ as well as in pre-
modern Japanese society.¹⁴ Indeed, nearly all legal systems seem to
have sprouted from the seed of satisfying the pre-legal desire for
vengeance.¹⁵ Nearly the same happened in Roman law, which later
developed a set of rules that has served as a model source of law in
several civil law countries for many centuries after the fall of the
Roman Empire, until the age of the modern codifications.¹⁶ At the
beginning of the Roman legal experience the victim had the right to
revenge, although retaliatory justice was administered under the
control of the community.¹⁷
This begs the inquiry: how exactly did the desire for vengeance
mature into a social norm and later into a basis for law? In addition to
the innate desire for revenge identified by evolutionary psychology,¹⁸
we observe notions of desert and retribution taking shape in early
societies, reinforced intra-tribally through “duty” and “honor” norms,
often bolstered by religious rules.
⁹ Though for present purposes, we may describe the desire for revenge as retribu-
tive, it bears distinguishing revenge from retribution. Retributivists in the philosophy
of punishment nearly unanimously distinguish retribution from revenge. See Moore
(1997: 89). We do not wish to err in conflating revenge and retribution. Therefore, we
intend special emphasis to be understood in our use of the qualifier “protean” here.
Though retribution is not conceptually identical with revenge-seeking, we think
notions of retributivism in moral philosophy arose from the pre-philosophical desire
for revenge as a historical matter.
¹⁰ See e.g. Reid (1970); Hudson (1976); Blackburn (1980); Chagnon (1988).
¹¹ e.g. Genesis 4:10–24; Proverbs 6:13; II Samuel 12:13–18.
¹² Allen (2001) (“Anger was thus assumed to be not only the source of particular
punishments but also at the root of law itself. The Athenians accordingly felt relatively
little uncertainty or unease about why (that is, in response to what causes) they
punished: they acted in response to anger”).
¹³ Ginat (1997). ¹⁴ Mills (1976).
¹⁵ An interesting exception may be observed in Indian culture, where incapacita-
tive objectives seem to have been primary in the development of criminal and tort law.
See Das Gupta (1930); Doongaji (1986); Olivelle (2011).
¹⁶ See Manthe (2000: 55–6). ¹⁷ Jhering (1906: 118).
¹⁸ See Joyce (2006).
Though the evolutionary basis for revenge-seeking is clear from the
game-theoretic analysis, this alone is not sufficient to explain why
communities embraced it as a desirable practice. It does not follow
that activities which are explainable through biological evolution are
necessarily reified as social norms.
To understand a community’s motivation for adopting a social
norm of revenge-seeking, we note that wrongs were not regarded,
insofar as “justice” was concerned, as perpetrated by or against
individuals. Instead, the family, clan, or tribe was regarded as being
collectively responsible for wrongs committed by members and for
the retaliation of wrongs committed by other tribes.¹⁹ Tribe members
who failed to exact vengeance for a wrong were regarded by their
community as dishonorable and faced intra-tribal social sanctions.
Suppose that a member of a revenge-seeking tribe had the utility
function described in equation 2. That individual could free-ride on
the revenge efforts of his fellows, enjoying the benefit of the deter-
rence that they generated by investing in revenge, while expending
less or no κ himself.
This situation is analogous to a cartel game. For any given indi-
vidual, the benefit of defection will tend to generate greater benefits
(i.e. the effort not dissipated by participating in communal revenge)
than the loss due to decreased deterrence (i.e. the reduction in
protection from deterrence). By not participating in revenge-seeking,
such an individual could improve his utility at a cost to the tribe.
Thus, the social norm of revenge-seeking reinforces the utility derived
from revenge (for individuals who possessed idiosyncratically weak
revenge-seeking preferences), while reducing the utility of free-riding
(by imposing intra-tribal social sanctions).
This process of social reinforcement begins the transformation
from biologically motivated preference into social norm. Yet the
road to legal rule is only just begun. Historically, the bare retaliation
norm tended to be a short-lived stage of development.²⁰ Bare revenge,
though it solves the repeated game problem, is a potentially
¹⁹ See Posner (1980) for a detailed description of the communal unit in archaic
societies and an economic explanation for its development. For the role of commu-
nities in specific archaic societies, see also Reid (1970); Hudson (1976); Mills (1976);
Blackburn (1980); Chagnon (1988).
²⁰ At this stage of development, it is dubious whether retaliation is even a social
norm—we think it something intermediate between an evolutionary instinct and a
social norm.

catastrophic norm. When a multiplier is attached to retaliation (i.e.
when ρ ¼ aH, where a > 1), a revenge act may be regarded as a fresh
offense by the original injurer, leading to spirals of escalating revenge.
Initially, a “kind-for-kind” restriction arose to constrain the type of
retaliation warranted by the harm suffered. The kind-for-kind con-
straint restricts the scope of available retaliatory options to like harms.
For instance, theft warranted revenge-theft, and killing warranted
revenge-killing. The kind-for-kind constraint is later reinforced by
the notions of “desert” and “talis” (equal in kind) captured by the
etymology of the word “talio” (retaliation) in ancient Roman law.
These social norms helped to tame the useful but potentially volatile
revenge impulse. Yet, though early norms constrained retaliation as
kind-for-kind, limitations on magnitude arose only later.²¹
Legal scholars describe this initial stage of development, when the
norm of kind-for-kind retaliation became established, as “discretionary
retaliation.” If the member of one tribe harmed the member of another
tribe, the offended tribe was obliged to seek like vengeance (of variable
magnitude) upon the offending tribe. The retaliatory conduct at this
stage was still often subject to an arbitrary multiplier. For instance,
Genesis 4:14 states, “[W]hosoever slayeth Cain, vengeance shall be
taken on him sevenfold.” Similarly, II Samuel 12:13–18 prescribes a
four-fold measure, while Genesis 9:5 prescribes a two-fold measure.²²
Though it constrained the type of revenge-harm, discretionary retali-
ation could still lead to an escalating infliction of reciprocal harms. For
example, one group, regarding the retaliation of another group as
unwarranted or excessive, might counter-retaliate with vengeance
of greater magnitude, inciting the other group to counter-counter-
retaliate with vengeance of yet greater magnitude.²³ Such cycles of
increasing violence would clearly be undesirable for the participants
in the long run, and discretionary retaliation norms tended to evolve
limitations on the permissible magnitude of retaliation.²⁴
²¹ Blau (1916: 7).

²² Interestingly, Sulzberger (1915: 33) analyzes the two-fold measure as arising
from the doctrine of “double blood-guilt.” The retaliation is given twice: once for the
individual injurer, and again for the injurer’s tribe for failing to prevent or punish the
crime internally.
²³ See Parisi (2001).
²⁴ The escalation of blood feuds is commonly observed in societies with discre-
tionary retaliation norms. See e.g. Richter (1992) on escalating “mourning wars”
between Iroquois tribes. Parisi (2001) suggests several elements that may lead to
The second stage of development is characterized by the emergence
of the lex talionis norm—that retaliatory harms should be equal in
magnitude to the harms suffered, popularly captured by the phrase,
“an eye for an eye.”²⁵ This development had the obvious benefit of
preempting the mechanisms that led to devastating spirals of recip-
rocal violence. It is at this stage of social development that moral
conceptions of desert and retribution began to take shape. The impulse
to strike back transforms into a sense that harms are “moral wrongs,”
for which an equivalent retaliatory harm is “owed.” In economic terms,
the inchoate moral calculus implicit in this new norm acted as an upper
bound on ρ, such that ρ H.
It is easy to see why the upper bound of lex talionis is an improve-
ment over unbounded ρ. First, it reduces excessive expenditure on κ
on the victim’s side, which generates over-deterrence when ρ > H.
Second, it sets a benchmark for proportional retaliation, against
which injurers are neither inclined nor obliged to counter-retaliate.
Third, it eliminates the mechanism for spiraling retaliations of
increasing magnitude.
Talionic law was an important development in the formalization of
early legal institutions. In addition to specifying the magnitude of
liability (the distinction between criminal liability and tort liability
had not yet developed), procedural clarifications facilitated the
orderly execution of talionic retaliation. In the biblical tradition,
the institution of Go’el (nearest of blood) determined the right of
the victim and his extended family to exact vengeance.²⁶ This effect-
ively reduced coordination problems that arose in the discretionary
retaliation paradigm, when it was often unclear upon whom the duty
to exact revenge fell, possibly leading to duplicative efforts/harms.
The exceptional case of ancient Indian law follows a similar history
of formalization. Though Indian law was anomalously based not on
retributive, but rather incapacitative grounds, its early development
included a comparable systemization of consequences, specifying
upper bounds for punishments. For example, the harshest punish-
ment for theft was dismemberment of the hands. Though this may
seem draconian by modern standards, it is important to recognize
that the function of this formal rule was not to expand, but rather to
escalating retaliatory actions, including variations in the perceptions of blameworthi-

ness of a wrongdoer, and overestimation of the gravity of harm suffered.
²⁵ See Miller (2007). ²⁶ Good (1967).

delimit the magnitude of available remedies. The king (ideally an
impartial official) was also specified as the legal executor of punish-
ments, with procedural constraints imposed by the Brahmins, further
reducing the risks of overinvestment in punishment.²⁷
Although the second stage of development is characterized by the
emergence of the principle of lex talionis, several other important
changes in the proto-legal institutions of early societies may be
observed. First are the codification and increasing formalization of
the law, which not only specifies the measure of retaliation due, but
also delimits the duties and rights of parties after the occurrence of a
harm. Second, and more abstractly, it should be observed that the
rationale underlying retaliation transformed from the pre-legal, prim-
ordial desire for revenge into a moral prescription for desert. It is also
important to notice that in economic terms, the principle of lex
talionis crudely assigns a measure of retaliation equal to the harm
suffered, ρ H, thus forcing potential injurers to internalize the cost
of their “wrongdoing.” It thus approximates, however crudely, some-
thing approaching an efficient measure of deterrence.
In its earliest period, Roman law applied talionic law for cases of
iniuria, including the case of intentional injury to another with the
breaking of a limb or a bone or insult.²⁸ But in a different case
Romans also admitted a non-proportional retaliation. The Twelve
Tables,²⁹ an archaic law promulgated in the fifth century BC, pre-
scribed that the victim could lawfully kill the thief, if he had been
caught in the act and was operating at night.³⁰ The murder was the
lawful reaction for theft.
It is also possible to ascribe to a legalization of vengeance the first
personal compulsory settlement that the Romans ordered in the law
of Twelve Tables, the manus iniectio (laying of a hand upon a
person).³¹ This procedure authorized a private citizen to formally
²⁷ See Das Gupta (1930); Doongaji (1986); Jaishankar and Haldar (2004); Olivelle
(2011).
²⁸ Law of the Twelve Tables, Table 8.2: Si membrum rup[s]it, ni cum eo pacit, talio
esto (“If a victim is maimed, then in the absence of compensation, retaliation is in
kind”).
²⁹ See Humbert (2005).
³⁰ Law of the Twelve Tables, Table 8.12: Si nox furtum faxsit, si in occisit, iure
caesus esto (“If the theft has been done at night, if [the owner] kills the thief, [the thief]
shall be considered to have been lawfully killed”).
³¹ Law of Twelve Tables, Table 1.1: Si in ius vocat,<ito>. Ni it, antestamin[o]: igitur
em capito (“When [the plaintiff] brings a case against [the defendant] before the court,
arrest the debtor, if—without giving a guarantor (vindex)—he did not
present himself at the trial as defendant or, independently from the
trial, he did not fulfill his debt.³² The creditor could sell the debtor as
addictus (slave until the discharge of his debt), but if there was no
buyer found at the market, the creditor could lawfully kill him.³³ The
cruel eventuality to cut the body of the debtor into pieces was
probably related to the case in which there was more than one
creditor.³⁴
These conceptual transformations were important ones. Recall that
the predisposition to seek revenge seemed best explained in evolu-
tionary terms as a means of effecting deterrence. This biologically
successful phenotype was tamed by social practice into a retributive
ground for retaliation. Yet this transformation brought us full circle
back to what appears to be a crude attempt at effecting optimal
deterrence, revealing the mechanism likely at work in effecting the
institutional changes.
This evolution of the grounds for punishment suggests a closer
relationship between retributive and deterrence theories than many
scholars have hitherto supposed.³⁵ Rather than being two competing
objectives of criminal law (and to a lesser extent, tort law), “retribu-
tion” may thus be regarded as a refinement of the deterrence objective
achieved through social norms.
22.2. “WRONGDOING” AND DETERRENCE
Many features of lex talionis suggest that it was a far more dynamic
system than often supposed, designed to take into account more
complex elements, such as general deterrent value, enforcement
[the defendant] should appear. If [the defendant] doesn’t appear for trial, [the plaintiff]
should call the witnesses. Afterwards he should catch him [the defendant]”).
³² Kaser (1971: 132).
³³ Law of Twelve Tables, Table 3.6: Tertiis nundinis partis secanto. si plus minusve
secuerunt, se fraude esto (“On the third market day they ought to cut off pieces of the
corpse. Whether they have cut off too much or too little, should be left unpunished”).
³⁴ Talamanca (1990: 293).
³⁵ Though a substantial number of “mixed” theories of punishment have recog-
nized the close relationship between retribution and deterrence rationales. Duff
(1986); Hirsch and Ashworth (2005); Moore (2007).

errors, and specific enforcement problems. As we remarked earlier,
the ancient body of laws dealing with “wrongdoing” treated both
tortious and criminal behavior as one, often without distinguishing
between intentional, culpable, or accidental wrongdoing (Cairns
and Robinson 2001). Modern legal systems, with their separate
treatment of torts and crimes, tailor remedies according to the
intentional or non-intentional nature of the act, generally pursuing
goals of absolute deterrence through criminal remedies (teaching
prospective wrongdoers that “crime does not pay”), and goals of
relative deterrence through tort remedies (teaching prospective
tortfeasors that “precautions pay”). Specifically, modern tort law
forces wrongdoers to compensate the victim for the harm suffered
(H). By contrast, modern criminal law creates deterrence by impos-
ing sanctions S based on the benefit enjoyed by the injurer B and the
probability of detection, ρ, oftentimes irrespective of H. In modern
criminal law the sanction is tailored to the incentives of the criminal.
For example, thefts of sheep for slaughter or for work are both
punished as thefts; that is, the punitive element of the sanction is
the same. Deterrence is achieved when the benefit to the wrongdoer
is less than the probability of enforcement multiplied by the sanc-
tion, i.e. B < pS, where S is independent of H. This difference
between criminal law and tort law influences the effects of the two
remedial regimes: modern tort law pursues goals of relative deter-
rence (“precautions pay”) by forcing the internalization of the loss
H, while criminal law pursues goals of absolute deterrence (“crime
does not pay”) taking into account the criminal’s temptation to
commit the crime, B.
Talionic regimes do not bifurcate remedial instruments and tackle
both objectives of criminal and tort remedies at once. By replicating
the harm suffered by the victim onto the wrongdoer, and giving the
parties an option to settle through the payment of blood-money
(kofer), lex talionis provided a hybrid remedy that led parties to
take into account both the harm to the victim and the benefit to the
wrongdoer.
As will be discussed below, the effectiveness of the talionic regime
hinges upon the ratio of the benefit to the offender to the harm that he
caused. Talionic law looks at the ratio B/H to achieve punitive and
compensatory goals. Since the sanction is tied to the loss to the victim,
and because it is the victim’s (or his tribe’s) right to impose this
penalty, both the victim’s loss and the benefit to the wrongdoer are
Unified Law of Wrongs
(B/H)
Tort Law (B) Criminal Law (H)
Figure 22.1. Deterring wrongdoing.
ultimately taken into account, balancing objectives of absolute and

relative deterrence in view of the ratio between B=H. Figure 22.1
illustrates how the talionic system relates to its tort and criminal
law alternatives.
When different instruments are available to pursue relative deter-
rence of unintentional wrongdoing through tort remedies and absolute
deterrence of intentional wrongdoing through criminal remedies, prob-
lems of deterrence spillovers across different categories of wrongdoing
can be reduced and enforcement levels can be tailored to the respective
policy objectives (Parisi 2001: Dari-Mattiacci and Parisi 2004).
When a single set of remedies is adopted, the choice of optimal
enforcement levels becomes essential for the pursuit of effective
deterrence. Prior to the lex talionis regimes, where kind-for-kind
was the only requirement for punishment, multipliers were used for
retaliation: the punishment imposed on the wrongdoer was more
severe than the harm originally suffered by the victim (Sulzberger
1915). When the lex talionis imposed a measure-for-measure limit to
the punitive remedy, possible situations of underdeterrence could
emerge.³⁶ With enforcement errors, only a fraction of wrongdoers
would be caught and punished, and prospective wrongdoers could
expect to face retaliatory punishment only a fraction of the time.
The relationship between retaliatory punishment and deterrence is
particularly interesting when considering enforcement levels.³⁷ With
retaliatory punishment, a lower level of enforcement is needed to
deter “inefficient crimes.” Retaliatory punishment instead becomes
less effective in deterring “efficient crimes” (i.e. those crimes for
which the benefit to the injurer is equal or greater than the harm to
³⁶ Blau (1916: 7).

³⁷ See Czabanski (2008: ch. 4) for a discussion of the relationship between retali-
ation and deterrence in ancient law. See also Stein (1999).

the victim). To understand the “deterrence paradox” of retaliatory
punishment with enforcement errors, it is necessary to take into
account the disparity between the benefit achieved by the wrongdoer
and the expected cost of punishment. With retaliatory punishment,
deterrence is preserved for inefficient crimes, but underdeterrence
may arise for relatively efficient crimes. Retaliatory punishment thus
links the harm to the offender to the harm suffered by the victim.
When the multiplier α is equal to one and with a less than 100 percent
enforcement rate, deterrence cannot be achieved for efficient crimes.
Unlike the internalization of the harm in tort law, absolute deter-
rence of crime requires offsetting the offender’s benefit with a threat
of punishment. Deterrence should teach the wrongdoer that crime
does not pay. A rational criminal should be adequately deterred if the
benefit from the crime, B, is less than the expected sanction, equal to
the probability of enforcement, p, times the sanction S.
B < pS
Under lex talionis regimes, the sanction S is set equal to the victim’s
loss H. Hence, enforcement should occur with a probability p,
such that
p > B=H
Crimes are generally inefficient. This is because the benefit that the
wrongdoer captures from committing the wrong is generally less than
the cost he imposes on the victim, B < H. As a result, a wrongdoer
would not generally agree to suffer retaliatory punishment in
exchange for the right to impose harm on his victim. The degree of
inefficiency of a crime can vary greatly, 0 B=H < 1. Lower values of
B=H indicate “more inefficient” crimes.
Despite the mechanical proportionality of retaliatory punishment
(i.e. more malicious crimes lead to more malicious sanctions), S ¼ H,
it is interesting to observe that, as H increases relative to B, the
enforcement level necessary to preserve adequate deterrence actually
decreases. When enforcement is imperfect, it becomes rational for
some wrongdoers to carry out some wrongful activities, and the
wrongs that will be rationally perpetrated will be those that are
generally less inefficient (i.e. those characterized by higher values
of B=H).
The deterrence paradox is easily found in early legal rules.³⁸
The crimes mentioned in the early codes that impose measure-for-
measure retaliatory punishment (murder, mutilation, wounding, etc.)
are crimes that generally take a lot from the victim and give a smaller
benefit to the perpetrator. For example, the crimes covered by the lex
talionis are characterized by lower values of B=H. The optimal
enforcement necessary to ensure deterrence ends up being lower for
the most inefficient crimes. For these crimes, enforcement level could
be kept low, without compromising effective deterrence. Policing and
enforcement of brutal crimes could be kept low, while policing of less
violent crimes would paradoxically need to be boosted.
Contrary to a retributivist view of punishment, given that policing
and enforcement are costly, when retaliatory penalties are utilized,
more serious crimes should be enforced less rigorously. As the dead-
weight loss of the crime becomes smaller (i.e. when B=H goes up),
society should instead strengthen policing and enforcement meas-
ures. Paradoxically, this means society should devote more resources
to deter efficient (i.e. less inefficient) crimes, or else adopt a different
standard of punishment.
The foregoing analysis further explains why early legal systems did
not apply the measure-for-measure cap on punishment to crimes that
entail a larger prospective benefit for the wrongdoer (Parisi 2001).
The optimal enforcement necessary to ensure deterrence in such
cases would be higher. One could imagine crimes such as theft
where the benefit to the victim approaches the sanction in a kind-
for-kind regime such as lex talionis. To promote adequate deterrence,
early legal systems adopted higher α > 1 in the case of theft, to account
for the lack of deterrence inherent to theft.³⁹ This ensured that
the ratio B/S would still be below the probability of enforcement.
A measure-for-measure requirement for theft would have been futile
here, while for violent crimes it would have been effective. Within the
confines of theft, further adjustments were made to correct for pos-
sible detection problems. These deterrent aspects of lex talionis are
further implicated by the higher penalties for animals marked for
slaughter than animals for work. For example, animals intended for
³⁸ See Kavka (1978) for a discussion on the emergence of deterrence paradox. See
also Fagan and Meares (2008) focusing on the deterrence paradox in minority
communities.
³⁹ e.g. Maine (1861: 378).

slaughter had a higher multiplier than those marked for work. Sheep
had a multiplier of 5:1, while oxen had a multiplier of 4:1.⁴⁰ This is
due to the specific detection problems with respect to animals
intended for slaughter. Animals for slaughter may likely be stolen
and perhaps eaten shortly after their theft. The perpetrator could
literally swallow the evidence. Animals for work, however, may be
recognized by a unique brand or other individual characteristics of
the animal. Since the probability of enforcement was necessarily
lower for slaughter animals, the lex talionis regime compensated by
increasing the ratio of B to H, making enforcement less necessary for
adequate deterrence, once again suggesting an understanding of the
role of B/H in ancient times.
Yet even within the category of work animals, special exceptions
were made when they considered two animals for work, both of
which need to be retrained.⁴¹ Here the animals are differentiated by
loyalty. One only takes orders from the owner; thus, instructions from
strangers will not be executed. Other animals would listen to the new
owner immediately. A corollary today would be the difference
between a dog, loyal to its owner and a horse, generally more likely
to obey whoever is riding it.
Considering this problem in terms of social efficiency, we can see
that the theft of an animal that one would have to retrain is less
efficient. Intuitively, one would expect that theft to be more socially
undesirable and it would have to be deterred more severely. However,
that may still encourage crimes of animals that more easily adapt to a
new owner.
In fact, to discourage theft, lex talionis punished thefts that
required no retraining with greater multipliers. Again the B/S ratio
does not look at loss to the victim, but it rather looks at the benefit to
the perpetrator relative to the sanction imposed on the same perpet-
rator. If the prospective perpetrator can capture the full value of an
animal, that is a very temping theft, or a high B value. The animals
that needed to be retrained had a lower B value. To compensate for
the higher B value of non-loyal animals, a higher S value was imple-
mented as well. Again, deterrence played the primary role in norms
dictating the penalties of the time.
⁴⁰ Exodus 22:3–4. ⁴¹ The Code of Hammurabi, Par. 8.

