[St Antony’s Series] Enrique Cárdenas, José Antonio Ocampo, Rosemary Thorp (eds.) - An Economic History of Twentieth-Century Latin America_ Volume 1 The Export Age_ The Latin American Economies in the Late Nineteen.pdf

St Antony’s Series
General Editor: Richard Clogg (1999– ), Fellow of St Antony’s College, Oxford
Recent titles include:

Enrique Cárdenas, José Antonio Ocampo and Rosemary Thorp (editors)
AN ECONOMIC HISTORY OF TWENTIETH-CENTURY LATIN AMERICA
Volume 1: The Export Age
Volume 2: Latin America in the 1930s
Volume 3: Industrialization and the State in Latin America
Jennifer G. Mathers
THE RUSSIAN NUCLEAR SHIELD FROM STALIN TO YELTSIN
Marta Dyczok
THE GRAND ALLIANCE AND UKRAINIAN REFUGEES
Mark Brzezinski
THE STRUGGLE FOR CONSTITUTIONALISM IN POLAND
Suke Wolton
LORD HAILEY, THE COLONIAL OFFICE AND THE POLITICS OF RACE AND
EMPIRE IN THE SECOND WORLD WAR
The Loss of White Prestige
Junko Tomaru
THE POSTWAR RAPPROCHEMENT OF MALAYA AND JAPAN, 1945–61
The Roles of Britain and Japan in South-East Asia
Eiichi Motono
CONFLICT AND COOPERATION IN SINO-BRITISH BUSINESS, 1860–1911
The Impact of the Pro-British Commercial Network in Shanghai
Nikolas K. Gvosdev
IMPERIAL POLICIES AND PERSPECTIVES TOWARDS GEORGIA, 1760–1819
Bernardo Kosacoff
CORPORATE STRATEGIES UNDER STRUCTURAL ADJUSTMENT IN ARGENTINA
Responses by Industrial Firms to a New Set of Uncertainties
Ray Takeyh
THE ORIGINS OF THE EISENHOWER DOCTRINE
The US, Britain and Nasser’s Egypt, 1953–57
Derek Hopwood (editor)
ARAB NATION, ARAB NATIONALISM
Judith Clifton
THE POLITICS OF TELECOMMUNICATIONS IN MEXICO
Privatization and State–Labour Relations, 1928–95
Cécile Laborde
PLURALIST THOUGHT AND THE STATE IN BRITAIN AND FRANCE, 1900–25
Craig Brandist and Galin Tihanov (editors)
MATERIALIZING BAKHTIN
C. S. Nicholls
THE HISTORY OF ST ANTONY’S COLLEGE, OXFORD, 1950–2000
Anthony Kirk-Greene
BRITAIN’S IMPERIAL ADMINISTRATORS, 1858–1966
Laila Parsons
THE DRUZE BETWEEN PALESTINE AND ISRAEL, 1947–49
M. K. Flynn
IDEOLOGY, MOBILIZATION AND THE NATION
The Rise of Irish, Basque and Carlist Nationalist Movements in the
Nineteenth and Early Twentieth Centuries
Karina Sonnenberg-Stern
EMANCIPATION AND POVERTY
The Ashkenazi Jews of Amsterdam, 1796–1850
Shane O’Rourke
WARRIORS AND PEASANTS
The Don Cossacks in Late Imperial Russia
Azar Gat
BRITISH ARMOUR THEORY AND THE RISE OF THE PANZER ARM
Revising the Revisionists
St Antony’s Series
Series Standing Order ISBN 978-0-333-71109-5
(outside North America only)
You can receive future titles in this series as they are published by placing a standing order.
Please contact your bookseller or, in case of difficulty, write to us at the address below with
your name and address, the title of the series and the ISBN quoted above.
Customer Services Department, Macmillan Distribution Ltd, Houndmills, Basingstoke,

Hampshire RG21 6XS, England
An Economic History of
Twentieth-Century
Latin America
Volume 1
The Export Age: The Latin American Economies in
the Late Nineteenth and Early Twentieth Centuries
Edited by
Enrique Cárdenas
Professor of Economics and Rector
Universidad de las Américas-Puebla
José Antonio Ocampo

Executive Secretary
United Nations Economic Commission for
Latin America and the Caribbean (CEPAL/ECLAC)
and
Rosemary Thorp
Reader in the Economics of Latin America
St Antony’s College
Oxford
in association with
St Antony’s College, Oxford
Editorial matter, selection and Chapter 1 © Enrique Cárdenas, José Antonio
Ocampo and Rosemary Thorp 2000
Chapter 3 © José Antonio Ocampo and Maria Mercedes Botero 2000
Chapters 2, 4–10 © Palgrave Publishers Ltd 2000
Softcover reprint of the hardcover 1st edition 2000 978-0-333-91304-8
All rights reserved. No reproduction, copy or transmission of

this publication may be made without written permission.
No paragraph of this publication may be reproduced, copied or
transmitted save with written permission or in accordance with
the provisions of the Copyright, Designs and Patents Act 1988,
or under the terms of any licence permitting limited copying
issued by the Copyright Licensing Agency, 90 Tottenham Court
Road, London W1P 0LP.
Any person who does any unauthorised act in relation to this
publication may be liable to criminal prosecution and civil
claims for damages.
The authors have asserted their rights to be identified
as the authors of this work in accordance with the
Copyright, Designs and Patents Act 1988.
First published 2000 by

PALGRAVE
Houndmills, Basingstoke, Hampshire RG21 6XS and
175 Fifth Avenue, New York, N. Y. 10010
Companies and representatives throughout the world
PALGRAVE is the new global academic imprint of
St. Martin’s Press LLC Scholarly and Reference Division and
Palgrave Publishers Ltd (formerly Macmillan Press Ltd).
ISBN 978-1-349-42274-6 ISBN 978-0-230-59965-9 (eBook)
DOI 10.1057/9780230599659
This book is printed on paper suitable for recycling and
made from fully managed and sustained forest sources.
A catalogue record for this book is available
from the British Library.
Library of Congress Cataloging-in-Publication Data
An economic history of twentieth-century Latin America / edited by Enrique
Cárdenas, José Antonio Ocampo, and Rosemary Thorp.
p. cm.
Includes bibliographical references and index.
Contents: v. 1. The export age / edited by Enrique Cárdenas, José Antonio
Ocampo, and Rosemary Thorp — v. 2. Latin America in the 1930s : the
role of the periphery in world crisis / edited by Rosemary Thorp — v. 3.
Industrialization and the State in Latin America : the postwar years /
edited by Enrique Cárdenas, José Antonio Ocampo, and Rosemary Thorp.
ISBN 978-1-349-42274-6 (v. 1 : cloth) — ISBN . 978-0-333-63341-0 (v 2 : cloth) –
ISBN 978-0-333-63342-7 (v. 3 : cloth)
1. Latin America—Economic conditions. 2. Latin America—Economic
conditions—1918– I. Cárdenas, Enrique. II. Ocampo, José Antonio.
III. Thorp, Rosemary.
HC125 .E37365 2000
330.98'0033—dc21
00–040449
10 9 8 7 6 5 4 3 2 1
09 08 07 06 05 04 03 02 01 00
For Carlos DõÂaz Alejandro
This page intentionally left blank
Contents
List of Tables and Figures viii
Foreword to An Economic History of Twentieth-Century Latin America xi
Notes on the Contributors xii
1 Introduction
The Export Age: The Latin American Economies in the Late
Nineteenth and Early Twentieth Centuries 1
Enrique CaÂrdenas, JoseÂ Antonio Ocampo and Rosemary Thorp
2 Brazil as an Export Economy, 1880±1930 32

Marcelo de P. Abreu and Afonso S. Bevilaqua
3 Coffee and the Origins of Modern Economic Development in

Colombia 55
JoseÂ Antonio Ocampo and MarõÂa Mercedes Botero
4 The Economies of Central America, 1860±1940 85

HeÂctor PeÂrez Brignoli
5 Export-led Growth in Mexico, c. 1900±30 119

Alan Knight
6 Peru, 1884±1930: A Beggar Sitting on a Bench of Gold? 152

Paulo Drinot
7 Bolivia, 1900±39: Mining, Railways and Education 188

Manuel E. Contreras
8 Trying to `Tax and Spend' Oneself out of the `Dutch Disease':

The Chilean Economy from the War of the Pacific to the Great
Depression 217
Gabriel Palma
9 The Vicissitudes of an Exporting Economy: Argentina, 1875±1930 265

Roberto CorteÂs Conde
10 Alteration, Crisis and Adjustment in the Cuban Export Economy,

1898±1939 299
Antonio SantamarõÂa GarcõÂa
Index 323
vii
List of Tables and Figures
Tables
1.1 World economic growth, 1820±1929 3

1.2 Growth of Latin American exports, 1860±1929 11
1.3 Commodity composition of Latin American exports 12
1.4 Latin America: economic and social indicators, circa 1913 26
2.1 Brazil: GDP shares 1910±93 33
2.2 Brazil: population data, 1872±1960 34
2.3 Brazil: commodity export shares, 1850±59 to 1950±59 35
2.4 Brazil: foreign capital stock, 1840±1930 40
2.5 Brazil: British direct investment and holdings of sterling-
denominated corporate securities by sector, 1865±1913 40
2.6 Brazil: coffee price regressions, 1852±1930 45
2.7 Brazil: coffee price regressions with tariff, 1880±1930 47
3.1 Nineteenth-century Colombian foreign trade 57
3.2 Coffee production by department, 1874±1932 60
3.3 Exports and imports, 1900±29 68
3.4 Concentration of coffee production, 1923 and 1932 72
3.5 Concentration of production in main departments, 1923 and 1932 73
3.6 Share of net worth of industrial firms in 1945 by date of
establishment 78
3.7 Employment in handicraft and manufacturing activities, 1870
and 1918 80
4.1 Years in which there was no payment of interest on external public
debt 99
4.2 State intervention in banking development: central bank and rural
credit 100
4.3 Money supply per capita in 1939 101
4.4 El Salvador and Costa Rica: composition of imports, 1946 102
4.5 Rates of growth of physical volume of coffee exports and
variability of annual fluctuations 108
4.6 Central America around 1925, selected indicators 114
5.1 Mexico: population 133
5.2 Mexico: federal government spending 133
5.3 Mexico: GDP and federal government spending 133
6.1 Peru: growth rates of exports by volume, 1884±1930 155
6.2 Peru: growth rates of exports by value, 1884±1930 156
viii
List of Tables and Figures ix
6.3 Peru: composition of exports by value, 1890±1930 157

7.1 Bolivia: tin exports in four decades, annual averages, maximum,
minimum and annual rates of growth, 1900±39 191
7.2 Bolivia: exports and imports by country of destination 193
Äo
7.3 Bolivia: detail of average taxes for selected periods in Patin
mines, 1924±39 200
9.1 Argentina: production of cereals, fodder and cotton, 1900±29 271
9.2 Argentina: wheat and cotton 271
9.3 Argentina: cultivated areas of cotton and textile production 272
9.4 Foreign investment in Argentina, 1917 273
9.5 Argentina: variation in the capital stock, 1914±35 276
9.6 Argentina: some items of the balance of payments, 1920±29 277
9.7 Argentina: international migration 278
9.8 Argentina: exports 282
10.1 Basic indicators of the intensification of sugar production,
1913±39 312
10.2 Annual variation of GDP and its causes, 1930±34 313
Figures
1.1 Net barter terms of trade between primary products and

manufacture, 1871±1913 13
2.1 Coffee prices, 1850±1960 35
2.2 Coffee terms of trade, 1850±1960 36
2.3 Rubber prices, 1887±1917 38
3.1 Real external price of coffee 60
3.2 Average collected tariff, 1833±1945 64
3.3 Equivalent ad-valorem tariffs on cotton textiles, 1828±1947 64
4.1 Coffee exports, 1840±1940 88
4.2 Banana exports, 1883±1940 93
4.3 Guatemalan government expenditure, 1883±1935 103
4.4 Costa Rican government expenditure, 1870±1940 104
4.5 Exports, US$ millions, 1926 prices 107
4.6 Annual fluctuation of coffee exports, Guatemala, El Salvador,
Costa Rica, Nicaragua 110-11
4.7 Coffee price in New York 112
8.1 Chile: productivity in the nitrate industry, 1880±1930 225
8.2 Chile: quantum of exports, 1865±1935 226
8.3 Chile: exports, 1865±1935 229
8.4 Chile: terms of trade, 1865±1935 232
8.5 Chile: annual precentage variation of terms of trade, 1865±1935 233
8.6 Chile: real rate of exchange, 1865±1935 244
x List of Tables and Figures

8.9 Chile: exports and imports, 1865±1935 247
8.10 Chile: public expenditure and exports, 1865±1930 248
8.11 Chile: public deficit, 1870±1930 251
9.1 Argentina: GDP adjusted by terms of trade, 1913±39 270
9.2 Argentina: saving trends in public bonds, mortgage bonds, shares
and deposits in savings accounts 274
9.3 Real wages in UK and Argentina, 1885±1938 279
9.4 Real Wages in Argentina and Italy, 1885±1938 279
9.5 Argentina: the rate of protection, 1910±40 283
9.6 Prices in the UK and Argentina 286
10.1 Cuban sugar: output, exports to the USA and price of raw sugar,
1850±1939 304
10.2 Index of sugar output per active mill and effective zafra day,
1901±39 311
Foreword to An Economic History of
Twentieth-Century Latin America
The three volumes of this mini-series reflect a collaborative effort that began in
1982, when a group of economists and economic historians met to consider
the effect of the 1929 Depression on Latin America, and the explanation of the
continent's rapid recovery. Carlos Diaz Alejandro was the leading light of
the group until his premature death in 1985. This initial group effort led to
the study Latin America in the 1930s, published in 1984. Subsequent collabora-
tion was made possible by the inspiration of Enrique V. Iglesias, President of the
Inter-American Development Bank, whose idea it was to fund a major project to
write an economic history of Latin America for the twentieth century, access-
ible to a wide audience but based on the best Latin American scholarship. His
invitation to me to lead the project and write the overview volume gave me the
most stimulating 16 months of my academic career. New friends joined the
group, and we came together in the course of 1997 in a series of workshops
organized to support the project. These workshops resulted in the books pub-
lished here as Volumes 1 and 3. Volume 2 is a reprint of the 1930s study. The
overview volume was published by the IDB in 1998 as Progress, Poverty and
Exclusion: An Economic History of Latin America in the Twentieth Century.
(available in English, Spanish, French and Portuguese from IDB bookstore,
www.iadb.org/pub, and in English through Johns Hopkins University Press,
www.press.jhu.edu).
My thanks for the opportunity to lead this stimulating project go first and
foremost to Enrique Iglesias himself, to the Inter-American Development Bank
and to the European Union for funding. Tim Farmiloe was our enthusiastic and
patient editor. I owe a particular vote of thanks to my co-editors, who made
themselves available against exceptional pressures and were remarkable part-
ners. Margaret Hancox, Catalina Sanint and Elvira Ryan bore the huge burden
of the final editing and processing.
ROSEMARY THORP
Latin American Centre
Oxford University and St Antony's College
xi
Notes on the Contributors
Marcelo de Paiva Abreu is Professor of Economics at the Department of

Economics, Catholic University of Rio de Janeiro, and former Director of the
Department, 1990±97.
Afonso S. Bevilaqua is a member of the Department of Economics, Catholic

University of Rio de Janeiro, with a PhD in Economics from the University of
California, Berkeley.
MarõÂa Mercedes Botero has a Master's degree from the Universidad Nacional
de Colombia and is a doctoral student at the University of London.
Enrique Ca Ârdenas is Professor of Economics and Rector at Universidad de las

AmeÂricas-Puebla, since 1985.
Roberto CorteÂs Conde is Professor, Department of Economics, Universidad

de San AndreÂs, Buenos Aires, Argentina. He is a member of the National
Academy of History, Argentina, and President of the International Association
of Economic History.
Manuel E. Contreras is Dean of the Graduate School of the Universidad

CatoÂlica Boliviana (UCB), La Paz, Bolivia. He wrote this chapter while working
as Project Associate for the Harvard Institute for International Development in
the UCB.
Paulo Drinot is a doctoral candidate in Modern History at St Antony's Col-

lege, University of Oxford.
Alan Knight is Professor of the History of Latin America at the University of

Oxford and a Fellow of St Antony's College.
JoseÂ Antonio Ocampo is currently the Executive Secretary of the United

Nations Economic Commission for Latin America and the Caribbean (CEPAL/
ECLAC). He is a former Finance Minister of Colombia, Director (Minister) of the
National Planning Department of Colombia and Minister of Agriculture of
Colombia. He has also been the Director of research centres in Colombia
(FEDESARROLLO and CEDE).
xii
Notes on the Contributors xiii
Gabriel Palma is Lecturer in Economics at the University of Cambridge.
HeÂctor PeÂrez Brignoli is Professor of History at the University of Costa Rica.
Antonio SantamarõÂa GarcõÂa has a PhD in history, and was a Senior Associ-
ate Member of St Antony's College, Oxford 1997±98. He was formerly Lecturer
at Universidad Carlos III in Madrid and Researcher at the Instituto Universitario
Ortega y Gasset (Universidad Complutense de Madrid).
Rosemary Thorp is Reader in the Economics of Latin America at the Uni-

versity of Oxford, Fellow of St Antony's College and Director of the Latin
American Centre. She is also a member of Queen Elizabeth House, Oxford.
1
Introduction
The Export Age: The Latin American
Economies in the Late Nineteenth and
Early Twentieth Centuries
Enrique CaÂrdenas, JoseÂ Antonio Ocampo and Rosemary Thorp
This book explores the impact on Latin America of the extraordinary trans-
formation in the international economy ± in kind and degree ± that took place
in the half-century or so which preceded the World Depression of the 1930s.
The period covered in the studies varies, however, from country to country,
reflecting the different experiences in the process of integration to the world
economy. The book also studies the internal transformation that the various
economies experienced as a consequence of successive external shocks. Some of
these had rather similar effects throughout: the First World War and the post-
war boom and crash, and the Great Depression of the 1930s. Others did not: the
Baring crisis of the early 1890s, the collapse of coffee prices at the turn of the
century, other events in specific commodity markets, and even the recovery of
the 1920s, the characteristics of which varied considerably from country to
country. Domestic events which affected economic development throughout
this period are also taken into account. The most remarkable of them was,
obviously, the Mexican Revolution.
Although the export sectors were not necessarily the most relevant in terms
of employment or even share in global economic activity, they usually repre-
sented the most dynamic element of the economies and constituted the chan-
nel by which the countries in the region were related to the external world.
With very few exceptions, the export sectors in the Latin American context of
the nineteenth and early twentieth centuries were intimately related to the base
of natural resources, or to specific products that had world demand and were
thus cash commodities. Foreign interests were involved in the trading, trans-
portation and sometimes in the production of these commodities, as in invest-
ment in some domestic economic activities. The rise of a modern international
monetary system contributed to the expansion of trade and the financing of
1
2 Introduction
foreign investments in the region. International migration was intensive and

constituted another means of interrelationship between countries, as migrants
became an active segment of society. The breadth of international trade, invest-
ment and migration varied, however, from country to country, and gave rise to
distinctive national experiences.
States also saw in international trade their most important source of rev-
enues, primarily through tariffs on imports, but also, in a few cases, through
explicit taxation of the dynamic export activities. Access to international capi-
tal markets allowed governments to complement tax revenues, using enhanced
resources to support economic growth, primarily but not only through the
modernization of transportation. Domestic capital accumulation was also
strongly linked to the export sector or the domestic sectors to which it was
associated, including commerce and transportation, domestic agriculture,
urban services and modern manufacturing. National differences were also
important in all these regards.
By way of introduction to the chapters which follow, we first briefly explain
the international context of the export economies. In the sections following we
then resume the themes which are highlighted in the case studies themselves
and which were developed in the discussions at and following the conference
where they were presented. First we review the themes which concern meth-
odology, particularly path dependence and institutional development, fol-
lowed by the issue of preconditions. We then look at export dynamics,
including the determinants of expansion and the structures of export econo-
mies and their consequences. The consequences of export dynamics for domes-
tic linkages are then explored, and in the final section the role of the state in all
these processes is considered, with associated political economy themes.
The development of an international economy
The rapid expansion of world trade was one of the outstanding features of the
acceleration of economic growth in the industrial `centre' in the nineteenth
century. This process continued in the early twentieth century, but experienced
a sharp slowdown from 1914, when the eruption of the First World War
generated a significant break in world political and economic history. At
1990 prices, world exports increased from US$7.3 billions in 1820 to
US$56.2bn. in 1870, US$236.3bn. in 1913 and US$334.4 billions in 1929
(Maddison, 1995, table I-4). Trade expanded at a faster rate than GDP in the
industrial centre and its offshoots in North America and Oceania up to 1913;
both experienced a slowdown in 1913±29, which was, nonetheless, sharper
for trade (Table 1.1). The large share of raw materials was a distinguishing
feature of world trade throughout the period of analysis. The dominance of
Western Europe, particularly of Great Britain, was also a major characteristic
Enrique CaÂrdenas, JoseÂ Antonio Ocampo and Rosemary Thorp 3
Table 1.1 World economic growth, 1820±1929 (%)
1820±70 1870±1913 1913±29
GDP
Western Europe 1.7 2.1 1.5
Western offshoots 4.3 3.9 3.0
Western Europe and offshoots 2.1 2.7 2.2
Latin America 1.5 3.3 3.3
World 1.0 2.1 1.9
World exports in real terms 4.2 3.4 2.2
Real exports of major Latin American countries 3.4 4.0
in real termsa
a
Argentina, Brazil, Chile, Colombia, Mexico, Peru and Venezuela.
Source: Maddison (1995), tables G-2, I-2 and I-4.
up to the First World War, but the balance gradually shifted towards the USA in
the interwar period (Ashworth, 1975; Kenwood and Lougheed, 1971; Maddi-
son, 1991 and 1995).
Behind the significant growth in world trade lay a revolution in technology
and transport, supported by major institutional developments. The break-
throughs in technology, in particular the steam engine and the use of coke
in the production of iron, that were at the heart of the Industrial Revolution
in Britain, had already taken place by the early nineteenth century. The
applications of steam and iron to transport, through railways and steam-
ships, revolutionized transportation, generating major impacts in the integra-
tion of internal markets and, in due course, of the international economy.
However, the spread of these innovations to industry and transportation was
gradual. Sailing ships, for example, were still dominant in world transportation
in the mid-nineteenth century. Many technological landmarks were added
throughout the nineteenth century, including in the last decades of the century
major breakthroughs in steel processing and chemistry and the discovery of
electricity. The internal combustion engine was also to revolutionize trans-
portation and industry in the early decades of the twentieth century. It was
spreading at a rapid rate worldwide in the 1920s, at the end of our period of
analysis.
However, it took the coming together of these technological developments
with institutional changes to generate the growth in production and trade
which was to have such an impact for Latin America as well as almost every
area of the world. The obvious institutional innovations that allowed produc-
tivity to multiply were in the organization of firms, and credit and capital
markets. Joint stock companies, stock exchanges and banks and the accompa-
nying regulatory legislation spread throughout much of Europe and the USA
and eventually worldwide. Landmarks such as the UK Bank Charter Act of 1844
4 Introduction
and the invention of the Registrar General of Companies were the unglamorous
but necessary background to rising prosperity.
Technological developments meant an extraordinary demand for minerals,
including some which up to then had little economic value. The demand for
nitrates, tin, copper and oil was to have a major impact on Latin American
exports. Increasing division of labour and rising incomes meant an increasing
demand for food, including èxotic' tropical products, as well as for agricultural
and forest raw materials. The growth in the demand for temperate products,
such as wool, wheat and meat, motivated the settlement of new lands, includ-
ing Argentina. The growing demand for cotton, henequen and rubber, among
agricultural raw materials, and sugar, coffee, bananas and cocoa, also pene-
trated the continent. Some of these mineral and agricultural raw materials,
which are viewed today as characterized by low-income elasticities, in fact
had the opposite feature during our period of analysis. For example, the
world production of coffee, a typical èxotic' tropical beverage, grew at an
annual rate of 2.7 per cent in 1830±1900, a rate much faster than that of GDP
and GDP per capita in the industrial centre (Ocampo, 1984, ch. 7).
The response of Latin America to the demand for minerals and both temper-
ate and tropical products required movements of capital and labour. Capital
moved primarily from Britain, Germany, France and the USA. It financed the
trading and transportation of commodities, and sometimes their production,
particularly in minerals and bananas. It also financed investments in railroads,
banking and modern public utilities. By 1913, one-fifth of UK capital overseas
was in Latin America, and half of that in Argentina (ECLA, 1965). But capital
also began to move increasingly with no necessary connection to prospective
export expansion. Its general availability responded to developments in the
centre countries. Fluctuations in capital flows were to become one of the most
threatening aspects of the new international economic order.
Furthermore, between 1820 and 1930, 62 million people emigrated from
Europe and Asia, mostly between 1860 and 1914. Of these, 61 per cent went
to the USA and 12 per cent to Canada, but Argentina and Brazil came next with
10 per cent and 7 per cent respectively (Hatton and Williamson, 1994). Chile
and Uruguay also received many newcomers, at least relative to their smaller
populations. Foreign migrant entrepreneurs and technicians, particularly from
Europe, were also important, even in countries which were not recipients of
massive immigration. Tropical and subtropical agriculture needed labour too,
but when it came from abroad it did so mostly under various types of indenture
to take forward plantation-style agriculture throughout the Caribbean, the
Caribbean coasts of Latin America and Peru. Part of these labour flows
were actually within the continent, as black labour from the Antilles moved
to the new banana plantations of Central America (see Chapter 4 in this
volume), the Cuban sugar industry and the construction of the Panama Canal.
Trade and factor movements also needed their institutional support. As