22.3. THE EVOLUTION OF
INDIVIDUAL LIABILITY
The lex talionis principle marks a stage in the evolution of ancient

law, after which civil and criminal penalties begin to diverge. Before
exploring these two branches of development, we pause to consider
an orthogonal development: the evolution of individual liability.
In nearly all early societies, there existed a clear division between
intertribal and intra-tribal norms. The revenge laws governing
the interaction between tribes did not necessary apply to the internal
interactions of members within the same tribe. The intertribal revenge
norms gradually evolved into modern legal systems. Meanwhile, intra-
tribal norms grew less relevant as the influence of tribal units deteri-
orated. The mechanisms and relationships between these transform-
ations are not immediately obvious and require some explication.⁴²
It is important to observe that the tribal unit was essential to the
development of intertribal norms, inasmuch as its dissolution was
essential to the further development of individual liability. Posner
(1980) discusses the dynamics of intra-tribal interactions in great
detail, identifying important characteristics of early societies that
distinguish them from modern societies. First, life in early societies
was characterized by an extreme scarcity of resources. Second, infor-
mation costs were prohibitive in many situations. And third, govern-
ment was either weak or nonexistent.
These factors entailed important consequences. Due to the scarcity
of resources and high information costs, tribes served an important
and desirable insurance function. By pooling resources, individuals
were better equipped to weather risks. To counteract the lack of
centralized government and high information costs, tribes lived in
close-knit communities. By living in crowded conditions lacking
privacy, where each individual knew what other members of the
tribe were doing (while also being monitored by other members of
the tribe), information costs were effectively reduced, leading to an
informal, decentralized reciprocal surveillance.⁴³
⁴² For a detailed formal discussion of this topic, see Dari-Mattiacci and

Parisi (2004).
⁴³ However, due to the strongly aligned interests of individuals and their tribes, it is
doubtful whether internally committed wrongs were punished. See Faris (1915), pointing
out that the patriarch of a tribe has little incentive to punish members of his own tribe.

It is the dual functions of insurance and self-policing that allowed
intertribal revenge norms to develop. Assigning liability not upon
individual wrongdoers, but instead upon the wrongdoers’ tribes
allowed ancient societies to overcome the detection and solvency
problems that would have attended individual liability. Moreover,
the structure of early societies amplified the deterrence effect of
intertribal retaliation, because tribes were incentivized to exercise
collective “self-control” over their members to avoid the shared
burden of retaliation.
Historically, the constraints of lex talionis tended to be pursued
with an absurd degree of fidelity to the rule. For instance, in the
Native American “law of revenge,” it was the right of the wronged
tribe to inflict retaliation—identical in kind and magnitude—for
harms suffered. If this retaliation were executed by anyone except a
member of the injured tribe, however, that act would not be recog-
nized as retaliatory, but regarded as a fresh wrong requiring its own
retribution.⁴⁴ Biblical and Babylonian implementations of lex talionis
required an even more extreme formal symmetry in the acceptable
retaliatory act: a man who causes the death of another man’s child is
to lose his own (Daube 1947: 169);⁴⁵ the wife of a rapist is to be raped
by the victim’s family (Middle Assyrian Laws, Paragraph 55);⁴⁶ and if
a building collapses, killing the owner’s son, then the builder’s son is
to be put to death (Code of Hammurabi, Paragraph 230).
However, as the environmental constraints motivating communal
liability, identified by Posner (1980), diminished over time, so too did
the effectiveness of intra-tribal punishments in deterring harms. For
example, Dari-Mattiacci and Parisi (2004) formally deduced that as a
tribe’s wealth and population increased, common pool problems (the
tendency of one person to use commonly owned property without
regard for the interests of other owners), moral hazard, and free-rider
problems (when people benefit by the efforts or expenditures under-
taken by others without contributing themselves) tended to increase
as well, diminishing the deterrence effect of communal liability.
⁴⁴ Thus, the law of revenge was intransitive. For example, if Tribe A injures Tribe B,
and Tribe B injures Tribe C, and Tribe C injures Tribe A, then the owed retaliations do
not cancel out. Rather, Tribe B must retaliate against Tribe A Tribe C must retaliate
against Tribe B and Tribe A must retaliate against Tribe C for “justice” to be satisfied.
⁴⁵ Similarly, see Code of Hammurabi, Par. 210 (requiring that the daughter of a
man who strikes a pregnant woman and causes a miscarriage be put to death).
⁴⁶ See Roth (1995: 174–5).
This conclusion is easy to grasp intuitively. While we may suppose
that the utility that an individual gains from harming (or risking
harm to) individuals of another tribe is indifferent to changes in
wealth or population, the decreasing share of communal liability for
each individual as the population increases in size reduces that
individual’s expected costs from committing the wrong.
To illustrate the point by use of an extreme case, contrast the
impact of retaliation against a tribe consisting of a dozen individuals
who all provide essential functions to the community, versus retali-
ation against a wealthy tribe, consisting of many hundreds of indi-
viduals, some of whom are only distantly acquainted with the would-
be injurer. It should be intuitively obvious that the cost of communal
liability in the latter case is substantially less than in the former. Thus,
as a consequence of increased wealth and population, the individuals’
interests will tend to diverge from the social optimum, as incentives
become misaligned.
Unsurprisingly, we do in fact observe communal liability collaps-
ing into individual liability in our historical specimens. For example,
by the sixth century BCE, communal liability was renounced in the
biblical tradition: “The fathers shall not be put to death for the
children, neither shall the children be put to death for the fathers”
(Deuteronomy 24:16).⁴⁷
22.4. THE COMMODIFICATION OF

CIVIL LIABILITY
Up to this point, ancient legal systems treated liability monolithically,

failing to distinguish between intentional and accidental harms. The
division of tort liability and criminal liability begins to emerge with
the commodification of retaliation.
Readers familiar with Coase (1960) will not be surprised, given the
retaliatory right under lex talionis regimes, to learn that the kind-for-
kind constraint eventually gave way to pecuniary compensation. In
Coasean terms, the initial allocation of the retaliatory right was not
⁴⁷ See generally Parisi (1992, 1997).


assigned to the highest-valuing party. Thus, we expect the injurer and
victim would bargain for a transfer of the right.
Several very plausible assumptions are implicit in this claim. First,
we assume that injurers value the retaliatory right more than victims.
Though individuals are indubitably predisposed to seek retribution
for harms suffered (recall the discussion from §22.1), it seems likely
that the desire of injurers to avoid retaliation will tend to be greater
than victims’ desire to exact vengeance. For example, if A harms B,
resulting in the loss of a limb, then by lex talionis, B obtains the right
to dismemberment of A. Yet it seems quite likely that although B’s
desire to retaliate may be great (call the magnitude of his desire x), A’s
desire to keep his body intact will be greater still (call it y½> x). Thus,
there will be a price ρ in the range p ∈½x; y, such that A can simply
pay B not to exercise the retaliatory right, whereupon A enjoys the
surplus y p, and B enjoys the surplus p x, assuming zero (or
sufficiently low) transaction costs,⁴⁸ leading to a Pareto superior
outcome: B gets more than the revenge was worth to him, and A
gets to keep his limb.
In fact, we may observe such bargaining in all lex talionis regimes
throughout history. For example, in times of peace, ancient Athenians
tended to exercise compensation alternatives;⁴⁹ blood revenge was
rarely carried out in practice (McHardy 2008: 9). Interestingly, how-
ever, blood revenge remained common during times of war, when
vengeance was exacted on the battlefield.⁵⁰ We may speculate that this
was due to two factors. First, transaction costs during times of war
may have been prohibitive. Second, since the legal right to kill enemy
combatants during war was redundant with the right to retaliation,
paying off the victim in exchange for one right did not prevent the
victim from exercising the wartime right to kill the injurer anyway. At
any rate, the Athenians seem clearly to have practiced Coasean
bargaining of the right to retaliation.
Likewise, later Talmudic interpretations of lex talionis understood
“an eye for an eye” in compensatory terms as “the value of an eye for
⁴⁸ Given the stakes in the hypothetical, it is plausible that transaction costs will not
exhaust the bargaining space in most situations.
⁴⁹ Demosthenes, Orationes 43.57.
⁵⁰ It is also interesting that Athenian women were less willing to accept compen-
sation for harm, insisting on “specific performance” of the blood revenge remedy.
McHardy (2008: 9). We speculate that this may be due to qualitative differences in the
crimes committed against women in ancient Greece.
an eye.” The rationale evolved from an apparent paradox in the strict
application of the kind-for-kind constraint. The rabbinic scholars
point out that Leviticus 24:22 requires, “Ye shall have one manner of
law,” and yet the requirement, “an eye for an eye,” interpreted literally
would be impracticable in many cases: “What then will you say where a
blind man put out the eye of another man, or where a cripple cut off the
hand of another, or where a lame person broke the leg of another?”
(Talmud, Baba Kamma 84a). Thus it was recognized that “an eye for an
eye” must, on pain of self-contradiction, be interpreted to mean “the
value of an eye for an eye.” Pecuniary compensation was eventually
established as the exclusive form of retaliatory right, rather than a mere
alternative to retaliation, in Jewish law.⁵¹
We observe similar commodification of the lex talionis principle
universally in the development of other ancient legal systems, includ-
ing the Roman law of the Twelve Tables, the Visigothic Code,⁵² the
Salic Code,⁵³ and informally in the Native American law of revenge.⁵⁴
Thanks to the numerous sources that survived throughout the cen-
turies, we can observe a similar—and very well-defined—pattern of
evolution in the history of Roman law. The Roman law of the Twelve
Tables includes a provision that an agreement between the parties can
obviate the literal application of the lex talionis. The option to buy out
the right to talionic revenge through the payment of a sum of money
in cases of iniuria was an important step in this evolution,⁵⁵ which is
often regarded as the genesis of the concept of delictual obligations.⁵⁶
The right to retaliation in kind outlives this early period throughout
classical Roman law. An example is found in the case of manifest theft
(the so called furtum manifestum, when the person is caught in the
act; a theft was manifest if the thief was caught on the day of the theft
with the stolen property before reaching the place where he intended
to take it). This became a delict that created an obligation for the
⁵¹ Pasachoff and Littman (2005).

⁵² Visigothic Code at 7.3.3 (“Should they wish to do so, they may exact from the
kidnapper, the legal compensation for homicide; that is to say, three hundred solidi . . . ”).
See Scott (1910: 248).
⁵³ Henderson (1910: 176–89): “Title LXII. If any one’s father have been killed, the
sons shall have half the compounding money (wergeld); and the other half the nearest
relatives, as well on the mother’s as on the father’s side, shall divide among them-
selves.” See also Hessels and Kern (1880).
⁵⁴ See Reid (1970); Hudson (1976); and Richter (1992).
⁵⁵ Law of the Twelve Tables, Table 8.2.
⁵⁶ Talamanca (1979: 2–4); Scherillo and Gnoli (1994: 127).

wrongdoer to pay a sum of money, which was a multiple of value of
property stolen. The thief could no longer be lawfully killed, but he
was required to pay back the double of the value of the stolen
property (or four times that value, if he was caught in flagrante).
An important turning point in the Roman history of tort was the
Lex Aquilia, a statute from the time of the Republic. This statute for
the first time created the opportunity to receive damages (that is to
receive money as a compensation for damage that one has suffered),
laying the foundations for modern tort liability.⁵⁷ A distinction arises
at this point between more serious crimes, such as murder or high
treason, that were publicly sanctioned,⁵⁸ and other delicts, like theft,
assault, or wrongfully inflicted loss, that were privately prosecuted by
the victim, creating an obligation for the offender to pay the victim a
certain sum of money.⁵⁹
From the compensatory scheme that evolved from the lex talionis
principle, we begin to see the outlines of modern tort law taking
shape. In economic terms, the compensatory modification approxi-
mates efficient injurer incentives.
We now analyze the efficiency of the “blood money” rule that
gradually replaced lex talionis in criminal and tort cases. We will
examine the efficiency property of blood money, distinguishing
between intentional and negligent crimes. A general result is that
bargaining creates an inefficient compensation measure and induces
underdeterrence of criminal activity.
In the case of intentional crimes, bargaining between the offender
and the victim (or victim’s family) will take place only in the case of
inefficient crimes, i.e. when the criminal’s benefit from committing
the crime B is lower than the harm H imposed on the victim. In the
absence of any enforcement error, a kind-for-kind lex talionis regime
will provide efficient deterrence, since it imposes on the criminal a
sanction equal to the harm suffered by the victim, and the offender is
not willing to pay to avoid the sanction. However, with inefficient
crimes the offender is willing to avoid the kind-for-kind sanction
since it would impose on him a greater harm than the benefit gained
from the crime.
⁵⁷ See Lübtow (1971: 7); Frier (1989: 3); Hausmaninger (1990: 3).
⁵⁸ Santalucia (1989: 42–3); Kunkel and Schermaier (2005: 41–2).
⁵⁹ Kaser (1971: 609–10).
Let γ denote the victim’s bargaining power. The offender is
willing to pay up to B þ γðH BÞ in order to avoid the application
of the kind-for-kind sanction required by the lex talionis. This
bargaining cannot take place for efficient crimes, since the criminal
will still have a positive payoff, equal to the net benefit, B–H, after
the kind-for-kind sanction. Bargaining between the offender and the
victim takes place in case of negligent crimes, providing an insuffi-
cient compensatory measure and hence underdeterrence. In both
cases, the criminal is not willing to pay more than γH, escaping the
full kind-for-kind sanction, equal to H. This creates room for under-
deterrence of delinquent behavior, and inefficient incentives to take
optimal precautions.
In the presence of enforcement errors, the underdeterrence gener-
ated by an inefficient compensatory damage measure is reinforced.
Underdeterrence emerges at equilibrium and the incentive to under-
take efficient precautions is diluted in the case of negligent crimes.
While underdeterrence is the only possible equilibrium for negligent
crimes, optimal deterrence of crime could still be achieved in case of
intentional crimes. This is solely the case for inefficient crimes, when
the enforcement error e is lower than the ratio of benefit to harm, i.e.
B
e < BþγðHBÞ . Note that blood money is less frequently observed in the
presence of enforcement error since the range of values that support
efficient deterrence is reduced after the introduction of bargaining.
For efficient intentional crimes, however, we will not observe any
bargaining and the only possible way to achieve higher efficiency in
the deterrence of crime is to adopt more severe sanctions, replacing a
kind-for-kind sanction with higher multipliers α in order to correct
for the enforcement error e. Higher multipliers α should be optimally
implemented for most efficient crimes in order to counteract
underdeterrence.
Several conceptual steps remain in the evolution of tort liability,
separating the compensatory modification of lex talionis from
modern tort law. These include: (1) recognition of the distinction
between intentional and accidental wrongs; (2) elimination of the
bargaining element via the introduction of fixed remedies; (3)
identification of compensation with harm; and (4) recognition of
victim precautions as a fundamental element in the calculus of tort
liability. However, these sophisticated changes occurred in rela-
tively developed legal systems, which fall outside the scope of our
present inquiry.

Yet it bears mentioning the durability of the blood money stage of
legal development. Unlike bare vengeance, discretionary retaliation,
and lex talionis, the mechanisms forcing further evolution from blood
money are subtler and less forceful. In regimes where legal change is
difficult, the relative efficiency of blood money represents a possible
stopping point in the refinement of tort liability rules. Indeed, though
the context of our discussion is ancient law, it is worth pointing out
the persistence of blood money in modern day Sharia in the concept
of Qis.ās..⁶⁰ Likewise, in the customary law of Somalia, blood money
compensation continues to be practiced. And informally, blood
money compensation remains a common extralegal social practice
in both Japan (mimaikin) and Korea (hapuigeum). For a more
detailed analysis of the evolution of tort liability in ancient law, see
Parisi (2001).
22.5. THE CENTRALIZATION OF CRIMINAL

LIABILITY
The literature distinguishing criminal law from tort law is vast.⁶¹ In

the philosophy of punishment, retributivists (and mixed theorists)
identify a putatively moral dimension in criminal conduct as the
distinctive element separating crimes from torts.⁶² Scholars in this
field are interested in the distinction for normative reasons, seeking to
determine what acts the state ought to punish, and what justifies
punishment from an ethical standpoint. However, careless scholars
may try to transform the claim into a descriptive one. If moral
considerations are given as a descriptive distinction between the
criminal law and the law of torts, it is difficult for such claims to
avoid endogeneity problems. Even assuming that crimes necessarily
punish moral wrongs, and that torts do not (already a disputable
point), it is not clear whether such activities are regarded as criminal
because they are immoral, or whether in at least some cases they are
seen as immoral because they are criminalized.
⁶⁰ El-Awa (1981).
⁶¹ See e.g. Morris (1933); Posner (1980); Coffee (1991); Epstein (1997); Weinstein
(2001); and Simons (2008).
⁶² See e.g. Hirsch and Ashworth (2005); Moore (2007); Husak (2008).
These questions raise meta-ethical problems about the existence of
moral facts, which are tangential to our present inquiry.⁶³ We think
that the historical dimension of the crime/tort distinction may be
more usefully seen in economic terms. Posner (1980) takes the strictly
instrumental view that the crime/tort distinction is a function of
wealth, government, and information costs. As we discussed earlier,
lex talionis norms taking the tribe as the bearer of liability developed
due to the high costs of detection, lack of state authority, and the lack
of wealth needed to compensate victims.⁶⁴ Given these constraints,
the legal instruments of modern criminal law and tort law would have
been impracticable.
As wealth increased in ancient societies, the possibility of compen-
satory payments emerged, and lex talionis was modified to allow
compensation as an alternative to reciprocal retaliation. As discussed
in the previous section, this innovation evolved into tort law. How-
ever, state authority tended to become established concomitantly with
increases in wealth, allowing state enforcement and sanctions to
emerge from the undifferentiated liability of lex talionis. This process
led to the development of the criminal law.
The relationships between economic development, population
growth, and the evolution of criminal law are complex. Increasing
wealth and population allowed for the establishment of more efficient
deterrence mechanisms, such as the compensatory scheme discussed
in the previous section. Yet these more efficient rules contributed to
greater predictability, stability, and wealth for those communities,
which in turn allowed for yet more efficient rules. However, the
effects of increasing wealth and population on the law were not all
positive.
As discussed in §22.3, the move from tribal liability to individual
liability was occasioned by increases in population and wealth to
avoid free-rider and common pool problems with respect to inter-
tribal offenses. However, it is not difficult to see how the misalign-
ment of interests resulting from population growth would also lead
to an increased need for intra-tribal mechanisms for deterring
inefficient conduct.
⁶³ The denial of moral facts is central to a large number of skeptical theories.

A particularly influential and clear account of one variant—moral error theory—may
be found in Mackie (1977), further developed in Joyce (2001, 2006).
⁶⁴ Posner (1980: 43).

In addition, although the increasing collective wealth of commu-
nities may have allowed the compensatory modification of lex talionis
to replace retaliatory violence prior to the dissolution of communal
liability, this process would not apply to intra-tribal remedies.
Though mostly tangential to our discussion, it is here worth observ-
ing that during this phase of legal development, individual property
rights would also be emerging, leading to heterogeneity in the per-
sonal wealth of community members. Thus, though relatively wealthy
communities might be capable of mustering pecuniary compensation
for victims of harm caused by its members, there would still be
relatively poor individuals within the communities incapable of mus-
tering compensation privately. Thus, as private incentives diverged
from the social optimum, these individuals would not be deterred by
damages-based deterrence measures.
Thus, as population and wealth increase, we expect that early
societies will tend to develop a system of criminal law, responding
to two pressures: first, the increased need for deterrence in cases of
intra-tribal harms; and second, the move from communal to individ-
ual liability in cases of intertribal harms. In both cases, the presence of
insolvent individuals will undermine the deterrent effect of the com-
pensation mechanism. But recall that the change from retaliatory lex
talionis to compensatory lex talionis was a Pareto improving move.
Compensation-based deterrence, by forcing injurers to internalize the
harms they cause, will tend to create efficient incentives (for wealthy
injurers, at least). Communities should therefore be reluctant to
abandon (or forgo) compensation-based deterrence, simply because
it may be ineffective in a subset of cases. And indeed, ancient civil-
izations did not do this, but instead invented different mechanisms to
effect a different kind of deterrence where compensation-based deter-
rence fails: the criminal law.
Additional factors may have contributed to the motivation for
development of a criminal law. For activities where victims are
unlikely or unable to demand compensation, such as incest or pollu-
tion, or where the harm is suffered by the public rather than a
particular individual, such as “witchcraft” or treason, the compensa-
tion system might not provide adequate (or indeed any) deterrence.
In yet other cases, the harm may be so great that the compensation
remedy is insufficient to force individuals to internalize the harm.
Finally, where injuries to others result in a reduction in tax revenue
(e.g when the injured party is less productive as a result of his injury),
then this represents an additional social harm (above the private
harm to the victim), and in such cases, criminal penalties may force
potential injurers to internalize this additional cost.⁶⁵
Criminal law as an instrument may also provide additional benefits
that civil liability is less capable of achieving. For example, assuming
that moral beliefs are at least partly shaped by the law, criminalization
may perform an important (albeit difficult to quantify) expressive
function—transforming the preferences of community members, and
generating a more efficient mix of norms. The expressive effect of the
criminal law may also generate focal points, resolving coordination
problems for various activities.⁶⁶
To summarize: the principle of lex talionis created a general deter-
rence against undifferentiated “wrongs.” As early societies developed,
more efficient compensatory mechanisms arose, which later developed
into the modern law of torts. However, the deterrent effect of civil
liability, while more efficient than that of lex talionis, is identical neither
in magnitude nor scope. Thus, we may regard the criminal law as
having arisen to fulfill those functions for which compensation-based
deterrence was either insufficient or inapplicable.
22.6. CONCLUSI ON
Historical overviews such as this chapter are not unlike theoretical

models. Models pick out features of their explananda as salient, while
turning a blind eye to others, to lay bare their essential elements and
relations. By forcing reality into a quantifiable framework, models
filter an undifferentiated mass of facts into something meaningful—
something accessible to analysis. Likewise, no telling of history can be
exhaustive; instead, we must choose which aspects of the past to
highlight, and which to disregard. In selecting certain features of the
past, while ignoring others, we begin to construct a story, abstracted
from reality to shine light upon some insight.
In this chapter, we have surveyed the evolution of deterrence in
ancient law. We began with the observation that individuals often
⁶⁵ See Posner (1980: 51–2).