late as the mid-nineteenth century, trade still ran predominantly through
bilateral channels. Under the stimulus of increasing specialization and new
needs and resources, trade became increasingly multilateral, supported by
sterling's role as an international means of payment and reserve currency
and by the stability of the international economy. The gold standard, we
now understand, was the product of that stability more than its cause. Its
spread in the last decades of the nineteenth century generated a unique episode
in world history, in which the rapid growth of world trade and investment
coincided with the stability of the exchange rates of major currencies. This
stability was extraordinarily important in allowing the development of inter-
national financial instruments and channels. Political stability helped too, as
the wars and disturbances between 1870 and 1913 were not enough to disrupt
the emerging international economy. This immediately makes clear why one
author characterized this expansion as `both unique and mortal' (Ashworth,
1975, p. 225).
The collapse of the gold standard during the First World War and the failure
to reintroduce it on a stable basis in the 1920s were major characteristics of the
world economy in the last decade and a half of our period of analysis (Eichen-
green, 1995). Major disruptions in world politics lay behind the slowdown and
increasing disorganization of the international economy. This process, as we
well know, accelerated the rise of the USA as a major player in the world
economy and as a source of capital and technology as the dollar competed
with the pound for world currency hegemony, processes which were already
evident at the turn of the century.
In the light of recent debates, it is important to emphasize that only in Britain
was this expansion supported by enthusiasm for free trade. In the rest of Europe
and in the USA, policies were protectionist. With or without protection, gov-
ernments in the UK and the USA tended to support their private sector's
economic expansion. For Latin America this might involve on some occasions
outright military intervention, undertaken for reasons both economic and
geopolitical; usually, however, it involved the deployment of economic power
within unequal relationships ± such as the USA and Cuba.
We can see, then, how profound the transformation was. However, it
varied considerably in its impact. Geography, natural resource endowments
and domestic factors mattered greatly in the way the transformation im-
pinged. As we will see, the experience of Argentina was profoundly different
from, say, that of Colombia and Central America. Within countries, regional
differences were also remarkable. The importance of migration as well as the
presence of indigenous and previously settled populations mattered greatly, as
did the land tenure system and other institutional heritages from the previous
period.
6 Introduction
A methodological interlude
The analysis of Latin America in the export age brings out interesting meth-
odological issues. Some of them are distinctly economic, and relate basically to
the linkages between the export sector and domestic economic activity. Tradi-
tional `staple theory' and the literature on the `Dutch disease' are important
contributions to this analysis. Other issues are more interdisciplinary in char-
acter. Some trends in economic analysis are relevant in this regard, and the
recent emphasis on institutional development as an essential determinant of
historical performance is particularly useful. Ìnstitutions', it should be remem-
bered, are used in this literature in a broad sense to include not only organiza-
tions, such as firms, producers' associations and state institutes, but also rules of
the game and intangibles like conventions and traditions ± that is, fiscal or
monetary conservatism. Also, the fresh notion of `path dependence' is stimu-
lating, but flesh must be put on this concept. Probably the biggest methodolo-
gical difficulty relates to the interaction between the state and economic
development. In this regard, although economists have abandoned the tradi-
tional approach to the state as exogenous and neutral, the insights on how to
endogenize it are partial and suggestive rather than complete, and in some
cases mechanical and simplistic.
Path dependence turns on two crucial ideas. The first is that the catalysing
role of the export economy operates on an economic and social structure that is
itself complex and changing. The second is that the trajectory which the
economy follows shapes, in turn, its own future. We found many instances in
our case studies of how the impact of external forces was conditioned by prior
history. The role of preexisting rural, social and economic structures is particu-
larly important. As we will see, they largely determined how labour was mobil-
ized to produce for world markets, and to what extent the rising demand for
agricultural goods for domestic consumption was supplied domestically.
A particular case in which path dependence and institutional development
came together was in the development of one of the most important ìnstitu-
tions' of modern capitalist economies, that of a free labour market. As we will
see in our comparative analysis, economic development during the export age
in Latin America was crucially dependent on access to a mobile labour force.
Slavery and indentured labour from abroad had been used by the dynamic
export economies in the past, but these alternatives were no longer available
in the late nineteenth century.
Access to the most developed free labour market in the world at the time, that
of Europe, and thereby to the massive currents of European migration, was thus
crucial to guarantee a dynamic response to the opportunities which the inter-
national economy offered. However, it also determined the ± path-dependent ±
evolution of social institutions and the conflicts which characterized the
economies and societies involved. The use of another pool of free labour, that
generated by the abolition of slavery in the Caribbean, was also important
though more limited, and its effects on the domestic economic and social
developments of the recipient countries totally different.
Countries which did not have access to international flows of labour had to
rely entirely on the domestic evolution of institutions to generate mobile
labour flows. Here, preexisting domestic labour conditions were crucial. As we
will see, how these conditions evolved over time and how labour was mobilized
to the dynamic export sector and its complementary activities largely deter-
mined both economic growth in the export age as well as the social and
political conflicts which characterized the different nations or colonies
involved.
Obviously, in traditional rural societies the development of a labour market
was intimately associated with that of land. How land ownership was managed
was, thus, also crucial in determining how economies evolved. In this regard,
the issue was not whether land was plentiful ± it was in the Latin American
economies ± but rather how institutions developed to eliminate traditional
ìnflexibilities' in the land market, many of them associated with the role of
the Church in traditional agrarian societies, and others to that of Indian com-
munities in those countries and regions in which traditional Indian structures
survived up to the nineteenth century. Equally crucial was the way new land
was granted by the state to private landowners.
Institution building during the period of analysis involved also the develop-
ment of other domestic markets ± for financial services, for foreign exchange,
and so on ± and a modern state. The development of markets and modern
relations between the state and economic activity was crucial in determining
each economy's `capacity for change'. This concept has purely economic
dimensions, such as supply-side elasticity, capacity to absorb technological
transfers from the advanced countries and capacity to innovate. But change
also requires transactions among social actors, and in this regard the relations
that emerged between, for example, business elites and government from a
particular export experience may or may not have proven functional as struc-
tures adjusted to shifting external conditions or internal pressures. The rela-
tively developed capacity for change of Colombia, Mexico and Brazil in the
1930s is illuminated in the next volume, where we analyse the varying speeds
of recovery from the 1929 crash.
The pivot here is the emerging relationship between key export sectors, the
non-export economy and the state, which includes the ability to shift resources
to new exports or domestic economic activities, and the capacity of the state to
develop and implement new policy instruments that empower it to play an
active role in the adaptation of the economy to changing circumstances.
Change, it must be said in this regard, has to be implemented and this is a
8 Introduction
matter of negotiation, but also of the legitimacy and efficiency of public and
private-sector institutions, as well as of the capacity of the private sector to react
in the face of new conditions. The policy machinery and interest-group chan-
nels created during a specific stage of development, say to manage a drop in
commodity prices, may serve as instruments for change when the next shock
comes. But size and the structure of output is important too: shocks require
supply-side flexibility to produce either exportable or domestic alternatives,
and probably prior investment in these alternatives.
The reverse side of the coin might be usefully termed vulnerability. It may be
associated to legitimate lack of alternatives, as the story of the smaller econom-
ies may reveal. But it may be associated to perverse forms of path dependence:
previous success may build up inertia or dynamics which block the adoption of
alternatives when a shock comes. The success of sugar in Cuba in the first
quarter of the twentieth century had quite adverse effects on that nation's
development from the mid-1920s, given extreme dependence and proximity
to the USA: despite a successful restructuring of her sugar industry, her growth
was limited by US protectionism and by the lack of domestic diversification.
Also, the combination of the loss of the dynamism of external markets since
the First World War and the social conflicts which built up simultaneously
explains the vulnerability of the most successful of all Latin American export
economies, Argentina, as CorteÂs Conde argues in his contribution to this
volume (Chapter 9).
The following sections outline the comparative analysis which emerges from
the different country studies included in this volume. They consider the char-
acteristics of export expansion, the linkages of the export sector to the rest of
the economy, the role of the state and some distributive issues.
Preconditions
As the concept of path dependence makes explicit, export expansion in the late
nineteenth and early twentieth centuries built on various preconditions. The
first one is the independent political status of most of Latin America prior to the
export age. The remarkable exceptions are, obviously, Cuba and Puerto Rico,
which continued to be Spanish colonies until the end of the nineteenth cen-
tury, and Panama, which was part of Colombia up to 1903. Major border
disputes were involved throughout the nineteenth century, many of them
continuing up to the present. At least one of them had a major impact on
export development: the War of the Pacific between Chile, Peru and Bolivia
(1879±81). It meant the gain of nitrate deposits for Chile at the cost of Peru.
Whereas this meant the beginning of a spectacular export boom for Chile
(Chapter 6), for Peru it meant a collapse from which it would take more than
a decade to recover.
More important than that, national unity remained problematic in at least

two different ways. First, the construction of a true national unity in the political
sense meant continuous civil strife in many countries, which was associated to
the unsettled development of stable national hegemonies. In economic terms,
the lack of modern communications meant that countries were really a set of
regions, economic archipelagos, in which regions within countries were often
better integrated to the ports of Europe or the USA than among themselves.
Indeed, the initial development of railroads and river steamship navigation
during the nineteenth century sometimes reinforced this fragmentation, as it
improved communications with ports whereas regions in the interior contin-
ued to be linked by traditional transportation. Notable exceptions in this regard
were Mexico and Chile, where policies of domestic integration were a central
element of railroad developments in the last decades of the nineteenth century.
Obviously, the coincidence of the national capital with the major port facil-
itated the integration of Argentina and Uruguay, though not of Peru. In other
countries, as the railroad network was extended and complemented by later
development of motor vehicle transportation, such developments finally
helped to build a true internal market, sometimes rather late in the export
age or even in later phases of development.
In other terms, Latin American nations were truer economic units. Clear
manifestations of such unity were a unique monetary system and legal frame-
work, common national taxes and tariffs with no internal customs, and a
government around which certain economic disputes could be settled ± such
as the level of tariffs and monetary stability. Even in these areas, however, there
were exceptions: mixed currency circulation was typical in border regions,
federalism meant some legal differences within countries, and internal customs
were eliminated somewhat late in some nations.
A second set of preconditions was related to previous economic expansions.
Silver and gold mining were an element of continuity with the colonial past for
Mexico, Peru, Bolivia and Colombia. The new phase of mining exports partially
built on this prior experience. Economic diversification in the late eighteenth
century had also generated new agricultural exports, some of which continued
to be exported throughout the nineteenth century. However, the half century
or so which followed the political events of the 1810s and 1820s was, in general,
a period of stagnation or very slow growth, with a few remarkable exceptions.
The guano boom of Peru, the expansion of coffee in Rio de Janeiro in Brazil and
in Costa Rica, and the sugar boom of Cuba from the 1790s to the 1860s are
some of these exceptions. Curiously, these early examples of export expansion
continued to depend on slave labour in Brazil and Cuba. This helps to explain
why these two regions kept the illegal slave trade despite British opposition and
repression at sea, and were the last two regions of the world to abolish slavery ±
in 1888 and 1886, respectively.
10 Introduction
A third group of preconditions relates to preexisting rural structures. Wage

and, more generally, mobile labour were very scarce. Traditional social and
economic relations involved restrictions on labour mobility in Indian com-
munities and traditional haciendas. Coercive means to mobilize Indian
labour for mining and commercial agriculture and to keep tenants on haciendas
had a long history, going back to the early colonial period. Slave labour
had been the traditional way to generate labour mobility, but under British
pressure it was eliminated prior to the export age in most of the region
and finally disappeared throughout Latin America in the 1880s. Pockets of
peasant communities existed in most countries, though their relative im-
portance was constrained by land concentration, the source of economic,
social and political power in traditional societies. Land concentration was
crucial to guarantee that powerful landowners could dominate the scarce avail-
able labour force. Restrictions on land transfers were also common in the early
nineteenth century. They included Indian reserves and other communal lands,
particularly lands around urban areas, Church properties and debts to the
Church.
The liberal reforms of the nineteenth century, which encompassed a long
period from the 1820s to the 1880s and varied in speed and intensity from
country to country, focused on eliminating restrictions to land mobility, on the
abolition of slavery and on the elimination of state monopolies. Land grants
from the state were also a major feature throughout the nineteenth century.
With few exceptions, they reproduced an agrarian system based on large land-
holdings; these were sometimes used to attract foreign migrants, with rather
limited success. The struggle by domestic peasants sometimes led to the divi-
sion of landholdings among medium and smallholders. A remarkable case of
this took place in the regions of central Colombia where a coffee economy
sprang up in the early twentieth century.
Export dynamics
Major features of export expansion
Export expansions varied considerably throughout Latin America in their mag-
nitude, timing, stability and product composition. The `commodity lottery', to
borrow a concept used by Carlos DõÂaz-Alejandro, was the joint effect of all these
factors and the major determinant of domestic economic developments.
Indeed, most economic expansions in Latin America up to the 1920s were
èxport led', in the sense that export growth was faster than and determined
the cyclical patterns of global economic growth. As we will see in the next
section, diversification of economic activity varied considerably from country
to country, depending on the nature of the linkages of the export sector,
domestic policies and size. However, in only one case, Brazil in the first three
Table 1.2 Growth of Latin American exports, 1860±1929 (US$ millions at current prices
and exchange rates)
1859±61 1899±1901 1911±13 1927±29
Value of exports
Latin America a 292 664 1493 2954
Argentina 13 163 437 964
excluding Argentina 279 501 1055 1989
Annual growth
Latin America 2.1% 7.0% 4.4%
Argentina 6.5% 8.6% 5.1%
excluding Argentina 1.5% 6.4% 4.0%
Value of world trade 5132b 9299 17345 32148
Annual growth 2.0% 5.3% 3.9%
Share of third world exports 41.8% 37.4% 38.4% 36.4%
Share of world trade 7.1% 8.6% 9.2%
a
Data refer to developing countries of America excluding the British Antilles.
b
1870.
Sources: Latin American and third-world exports, Bairoch and Etemad (1985), table 5.1; value of
world trade, Maddison (1995), table I-3.
decades of the twentieth century, did GDP growth exceed export growth by a
considerable margin.
Over the long term, export expansion in Latin America was more or less
similar to the growth of world trade and slower than in Asia and Africa in the
last decades of the nineteenth century, but faster than world trade and similar
to other `peripheral' regions in the first three decades of the twentieth century
(Table 1.2, and Bairoch and Etemad, 1985). Indeed, the region as a whole did
not experience the slowdown which characterized world trade between the
First World War and the onset of the Great Depression. In fact, real export
growth may have accelerated (Table 1.1). By the late 1920s, Latin America's
share in world trade had reached 9 per cent, an increase over the 7 per cent
typical in the last decades of the nineteenth century (Tables 1.1 and 1.2).
Growth was accompanied by major variations in the composition of regional
exports (Table 1.3). Prior to the era of export-led growth, exports were domin-
ated by sugar, coffee, precious metals, natural fertilizers (guano), tobacco and
hides. Except for coffee, which continued to represent close to a fifth of Latin
American exports, they experienced a long-term relative decline in the export
age. Export growth in the late nineteenth and early twentieth century was
characterized by the relative growth of exports of temperate zone products ±
cereals, wool and meat ± and minerals ± nitrates, tin, copper and oil. Most of
these products experienced a steady increase; the two exceptions were wool and
nitrates which reached their peak share in Latin American exports just before
the end of the nineteenth century and the First World War, respectively. Also, at
12 Introduction
Table 1.3 Commodity composition of Latin American exportsa (percentage shares of

total exports, unclassified products excluded)
Products 1859±61 1899±1901 1911±13 1927±29
Meat 1.1 4.1 5.0 6.7

Cereals 0.9 8.8 13.6 17.0
Sugar 26.6 10.1 9.7 10.6
Coffee 18.2 18.5 18.6 18.0
Tobacco 7.8 5.5 3.0 1.7
Leather 6.8 6.5 6.0 4.2
Rubber 6.4 5.8 0.4
Wool 1.9 9.3 5.8 4.0
Natural fertiliser (guano) 8.3 0.0 0.1 0.0
Mineral fertiliser (nitrates) 0.7 6.2 7.6 3.9
Oil 0.1 2.1 7.6
Non-ferrous metals (copper and tin) 0.2 1.1 2.6 6.6
Precious Metals 9.8 8.0 5.3 2.7
Others 17.6 15.4 15.0 16.6
Total 100.0 100.0 100.0 100.0

Unclassified products 11.6 7.8 6.1 5.2
a
Includes the British Antilles.
Source: Bairoch and Etemad (1985), table 5.3.
the beginning of the twentieth century there was a brief but intense rubber
boom, which was followed by rapid decline in the 1920s.
Export booms did not always lead to steady growth. A first factor which lay
behind this pattern was the failure to develop a stable export base. The most
spectacular case was that of the forest rubber boom which countries in the
Amazon basin experienced in the late nineteenth and early twentieth centuries,
which was followed by collapse when plantations in the east were developed on
a large scale in the 1920s. On a smaller scale, reliance on an unstable export
base was quite common in Colombia and Central America prior to the large-
scale development of coffee (Chapters 3 and 4). The exhaustion of the natural
resource on which the boom depended was also an important factor. On a large
scale, this was the case with oil in Mexico in the 1920s, but it was quite a
common phenomenon in mining countries, which depended on the discovery
and development of new resources to keep up the pace of expansion. Due to an
inadequate management of disease controls, banana production was also char-
acterized in the first decades of the twentieth century by continuous movement
of the production frontier (Chapter 4).
Substitution by new products was also significant. The most important was
the substitution of guano by mineral nitrates in the 1870s, and of the latter
by chemical fertilizers from the 1910s. Substitution by beet sugar, largely
facilitated by protectionism, affected all sugar-cane economies in the last dec-

ades of the nineteenth century. Protectionism was again a major factor for the
sugar economies, primarily Cuba, in the 1920s and 1930s, and explains why
Cuba experienced one of the earliest and sharpest depressions during these
years (Chapter 10). It was also a crucial determinant of the slowdown in the
growth of markets for temperate zone products since the First World War, as the
Argentinean experience indicates (Chapter 9).
Geographical trade and investment links were also important. Being depend-
ent on the UK rather than the USA in 1914 meant that war and UK economic
stagnation in the 1920s had stronger consequences ± Argentina versus Mexico.
The proximity of the north of Latin America to the USA had an overwhelming
influence, not only in economic but also in political terms. This had contra-
dictory effects: strong dependence on the USA facilitated massive inflows of
capital and rapid growth of sugar markets for Cuba in the first two decades of
the century, but also the dramatic effects of USA protectionism in the 1920s.
Membership of informal empires constrained and differentiated countries even
more vividly once the shock of the 1930s came (see the next volume, especially
the chapters by Abreu and O'Connell).
Aside from the aforementioned factors, the export economies were subject ±
as they are today ± to significant terms-of-trade shocks. Figure 1.1 summarizes
the global evolution of terms of trade for primary commodities from the 1870s
to the 1930s. A deterioration in the last three decades of the nineteenth century
was followed by an improvement in the early part of the twentieth century.
These trends are an important determinant of the evolution of export values
140
130 Grilli and Yang (1988)

Lewis (1952)
120
(Index, 1913 = 100)
110
100
90
80
70
60
50
1870
1873
1876
1879
1882
1885
1888
1891
1894
1897
1900
1903
1906
1909
1912
1915
1918
1921
1924
1927
1930
1933
1936
1939
Figure 1.1 Net barter terms of trade between primary products and manufacture, 1871±
1913
14 Introduction
reported in Table 1.2. Individual commodities often experienced quite different

patterns. Coffee, for example, faced rising real prices in the last decades of the
nineteenth century followed by collapse at the turn of the century when the
development of coffee in the SaÄo Paulo region brought massive overproduc-
tion. From the First World War, commodity prices became extremely unstable,
as Figure 1.1 indicates: falling prices at the beginning of the war were followed
by an extremely sharp boom-bust cycle at the end of the war and in the early
postwar period, rising prices in the 1920s and collapse in the early 1930s.
In three cases, specific Latin American countries had a dominant position in
world markets: guano in Peru, mineral nitrates in Chile and coffee in Brazil. In
the first two cases this facilitated high taxation of export commodities (see, on
guano, Hunt, 1985, and on nitrates, Chapter 6). In Brazil, this led to the earliest
attempt to manage a commodity market (see below). Also, as the chapter by
Abreu and Bevilaqua in this volume shows, this monopolistic position was
exploited in yet another way: it allowed the cost of Brazilian protectionism to
be paid by foreigners through higher international coffee prices (Chapter 2).
The cost for Brazil of this policy was obviously the long-term loss of market
share in the coffee market, a phenomenon which had interesting spillover
effects on other Latin American countries, particularly the growth of coffee
in Colombia.
Changing export composition also reflected a quite diverse regional perform-
ance. By far the most spectacular growth process was that experienced by
Argentina from the 1870s to the First World War, based on temperate zone
products, massive migration and large inflows of foreign capital. As a result of
this process, Argentine exports increased from 4.5 per cent of Latin American
exports around 1860 to 29.3 per cent in 1911±13 and 32.7 per cent just before
the Great Depression. Indeed, excluding Argentina, Latin American export
growth in the last decades of the nineteenth century was quite slow, as Table
1.2 indicates. This fits Bulmer-Thomas' picture of how disappointing export
expansion was up to the First World War, if Argentina is excluded from the story
(Bulmer-Thomas, 1994, Chapter 3).
Export growth in other countries and colonies was slower, less steady or came
late in the export age. The exports of Uruguay had many similarities with those
with Argentina, and also experienced fairly steady growth up to the First World
War, but export growth was only similar to the continental average. Chile
underwent an impressive boom of nitrate exports in the 1880s, following the
War of the Pacific, an expansion which continued at a slower rate up to the First
World War, after which exports experienced a sharp decline; they were
replaced, however, by growing copper exports in the 1920s (Chapter 6).
After a long period of stagnation, Mexico experienced rapid export growth
during the Porfiriato (1876±1910). Oil sustained export dynamism even during
the peak of the Revolution, but then experienced an early depression in the
1920s (Chapter 5). Brazilian coffee export growth in the last two decades of the
twentieth century was actually followed by very slow aggregate export growth
in the first three decades of the twentieth century. Thus, whereas the export
quantum increased at an annual rate of 5.5 per cent in the last two decades of
the nineteenth century, export growth was rather slow ± 1.7 per cent a year ± in
the first three decades of the twentieth century (Chapter 2). Coffee expansion
was also rapid in the last two decades of the nineteenth century in Central
America, particularly El Salvador and Guatemala, and in Colombia (Chapters 3
and 4). This period also saw the rise of cocoa exports from Ecuador, which
continued to grow up to the First World War (Chiriboga, 1988).
The first three decades of the twentieth century added several late but rapid
export booms in medium-sized countries. They included Peru, based on a
diversified export base of sugar, cotton, copper and oil (Chapter 8); Cuba,
based on sugar (Chapter 10); Bolivia, based on tin (Chapter 7); Colombia,
based on coffee (Chapter 3); and the latest but most spectacular, the oil boom
which started in Venezuela in the 1920s and continued at very rapid rates in the
following decades. Among smaller countries, banana exports were responsible
for export growth during the first three decades of the twentieth century in
Costa Rica, Panama and Honduras (Chapter 4).
The extent of export expansion relative to the domestic economies can be
best measured by the value of exports per capita at the end of the export age. At
the outbreak of the First World War, a definite pattern of regional disparities
had been built (see Table 1.4). The three Southern Cone countries and Cuba had
then the highest exports per capita in the region, which exceeded by a con-
siderable margin those of other countries. Also, within Central America,
exports per capita in Costa Rica exceeded by a considerable margin those of
other countries (Chapter 4). This picture continued with only slight differences
in the late 1920s, basically associated with the booms of the medium-sized
economies. As we will see below, as a reflection of the dominant role of exports,
this pattern was reproduced in many other economic indicators. Some major
regional differences are, thus, a legacy of relative success during the export age.
Nonetheless, relative success from the 1930s onwards was to have a pattern
quite different from that of the export age (see the next two volumes in this
series).
The productive structures of export sectors