⁶⁶ For an impressively general theory of law as a mechanism for generating focal
points to solve coordination problems, see McAdams (2000).

seem to desire revenge for perceived wrongs committed against them.
We offered an explanation for this apparently innate preference in
terms of evolutionary psychology: that a disposition for retaliatory
behavior acted as a deterrent against other tribes causing harm.
Though investment in revenge-seeking was undoubtedly costly,
when framed in game theoretic terms, we described how it could
represent an efficient equilibrium in repeated-games. We then
described the process through which this preference for revenge-
seeking transformed into a social norm, constrained by kind-for-
kind limitations of discretionary magnitude. Later, an additional
measure-for-measure constraint developed to limit spirals of escalat-
ing blood feuds. Concomitant with these changes, liability shifted
from tribes to individuals. Finally, more specialized tools for effecting
deterrence were developed, and the monolithic liability of lex talionis
branched out on two separate paths: tort liability and criminal
liability.
At each stage, we observed that early societies tended to adopt
norms (later laws) that were efficiency-improving. Indeed, each stage
of development seems to have been characterized by the adoption of
the ever more efficient rules, constrained by severely limiting circum-
stances. As ancient societies grew in wealth and population, and as
state power grew, more efficient rules became available.
Thus, when the broad sweep of legal history is viewed through an
economic lens, we are able to identify the deterrence objective under-
girding many of the major “paradigm shifts” in the law.⁶⁷ From its
antediluvian pre-legal origins through to what would eventually
become the modern law of torts and criminal law, we have seen the
deterrence objective effect an alignment of social and private incen-
tives through a gradual process of accretion. The resilience of the
deterrence objective is evident in the remarkable variety of its dispar-
ate manifestations in human interaction—from base instinctual
response to social norm to legal rule. That the evolution of these
⁶⁷ We are sensitive to possible criticism that applying Kuhn’s (1962) concept of

“paradigm” to a social practice, such as the law, is a misuse of the term. The
mechanisms effecting large-scale transformation in the sciences, identified by Kuhn,
are not directly transplantable to the law. Our use of the term should be understood as
a loose analogy, upon which nothing essential depends. Certainly, we do not intend to
imply the epistemic relativism that many of Kuhn’s followers (though not Kuhn
himself ) endorsed.
apparently distinct realms should share a common cause, to solve and
refine a common problem, is a testament to the robustness of the
economic forces driving efficiency.
REFERENCES
Allen, Danielle S. 2004. “Angry bees, wasps, and jurors: the symbolic politics
of ὀργή in Athens,” in Susanna Braund and Glenn W. Most, eds, Ancient
Anger: Perspectives from Homer to Galen. Yale Classical Studies (no. 32).
Cambridge: Cambridge University Press, 76–98.
Blackburn, Bob L. 1980. “From blood revenge to the Lighthorsemen: evolu-
tion of law enforcement institutions among the five civilized tribes to
1861.” 8.1 American Indian Law Review 49–63.
Blau, Joel. 1916. “Lex talionis,” in Isaac Rabbi, ed., 26 Yearbook of the Central
Conference of American Rabbis. Philadelphia: CCAR, 336–66.
Cairns, John W., and Robinson, Olivia F. 2001. Critical Studies in Ancient
Law, Comparative Law and Legal History. Oxford: Hart Publishing.
Camerer, Colin F., and Richard H. Thaler. 1995. “Anomalies: Ultimatums,
Dictators and Manners.” 9.2 Journal of Economic Perspectives 209–19.
Carlsmith, Kevin M., Timothy D. Wilson, and Daniel T. Gilbert. 2008. “The
paradoxical consequences of revenge.” 95 Journal of Personality and Social
Psychology 1316–24.
Chagnon, Napoleon. 1988. “Life histories, blood revenge and warfare in a
tribal population.” 239 Science 985–92.
Coase, Ronald H. 1960. “The problem of social cost.” 3 Journal of Law and
Economics 1–44.
Coffee, John C., Jr. 1991. “Does ‘unlawful’ mean ‘criminal?’: reflections on
the disappearing tort/crime distinction in American law.” 71 Boston
University Law Review 201–52.
Czabanski, Jacek. 2008. Estimates of Cost of Crime: History, Methodologies,
and Implications. Warszawa: Springer.
Dari-Mattiacci, Giuseppe, and Francesco Parisi. 2004. “The rise and fall of
communal liability in ancient law.” 24 International Review of Law and
Economics 489–506.
Das Gupta, Ramaprasad. 1930. Crime and Punishment in Ancient India. New
Delhi: Bharitiya Publishing House.
Daube, David. 1947. Studies in Biblical Law. Cambridge: Cambridge Univer-
sity Press.
Demosthenes. 2002. Orationes. Oxford: Oxford University Press.
Doongaji, Damayanti. 1986. Crime and Punishment in Ancient Hindu Society.
Delhi: Ajanta Publications, Distributors, Ajanta Books International.

Duff, Antony. 1986. Trials and Punishments. Cambridge: Cambridge
University Press.
Falk, Armin, and Urs Fischbacher. 2006. “A theory of reciprocity.” 54.2
Games and Economic Behavior 293–315.
Frier, Bruce W. 1989. A Casebook on the Roman Law of Delict. Atlanta:
Scholars Press.
El-Awa, Mohamed. 1981. Punishment in Islamic Law. Indianapolis: Ameri-
can Trust Publications.
Epstein, Richard A. 1997. “The modern uses of ancient law.” 48 South
Carolina Law Review 243–65.
Fagan, Jeffrey, and Tracey L. Meares. 2008. “Punishment, deterrence and
social control: the paradox of punishment in minority communities.” 6
Ohio State Journal of Criminal Law 173–229.
Falk, Armin, Ernst Fehr, and Urs Fischbacher. 2008. “Testing theories of
fairness—intentions matter.” 62.1 Games and Economic Behavior 287–303.
Faris, Ellsworth. 1915. “The origin of punishment,” in Albert Kocourek and
John H. Wigmore, eds, Primitive and Ancient Legal Institutions. Boston:
Little, Brown & Co. 151–61.
Farrington, David P., Patrick A. Langanand, and Per-Olof H. Wikström.
1994. “Changes in crime and punishment in America, England and
Sweden between the 1980s and the 1990s.” 3 Studies on Crime and
Crime Prevention 104–31.
Gendreau, Paul, Claire Goggin, and Francis T. Cullen. 1999. The Effects of
Prison Sentences on Recidivism. A Report to the Corrections Research and
Development and Aboriginal Policy Branch. Solicitor General of Canada.
Ottawa: Public Works & Government Services Canada.
Ginat, Joseph. 1997. Blood Revenge: Family Honor, Mediation and Outcast-
ing. Brighton: Sussex Academic Press.
Good, Edwin M. 1967. “Capital punishment and its alternatives in ancient
Near Eastern law.” 19 Stanford Law Review 947–77.
Güth, Werner, Rolf Schmittberger, and Bernd Schwarze. 1982. “An experi-
mental analysis of ultimatum bargaining.” 3.4 Journal of Economic Behav-
ior and Organization 367–88.
Hausmaninger, Herbert. 1990. Der Schadensersatz der lex Aquilia. Vienna:
Manz.
Henderson, Ernest F. 1910. Select Historical Documents of the Middle Ages.
London: George Bell and Sons.
Hessels, Jan H. 1880. Lex Salica: The Ten Texts with the Glosses and the Lex
Emendata. London: John Murray.
Hirsch, Andrew von. 1993. Censure and Sanctions. Oxford: Clarendon Press.
Hirsch, Andrew von, Anthony E. Bottoms, Elizabeth Burney, and Per-Olof
H. Wikström. 1999. Criminal Deterrence and Sentence Severity: An
Analysis of Recent Research. Oxford: Hart Publishing.
Hirsch, Andrew von, and Andrew Ashworth. 2005. Proportionate Sentencing.
Oxford: Oxford University Press.
Hudson, Charles M. 1976. The Southeastern Indians. Knoxville: University of
Tennessee Press.
Humbert, Michel, ed. 2005. Le Dodici Tavole—dai Decemviri agli Umanisti.
Pavia: IUSS PRESS.
Husak, Douglas. 2008. Overcriminalization: The Limits of the Criminal Law.
Oxford: Oxford University Press.
Jacoby, Susan. 1983. Wild Justice. New York: Harper & Row.
Jaishankar, Karuppannan, and Debarati Haldar. 2004. “Manusmriti: a cri-
tique of the criminal justice tenets in the ancient Indian Hindu code.” 1.3
ERCES Online Quarterly Review 1–15.
Jhering, Rudolf von. 1906. Geist des römischen Rechts, I./5th–6th edn. Leipzig:
Breitkopf und Härtel.
Joyce, Richard. 2001. The Myth of Morality. Cambridge: Cambridge Univer-
sity Press.
Joyce, Richard. 2006. The Evolution of Morality. Boston: MIT Press.
Kant, Immanuel. 1965 [1787]. Critique of Pure Reason, trans. Norman Kemp
Smith. New York: St. Martin’s Press.
Kaser, Max. 1971. Das römische Privatrecht. Erster Abschnitt: Das altrö-
mische, das vorklassische und klassische Recht. Munich: C. H. Beck.
Kavka, Gregory S. 1978. “Some paradoxes of deterrence.” 75.6 Journal of
Philosophy 285–302.
Kuhn, Thomas S. 1962. The Structure of Scientific Revolutions, 1st edn.
Chicago: University of Chicago Press.
Kunkel, Wolfgang, and Martin Schermaier. 2005. Römische Rechtsgeschichte.
Cologne/Weimar/Vienna: Böhlau.
Langan, Patrick A., and David P. Farrington. 1998. Crime and Justice in the
United States and in England and Wales, 1981–96. Washington, DC:
Bureau of Justice Statistics.
Lewis, Clive S. 1953. “The humanitarian theory of punishment.” 6 Res
Judicatae 224–30.
Lübtow, Ulrich von. 1971. Untersuchungen zur lex Aquilia de damno iniuria
dato. Berlin: Duncker & Humblot.
Mackie, John L. 1977. Ethics: Inventing Right and Wrong. New York: Penguin
Books (repr. in Penguin Books in 1990).
Maine, Henry J. S. 1861. Ancient Law: Its Connection with the Early History
of Society, and its Relation to Modern Ideas. London: John Murray.
Manthe, Ulrich. 2000. Geschichte des römischen Rechts. Munich: C. H. Beck.
McAdams, Richard H. 2000. “A focal point theory of expressive law.” 86
Virginia Law Review 1649–1729.
McHardy, Fiona. 2008. Revenge in Athenian Culture. London: Duckworth.

Miller, William. 2007. An Eye for an Eye. Cambridge: University Press.
Mills, Douglas E. 1976. “Kataki-uchi: the practice of blood-revenge in
pre-modern Japan.” 10.4 Modern Asian Studies 525–42.
Moore, Michael. 2007. Placing Blame. Oxford: Clarendon Press.
Morris, Raphael C. 1933. Law and the Social Order. New York: Harcourt
Brace & Company.
Olivelle, Patrick. 2011. “Penance and punishment: marking the body in crim-
inal law and social ideology of ancient Indi.” 4 Journal of Hindu Studies 1–19.
Parens, Erik. 2014. “Afterword: living with the ancient puzzle.” 45.2 Hastings
Center Report 50–2.
Parisi, Francesco. 1992. Liability for Negligence and Judicial Discretion, 2nd
edn. Berkeley: University of California Press.
Parisi, Francesco. 1997. “Liability without Fault and Communal Responsi-
bility.” 97–08 Ancient Legal Systems. GMU Law and Economics Working
Paper Series.
Parisi, Francesco. 2001. “The genesis of liability in ancient law.” 3 American
Law and Economics Review 82–124.
Parisi, Francesco, and Vincy Fon. 2005. “The behavioral foundations of
retaliatory justice.” 7 Journal of Bioeconomics 45–72.
Pasachoff, Naomi E., and Robert J. Littman. 2005. A Concise History of the
Jewish People. Lanham: Rowman and Littlefield Publisher Inc.
Posner, Richard A. 1980. “A theory of primitive society, with special refer-
ence to law.” 23.1 Journal of Law and Economics 1–53.
Posner, Richard A. 1988. Law and Literature: A Misunderstood Relation.
Reid, John P. 1970. A Law of Blood: The Primitive Law of The Cherokee
Nation. New York: New York University Press.
Richter, Daniel. K. 1992. The Ordeal of the Longhouse: The Peoples of the
Iroquois League in the Era of European Colonization. Chapel Hill: Uni-
versity of North Carolina Press.
Roth, Martha T., trans. 1995. Law Collections from Mesopotamia and Asia
Minor. Atlanta: Scholars Press.
Santalucia, Bernardo. 1989. Diritto e processo penale nell’antica Roma. Milan:
Giuffrè Editore.
Scherillo, Gaetano, and Franco Gnoli. 1994. Lezioni sulle obbligazioni: corso
di diritto romano. Milan: Monduzzi Editore.
Scott, Samuel P., trans. and ed. 1910. The Visigothic Code. Boston: Boston
Book Company.
Shalom, Paul. 1970. Studies in the Book of the Covenant in the Light of
Cuneiform and Biblical Law. Leiden: Brill.
Simons, Kenneth W. 2007–8. The crime/tort distinction: legal doctrine and
normative perspectives. 17 Widener L. J. 719–32.
Stein, Peter. 1999. Roman Law in European History. Cambridge: Cambridge
University Press.
Sulzberger, Mayer. 1915. The Ancient Hebrew Law of Homicide. Philadel-
phia: Julius H. Greenstone.
Talamanca, Mario. 1979. “Obbligazioni (dir. rom.),” in Enciclopedia del
diritto, vol. XXIX. Milan: Giuffrè Editore, 2–78.
Talamanca, Mario. 1990. Istituzioni di diritto romano. Milan: Giuffrè
Editore.
Tonry, Michael. 2008. Crime and Justice: An Annual Review of Research.
Chicago: University of Chicago Press.
Webster, Cheryl M., and Anthony N. Doob. 2012. “Searching for Sasquatch:
deterrence of crime through sentence severity,” in Joan Petersilia and
Kevin Reitz, eds, Oxford Handbook on Sentencing and Corrections. New
York: Oxford University Press, 173–95.
Weinstein, Mark I. 2001. “Limited liability in California: 1928–1932.” Paper
No. 00–17 USC Olin Research.
23
Collective Responsibility
Thomas J. Miceli
Collective guilt, whatever form or justification it takes, has to be

one of the most evil notions the human race has concocted,
most likely the cause of more suffering of innocents than any
other vile belief in history.
(Simic 2013: 22)
The principle of collective responsibility—so abhorrent to mod-
ern sensibilities—may be efficient in the conditions of primitive
society.
(Posner 1983: 194)
23.1. INTRODUCTION
Economic models of law enforcement beginning with Becker (1968)

have focused on the goal of identifying and apprehending the per-
petrator of a crime based on a principle of “individual responsibility.”
Throughout much of history, however, society operated under a
much different system that attributed responsibility for wrongdoing
to an entire community or group of which a wrongdoer was known to
be a member. Indeed, examples of “collective responsibility,” or group
punishment, are pervasive in ancient societies as well as biblical
stories and mythologies. And surprisingly, the practice persists in
various forms in modern society, such as when an entire class is
punished for the actions of a prankster, a corporation is held liable
Thomas J. Miceli, Collective Responsibility In: Roman Law and Economics: Exchange, Ownership, and Disputes,
Volume II. Edited by: Giuseppe Dari-Mattiacci and Dennis P. Kehoe, Oxford University Press (2020).
© Oxford University Press.
DOI: 10.1093/oso/9780198787211.003.0023
380 Thomas J. Miceli

for the actions of its employees, or a country is sanctioned for terrorist
acts by one of its citizens.
This chapter provides an economic analysis of the choice between
individual and group punishment and derives the conditions under
which the latter is an optimal strategy.¹ According to the theory, the
chief gains from group punishment are the certainty of punishing the
actual offender plus the saved detection cost, while the primary
drawback is aversion to wrongfully punishing innocent members of
the target group. The use of group punishment has therefore varied
over time and circumstance based on differing perspectives on the
attribution of guilt and the perceived cost of wrongful punishment.
Several important examples of collective responsibility are found in
Roman law, as described in §23.2. §23.3 offers other examples from
the ancient world, including the Bible, and §23.4 provides some
modern uses of the practice. §23.5 then develops the theoretical
analysis and derives the principal conclusions, the formal details of
which are provided in the Appendix. Finally, §23.6 concludes.
23.2. EXAMPLES FROM ROMAN LAW
Probably the best example of collective responsibility in Roman law is

the Senatus Consultum Silanianum, a resolution passed by the Roman
senate around 10 CE concerning the treatment of the slaves of a
murdered master. The resolution provided that any slaves under the
roof of the master at the time of his murder, or accompanying him if
he was murdered elsewhere, were implicated in his death, with the
punishment being torture, or, if it was determined that the slaves had
not done enough to prevent the murder, death. However, a slave who
discovered his master’s murderer was given his freedom. The early
third-century jurist Ulpian explains that masters could not otherwise
be safe unless slaves were compelled to provide protection at the risk
of their own lives against threats from within the household and from
¹ Previous economic models of individual versus group punishment are found in

Heckathorn (1990); Parisi and Dari-Mattiacci (2004); and Miceli and Segerson (2007).
Informal treatments of this question from a legal and economic perspective include
Posner (1983); Levmore (1995a, 1995b); and Levinson (2003). For a philosophical
perspective on collective responsibility, see the various readings in May and Hoffman
(1991).
Collective Responsibility 381

without (D. 29.5.1 pr.; Ulp. 50 ad ed.). Likewise, Modestinus specif-
ically imposed on slaves the duty of protecting a master with
weapons, their hands, shouting, and throwing their own bodies in
the way; if they failed in this duty, capital punishment would rightly
be exacted from them (D. 29.5.15; 8 pandect.). In addition, the law
provided that the will of a slain person could not be opened until the
murder had been investigated; heirs who did not avenge (through a
criminal case) a murder would lose their inheritance, which would be
confiscated by the imperial treasury. Much of the juristic discussion
as it pertains to slaves concerns defining when they were in a position
to come to the aid of their master, what type of aid certain slaves, such
as women, children, or the infirm, would be expected to provide, and
under what circumstances they might be exempt from questioning
under torture and punishment.
Punishments for slaves who failed to protect a master apparently
were first established in a law of Sulla under the Republic. The senate
in 57 CE revised the SC Silanianum by subjecting slaves who had been
freed in the master’s will to its sanctions (Tac., Ann. 13.32). The
sanctions of the SC Silanianum applied no matter who killed the
slaves’ master, but the notorious killing by his own slave of the city
prefect Pedanius Secundus in 61 CE touched off a riot when the senate
debated the execution of the numerous slaves in his household
(Tac., Ann. 14.42–5). Tacitus depicts a forceful speech by the jurist C.
Cassius Longinus, who claimed that the four-hundred-odd slaves in
his household had failed to protect Pedanius, and that without the
implementation of the death penalty, no senator would be safe.
His view prevailed in the senate, although there was significant
opposition.² Pliny the Younger likewise reports about a vigorous
debate in the senate on the fate of the freedmen of the senator
Afranius Dexter, who had died under suspicious circumstances, pos-
sibly by his own hand (Ep. 8.14). The senate was divided on whether
the freedmen should be killed, relegated (a form of exile), or freed;
although Pliny does not explicitly state this, the penalty was likely to
have been relegation.³
² For the details of the laws and the ancient sources, including D. 29.5 and C. 6.35,
see Kaser (1971: 283 and n. 3, 293 and n. 3), as well as Sherwin-White (1985: 463). For
discussion, see Gardner (2011: 430–1); Harries (2013).
³ Harries (2013: 66).

Another form of collective responsibility in Roman law is related to
delicts, which represented the principal form of wrongfully imposed
harm. The delict is most similar to a tort in modern law in the sense
that enforcement was by a private action, and the usual remedy was
payment of damages by the responsible party.⁴ The tort analogy is not
exact, however, as in some cases the assessment of damages had more
of the character of a criminal fine (or punitive damages) than of mere
indemnification for the victim’s losses (du Plessis 2015: 325). Indeed,
Buckland et al. (1952: 344–5) note that the “[d]elict is imbued with
the idea of vengeance,” prompting the authors to argue that “[t]he
Roman law of delict has far more affinity to the criminal law than to
the law of tort . . .”; du Plessis (2015: 325) makes a similar argument.
Roman law did include a distinct concept of crime (crimen), which
gave rise to a criminal (public) rather than a civil action, but to
modern eyes “the conceptual difference between some delicts and
public offenses would have been hard to establish” (Harries 2007: 58),
with the two concepts apparently lying on a continuum that
depended on the seriousness of the act (Watson 1991: 212). Many
offenses that are nowadays considered crimes would have been clas-
sified as delicts in Roman law, as is apparent in the cases of theft
(furtum), robbery and violence (rapina), as well as in some instances
of damage to property (damnum iniuria datum), and personal injury
or insulting behavior (iniuria).⁵ In early Roman law, criminal offenses
were generally limited to truly public harms like treason and sacral
wrongs. Only later were murder and other violent acts included as
crimes, but even in genuine criminal cases, private prosecution
remained the rule (Zimmermann 1996: 918). The fuzzy line between
crimes and torts in Roman law is a common feature of ancient legal
systems (du Plessis 2015: 325).
The actual perpetrator of harm was responsible under the delict,
but there are elements of collective responsibility in a related category
of obligation in Roman law, namely the so-called quasi-delict. The
term quasi-delict refers to remedies that were created to address
situations in which harm arose “as if from a delict,” but the remedies
available in the actions on delicts did not apply because of a lack of
direct fault on the part of the responsible party (Watson 1991: 75;
Cursi 2002). Examples included liability of owners of dwellings from
⁴ See Friedman in this volume.

⁵ See, generally, Watson (1991); du Plessis (2015), ch. 10.

which something was thrown or poured so as to cause damage to
passersby; liability of ship’s captains, innkeepers, or stable owners
whose employees caused harm or committed fraud; and liability of
judges who, through breach of their official duty, caused damage to
others.⁶ In all of these cases, the person being held liable need not
have committed or authorized the harmful act himself, or even have
known about it, to be found responsible for the damages. One
interpretation of quasi-delicts is that they represent a form of strict
liability on the part of individuals who are “in control of a potential
source of danger to other peoples’ lives, health and property”
(Zimmermann 1996: 17). In this sense, quasi-delicts share elements
with the modern concept of vicarious liability (du Plessis 2015: 359).
The related concept of noxal liability provided a way of establish-
ing liability for delicts committed by children in the power of a
paterfamilias (head of the household) or his slaves. Originally, the
paterfamilias could discharge any liability for such wrongs by sur-
rendering his dependent (noxal surrender), but in classical law the
paterfamilias could avoid this step by compensating the injured
party. The surrender option thus effectively limited the amount of
the paterfamilias’s liability, though if the paterfamilias himself had
been involved in the wrong, he usually forfeited his right of surren-
der.⁷ Parisi (2001: 112–13) argues that noxal surrender represented a
kind of transitional phase from a retaliatory, or talionic, law enforce-
ment regime (“an eye for an eye”) to one based on compensation.⁸
Generally, all cases of quasi-delict and noxal liability reflect a
separation of responsibility and fault, which is a distinguishing fea-
ture of group punishment and vicarious liability. As will be discussed
in detail below, there are often good reasons for this, including the
“practical inability to get evidence of fault,” and “the need to get a
better defendant, for the workman in charge of dangerous things is
not likely to be able to compensate for the damage he does”
(Buckland et al. 1952: 397).
A final example of group responsibility in Roman times is found in
the practice of decimation, which involved the random execution of
one in ten soldiers as a means of maintaining military discipline.⁹ The
⁶ For example, a judge “become[s] (emotionally) so entangled in the case that he

lacks the necessary impartiality . . .” (Zimmermann 1996: 16 n. 87).
⁷ Watson (1991: 75); Zimmermann (1996: 916–17). See also Kaser (1971: 631–2).
⁸ Also see the chapter by Parisi et al. in this volume. ⁹ Lintott (1999: 41–2).

Italian army employed a similar procedure in World War I to punish
companies that broke and ran during combat, a practice described by
Ernest Hemingway in his autobiographical novel about World War I,
A Farewell to Arms. The example of decimation differs from vicarious
liability, however, in that there is a sense in which all members of the
targeted group are equally responsible, but punishing each one of
them would obviously be impractical in a military context. As will be
seen below, however, random individual punishment is theoretically
equivalent to a scaled-down version of group punishment, and there-
fore has many of the same attributes (both positive and negative).
23.3. OTHER ANCIENT EXAMPLES
As several scholars have pointed out, a defining feature of ancient law

enforcement was its reliance on collective or group responsibility
(Posner 1983:193–5; Parisi and Dari-Mattiacci 2004). If a person
committed a harmful act, the victim’s kinship group often retaliated
against the injurer or his kinship group. Even states as civilized as
ancient Greece believed that a murderer “polluted” his city and
family, thereby creating a form of guilt that fell upon all citizens of
the city as well as descendants of the family (Posner 1983: 217). Of
course, the most famous example of this is the pollution of Thebes by
Oedipus’ murder of his father, as depicted in Sophocles’ play Oedipus
Rex. Two more modern examples of the same concept are found in
Samuel Taylor Coleridge’s poem, “The Rime of the Ancient Mariner,”
which, according to Kitson (1989), symbolizes collective guilt for the
unrealized promise of the French Revolution; and the work of
Nathaniel Hawthorne, which reflects “the theme that the sins of the
father are visited upon the children” (Buel 1986: 360). (Hawthorne, of
course, believed that he carried the guilt of his forebear, who was a
judge at the Salem witch trials.)
The Hammurabi Code contained a provision that allowed a victim
of robbery, if unable to identify the guilty party, to seek compensation
from the city in which the robbery occurred (Levmore 1995a: 117).
Likewise, the kinsmen of a murder victim could seek compensation
from the city in which the murder took place. Old English law
contained a similar provision, as chronicled by Blackstone (1766,
Book 3: 161) in his Commentaries, that obliged residents of a village

in which a man was robbed to “make hue and cry after the felon,” and
if they failed to identify him, the residents were collectively liable for
“damages equivalent to his loss.”¹⁰
The Bible includes numerous allegories centering on group pun-
ishment. The stories of Noah’s flood (Genesis 7) and Sodom and
Gomorrah (Genesis 18–19) dramatically portray the idea that the sins
of a few cast guilt upon an entire community for which all citizens,
the innocent and guilty alike, must pay. (Although, as Levmore 1995a:
95 points out, if all citizens were truly guilty, then these are not really
examples of collective responsibility.) The Old Testament also
includes an example of random punishment in the story of Jonah,
who boards a ship to escape God’s command to go to the wicked city
of Nineveh to serve as a missionary. As punishment for Jonah’s
defiance, God creates a storm that threatens to destroy the ship and
its crew. The mariners then draw lots to “detect” the person whose
transgression was responsible for the storm, and when Jonah is
miraculously identified as the guilty party, he is thrown overboard,
where he is swallowed by a whale (Jonah 1:1–17). The fact that Jonah
was correctly identified as the wrongdoer by the lottery is clearly
meant to convey to readers the idea that the truly guilty can never
escape the ever-present surveillance of God, which, in a world where
detection and punishment of criminals is difficult, provides an
important source of social control.¹¹
The pervasiveness of group punishment in ancient society reflects
two characteristics of law enforcement in that era: the idea of
punishment as revenge (Holmes 1881 [1963]: 6; Posner 1983: ch.
8),¹² and the absence of an effective centralized enforcement mech-
anism. The former requirement meant that someone had to pay in
order to satisfy and/or compensate victims and their sympathizers
¹⁰ Also see Feinberg (1991: 65–6).

¹¹ On the role of religion in monitoring behavior, see Wade (2009); Johnson
(2016). A modern version of this same idea is found in Adam Smith’s construct of
an “impartial spectator,” which he argued was the basis for moral behavior by
individuals (Smith 1759 [1976]).
¹² As Parisi and Dari-Mattiacci (2004: 497) note, “Several ancient rules of talionic
justice require that the penalty be imposed with uncompromising symmetry, repli-
cating the harm suffered by the victim and his or her family.” For discussion of the
circumstances in which collective sanctions could play a positive role in an ancient
polity, see Lanni (2017).

(a retributive motive), while the latter reflected the impracticality of
identifying and exacting punishment or compensation from the
actual offender.
23.4. EXAMPLES FROM MODERN LAW
Although collective responsibility is commonly associated with

ancient law enforcement, vestiges of it remain in contemporary
doctrines and practices.¹³ The principal objection to collective
responsibility is the idea that innocent parties are punished. Indeed,
Parisi and Dari-Mattiacci (2004: 502–3) link the evolution from
group to individual responsibility in the Old Testament to the emer-
gence of the idea that “the collective guilt referred to in the earlier
sources was no longer the basis of the communal responsibility for
wrongs. An individual’s sin did not place any blame on the clan or
class to which the wrongdoer belonged.” Modern examples of group
responsibility thus tend to involve situations in which “innocent”
members of the target group are nevertheless seen as being at least
partially culpable for the offender’s action. In other words, the cost of
punishing them is not seen as being high. In that case, the savings in
detection costs under group punishment make it an attractive
strategy.
A timely example of this view in modern times is the imposition of
sanctions, or the waging of warfare, against nations known to harbor
terrorists or other wrongdoers. Historical examples of this same
strategy include the attack on Troy by the Greeks in retaliation for
the kidnapping of Helen (as chronicled in the Iliad);¹⁴ the policy of
total destruction wreaked by Union general William T. Sherman in
his infamous “march to the sea” during the American Civil War; the
declaration of war against Germany and Japan by the United States
following the bombing of Pearl Harbor (which included the bombing
of cities); and the resulting internment of Japanese Americans during
that war. As Levmore (1995a: 99) notes, “We are accustomed to the
fact that war claims the lives of innocent people,” and “one nation
¹³ See, for example, the surveys in Levmore (1995a, 1995b); and Levinson (2003).
¹⁴ Archaeologists have concluded that the mythological story of the attack of Troy
was likely based on an actual historical event (Wood 1998).

may be justified in warring against another when the other nation is
unable to control the actions of some of its inhabitants.” In these
cases, citizens of the target country are viewed as somehow collabor-
ating (or at least sympathizing) with the wrongdoer and are therefore
partially culpable.
Most modern examples of group responsibility, however, are found
in the context of tort law, and can be justified in various ways. One is
as a form of insurance for liability, which spreads the cost of acci-
dental harms across an injurer’s risk group (Levinson 2003: 371). In
this case, individuals pay premiums in proportion to their expected
contribution to the risk of harm, and, in this sense, are “punished”
appropriately in relation to their individual “guilt.” Thus, there is no
perceived cost of erroneous punishment, and, as long as premiums
are calibrated to individual risk (i.e. are actuarially fair), there is no
loss of deterrence. However, if such calibration is costly or impossible,
deterrence may be sacrificed owing to the usual moral hazard prob-
lem (though collective governance within the group, or by the insurer,
can help to restore incentives) (Levinson 2003: 372).
Group monitoring of individual members is undoubtedly the prin-
cipal justification for other recognizable forms of group punishment
in modern law. May and Hoffman (1991: 1) call this the “conspiracy
model” of collective responsibility. Examples include various forms of
vicarious liability like the doctrine of respondeat superior, under
which an employer can be held responsible for torts committed by
his employees; the related criminal law rule that a corporation can be
held criminally liable for crimes committed by its employees;¹⁵ and
joint and several liability, under which any one of a group of injurers
can be held responsible for the victim’s entire loss.¹⁶ These rules help
to promote deterrence in those circumstances where offenders are out
of reach of victims or law enforcers, but are under the surveillance of
other members of the group, who, it is hoped, will either impose their
own form of discipline or turn the offender over to plaintiffs (Varian
1990). The Roman example of the Senatus Consultum Silanianum is a
good example of this, as those accompanying a murder victim are
¹⁵ As Garoupa (2000: 244) notes, “Within the context of corporate liability,

shareholders become quasi-enforcers.”
¹⁶ Menell (1991: 109) notes that under joint and several liability for environmental
harms, “disposal of a thimbleful of hazardous waste at a large disposal site exposes an
entity to enormous potential liability.”

likely in a good position to identify, and perhaps thwart, the offender.
In addition, as noted above, vicarious liability rules allow victims (or
prosecutors) to pursue “better defendants” (i.e. deeper pockets) in the
sense of having sufficient resources to fully compensate victims.
In discussing the information revelation function of group punish-
ment, Levmore (1995b) describes the well-known case of Ybarra vs
Spangard,¹⁷ in which a court allowed a plaintiff to jointly collect from
several health care professionals for harm suffered during an oper-
ation. This case reflects a general class of torts in which courts
apportion damages in the presence of uncertainty over causation
(Shavell 1985), but it is unique in the fact that the members of the
group held liable presumably knew which of them actually caused the
injury. The court’s reasoning in holding them all liable was that the
absence of proof that any one person’s negligence caused the acci-
dent, which ordinarily would have resulted in a finding of no liability,
should not be grounds for denying any compensation to the victim,
especially given that the absence of evidence was apparently due to
the unwillingness of the parties to testify against one another. The
threat of group punishment in such a setting may therefore be an
effective way to induce one or more of the members to reveal the true
offender’s identity. At the same time, failure of any to do so, thus
triggering group responsibility, would not be seen as imposing
wrongful punishment because, as in the terrorism example, they are
all then seen as “harboring the guilty party,” which is itself a form
of guilt.
Levmore (1995b) further speculates on why the court did not
impose liability on each member of the group in excess of the harm
suffered by the victim so that even the guilty party would have had an
incentive to confess. For example, if the total harm was $500, then
threatening to impose damages of, say, $600 on each person in the
operating room would presumably induce the responsible party to
confess so that he or she would only have to pay $500. In other words,
the guilty party would have saved $100 by confessing. Such an “over-
extraction” rule would, in addition, give other members of the group
an enhanced incentive to turn over the offender. The obvious disad-
vantage of such a strategy is that it potentially imposes high error
costs if the scheme fails to reveal the offender. Further, as Levmore
¹⁷ 154 P.2d 687 (Cal. 1944).


(1995b: 1595) observes, the offender may anticipate that society
would not tolerate the actual imposition of such disproportionate
penalties on innocent offenders, making the scheme non-credible.
Indeed, it is this intolerance of wrongful or disproportionate punish-
ments that makes group punishment objectionable in all but a few
settings nowadays, as described above.
23.5. THEORETICAL ANALYSIS
This section describes the results of a simple theoretical model

devoted to answering the question of when group punishment is
preferred, on economic grounds, to individual punishment. (The
formal derivations of most of the results are contained in the Appen-
dix.) The analysis is based on the following scenario. A harmful act
has been committed by a single individual who is known to be a
member of a group of size n. In other words, n is defined to be the
smallest group of which the offender is known to be a member with
certainty. The enforcement authority—whether the victim, the vic-
tim’s kinship group, or the state—considers two punishment strat-
egies: (1) seek out and punish a single member from the group
(individual punishment), or (2) impose a uniform sanction on the
whole group (group punishment). The enforcer’s objective in choos-
ing between these two strategies is taken to be maximization of the
net benefits from punishment, which consist of the gains from cor-
rectly punishing the true offender, less the costs of punishment
(including the cost of wrongful punishment, if any) and detection.
The gains from punishment may come from various sources,
including optimal deterrence, retribution, and/or compensation of
victims.¹⁸ Whereas optimal deterrence is the primary motivation in
economic models of crime (Becker 1968; Polinsky and Shavell 2000),
ancient societies seem also to have valued proportionality of punish-
ments to crimes (based on the talionic dictum of “an eye for an eye”),
which originally had its origins in notions of vengeance (Posner 1983;
Parisi 2001). But even in modern times, there seems to be a strong
inclination toward proportional penalties, based primarily on notions
¹⁸ Parisi et al. (in this volume) specifically discusses the goal of deterrence in
ancient law.

of fairness, regardless of whether or not they further the goal of
deterrence.¹⁹ The basic conclusions of the following analysis will
not, however, depend on the specific motivation for punishment.²⁰
In comparing the two punishment strategies, the primary trade-off
is between certainty of punishing the true offender and the costs of
erroneously punishing the innocent. Thus, modern examples of
group punishment seem to be found in contexts where wrongful
punishment is not a large concern, either because the sanction is
not really deemed to be punishment (as in the case of the insurance
examples), or because all members of the target group are seen as at
least somewhat culpable for the underlying offense (as in the cases of
military actions).
To the contrary, in cases of true criminal punishment, especially
when imprisonment is involved, the predominant reliance in modern
society on individual punishment reflects an overriding concern for
avoiding wrongful punishment. Indeed, the significant safeguards of
criminal defendants’ rights in the Anglo-American and other modern
criminal justice systems are a manifestation of this social value. The
other factor that has contributed to (or, more properly, allowed) the
rise of individual criminal punishment, of course, is the emergence of
centralized enforcement authorities (principally governments) and
the development of technologies (like fingerprinting and DNA) that
allow the (reasonably) accurate detection of offenders.
Prior to the emergence of a “science of detection,” individual
punishment, if employed, often amounted to little more than random
punishment, as in the Roman practice of decimation (which, as
noted, continued to be used by the Italian army into the twentieth
century). As shown in the Appendix, however, if there are constant or
increasing returns to scale in punishment costs (including error
costs), then random individual punishment—that is, punishing one
¹⁹ For example, one of the key prescriptions of Becker’s model, which is based
solely on deterrence, is that punishments should be raised to the maximum extent
possible, and apprehension probabilities scaled back correspondingly, so as to save on
overall enforcement costs. Actual punishment schemes, however, rarely if ever imple-
ment such a policy. On the inclusion of fairness in Becker’s model, see Miceli (1991).
²⁰ See Miceli and Segerson (2007) for a formal analysis that explicitly distinguishes
between deterrence and retribution as motives and shows how they affect the choice
between individual and group punishment. The principal conclusion is that when
retribution is the primary goal, either punishment strategy may dominate, but when
deterrence is the goal, individual punishment dominates.

randomly selected individual from the group—is dominated by
group punishment. The reason is that, if punishing one person in the
group is socially advantageous (i.e. if it yields a net social benefit), then
punishing all members must yield at least n times that benefit. This logic
is most easily seen in the case where the enforcer incurs no punishment
costs. When that is true, it makes no sense to punish only one member
of the group since by punishing all of them, the true offender is
punished with certainty, thereby achieving the desired level of deter-
rence, revenge, or incapacitation. Adding punishment costs does not
alter this conclusion as long as the gain from punishment exceeds the
costs on a per-person basis as the scale of punishment is increased.
Random punishment does, however, possibly dominate group
punishment if punishment costs are increasing at the margin. Thus,
in cases where detection is either very difficult, or where all group
members are seen as culpable but there are decreasing returns to scale
in punishment costs, random individual punishment may make
sense. The case of decimation is a good example of this scenario
since the imposition of some punishment no doubt provided a strong
deterrent against mutiny or flight in battle, but execution or incap-
acitation of all members of an offending company would have been
impractical for obvious manpower reasons.
The other situation where group punishment possibly offers strong
advantages, as noted above, is when the members of the group can
thereby be induced to reveal the identity of the true offender, or to
impose their own form of punishment. In this case, the threat of
group punishment effectively enlists the members of the group as
surrogate enforcers, thus minimizing the cost of enforcement to the
central authority.
Economic theory provides further insight into this particular
motivation for group punishment. As shown in the Appendix, the
size of the group likely has an important impact on the efficacy of
group punishment for inducing the group to impose internal sanc-
tions, but the effects are cross-cutting. On one hand, it would not
seem to be effective if the offender is a member of a large group
because then it would be fairly easy for the latter to remain anonym-
ous to other members, and free-rider problems would preclude them
from investing much effort in identifying him. Based on this logic,
Parisi and Dari-Mattiacci (2004: 504) argue that growth in the size of
communities over time “might explain the historical evolution
towards systems of individual responsibility and in general towards

the limitation of the size of the group that was to be held collectively
responsible for the wrongs of its members.”
On the other hand, small groups would likely have significant
group solidarity, and therefore would be reluctant to turn over one
of their own members. Thus, for example, sanctioning of fraternities
for campus pranks would not likely be an effective way to discover the
offender’s identity (as exemplified by the movie Animal House).²¹
Another example, suggested by Akerlof and Yellen (1994) in the
context of gang behavior, emphasizes the importance of a commu-
nity’s willingness to cooperate with police in identifying offenders. In
this case, the use of group sanctions may significantly inhibit such
cooperation by fostering an “us vs them” attitude among gang mem-
bers. (Although in this case, members of the group in some sense
inherit the offender’s guilt by their refusal to “out him.”) In the same
vein, some contend that harsh sanctions against countries harboring
terrorists might actually incite more terrorist acts. Taken together,
these arguments suggest that the information-revealing function of
group punishment might best be implemented for moderately sized
groups and those with memberships that are not long-lasting so as to
avoid the emergence of strong intra-group solidarity.²²
The use of vicarious liability in Roman law and the modern
manifestations of this practice represent a closely related form of
group responsibility. Under this strategy, a principal (an employer
or head of a household) is sanctioned in the hopes that he or she will
pass the sanction on to an offending agent who is either in the
principal’s employ or is otherwise under his or her control (as in
the case of noxal liability). As Levinson (2003: 349) notes, “Vicarious
liability might be understood as a special type of collective sanction
regime, characterized by a target group consisting of (only) two
individuals bound together in a contractual relationship.”²³ In
²¹ But see Feinberg (1991: 61), who argues that collective responsibility would be
most effective in groups with a “high degree” of solidarity because such groups would
be more likely to share common interests and thus to suffer sanctions in common.
Although this might promote mutual restraint before an act is committed, it will also
likely prevent the group from turning over a guilty party after the fact.
²² See Heckathorn (1990) for a formal analysis of intra-group control norms in the
presence of group sanctions.
²³ For a formal model of this in an environmental context, see Segerson and
Tietenberg (1992). For more general treatments, see Kornhauser (1982); Sykes
(1984); Feinberg (1991); and Dari-Mattiacci and Parisi (2004).

addition to providing incentives for offenders to refrain from harmful
acts, vicarious liability can also increase the chances of obtaining
compensation for victims (when that is one of the social goals) by
correcting for judgment-proof problems, such as when employees,
minor family members, animals, or slaves lack the personal resources
to cover any damages that they impose (Parisi 2001: 111).²⁴ An
offsetting effect, noted by Arlen (1994) in the context of corporate
liability, is that when corporations face the threat of vicarious liability
they might actually refrain from policing their employees for fear that
doing so will increase the chances that crimes will be uncovered,
thereby potentially raising the company’s expected liability.²⁵
Notwithstanding the foregoing benefits from group responsibility,
it is undoubtedly the case that the primary objection in modern eyes
to this form of punishment in most situations is the attribution of
guilt to all members of the group. In addition to the stigma that this
creates, especially when the punishment is criminal in nature, there is
a risk that it might justify other forms of punishment or discrimin-
ation against the group. For example, the practice of statistical
discrimination (or profiling) based on race, gender, or ethnic back-
ground is a form of group punishment that may be justified on purely
efficiency grounds due to imperfect information, but is in many
(though not all)²⁶ cases viewed as offensive and possibly unconstitu-
tional (Posner 2003: 689–90). This has not always been the case,
however, as exemplified by the example, noted above, of the intern-
ment of Japanese Americans during World War II, an action that was
accepted as a necessary security measure at the time, but which would
be unthinkable today, only two generations later.
What these examples reveal is that the acceptability of group
responsibility as an enforcement strategy depends on factors that
vary with both time and circumstance, and reflect a balancing of
social values about guilt and punishment. As Feinberg (1991: 67)
observes, the general demise of collective responsibility has not
occurred because “individual responsibility is an eternal law of
²⁴ On the judgment-proof problem in a tort context, see Shavell (1986); Miceli and
Segerson (2003).
²⁵ For further discussion of collective responsibility in the context of corporations,
see French (1991); Velasquez (1991).
²⁶ The setting of different auto insurance rates based on age and gender, for
example, apparently remains acceptable.

reason,” but because the conditions that made it reasonable or
necessary in ancient societies are rarely present in modern times.
23.6. CONCLUSI ON
Collective responsibility for wrongdoing is a law enforcement practice

that is often associated with ancient society, or at least with notions of
vengeance, as exemplified by the infamous Hatfield–McCoy feud
(Alther 2012). A careful examination of the practice, however, reveals
that several forms of group punishment remained in use in highly
civilized societies like ancient Greece and Rome, and continue to be
used in various ways in modern society, both informally and doctri-
nally. The persistence of collective responsibility, or group punish-
ment, suggests that the strategy must serve a useful law enforcement
function in at least some circumstances. Prompted by this insight, this
chapter has examined the economic logic underlying the use of group
punishment by focusing on the costs and benefits of the practice in
relation to individual punishment. The conclusions reveal that col-
lective responsibility is often a rational social response to prevailing
circumstances, such as when individual punishment is inefficient or
impractical, or when members of the group are harboring the offend-
ing party or otherwise share in his or her responsibility for the
wrongful act.
A P P E N DI X : A S I MPL E M O DE L O F I N DI V I DU A L
VS GROUP PUNISHMENT
This appendix derives the formal results that are referred to in the body of
the text. The analysis will make use of the following notation:
B = social benefit of punishing the true offender, reflecting either
deterrence, retribution, or the desire for compensation of victims;
k = social cost incurred for each person wrongfully punished;
n = smallest group of which the offender is known to be a member, n > 1;
p = probability of correctly apprehending the true offender;
c(p,n) = cost of detection, where cp > 0, cpp > 0, cn > 0, and cpn > 0.

The cost of detection function reflects the best available technology for
identifying the true offender under a strategy of individual punishment. It
seems plausible to assume that both the overall cost, and the marginal cost of
increasing p (i.e. of achieving greater certainty), are increasing in the
group size.
Given this notation, we can write the expected social benefits from
individual (I) and group (G) punishment as:
WI ¼ pB ð1 pÞk cðp; nÞ; ð1Þ
WG ¼ B ðn 1Þk: ð2Þ
For sake of comparison, we assume equal punishments under the two regimes,
given that both B and k will depend on the nature and magnitude of the
sanction.²⁷ In the case of individual punishment, the enforcer will ordinarily be
able to choose p to maximize WI, which yields the first order condition:
B þ k ¼ cp : ð3Þ
According to (3), the optimal probability of apprehension, denoted p*,

equates the marginal benefit of correct punishment, consisting of the direct
benefit B plus the savings in error costs k, to the marginal cost. Totally
differentiating (3) and using the assumption that cpn > 0 shows that p* is
decreasing in n and increasing in k and B. Thus, under individual punish-
ment, the enforcer devotes less effort to identifying the true offender as the
group size increases, and more effort as the cost of erroneous punishment
and/or the benefit of conviction increases.
Using this model, we can prove several of the claims asserted in the text.
Result 1: When the cost of erroneous punishment is zero, group pun-
ishment is preferred to individual punishment for any choice of p.
This follows immediately from a comparison of (1) and (2) when k = 0.
Intuitively, group punishment both ensures that the true offender is pun-
ished and saves enforcement costs compared to individual punishment. This
establishes that the primary justification for individual punishment is the
aversion to wrongful punishment.
²⁷ See Miceli and Segerson (2007) for a model that allows an endogenous choice of
the punishment under each regime, and also formally distinguishes between deter-
rence and retribution as social goals of punishment.

One step on the road to individual punishment is punishment of a
randomly chosen member of the group known to harbor the true offender.
Result 2: Random individual punishment achieves the same level of
social welfare, on a per-person basis, as group punishment.
To prove this, note first that under random individual punishment, p = 1/n
and c(1/n,n) = 0 (given that no enforcement costs are expended). Thus, (1)
reduces to:
WI R ¼ B=n ð1 1=nÞk
ð4Þ
¼ ½B ðn 1Þk=n:
Comparing (4) to (2) shows that WI R ¼ WG =n. Further, assuming that

B ðn 1Þk > 0; WG > WI R provided n > 1. It therefore follows that:
Result 3: Group punishment is preferred to random individual punish-
ment, provided that punishment is desirable at all.
Thus, if random individual punishment is desirable in the sense that it yields
a net social gain, then punishing all members of the group must yield
proportionately greater social benefits, given constant returns to scale.
The preceding conclusions imply that the benefits of individual punish-
ment must come from optimal detection efforts. Indeed, when p is chosen
optimally we have:
Result 4: As the cost of erroneous punishment rises, individual punish-
ment will eventually be preferred to group punishment.
The effects of an increase in k on welfare under the two punishment regimes
are given by:
@WG
¼ ðn 1Þ < 0 ð5Þ
@k
@WI
¼ ð1 p Þ < 0; ð6Þ
@k
where the Envelope Theorem was invoked to obtain (6). From Result 1
above, we know that WG > WI when k = 0, but both expressions are
decreasing in k. However, WG declines linearly whereas WI is convex
given that p* rises with k. Thus, for large enough k, WI will cross WG
from below.
One claimed advantage of group punishment is that it may induce
members of the group to identify the offender.