The nature of the commodities sold abroad, interactions of export expansions
with domestic economic and social preconditions, and the ways such expan-
sions were supported ± or not ± by foreign capital and labour determined the
productive structure and the typical distributive issues and conflicts which
surrounded the export sectors. Agriculture was more land and labour intensive
than mining. Conflicts which surrounded the agricultural export economies
16 Introduction
were thus generally associated with labour issues ± which, as we have seen, were
closely intertwined with land distribution issues. On the other hand, mining
was capital and technology intensive. In this case, the major distributive issues
were associated with the dominant role of a few large national and interna-
tional firms and the control over rents from non-renewable resources.
Different products also have different processing and transportation needs,
and such needs determined the extent of domestic forward linkages. Mineral
ores must generally be processed close to the point of production to minimize
transportation costs. Thus, they generally required the development of smelt-
ing and sometimes refining facilities, which lay behind the early industrializa-
tion efforts of mining economies. This is not true of oil, which does not require
processing before it is exported but needs a special, capital-intensive, trans-
portation network ± pipelines and specific port facilities.
As a rule, agricultural goods also require processing and some of them special
transportation facilities. Sugar shares with mining ores the need to be processed
close to the production of its raw material, a fact which accounts for the
impressive development of modern sugar manufacturing in Cuba since the
nineteenth century. Meat exports are also a remarkable case in this regard,
particularly when refrigerated transportation abroad became widespread in
the early twentieth century. They required the development of packing houses
in the export economies, a major factor behind early industrialization in
Argentina. Modern cotton gins are also crucial for the development of stable
exports of this product, as the experience of Peru since the 1890s indicates. In
other cases, agricultural processing was simpler. Such is the case with coffee,
although processing facilities were, nonetheless, an essential element of early
industrialization efforts in coffee producing countries. By contrast, bananas
require no processing before they are exported but need access to a quick,
capital-intensive transportation network, a factor which lay behind the early
penetration of foreign capital in their production.
Processing requirements may turn the export sector into a major instrument
of technological transmission and diffusion. Where industrial processing was
required, the export sector became a channel for the diffusion of foreign
capital-intensive technologies and the training of skilled labour. Processing
was also a major mechanism for technology diffusion in coffee economies,
particularly when part of the processing was done directly by smallholders, as
was the case of Colombia.
Agricultural exports also directly involved mechanisms of technological inno-
vation. More intensive land use was a major characteristic of agricultural export
developments in Argentina in the late nineteenth and early twentieth century
(Chapter 9). Also, the introduction of new varieties of cotton was crucial for that
sector's development in Peru in the 1910s, just as new varieties of sugar were
crucial for the rapid growth of that sector in Cuba. The introduction of coffee in
a slash-and-burn agriculture was itself a major technological innovation, as it

allowed an intensive and permanent rather than an extensive and shifting use of
land. On the other hand, the story of rubber is a remarkable example of lack of
innovations, as forest exploitation was not followed by plantation agriculture,
except for some isolated experiences. On the contrary, systematic research
efforts by European governments were required by the successful development
of rubber plantations in their colonies in the East.
The nature of production, processing and transportation largely determined
the industrial structure of the export sectors. Large-scale capital penetrated
those activities where fixed capital and economies of scale were more import-
ant. This was the case of mineral exploitation and banana plantations. Foreign
capital played a dominant role in both. There were obviously exceptions to the
leading role of foreign capital, the most notable of which was the Patin Äo tin
empire in Bolivia. In other cases, large-scale capital was usually happy to con-
trol marketing and processing, but did not reach the raw material production
stage.
The nature of industrial concentration was not always dictated by technolo-
gical imperatives, however. The contrast between the large scale-coffee planta-
tion which developed in most Latin American countries, despite the lack of any
economies of scale in production, and the small and medium-sized farms of
Costa Rica and western Colombia, is one of the most remarkable cases in this
regard. It indicates to what extent other social factors ± land concentration to
guarantee the control of the labour force ± played a crucial role. The restrictions
imposed on Brazil to avoid migrants from acquiring land is also a notable
example in this regard (Chapter 2).
Labour relations varied considerably from country to country. Access to the
free labour market of Europe was essential to determine the particular pattern of
development of the Southern Cone countries. It facilitated rapid economic
growth based on a modern labour market and higher standards of living than
in the rest of region, but also the early rise of social conflicts associated with
strong labour unions. As the chapter by CorteÂs Conde in this volume indicates,
the last phases of the export age in Argentina may be viewed as characterized by
the building up of social conflicts which would be a major source of difficulties
in later phases of development in that country (Chapter 9). The fact that the
export products were at the same time major items in domestic consumption,
also generated a particular modus operandi: fluctuations in exports could be
accommodated by more (less) sales to the domestic market at some loss of
profitability and the gain (loss) of consumers.
The access to the European labour market was only partly captured in coffee
development in Brazil, where it actually became part of a mixed strategy which
maintained some of the old elements of land and labour control. This case
may thus be considered an intermediate form of labour relations. Another
18 Introduction
interesting intermediate case was that of migration of black workers from the
Caribbean to the banana plantations of Central America and the sugar planta-
tions in eastern Cuba, a response to the limited local supply of mobile labour
using up the emerging pockets of free labour markets in the region. Here, in
addition to traditional labour conflicts, frictions also built up with domestic
residents, with racial connotations.
As we saw, although slavery was abolished very late in those countries which
relied heavily on it for their export and domestic activities, no country or
colony could count on it for export expansion even before the middle of
the nineteenth century. Chinese-indentured labour was then imported to
work in Cuban sugar and in Peruvian sugar and cotton plantations, but this
mechanism of labour mobilization ceased to be available by the end of the
century.
Thus, in many countries, labour for export and complementary activities had
to be recruited domestically. This form of labour mobilization was many times
suboptimal, leading to a de facto model of economic growth limited by the
supply of `mobile' labour, a factor which was emphasized by many entrepren-
eurs at the time. In this regard, the stories are strongly regional in character, as
there were often substantial differences between regions within a specific
country. In some areas, a modern labour market of some sort had started to
build up and its development may have been accelerated by the liberal reforms
of the nineteenth century. Population pressures, generally, and pockets of
surplus population in peasant economies, in particular, also facilitated this
development. These `free' workers could be mobilized as temporary or perman-
ent wage labourers or, more frequently in export agriculture at the time, as
tenants. In this case, non-economic restrictions on labour mobility ± including
debt peonage ± were often involved. In the opposite case, labour mobilization
mixed cash incentives with outright coercion. This was generally the rule when
mobilization of the Indian population was involved. Frequently, this system
evolved into a regular flow of labour of the first type. The contrast between
these opposite forms of labour mobilization was present in Mexico, as the
development patterns of the North and the central highlands, on the one
hand, and the South, on the other, indicates (Chapter 5). It was also present,
in Central America, in the contrasting development of coffee plantations of El
Salvador and Guatemala, respectively (Chapter 4).
Either because exploitation tended to increase, as argued by Knight in his
chapter on Mexico (Chapter 5), or because the increasing demand for labour
actually increased the opportunities for workers and restricted the ability of
landlords to exploit them ± as argued, for instance, by Palacios (1983) for the
coffee haciendas in eastern Colombia ± major social conflicts erupted. They
sometimes interacted with those arising from the illegal usurpation of state or
communal lands. The most important of them was obviously the Mexican
Revolution. The peasant uprisings of Peru in the 1910s and in Colombia in the
1920s and 1930s followed a similar pattern. Obviously, conflicts of this sort had
major political implications. Alliances could be built up with labour unions,
with Indian communities in their traditional revindication for land, or with
peasants involved in land conflicts in frontier regions, for example, and these
alliances could be used by emerging new political forces. The most remarkable
case was, again, Mexico, where the Revolution led to major change in social and
political relations.
There was, of course, another possible source of domestic labour supply:
peasant labour. Colombia in the early twentieth century demonstrated how a
significant pool of peasant labour could be used to produce for the interna-
tional market. Similar though more restricted phenomena were experienced in
the expansion of coffee in Costa Rica from the early Republican period
onwards, and in the Cuban tobacco sector. It also played a role in the supply
of food for the cities or for the export sector, as for instance in supplementary
activities to copper mining in the Central Peruvian Sierra. For the region as a
whole, however, this form of labour mobilization was rather limited in scope.
The control of rents was also a major source of disputes and outright con-
flicts. In a sense, all export expansion in Latin America at the time was based on
the exploitation of previously underutilized resources, an idea that fits neatly
with the old `vent for surplus' trade models inspired by Adam Smith (Myint,
1958). There were indeed significant rents to be captured. In agriculture, the
capturing of rents by producing firms was seen as `natural'. There was, however,
an indirect way to capture them for domestic agricultural and industrial devel-
opment: protection, which increased the terms of trade at which export
incomes could be converted into other goods.
On the contrary, in mineral economies the issue of control over rents was
always in the forefront of domestic political discussions, particularly as they
necessarily involved the negotiation of rights to the exploitation of specific
mineral deposits, generally with fairly large firms. The willingness to attract
new companies by giving up royalties, taxation of mineral exports or direct
taxation of profits was one strategy; the alternative route generally involved the
taxation of exports and/or profits. The strategy followed determined the
returned value from those exports and the linkages to the rest of the economy
(see below).
Conflicts in export economies were magnified in the case of forest exploita-
tion, when they easily became very violent, as they generally involved at the
same time the extraction of rents, the forceful mobilization of labour ± includ-
ing the return to outright slavery ± and the lack of rule of law typical of frontier
regions. The history of rubber exploitation in all countries of the Amazon basin
is the clearest manifestation of how violent export economies could become
when all these conditions were mixed.
20 Introduction
Domestic linkages
The links between the export sector and other activities in the Latin American
economies involved three major issues. The first was the way the surplus of the
export sector was taxed and how the resulting tax proceeds were used domes-
tically. The second was the macroeconomic instability associated to the pro-
pensity of export economies to face strong boom and bust cycles, which could
be compounded by the instability of foreign lending. The third were the
forward and backward linkages generated by export activities and how they
built up ± or not ± the basis for a more diversified economy. As we will see, this
involved other macroeconomic issues, particularly those emphasized by tradi-
tional `staple' theory on the one hand, and the `Dutch disease' literature on the
other.
A major issue in relation to the export economy model was the concept of the
`returned value' of exports, namely the share of gross output that remained
within national boundaries and therefore had a linkage effect with the domes-
tic economy. This percentage depended on the nature of the product and on
the capacity of the state to effectively tax the producer and extract part of the
rent. The latter point becomes crucial for economic reasons, since such taxation
is usually a major means by which a nation captures the rents from its mineral
resources, and because it exposes the interplay between the state and large
investors, often foreigners. The political economy of this interrelationship
usually shows the means and ways by which external economic and political
interference occurred. Taxing of tin mining in Bolivia, nitrate and copper in
Chile, and silver and oil in Mexico are notable examples, as are the absence of
similar efforts in Peru in the post-guano era and the lack of any taxation of
banana exports in Central America and Colombia. As we pointed out, in the
cases of Chilean nitrates and Peruvian guano, export taxation was facilitated by
a high world market share.
As we pointed out above, domestic protection was another major way of
indirectly taxing export activities, as it thereby reduced the real income of
exporters. It was actively used by several countries and, as we will see, partly
determined the linkages of the export sector to the rest of the economy. In the
case of Brazil, as Chapter 2 in this volume shows, protection was partly trans-
ferred to world coffee prices and thus paid by coffee consumers worldwide. It
also redistributed income from those consumers to coffee producers outside
Brazil, notably Colombia and Central America.
Depending on the commodity lottery, including the degree of diversity,
countries had varying tendencies to boom and bust with resulting variances
in the degree of instability. The variability of Chile's terms of trade was so high
that it was equivalent on average to a shock on GDP of 1.5 per cent a year
during 1865±95, and 3.5 per cent a year during 1886±1929. Three major
macroeconomic consequences were the instability of fiscal revenues, given the

strong dependence on import tariffs, of foreign exchange and the money
supply. It is perhaps unsurprising that Latin America generated extensive
experiences of currency inconvertibility during the export age (see below).
Additional macroeconomic consequences were strong cycles in a country's
capacity to import, domestic investment and GDP, as well as micro effects
such as the impact on attitudes and the investment climate.
The instability of foreign lending added a component to this cyclical pattern.
After the debt boom which characterized the Independence War period and the
generalized moratoria which followed, access to foreign lending remained
problematic. For some countries, the nineteenth-century story was thus one
of rescheduling, short periods of access to capital markets and renewed
moratoria and rescheduling. Those which had more stable access were in any
case subject to waves of lending which are not, or not solely, related to the
specific conditions of the export economies or even to regional patterns. Two
notable examples, following the Independence War boom and crash, were the
loan frenzy of the 1880s followed by the 1890 Baring crash, and the Wall Street
bond boom of the 1920s followed by the 1929 crash (Marichal, 1989).
Under these conditions, one member of the group formulated a `general
thesis' that countries always borrow where they can ± others considered it
incapable of disproof and thus meaningless. Obviously, access was not inde-
pendent of exports. Indeed, a prime characteristic of an export-economy boom
was that it made countries eligible for foreign borrowing. Thus, even a country
which had persisted in moratoria and rescheduling, such as Colombia, could
become eligible for external borrowing when exports performed well. Sharp
debt boom±busts enhanced the cyclical pattern characteristic of export econ-
omies, with additional effects: they weakened the development of the internal
tax system during booms and enhanced the burden of fiscal adjustment during
the succeeding crisis.
External capital flows had additional consequences, the result of the inter-
play of interests of lenders and investors and those of governments, which were
either positive or negative. The positive took the form of institution building.
For example, foreign bankers insisted on the strength of the Bolivian Fiscal
Commission, with remarkable results in terms of the development of a modern
tax system, despite the opposition of tin producers (Chapter 7). The negative,
depending on the vulnerability given by geography and other related factors
such as the size of the economic interests at stake, took the form of political
interference, to the extreme of invasion.
The third type of interactions between the export sector and the rest of the
economies was the forward and backward linkage generated by export activ-
ities. The net effect on the economy was the balance between direct effects at a
macro and micro level, which were positive, and the negative relative price
22 Introduction
shifts which export expansion generated. Two analytical frameworks compete

for the analysis of the dynamics of export economies, by emphasizing either
the positive or the negative effects. On the one hand, traditional staple theory
highlights the positive forward and backward linkages and the domestic supply
response to the demands generated by the export sector (see, on the linkage
approach, Hirschman, 1981). On the other hand, the `Dutch-disease' literature
emphasizes the limited factor availability faced by the export economy: the
increasing demands thus generate a reallocation of scarce productive factors
towards the expanding sectors at the cost of other domestic activities. Relative
price shifts are the mechanism by which this reallocation takes place (Corden
and Neary, 1982; Neary and van Wijnbergen, 1984). The domestic-supply
response to the demand generated by export sectors is, thus, the clue to the
predominance of one or other of these effects. As we will see, this response
depends on the nature of the export sectors, size, government policies and the
ability to attract productive resources into the expanding sectors.
The direct effects at the macro level were, first of all, the result of rising
income. Comparatively, the strength of this effect was determined by the size
of returned value, which depends on the use of local factors and inputs and the
degree to which the sector was successfully taxed without limiting its rate of
growth. As we have seen, the nature of the export product was a critical
determinant of some of these effects, as it defined the labour demand and the
processing and transportation requirements. Export taxation was, by contrast, a
political economy issue. Also, the nature of the labour mobilized for export
activities, and the manner of its mobilization, were crucial. Stronger demand
effects were present when either European wage labour or peasant farming were
involved, whereas the weakest demand effects were characteristic of those
economies where forced mobilization mechanisms were used to guarantee the
labour force required by the expanding export activities.
The demand linkages are obviously more encompassing. A particularly relev-
ant issue is whether the rising demand for consumer tradable goods generated
directly or indirectly by export expansion was met by domestic supply. A very
important part of the modus operandi of an export economy in this regard was
associated with its impact on food supplies for the domestic market. As the two
subsequent volumes show, dependence on food imports became with time a
serious point of vulnerability in foreign exchange crises, to which many coun-
tries responded with import substitution. This vulnerability has its roots in this
period, as domestic food supplies were inadequate, particularly to feed the
growing cities.
Experiences varied considerably in this regard. Some agricultural economies
were exporters of food ± Argentina and Uruguay, in particular. In others, how-
ever, food imports became sizeable and import substitution of those foodstuffs
became an important issue at some stage of development ± Colombia and
Central America. The domestic sugar and flour industries were protected by
high tariffs in Colombia in the early twentieth century. Cuba is probably the
most striking case of an economy which was at the same time a large exporter
and importer of food. Competition for land was in this case an important
matter, a fact which was not true in other agricultural economies.
Mineral economies provide even more interesting contrasts. The positive
response in Chile to the consumer demands from nitrate workers is a remark-
able example; it also illustrates the importance of state intervention to guaran-
tee adequate domestic transportation and investments in human capital
(Chapter 6). The Peruvian copper story under the Cerro de Pasco Corporation
is quite different, as it began generating significant linkages to the Central
Sierra peasant communities. The contamination of agricultural land by fumes
from the copper smelter led to the purchase of land, the creation of a cattle-
raising division and a delinking from the regional economy, which led it into
stagnation in the 1920s (Chapter 8). Indeed, in several countries, many large
mining companies became importers of foodstuffs for their workers, a beha-
viour which was also typical of banana plantations.
The story of manufacturing during the export age is an equally fascinating
subject. Rising exports were matched by rising imports, mainly manufactures.
However, as several research efforts in recent decades and the contributions to
this volume indicate, manufacturing development clearly predated the 1930s
and the postwar era of state-led industrialization. Manufacturing growth was
induced in several countries through four different channels. The first was the
forward linkage of export activities: as we have seen, some exports required
significant processing and thus generated numerous manufacturing establish-
ments in many countries. The second was the combination of backward demand
linkages and high transport costs, which generated a `natural import substitu-
tion' of some manufactured foodstuffs, beer, printing and, later, cement, among
other manufactures. Indeed, the term import substitution may be inappropriate
in this case, as manufacturing production directly accompanied in several cases
demand growth without significant imports ever developing.
The two additional channels were associated with tariff policy. As we have
seen, all countries in the continent relied heavily on import tariff revenues
during the export age, and this necessarily provided protection. The tariff
schedules of the time generated some peculiar distortions, which are behind
the scenes in the stories of early manufacturing development. They were either
specific tariffs based on the weight of merchandise, or ad-valorem tariffs paid
on the basis of an official price list. Specific weight tariffs gave high protection
to the production of industrial goods with low value per weight: for instance,
the system gave higher protection to simple rather than sophisticated textiles.
Also, under both systems, the real protectionist effect depended on interna-
tional and domestic processes: inflation eroded protection, but deflation ±
24 Introduction
which was quite common during world and domestic crises up to the 1930s ±
increased protection. This generated a peculiar cyclical pattern of protection,
which was compounded with that of exchange rates in those countries which
did not follow gold-standard rules. During external booms, manufacturing was
discouraged by falling ad-valorem tariffs and real appreciation ± but was pro-
moted by rising demand ± but tariff and exchange-rate protection were auto-
matically increased during crises. The coincidence of import substitution
efforts in several Latin American countries during the First World War was a
precedent to a similar story during the 1930s, which is explored in the next
volume. During the war, the physical scarcity of some manufacturing goods
imported from Europe generated an additional encouragement to produce
them domestically.
Beyond this, protectionism was quite common in Latin America at the time.
This obviously matched worldwide trends, as free trade was in fact an exotic
policy practice even in the industrial centre, as we have seen. Thus, contrary to
modern intellectual trends, export promotion and protectionism were not
viewed in several countries as opposite, but rather as complementary strategies,
elements of one single modernization drive. Brazil, Mexico, Chile and
Colombia are the outstanding examples of Latin American countries which
practised protectionism long before ìnward-looking' development, and where
early industrialization was closely associated to protection. Argentina was a
different case, in which early industrialization was associated much more
with the other mechanism we have outlined. Interestingly, the rising labour
movement in this country was distrustful of both protectionism and exchange-
rate activism, both of which were seen as ways to depress real wages.
The rise of the textile industry in this country prior to the Great Depression
provides a possible fifth link: its complementarity with dynamic cotton cultiva-
tion. However, this was not common at the time in the region. Indeed, the
processing of imported low-tariff raw materials ± that is, high effective protec-
tion ± was very important for early industrialization in other countries.
The success of manufacturing development depended, of course, on other
factors, two of which were crucial. The first was size, and the second was
domestic market integration. Indeed, early industrialization critically depended
on this market integration, which was the accumulated effect of export growth,
urbanization and the development of modern infrastructure. Obviously, a large
country such as Brazil only depended on this process taking place around its
most dynamic growth pole, the SaÄo Paulo region, delaying true national integ-
ration. By contrast, in medium-sized economies, such as Colombia, successful
industrialization was delayed by lack of domestic market integration, despite
rapidly rising domestic incomes and protectionism. Late domestic market
integration would, nonetheless, become a facilitator of very rapid industrial
growth in Colombia since the 1930s.
The development of a modern transport infrastructure was one of the most

important linkages of exports to the domestic economy. Independently of size,
successful export growth required a modern transportation network. In this
sense, this could be regarded as one of the crucial forward linkages of export
growth. It also operated as a backward linkage, however, as rising domestic
demand for goods also generated a demand for modern transport services. The
complementarity between export growth and foreign capital was also crucial in
this regard, as we have seen, as was efficiency in the use of rising public sector
revenues (see below). The positive effects of infrastructure development could
be maximized when a national integration strategy lay behind it, as in Mexico
and Chile. As we have seen, however, even in countries where it tended to
further disintegrate the domestic economy, by improving the integration of
different regions to the rest of the world before they were integrated among
themselves, in the long run it helped to integrate the domestic market.
Finally, the linkages of the export economy were sometimes microeconomic
and institutional. At a microeconomic level, the export sector could also
become a major instrument of technological transmission and diffusion, the
major building block of a modern entrepreneurial class and an instrument for
the creation of human capital as a training ground for a skilled labour force. At
an institutional level, the development of mining and trade codes, and
improved banking and currency regulations, were closely associated to export
growth.
`Dutch-disease' effects were present whenever domestic factor availability
became a major constraint. Land was a critical factor in only a few cases,
notably Uruguay and Cuba; in the latter case, transport development facil-
itated, nonetheless, the incorporation of the east of the island to sugar devel-
opments in the early twentieth century. More important were labour and
capital. Mobile labour was crucial for economic growth during the export age,
as we have seen throughout this chapter; foreign capital was particularly
important for infrastructure. Very rapid rates of growth could thus be achieved
where an ample supply of land was mixed with massive labour immigration
and capital inflows, as the spectacular rates of growth of Argentina indicate: 7.4
per cent a year, 3.8 per cent per capita, at its peak in 1881±1913, according to
CorteÂs Conde. Chile comes next, but here government policies were crucial,
particularly the ability to extract a large part of the surplus of the nitrate sector
and invest it successfully in transport infrastructure and education. This `tax
and invest' policy was the clue to avoiding the `Dutch disease', as Palma con-
vincingly argues in his contribution to this volume (Chapter 6).
Overall, export growth was the crucial factor behind economic development
at the end of the export era. The correlation between exports and GDP per
capita prior to the First World War and the Great Depression is remarkable, as
Bulmer-Thomas (1994, Chapter 5) has argued. National differences were
26 Introduction
Table 1.4 Latin America: economic and social indicators, circa 1913
Exports GDP Rail ways % of labour Literacy Life-

per capita per capita per capita force in expectancy
agriculture
Argentina 67 188 4.4 34.2 63 46

Bolivia 20 0.7 21 29
Brazil 13 44 1.0 66.7 35 31
Chile 75 140 2.3 37.7 56 30
Colombia 6 45 0.2 70.5 41 31
Costa Rica 28 76 1.6 50 34
Cuba 69 148 1.5 48.9 59 36
Dom. Rep. 13 0.3 67.6
Ecuador 12 0.4 39
El Salvador 13 39 0.9 14 24
Guatemala 6 65 0.1 8
Haiti 5 0.4 31
Honduras 9 67 1.4 31 30
Mexico 13 78 0.6 63.7
Nicaragua 14 54 0.2 83.7 31
Panama 9 0.6 40 30
Paraguay 10 0.8 30
Peru 9 37 0.3 27
Uruguay 59 195 2.1 28.0 67 52
Venezuela 9 0.3 72.0 30 30
Notes: Exports per capita: current USA dollars per capita; railways per capita: kms of track per
thousand persons; GDP per capita: GDP per capita based on current prices, in USA dollars; literacy:
proportion of literates in population 15 years and over; life-expectancy: years at birth.
Sources: Exports from Thorp (1998); railways from Mitchell (1993); literacy from Hunt (1998); GDP
figures from Bulmer-Thomas (1994, table A.2.1, p. 439); agricultural labour force from Bulmer-Thomas
(1994, table 5.1, p. 122).
closely associated with differences in export performance. Argentina, Chile,

Uruguay and Cuba were highest in the export ranking; they also had the high-
est GDP per capita, the highest rate of railroad tracks to population, and the
highest literacy and life expectancy ratios (Table 1.4). On the basis of limited
data, they also had the highest manufacturing output per capita and the lowest
proportion of the labour force in agriculture. In Cuba, however, global perform-
ance was relatively poorer, as exports were not accompanied by a similar
growth of non-export activities. Also, in relative terms, Costa Rica was above
other Central American countries in terms of export and economic and social
performance generally. Mexico is another interesting intermediate case, which
was not associated with particularly successful export performance but rather
with above-average development of transportation infrastructure.
Conditions were changing towards the end of our period of analysis. Some
medium-sized countries ± Colombia, Peru and Venezuela ± had a late export
boom, as we have seen, which then improved their relative standing. Moreover,
the basis for the dynamism of Argentina was being eroded. Its rate of growth fell
to 2.3 per cent in 1913±29, slightly below population growth. At the same time,
Brazilian economic growth had accelerated. After a rather dismal record in the
nineteenth century, this country's GDP grew at 4.3 per cent a year between
1900±04 and 1925±29 (Haddad, 1980). Moreover, this was the first case in the
region in which GDP growth was not accompanied by rapid export growth ± 1.7
per cent a year in the first three decades of the century, as we have seen. Inward-
looking development had made its first appearance, facilitated nonetheless by
the rapid growth of coffee in the SaÄo Paulo region in the late nineteenth
century and by the transfer of resources from foreign coffee consumers to
protected industrial producers in Brazil. These were the first signs of the new
era to come.
The role of the state and the political economy of export

development
The state played an essential role in export development and the ways the
export sector was linked to the rest of the economy. Although it was certainly
not the type of interventionist state which built up from the 1930s in Latin
America ± as in the rest of the world ± it also did not fit the laissez-faire
image constructed by some nostalgic analysts viewing the export era in Latin
America.
In passing, it should be said that the traditional `law and order' functions
were performed rather poorly in many cases, particularly with respect to òrder'.
As we pointed out above, the building of a national, republican order was
painful in most nations. One of the major consequences of this fact was that
the victory of economic liberalism was not necessarily accompanied in the
region by the development of liberal `political' institutions. Indeed, economic
liberalism happily, even functionally, coexisted with many authoritarian
regimes.
State intervention played an essential role in five different areas. The first
was, as we have seen, land and labour policies. As labour or peasant movements
developed, this also meant either repression, the promotion of social legislation
to accommodate the claims of labour or the peasantry, or a mixture of both.
The second was taxation. Although taxation of exports was adopted in some
mineral economies, the most common way to tax at the time was to impose
import tariffs. Whether intentionally or not, this meant protection for domes-
tic agriculture and manufacturing. Indeed, as we have seen, protectionism was
widely practised in the region long before import substitution became an
official policy. Towards the end of our period of analysis new forms of taxation
were introduced on an important scale in the most successful export
28 Introduction
economies, but fiscal dependence on such taxation became common only in

the later phases of development.
Closely associated with the former issue was that of the allocation of govern-
ment revenues. In this regard, there is a clear differentiation within the region,
and in particular countries over time, in the pattern of government expendi-
ture. As several studies in this volume indicate, the decision to allocate increas-
ing resources to transportation and education vs. more traditional areas of
expenditure ± general administration, defence and debt service ± was crucial
to economic growth. Indeed, although private capital was closely involved in
the development of the domestic transportation network, particularly railways,
state investments in this area were very important in most countries. Towards
the end of our period of analysis, with the rise of motor vehicle transportation,
state investments became even more important for the development of road
networks.
Governments also played a central role in monetary developments and in the
early development of domestic banking. The establishment of a unique
national currency remained problematic for several countries after Independ-
ence. More than that, however, in the last decades of the nineteenth and in the
early twentieth century, two issues were relevant. The first was the very gradual
abandonment of the silver standard which Latin America as a whole had
inherited from colonial times. This was a crucial issue after the depreciation
of silver which took place after the 1870s, as the gold standard made a definite
advance in the centre countries. As many countries, notably Mexico, were
major exporters of silver, this phenomenon also had major direct implications
for foreign trade. The second issue was the fairly common use of inconvertible
currencies during the export age. Among countries studied in this volume,
Argentina, Brazil, Chile and Colombia experienced long episodes of inconvert-
ibility. These episodes were associated with public sector financing, with exter-
nal crises or with both, as fiscal crises tended to coincide with external crises
due to the strong fiscal dependence on customs revenues.
The late abandonment of the silver standard and the episodes of inconvert-
ibility generated an inflationary propensity relative to world levels. Monetary
expansion even led to `hyperinflation' in two cases, both associated with civil
strife ± Colombia during the War of the Thousand Days, at the turn of the
century, and Mexico during the Revolution. It also meant that nominal depre-
ciation could be partly counted upon as both a protectionist and an export-
promoting device, to the extent that domestic prices or wages responded
partially or with a lag to such depreciation. In turn, this meant that the estab-
lishment of the gold standard, which became increasingly common in the early
twentieth century, prior to the First World War, and again in the 1920s, was not
always easy. To establish or reestablish the gold standard meant both the use of
scarce fiscal resources to guarantee the reserves necessary to make the gold
standard viable, and also required domestic relative price adjustments which
had adverse effects on import-competing and export activities.
Finally, aside from taxation issues, governments were also involved in the
direct regulation of exports. The early and most notable example were, of
course, Brazilian coffee regulations. They included prohibitions of new plant-
ings in the first decades of the century, three valorization schemes ± 1906, 1917
and 1921 ± which aimed to accumulate stocks within Brazil to defend prices,
and the `permanent defence' of coffee introduced by the SaÄo Paulo government
in 1925 which led to other regulations (Rowe, 1932). From the mid-1920s, the
Cuban government was also active in the regulation of domestic sugar produc-
tion to manage its own excess supply, and in the promotion of partial interna-
tional agreements (Chapter 10). Indeed, domestic and international regulation
of commodity markets became an attractive alternative since the First World
War as a way of managing overproduction; it became widespread in the 1930s
(Rowe, 1965, part IV).
The specific response of the state depended obviously on the character of the
export elites and to what extent they were articulated or disarticulated in social,
economic and political terms, among themselves, and with other groups. Some
hypotheses can be advanced in this regard. Strong domestic ownership of a
dominant export product but diverse regional bases led to political trading
mechanisms between regional elites, as in Brazil and Colombia. A system
where concessions for the exploitation of natural resources were required
usually meant a close relationship between the state and the elites, their inter-
play and often their overlapping. For instance, members of the government or
the legislature in Mexico and Chile participated in mining companies as share-
holders, and had interests in other activities, including the commercial banks.
Second, the needs of the sector were influential in the demands the elites
imposed on the state. They may or may not have required some degree of state
intervention that went beyond the guardian state. Moreover, this may or may
not have provoked some degree of institutional development in the state, the
issue of laws, norms and other regulations, the prime positive examples being
Colombia and Mexico. Again, export diversity and a strong presence of foreign
capital substituting for the state made for a weak link in this regard (Peru).
The third area was the elite's ability to respond to the social and economic
challenges of the period, with varying degrees of consensus and violence. This
included the ability to respond to the social conflicts which arose during the
export age, including the development of organized labour or the conflicts over
labour and land relations. In Brazil and Chile, elites showed a response capacity,
and to a lesser degree in Colombia, with considerable degrees of violence in the
latter case. Elite capacity to respond was not a characteristic of Peru or Mexico.
This introduction has attempted to draw out the rich themes stimulated by
the case studies and by the workshop where they were discussed. Each theme is
30 Introduction
significant not only for the understanding of the period which it yields, but for
the way export dynamics, linkages and the consequences for the state, the
evolving political economy of policy-making and public±private relations,
were to condition and shape the decades that followed. This theme is devel-
oped in the overview volume that is also a part of this project (Thorp, 1998).
The rest of the present volume contains the country-specific studies in which
the overview is grounded.
References
Ashworth, W. (1975) A Short History of the International Economy since 1850, London:
Longman.
Bairoch, P. and Etemad, B. (1985) Structure par produits des exportations du Tiers-Monde,
Geneva: UniversiteÂ de GeneÁve: Centre d'Histoire Economique Internationale.
Bulmer-Thomas, V. (1994) The Economic History of Latin America since Independence, Cam-
bridge: Cambridge University Press.
Chiriboga, M. (1988) Àuge y crisis de una economõÂa agroexportadora: El perõÂodo cacao-
tero', in E. Ayala Mora (ed.), Nueva Historia del Ecuador, Vol. 9, Grijalbo.
Corden, W.M. and Neary, J.P. (1982) `Booming Sector and De-Industrialization in a Small
Open Economy', Economic Journal, December.
ECLA (1965) External Financing in Latin America, New York: United Nations.
Eichengreen, B. (1995) Golden Fetters: The Gold Standard and the Great Depression 1919±
1939, New York: Oxford University Press.
Grilli, E.R. and Maw Cheng Yang (1988) `Primary Commodity Prices, Manufactured
Goods Prices and the Terms of Trade of Developing Countries: What the Long Run
Shows', The World Bank Economic Review, January.
Haddad, C.L.S. (1980) `Crecimiento Economico do Brasil, 1900±76', in P. Neuhaus (ed.),
Economia Brasileira: Uma Visao HistoÂrica, Rio de Janeiro: Editora Campus.
Hatton, T. and Williamson, J.G. (1994) Ìnternational Migration 1850±1939: An Economic
Survey', in T. Hatton and J.G. Williamson (eds), Migration and the International Labour
Market 1850±1939, London/New York: Routledge.
Hirschman, A.O. (1981) À Generalised Linkage Approach to Development, with Special
Reference to Staples', in Essays in Trespassing: Economics to Politics and Beyond, New York:
Cambridge University Press.
Hunt, S.J. (1997) `The Human Conditions in Latin America, 1900±95', mimeo, Consult-
ancy paper for this project.
ÐÐ (1985) `Growth and Guano in Nineteenth Century Peru', in R. CorteÂs Conde and S.J.
Hunt (eds), The Latin American Economies: Growth and the Export Sector 1880±1930, New
York: Holmes & Meier.
Kenwood, A.G. and Lougheed, A.L. (1971) The Growth of the International Economy 1820±
1960, London: George Allen & Unwin.
Lewis, W.A. (1978) Growth and Fluctuations 1870±1913, London: George Allen & Unwin.
Maddison, A. (1995) Monitoring the World Economy 1820±1992, Paris: OECD Development
Centre.
ÐÐ (1991) Dynamic Forces in Capitalist Development: A Long-Run Comparative View, New
York: Oxford University Press.
Marichal, C. (1989) A Century of Debt Crisis in Latin America: From Independence to the Great
Depression, 1820±1930, Princeton: Princeton University Press.
Mitchell, B.R. (1993) International Historical Statistics: The Americas, 1750±1988, Basing-
stoke: Macmillan.
Myint, H. (1958) `The Classical Theory of International Trade and the Underdeveloped
Countries', Economic Journal, June.
Neary, J. P. and van Wijnbergen, S. (1984) `Natural Resources and the Macroeconomy: A
Theoretical Framework', in J. P. Neary and S. van Wijnbergen (eds), Natural Resources
and Macroeconomy, Oxford: Basil Blackwell.
Ocampo, J. A. (1984) Colombia y la EconomõÂa Mundial 1830±1910, BogotaÂ: Siglo XXI.
Palacios, M. (1983) El cafeÂ en Colombia 1850±1970: Una historia econoÂmica, social y polõÂtica,
Mexico/BogotaÂ: El Colegio de MeÂxico/El Ancora Editores.
Rowe, J.W.F. (1965) Primary Commodities in International Trade, Cambridge: Cambridge
University Press.
ÐÐ (1932) Studies in the Artificial Control of Raw Material Supplies, No. 3, Brazilian Coffee,
London and Cambridge Economic Survey, January.
Thorp, R. (1998) Progress, Poverty and Exclusion: An Economic History of Latin America in the
20th Century, Washington: Inter-American Development Bank and Johns Hopkins Press.
2
Brazil as an Export Economy,
1880±1930*
Marcelo de P. Abreu and Afonso S. Bevilaqua
Introduction
Conventionally, the year 1930 is taken as a watershed in Brazilian history. It

marked the fall of the political regime of the Old Republic and the transition
from an export-led economy to a long period of growth led by import substitu-
tion. A closer look at the evidence makes this transition far less clear-cut.
Coffee's contribution to the generation of foreign exchange earnings remained
crucial until the 1960s when import substitution had reached maturity. On the
other hand, domestic industrial production had already become relevant as a
share of domestic supply at the turn of the century. While attention here will be
concentrated on the export economy in the 1880±1930 period, the links
between the export economy and import substitution should be kept in
mind. Table 2.1 shows the gradual replacement of agriculture by industry in
GDP. While there is no information for the pre-1910 period, it is known that
industrial activity was rather limited before the 1870s, and the first important
industrial investment boom was in the early 1890s.
It is also important to keep in mind that after independence Brazil preserved
its integrity as a country of continental size, with an area of about 3.3 million
square miles. The Brazilian economy has often been described as a number of
islands, with its main regions having closer links with markets abroad than
with those of other regions. Coffee culture has been mainly limited to the
south-east, migrating from the mid-nineteenth century from the neighbour-
hood of Rio de Janeiro along the Paraiba valley in the direction of SaÄo Paulo,
* Revised version of a paper presented in the workshop on the Latin American Export Economies for
the project on the Economic History of Latin America in the 20th Century, held at Panajachel,
Guatemala, 9±11 December 1996. The authors wish to thank Elisa Reis, AndreÂ Villela, RogeÂrio
Werneck and Fernando Blanco as well as participants in the Panajachel meeting for their help and/
or comments. The competent research assistance of Fernando H. Alvares, Alvaro B. A. Motta and Luiz
Gustavo Cherman is acknowledged. The FINEP (Financiadora de Estudos e Projetos) and the Conselho
Nacional de Desenvolvimento CientõÂfico e TecnoloÂgico are thanked for financial support.
32
Marcelo de P. Abreu and Afonso S. Bevilaqua 33
Table 2.1 Brazil: GDP shares, 1910±93 (%)a
Year Agriculture Industry Services
1910 35.8 14.0 50.2

1920 31.9 17.1 50.9
1930 30.6 16.5 52.9
1940 25.0 20.8 54.2
1950 24.3 24.1 51.6
1960 17.8 32.2 50.0
a
GDP at factor prices, including financial intermediation;
services include government.
Sources: 1910±47: Haddad (1978), sectoral data aggregated
using 1947 weights; 1947±80: Brazil (1990).
then spreading in the Paulista plateau; Minas Gerais became an important

producer from the turn of this century; sugarcane agriculture and cotton pro-
duction were concentrated in the north-east until the 1930s; rubber was
produced in the Amazon region; and the south, with weak links with the core
coffee economy, traditionally exported hides, skins and jerked beef, and later
on timber and matte.
To put the economic integration issue in perspective, it is important to take
the geography of the country into account. The distance from Rio to Recife is
1125 miles and to BeleÂm, in the mouth of the Amazon river, is 2280 miles.
Manaus is a further 924 miles upriver from BeleÂm. Rio Grande, the main
harbour in Rio Grande do Sul, is 875 miles from Rio. Transportation costs
other than by ship were extremely high before the introduction of railways,
as rivers generally flow in the north±south direction and the Brazilian plateau
rises abruptly to 3000 feet roughly 50 miles from the coastline in most of the
south-east.
The country was very sparsely populated. In 1872 the population was 10.1
million, barely three inhabitants per square mile (see Table 2.2). There are
indications that the 1920 census data seriously overcounted Brazil's actual
population at the time, but there is a clear picture of high and accelerating
population growth, especially after the 1940s, as a result of a rapid fall in
mortality rates and persistently high birth rates.
In this chapter, attention will centre mainly on coffee due to its dominant
role in the Brazilian economy and to the persistent high share of Brazilian
exports in the world market, which allowed Brazil to try to exploit its market
power. The chapter is divided into five sections besides this introduction. The
next section focuses on the core export economy, from the 1880s to 1930, as
well as on the commodity exports outside the core in the same period: rubber,
sugar and cotton. The third section analyses the implications of the position of
Brazil as the main world supplier of coffee, particularly the links between
34 Brazil as an Export Economy, 1880±1930
Table 2.2 Brazil: population data, 1872±1960
Population at Population yearly Share of population

census date growth rates in cities of more than
(millions) between censuses 50 000 inhabitantsa
1872 10.1 ± 5.9

1890 14.3 2.5 6.8
1900 17.3 2.2 9.4
1920 30.6 3.8 10.7
1940 41.2 1.7 12.6
1950 51.9 2.6 16.3
1960 70.1 3.7 22.9
a
Data not strictly comparable, as up to 1920 rural population in cities of more than 50 000
inhabitants was also included.
Source: Lopes, 1968, pp. 13±16.
domestic policy and world coffee prices. We than consider the linkages of coffee
agriculture with other sectors, especially infrastructure and industry, followed
by the export economy between 1930 and 1960, included as a coda. But this is
an important coda, due to the role of coffee policy in Brazil's recovery from the
Great Depression and the continued importance of commodity exports in
foreign exchange earnings, even if the economy became increasingly closed.
In spite of the erosion of its dominant past position, in 1960 Brazilian coffee
exports still corresponded to almost 37 per cent of world exports. The final
section concludes the chapter.
The export economy, 1880±1930
Coffee became the leading Brazilian export commodity very early in the nine-
teenth-century; by the early 1830s it already accounted for nearly 30 per cent of
total exports. From the 1850s it reached almost 50 per cent of total exports, on
average, and remained in the 50±65 per cent range until the early 1960s. Its
share fell slightly in the 1860s due to the US Civil War, and more significantly
in the 1940s as the Second World War closed most Brazilian coffee markets
outside the USA (see Table 2.3).
The Brazilian share of world coffee production, which was already in excess of
50 per cent in the 1850s, fell slightly in the 1870s, but then, as competitors in
Asia were hit by plant disease and price support schemes were introduced, it
increased to reach more than 60 per cent in every year from 1896 to 1942 - more
than 70 per cent in almost half of these years and more than 80 per cent in
1906. It fell from the early 1940s, but it was still around 47 per cent in the
1950s. The long-term evolution of coffee prices is shown in Figure 2.1. Figure
2.2 shows the terms of trade between coffee prices and import prices. As the
Table 2.3 Brazil: commodity export shares, 1850±59 to 1950±59a
Coffee Sugar Cotton Rubber Total
1850±59 48.7 21.3 6.3 2.2 78.6