Result 5: Social welfare from group punishment is increasing in the
probability that the group will turn over the true offender.
Let q be the probability that the group will identify the true offender when
faced with group punishment. In that case, expected social welfare under
group punishment becomes:
WG ¼ B ð1 qÞðn 1Þk: ð7Þ
This expression is clearly increasing in q, reflecting the savings in the

expected costs of erroneous punishment when the group may turn over the
offender. Generally, we would expect q to be a function of the group size, but
the sign of q0 (n) may be positive or negative. On one hand, q0 < 0 because an
offender can avoid detection more easily in a larger group, but on the other,
q0 > 0 if smaller groups have greater solidarity and hence are less willing to
turn over one of their own. Thus, we have:
Result 6: If q0 < 0, welfare under group punishment is decreasing in
group size, but if q0 > 0, welfare under group punishment may be increas-
ing or decreasing in group size.
Differentiating (7) yields:
@WG
¼ ð1 qpÞk þ ðn 1Þkq0 : ð8Þ
@n
The first term is the direct increase in error costs from an increase in n, while
the second reflects the effect of n on q. If q0 < 0, this term reinforces the first
term and the entire expression is negative, but if q0 > 0, this term offsets the
first term and the entire expression is ambiguous in sign. It follows from
these results that the most effective use of group punishment may be for
moderately sized groups. One may also infer that group punishment would
be most effective for groups that endure only temporarily so that members
don’t have an opportunity to establish significant intra-group solidarity (in
which case q would be high).
Finally, suppose that punishment is costly to impose (apart from error costs),
and let s be the social cost per person punished. Now (1) and (2) become:
WI ¼ pB ð1 pÞk s cðp; nÞ; ð9Þ

WG ¼ B ðn 1Þk ns: ð10Þ
Clearly, the higher cost of punishment under group punishment is now a

significant disadvantage compared to individual punishment, all else equal.

This will be especially true if s is itself increasing in n; that is, if marginal
punishment costs are increasing. In this case, random individual punishment
is no longer simply equal to WG =n due to decreasing returns to scale in
punishment costs. Rather, WI R > WG =n, and it is no longer necessarily true
that group punishment dominates random individual punishment. Beyond
these factors, the other results derived above for the case where s = 0 remain
largely unaffected.
REFERENCES
Akerlof, George, and Janet Yellen. 1994. “Gang behavior, law enforcement,
and community values,” in Henry Aaron, Thomas E. Mann, and Timothy
T. Taylor, eds, Values and Public Policy. Washington, DC: Brookings,
173–209.
Alther, Lisa. 2012. Blood Feud: The Hatfields and the McCoys: The Epic Story
of Murder and Vengeance. Guilford: Lyons Press.
Arlen, Jennifer. 1994. “The potentially perverse effects of corporate criminal
liability.” 23 Journal of Legal Studies 833–67.
Becker, Gary. 1968. “Crime and punishment: an economic approach.” 76
Journal of Political Economy 169–217.
Blackstone, William. 1766. Commentaries on the Laws of England, 1st edn.
Oxford: Clarendon Press.
Buckland, William W., Arnold McNair, and Frederick H. Lawson. 1952.
Roman Law and Common Law: A Comparison in Outline. Cambridge:
Buell, Lawrence. 1986. New England Literary Culture. Cambridge: Cam-
Cursi, M. Floriana. 2002. Iniuria cum damno: Antigiuridictia e Colpevolezza
Nella Storia del Danno Aquiliano. Milan: Giuffrè.
Dari-Mattiacci, Giuseppe, and Francesco Parisi. 2004. “The cost of delegated
control: vicarious liability, secondary liability and mandatory insurance.”
23 International Review of Law and Economics 453–75.
du Plessis, Paul. 2015. Borkowski’s Textbook on Roman Law, 5th edn. Oxford:
Feinberg, Joel. 1991. “Collective responsibility,” in May and Hoffman, eds,
53–76.
French, Peter. 1991. “The corporation as a moral person,” in May and
Hoffman, eds, 133–49.
Gardner, Jane. 2011. “Slavery and Roman law,” in Keith Bradley and Paul
Cartledge, eds, The Cambridge World History of Slavery. Cambridge:
Garoupa, Nuno. 2000. “Corporate criminal law and organization incentives:
a managerial perspective.” 21 Managerial and Decision Economics 243–52.
Harries, Jill. 2007. Law and Crime in the Roman World. Cambridge:

Harries, Jill. 2013. “The Senatus Consultum Silanianum: court decisions and
judicial severity in the early Roman Empire,” in Paul du Plessis, ed., New
Frontiers: Law and Society in the Roman World. Edinburgh: Edinburgh
Heckathorn, Douglas. 1990. “Collective sanctions and compliance norms: a
formal theory of group mediated social control.” 55 American Sociological
Review 366–84.
Hemingway, Ernest. 1929. A Farewell to Arms. New York: Scribner.
Holmes, Oliver Wendell. 1881 [repr. 1963]. The Common Law. Boston:
Little-Brown.
Johnson, Dominic. 2016. God Is Watching You: How the Fear of God Makes
Us Human. New York: Oxford University Press.
Kaser, Max. 1971. Das römische Privatrecht. Erster Abschnitt: Das altrö-
mische, das vorklassische und klassische Recht, 2nd edn. Munich: Beck.
Kitson, Peter. 1989. “Coleridge, the French Revolution, and the ‘Ancient
Mariner’: collective guilt and individual salvation.” 19 Yearbook of English
Studies 197–207.
Kornhauser, Lewis. 1982. “An economic analysis of the choice between
enterprise and personal liability.” 70 California Law Review 1345–92.
Lanni, Adriaan. 2017. “Collective sanctions in classical Athens,” in Dennis
P. Kehoe and Thomas A. J. McGinn, eds, Ancient Law, Ancient Society.
Ann Arbor: The University of Michigan Press, 9–31.
Levinson, Daryl. 2003. “Collective sanctions.” 56 Stanford Law Review
345–428.
Levmore, Saul. 1995a. “Rethinking group responsibility and strategic
threats in biblical texts and modern law.” 71 Chicago-Kent Law Review
85–121.
Levmore, Saul. 1995b. “Gomorrah to Ybarra and More: overextraction and
the puzzle of immoderate group liability.” 81 Virginia Law Review
1561–1604.
Lintott, Andrew. 1999. Violence in Republican Rome, 2nd edn. Oxford:
May, Larry, and Stacey Hoffman, eds. 1991. Collective Responsibility: Five
Decades of Debate in Theoretical and Applied Ethics. Savage: Rowan and
Littlefield.
Menell, Peter. 1991. “The limitations of legal institutions for addressing
environmental risks.” 5 Journal of Economic Perspectives 93–113.
Miceli, Thomas. 1991. “Optimal criminal procedure: fairness and deter-
rence.” 11 International Review of Law and Economics 3–10.
Miceli, Thomas, and Kathleen Segerson. 2003. “A note on optimal care by
wealth-constrained injurers.” 23 International Review of Law and Eco-
nomics 273–84.
Miceli, Thomas, and Kathleen Segerson. 2007. “Punishing the innocent
along with the guilty: the economics of individual versus group punish-
ment.” 36 Journal of Legal Studies 81–106.

Parisi, Francesco. 2001. “The genesis of ancient law.” 3 American Law and
Economics Review 82–124.
Parisi, Francesco, and Giuseppe Dari-Mattiacci. 2004. “The rise and fall of
communal liability in ancient law.” 24 International Review of Law and
Economics 489–505.
Polinsky, A. Mitchell, and Steven Shavell. 2000. “The economic theory of
public enforcement of law.” 38 Journal of Economic Literature 45–76.
Posner, Richard. 2003. Economic Analysis of Law, 6th edn. New York: Aspen
Publishers.
Posner, Richard. 1983. The Economics of Justice. Cambridge, MA: Harvard
University Press.
Segerson, Kathleen, and Thomas Tietenberg. 1992. “The structure of penal-
ties in environmental enforcement: an economic analysis.” 23 Journal of
Environmental Economics and Management 179–200.
Shavell, Steven. 1985. “Uncertainty over causation and the determination of
civil liability.” 28 Journal of Law and Economics 587–609.
Shavell, Steven. 1986. “The judgment proof problem.” 6 International Review
of Law and Economics 45–58.
Sherwin-White, Adrian N. 1985 [1966]. The Letters of Pliny: A Social and
Historical Commentary. Oxford: Oxford University Press.
Simic, Charles. 2013. “Oh, what a lovely war.” 40.15 New York Review of
Books, 21–3.
Smith, Adam. 1759 [1976]. The Theory of Moral Sentiments. Indianapolis:
Liberty Classics.
Sykes, Alan. 1984. “The economics of vicarious liability.” 93 Yale Law
Journal 1231–80.
Varian, Hal. 1990. “Monitoring agents with other agents.” 46 Journal of
Institutional and Theoretical Economics 153–74.
Velasquez, Manuel. 1991. “Why corporations are not morally responsible for
anything they do,” in May and Hoffman, eds, 111–31.
Wade, Nicholas. 2009. The Faith Instinct: How Religion Evolved and Why it
Endures. New York: Penguin Books.
Watson, Alan. 1991. Roman Law and Comparative Law. Athens: University
of Georgia Press.
Wood, Michael. 1998. In Search of the Trojan War. Berkeley: University of
California Press.
Zimmermann, Reinhard. 1996. The Law of Obligations: Roman Foundations
of the Civilian Tradition. Oxford: Oxford University Press (first publ. Cape
Town: Juta & Co., Ltd., 1990).
24
The Dual Origin of the Duty to

Disclose in Roman Law
Barbara Abatino and Giuseppe Dari-Mattiacci
24.1. INTRODUCTION
To guarantee the smooth functioning of markets, it is crucial that

buyers have the opportunity to invoke legal protection if the goods
they have purchased turn out to lack essential characteristics. Two
different sets of officials, the aediles curules (having jurisdiction over
regulated cattle and slave markets) and the praetor (having general
civilian jurisdiction over contracts), developed remedies for noncon-
formity in sale contracts in Roman law. The aedilician remedies gave
buyers a choice between an off-the-contract remedy, determining the
restitution of both the good and the price paid (actio redhibitoria),
and an on-the-contract remedy, affirming the contract while reducing
the price (actio quanti minoris).¹ In contrast, the praetor initially
provided and developed a single on-the-contract remedy (actio ex
empto), affirming the contact but allowing the buyer to claim dam-
ages (damnum emergens, close to the modern reliance damages),
which could possibly go beyond the restitution of the price paid.²
¹ For an economic analysis of modern on- and off-contract remedies, see Brooks
and Stremitzer (2011, 2012).
² See Arangio-Ruiz (1956: 237–9 and 242–3); Talamanca (1990: 657–8). An actio
empti ad redhibendum (or actio empti ad resolvendam emptionem, rescinding the
contract) is first attested in the imperial period. We will provide details on this remedy
below in the text.
Barbara Abatino and Giuseppe Dari-Mattiacci, The Dual Origin of the Duty to Disclose in Roman Law In: Roman Law and
DOI: 10.1093/oso/9780198787211.003.0024
402 Barbara Abatino and Giuseppe Dari-Mattiacci

Under pre-Justinianic Roman law, aedilician and praetorian
remedies were strikingly different along three additional dimensions:
first, the aedilician remedies could be claimed only for a limited set of
clearly enumerated defects, while the praetorian remedies were open-
ended; second, the aedilician remedies applied irrespective of the
fraudulent conduct used by the seller to deceive the purchaser
(dolus), while the praetorian remedies required proof of dolus; third,
the aedilician remedies were subject to short statutes of limitations—
six months for the actio redhibitoria and one year for the actio quanti
minoris³—while the praetorian remedies were not subject to
prescription.
In this chapter, after offering some details on the lawmaking
institutions and the remedies (§§24.3 and 24.4), we will show that
the differences in the remedies were designed to maximize the value
of transactions. The aedilician remedies (actio redhibitoria and actio
quanti minoris) maximized the value of a typical “market transac-
tion” concluded between a professional seller and a private buyer in
the markets that were under the jurisdiction of the aediles. In con-
trast, the praetorian remedy (actio ex empto) maximized the value of
the typical “private transaction” concluded between two private
parties outside a regulated market. In addition, the differences
between these two sets of remedies allowed buyers to optimally
choose between private and market transactions, thereby optimally
opting into the two different sets of remedies and hence further
enhancing their efficiency (§24.5). The availability of efficient boiler-
plate prescriptions reduces the need for ad hoc contractual arrange-
ments and hence reduces transactions costs.
We focus our attention on the city of Rome.⁴ The remedies we
study developed at a time when the jurisdictions of the two magis-
trates were clearly distinct. The curule aediles were magistrates in
charge of public order in markets where cattle and slaves were sold
through auctions. As part of the cura annonae—concerning the smooth
functioning of the markets and the distribution of foodstuffs—they
had the power of iurisdictio and adjudication only concerning
³ These actiones could also be used to obtain a warranty against eviction from the
seller (through stipulatio); in this case the terms were shorter. See Arangio Ruiz (1956:
367–8).
⁴ But see TPSulp. 42, l. l–2, and TPSulp. l. 3–4 with references to the aedilician
edict, showing the validity of the remedies within the Italian territory.
The Dual Origin of the Duty to Disclose in Roman Law 403

controversies arising in the regulated markets. Most likely, they
developed the actio redhibitoria first (plausibly, in the mid-second
century BC) and later the actio quanti minoris (plausibly before the
first century BC).⁵ The praetor, by contrast, had general jurisdiction
over contracts, not limited to specific types of transactions. The actio
ex empto was possibly introduced as early as the third century BC.⁶ Later,
in the Augustan period, the praetor also introduced an off-the-contract
remedy, the actio ex empto ad redhibendum, including the possibility of
obtaining restitution. By the time of Justinianic Corpus Iuris Civilis, both
the aediles and the praetor gradually lost their initial independence in
the administration of justice and the aedilician remedies were extended
to all kinds of sales. As a result, the aedilician and praetorian remedies
converged and blurred together. The timing of emergence and evolu-
tion of the different remedies, and especially the fact that in the aedili-
cian remedies the on-the-contract remedy followed the off-the-contract
remedy, will be shown to be consistent with the view that the remedies
were designed to maximize the value of transactions. We will conclude
with some considerations on why both the aediles and the praetor so
effectively catered to the interests of the contracting parties to whom
their remedies were addressed (§24.6).
The core of our argument is as follows. Remedies for defects in sale
contracts address a fundamental economic problem: the seller is
usually better informed than the buyer about the characteristics of
the goods for sale and might have incentives to conceal defects or
misrepresent other characteristics. Both the aedilician and the prae-
torian remedies addressed this problem but did so in different ways.
The aediles allowed buyers a remedy only if the seller had not
provided the correct information about the quality of the goods. All
buyers value goods of higher quality (for instance, a trustworthy
slave) more than goods of lower quality (for instance, a slave prone
to theft). Other idiosyncratic characteristics of the goods (for
instance, specific slave skills, such as baking) that might be important
for some individual buyers but not for others were left out of the
reach of the aedilician remedies. This allowed for quick transactions
(only information about quality was exchanged) and fast dispute
resolution (only some predefined quality markers were subject to
scrutiny ex post). In addition, the buyer did not need to prove the
⁵ See Donadio (2004: 40–54, esp. 45).

⁶ See Donadio (2004: 37–8 and 2007: 510–11). Cf. Watson (1987: 167–75).

dolus (fraud) of the seller, which further simplified and accelerated
dispute resolution. Finally, quality was relatively easy to verify after
the purchase, so that the prescription for the aedilician remedies
could be short. In turn, a short prescription reduced the cost of having
some goods returned to innocent sellers—those who had misled the
buyer in good faith. The downside of the aedilician remedies was that
they did not allow for an effective exchange of information about
idiosyncratic characteristics of the good prior to the contract, which
had a negative effect on the value of the transaction in all those cases
in which a buyer valued such characteristics.
In contrast, the remedies applied by the praetor allowed for the
verification of both quality and idiosyncratic characteristics but
limited the liability of the seller to cases of dolus. This made trials
potentially longer and more expensive but had two positive effects.
On the one hand, sellers had incentives to reveal more information
than under the aedilician remedies, which increased the value of the
transaction for those buyers who valued idiosyncratic characteristics
besides quality. On the other hand, sellers were not liable for innocent
misrepresentation, which had a dampening effect on prices as the
contract did not include an implicit insurance against such an even-
tuality. The praetorian remedies were not subject to prescription. This
was likely due to the fact that some characteristics might become
evident after a long time—hence the buyer needed to be given more
time in order to have effective protection. Moreover, since goods were
not returned to the seller under the actio ex empto, the absence of a
prescription period did not expose the seller to resale risk. The
measure of compensation could not be established in relation to a
market price due to the private nature of the transaction; therefore,
damages were calculated according to the buyer’s negative interest,
that is, his situation before the contract.
Typically, buyers involved in market transactions were most likely
those with little interest in idiosyncratic characteristics, while those
looking for specific slaves or cattle would probably self-select into
private transactions. As a result, the transactions that fell under the
jurisdiction of the aediles were those for which the aedilician remed-
ies were most effective—and, hence, buyers had, even if given a
choice, no incentives to opt for a praetorian remedy. Likewise, the
transactions that fell under the jurisdiction of the praetor were those
involving idiosyncratic buyers who valued the extended protection
given by the praetorian law. In the following, we will elaborate upon
and demonstrate these claims.

24.2. INSTITUTIONS AND REMEDIES
24.2.1. Institutions
The praetorship was created in 367 BC to take over some of the duties
of the consuls. In exchange for opening the consulship to plebeians,
the patricians obtained that the praetor would be a patrician. Yet, the
first plebeian praetor was elected as early as 337 BC, so the praetorship
cannot be defined as a patrician institution. The praetor had the
power of iurisdictio, that is, to resolve disputes between litigants⁷
and of ius edicendi, that is, the power to issue an edict listing the
remedies available to litigants.⁸
The aediles were lower-ranking plebeian magistrates created in 449
BC⁹ and elected by the plebeian council (concilia plebis) to collaborate
with the main plebeian magistrates, the tribunes, in overseeing and
policing markets. Two additional patrician aediles elected by the
comitia tributa were added probably after 367 BC. Like the praetor,
the aediles also held office for one year and their functions included
the cura urbis (the management of the city roads, baths, and build-
ings), the cura ludorum (the management of the public games), and
the cura annonae (the control and policing of the city markets).¹⁰ The
aediles had limited coercive powers, including the power to issue fines
and, within their prerogatives concerning markets, iurisdictio and ius
edicendi, the power to resolve disputes and issue an annual edict
listing remedies available to litigants. However, as opposed to the
praetor’s edict, the edict of the aediles was limited to the remedies
necessary to resolve the disputes arising in the markets under their
jurisdiction.¹¹
Therefore, the aedilician remedies and, in particular the actio red-
hibitoria, were created as specific solutions for cases in which the
seller had failed to disclose a particular defect (vitium) to the buyer
⁷ The praetor urbanus had jurisdiction over disputes between Roman citizens.
Proceedings before the praetor were in iure: the praetor would typically give the
parties a formula, stating the parties’ claims and instructing the iudex to verify the
facts and award a remedy accordingly. From the third century BC, a praetor peregrinus
with jurisdiction over disputes involving foreigners was added to the praetor urbanus.
⁸ See de Ligt in volume I.
⁹ See the Lex Valeria Horatia de tribunicia potestate.
¹⁰ On permanent and periodic markets, see de Ligt and de Neeve (1988: 391–416);
de Ligt (1993). Cf. also Lo Cascio (2000).
¹¹ On the aedilician edict as portio iuris honorarii, see Guarino (1955, 1956). Cf.
Volterra (1955, 1956). Recently see Chevreau (2012: 225–9).

prior to a sale effected in markets where cattle and slaves were
traded.¹² Sales in these markets were typically executed through an
open ascending price auction (now known as an English auction):
goods were offered for sale by an auctioneer (praeco) for a base price
and assigned to the highest bidder.¹³ The auctions were announced in
advance (TPSulp. 83–6, 88, 90–3)¹⁴ and, at least at Pompeii and
Puteoli in the early imperial period, organized by a banker (argenta-
rius or coactor)¹⁵ who could also lend money to the buyer.¹⁶ If the
banker was hired by the seller—normally through a contract of
hire, locatio conductio¹⁷—he could also guarantee (promissio auctio-
natoris; Donadio 2011) the payment of the price after having
deducted the expenses for the organization of the auction (TPSulp. 81;
D. 46.3.88.5). As with the praetor, the edict of the aediles also
gradually lost importance during the imperial period.
24.2.2. Remedies
It is not known when exactly the aedilician remedies were introduced.
However, it seems certain that the aediles introduced first the actio
redhibitoria (most likely before the second century BC; Donadio 2004:
45) and later the actio quanti minoris (most likely before the first
century BC; Cic., De off. 3.17.71 and 3.23.91; Jakab 1997: 126–7). Both
remedies imposed on the seller a duty to disclose whether the good
for sale was affected by any of the defects indicated in the aedilician
edict. Failure to do so gave rise to the buyers’ right to ask for a
remedy. Note that not all defects were relevant but only those listed,
¹² See Serrao (2000); cf. Donadio, (2007: 464 and n. 11) with additional literature
references. On the economic analysis of the aedilician remedies, see Kupisch (2002:
21–54, esp. 41 n. 74).
¹³ On modern auction theory, see McAfee and McMillan (1987); Krishna (2002).
On the function of auctions in ancient Rome, see Malmendier (2002: 101–5); cf.
Ankum (1972: 377–93).
¹⁴ TPSulp. 81 describes a case in which the intermediary lends money to the buyer
to pay the seller. TPSulp. 90–3 documents instead cash payments by the buyer. For
epigraphic evidence, see Gröschler (1997: 35–8); Camodeca (1999: 185–206). Cf. Bove
(1975: 322–31); Wolf (2010: 105–21).
¹⁵ See Andreau (1974: 77–81).
¹⁶ See the apochae Iucundianae (e.g. CIL IV 3340, 1–137) and TPSulp. 82. For
more details, see Camodeca (1999: 185); Lerouxel (2016: 204–9, 281–7).
¹⁷ On locatio conductio, see Thielmann and Bettermann (1961); cf. Thomas (1966)
and Bove (1975: 327).

although later the jurists tried to give an expansive interpretation of
such list.
Legal and epigraphic sources provide examples of slaves who are
considered defective (wandering or runaway slaves) or diseased
(dumb, shortsighted, or dim-sighted: D. 21.1.9; Ulp. 44 ad Sab.;
D. 21.1.10.3 and D. 21.1.10.4; Ulp. 1 ad ed. aed. cur.). A well-known
text by the jurist Ulpian (D. 21.1.1.1; Ulp. 1 ad ed. aed. cur.) states that
“[t]hose who sell slaves are to apprise purchasers of any disease or
defect in their wares and whether a given slave is runaway, a loiterer
on errands, or still subject to noxal liability. . . . Again vendors must
declare at the time of sale all that follows: any capital offense com-
mitted by the slave; any attempt which he has made upon his own life;
and whether he has been sent into the arena to fight wild animals.”¹⁸
With the actio redhibitoria¹⁹ the buyer returned the good and
asked for the restitution of the price paid, obtaining a result that
tended to restore the status quo ante (in integrum restitutio;
D. 21.1.23.7; D. 21.1.60). The responsibility of the seller was inde-
pendent of his actual knowledge of the defect and hence was not
based on dolus.²⁰ Yet, Ulpian adds that “[i]f a defect in or disease of
the slave be perceptible (and defects reveal themselves generally
through symptoms), it may be said that the edict has no place; its
concern is simply to ensure that a purchaser is not deceived”
(D. 21.1.1.6; Ulp. 1 ad ed. aed. cur.). The ordinary prescription period
was six months from the day on which the contract of sale was
concluded.
Instead, with the actio quanti minoris,²¹ which was introduced with
respect to the sale of cattle and then extended to the sale of slave, the
buyer affirmed the contract but asked for the reduction of the price.
A iudex would then calculate the new price with reference to the
going market price for a similar (but defective) good. In this case as
well, the responsibility of the seller was not based on dolus. The
¹⁸ For this and other quotations we have adapted the English translation of the
Digest by Watson (2009).
¹⁹ See Guarino (1955: 295–6; 1956: 352–7); Volterra (1955: 3–7); Arangio-Ruiz
(1956: 353–8); Watson (1965: 86ff.); Burdese (1975: 594–600); Memmer (1990: 1–45);
Manna (1994: 137–46); Jakab (1997: 127–9); Garofalo (1998: 57–8); Donadio (2004:
173–7).
²⁰ See Hallebeek (2009: 10–179).
²¹ Giffard (1931: 682ff.); Arangio Ruiz (1990: 384, 391); Donadio (2007: 518–22).

ordinary prescription period was one year from the day the contract
was concluded.²²
The praetorian remedy, the actio ex empto, was older than the
aedilician remedies (third century BC) and gave the buyer the possi-
bility to ask for damages, while affirming the contract. Understand-
ably, since there was no market price to refer to, damages had to be
calculated according to the buyer’s negative interest (damnum emer-
gens) and were meant to make the buyer whole with respect to his
position prior to the contract. Damages did not include the buyer’s
lost profits (lucrum cessans) but only his reliance expenditures and
the difference between the price paid and his valuation of the good,
which approximated the price that the buyer would have paid rather
than the objective market price. Differently from the aedilician rem-
edies, the actio ex empto required proof of the seller’s dolus but could
be used to claim any kind of defect and was not subject to prescrip-
tion.²³ As a result, the praetor could consider the idiosyncratic inter-
ests of the buyer with respect to the good purchased and not only the
quality indicators listed by the aediles.
The actio ex empto affirmed the contract. During the Augustan
period there are traces of an actio empti ad redhibendum,²⁴ giving the
buyer the possibility of asking for restitution, an innovation first
advocated by the jurist Labeo (D. 19.1.11.3; Ulp. 32 ad ed.). This
was a later innovation, anticipating a gradual convergence of the
remedies as the functions of both aediles and praetor were absorbed
by the imperial administration of justice. Table 24.1 summarizes the
differences between the three original remedies.
Although there has been a long discussion as to the relationship
among these remedies,²⁵ it seems clear that the aedilician remedies
could be given only by the aediles within their jurisdiction and,
likewise, the praetorian remedy could only be awarded by the praetor.
Thus, buyers seeking a specific form of protection had to bring an
action in the corresponding jurisdiction. A buyer in a private sale
could only bring an action against the seller under the praetor’s
jurisdiction using the actio ex empto, since the aediles had no juris-
diction outside the regulated markets. A buyer in a market sale could
undertake a lawsuit with the aediles and choose between restitution
²² See Talamanca (1990: 591). ²³ Donadio (2007: 458 n. 7 and 510–18).