1860±69 45.9 12.3 17.7 3.1 79.0
1870±79 56.3 11.8 9.7 5.5 83.3
1880±89 60.5 10.6 4.4 7.6 83.1
1890±99 65.4 6.1 2.5 14.2 88.2
1900±09 53.1 1.5 2.3 25.6 82.6
1910±19 52.1 2.4 1.7 16.4 72.6
1920±29 67.2 2.3 2.7 2.7 75.0
1930±39 56.3 0.5 11.1 1.0 68.9
1940±49 36.3 0.8 11.6 1.6 50.3
1950±59 57.9 2.4 7.0 0.2 67.4
a
10-year average of yearly shares. 1850±51 to 1887±88: fiscal years.
Source: Computed from AnuaÂrio EstatõÂstico do Brasil, various years, and Brasil em NuÂmeros, various
years.
80.00
60.00
40.00
20.00
0.00
1850
1855
1860
1865
1870
1875
1880
1885
1890
1895
1900
1905
1910
1915
1920
1925
1930
1935
1940
1945
1950
1955
1960
Year
Figure 2.1 Coffee prices, 1850±1960a

a
US cents per pound, calendar years.
Source: US Department of Commerce.
Brazilian series of import prices are not very reliable, export prices for the USA
and the UK were used as proxies. Export quantities increased modestly until
1880, at 0.9 per cent yearly, then very rapidly until 1900 at 5.5 per cent yearly,
slowing down to 1.7 per cent in 1900±30. By the end of the 1920s capacity to
import (due to coffee exports) had increased more than fourfold in relation to
1851, at roughly 1.8 per cent yearly (see Bacha and Greenhill, 1992, appendix).
The share of Brazil in the world market was lower than the share in produc-
tion as Brazilian coffees tended to command lower prices, and 78 million bags
corresponding to three years of world consumption were destroyed under
300
UK
250
US
200
150
100
50
0
1850
1858
1866
1874
1882
1890
1898
1906
1914
1922
1930
1938
1946
1954
Year
Figure 2.2 Coffee terms of trade, 1850±1960a

a
Coffee prices over US/UK export prices (1913 100).
Sources: Mitchell and Deane (1971) and US Department of Commerce.
coffee support programmes, mainly in the 1930s. The share of Brazilian coffee
exports in total world exports was nearly 40 per cent in the 1850s, rose to more
than 50 per cent in the 1860s, receded to 40 per cent in the 1870s, and then
increased to reach almost 75 per cent at the turn of the century and more than
80 per cent in 1911. It then fell, but remained above 50 per cent until 1929.
With the Great Depression it fell further to reach a value lower than 30 per cent
in 1938±39, and then rose again to almost 50 per cent until the early 1950s. In
1960 it was still above one-third of the market.
Brazilian coffee exports increased with the dissemination of world coffee
consumption. Growth was very rapid in the early nineteenth century: fourfold
by volume in the 1820s, twofold in the 1830s, and 1.7 times in the 1840s. Then
there was stagnation in the 1850s, further increase in the 1860s ± by 70 per cent
± and stagnation in the 1870s. Growth was resumed in the 1880s, by 30 per
cent, and spectacularly in the 1890s as the export quantum increased by 170
per cent. From 1821±25 to 1901±05, the volume of coffee exports increased
at 5.2 per cent a year. From 1851±55, when production was already well-
established, to 1901±05 it increased at a yearly rate of 3.2 per cent. In the late
1860s the USA market was already absorbing some 50 per cent of the Brazilian
coffee output. This share was the same in the early 1900s and in the late 1920s.
The British share of Brazilian total exports fell sharply due to the insignificant
British coffee imports. Before the First World War Britain had been overtaken by
Germany as a market for Brazilian exports and this share had been reduced to
13 per cent (see Delfim Netto, 1979, p. 9 and Brazil, 1941a, pp. 299±304).1
The domestic consumption of coffee was relatively unimportant when com-

pared to production.2 This makes the ratio between coffee exports and GDP a
reasonable indication of the direct weight of coffee in economic activity. How-
ever, there are no reliable series on GDP and GDP deflators for the period before
1908. From the available data it seems that the coffee exports±GDP ratio was
around 9 per cent in 1850, and reached more than 10 per cent in 1900 and
1913. In the last `normal' pre-depression year of 1928 this ratio was slightly
above 9 per cent, but it fell with the collapse of coffee prices and the reduction
of the importance of exports as a share of GDP, reaching 4.8 per cent in 1939
and 4.3 per cent in 1947. In 1960 it was 2 per cent.3
While coffee production corresponded to the core of the Brazilian export
economy for the first 140 years of its independence, other commodities such as
sugar, cotton and rubber were also important in specific periods of Brazilian
history. The decline of sugar exports was steady from the 1820s, as the sugar
economy in the north-east became increasingly less efficient in spite of a spurt
of modernization in the 1880s.4 Sugar production in the south-east became
important only after 1914, but Brazilian exports never played an important role
again.
The share of cotton exports in total exports also declined, but less smoothly
as the trend was reversed in several periods. In the 1860s, cotton exports peaked
to reach nearly 30 per cent of exports in 1864±65 during the `cotton famine'.
More than 80 per cent of the exports were from the north-east and the northern
provinces.5 In the second half of the 1930s, due to the side effects of USA cotton
price support programmes and the rise of the German compensation trade,
they peaked again at more than 20 per cent of total exports in 1939. There was a
sharp growth in cotton production in SaÄo Paulo: 97 per cent of the cotton
exports were embarked in north-eastern ports in 1928, 81 per cent in Santos in
1939. Cotton exports also rose, but more modestly in the second half of the
1940s due to the postwar cotton demand boom and the export of British-
controlled stocks accumulated during the war, only to fall again from the
early 1950s (see Ellis, 1971, pp. 246±55; and Abreu, 1990).
Exports of native rubber from the Amazon region became relevant in the
1890s, peaked in 1910 at nearly 40 per cent of total Brazilian exports, and fell
precipitously as exports of south-east Asian plantation rubber started to rise.6
By 1919, rubber exports were under 5 per cent of total exports. As in the case of
coffee, Brazilian output was a high proportion of world output: not less than
50±60 per cent in the period between 1895 and 1911 (see LeCointe, 1922, vol. I,
p. 333). Figure 2.3 shows the evolution of rubber prices from 1887 to 1917.
Rubber exports were an important source of foreign exchange earnings,
especially in the golden decade before the First World War, and their expansion
played an important role in the success of the 1898 stabilization programme
which led to exchange rate appreciation and the adoption of the gold standard.
30.00
25.00
20.00
15.00
10.00
5.00
0.00
1887
1889
1891
1893
1895
1897
1899
1901
1903
1905
1907
1909
1911
1913
1915
1917
Year
Figure 2.3 Rubber prices, 1887±1917a

a
Fine Para, francs per kilogram, London.
Source: LeCointe (1922).
The rubber economy was based on the extraction of native rubber with ex-
tremely limited pre-export processing. The boom attracted surplus manpower
from other parts of the country, especially from the declining north-east (see
Furtado, 1965, ch. 23). Its abrupt end resulted in a sharp contraction of eco-
nomic activity in the region as it returned to subsistence activities.
Coffee in Brazil was produced in plantations. Before the 1870s the region
around Vassouras, relatively near Rio de Janeiro in the ParaõÂba Valley, had been
the most important producing area, but the primitive planting methods and
the characteristics of the dominant topography led to the continuous destruc-
tion of productive soils. As free land was available at low cost, the coffee frontier
moved southwards in the direction of SaÄo Paulo leaving behind a trail of
destruction.7 In the Paulista plateau there were very good soils for coffee
together with a good topography. For many decades, from the 1870s until the
Great Depression (1928±33), SaÄo Paulo was the most dynamic coffee region in
the country. The Brazilian native population in expanding coffee culture areas
had already been unfavourably affected by the economic expansion before the
coffee boom, so the link between the expansion of the coffee culture and the
declining standard of living of the Indian population does not seem direct. But
the expansion of the coffee economy directly resulted in the destruction of
most of the native Atlantic forest, as burn and slash methods of clearance
remained the rule for the whole period.8
While land policies nominally protected the interests of all those seeking to
occupy free government land, actual policy prevented the establishment of
small-scale subsistence agriculture by immigrants for a long period and thus
the creation of an alternative to low-wage agriculture. The link between land
policy and the elasticity of supply of labour to the plantation-based coffee
agriculture was of crucial importance and was well-understood by coffee plan-

ters.9 The average size of coffee farms in Brazil declined over time, but produc-
tion in the 1920s was heavily concentrated in big farms. Even if it is accepted
that farms up to 20 000 trees were mostly worked by `peasant owners' and their
families with a minimum of hired labour, only 18 per cent of total coffee trees
in 1927 were not in `commercial' or big farms (see Rowe, 1963, p. 44, and Rowe,
1932, p. 89).10 The average coffee farm in Brazil in 1927, even taking as
representative the least productive group of farms, produced almost six times
the output of the average Colombian coffee farm in 1932 (see Rowe, 1932, pp.
88±9, and Palacio, 1983, pp. 437, 448).
Indeed, an essential long-term objective of economic policy in Brazil was to
assure an elastic supply of labour, so as to guarantee that wages remained low.
The government turned a blind eye to the illegal slave trade until 1850, and also
participated actively in the efforts to find alternative supplies of labour. Espe-
cially from the 1880s, as the abolition of slavery could be anticipated, a stable
substitute to slavery was found in the parceria system, which depended on the
attraction of European ± mainly Italian ± immigrants. By introducing a hybrid
system of labour relations including elements of share cropping and wages it
was possible to expand the coffee sector and to maintain low wages. There were
some federal subsidies to immigration, but the system was mainly financed by
the state of SaÄo Paulo. Brazil was the fourth major destination of European
immigration between 1815 and 1930, after the USA, Canada and Argentina.
Immigration peaked in the late 1880s, the 1890s, and immediately before the
First World War. After the turn of the century it was unfavourably affected by
restrictive legislation prohibiting subsidized emigration in Italy.11
The commercialization of coffee at the turn of the century was controlled by
foreign exporting houses which had circumvented Brazilian factors which had
traditionally acted as intermediaries between planters and coffee exporting
houses (see Stein, 1957, pp. 282ff, and Greenhill, 1977). While coffee produc-
tion was not directly capital-intensive, requiring little machinery, the inflow of
foreign capital was essential for the consolidation of the infrastructure related
to the expansion of the coffee culture.12 The expansion of the coffee industry
fostered, and was fostered by, investment in infrastructure. Table 2.4 shows the
concentration of foreign capital inflows in the golden decade before the First
World War.13 Investment in railways was attracted by the widespread use of
interest guarantees which were redeemed after the turn of the century. How-
ever, recent work on the economic impact of railways in Brazil has suggested
that coffee freight as a proportion of total freight in most big railway networks
in the south-east declined steadily from 1869 to 1913 (see Summerhill, 1995,
ch. 5). It is also surprising, as shown by data on British capital (see Table 2.5),
that investment was heavily concentrated in public services rather than in
railways, especially in the golden decade. Urbanization proceeded rapidly in
Table 2.4 Brazil: foreign capital stock, 1840±1930 (£ millions)
Public Public Non-public Non-public Total

portfolio portfolio investment investment
British othera British othera
1840 5.6 0 1.3 0 6.9

1865 13.0 0 7.3 0 20.3
1875 20.4 0 10.6 0 31.0
1885 23.2 0 24.4 ... 47.6
1895 37.5 1.0 40.6 ... 79.1
1905 83.2 5.0 41.1 34b 163.2
1913 129.1 36.9 135.2 120.7b 421.9
1930 163.0 91.4 118.6 123.8b 496.8
a
USA and French.
b
These estimates are particularly fragile given the weight of French investment in 1913 and its sharp
fall in the 1920s.
Sources: See sources in Abreu (1985, 1994 and 1995).
Table 2.5 Brazil: British direct investment and holdings of sterling-denominated

corporate securities by sector, 1865±1913 (£ millions)
1865 1875 1885 1895 1905 1913
Railways 5.4 6.4 17.1 33.1 24.0 59.1

Public utilities 0.8 2.7 3.0 3.4 6.6 55.0
Other 1.1 1.4 4.3 4.2 10.3 21.1
Total 7.3 10.5 24.4 40.6 41.1 135.2
Unallocated to particular Latin American countries
Banks 2.0 3.2 2.4 5.0 9.4 24.3
Shipping 2.9 4.8 3.0 3.1 6.0 18.3
Source: Stone (1987).
nineteenth-century Brazil, especially in the 1890s and after the 1940s when
rates of growth were two or three times those for the total population (see Table
2.2). It is as if most of the important infrastructure investments had been
completed before the turn of the century and urbanization had become the
focus of attraction of new investment (see Lewis, 1978, ch. 7).
The transition to a Republican regime in 1889 was intertwined with the
institutional strengthening of the military, the abolition of slavery, and the
erosion of the political coalition which supported monarchy and was heavily
tilted in favour of the north-east. The coup of 1889 resulted from an alliance
between the military and the Paulista coffee growers. The transition to the
stable political preeminence of the south-east was achieved towards the end
of the 1890s by a pact which assured the automatic election of the presidential
candidate of the party in power and the continuity of local political control by
the ruling oligarchy. It was possible to speak of a decentralized system as, in

contrast with the Empire, there was devolution of financial and administrative
power to the provinces. But this was a rather special type of decentralization,
being under strict central political control.14
Such an arrangement provided one of the elements required by the imple-
mentation of policies which tended to favour coffee interests, particularly the
accumulation of coffee stocks. The traditional interpretation, which underlined
the very strong influence exerted by the coffee oligarchy on economic policy
under the Old Republic, has been effectively qualified by the identification of
important contrasts in the policy stance adopted by different segments of
traditional agriculture (see Fritsch, 1988, passim). But even with important
qualifications, it is still true that to a large extent economic policy tended to
reflect the interests of the rising coffee oligarchy and the long-established
tradition of rent appropriation.
The historical precedent on policy formulation shows a tradition of rent-
seeking arrangements since colonial times. Colonial rule was marked by the
farming-out of a large number of contracts regulating the extraction of mono-
poly profits by private entrepreneurs from a wide range of activities. Trading
monopolies and detailed regulations constrained economic activity in the
colony. Manpower and land policies during the Imperial regime underlined
the ability of the rising coffee oligarchy to keep labour costs under control by
maintaining the slave trade, even if illegally for 20 years, and then mobilizing
subsidies partly paid by all taxpayers - mainly consumers of imports - and
blocking the legalization of land occupation by agricultural producers other
than those in large-scale plantations.15
What perhaps has not been sufficiently stressed in the literature is the whole
range of implications of the fact that Brazil was a price-maker in the coffee
market. Some of these implications are examined in later sections. This ability
to influence world coffee prices contrasts with the case of most relatively large
developing economies such as Argentina, Australia or Canada which faced
competition as price-takers in the markets for temperate agricultural products.
Brazil had, in principle, more degrees of freedom than other primary commod-
ity exporters.
Market power was used on several occasions to support world coffee prices,
sometimes by syndicates controlled by exporting houses, sometimes by the
Federal government, and sometimes by the state of SaÄo Paulo, most notably
in 1907±13 and 1923±30 in the Old Republic, but also after 1931. There was
practically no international cooperation in such support activities before the
end of the 1950s, with the exception of a short-lived agreement in 1936±37
with Colombia (see Ocampo, 1984, pp. 70±4). The result in the long-term, as
could be expected, was the gradual reduction of the Brazilian share of world
coffee output. The golden age of coffee `valorization' ± a word first used, as
recorded in the Oxford English Dictionary, in 1907 ± was from 1907 to 1930.
The first Brazilian valorization of 1907±13 was decided in the wake of the
considerable exchange rate appreciation which followed the Murtinho Plan of
1898, when the domestic currency rose from 7 pence to 12 pence, coupled with
a large expansion of coffee output in the 1890s and again after 1905. Provincial
authorities entered into a pact to restrain supply by purchasing stocks and to
freeze production. They needed federal backing to raise finance and added a
clause to their pact which demanded exchange rate stabilization. Foreign debt-
service was to be paid by an export tax which was thought would be mainly
borne by consumers, given the price inelasticity of coffee demand. After many
political difficulties (see Fritsch, 1988, ch. 2), coffee valorization came into
being, mainly controlled by foreign banks and coffee importing houses. It was
a success until it was disorganized by compulsory sales of coffee stocks in the
USA in 1912 due to legal action under the Sherman Act. In parallel with
valorization, Brazil adopted the gold standard, an experiment which survived
until the outbreak of the war in July 1914.
The valorization experiment relying on foreign finance ± half-hearted domes-
tically financed attempts were undertaken during the First World War ± was
repeated in the 1920s. But after the difficulties with USA legislation, stocks were
held near producing regions and responsibility for coffee `defence' was trans-
ferred to the state of SaÄo Paulo so that the federal government could retain a
free hand in relation to initiatives to raise finance in New York without fearing
the USA government veto, as happened early in the 1920s. In the wake of the
international trend to return to gold, Brazil adopted the gold exchange stand-
ard in 1926. So there was a close parallel between economic policy develop-
ments before the First World War and before the Depression, with an active role
played by the government. The lack of a laissez-faire tradition, which was
already well-established when the Republican regime was introduced in 1889,
was only reinforced by developments under the Old Republic.
The vulnerabilities of coffee valorization have been examined in the literature.
To the extent that it relied on foreign borrowing it was vulnerable to the cyclical
behaviour of international capital markets. The success of such schemes
depended on the ability to establish barriers to entry, that is, on successfully
limiting the planting of new coffee trees as this would make oversupply even
more serious and turn valorization into a snowball. The costs of price valoriza-
tion were paid only by Brazil as the dominant supplier; competitors were free-
riders as they simply reaped the benefits of higher prices at no cost. In the long
run Brazil's umbrella effect would naturally entail the erosion of its market share.
It is not easy to assess the impact of the coffee export economy on the
standard of living of the Brazilian population as there is very limited reliable
and comprehensive information on social conditions in Brazil before the 1940s.
But what exists underlines the widespread poverty. Average life expectancy at
birth in the 1930s was still 42.7 years (the extremes were 33.5 years in the state
of Rio Grande do Norte and 52 in Rio Grande do Sul), and infant mortality
stood at 158.3 per thousand live births (extremes of 201.1 and 119.4 in the
same states). The coffee states and, specifically, SaÄo Paulo were very near the
average (43.6 years and 154.7 per 1000 live births in SaÄo Paulo). However,
improvement by the end of the 1950s was considerable and concentrated in
the south and south-east. The illiteracy of the population over five years of age
was 61.2 per cent in 1940 (65.3 per cent for women), the first year for which
there is reliable data (see Brazil, 1990, p. 52,).16
There is much controversy on the direct effect of the expansion of coffee
production on the standard of living of agricultural labourers and small land-
owners, but there is no hard evidence favouring any of the contrasting inter-
pretations (see Stolcke, 1986, pp. 78ff ). The main mechanism working in favour
of an improvement of the standard of living was the reduction in the share of
employment in agriculture and a corresponding increase in industrial employ-
ment as the productivity in industry was about 3.5±4 times the productivity in
agriculture both in 1920 and 1940. The increase in industrial employment of
almost 1.3 million between these two years corresponded to more than 8.5 per
cent of the working population in 1940.
The implications of being the main world coffee supplier
The links between world coffee prices and the Brazilian exchange rate, or for
that matter world commodity prices and the exchange rate in any commodity
exporter which holds a substantial share of the relevant market, were ac-
knowledged early in the literature on Brazilian economic history.17 Brazilian
supply had an important impact on world coffee prices; due to the low price
elasticity of coffee demand, supply shifts had significant consequences on the
price level, and increased output meant a heavy fall in prices. Retention of
stocks could hold or raise world prices without much impact on quantities
demanded.
The fact that Brazil was the leading coffee producer and exporter implies that
production costs in Brazil had a significant influence in the determination of
world coffee prices.18 Foreign exchange rate fluctuations affected the supply
of coffee in Brazil and consequently world coffee supply due to the weight of
production in Brazil. In the short run, devaluation tended to weaken world
prices as there was an inducement to dump stocks in the market. In the long
run, the net effect of a devaluation will depend on the relative importance of
the impact on production costs and on revenues of coffee growers both instant-
aneous and expected in the long term.
In order to test the empirical relevance of this hypothesis, a standard
reduced form equation for the determination of world coffee prices was
estimated in logarithmic form, using annual data for 1852 to 1930. The Brazi-
lian share of world exports was substantial throughout the 1852±1930 period.19
In the basic specification, real coffee prices (PRICE) are a function of their own
lagged values and of the lag of a variable that tries to capture the supply±
demand balance in the coffee market (MARKET ). Here, this variable is con-
structed as the ratio of world coffee consumption to the sum of world coffee
supply and world coffee stocks. Therefore, an increase in this variable should
have a positive impact on real coffee prices. The inclusion of this explanatory
variable is justified by the special characteristics of the coffee market. Since the
product can be easily stored and production responds to prices with a lag of
several years, a standard model where supply and demand are functions of
current prices and determine the market price through a clearance condition
is not appropriate for the case of coffee (see De Vries, 1975). An additional
explanatory variable, the real exchange rate (RER), is included to capture the
increase in production costs associated with exchange rate fluctuations. The
variable is defined so that an increase in the index corresponds to a deprecia-
tion of the domestic currency and enters the equation with lags of one and five
years in order to capture both short-term and long-term effects. The reason for
the five-year lag is that at the beginning of the century production started, in
general, five years after coffee trees were planted, and some three-fourths of
total costs in the coffee sector were associated with fixed costs (see Rowe 1936,
p. 37).
Table 2.6 presents the main estimation results. Equation 1 displays the base-
line coffee price equation for the entire 1852±1930 period. The short-term
effects of the real exchange rate are not significant, neither are the effects of
the variable MARKET. Long-term real exchange rate fluctuations are shown to
have a significant effect on real coffee prices, with an estimated coefficient
significantly different from zero at standard confidence levels. A real devalua-
tion results in an increase in world coffee prices. In Equation 2 the short-term
RER variable is omitted and then the significance of the RER variable in the long
term is unfavourably affected. Results for subperiods such as 1852±1913 and
1930±60 proved to be disappointing. For the earlier period it is believed that
measurements may have been affected by the shortcoming of the statistical
series. Alternatively, the recession in the 1870s and early 1880s may also have
affected the results. For the period after 1930 the main difficulty is that the
exchange rate ceases to be as relevant as before, since foreign exchange controls
became the rule. However, for the golden age of coffee, the subperiod 1880±
1930 which is the core of this study, the results are much better. In Equation 3
of Table 2.6 it is shown again that the RER variable with a lag of one period is
not statistically significant. When RER 1 is omitted (Equation 4), all estim-
ated coefficients have their expected signs and are statistically different from
zero at standard confidence levels.
Table 2.6 Brazil: coffee price regressions, 1852±1930a
Variable Equation 1 Equation 2 Equation 3 Equation 4

1852±1930 1852±1930 1880±1930 1880±1930
CONSTANT 0.61 1.07 1.54 1.73

( 0.97) ( 1.99) ( 1.76) ( 2.25)
PRICE ( 1) 0.73 0.76 0.65 0.65
(5.98) (5.66) (4.80) (4.66)
MARKET ( 1) 0.14 0.16 0.54 0.56
(1.01) (1.09) (2.56) (2.62)
RER ( 1) 0.18 ± 0.09 ±
( 1.18) ( 0.41)
RER ( 5) 0.25 0.18 0.39 0.34
(2.05) (1.55) (1.94) (2.02)
Rho 0.29 0.36 0.26 0.29
(1.68) (1.82) (1.25) (1.43)
Adjusted R2 0.64 0.58 0.64 0.62
Number of observations 75 75 50 50
Standard error 0.16 0.16 0.17 0.17
a
Corrected for first-order serial correlation.
Notes: Figures in brackets are the t-statistic values of the estimations of the coefficients. Data are
adjusted to calendar years.
Sources: Coffee prices: imports into USA, United States (1975); world coffee production,
consumption and stocks: Bacha and Greenhill (1992), statistical appendix; domestic prices USA:
United States (1975); exchange rates: Brazil (1941b); average tariffs: computed from Brazil (1941b),
Brazil (1990) and Fritsch (1988); domestic prices Brazil: Goldsmith (1986), CataÄo (1992), Haddad
(1978) and Brazil (1990).
Much emphasis has been placed on `socialization of losses' as an important

mechanism to explain the political economy of the Brazilian export economy
in the pre-1930 period (see Furtado, 1965, ch. 28). But such an emphasis does
not take into account Brazil's share in the world coffee market. The gist of the
`socialization of losses' argument is that through exchange rate devaluation
coffee growers were able to significantly recoup the losses entailed by the fall
in the world prices of their commodity exports. This argument is likely to be
much more relevant for a small commodity exporter than for a price-maker
such as Brazil. In the case of a price-maker, the devaluation of the exchange
rate increases the amount of domestic currency generated by each unit of
foreign exchange received by exporters in the short run; but it also weakens
world prices denominated in foreign currency through the inducement to a
release of stocks. Robust results for the short-run effects have proved to be
elusive to capture. In the long run, however, it has been shown that foreign
exchange devaluation resulted in a lagged increase in world coffee prices
denominated in foreign currency. That this effect was strong has been empiric-
ally shown above.
The coffee economy: linkages and diversification
There was little scope for forward and conventional backward linkages in a
coffee-based export economy. Soluble coffee became important only in the
1950s and there is no mass market for processed coffee products. Traditional
backward linkages were also weak due to the rudimentary techniques used in
agriculture.20 Consumption linkages, however, were powerful and of special
interest. From rather early days there was involvement of coffee growers in the
building-up of export infrastructure, especially railways from the early 1870s,
and somewhat later, from the 1890s particularly, in the development of a
broadly-based import substitution industrialization process. This is a substan-
tial qualification of traditional views which underline the more or less perman-
ent opposition between the interests of coffee growers and industrialists as
providing the economic rationale for the accumulation of tensions, which
were eventually solved by a shift of policies in 1930 in favour of emerging
domestic industry and against coffee interests.
It is true that industrial interests in Brazil had considerably more political
weight in the Old Republic than in other developing economies. Industrial
interests were able to obtain important concessions from the government, in
particular a very protective tariff.21 It has been shown that due to the weight of
Brazilian coffee exports in world exports and to the price elasticity of world
coffee demand a high tariff in Brazil was transferred with a lag to world coffee
prices.22 The basic mechanism at work is through the increased production
costs in Brazil which, given Brazil's very high share of the world coffee market,
affect world prices.
Tariffs were also, of course, very important as a share of government revenue,
as alternative forms of taxation such as taxes on land and on income were
politically unpopular with the agricultural oligarchy, or difficult to implement
as in the case of excise or consumption taxes. During the later years of the
Empire the share of import duties in total central government revenue fluctu-
ated around 60 per cent. It fell slightly in the beginning of the century and
more substantially during the First World War. In the 1920s it was still around
40 per cent of total revenue.
What is argued here is not that high import tariffs were sought because it was
perceived that increased production costs would be transmitted to world coffee
prices. It is rather that once the political economy of protection, as for instance
lobbying by industrial interests, resulted in increased tariffs, there was no
significant deterioration in the performance of exports to justify countering
such tariff increases. Coffee growers were generally keen on maintaining low
production costs, as exemplified by their stance on labour costs since very early
in the nineteenth century. One cannot thus claim that a high tariff was adopted
because it was perceived that the foreign consumer would bear the increased
Table 2.7 Brazil: coffee price regressions with tariff, 1880±1930a
Variable Equation 1 1880±1930 Equation 2 1880±1913
CONSTANT 3.43 ( 3.54)