²⁴ Serrao (1959: 267–71); Vacca (1994: 60ff.).
²⁵ See Donadio (2004) for a detailed review of this debate.
Table 24.1. Aedilician and praetorian remedies.

Aedilician remedies Praetorian remedies
Actio redhibitoria Actio quanti minoris Actio ex empto
Effect Contract undone Contract affirmed Contract affirmed

Seller’s liability Restitution of the price Reduction of the price Reliance damages
Seller’s dolus Not relevant Not relevant Relevant
Type of defect Only listed defects Only listed defects Any defect
Prescription Six months One year No prescription
(actio redhibitoria) and price reduction (actio quanti minoris). In

addition, given the general jurisdiction of the praetor it seems plaus-
ible that a buyer in a market sale could also initiate a lawsuit with the
praetor using the actio ex empto (Donadio 2004: 34). Therefore, a
buyer who felt that the aedilician remedies did not fully protect his
interests could choose to pursue legal action with the praetor. Yet, as
we will argue, this latter option was probably not in the interest of
buyers and hence, although there are no data to support this claim,
was probably used infrequently.
24.3. ASYMMETRIC INFORMATION AND

SYMMETRIC IGNORANCE IN SALE CONTRACTS
In a typical sale contract, sellers have more information than buyers

about the quality of goods for sale, a problem known as asymmetric
information.²⁶ A feature of “quality” is that it has a common ranking
for buyers. This is not to say that all buyers attach the same value to
goods—valuations do indeed typically differ across individuals—but
rather that all buyers value high-quality goods more than low-quality
goods. As a result, sellers can exploit their informational advantage to
the detriment of buyers, by selling low-quality goods as if they were
²⁶ On the historical development of the legal relevance of information disparities,

see Decock and Hallebeek (2010: 89–133).

high-quality goods. Uncertainty about quality makes buyers wary and
unwilling to pay high prices. This in turn makes it unattractive for
sellers to sells products of high quality and generates a downward
spiral that drives high-quality products out of the market (adverse
selection: Akerlof 1970; cf. Frier and Kehoe 2007: 119–22). This well-
known problem can be solved by imposing a legal duty to disclose
information about quality on sellers, which in turn can be sustained
by penalties for failure to do so (Kronman 1978). We will show that
both the aedilician remedies and the praetorian remedies addressed
this problem.
Yet, reality is often more complex than that. Goods for sale might
have other “idiosyncratic characteristics,” which, in contrast to qual-
ity, buyers rank differently. To clarify, a buyer might be willing to pay
more for a slave baker rather than for a slave actor, while another
buyer might have the reverse preferences. Buyers’ preferences are
typically private information. Thus, as long as a seller does not
know a specific buyer’s idiosyncratic preferences, he cannot exploit
his informational advantage about the good’s idiosyncratic charac-
teristics. This state of “symmetric ignorance” (Barzel et al. 2006)
might actually facilitate trade if compared to asymmetric informa-
tion: since both parties are uninformed, they could trade all goods for
their expected value, without fear of being exploited. Ex ante, neither
party has an incentive to strategically withdraw from trade and,
hence, the adverse selection problem does not arise.²⁷
In contrast to asymmetric information, there are two ways to deal
with symmetric ignorance. One solution requires inducing buyers to
reveal their type to sellers and, subsequently, imposing a duty to
disclose upon sellers. As we will demonstrate, this was the effect of
the praetorian remedy. Another solution consists of preventing the
parties from exchanging information, thereby avoiding that a situ-
ation of symmetric ignorance might degenerate into one of asym-
metric information. Yet, in this case, a buyer might end up with a
good having undesired characteristics. We will show that this was the
effect of the aedilician remedies. It is worth nothing that both sets of
²⁷ It is worth noting that, if the buyers learned about the good’s characteristics they
would be better informed than sellers and an inverse-adverse-selection problem
would arise, since buyers could exploit their informational advantage to the detriment
of sellers (Barzel et al. 2006).

remedies were designed to assure that trading parties had the same
information (or lack thereof).
24.3.1. A Simple Model

In this section, we sketch a very simple model of sales. The model is
meant to capture the essential features of the transactions that gen-
erated the cases brought before the Roman magistrates and hence
disregards many important details of these transactions. Yet, the
stylized transactions that we describe in the following provide the
basic ingredients of our argument.
We consider a typical transaction between two parties, a seller and
a buyer, for the sale of a good. Sellers are homogeneous, while buyers
are of two types: one half of the buyers are of type A (for instance,
they run a business in which they employ actors) while the other half
are of type B (for instance, they run a business in which they employ
bakers).
Goods for sale can be distinguished along two dimensions: quality
and idiosyncratic characteristics. Quality can be either high (H) or
low (L), so that one half of the goods have high quality (for instance, a
healthy slave) and one half of the goods have low quality (an
unhealthy slave). The idiosyncratic characteristic of the good can be
either A (for instance, the slave is an actor) or B (the slave is a baker).
The crucial difference between quality and idiosyncratic characteris-
tics is that all buyers value high quality more than low quality, while
they value the idiosyncratic characteristic in a idiosyncratic way:
buyer A prefers good A while buyer B prefers good B. To operation-
alize these ideas, let us specify the buyer’s valuations as in Table 24.2:
high quality is valued at q by both buyers; further, each buyer attaches
a value c to his preferred characteristic. The valuation of a good by a
certain buyer is simply assumed to be the sum of these two values.
Ideally, an efficient market should allocate goods with characteristic
A to buyers of type A and goods with characteristic B to buyers of type
B. Finally, the seller values high-quality goods at h and low-quality
goods at l, irrespective of their characteristics.
In an ideal world with complete information, all transactions
would be efficient. However, in the real world, inefficiencies can
arise because of asymmetric information. Typically, sellers know
more about the goods for sale than buyers; hence, we assume that
Table 24.2. Buyers’ and seller’s valuations of the goods.
Idiosyncratic Idiosyncratic Idiosyncratic

characteristic characteristic characteristic
A B A B A B
Quality
Quality
Quality
H q+c q H q q+c H h h
L c 0 L 0 c L l l
I.1.1.3 I.1.1.6 I.1.1.9
Buyer A Buyer B Seller
sellers know the quality level and the idiosyncratic characteristics of

the goods, while the buyers only know their probabilities. This is a
plausible assumption because learning about, say, a slave takes time
and sellers might have acquired this information prior to the sale.
Conversely, buyers know their own preferences. This bilateral asym-
metric information can lead to inefficiencies. To illustrate, consider
an example.
If goods were sold without regard to asymmetric information, the
price for a representative good would be somewhere between the
buyer’s expected valuations, which is equal to (q + c)/2,²⁸ and
the seller’s average valuation, which is equal to (h + l)/2. If,
for instance, q = 120, c = 60, h = 100, and l = 20, all goods could in
theory be sold irrespective of quality and idiosyncratic characteristics,
because the expected buyer’s valuation, (120 + 60)/2 = 90, is higher than
the average seller’s valuation for the same goods, (100 + 20)/2 = 60, and
hence there is a price in between these values (say, 75) that would make
both of them better off.
However, the well-known problem of adverse selection (Akerlof
1970) might impair some or all of these transactions. To see why,
consider that the seller knows the quality of the good and he values
high-quality goods at 100. Hence, although on average the seller
makes a gain by selling all goods at, say, 75, he makes an even bigger
gain if the good is of low quality; this gain is partially offset by a loss
²⁸ The buyer does not know the level of quality and the idiosyncratic characteristic
of the good but knows that half of the goods have high quality and that half of
the goods have his preferred characteristic. Therefore, the expected value of a good is
q/2 + c/2.

associated with goods of high quality, which are sold below value.
Thus, if the price is less than 100, the seller, having information on
quality, will withdraw high-quality goods from the pool of goods for
sale. Since the maximum price that a buyer is willing to pay for a good
of unknown quality is 90, the only goods that the seller will put on the
market are low-quality goods, those goods that the seller values at less
than 90. If buyers anticipate the seller’s behavior, as they should
rationally do, they will expect to be sold only low-quality goods and
hence their willingness to pay will drop to c/2 = 30.²⁹ In turn, since the
seller values low-quality goods at 20, the price will be between 20 and
30 (say, 25). The result is that the market shrinks, with only low-
quality goods being sold due to asymmetric information and the
ensuing adverse selection problem—which, in turn, drives high-
quality goods out of the market and pushes prices down.
Note that if the seller could communicate the information he
possesses about quality, high- and low-quality goods could be sold
at different prices. Buyers are willing to pay 120 + 60/2 = 150 for a
high-quality good with uncertain idiosyncratic characteristics, which
the seller is willing to sell for at least 100. The problem with this is that
information exchange needs to be supported by penalties for reticent
or fraudulent behavior. Absent such penalties, the seller’s claims
about the quality of the goods for sale are not credible and hence
information cannot be effectively exchanged. Both the praetor and
the aediles developed remedies to address this problem and, more
specifically, to allow information exchange about quality.
Note further than exchanging information about quality is not
enough to ensure perfect matching, since A-goods could still end up
in the hands of B-buyers and B-goods in the hands of A-buyers. If
buyers knew the idiosyncratic characteristics and sellers knew the
buyers’ types, high-quality goods could be traded for a price between
120 + 60 = 180 (the buyer’s valuation of his preferred good) and 100
(the seller’s valuation), while low-quality goods could be traded for a
price between 60 and 20. With full information, A-buyers would buy
A-goods and B-buyers would buy B-goods. By ensuring perfect
matching ex post, full information would enhance the value of the
transaction for both parties. The remedies introduced by the praetor
and the aediles addressed this problem in diametrically different
²⁹ Recall that the buyer still buys without knowing the idiosyncratic characteristic
in this basic setup.

ways. While the praetorian remedies applied by the praetor stimu-
lated exchange of information about idiosyncratic characteristics—
striving to achieve symmetric information on both dimensions—the
aedilician remedies impaired it—leaving the parties symmetrically
ignorant about idiosyncratic characteristics but symmetrically
informed about quality.
24.3.2. Transactions before the aediles

The typical transaction within the jurisdiction of the aediles involved
a professional seller selling goods through an auction. Since an
auction does not involve negotiations, the seller did not learn the
buyer’s type. Similarly, the buyer remained uninformed about the
goods’ idiosyncratic characteristics and quality, since transactions
were closed rapidly and the buyer normally did not have enough
time to thoroughly inspect the good. Therefore, asymmetric informa-
tion persisted both with respect to quality and with respect to idio-
syncratic characteristics.
The extent to which parties could credibly exchange information
depended on the penalties that the aedilician remedies placed on false
or reticent information. The liability regime of the aediles was as follows:
1. Scope: a seller was liable only for defects pertaining to quality,
which were listed by the aediles.
2. Subjective knowledge: the seller’s subjective knowledge was
irrelevant (no inquiry about dolus).
3. Remedy: the buyer could choose between restitution (actio red-
hibitoria) and reduction of the price (actio quanti minoris).
4. Prescription: short prescription terms.
The aedilician remedies supported information exchange only with
respect to quality, so that sellers had incentives to reveal information
about quality and buyers could rely on it. This solved the adverse
selection problem examined above. With respect to the idiosyncratic
characteristics, both parties lacked some piece of information and
hence neither party was in a position to exploit his superior know-
ledge. Since the buyer was ignorant of the good’s idiosyncratic char-
acteristic and the seller was unaware of the buyer’s type, the parties
were in fact in a position of symmetric ignorance. Therefore, goods
were mismatched with some probability. In our simple model, half of

the goods are sold to the “wrong” buyer: an A-buyer could end up with
a B-good or vice versa. The aedilician remedies induced transactions
that were partially inefficient. In our simple model, we can measure this
inefficiency precisely: it is c/2, because in half of the cases the good is
sold to the buyer who values it the least and hence a value c is lost.³⁰
Yet, this allocative disadvantage was balanced by a reduction in the
overall costs of the transactions, due to the other features of the
aedilician remedies:
• Market expediency: transactions were fast and did not involve
lengthy negotiations precisely because information exchange was
limited to important quality features and could be standardized
through, for instance, the use of placards at auctions, the tituli.
• Market effectiveness: if quality turned out to be low, buyers
could retain the good for a lower price, if they wished to do
so; this reduced the need for additional transactions ex post
since some buyers might value the good even if the good was of
low quality and hence they would have bought it (for the lower
price) even if they had been informed. In such cases, restitution
is inefficient and price reduction is a better alternative.
• Low litigation costs: if problems arose, litigation was rapid and
probably not very costly, since it only involved the verification
of predetermined quality markers and did not require an
inquiry into the seller’s subjective knowledge.
• Convenient allocation of risk: the seller bore the risk of
unnoticed quality failures because he was liable even if he did
not know that quality was low. This was optimal, given that a
professional seller was better placed to resell goods than a buyer
and because this risk was reduced by the short statutes of
limitation, which allowed for the quick reselling of the goods
that were returned to the seller through the actio redhibitoria.
24.3.3. Transactions before the praetor

The typical transaction before the jurisdiction of the praetor
involved a seller who sold a good directly to a buyer. During the
³⁰ We implicitly assume that all goods offered by sellers would be traded under
perfect information. This assumption is inessential for the qualitative results of the
analysis.

negotiations prior to the contract, the seller could learn the
buyer’s type. For instance, the buyer searched for a good with
specific characteristics (say, a slave baker) and thereby revealed
his type. In this case, the seller became asymmetrically better
informed along both dimensions and hence had an incentive to
convey incorrect information both about quality and about idio-
syncratic characteristics. As a result, the praetorian remedies were
designed to induce information exchange by sellers along both
dimensions:
1. Scope: a seller was liable for all defects, not only those pertaining
to quality.
2. Subjective knowledge: the seller’s subjective knowledge was
relevant (the buyer needed to prove dolus).
3. Remedy: the actio ex empto allowed the buyer to receive dam-
ages equal to the difference between the price paid and the value
of the good to him (damnum emergens).
4. Prescription: no prescription term.
The praetorian remedy induced information exchange along both
dimensions and hence obtained a more efficient matching between
goods and buyers, thereby removing the efficiency loss typical of the
aedilician remedies. This was done at the price of inferior market
expediency, because negotiations took probably longer and were
more involved, and of higher litigation costs, because the inquiry
also concerned the idiosyncratic characteristics of the good and the
subjective knowledge of the seller.
Since the remedy of the actio ex empto, at least until the intro-
duction of an actio empti ad redhibendum, did not allow for the
restitution of the good but only for the payment of damages, ex post
market effectiveness was also lower. Some goods that the buyer
would not have bought under perfect information remained never-
theless in his possession—at least until they were resold. Moreover,
the praetorian remedies induced a different allocation of risk than
the aedilician remedies. Since the seller was only liable for dolus, the
buyer bore the risk for quality failures or of unsatisfactory charac-
teristics that the seller innocently ignored. Assigning more risk to
the buyer is consistent with the fact that sellers were not necessarily
professionals and hence it was not necessarily optimal to shift
liability to them.

24.3.4. The Buyers’ Choice of Market vs Private
Transactions
In the model, buyers with a large appreciation of idiosyncratic charac-
teristics compared to quality (large c relative to q) have an incentive to
self-select into personal transactions. They care about obtaining the
good of their preferred type and are willing to bear the higher quality
risks (as liability is only for dolus) and the larger negotiation and
litigation costs associated with operating under the praetorian law. In
contrast, buyers characterized by a low appreciation of idiosyncratic
characteristics relative to quality (small c relative to q) have an incen-
tive to self-select into market transactions and hence accept the risk of
not obtaining the good of their favorite type and, in exchange, face less
risk about quality and smaller negotiation and litigation costs.
The buyers’ behavior arguably enhanced the efficiency of the
remedies. The aedilician remedies suffered from inefficient matching,
but this problem would have been less severe if buyers who chose
market transactions were, as we postulate, characterized by small
appreciation for idiosyncratic characteristics. Hence, the cost-saving
properties of the aedilician remedies could be advanced without
much loss in terms of allocative efficiency. This argument also sug-
gests that buyers who initially chose market transactions were
unlikely to be willing to opt for the praetorian remedies and their
higher resolution costs. In turn, buyers who chose personal transac-
tions were characterized by a large appreciation for idiosyncratic
characteristics, which made the costs associated to the praetorian
remedies worth bearing.
24.4. ANALYSIS
24.4.1. The Aedilician Remedies

The aedilician remedies possibly applied to contracting parties who
interacted in the cattle and slave markets, in which sales were effected
through auctions. Since sellers did not generally have the opportunity
to interact with buyers prior to the sale, they were not in a position to
learn about a buyer’s preferences and hence could not easily exploit
their private information about the idiosyncratic characteristics of the

goods for sale. The aedilician remedies left buyers’ interests in idio-
syncratic characteristics unprotected: the buyer could only claim the
remedies if the seller had withheld information about the quality
characteristics listed in the aediles’ edict, which did not allow the
aediles to consider a buyer’s idiosyncratic interests.
Therefore, sellers could, if they wanted, give misleading informa-
tion about characteristics of the goods other than quality. This appar-
ently inefficient solution in fact allowed for quick transactions. Sellers
would expose the goods for sale (cattle or slaves) together with a
placard (titulus) indicating possible defects, the slave’s nationality,
and other standard quality markers.³¹ Since the only meaningful
exchange of information pertained to quality, sellers and buyers
would not engage in lengthy negotiations but rather would trade on
the basis of standardized procedures.
The aedilician remedies, however, gave sellers incentives to reveal
information about quality truthfully and, in turn, induced buyers to
rely on such information. If a buyer discovered that the good he
purchased had one of the defects listed in the edict, which the seller
had not disclosed, he could choose between the actio redhibitoria
(restitution) and the actio quanti minoris (price reduction). This
choice assured that buyers never paid more than what they were
willing to pay, irrespective of the quality of the goods purchased. To
clarify, the market generated, say, two prices: a high price for high-
quality goods and a low price for low-quality goods. The actio quanti
minoris adjusted downwards the price of goods with undisclosed
defects, so that the seller had to pay back to the buyer the difference
between the high price and the low price. Since buyers’ valuations
vary, a buyer who had purchased a good for a high price under the
assumption that the good was of high quality might not want to retain
the good once he discovered that the good was of low quality even if
the price was reduced. In this case, the buyer could opt for the actio
³¹ About tituli Gellius (Noctes Atticae 4.2.1) reports that Titulus servorum singu-
lorum scriptus sit curato ita, ut intellegi recte possit quid morbi vitiive cuique sit, quis
fugitivus errove sit noxave solutus non sit (“See to it that the sale ticket of each slave be
so written that it can be known exactly what disease or defect each one has, which one
is a runway or a vagabond, or is still under condemnation for some offence,” trans.
Rolfe 1927.) The text suggests that, by reading the titulus, purchasers obtained
information about diseases, defects, noxal liability, propensity for stealing, or loitering,
which might reduce the value of the slave.

redhibitoria. If instead the buyer was willing to buy a low-quality
good, he could choose the actio quanti minoris.
These considerations also explain why the actio redhibitoria
emerged earlier than the actio quanti minoris. Restitution is a safer
remedy from the perspective of buyers as it brings the buyer back to
the status quo ante. In contrast, price reduction could make some
buyers worse-off with respect to the status quo, because it might
impose a loss on buyers who have a very low valuation for low-
quality goods, and hence are unwilling to buy a low-quality good
for a low price. Assuming plausibly that buyers faced high transaction
costs when reselling goods, this outcome would be inefficient ex post
as would not optimally allocate goods to buyers.
Yet, in combination with the actio redhibitoria, the actio quanti
minoris enhanced the overall efficiency of the aedilician administra-
tion of justice. In fact, it allowed buyers who were willing to retain
low-quality goods to do so, thereby saving on the transaction costs
generated by returning the goods and offering them for sale at the
next market. On the seller’s side, the buyer’s choice was relatively
inconspicuous since most sellers were professionals who sold both
high-quality and low-quality goods for the market price. Therefore, if
the buyer decided to retain the good, we can assume that the seller
was still willing to sell for the low price.
Both remedies had short prescription periods. This was possible
since only certain basic qualities needed to be verified by buyers. In
addition, this allowed returned goods to be resold relatively quickly
and hence reduced the resale costs borne by sellers. This was an
important feature of the aedilician remedies. Since both remedies
applied irrespective of dolus, sellers were in fact obliged to offer an
implicit insurance against the eventuality that an unnoticed defect
might materialize. It was probably efficient to allocate the risk of
unnoticed defects to sellers due to the professional nature of most
of them, and also because this risk was reduced by applying relatively
short prescription periods.
Summing up, the aedilician remedies allowed for quick
transactions—because they reduced the amount of information that
the parties exchanged—and for quick, relatively inexpensive trials—
since they only allowed litigation on quality and did not examine
dolus. Moreover, by offering buyers a choice between undoing and
affirming the contract, they optimized the ex post allocation of goods
to buyers with respect to quality. Short prescription periods reduced

the risk to sellers. Yet, the downside was a loss of efficiency ex post:
since sellers could not credibly reveal information about idiosyncratic
characteristics, buyers might end up with goods having undesired
characteristics. The magnitude of this loss depended on how much
buyers valued idiosyncratic characteristics. This problem was
addressed by the praetorian remedy.
24.4.2. The Praetorian Remedies

The praetor initially offered only one remedy, the actio ex empto,
affirming the contract and awarding damages to the buyer. This
remedy principally concerned private transactions—where buyers
and sellers negotiated and exchanged information more intensely
than in an auction—and only secondarily pertained to transactions
at a public market. In private transactions, a buyer looking for a
specific good (for instance, a slave actor) would implicitly reveal his
type to the seller and hence put the seller in a position to exploit his
informational advantage both about the good’s quality (for instance,
whether the slave was prone to theft) and about its idiosyncratic
characteristics (for instance, whether the slave was a good actor).
Therefore, the buyer needed protection along both dimensions.
Accordingly, the actio ex empto could be used to claim damages for
undisclosed information about any characteristic of the good.
Since both parties in a private transaction were likely to be private
individuals rather than professional sellers, it is not obvious that the
seller was in the best position to bear the risk of unnoticed negative
characteristics. The actio ex empto awarded damages only in case of
dolus, thereby allocating this risk to the buyer. Since unnoticed
characteristics could also be of positive value, the buyer bore in fact
offsetting risks of both negative and positive discoveries.
This also explains why there was no prescription period: a buyer
needed time to verify the characteristics of the good and to collect
information about the seller’s prior knowledge. The natural decay of
evidence over time implicitly constrained the date within which a
successful suit could be brought.
Moreover, since he was dealing with private transactions, the
praetor could not easily refer to the market price of a good having
different characteristics from those claimed by the seller. The very
fact that the parties engaged in a private transaction suggests that they

may have intended to cover for idiosyncratic needs and hence that
their exchange was not readily comparable to others occurring in the
market. Hence, the compensation paid to the buyer was in the form of
negative-interest damages, designed to bring the buyer back to his
position before the contract and including damages beyond the price
paid. This also shows that the actio ex empto did not have the same
problems observed for the actio quanti minoris because it took into
account each buyer’s private valuation for the good.
This system made negotiations and trials more complex and pos-
sibly more expensive, but it allowed a more effective exchange of
information and hence supported private transactions that allocated
goods ex post to buyers more efficiently than the aedilician markets
did, especially with respect to those buyers who valued idiosyncratic
characteristics.
24.5. FORUM SHOPPING
Contractual parties had an opportunity to choose between prae-

torian and aedilician remedies at two points in time. First, buyers
and sellers could decide whether to trade in a market or through
private transactions. If they opted for a private transaction, they
would then fall under the exclusive jurisdiction of the praetor. If
instead they opted for a market transaction, the buyer could
possibly have a choice between filing a case with the aediles or
with the praetor.
We have shown that the aedilician remedies allowed for less
expensive negotiations and litigation and involved less quality risk
for buyers than the praetorian remedies. Yet, they generated larger
allocative losses ex post due to the fact that idiosyncratic buyers’
interests were left unprotected. Therefore, buyers who valued a
good’s idiosyncratic characteristics (for instance, a buyer looking for
a slave with very specific skills) probably preferred to trade through
private transactions. In contrast, buyers looking for a standard good
and only interested in quality were most likely better off when
purchasing in a market.
The self-selection of different types of buyers into different types of
transactions made the typical transaction that fell under the jurisdic-
tion of the praetor essentially different from the typical transaction

falling under the aediles and reinforced the overall efficiency of the
system. Since the buyers operating in markets were not too interested
in idiosyncratic characteristics, the actual ex post efficiency losses
were likely to be small. Moreover, these buyers were unlikely ex
post to opt for the praetorian remedy, as it involved higher litigation
costs and more quality risk and only a (small, given the types of
buyers) advantage due to the better protection of idiosyncratic inter-
ests. Conversely, buyers interested in idiosyncratic characteristics
were more likely to opt for private transactions and thereby the
higher negotiation and litigation costs balanced large gains due to
the more efficient allocation of goods ex post.
24.6. CONCLUSIONS
In this chapter, we have shown how the aedilician and the prae-
torian remedies for defects in sale contracts addressed the contrac-
tors’ needs to efficiently address the problems created by
asymmetric information. While the aedilician remedies catered to
buyers purely interested in quality and induced quick, standardized
transactions and inexpensive litigation, the remedies applied by the
praetor catered to buyers interested in idiosyncratic features of the
goods for sale and effectively protected those interested at the price
of lengthier negotiations and more expensive litigation. These
remedies slowly converged.
An interesting question is how and why the praetor and the aediles
fostered the economic interests of the parties over whom they had
jurisdiction. The remedies they applied accumulated into an edict
that, for the most part, was passed over from one magistrate to the next.
In both cases, magistrates holding these functions were ascending the
cursus honorum (the customary career progression) and aspired to
higher functions. Since both the praetor and the aediles were in charge
for only one year and were assisted by a council of jurists, they had little
opportunity to entrench themselves in office and had instead clear
incentives to do a good job in order to be subsequently elected to a
higher office. The edict was announced at the beginning of the year and
there was little room for subsequent adjustments. As a result, the
lawmaking effort by both magistrates was directed to the general
interests of contracting parties rather than to special interests

of individual litigants. This might have well been the engine that moved
lawmaking in ancient Roman and produced such remarkable legal
solutions to omnipresent problems.³²
REFERENCES
Akerlof, George A. 1970. “The market for ‘lemons’: quality uncertainty and
the market mechanism.” 84 Quarterly Journal of Economics 488–500.
Andreau, Jean 1974. Les affaires de Monsieur Jucundus. Paris/Rome: Collec-
tion de l’Ecole française de Rome.
Ankum, Hans 1972. “Quelques problèmes concernent les ventes aux
enchères en droit romain classique,” in Studi in onore di Gaetano Scherillo.
Milan: Istituto editoriale cisalpino—La Goliardica, 377–93.
Arangio-Ruiz, Vincenzo 1956 [repr. 1990]. La compravendita in diritto
romano, vols I–II. Naples: Jovene.
Barzel, Yoram, Michael A. Habib, and D. Bruce Johnsen 2006. “Prevention is
better than cure: the role of IPO syndicates in precluding information
acquisition.” 79 Journal of Business 2911–23.
Bove, Lucio 1975. “Rapporti tra dominus auctionis, coactor e emptor in Tab.
Pomp. 27.” 21 Labeo 322–31.
Brooks, Richard R. W., and Alexander Stremitzer. 2011. “Remedies on and
off contract.” 120 Yale Law Journal 690–727.
Brooks, Richard R. W., and Alexander Stremitzer. 2012. “On and off contract
remedies inducing cooperative investments.” 14 American Law and Eco-
nomics Review 488–516.
Burdese, Alberto. 1975. Vendita (diritto romano), in 20 Novissimo Digesto
Italiano. Turin: UTET, 594–600.
Camodeca, Giuseppe. 1999. Tabulae Pompeianae Sulpiciorum. Edizione cri-
tica dell’archivio puteolano dei Sulpici. Rome: Quasar.
Chevreau, Emanuelle. 2012. “L’édit des édiles curules: un droit des marchés
avant la lettre?,” in Laurent Capdetrey et Claire Hasenohr, eds, Agora-
nomes et édiles: institutions des marchés antiques. Bordeaux: Ausonius,
223–33.
de Ligt, Luuk. 1993. Fairs and Markets in the Roman Empire: Economic and
Social Aspects of Periodic Trade in Pre-Industrial Society. Amsterdam:
Gieben.
³² The authors would like to thank two anonymous referees, the participants in the
Forum Romanum seminar at the University of Amsterdam, and the 2013 conferences
of the International Society for New Institutional Economics in Florence and the
Italian Society for Law and Economics in Lugano for helpful comments.

de Ligt, Luuk, and Pieter W. de Neeve. 1988. “Ancient periodic markets:
festivals and fairs.” 66 Athenaeum 391–416.
Decock, Wim, and Jan Hallebeek. 2010. “Precontractual duties to inform in
early modern Scholasticism.” 78 Tijdschrift voor Rechtsgeschiedenis 89–133.
Donadio, Nunzia 2004. La tutela del compratore tra actiones aediliciae e actio
empti. Milan: Giuffrè.
Donadio, Nunzia. 2007. “Azioni edilizie e interdipendenza delle obbligazioni
nell’emptio venditio. Il problema di un giusto equilibrio tra le prestazioni
delle parti,” in Luigi Garofalo, ed., La compravendita e l’interdipendenza
delle obbligazioni nel diritto romano, vol. 2. Padua: Cedam, 457–537.
Donadio, Nunzia. 2011. “Promissio auctionatoris.” 39 Index 524–57.
tions,” in Walter Scheidel, Ian Morris, and Richard Saller, eds, The Cam-
bridge Economic History of the Greco-Roman World. Cambridge:
Hallebeek, Jan. 2009. “The ignorant seller’s liability for latent defects: one
regulator various sets of rules?,” in John W. Cairns and Paul J. du Plessis,
eds, The Creation of the Ius Commune from Casus to Regula. Edinburgh:
Edinburgh University Press, 175–218.
Garofalo, L. 1998. “Redhibitoria actio duplicem habet condemnatio (a pro-
posito di Gai. 1 ad ed. aed. cur. D. 21.1.45),” in Atti del II Convegno sulla
Problematica contrattuale in Diritto Romano (Milano 11–12 maggio 1995).
In onore di Aldo dell’Oro. Milan: LED, 57–86.
Giffard, André E.V. 1931. “L’action édilicienne quanti minoris (D. 21.1.38
pr.; 13 et 14).” 12 Revue historique de droit français et étranger 682–87.
Gröschler, Peter. 1997. Die tabellae-Urkunden aus den pompejanischen und
herkulanensischen Urkundenfunden. Berlin: Duncker & Humblot.
Guarino, Antonio. 1955. “L’editto edilizio e il diritto onorario.” 1 Labeo
295–307.
Guarino, Antonio. 1956. “Ancora sull’editto edilizio.” 2 Labeo 352–7.
Guarino, Antonio. 1995. “Tagliacarte.” 41 Labeo 304–5.
Impallomeni, Giovanni B. 1955. L’editto degli edili curuli. Padua: Cedam.
Jakab, Éva. 1997. Praedicere und cavere beim Marktkauf. Munich: Beck.
Krishna, Vijay. 2002. Auction Theory. San Diego: Academic Press.
Kronman, Anthony T. 1978. “Mistake, disclosure, information, and the law
of contracts.” 7 Journal of Legal Studies 1–34.
Kupisch, Bertold. 2002. “Römische Sachmängelhaftung: ein Beispiel für die öko-
nomische Analyse des Rechts.” 70 Tijdschrift voor Rechtsgeschiedenis 21–54.
Lerouxel, François. 2016. Le marché du crédit dans le monde romain (Égypte
et Campanie). Bibliothèque des Écoles françaises d’Athènes et de Rome
374. Rome: École française de Rome.
Lo Cascio, Elio, ed. 2000. Mercati permanenti e mercati periodici nel mondo
romano. Bari: Edipuglia.