3.54 ( 3.48)
PRICE ( 1) 0.77 0.81
9.98 (8.96)
MARKET (-1) 0.48 0.43
2.77 (2.16)
RER (-5) 0.57 0.46
3.50 (2.11)
TRF (-5) 0.23 0.38
2.26 (3.58)
Adjusted R2 0.78 0.88
Number of observations 51 34
Standard error 0.17 0.13
a
Figures in brackets are the t-statistic values of the estimations of the coefficients.
Sources: See Table 2.6.
costs. It is simply that the policy of high protection was not countered by an
export lobby interested in maintaining low production costs as was the case in
most big primary commodity exporters.
The empirical relevance of this hypothesis for the period of this study is
tested with an extension of the framework for the determination of world
coffee prices used in the previous section. The analysis is based on annual
data for the subperiods 1880±1930 and 1880±1913.23 Real coffee prices are
explained by their own lagged values, the lag of the supply - demand balance
variable, the real exchange rate lagged five years, and the implicit tariff (TRF )
also lagged for five years. The lag in the TRF variable is explained by the same
reasons as the lag in the RER variable. As the results in Table 2.7 indicate, all
signs are as expected and all variables are significant at standard significance
levels for both periods. This means that Brazil was able to pass increased import
costs to higher world coffee prices.
Coda: the export economy, 1930±60
The second internationally-financed Brazilian coffee valorization cum foreign

exchange stabilization programme started to break down in mid-1928 with the
end of capital inflows. After the second half of 1929 coffee prices collapsed and
by mid-1930 Brazil had de facto abandoned the gold exchange standard. The
difficulties in the early 1930s were compounded by the important increase in
coffee output following the failure of the efforts of the Paulista authorities to
freeze output capacity in the late 1920s.
Government policies in the wake of the depression relied on a package which

included expenditure-switching policies, through devaluation and import con-
trols, and coffee price support which was taken over from the state of SaÄo Paulo.
It involved the purchase of coffee by the government, accumulation of stocks
and destruction of coffee equivalent to three good crops, over 70 million bags
mainly during the 1930s. Perhaps these policies were not so original in their
instinctive Keynesiasm (see Furtado, 1965, ch. 31), since other big spenders can
be found in Brazilian Republican history.24 But it has been shown that transfers
from federal government had an important role in financing the coffee price
support, and that the mildness of the depression in Brazil was partly due to
expansionary fiscal policies adopted by the government from 1931 (see Fish-
low, 1972, pp. 327±31). Besides purchasing excessive coffee, the government
reduced by 50 per cent the debt of coffee growers and stimulated output and
export diversification in the south-east. This was especially successful in the
case of cotton, as there was rising world demand with high prices assured by
USA support policies.
There is little doubt that 1930 represented a watershed in Brazilian economic
history. The importance of foreign trade and foreign capital was significantly
reduced as a proportion of output and investment. The share of domestic
industrial output in total supply rose, but commodity export sectors and espe-
cially coffee remained important as a share of output and as an outlet for the
much increased domestic industrial output. Export sectors remained vital as a
source of foreign exchange, especially so as capital flows dried up almost
entirely.
The increased role of the state, and specifically of the federal government, is
evident in coffee policy. But from another crucial point of view there was a
radical qualitative change of the role of the state, a change that has persisted
until present times. This was the intervention in the foreign exchange market.
The foreign exchange regime was to play a vital role as an instrument of
economic policy for the remainder of the period covered in this chapter. With
the exception of a short period in 1946±47, a government monopoly in the
foreign exchange market was exerted indirectly by the federal government for
the whole period. Policies varied considerably, but the common feature was the
maintenance of an overvalued exchange rate. This was the case during periods
when a single exchange rate regime was in place25 and also at times when
multiple rates were adopted.26
A single overvalued exchange rate required foreign exchange rationing to
cope with the excess demand for imports and the operation of the system of
distribution of cheap imports to economic agents. Both mechanisms played a
crucial role in industrialization based on import substitution. To absolute
protection for domestically produced goods was added secure access to cheap
imports of inputs and capital goods. The core of rent-seeking interest was thus
moved away from the export economy and in the direction of industry. The
rationale for maintaining a single overvalued exchange rate was complex. The
traditional fiscal argument was, of course, that a sizeable share of government
expenditure was indexed to the exchange rate and thus increased automatically
with devaluation, while government revenues relied importantly on imports
which had been curtailed following the Great Depression. The government
through its monopoly power in the foreign exchange market had access to the
relatively cheaper (official) foreign exchange rate. The other reason was the fear
that a further devaluation would weaken coffee prices even more.
It is reasonable to believe that after 1930 Brazil continued to be able to pass
through increased production costs entailed by high tariffs to coffee prices, but
empirical work is prevented by the lack of data on the tariff equivalents of non-
trade barriers, particularly import rationing. As a last resort, the foreign
exchange reserves±imports ratio was used as a proxy for such non-tariff barriers,
but the empirical results were not satisfactory.
Multiple foreign exchange regimes were introduced in recognition of the
undesirable impact of a single official exchange rate regime on non-traditional
exports and on the cost of essential imports. Since the 1930s the Brazilian
government tried to stimulate non-traditional exports by assuring sales of
foreign exchange at more favourable rates.
The Brazilian exchange rate remained nominally stable from 1939 to 1953 in
spite of a difference in inflation relative to the USA which reached almost 140
per cent. This was possible, in spite of balance of payments problems from 1947
on, because of the sharp rise of coffee prices after 1945, but more especially after
1949. Coffee oversupply in the 1930s and early 1940s had become coffee
scarcity by the late 1940s, and the high coffee prices were to persist until
1954. After 1953 the multiple exchange rate regime also included multiple
export and multiple import rates and the government also used the exchange
regime as a rudimentary import tariff, since the Brazilian specific tariff had been
eroded by inflation. The wedge between the average import and export rates
was an important instrument fiscal for the government in the second half of
the 1950s, especially so as it escaped congressional control.
In the second half of the 1950s, as exports declined continually, foreign direct
investment was attracted by heavy subsidies and room to reap monopoly or
oligopoly profits behind a wall of absolute protection. Coffee prices fell after
1954 and the coffee industry was to again face a period of oversupply. While
there was no return to the policy of burning coffee stocks, there was a policy of
fostering the eradication of coffee trees in an effort to control supply. At the
turn of the decade, coffee control was to become an international concern after
many years of single-handed supply restraints by Brazil.
It is not easy to assess the impact of overvalued exchange rates on coffee
interests. In the short term it was likely that at the same time there was a
reduction of income in domestic currency per unit of foreign exchange, while

world coffee prices were supported by an overvalued exchange rate. This link
was reversed in the long term. The net extraction of income from coffee growers
in the 1950s was not that evident. Although it is true that in the short run the
government skimmed away income from exporters by making compulsory the
tendering of a certain share of their export earnings at a grossly overvalued
exchange rate, there was also an important programme of transfers to the coffee
sector in an effort to reduce productive capacity.
Conclusions
Brazil was a price-maker in the international coffee market for the whole period
during which the economy can be reasonably described as an export economy.
This is a crucial element for understanding how policies were designed and
implemented. Even before coffee oversupply and the adoption of coffee valor-
ization policies there was no laissez-faire tradition in Brazilian economic his-
tory. Land and labour policies were designed and implemented to maintain the
profitability of the plantation-produced coffee export economy; valorization
depended directly or indirectly on financial guarantees by the federal govern-
ment. There was also complementarity between coffee price support and
Brazil's adoption of the gold and gold-exchange standards immediately before
the First World War and the Great Depression. There was, however, an inbuilt
mechanism in valorization which made inevitable the steady decline of Brazil's
share of the international coffee market as the support of artificially high prices
attracted high-cost competitive suppliers which otherwise would not have been
able to compete.
Government intervention was already important before 1930 also in relation
to protection of the domestic industry against the competition of imports. A
commercial policy based on a very high tariff made import substitution already
relevant from the turn of the century. The impact of such tariffs on coffee
production costs had as a consequence higher world coffee prices due to Brazil's
market power as the main producer and the price inelasticity of world coffee
demand.
The diversification of economic activity was the main source of improve-
ment of the standard of living. This was related to urbanization, the increased
role of government and the rise of industry. It was thus not surprising
that improvement in living standards affected regions very heterogeneously
with the north-east trailing badly behind the south-east, especially in the
1940s and 1950s. Until 1930, however, disparities were surprisingly small,
at a very low standard-of-living level. Coffee culture remained based on low-
wage manpower and concentrated in plantations with very little change in the
techniques used since the nineteenth century. This also applied to the use made
of land resources as there were no limits to the advancing land frontier incor-
porated in the coffee economy within the relevant time span. Traditional tech-
niques which involved slash and burn remained the rule.
The scope of state intervention became even wider after 1930 when the
federal government took over coffee policies. Exchange regimes based on
exchange rate overvaluation and multiple rates became the rule, and the
wedge between average export and import exchange rates became important
for fiscal reasons. In some periods the government imposed additional levies on
exporters of traditional commodities, especially coffee, sugar and cotton.
Until the 1930s the protection to industry relied until the 1930s on high
tariffs, but then increasingly on access to low-cost imported inputs and capital
goods coupled with absolute protection assured by exchange controls. Brazil
`made the consumer pay' for its commercial policy based on high tariffs before
1930, and it is likely that it continued to do so after 1930. Exports remained
concentrated in a few agricultural commodities, especially coffee, until very
late, as industrial exports except in very special conditions such as during the
Second World War were unable to compete in international markets. There was
no introduction of new commodity exports until the soya boom in the late
1960s and early 1970s.
Notes
1 The British share of Brazilian total imports fell less markedly, but in the 1900s it was
down to 30 per cent compared to more than 50 per cent before 1880.
2 Brazil imported some foodstuffs, especially before 1930, but only wheat was an
important item of the import bill.
3 Our own estimates. For pre-1913 GDP and deflator data see Goldsmith (1986), pp. 22±
3 and 82±3.
4 See Eisenberg (1974), ch. 1 for a general view of sugar production and exports in
Brazil.
5 See Canabrava (1984) on the short-lived cotton boom in SaÄo Paulo in the 1860s and
1870s.
6 Weinstein (1983) and Schurtz et al. (1925) are the standard sources on the rubber
boom and collapse in the Amazon region.
7 See Stein (1957) for the boom and downfall of Vassouras as a coffee region in the
nineteenth century.
8 See Dean (1995) chs 8±9, for the environmental consequences of coffee culture
expansion.
9 See Dean (1971) and Reis and Reis (1988) for land policies and labour policies in Brazil
and the fragmentation of oligarchic interests.
10 Rowe (1936), p. 29 suggests a much lower threshold of 4000 trees as compatible with
operation of a farm by a family without hiring labour. The importance of the small-
scale coffee production sector would be even smaller if this was the case.
11 See Hatton and Williamson (1994), pp. 7, 44. For immigration to SaÄo Paulo see
Holloway (1980), especially ch. 3.
12 See LalieÁre (1909), part III, for a comprehensive explanation of machinery used in
best-practice farms for the preparation of coffee.
13 Even if the suspiciously high stock of French capital is taken at face value.
14 See Cardoso (1975) for the classic treatment of this issue.
15 See Reis and Reis (1988), for a discussion of the sharply contrasting stance of different
segments of the agricultural oligarchy on land policies, slavery and immigration in
late nineteenth century Brazil.
16 To put these numbers in historical comparative perspective, infant mortality in
England and Wales was 154/1000 live births in 1900, and life expectancy at birth
was 48 years.
17 Granger causality tests show that for the 1850±1930 period as well as for all the
relevant subperiods one cannot reject the hypothesis that there is no causal link
between changes in real coffee prices and changes in the Brazilian real foreign
exchange rate. For early perceptions of the importance of the exchange rate in the
determination of coffee prices see Gudin (1933) and Williams (1934). As noted by
Furtado (1965), Wileman (1896) detected the link between coffee prices and the
exchange rate. Such a link was investigated empirically by Cardoso (1983).
18 For conditions of coffee planting and costs in the 1880s see LaeÈrne (1885), ch. X; for
current costs in the 1900s see LalieÁre (1909), Part 4; on the importance of fixed over
current costs see Rowe (1936), p. 37, and for detailed costs for plantations at different
stages of maturity of coffee trees in the 1920s see Rowe (1932), pp. 88±9.
19 During 1852±1930 the average Brazilian share of world coffee exports was about 51
per cent. It was about 55 per cent in 1880±1930.
20 See Hirschman (1958), ch. 6, and Hirschman (1981) for linkage taxonomies.
21 For the history of protection under the First Republic (1889±1930) see Villela (1993).
22 See Abreu, Bevilaqua and Pinho (1997) for a treatment of import substitution in Brazil
from the 1890s to the 1960s.
23 After the balance of payments crisis following 1930, tariffs are a poor measure of
protection since exchange controls became the rule.
24 There was no scarcity of big spenders, but EpitaÂcio Pessoa and the building of dams in
the north-east in the early 1920s can be singled out.
25 As it was in 1931±34, 1937±39 and 1947±53.
26 This was so in 1934±37, 1939±46 and 1953±64.
References
Abreu, M. de P. (1995) `British Business in Brazil, 1850±1950', mimeo, Rio de Janeiro.

ÐÐ (1994) À DõÂvida Externa Brasileira em Francos, 1888±1956', Revista de Economia
PolõÂtica, vol. 14 (4), October.
ÐÐ (1990) `Crise, crescimento e modernizacÄo Ëa autoritaÂria: 1930±1945', in M. de P. Abreu
(ed.), A Ordem do Progresso. Cem Anos de PolõÂtica EconoÃmica Republicana 1889, Rio de
Janeiro: Campus.
ÐÐ (1985) À DõÂvida PuÂblica Externa do Brasil. 1824±1931', Estudos EconoÃmicos, vol. 15
(2), May.
Abreu, M. de P., Bevilaqua, A. and Pinho, D.M. (1997) Ìmport Substitution and Growth in
Brazil, 1890s±1970s', mimeo, Rio de Janeiro.
Bacha, E. and Greenhill, R. (1992) 150 Anos de CafeÂ no Brasil, Rio de Janeiro: Marcellino
Martins and E. Johnston.
Brazil (1990) FundacËaÄo Instituto Brasileiro de Geografia e EstatõÂstica, EstatõÂsticas HistoÂricas

do Brasil. SeÂries EconoÃmicas DemograÂficas e Sociais de 1550 a 1988, 2nd edn, Rio de Janeiro.
Brazil (1941a) Departamento Nacional de CafeÂ, AnuaÂrio EstatõÂstico do CafeÂ 1939/40, Rio de
Janeiro.
Brazil (1941b) Instituto Brasileiro de Geografia e EstatõÂstica, AnuaÂrio EstatõÂstico do Brasil
Ano V 1939/1940, Rio de Janeiro.
Canabrava, A. (1984) O Algoda Äo Paulo 1861±1875, 2nd edn, SaÄo Paulo: T.A. Queiroz.
Äo em Sa
Cardoso, E. (1983) Èxchange Rates in Nineteenth-Century Brazil: An Econometric
Model', Journal of Development Studies, vol. 19 (2), January.
Cardoso, F.H. (1975) `Dos Governos Militares a Prudente-Campos Sales', in B. Fausto (ed.),
HistoÂria Geral da CivilizacÄo
Ëa Brasileira, tomo III, vol. 1, SaÄo Paulo: Difel.
CataÄo, L. (1992) À New Wholesale Price Index for Brazil during the Period 1870±1913',
Revista Brasileira de Economia, vol. 46 (4), October.
Dean, W. (1995) With Broadax and Firebrand: The Destruction of the Brazilian Atlantic Forest,
Berkeley: University of California Press.
ÐÐ (1971) `Latifundia and Land Policy in Nineteenth-Century Brazil', Hispanic American
Historical Review, November.
Delfim Netto, A. (1979) O Problema do CafeÂ no Brasil, reprint of 1959 edition, Rio de
Janeiro: FundacÄo Ëa GetuÂ lio Vargas.
De Vries, J. (1975) `Structure and Prospects of the World Coffee Economy', World Bank
Staff Working Paper no. 208, June.
Eisenberg, P.L. (1974) The Sugar Industry in Pernambuco. Modernization without Change,
1840±1910, Berkeley: University of California Press.
Ellis, H. (1971) Exchange Control in Central Europe, reprint of 1939 edition, Westport:
Greenwood.
Fishlow, A. (1972) Òrigins and Consequences of Import Substitution in Brazil', in L.E.
DiMarco (ed.), International Economics and Development. Essays in Honor of RauÂl Prebisch,
New York.
Fritsch, W. (1988) External Constraints on Economic Policy in Brazil, 1889±1930, London:
Macmillan.
Furtado, C. (1965) The Economic Growth of Brazil, Berkeley: University of California Press.
Goldsmith, R. (1986) Brasil 1850±1984: Desenvolvimento Financeiro sob um SeÂculo de Infla-
cÄo,
Ëa Rio de Janeiro: Harper & Row do Brasil.
Greenhill, R.G. (1977) `The Brazilian Coffee Trade', in D.C.M. Platt (ed.), Business Imperi-
alism, 1840±1930: An Inquiry Based on British Experience in Latin America, Oxford:
Clarendon Press.
Gudin, E. (1933) CaÃmbio e CafeÂ, Rio de Janeiro: Laemmert.
Haddad, C. (1978) Crescimento do Produto Real Brasileiro, 1900±1947, Rio de Janeiro: FGV.
Hatton, T. and Williamson, J.G. (1994) Ìnternational Migration 1850±1939: An Economic
Survey', in T. Hatton and J.G. Williamson (eds), Migration and the International Labor
Market 1850±1939, London/New York: Routledge.
Hirschman, A.O. (1981) À Generalized Linkage Approach to Development, with a Special
Reference to Staples', in A.O. Hirschman, Essays in Trespassing. Economics to Politics and
Beyond, Cambridge: Cambridge University Press.
ÐÐ (1958) The Strategy of Economic Development, New Haven: Yale University Press.
Holloway, T.H. (1980) Immigrants on the Land. Coffee and Society in SaÄo Paulo, 1886±1934,
Chapel Hill: The University of North Carolina Press.
LaeÈrne, C.F. Van Delden (1885) Brazil and Java. Report on Coffee-Culture in America, Asia and
Africa to H.E. the Minister of the Colonies, London/The Hague: W.H. Allen and Martinus
Nijhoff.
LalieÁre, A. (1909) Le cafeÂ dans l'eÂtat de Saint Paul (BreÂsil), Paris: A. Challamel.
LeCointe, P. (1922) L' Amazonie breÂsilienne, 2 vols, Paris: A. Challamel.
Lewis, W.A. (1978) The Evolution of the International Economic Order, Princeton: Princeton
University Press.
Lopes, Juarez (1968) Desenvolvimento e MudancËa Social. FormacÄo Ëa da Sociedade Urbano-
Industrial no Brasil, SaÄo Paulo: Companhia Editora Nacional.
Mitchell, B.R. and Deane, P. (1971) Abstract of British Historical Statistics, Cambridge:
Ocampo, J.A. (1984) `La economõÂa colombiana en la deÂcada del treinta', in J.A. Ocampo
and S. Montenegro, Crisis mundial, proteccioÂn e industrializacioÂn. Ensayos de histoÂria
econoÂmica colombiana, BogotaÂ: CEREC.
Palacios, M. (1983) El CafeÂ en Colombia. 1850±1970. Una HistoÂria EconoÂmica, Social y
PolõÂtica, Mexico: El Colegio de MeÂxico/El A Â ncora.
Reis, E. and Reis, E. (1988) Às Elites AgraÂrias e a AbolicËaÄo da EscravidaÄo no Brasil', Dados,
vol. 31 (3).
Rowe, J.W.F. (1963) The World's Coffee. A Study of the Economics and Politics of the Coffee
Industries in Certain Countries and the International Problem, London: HMSO.
ÐÐ (1936) Markets and Men, A Study of Artificial Control Schemes in some Primary Indus-
tries, Cambridge: Cambridge University Press.
ÐÐ (1932) Studies in the Artificial Control of Raw Material Supplies. No. 3 Brazilian Coffee,
Memorandum No. 34, Royal Economic Society, London, February.
Schurtz, W.L., Hargis, O.D., Marbut, C.F. and Manifold, C.B. (1925) Rubber Production in
the Amazon Valley, US Department of Commerce, Bureau of Foreign and Domestic
Commerce, Washington, DC: Government Printing Office.
Stein, S. (1957) Vassouras, A Brazilian Coffee County, 1850±1900, Cambridge, Mass.:
Harvard University Press.
Stolcke, V. (1986) Cafeicultura. Homens, Mulheres e Capital (1850±1980), SaÄo Paulo: Brasi-
liense.
Stone, I. (1987) The Composition and Distribution of British Investment in Latin America, 1865
to 1913, New York: Garland.
Summerhill, W.R. (1995) `Railroads and the Brazilian Economy before 1914', doctoral
dissertation, Stanford University.
United Kingdom (1987) Central Statistical Office. Social Trends no. 17, London: HMSO.
United States Department of Commerce (1975) Historical Statistics of the United States.
Colonial Times to 1970, Washington, DC: Government Printing Office.
Villela, A.A. (1993) `PolõÂtica comercial e importacÄes Ëo na Primeira RepuÂ blica: 1889±1930',
M.A. dissertation, Department of Economics, Catholic University, Rio de Janeiro.
Weinstein, B. (1983) The Amazon Rubber Boom, 1850±1920, Stanford: Stanford University
Press.
Wileman, J.P. (1896) Brazilian Exchange. The Study of an Inconvertible Currency, Buenos
Aires: Galli Brothers.
Williams, J.H. (1934) Àmerican Foreign Exchange Problems in Brazil, Argentina, Chile
and Uruguay, 1934', reprinted in Foreign Relations of the United States 1934, Washington,
DC: Government Printing Office.
3
Coffee and the Origins of Modern
Economic Development in Colombia
JoseÂ Antonio Ocampo and MarõÂa Mercedes Botero
Introduction
In comparison with Brazil, Mexico and the Southern Cone economies, Colom-
bia may be characterized as a latecomer to export and industrial development.
Throughout the nineteenth century, Colombia struggled to establish a solid
export base. Its efforts bore little fruit, however, partly because of the nearly
continuous political instability the country experienced throughout the cen-
tury, associated with the difficult process of building a national state after
achieving political independence in 1819. Moreover, the notorious geograph-
ical barriers which characterized the country tended to fragment the economy
into a set of semi-autonomous regions.
In a departure from the nineteenth-century legacy, the twentieth century
produced sustained economic development from the start. The basis for devel-
opment was the establishment of a coffee economy made up of small and
medium-sized producers in the west of the country. Based on coffee production,
and to a lesser extent on a few enclave sectors, Colombia experienced a major
export boom during the two decades prior to the Great Depression of the 1930s.
The expansion of coffee represented not only an appreciable increase in
foreign trade volumes, but also a qualitative change in Colombian develop-
ment patterns. The country's unstable export trade, which in the nineteenth
century depended on exceptional, transitory conditions in international mar-
kets, gave way to a stable and competitive export sector. More importantly,
coffee finally created a cluster of activities that were to break away definitively
from the pre-capitalist economic pattern the country had inherited from the
colonial period, a pattern which the export expansion of the nineteenth cen-
tury had barely touched.
Thanks to the expansion of domestic demand generated by the export boom,
the aggressively protectionist policy of General Rafael Reyes' Administration in
the second quinquenium of the century and the partial shutdown of import
55
56 Coffee and Economic Development in Colombia
markets during the First World War, the industrialization process took off.
However, manufacturing development was still in its infancy during the
1920s. Geographic fragmentation was undoubtedly the reason why manufac-
turing lagged behind in Colombia compared to other large and medium-sized
countries in the region. Nevertheless, the coffee boom, the establishment of the
basis of a modern transport network and the manufacturing experience accu-
mulated prior to the Great Depression created the foundations for the rapid
process of structural change that the Colombian economy was to experience
after 1930.
In this chapter, we analyse the features of Colombia's export expansion and
its relationship to economic development in the closing decades of the nine-
teenth century and the opening decades of the twentieth. The chapter is
divided into five sections, the first of which is this introduction. The second
briefly reviews nineteenth-century development. The third section explores the
most noteworthy aspects of the process of transformation that the Colombian
economy experienced in the first three decades of the twentieth century. The
fourth and fifth examine in greater detail the development of the coffee and
manufacturing sectors in the latter period.
The nineteenth - century legacy