Malmendier, Ulrike M. 2002. Societas publicanorum. Cologne: Böhlau.
Manna, Lorena. 1994. Actio redhibitoria e responsabilità per vizi della cosa
nell’editto de mancipiis vendundis. Milan: Giuffrè.
McAfee, R. Preston, and John McMillan. 1987. “Auctions and bidding.” 25
Journal of Economic Literature 699–738.
Memmer, Michael. 1990. “Der schöne Kauf des guten Sklaven. Zum Sach-
mängelrecht im Syrisch-römischen Rechtsbuch.” 107 Zeitschrift der
Savigny-Stiftung für Rechtsgeschichte, Rom. Abt. 1–45.
Monier, Raimond. 1930. La garantie contre les vices cachés dans la vente
romaine. Paris: Recueil Sirey.
Rolfe, John C., ed. 1927. Gellius. Attic Nights. Loeb Classical Library 195,
3 vols. Cambridge, MA: Harvard University Press.
Serrao, Feliciano. 1959. “Vacanza dell’edilità e competenza dei pretori nel
I sec. a.C.” 62 Bullettino dell’Istituto di diritto romano Vittorio Scialoja
267–71.
Serrao, Feliciano. 2000. “Impresa, mercato, diritto. Riflessioni minime,” in
Elio Lo Cascio, ed., Mercati permanenti e mercati periodici nel mondo
romano. Bari: Edipuglia, 31–67.
Talamanca, Mario. 1990. Istituzioni di diritto romano. Milan: Giuffrè.
Thielmann, Georg, and Bettermann Karl A. 1961. Die römische Privatauk-
tion: zugleich ein Beitrag zum römischen Bankierrecht, ed. Ulrich von
Lübtow. Berlin: Duncker & Humblot.
Thomas, J. A. C. 1966. “Le ‘auctiones’ private.” 12 Labeo 395–9.
Vacca, Letizia. 1994. “Ancora sull’estensione dell’ambito di applicazione
dell’actio empti in età classica.” 45 Iura 35–73.
Volterra, Edoardo. 1955. “Intorno all’editto,” in Scritti in onore di U. Borsi.
Padua: Milani.
Volterra, Edoardo. 1956. “Ancora sull’editto degli edili curuli.” 7 Iura 141–7.
Watson, Alan. 1965. The Law of Obligations in the Later Roman Republic.
Oxford: Clarendon Press.
Watson, Alan. 1987. “Sellers’ liability for defects: aedilician edict and prae-
torian law.” 37 Iura 167–75.
Watson, Alan, ed. 2009. The Digest of Justinian. Philadelphia: University of
Pennsylvania Press.
Wolf, Joseph Georg. 2010. Neue Rechtsurkunden aus Pompeji. Tabulae
Pompeianae novae. Darmstadt: Wissenschaftliche Buchgesellschaft.
Index
accounting 28, 35–9, 49, 52, 54–6, 85, assets 36–7, 57, 69, 72, 102, 117–19, 122,
175, 280 (see also rationes) 124–5, 127, 129, 132, 135, 137–8,
actio (legal remedy) 144, 148–50, 169, 202, 252–4, 256,
adiecticiae qualitatis 97, 278 268, 277–9
empti utilis 144 charged asset 117–18, 121–31, 145,
ex contractu 303 149–51
ex delicto 303 pledged assets 136, 137, 144–5
ex empto 401, 403, 404, 408, 416, Athens (classical) 16, 18
420, 421 auction 88–9, 97, 100–1, 106–7, 128–30,
institoria 95, 278 402, 406, 414, 417, 420
legis 333, 339, 340 Augustan age 13, 14, 17, 19, 31, 41, 45,
quanti minoris 401–3, 406, 407, 409, 69, 82, 93, 133, 179
414, 418, 419, 421 Augustus 20, 21, 47, 87, 192, 198,
redhibitoria 401–3, 405–7, 409, 414, 200, 223
415, 418, 419
Serviana 122, 137 bank/banking 57, 90–5, 97, 102–3, 105,
tributoria 97 107, 124, 175, 263
actus (land measurement) 218, 219 bankruptcy 104, 118
adrogatio 287 bibliotheke enkteseon (property
adverse selection 3, 120, 410, 412, registry) 132, 264, 290
413–4 bona fides 30, 48, 332
aedile 192, 402–6, 408, 410, 414, 415, bonding 3, 5, 119, 121, 285, 337
417, 419, 421, 422 borrowing 55, 70, 83–5, 98–9, 114,
edict of 418, 422 117–20, 125, 128, 145, 260,
curule 162, 401, 402 275–6, 285
plebeian 162 bundle of rights 71, 344
aequum 31 Byzantine period 21
Africanus 38
ager acrifinius 240 Caesar, Julius 196, 199, 223
agriculture 12, 17–18, 41, 116, 140, capital 22, 70, 85, 102, 150, 165, 172,
167, 174, 178, 181, 212, 236, 243, 175, 182, 282
259, 267 capital punishment 331, 334, 340, 343,
agrimensor 223, 225, 237, 241 381, 407
Agrippa, Marcus 198 capitalism 11, 14, 23, 32
American rectangular system 212, 221, Caracalla 91, 93, 142
223–7, 229–33, 237–9, 243 cardo maximus 218, 220, 221, 223
amicitia 113, 119, 134, 283 Cato 58, 82, 89
antichresis 127, 150 cautio praedibus praediisque 133
Antonine period 91 cenacula 162
arbitration 224–5, 328, 333, 337, 339 censor 196
argentaria 90, 91, 97 census 133, 188, 262, 264, 265, 291
argentarii 82, 85, 88–91, 97, 101, 105, centuriation 212, 219, 226, 233–8, 240
107, 129, 406 Cicero 18, 64, 82, 84–6, 89, 93, 101, 165,
receptum 105, 106 171, 179, 202
Aristotle 18, 24 claim priority 122–3, 250, 264, 277
428 Index
Claudius 45, 198 fixed cost 252, 254–5, 282, 290
clientela 284, 285 implementation cost 214, 231,
coactor 89, 406 254–5, 261
argentarius 89, 97, 101 monitoring cost 46, 57, 59, 62, 117,
cognitio 341 121, 226, 288
collateral 120, 127–30, 138, 149–50, 251, variable cost 254–5, 290
276, 285, 322 Crassus 165
collective responsibility 227, 352, credit 3, 48, 88–90, 92, 97–8, 100–1, 105,
379–80, 394 107, 114–16, 119, 121, 124–5, 129,
comitia tributa (tribal assembly) 405 175, 261, 283, 356
commodatum 87 access to credit 117, 120–1, 125, 127,
Commodus 55 128, 143
common law 290, 304–5, 308–9, 315, secured credit 115–16, 121, 123,
327, 343 128–33, 136, 142–5, 148–51
common pool problem 363, 370 creditor 49, 83, 85, 92, 98, 105, 116–24,
compensation 46, 90, 103, 104, 138, 258, 126–31, 135, 137–8, 141, 144–5,
327, 335, 365–6, 371, 383–4, 386, 149–51, 199, 248, 252, 279, 284,
388–9, 393, 404 287, 331, 336–7, 340
concilia plebis (plebeian assembly) 405 crimen 144, 382
consensus 30 criminal law, generally 5, 308, 327, 330,
contracts, generally 5, 26, 49, 70, 87, 92, 334, 342–3, 347–8, 356–8, 369–70,
96, 98–9, 104, 118, 120, 123, 125, 372–3, 380, 382, 387, 390
128, 132, 140, 144, 151, 183, 197, criminal liability 145, 354, 364, 373
227, 247–52, 256, 263, 266–8, 276–8, criminal sanction 5, 276, 329, 341,
280–1, 285–6, 288, 291, 306, 310, 357, 359, 361, 367–8, 370, 372,
318–9, 321–2, 335, 338, 392, 401–4, 380, 382, 390, 393
406, 408, 419–21 murder 311, 328, 334
breach 118, 190, 310–11, 383 cura annonae 402, 405
contract law 302–3, 308, 311, 320, cura ludorum 405
324, 339, 347 cura urbis 405
contract rights 2, 49, 268, 313 cursus honorum 422
contracts of sale 6, 322, 401, 403, 407, custom 13, 94, 142, 145, 187–8, 190–1,
409, 422 193, 198, 237, 239, 258, 322, 328,
freedom of contract 11, 183, 191–2 330, 333–4, 369
performance 30, 56, 66–7, 191, 198,
306, 310–11, 322, 323 damages 6, 56, 60, 190, 193, 285, 313,
contubernalis 54 341–3, 348, 382–3, 385, 388, 393,
conventus 94 401, 408, 416, 420
coordination 2, 56, 193–4, 214, 227–30, punitive 341, 382
232, 233, 235, 238, 320, 354, reliance 401, 408, 418
372, 392 restitution 5, 6, 401, 403, 407, 415–16,
corporation 252, 281–2, 379, 387, 393 418–19
business enterprise 1, 37, 58–9, 86, debt 28, 45, 49, 63–4, 82–5, 92, 96, 98,
88–9, 98, 104–5, 117, 126, 137–8, 104, 106, 120, 122, 124–8, 130, 134,
165, 277, 279–82, 286 138, 150, 161, 182, 190, 199–200,
costs, generally 4, 46, 54, 60, 62, 123–5, 252, 258, 260, 284–5, 306, 320–1,
128, 131, 134, 169, 171–2, 185, 189, 336, 340, 356 (see also reliqua)
194, 199, 225, 229–30, 250–2, debtor 49, 57, 116–19, 121–31, 134–5,
254–6, 261, 265, 271, 280, 289–91, 137–8, 142, 144–5, 149–50, 199,
365, 372–3, 386, 415, 419 256, 284, 331, 337, 356
demarcation cost 225–6, 228–9 default 128, 142, 260, 284, 331, 336
(see also transaction costs) decimation 283
Index 429
decumanus maximus 218, 220, 221, 223, equestrians 82, 83, 84, 85, 86, 105
237, 239 estates
deed 53, 142, 146–7, 251, 264, 266 landed 23, 26, 42, 82–3, 85, 99–100,
delict 5, 308, 382 122, 148, 188, 259
Delos 23 inherited 36, 66, 69, 138, 175, 184–5,
demurrer 304, 308, 321 336, 339
deposit accounts 90–1, 95–6, 104, 105–6 euergetism 198
(see also receptum argentarii) exchange 26, 81, 404, 413–14, 416,
depositum 98 418–21
deterrence 5, 143, 347–50, 354, 356–61, externalities 192–3, 195, 314, 349
363, 367–8, 370–3, 387, 389–91
Diocletian 50, 63, 123 faeneratores 86, 89, 100, 101
disclosure, generally 46, 143–4, 406, 410 familia 285–7, 291
failure to disclose 134, 144, 151, 318, family law 286, 291
405–6, 410, 420 fault 308–9, 316, 382–3
dolus 309, 402, 404, 407, 408, 414, 416, feud system 328, 338–9, 344
417, 419 fideiussor 335, 336 (see also surety)
dominium 190 fiducia 114, 123, 126, 138, 140, 141, 275
dominum ex iure Quiritium 138 finance 3, 46, 81, 84–105, 116, 119–22,
domus 162, 195 131, 149–50, 175, 196–9, 260
due process 303–4 forfeiture 129, 340, 383
duty of care 317, 322, 381, 383 forma censualis 265, 266
formal rules 115, 189, 280, 283–4, 291,
economic growth 11, 23, 119–20, 150, 338, 354
171, 175, 198–9, 243, 247–8, Forum, Roman 167
271, 289 fraud 69, 92, 134, 137, 141, 144, 151,
editio rationum 104, 105 262, 276, 286, 307, 318, 383, 402,
efficiency 2, 4, 5, 121, 123, 131–2, 197, 404, 413
200, 237, 258, 260–2, 271, 274, 290, free–rider problem 50, 352, 363,
302, 348, 355, 360–1, 367–71, 370, 391
373–4, 379, 391, 393–4, 402, freedman (libertus) 20, 40, 47, 54, 64, 65,
411–12, 415, 417, 419–22 69, 83, 165, 381
Egypt orcinus 36, 66
Ptolemaic 134, 263 operae 66–69
Roman 41, 42, 53, 61, 71, 91, 93, 94, furtum 136, 366, 382 (see also theft)
114, 132–4, 143, 148, 259, 261, 263,
264, 267, 268, 291 Gaius (jurist) 25, 95, 123, 178, 188, 302,
elite 15, 23, 30, 41, 43, 82, 84, 86, 90, 100, 304–6, 311
102, 113, 116, 170–1, 174, 183 game theory 349–50, 352, 373
emperor 87, 192, 198 behavior 413, 417
enforceability 6, 50, 56, 71, 127–9, 131, repeat game 166, 350, 352, 373
141, 145, 150, 227, 248–50, 255, gesta municipalia 146, 147
276, 284, 286, 288, 290–1, 306, 309, gift 53, 86–7, 166, 170, 321
321, 323, 327, 329, 335, 340 good faith 30, 37–8, 48, 57, 130, 134–8,
enforcement by law 344, 379, 383–7, 150, 186–7, 250, 273–4, 319, 322,
391, 394 332, 404
private enforcement 5, 308, 334, goods 12, 27, 87–8, 99–100, 119, 139,
338–9, 341–4, 382 161, 169, 173–6, 196–8, 247,
public enforcement 227, 308, 330, 249–51, 256–9, 261, 269, 274, 323,
334, 337, 339–41, 343, 370, 382 401, 410, 412, 414–21
strict enforcement 4, 248, 277, nonconforming goods 401, 403, 405,
289, 291 407, 416, 418–19, 422
430 Index
Gracchi 31 ius respondendi 95
groma 224, 225 iustum pretium 130, 200
guarantee 3, 60, 113, 119, 261, 276, 336,
339, 356 judicial forbearance 280, 285–6, 291,
306, 314
Hammurabi Code 361, 363, 364 Julio–Claudian period 99
Hellenistic period 71, 91, 268 jurist 13–16, 18, 20–1, 23–5, 28, 31–2,
heredium 219 37–9, 45, 52, 57, 60, 94–9, 104,
heres necessarius 45 142, 177, 179, 190–1, 196, 200,
hypothecae 141 214, 224–5, 249–51, 265, 267,
284, 303, 324, 330, 333, 380,
idiosyncrasy 4, 352, 403–4, 408, 410–14, 383, 407, 422
416–18, 420–2 Justinian 21–3, 106, 301, 329, 330, 336,
impetratio domii 129 341, 403
in iure cessio 255, 269, 271 Juvenal 167, 173, 179
incentives 37, 54, 59, 72, 190–1, 195,
230, 247, 249, 280, 286, 288, 339, kind–for–kind constraint 353, 360,
343, 347–8, 362, 364, 367, 373, 367–8, 373
387–8, 393, 403–4, 410, 414, Kingdom, Roman 217, 338
416–18, 422
indemnification 250, 315, 382 labor 17–8, 20, 25, 31, 36, 42, 43, 46, 58,
Industrial Revolution 11, 186, 281, 282 60–3, 66–9, 72, 172, 258, 280
infamia 134, 143, 284, 338 (see also operae)
information 64, 120, 173, 235, 265–6, land, generally 25–6, 57, 85, 100, 132,
273, 284, 403–4, 411–14, 418, 138, 140–2, 145–6, 148, 160, 165,
420–1 167, 170, 180–2, 184–8, 192, 198–9,
information asymmetry 247–51, 256, 212, 239–40, 254, 256, 261, 263–4,
265, 274, 280, 284–5, 316, 403, 409, 266, 269–71, 273–4, 276–7, 290,
410–14, 416, 422 310, 313–15
information costs 119, 131, 172, 185, land demarcation (see centuriation)
350, 362, 370 land records 53, 202, 237, 266
private information 6, 410, 417 land tenure 161, 190–1, 276–7
iniuria 309–11, 324, 382 land value 166–7, 214, 226–32, 240, 250,
innovation 13, 169, 175, 278, 370, 408 271–3
insula 162, 163 landlord 125, 164, 189–192, 195
insurance 182, 200, 362–3, 387, 390, landowner 100, 116, 160, 181, 187, 193,
404, 419 198–9, 235, 248, 266
interest 81–2, 84–6, 88–9, 92, 96–9, 101, lease 25–6, 125, 160, 170, 180–1,
116–17, 119, 135, 408, 421 189–91, 194, 195, 202
interest rate 82, 84–7, 89, 119–20, 125, legal claims, generally 50, 57, 66, 69, 71,
128, 133, 145, 199, 260, 263 97, 149, 195, 250, 265, 304, 328, 337,
investment 44, 58, 62, 98–9, 102, 118–9, 343, 401
165, 175, 190, 232, 235, 248, 250, cause of action 52, 65–6, 106, 118,
281, 350, 352, 373 144, 193, 223, 270, 273, 274, 279,
iurisdictio 402, 405 287, 303, 305, 307–38, 311, 337,
ius civile 31, 92, 115, 126, 403 408–9, 421
ius commune 115 claim elements 303, 308, 322
ius edicendi 405 legal defenses, generally 305, 309–10,
ius gentium 92, 93, 260 316–7, 321
ius honorarium 94, 330 excuse 308–11, 315–16, 318, 323
ius mercatorum 103 justification 308–11, 315–16,
ius Papirianum 332 318–19, 323
Index 431
legal obligation 21, 28–9, 36, 49–50, mancipatio 123, 124, 136, 138–40, 145,
54–6, 66, 129, 265, 277–8, 283, 321, 146, 186–8, 255, 269–74, 287
336–7, 341 mandatarius 277
legal sanction 283–4, 291, 339, 367, 380, mandatum 87
386–7, 389–92, 395, 410, 413 manufacturing 17, 84, 101, 106,
legal status 44, 56, 68, 105, 131, 144–5, 169–170, 173–4
281, 287–8, 334, 338–9 manumission 20, 21, 35–72, 194, 291
legal systems 2, 16, 63, 93, 114, 128–9, by fideicommissum 55, 60, 66, 68
185–6, 190, 211, 250–1, 257, 283–4, emptio suis nummis 55, 66, 69, 71,
301–2, 309, 324, 327, 330, 337, inter vivos 50–2, 67, 68, 71
339–40, 342–4, 347, 351, 357, iteratio 48
364, 368 iusta manumissio 53
leges publicae 95 pactiones pro libertate 49, 50, 54, 56
legislation 13, 63, 92, 160, 200, 251, 262, tax 40
330, 333–4 Marcus Aurelius 55, 332
lending, generally 3, 29, 57, 82–9, 91–2, markets 1, 12, 23, 25, 100, 106, 131, 176,
96, 98, 100, 120, 122, 125, 131, 182, 181, 200, 251–2, 255, 260, 277,
199, 251, 263, 275, 285, 406 401–3, 406, 408, 413, 419–22
loans 81–92, 96, 98–102, 114, 118–19, housing market 189–90, 200, 214,
125, 133, 182, 199–200, 260–1, 263, 226–7, 229–30, 232, 243, 258,
266, 275, 285 261, 271
loan agreements 83–4, 118, 199–200 market price 61, 63, 200, 404, 407–8,
secured loans 3, 83, 117, 119, 132–3 419, 420
Lex Aelia Sentia 42, 45, 64 market value 36, 60, 63, 117, 129
Lex Aquilia 308, 309, 314, 315, 318, 367 Martial 63
Lex Genucia 82 measure–for–measure punishment
lex talionis 342, 348, 354–72 360, 373
liability, generally 5–6, 56, 58, 105, 143, metator 223
145, 151, 190, 225, 277–9, 281, 284, metes and bounds 212, 227–31,
307, 309–11, 313–19, 321–2, 332, 238, 240
348, 354, 362–4, 369–70, 373, 379, monitoring 3, 70, 119, 121, 228, 284,
380, 382–90, 392–8, 404, 407, 414, 362, 387
416–17 moral hazard 3, 118–19, 363, 387
civil liability 5, 143, 225, 354, 372, 382 mortgage 138, 200, 251–2, 256, 258,
collective liability 6, 348, 352–3, 263, 291
379–99 mutuum 87, 96
individual liability 5, 6, 348, 362, 364,
370, 373, 380, 384, 386, 389–90, negatiatores 100
394–5, 397–8 negligence 305, 307, 309–10, 315,
joint and several liability 105, 387 317–19, 322, 367–8, 388
liability rules 190, 307, 369 negotiations 35, 48, 51–2, 55–6, 60, 68,
limited liability 58, 278–9, 383, 404 71–2, 170, 187, 193, 199–200, 414,
noxal liability 6, 383, 392, 407 416–18, 420–2
strict liability 307, 309–10, 314, 317, nemo plus principle 135
319, 321, 383 neo–institutionalism 106
vicarious liability 383–4, 387–8, Nero 20
392–3 network effects 230, 232
limites 218, 220, 239 norms 13, 42, 49, 57, 84, 89, 91–4, 105,
litigation 29, 31, 56, 83, 93, 104, 145, 185, 134, 143, 182–4, 189–90, 279,
237, 273, 281, 286, 303–7, 308, 328, 283–5, 291, 307, 312, 322–3, 348,
333, 338–9, 356, 404, 408–9, 421 351–4, 356, 361–3, 369, 372–3
Livy 81, 92 notice 278, 287, 303, 312
432 Index
nuisance law 193–4 praetor 174, 192, 260, 290, 303, 401–6,
nummularii 97, 106 408–10, 415–17, 420–2
nundinae 100 praetor’s edict 13, 29, 30, 55, 94, 95,
97, 278
obligation (contractual/delict) 6, 30, 84, preferences 132, 349, 352, 372–3, 410,
85, 87, 96, 129, 143, 176, 249, 264, 412, 417, 420
265, 277, 278, 313, 319–24, 336–7, premium 60–61, 387
340–2, 366–7, 382 prescription period 186, 269, 271,
involving slaves and freedmen 2, 21, 273–4, 402, 404, 407–8, 414–16,
28, 36, 38, 49, 50, 54–5, 56, 65–8 419–20
personal obligations, 255–6, 268, 275, principal–agent theory, generally 194,
278, 283–5, 288–91 277, 288
officium 261, 283 agency law 194, 277
omission 307–8, 318 agent 36, 166, 193–5, 227, 247–50,
operae 52, 65–96 (see also labor) 277–80, 285, 288, 327
optimality 24, 126, 272, 281, 356, 359, principal 102, 117, 248–9, 277–9,
368, 395, 402 281, 392
ownership 127, 129–30, 133, 135, 137, Principate 11, 16, 18, 83, 86–88, 93, 95,
139, 146, 148, 165, 166, 181, 186, 98, 99, 106, 114, 150, 162, 174, 177,
193–5, 202, 240, 269–70, 273–4, 179, 181, 195, 217
276–7, 312, 332, 361 prior tempore principle 135
private law 11, 96, 133, 183, 302, 309,
partnership 38, 102, 105, 281 (see also 313, 339
societas) procedural law 5, 22, 56, 95, 102, 187,
paterfamilias 184, 195, 278, 280, 281, 269, 271, 273, 301–3, 330, 333,
283, 285, 286, 288 338–9, 355
patronage 35, 65, 69, 70, 71, 166, 184, production 12, 17, 18, 25, 32, 41, 98,
284, 286 120, 131, 228, 259–60
Pax Romana 174, 179 productivity 23, 46, 72, 212, 226, 228–9,
peculium 27, 35, 37–9, 49, 51, 53, 231, 253, 259
55, 57, 58, 65, 69, 72, 102, 194, promise 306–7, 310, 320–1, 347
279, 280 property, generally 5, 14, 15, 48, 123–4,
pecuniae creditae ius 92 126–7, 130, 132, 134–6, 139–40,
peregrinus 91, 93, 94, 260 143–5, 148–9, 160, 165, 184, 187–8,
perticae 224 192, 232, 235, 242, 247, 251, 255,
pignoris capio 337, 340 257, 265–6, 268–9, 279, 284, 290–1,
pignus 83, 114, 115, 123, 124, 126, 127, 309, 332, 337–8, 340–1, 343, 363,
140, 141 (see also pledge) 366–7, 382
Plautus 53, 82, 89, 90 alienation of property 69, 180,
pleading rules 301–2, 307, 311, 317 182–3, 186
pledge 118, 122–3, 126–7, 131, 136–8, liability rules on 190, 307, 369
141, 143–7, 150–1, 336 (see also private property 160, 181, 183,
pignus) 196, 202
possessory pledge 118, 122, 123, 127, property law 135, 237, 267–8, 302,
141–2, 150 320, 324
pledge rights 122, 125, 144, 147 property ownership 15, 268, 276
Plutarch 98, 165 property rules 190, 249–51, 277, 279
Pompey 23 property registration system 131–4,
possession 130, 135, 136–8, 142, 146, 143, 146–50, 185, 202, 252, 255–6,
186, 269–70, 273–4, 276–7, 279, 261–4, 266–8, 275, 281–2, 291
312–13, 332, 339–40 property transfer (see transfer)
praepositio 102 public property 133, 196
Index 433
prosecution 327, 329, 334, 339, 341–2, resources 160, 165, 171, 224, 240, 243,
344, 367 253–5, 257–9, 279, 301, 362,
protopraxia 132 388, 393
provinces 15, 85, 259, 268 restitution 5, 6, 401, 403, 407, 415–16,
provincial governor 94 418–19
public goods 196–9, 201, 229–30, 235, retaliation 329, 335, 349, 351, 353–5,
258, 290 359–60, 362, 364–6, 369–71,
publica causa 105 373, 385
publicani 197, 281 retributivism 351, 354, 360, 363, 369,
Punic Wars, 241, 260 386, 389
rights, generally 2, 4, 66, 130–1, 135–6,
quaestor 23 139, 142, 192, 212, 214, 243, 247–8,
querela inofficiosi testamenti 184 250, 256, 264, 269, 271, 273, 277,
285, 303, 308, 313, 316, 334, 371
rapina 382 contract rights 2, 49, 268, 313
rationality 24, 32 in personam rights 250, 269, 283, 303,
rationes 37–8, 90, 96, 104, 105 (see also 308, 313
accounting) in rem rights 130–1, 135–6, 248, 250,
real estate 84, 122, 132–3, 142, 147, 150, 264, 268–9, 277, 283, 289, 291, 303,
160, 165, 169, 173, 175, 181, 183, 308, 313–4
185–6, 188–9, 193, 201–2, pledge rights 122, 125, 144, 147
258, 265 property rights 2, 4, 139, 142, 192,
reciprocity 11, 13, 25, 30, 171, 350, 212, 214, 243, 247–8, 256, 264, 271,
353–4, 362, 370 273, 277, 285, 316, 371
registration systems, generally 131–4, security rights 122–4, 131–132,
143, 146–50, 185, 202, 252, 255–6, 148, 150
261–4, 266–8, 275, 281–2, 291 vengeance right 330, 335, 351
of land 53, 202, 237, 266 risk 2, 4, 85, 104–5, 113, 117–21, 138,
public registration 130, 132–3, 145, 164, 168, 185, 188, 191, 195,
251–2, 291 240, 260, 275, 318, 323, 347, 355,
regulation 15, 20, 22, 82, 103–5, 362, 364, 380, 387, 404, 415–7,
195, 260 420–2
reliqua 37, 38 (see also debt) allocation of risk 118, 318, 415–16, 419
remedies 6, 144–5, 184, 250, 278, 280, assumption of risk 305, 317–20, 322
314, 321, 337, 355, 357–8, 368, 371, mitigation of risk 2, 113, 117, 120,
382, 401–6, 408, 410–1, 413–14, 240, 323
416–22
rent 62, 161, 164–5, 167, 170, 188–92, security, generally 38, 72, 113, 116–17,
199–200, 231, 258 120–30, 138–9, 141, 149, 182,
replication 301, 304–6, 323 191–2, 257, 266, 276, 290, 331,
Republic, Roman 339–41
early 161, 186, 202, 217, 330, 338 non–possessory security 114, 125,
late 11, 14, 21, 23, 82, 83, 86, 88–9, 92, 130, 133
95, 97, 98, 106, 150, 160, 162–4, 174, real security 3, 113–21, 124, 128, 143,
177, 179, 181, 195, 199, 217, 255, 149–50, 274–6, 283
278, 312 security law 114, 120–2, 124–5,
res locata 26 128–30, 143–4, 149–50
res mancipi 123, 124, 136, 138, security interest 116–18, 126, 131–3,
140, 285 135, 138, 143, 149, 187
res nec mancipi 123, 141 security rights 122–4, 131–2, 148, 150
res publica (see public property) seizure of security 190–2, 331–2, 337,
respondeat superior 387–8 339–41
434 Index
self–help 191, 284, 286, 330, 332–3, 335, technology 12, 131, 148, 163, 171, 197,
337, 339, 344 231, 256, 281–2, 390
senate 170, 381 tenants 162–4, 189–92, 194–5,
senators 82–4 200, 258
Senatus Consultum Silanianum 287, termini (land boundaries) 225
380, 381 testament 37, 49–52, 55, 64, 66, 68–9,
Senatus Consultum Turpillianum 342 71, 160–1, 184–5 (see also will)
Seneca (Elder) 63, 86, 98, 176 theft 54, 139, 332, 335, 341, 353–5, 357,
Septimius Severus 142 360–1, 366–7, 382, 403, 420
services 17, 35, 52, 66, 68, 87–8, 119, (see also furtum)
160, 169, 173–6, 180, 196–8, 249 Tiberius 179
Servius (Sulpicius Rufus) 21, 23–9, 31 title, generally 130, 136, 146, 149, 185–6,
Severan period 18, 21 188, 235, 249–50, 253–4, 256,
shareholders 248, 252, 279 265–6, 269–71, 273–4, 276–7,
slavery 11, 16–32, 35–72, 83, 95, 97, 102, 285, 287
148, 163, 165, 194, 202, 258, 278, title defect 269, 270, 273, 285
280, 285, 288, 314, 356, 380, 381, title purge 269, 273–4, 276, 285, 287
401, 403, 404, 407 tort law 302–3, 308–10, 312–3, 317,
actiones involving 29–31 319–20, 323–4, 327, 339, 343–4,
in the familia Caesaris 42 348, 356–9, 364, 367–70, 372–3,
revolts 19 382, 387–8
slave–run businesses 59, 194, trade 14, 19, 23, 100, 116, 124, 126, 132,
278, 280 163, 166, 170, 174–5, 182, 211,
statuliberi 47, 51, 52, 55, 66, 69 247–8, 250, 253, 258, 260, 263,
vicarii, 58, 59 268, 277–8, 280, 284, 289–90, 410,
Social War 91, 93 413, 421
societas 102, 105, 281 (see also export 170, 174–5, 180, 260
partnership) long–distance trade 166, 257, 259–60,
socii 102, 105 268, 272
socii (non–Roman) 91, 92 traditio 271–4
solvency 83, 114, 117–18, 121–2, 128, tragedy of the commons 252
131, 145, 250 Trajan 55
specialization 160, 172–3, 176, 211, 248, transaction, generally 2, 3, 6, 22, 25, 26,
250, 256, 258–60, 277 30, 46, 89–90, 95–6, 99–102, 118,
stipulatio 145, 269 120–2, 127, 131, 138, 140–7, 170,
Sulla 381 182, 187–8, 233, 237, 248, 252, 256,
summae honorariae 198 258, 260–1, 263–4, 266–9, 271, 273,
sumptuary laws 198 276, 278, 280, 283, 286–8, 290–1,
supply 23, 161, 171–2, 195, 197, 261 338, 402–4, 411, 413–15, 418–19,
surety 5, 86, 113–14, 261, 275–6, 283, 421–2
335–8 (see also fideiussor) financial transaction 46, 119–20, 122
surplus 60, 122, 128–30, 188, 259 impersonal transaction 4, 119–20,
131, 252–3, 255–7, 259–62, 268,
tabernae 162 282–3, 289–90
tabulae alimentariae 133 market transaction 4, 6, 232, 257, 289,
tabulae communes 133 291, 402, 417
tabulae publicae 133 personal transaction 6, 253, 255–7,
Tacitus 81 261–2, 282–3, 285, 289–91, 402,
takings 313, 332 404, 417, 420–2
tax 14, 36, 40, 42, 44–5, 86, 122–3, 132–4, transaction costs 4, 104–6, 123–4, 131,
146–7, 174, 198–9, 214, 237, 239, 172, 194, 199, 248, 250–2, 264, 280,
257, 259, 265–6, 281–2, 291, 371 289, 291, 365, 415, 419
Index 435
transfer utilitas publica 105
obligations of transferee 321 utility maximization 348–9, 389
of obligations 90, 92, 118, 249–50,
266–7, 269, 274, 364–65 Valentinian period 342
of property rights 116, 123, 126, vindex 337, 339, 340, 356
127, 128, 129, 135–41, 146, 170,
175, 182–4, 187–8, 202–3, 250, waiver 323, 329
261, 266, 269–70, 273, 274, warranty 141, 321
276, 342 wealth 1, 2, 12, 14, 23, 70, 82–3, 87, 121,
Triumvirate, Second 200 160, 164–5, 169–70, 173, 175, 178,
Twelve Tables, Law of 11, 47, 55, 81, 184, 198, 211, 243, 247, 258, 279,
160, 183, 184, 186, 187, 202, 336, 363–4, 370–1, 373
329, 330, 332, 334, 335, 337, will 21, 37, 51–2, 66, 184, 338, 381
355, 366 (see also testament)
witnesses 139–41, 146, 187–8, 263,
Ulpian 24, 27, 30, 49, 56, 95, 99, 179, 269–70, 284, 337, 339
265, 380, 407
usucapio 136, 137, 186, 273, 274, 277 zero–sum game 121, 198
usury 84–5, 260 zoning policies 192

Oxfordstudiesinroman Societyandlaw: General Editors . . C

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Oxfordstudiesinroman Societyandlaw: General Editors . . C

Uploaded by

Copyright:

Available Formats

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

OXFORD STUDIES IN ROMAN

OXFORD STUDIES IN ROMAN SOCIETY AND LAW

The aim of this monograph series is to create an interdisciplinary forum

Roman Law and

In recent years, historians of Roman law and the Roman economy

18.1. Cesena, Italy (44°N, 12°E), showing RS demarcation

14.1. Manumission terms in the Digest (# of fragments). 52

Barbara Abatino, Research Fellow in the Faculty of Law at the

xviii List of Contributors

List of Contributors xix

Rome and the Economics

The ﬁrst volume of this collection enlists economic theory and

2 Geoffrey Parsons Miller

Rome and the Economics of Ancient Law II 3

4 Geoffrey Parsons Miller

Rome and the Economics of Ancient Law II 5

6 Geoffrey Parsons Miller

Rome and the Economics of Ancient Law II 7

Law, Slaves, and Markets in the

The Roman Mediterranean Sea is the historical and natural

13.1. LAW, ECONOMY, AND POWER

The Roman Mediterranean between the end of the second century BC

¹ The economic history scenarios outlined in the text presuppose my work in

Law, Slaves, and Markets in the Roman Imperial System 13

Law, Slaves, and Markets in the Roman Imperial System 15

⁵ See Schiavone (2012: 245).

13.2. THE SLAVE SYSTEM

Law, Slaves, and Markets in the Roman Imperial System 17

⁹ Arist., Politics 1253b–1255a. ¹⁰ Cic., De rep. 3.25.37.

Law, Slaves, and Markets in the Roman Imperial System 19

¹¹ D. 1.1.4 (Ulp. 1 Inst.) and D. 50.17.32 (Ulp. 43 ad Sab.); D.1.5.4.1 (Florentinus 9

Law, Slaves, and Markets in the Roman Imperial System 21

13.3. LEGAL THOUGHT

In Roman jurisprudential texts from the time of Servius Sulpicius

¹⁸ Buckland (1908: v). See also Watson (1987).

Law, Slaves, and Markets in the Roman Imperial System 23

¹⁹ Strabo 14.5.2 (Meineke).

²⁰ Lenel, Pal. I. 38–54. ²¹ Schiavone (2012: 522 n. 22).

Law, Slaves, and Markets in the Roman Imperial System 25

13.5. A DUAL REGIME

Law, Slaves, and Markets in the Roman Imperial System 27

²⁹ D. 15.1.9.2–3 (Ulpian, 29 ad ed.): Peculium autem deducto quod domino debetur

Law, Slaves, and Markets in the Roman Imperial System 29

13.6. PRAETORS AND SLAVES

Another insuperable obstacle imposed by the persistence of the

³² D.14.3.5.2 (Ulp. 28 ad ed.).

Law, Slaves, and Markets in the Roman Imperial System 31

In appearance, it seemed that not so much was required for the

³⁶ Schiavone (2012: 145–53); see also Cardilli (2014).

Law, Slaves, and Markets in the Roman Imperial System 33

The Practice of Manumission

Roman slaves often had to meet expressly negotiated conditions to

In  142, the cavalry trooper Antonius Silvanus drew up his will at

The Practice of Manumission 37

The Practice of Manumission 39

14.3. THE FREQUENCY OF MANUMISSION

It would be amiss to suggest that Cronion or the dispensator Calenus

The Practice of Manumission 41

¹⁹ Gal., Propr. an. 9.13 (Corpus medicorum Graecorum V 4.1.1).

²¹ As much as 67 percent in Rome itself, though regional differences should be kept

The Practice of Manumission 43

14.4. ECONOMIC CONSIDERATIONS