Export development
Attempts to diversify the export base, which had began with the Bourbon
reforms at the end of the colonial era, persisted into the first decades of
the Republic. However, for a long while the results were frustrating. During
the first half of the nineteenth century, gold continued to represent three-
quarters of the country's exports. Only by mid-century, thanks to the privatiza-
tion of the tobacco monopoly and the introduction of steam navigation on
the Magdalena river, was the country able to break out of its colonial single-
export mould.
The stagnation which characterized the first half of the nineteenth century
was succeeded by a sustained increase in exports in the second half. Growth was
rapid from the early 1850s to the 1880s, during which period real exports
increased at an annual rate of 4 per cent (Table 3.1). Nevertheless, this period
was marked by an unstable export base. Expansion was, in fact, based on a series
of regional booms in agricultural, livestock, forestry and handicraft products
(tobacco, cotton, indigo, leather, quinine, rubber, straw hats, among others)
associated to a large extent with temporary shortages of the various products in
the international market. Most of the booms evaporated without creating a
stable productive base (Ocampo, 1984).
This period ended in a serious crisis in 1883±85, associated with the collapse
of quinine exports owing to the emergence of quinine plantations in the Far
JoseÂ Antonio Ocampo and MarõÂa Mercedes Botero 57
Table 3.1 Nineteenth-century Colombian foreign trade
Exports Purchasing Imports

power of
Value Volume exports Value Volume
(thousands (1865±70100) (1865±70100) (Thousands (1865±70100)
gold pesos) gold pesos)
1834/35±38/39 3 261.6 44 31a 2 985.3 37

1840/41±44/45 3 306.5 47 42b 3 501.5 48e
1854/55±57/58 6 353.0 82 102 3 767.2c n.a.
1864/65±69/70 7 394.0 100 100 6 419.7b 100d
1870/71±74/75 9 888.2 137 149 8 818.7 141
1875/76±77/78 10 105.5 127 171 7 713.2 142
1878/79±80/81 13 689.1 168 247 10 527.3 191
1881/82±82/83 15 430.5 n.a. 280 11 929.8 231
1888±91 12 165.1 170 255 12 119.0 286
1894±97 16 533.3 n.a 385 14 872.3 305f
1898 19 154.1 269 447 11 052.0 397g
a
Based on purchasing power of exports in terms of textiles, with respect to 1855±58.
b
Excludes 1844/45.
c
1855/56±1858/59.
d
Excludes 1864/65.
e
Based on volume of imports for 1845±49 and trends in textile prices.
f
1892±95.
g
1896±98.
Source: Ocampo (1984). The value of imports for the first period has been reestimated using original
sources.
East. From the late 1880s to 1898 there was another phase of rapid export
growth (5.5 per cent annually between 1888±91 and 1898) based on the only
two products that showed sustained growth in the second half of the century:
precious metals and coffee. Both booms ended in bust around the turn of the
century. Nonetheless, the first played a key role in the economic development
of Antioquia, which from the end of the century onwards was to become one of
the pillars of modern industrial development in Colombia. By bringing coffee-
growing to the west of the country, the second was to lay the foundations for
export development in the twentieth century.
The expansion of gold and silver production in Antioquia and other regions
of the country in the 1870s reversed the downward trend in precious metals
extraction that had begun at the end of the colonial period. The freeing of the
slaves in mid-century had dealt the final blow to the already weakened slave-
based economies of ChocoÂ and Cauca in the Colombian Pacific. By contrast,
starting in the eighteenth century, gold-mining had become the mainstay of
economic expansion in Antioquia. This process, far from slowing down after
independence, continued throughout the nineteenth century and underwent
significant modernization. According to Brew (1977), vein-mining was unques-
tionably a seedbed of technical innovation and a channel for modern industrial
development in Antioquia.
Gold-mining in Antioquia was basically of two kinds (Brew, 1977; M.M.

Botero, 1994 and 1995). The first, with strong colonial roots, was small-scale
mining that took the form of panning in alluvial placer deposits. The second,
which throughout the nineteenth century assumed increasing importance,
involved the formation of corporations, most set up with local capital, which
made the major investments in vein-mining. Some of them were to become
large firms with big payrolls and substantial investments in modern machinery.
Two of them employed over a thousand workers and had sales of nearly one
million pesos annually at their production peak in the last decades of the
century ± equivalent to over 5 per cent of total Colombian exports at that
time. Some of these companies were foreign-owned. However, the capital of
the El Zancudo company, which established itself as the largest company in
Antioquia and the country by the end of the century, was wholly owned by
Colombians.
The modernization of the mining industry was not limited to the extraction
phase. From 1850 onwards, advances were made in the treatment of the raw
ores, that is, casting and assaying. Some of the large mining companies, such as
El Zancudo and the Western Andes Mining Company, which exploited the
Marmato mines, installed their own casting and assaying plants in the mining
areas. Their production volumes justified the installation of industrial facilities
for ore processing.
For most of the companies, however, that kind of investment was beyond
their means. In order to process the output of their mines, therefore, they were
obliged to turn to the facilities installed in MedellõÂn by large commercial
metals-exporting firms. In 1858, the Restrepo brothers founded the first estab-
lishment in MedellõÂn for assaying and casting metals.
By the 1880s, three such laboratories were already in operation in MedellõÂn.
These establishments provided the owner of the ore with an assay slip stamped
with the seal of the laboratory, a certificate attesting to the purity of the gold
and its value in pounds sterling. Some of the large and medium-sized mining
companies of Antioquia and miners located in more remote areas began send-
ing their ore to these laboratories to have it cast, assayed and then sold to the
export firms. To be able to export ingots properly cast and assayed was a great
advantage to the exporters. Before the establishment of the laboratories, traders
who sent unrefined gold to Europe sometimes suffered losses when the ore
proved to be of poor quality.
Gold production was thus controlled from MedellõÂn by big trading-banking
firms with ties to international markets. These firms arranged for the procure-
ment of all the operating supplies the mining companies needed, in return for a
commission. They were also authorized by the mining firms to make the
decision that seemed most advisable at any given time, either to sell the gold
bars in MedellõÂn and deposit the funds in a local bank, or to sell the gold
abroad. This division of labour worked to the advantage of the big MedellõÂn
merchants, who even ended up managing the ore produced by foreign firms.
The only exception to the rule were some of the large locally-owned mining
companies, who sold gold bars (and/or bills of exchange) in MedellõÂn to finance
their operating costs or exported them to their own agents abroad. There is, on
the other hand, no indication that foreign capital was involved in the market-
ing of the gold.
Along with the expansion of mining came development in other areas of
activity. As the region's capacity to pay for imports increased, other commercial
firms appeared in the state ± later departmental ± capital and some other
localities. MedellõÂn became an important mercantile and financial centre, and
between 1870 and 1883, three banks serving the public and seven private bank-
ing firms were established. In addition, commercial cattle-raising underwent
substantial development, since meat formed part of the miners' staple diet.
The laws and official policy of the state of Antioquia offered great support to
the mining industry. The regional governments set about improving and devel-
oping the transport network, constructed a telegraph system linking MedellõÂn
with the mining districts, promulgated a mining code and established a mint.
Lastly, and perhaps most importantly, the development of mining gave impetus
to technical education and the training of a workforce with the skills to handle
modern machinery. These elements were crucial for the industrial development
of MedellõÂn later on.
As for coffee, in 1860 it was still a minor item in Colombia's foreign trade,
representing only 4 per cent of total exports.1 From then on, however, coffee
production underwent three successive booms in the nineteenth century
which generated a significant shift of the production frontier in the interior
of the country. The first of these expansions took place during the 1860s.
Exports increased to some 100 000 bags, but production remained concentrated
in North Santander, near the Venezuelan border, where coffee-growing had
been introduced in the early decades of the Republic.
The second expansion occurred during the 1870s. Coffee exports increased
from 100 000 to 200 000 bags, and became one of the country's major export
staples. By the end of the decade it accounted for more than 20 per cent of
Colombia's external sales. This period of expansion involved a major shift in
the production frontier. Although North Santander continued to increase acre-
age planted and raised its output to 120 000 bags per year, most of the new
production came from Santander and Cundinamarca, which by mid-decade
were producing 60 000 and 40 000 bags, respectively (Table 3.2). In both these
regions the expansion was interrupted by falling international prices in the
early 1880s and later by civil war.
Soon after public order was restored in the mid-1880s, the largest coffee
boom of the century began. The country's coffee production expanded rapidly;
Table 3.2 Coffee production by department, 1874±1932 (thousands of 60 kg bags of

green coffee)
Departments 1874 1890 1900 1913 1932
Antioquia 1 6 90 185 618

a a a
Caldas 199 1 004
Cundinamarca 3 40 200 200 406
North Santander 95 120 150 200 270
Santander 10 60 120 105 150
b b b
Tolima 60 448
Valle del Cauca 1 4 20 50 354
Other 70 204
Total 110 230 580 1 069 3 454
a
Included in Antioquia.
b
Included in Cundinamarca.
Sources: 1874±1900: Ocampo (1984); for 1913: Bell (1922); for 1932: Coffee census, Revista Cafetera de
Colombia, nos 34±35, January±February 1932.
USc/lb (1980 dollars)
240
220
200
180
160
140
120
100
80
60
40
1830
1833
1836
1839
1842
1845
1848
1851
1854
1857
1860
1863
1866
1869
1872
1875
1878
1881
1884
1887
1890
1893
1896
1899
1902
1905
1908
1911
1914
1917
1920
1923
1926
1929
1932
1935
1938
1941
1944
Figure 3.1 Real external price of coffee (USc/lb in 1980 dollars)

Source: Ocampo (1984).
by the time the last and most serious of the nineteenth century civil wars broke
out, the Thousand Days' War (1899±1902), coffee output had reached 600 000
bags and represented nearly half of Colombia's exports.
The expansion coincided with a period when real international coffee
prices were at their highest levels ever (Figure 3.1), which was no doubt the
fundamental reason for the boom. Other factors, however, influenced the
process. As a reflection of the general crisis in Colombia's foreign trade in
the early 1880s, the declining price of silver in international markets and the
subsequent introduction of paper money, the peso underwent a sharp devalua-
tion. During the initial phase of this process, coffee production costs tended to
decline in terms of gold, which encouraged planting.2 No less important was
the fact that landowners had access to credit from foreign trading companies
on terms that were attractive at that time: an average interest rate of 6.6 per
cent per annum ± with 1.5 or 2 additional points by way of commission ±
maturity of two years and repayment in coffee.
The boom of the 1890s was the most diversified of the century in regional
terms. The key expansion area was western Cundinamarca ± and localities in
neighbouring Tolima ± within the sphere of influence of BogotaÂ. In addition,
production began to expand appreciably in Antioquia, which was producing
some 90 000 bags by the end of the century (Table 3.2).
In Santander and North Santander, coffee production took a variety of
forms and lacked some of the semi-servile aspects that characterized labour
relations in the large plantations in Cundinamarca. This divergent develop-
ment reflected the history of each of these regions. In Cundinamarca, the
evolution of social relations was imprinted by three centuries of domination
of the indigenous population. In the Santanders, in contrast, smallholders
and craftsman had played an important role in the economy since colonial
times.
In Cundinamarca, although small plantations were not unknown, the pre-
dominant model was the large estate, or hacienda, with a permanent workforce
consisting primarily of tenant farmers who were in some respects in a servile
position vis-aÁ-vis the landowner. The tenants constituted the core of the
hacienda workforce. In exchange for a dwelling and an allotment of land for
subsistence farming, the tenant was required to work a certain amount of time
on the plantation, usually at half of what would ordinarily be a day's pay. For
the seasonal work of harvesting, the plantation depended on temporary day
labourers, most of whom, like the tenants, came from the high mountain
plateaux, or altiplano (Palacios, 1983).
Although this form of social organization had elements of serfdom and relied
upon coercion of the labour force with the support of local political authorities,
it cannot be described as a step backward with respect to the labour arrange-
ments on the altiplano haciendas, as some authors have suggested (Kalmanovitz,
1985). On the contrary, Palacios (1983) has shown that, due to the mix of
permanent labour scarcity with an open agrarian frontier, the basic tendency
in western Cundinamarca was towards the strengthening of a peasant economy
on the plantations, a situation that was to lead to the agrarian conflicts of the
1920s (see below).
The processing methods ± depulping, washing, drying and hulking ±

remained generally backwards, although in the last two decades of the nine-
teenth century some technological improvements were introduced in the larger
plantations. Cundinamarca became the leader in processing methods, at least
until 1920. The least developed process in the country was hulking, in which
there was still widespread use of primitive techniques.
Transportation was a major obstacle to the development of the coffee indus-
try in the nineteenth century. The lack of modern transportation was much
more detrimental to coffee growers than to producers of other export products,
for the simple reason that a ton of coffee was worth much less in Colombian
ports than the same weight of tobacco, quinine or precious metals, the other
main export staples at the time; the same freight cost thus cut much more
deeply into coffee profits. The cost of transporting and trading the coffee thus
represented a substantial percentage of the price, especially in the early years.
In the nineteenth century, coffee marketing was not a specialized business.
The trading firms dealt with a variety of import and export products, employing
commission agents in foreign ports for the purpose. Some large plantation
owners, however, acted as their own exporters and even managed to create
brand recognition in international markets for the coffee from their particular
plantations.
The collapse of coffee prices at the turn of the century (Figure 3.1) and the
Thousand Days' War put an end to this phase of coffee development. The
profitability of coffee-growing, which had been high during the 1890s, evapor-
ated. Moreover, the impossibility of harvesting all the coffee and keeping
plantations in good condition during the war led in many cases to irreparable
damage. All the coffee plantations in the country suffered from the difficulty of
exporting during the years of conflict, and during that time the beans were
stored in less than optimum conditions. Landowners who had taken out for-
eign loans to extend their plantings found themselves in dire financial straits as
a result, and some lost their properties. At the start of the new century, it was
evident that even if the traditional coffee plantations would not immediately
die out as other nineteenth-century exporting ventures has done, they would
nevertheless be facing a gradual decline.
Exports, economic policy and their impact on the rest of the economy3
At the start of the twentieth century, both the large precious metal mining
operations and the large coffee plantations suffered a collapse, bringing to a
close the longest-lasting export experiments of the nineteenth century. The
inability to develop a solid export base was undoubtedly the main cause of
Colombia's slow economic development in the last century, which left the
country on the periphery of international flows of trade, capital and labour.
Despite these limitations, the expansion of exports in the second half of the
century did create general economic growth, which manifested itself in a

number of different ways: in significant urban growth in the last three decades
of the century,4 the accumulation of capital in the largest cities, the increase in
the living standards of the upper classes, the establishment of modern banking
activities, the introduction of some technological innovations, particularly in
mining, and in other ways.
Moreover, the nineteenth century was characterized by rapid population
growth ± 1.5 per cent annually ± which fuelled an active internal migration
throughout the country. By the end of the century, population growth and
internal migration had filled in many of the empty spaces separating the main
urban centres. The most important movement of population was the Antio-
quian migration to the rich volcanic soils of the central Andean range, which
was to become the heartland of coffee development in the twentieth century
(Parsons, 1968; LoÂpez Toro, 1970). This wave of settlement also meant that by
the turn of the century the area that would become the `golden triangle' in the
new century, formed by BogotaÂ, Cali and MedellõÂn, was relatively densely
populated. In the north, Barranquilla was transformed from an unimportant
village to the country's main shipping port (Posada-CarboÂ, 1987).
The fact that the transportation system underwent few improvements was
perhaps the clearest evidence that the export booms of the nineteenth century
made only a limited contribution to the country's overall development. The
expansion of coffee benefited to some extent from the mid-century tobacco
boom, which had led to some important improvements in the internal trans-
port system, notably steam navigation on the Magdalena river. Towards the end
of the century, significant progress was also made in expanding railroads and
port facilities, which lowered freight costs. Nevertheless, by 1898 there was still
only 498 kilometres of rail track scattered throughout the territory (McGreevey,
1971). Until well into the twentieth century, the mule continued to be the chief
means of transport in the country; until early in the new century, mules were
even the means by which the first industrial firms transported their imported
machinery, with considerable difficulty and loss.5
In a country of Colombia's geographic complexity, the meagre development
of modern transport not only perpetuated but at first even exacerbated the
strong regional fragmentation, which the well-known Colombian writer Luis
Eduardo Nieto Arteta called the àrchipelago economy'. In many cases, the
segmented infrastructure that began to develop in the country made it, in
effect, much easier and cheaper to import goods from abroad than from other
regions within the country, even from places only a few kilometres away, if the
only connecting link was along a rough mule track.
The economic ideas of the early governments of the Republic were generally
protectionist. This was reflected not only in tariff policy, but in government
attempts to promote, in the 1830s, some new industries by granting them
temporary monopolies on the use of technologies and by authorizing subsidies

and loans. Even during the liberal period that began shortly before the middle
of the century, when free-trade ideas flourished, tariffs were not necessarily low.
On the contrary, customs duties tended to increase and by the end of the free-
trade period (in the early 1880s) stood at their highest levels in the century,
considerably higher than during the protectionist period prior to 1847 and the
final decades of the century (Figures 3.2 and 3.3).
60
% 55
50
45
40
35
30
25
20
15
10
1867/8
1871/2
1875/6
1881/2
1850/1–54/5
1865/6
1873/4
1886
1888
1890
1892
1894
1896
1898
1900
1902
1904
1906
1908
1910
1912
1914
1916
1918
1920
1922
1924
1926
1928
1930
1932
1934
1936
1938
1940
1942
1944
1832/3–35/36
1883/4
1879/80
1877/8
1869/70
1840/–43/4
Figure 3.2 Average collected tariff, 1833±1945

Source: Ocampo and Montenegro (1984).
200
% 180
160
Raw cloth
140
120
Ordinary
cloth
100
80
60
40
20 Coloured
0
1911
1931
1921
1941
1828–31
1848–51
1905
1929
1913
1915
1933
1935
1937
1919
1925
1945
1917
1927
1939
1885–86
1943
1871–73
1890–92
1923
1856–59
1947
1842–44
1865–67
1896–98
1877–80
1835–37
Figure 3.3 Equivalent ad-valorem tariffs on cotton textiles, 1828±1947

In fact, all the major liberal tariff reforms ± 1847, 1861, 1870 and 1873 ± were
followed a few years later by large customs surcharges that in effect consider-
ably raised taxes on imports. The basic reason for the contradiction between
free-trade principles and the actual trend in tariffs was the increasing fiscal
dependence on customs revenues (Ocampo and Montenegro, 1984). Except
for a few brief periods during which the revenues from tobacco, salt, confis-
cated church property or issuance of money ± during the Thousand Days' War ±
exceeded customs revenues in importance, over the rest of the period in ques-
tion customs were the national government's chief source of revenue.
In the 1880s, under Rafael NuÂ n Äez' Regeneration movement, protectionist
ideas reemerged with renewed force. Contrary to what is generally believed,
however, the reforms carried out during the Regeneration in the long run
actually reduced existing duties. The basic explanation is that the specific ad-
valorem duties levied on imports were eroded by the inflation that the country
experienced following the devaluation first of silver and later of paper money.
Although the ad-valorem duties reached their height in the period 1885±86,
the average tariff and most of the ad-valorem duties stabilized in the rest of the
decade at levels comparable to those at the close of the 1870s and fell in the
1890s. The supposed protectionist effect of tariff reforms under the Regeneration
was thus largely illusory, although it did result in greater tariff dispersion, which
generated greater effective protection for the same level of nominal protection.
Trends in Colombian tariff policy in the nineteenth century thus had very
little to do with modern controversies about free trade and protectionism. In
the absence of an alternative fiscal base, the need for a sufficient level of public
revenue was the main factor determining tariff movements.
Economic policy played a very modest role in the decline of the handicraft
industry, especially textiles, during the nineteenth century. The backwardness of
the transport system and the effective increase in tariff protection tended to
counteract the effect of sharp reductions in international prices for some
imported goods, especially textiles. Moreover, as some nineteenth century ana-
lysts pointed out at the time, economic growth offered an opportunity of expan-
sion for handicraft activities that were not in direct competition with imports to
expand ± dressmaking, furniture production and printing, to name a few.
The marked fragmentation of the domestic market, the protection generated
by high transport costs and the attempts at protectionism in the early years of
the Republic encouraged a number of experiments in industrialization. Most of
these pioneering efforts were unsuccessful or prospered only on a modest scale.
The first initiatives of importance were undertaken in and around BogotaÂ in the
1830s, when a number of small factories ± producing textiles, pottery, glass,
paper and iron ± were set up. Nearly all these enterprises quickly failed owing to
a lack of markets, adequate raw materials and skilled labour (Ospina VaÂsquez,
1955; Safford, 1965).
Certain agricultural processing industries, such as sugar mills, made some

technological advances during the century, as evidenced by their importance
as a market for the iron-working enterprises of the time. One of the sugar
mill ventures, Manuelita, became the only important firm established prior
to 1880 to make the transition to a large manufacturing enterprise. The
firm was founded by a foreign immigrant in Palmira ± Valle del Cauca ± in
1864, but became a modern sugar refinery only after the turn of the century
(Eder, 1959).
Industrial ventures became more common in the last two decades of the
nineteenth century, when firms emerged that were to make a successful transi-
tion to modern industrial development. The economic expansion of the pre-
vious decades undoubtedly paved the way for the establishment of these
enterprises, chiefly in MedellõÂn, but also in BogotaÂ and Barranquilla. The ven-
tures were financed not only by local, but also, in some cases, by foreign capital,
especially in the latter two cities. At the turn of the century, however, these
businesses were for the most part still modest in size, little more than semi-
mechanized workshops.
The strong swings in exports characteristic of the past century generated
significant macroeconomic stresses. Nevertheless, the Colombian economy
from the mid-1850s to the early 1880s displayed the necessary flexibility for
the functioning of a metal standard, which despite its legal definitions was
in effect a silver standard. These conditions included sharp fluctuations in
imports, great flexibility in money demand, rapid cyclical adjustments
in domestic bank rates ± after their appearance in the 1870s ± and flexible
domestic prices that fell during economic downturns (Ocampo, 1994).
The main difficulty lay in the fiscal instability that such an adjustment
required. Fiscal problems led to the issuance of treasury bills in the early
1860s, during a period of external crisis and internal political unrest. Similar
circumstances led to the abandonment of the silver standard in 1885 and the
adoption of paper money issued by the Banco Nacional as legal tender. The bank
had been created in 1880 as part of the Regeneration agenda of economic and
political reforms, which also, as we have seen, included protectionist ideas.
The inflationary period that followed the introduction of paper money was
short, and the inflation moderate. In fact, during most of the 1890s, not only
was the inflation rate low but there was an explicit plan to return to the silver
standard. This goal was on the point of being reached when the Thousand
Days' War broke out. The issuance of paper money then became the national
government's major means of financing, so that inflation rates went into three
digits during the critical years of the war. The Colombian economy thus closed
the nineteenth and entered the twentieth century suffering not only from a
serious collapse of its coffee industry, but from the monetary upheaval and
physical devastation caused by civil war.
Structural transformation of the Colombian economy during the

first three decades of the twentieth century
The return to peace and the rapid economic stabilization achieved after the
Thousand Days' War, together with the vigorous export promotion and protec-
tion policy introduced by General Rafael Reyes (1904±09), began a period of
rapid economic growth which lasted until the late 1920s and laid the founda-
tions for the structural changes that were to take place in the country's eco-
nomy from the Great Depression onwards.
The economic growth that took place during this period was based on a
strong export expansion. Between 1905±09 and 1925±29, real exports grew at
an annual rate of 7 per cent. The basis of that expansion was the export of
coffee, which grew by 7.3 per cent per year during this period (Table 3.3). Two
other export products contributed to the expansion: bananas, from the first
decade of the century, and oil, starting in the 1920s. Other export products,
including precious metals, lagged behind. The export structure which had
remained diversified until the turn of the century thus became much more
concentrated which, of course, made the whole economy much more vulner-
able to coffee price fluctuations but reflected the great competitiveness of the
coffee industry that developed during the period.
There are no precise estimates of overall economic growth during the period.
On the basis of very incomplete information, Maddison (1995) and Hofman
(1998) estimate that the Colombian economy grew at 3.9 to 4.5 per cent a year
between 1900 and 1929,6 which would indicate that exports made up 13.5 to
15 per cent of GDP in 1905±09 since existing figures show that share having
reached 24 per cent in 1925±29 ± estimated at 1975 prices. Growth was prob-
ably in the higher part of the range;7 it was certainly higher than the regional
and worldwide and regional averages for the first three decades of the century ±
2.2 per cent and 4.1 per cent respectively, according to the same authors ±
although it was lower than that of the fastest-growing Latin American eco-
nomies.
However, economic expansion was uneven over time; its ups and downs were
basically due to fluctuations in the external sector. The growth which charac-
terized the period from 1905 to 1913 was reflected in annual growth of 22 per
cent in deposits in the three major banks of BogotaÂ (Romero, 1994). The period
ended with a sharp increase in the terms of trade and external purchasing
power (see Figure 3.1 and Table 3.3). During that period, as had been the case
during the boom periods of the nineteenth century, there was an influx of gold
coins which facilitated the growth of currency in circulation.
From 1914 to 1922, on the other hand, the country went through a period of
slowdown and sharp fluctuations associated with political and economic
events at the international level. The terms of trade were highly unstable during
68
Table 3.3 Exports and imports, 1900±29
Period Coffee exports Export volumes Purchasing Composition of exports (in Import volumes Machinery
(thousands of (1925±29100) power of exports dollars at current prices)b (1925±29100) imports
bags) (1925±29100) (1925±29100)
Coffee Mineralsa Other
1900±04 542
1905±09 604 25.7 39.0 19.7 41.3
1910±14 837 37.3 38.4 47.4 16.3 36.3 29.0
1915±19 1 244 50.2 39.9 55.1 7.9 37.1 25.4 6.8
1920±24 1 906 74.3 52.8 75.5 5.6 8.9 41.1 26.4
1925±29 2 451 100.0 100.0 67.7 17.9 14.3 100.0 100.0
a
Oil, coal, nickel and gold.
b
Excludes other.
Sources: Coffee exports: 1900±10, Ocampo (1984); 1910±29, National Federation of Coffee Growers; export volumes and exports' purchasing power: Ocampo and
Montenegro (1984); exports as % of GDP: CEPAL (1957); composition of exports: Foreign Trade Yearbooks and records of purchases of monetary gold by the Banco de
la RepuÂblica; import volumes and imports of machinery: Ocampo and Montenegro (1984).
that period, and external purchasing power was on average relatively stagnant.
The annual rate of growth in deposits at the main BogotaÂ banks was only 2.6
per cent, at a time when their productive assets were stagnant (Romero, 1994).
In critical years, particularly 1914 and 1920, they suffered severe contraction.
During the first years of that period, surpluses accumulated by the Amortization
Board were used to finance public spending, postponing the possibility of
giving metal backing to the currency in circulation. Export restrictions imposed
by the belligerent states were reflected in a balance of payments surplus, but the
inconvertibility of their currencies, which continued during the postwar per-
iod, prevented the surplus from increasing the currency in circulation; a tem-
porary appreciation of the peso therefore ensued. Faced with fiscal problems
again in 1919 and 1921, the government issued treasury bonds. In addition, the
sharp postwar fluctuations in international prices led to the bankruptcy of
major coffee traders and some banks.
Beginning in 1923, the economy entered a phase of faster expansion led by
strong export growth and improved terms of trade, together with the payment
by the USA between 1923 and 1925 of US$25 million in compensation for the
independence of Panama, and the receipt of US$200 million between 1926 and
1928 in net foreign borrowing. The latter was equivalent to 35 per cent of
exports for 1925±29. The country finally had access to the international capital
market, a source of income that had been unavailable for more than a century
owing to the country's inability to pay off its foreign debt left over from the War
of Independence (Junguito, 1995). All this was reflected in an unprecedented
boom in imports, and in levels of public and private capital formation which
had never before been experienced in the country (Table 3.3). Despite increased
imports, the balance of payments surplus led to substantial growth in interna-
tional reserves and currency in circulation; the latter grew by about 20 per cent
per year from 1923 to 1928. Annual GDP growth was 7.4 per cent in 1925±29,
the first years for which such estimates are available. Thus, the country ended
this phase of òutward-oriented development' with a spectacular surge in for-
eign borrowing, imports and investment (CEPAL, 1957; Ocampo and Montene-
gro, 1984).
Urbanisation progressed rapidly during this period: while the country's popu-
lation grew at an annual rate of 2.1 per cent between 1912 and 1938, the 12
main cities grew at a rate of 3.9 per cent. There was strong growth in all urban
activities, including manufacturing (see later), commerce, banking, construc-
tion and government services.
Transport also made notable progress. The railway network expanded con-
siderably, steam-powered shipping flourished and the first highways were built.
In the early 1930s, Colombia's railway network (3262 km) was 6.6 times more
extensive than in the late nineteenth century (McGreevey, 1971). The Pacific
and Caldas railways had created the first integrated network, linking the
coffee-growing areas to the Pacific coast. The opening of the Panama Canal in
1914 justified the development of a modern seaport on the Pacific coast, which
was the natural outlet for the new coffee-growing areas in the west of the
country. Steam-powered shipping on the Magdalena and Cauca rivers and the
gradual introduction of automobile transport also contributed to this revolu-
tion in domestic transport systems. However, most infrastructure investments
came late, following the 1925±29 investment boom. Meanwhile, although the
modern transport network was still far from complete and its separate compo-
nents had been developed mainly to link the various regions of the country to
the outside world, Colombia had, for the first time, the basis for a minimum
level of domestic market integration.
In any event, the country's productive structures were still very backward at
the end of the coffee boom. In 1925±29, manufacturing accounted for less than
8 per cent of GDP, estimated in 1975 prices, whereas the share of the agricul-
tural sector was still 48 per cent. In comparison with international patterns
there was relative òver-expansion' of agriculture and a considerable lag in
manufacturing (London Äo, 1989). This was doubtless attributable to the geo-
graphical complexity of Colombia and to the very late development of modern
means of transportation. Manufacturing could not thrive in an economy
which, aside from the low standards of living of the population, was still
characterized by a considerable fragmentation of the domestic market.
The expansion of government activities was a decisive factor in the trans-
formation process. However, as was common at the time, the state was still
relatively small and had limited means of action. The most significant state
reform was that which took place in 1923 as a result of the Kemmerer Mission.
In that year, the Banco de la RepuÂblica was founded; it was the first central bank
in the modern sense, in that it had a monopoly on the issue of money, acted as
the central repository for bank reserves and was the lender of last resort (Banco
de la RepuÂblica, 1990). Its creation was also the culmination of a long transition
towards a stable gold standard, which had begun after the Thousand Days' War;
the process had involved various mechanisms for guaranteeing the convert-
ibility of the currency and had involved a number of initiatives to replace the
nineteenth century Banco Nacional and the Banco Central created in the early
twentieth century by General Reyes (Avella, 1987). The founding of the new
central bank was accompanied by the creation of a banking supervisory body,
the Superintendencia Bancaria; until then, private financial institutions had not
been regulated. Finally, the Office of the Controller-General of the Republic was
created as a modern instrument of fiscal supervision.
In social terms, the country also underwent significant changes during the
period. The relative size of the wage-earning workforce increased, providing the
basis for trade-union movements which began to form in the 1910s and experi-
enced their first period of expansion in the 1920s. The main areas in which
trade union struggles developed were steam-powered shipping and the export
sectors (Urrutia, 1969; Archila, 1989). The most significant event in terms of its
political implications was, unquestionably, the 1928 banana strike (White,
1978; LeGrand, 1989), immortalized by Gabriel GarcõÂa MaÂrquez in his classic
novel, One Hundred Years of Solitude. At the same time, the availability of alter-
native employment exacerbated conflicts which had long existed in certain
rural areas; as we will see, these were particularly significant in the large
plantations of Cundinamarca, which had been the heart of the expansion of
coffee-growing in the late nineteenth century.
There are few quantitative indicators of the trend in the living conditions of
the population. The first available data for most social indicators date from the
mid-twentieth century and are all unsatisfactory. It is clear that limited progress
was made in the first half of the century. Nevertheless, the acceleration in
population growth from 1.5 per cent per year in the nineteenth century to
2.1 per cent in the first decades of the twentieth century was due in part to
falling death rates, no doubt a reflection of an improvement in living condi-
tions. The increase in the average height of the population since the turn of the
century is also evidence of a gradual improvement in nutrition among the
population as a whole (Ordon Äez and PolanõÂa, 1992).
Changes in the coffee economy8
The coffee boom of the early 1900s entailed significant changes in regional
terms and in the organization of production. As we have seen, most of the
traditional large plantations in eastern Colombia were profoundly affected by
the Thousand Days' War and by the collapse of international prices at the turn
of the century and were unable to expand output once peace was restored and
coffee prices recovered. As a result, production in the east of the country, where
plantations were concentrated, rose very slowly from some 500 000 bags in
1890 to 825 000 bags in 1932, a 1.6 per cent annual increase (Table 3.2). The
main problem they faced was their inability to change their labour relations,
and they remained trapped in servile land tenancy arrangements. In the 1920s,
these plantations became sites of agrarian conflict and, in the following decade,
one of the key targets of the first wave of agrarian reforms, introduced by the
reformist liberal governments of the time (Bejarano, 1979; Palacios, 1983).
At the same time, the coffee cultivation began to flourish in the west of the
country, in the steep but rich volcanic soils of the central mountain range
where Antioquen Äo migration had settled down during the course of the nine-
teenth century. Indeed, the expansion of production centred in the Antio-
quenÄo-settled areas in Caldas, the south of the department of Antioquia, and
the north of Tolima and the Cauca valley. The settlers in those areas had
originally employed slash-and-burn methods of agriculture similar to those
Table 3.4 Concentration of coffee production, 1923 and 1932

(by size of mature coffee trees)
Number of plantations % of production
1923 1932 1923 1932
Up to 3 hectares 34 841 113 950 24.4 26.1

3.1±12 hectares 8 544 31 178 32.0 33.4
12.1±35 hectares 1 544 3 255 20.1 22.3
Over 35 hectares 560 823 23.5 18.2
Total 45 489 149 206 100.0 100.0
Source: Machado (1988), tables 12 and 13.
used in other parts of the country, while also engaging in some supplementary
income-producing activities such as pig-raising and searching for archaeolog-
ical treasures. Coffee-growing had not been a major part of their livelihood,
even in those regions that were settled late (QuindõÂo). Nevertheless, from the
turn of the century onwards, coffee-growing expanded rapidly in the small and
medium-sized holdings that had developed in the region as a result of a pro-
tracted struggle over land between settlers and the owners of the idle land.
Coffee-growing expanded at a remarkable pace in the region, with coffee pro-
duction increasing at rates close to 10 per cent a year during the first three
decades of this century so that by 1932 the three main coffee-growing depart-
ments in western Colombia ± Antioquia, Caldas and Valle del Cauca ±
accounted for 57 per cent of domestic output (Table 3.2).
In sharp contrast to the patterns that characterized the central and eastern
parts of the country, production in the west concentrated in small and med-
ium-sized holdings. Consequently, in 1932, 59.5 per cent of coffee bean pro-
duction came from plantations of 12 hectares or less. The proportion was
higher in Caldas and the Cauca valley, while in Cundinamarca and Santander
larger-sized holdings continued to be the norm at that time (Tables 3.4 and 3.5).
In any event, the egalitarian nature of production should not be overstated. The
majority of producers (76 per cent in 1932) had coffee plantations of less than 3
hectares, but their share of production was relatively small (26 per cent accord-
ing to Table 3.4). Social differentiation in the rural economy was as important a
reality as growth in the number of small-scale producers, as evidenced by the
increasing importance of medium-sized farms (Machado, 1988).
The workforce on the smallest holdings was made up exclusively of family
members, although the use of family labour was not highly intensive, except at
harvest time. Coffee-growing methods were simple, and coffee was supplemen-
ted by other crops and animal husbandry. On the largest farms outside labour
was used, especially during harvesting. Hiring practices in Caldas included
Table 3.5 Concentration of production in main departments, 1923 and 1932 (by size of
mature coffee trees)
Department Up to 12 hectares (%) 12.1±35 hectares (%) Over 35 hectares (%)
1923 1932 1923 1932 1923 1932
Antioquia 57.4 56.6 23.0 26.1 19.6 17.3

Caldas 69.2 64.1 20.4 25.2 10.4 10.7
Cundinamarca 29.0 47.2 16.0 14.9 55.0 37.9
North Santander 56.3 70.4 25.0 14.1 18.7 15.5
Santander 35.6 39.7 16.4 23.5 48.0 36.8
Tolima 41.7 46.0 19.8 26.4 38.5 27.6
Valle del Cauca 80.7 72.8 12.1 22.8 7.2 4.4
Source: Machado (1988), tables 12 and 13.
payment on a per diem basis and `company systems'. The latter were not very
different from the share-cropping arrangements in eastern Colombia. The
workers received 50 per cent of the coffee harvested, with an additional deduc-
tion if it were delivered to the landowner unprocessed. As in the east of the
country, a company system was also used for coffee planting. Under this
arrangement, the worker could make full use of the land during the term of
the contract (generally four years) and would be required to sell the mature
trees to the owners.
One of the most important innovations in the coffee industry in the first
decades of the twentieth century was the advent of the manual pulper. Pulping
machines became the main product of the AmagaÂ ironworks in Antioquia and
became indispensable on all coffee farms. In 1915, 70 per cent of Antioquian
farms had such a machine, a figure which had risen to 97 per cent by 1922
(Palacios, 1983). This innovation made it possible to shift the main processing
tasks other than hulking ± depulping, washing and drying of the beans ± from
the large plantation estates and small processing centres to the smaller farms,
thus bringing the farmers into direct contact with urban traders.
In the areas settled by AntioquenÄos, the relationship between the peasant
and the store-owner ± fondero ± replaced that between tenant and plantation
owner more typical of the eastern part of Colombia. The store-owner would
advance the peasant the credit he required for purchases throughout the year
and in return would take the coffee crop at a significant discount. The interest
rates applied in these transactions reached usurious levels and were a major
bone of contention until the Agrarian Bank and the National Federation of
Coffee Growers intervened in the 1930s to put an end to this unfair practice.
The store-owner was only the first link in the commercialization of coffee.
Coffee beans were delivered to the agents of export firms located in the more
important towns; these agents would then send the beans on to commercial
centres for hulking and subsequent shipping to their final destination. As

Palacios (1983) has noted, the export firms sometimes colluded with their
agents to reduce the purchase prices and control mule freight charges and the
like. The primitive communication networks of the time allowed for huge
disparities between the local purchase price of coffee beans and the high trad-
ing margins. Only action by the National Federation of Coffee Growers in the
1930s brought some limits to those local market imperfections.
Another great innovation of the twentieth century was the introduction and
widespread use of coffee hulking. This activity became one of the main indus-
tries in MedellõÂn and Cali and, to a lesser extent, in other cities in the western
part of the country. In 1923 in MedellõÂn, coffee-hulking factories employed
1823 people, or 32 per cent of the 5702 office and industrial workers in the city.
In Cali, in 1925, the three main industrial facilities were all coffee-hulking
factories. The seven such plants then in operation employed 41 per cent of
the city's industrial workforce of 1504 (Arango, 1977; Ocampo and Montene-
gro, 1984).
Until the coffee crisis of 1920±21, the coffee export trade continued to be
based on the consignment system prevalent in the nineteenth century. In the
first two decades of the twentieth century, the export trade underwent two
basic changes. First, exports increasingly went to the USA. Even before the First
World War, nearly 70 per cent of Colombia's exports went to that market, and
this figure rose to over 90 per cent when access to the European market was cut
off during the war. The second major change was the appreciable rise in the
scale of operations, which enabled the largest trading houses to open branches
in New York. Many of these firms went bankrupt after the coffee business
experienced wild fluctuations in 1920±21, which in turn opened the door to
new trading firms and the entry of foreign capital into the coffee trade. Foreign
firms derived market power from their vertical integration with the shipping
companies, or with the coffee-roasting and wholesale trade, chiefly in the USA.
The coffee boom was not based on particularly favourable international
prices, except in a handful of years in the 1910s and 1920s. In point of fact,
planting began in 1905±10 when export prices were still depressed. This situa-
tion undoubtedly reflected the wealth of natural resources upon which the
boom was based and the possibility of permanent access to land and labour at
minimal monetary costs,9 which had up till then been partially or completely
underutilized. Indeed, the mere introduction of coffee represented a funda-
mental technological innovation, as it provided for permanent use of land in
hilly areas where until then only slash-and-burn agriculture had been
employed. It also offered an opportunity to productively employ the available
labour force.
Coffee production was also boosted by the development of a series of supple-
mentary activities: an extensive financing and trading network serving the
coffee trade, the manufacture of bags and pulping machines and the growth
of the hulking industry. The critical difference, however, compared to the
nineteenth century, was the development of a modern transport network.
The competition between MedellõÂn, Manizales and Cali to control the coffee
business depended on upgrading transport systems. In this process, MedellõÂn
increasingly lost ground to Cali, thanks to the integrated transport net-
work developed between the coffee-growing area and Buenaventura on the
Pacific which was controlled by Cali interests. This certainly put pressure on
MedellõÂn to seek out new economic activities, especially in the manufacturing
sector.
The creation in 1927 of the National Federation of Coffee Growers consolid-
ated coffee development by facilitating public action in favour of the sector.
However, the Federation's intervention in the domestic coffee market only
really began during the early years of the Great Depression, when it built the
first warehouses and stepped in to purchase crops in regions where traders'
profit margins were excessive. The Federation was a private organization, but
this did not prevent it from gradually assuming public functions. From the time
the Federation was founded, it adopted a special mechanism which enabled the
government to participate in policy-making; the National Coffee Committee,
on which the government was represented, took policy decisions which the
Federation subsequently carried out.
The emergence of manufacturing
The impetus generated by the export sector was sufficient to consolidate early
attempts at industrialization prior to the crisis of the 1930s. By that time, as the
eminent historian of manufacturing and protectionism in Colombia pointed
out, the era of industrial pioneering was already a thing of the past (Ospina
VaÂsquez, 1955). Thus, as in other Latin American countries similar or larger
in size, the first phase of industrialization was closely linked to export
expansion.10
As one might expect, the relationship between the growth of the coffee
industry and the origins of industrialization has been one of the most popular
topics in Colombian economic historiography. The most widespread theory has
emphasized the role of small and medium-sized coffee producers in the west of
the country in modern industrial development. According to this view, the
widespread distribution of the benefits of expanded coffee cultivation to
the peasant economy in the west of the country laid the foundations for the
development of a domestic market which was to sustain the successful indus-
trialization efforts pursued in Colombia in the early decades of the twentieth
century (Nieto Arteta, 1975; Urrutia, 1979, ch. 3). Arango (1977) has countered
this theory by showing that traders' profit margins in areas of small-scale coffee
production areas were very high, a state of affairs not effectively corrected until
the beginning of the 1930s when the National Coffee Growers Federation
intervened. Hence, according to Arango, the contribution of the coffee expan-
sion was not so much a matter of generating a peasant market for industrial
goods as it was of accelerating capital formation. Coffee also contributed
directly to the development of manufacturing firms and a modern proletariat
in the cities through the spread of industrial hulking, which tended to be
concentrated in urban areas in the west of the country in contrast to the typical
pattern in the regions in which large plantations predominated.
These alternative interpretations of the relationship between coffee produc-
tion and manufacturing focus on certain key aspects of Colombian economic
development, but ignore other equally important elements. With a few excep-
tions ± textiles may have been one ± most consumption of industrial goods was
essentially urban from the start.11 Moreover, there is no clear evidence that
commercial capital associated with coffee participated significantly in the
founding of industrial enterprises in Antioquia (F. Botero, 1984). Lastly,
although employment in the coffee-hulking plants accounted for a large per-
centage of manufacturing jobs in the initial stages of industrialization, espe-
cially in the west of the country, the number of workers involved was too low to
be considered the source of the modern industrial proletariat.
Thus, coffee expansion boosted early attempts at industrialization not so
much through the creation of a rural market for industrial goods or capital
formation in activities directly related to marketing of coffee beans, but
through essentially macroeconomic linkages: capital formation in general, urba-
nization, the development of a modern road system and other factors asso-
ciated with this period of prosperity. Moreover, other factors having to do with
economic policy and various external phenomena contributed to the industrial
take-off.
Among the policy factors, the most important was undoubtedly the impetus
the Reyes Administration gave to early manufacturing. Protectionism, which
had been firmly rooted in the nation's economic thinking since the Regenera-
tion, found one of its most thorough expressions during the Reyes Administra-
tion (Ospina VaÂsquez, 1955; Ocampo and Montenegro, 1984). The tariff
reforms of 1905 and 1913 reinforced and in effect made explicit the protec-
tionist tendency that had been manifest since the 1880s; the reforms raised the
average tariff collected to the highest level yet seen in the country's history
(Figure 3.2). At the same time, the spreads between duties on goods for final
consumption and on inputs widened, and the wave of protectionism extended
to some processed agricultural products ± sugar and flour, among others. In
addition, the Reyes Administration fostered the infant industries ± both import
substitution and export ± through government purchases and loans to enter-
prises, direct subsidies and tariff exemptions on imported machinery.
Although protectionist sentiment did not disappear in the decades that

followed, the few surcharges on specific duties were insufficient to make up
for the inflation prevailing during those years. The average tariff declined
substantially (Figure 3.2). Moreover, some of the reforms of the period tended
to increase tariffs on inputs for some industrial sectors and ultimately to reduce
effective protection. The impact of these measures was not even, however:
while the textile industry suffered an appreciable drop in protection levels
(Figure 3.3), other industrial sectors ± beer brewing, cigarette manufacturing
and sugar refining, in particular ± continued to enjoy high levels of protection
(EchavarrõÂa, 1992; Ocampo and Montenegro, 1984).
During the 1910s and 1920s, external economic factors acted as a catalyst for
new industrial developments. The most significant of these were undoubtedly
the scarcity of foreign products during the First World War and the import
boom during the five-year period preceding the Great Depression. The 1945
industrial census indicates that most of the companies established during the
first thirty years of the century did so during the Reyes Administration, the First
World War and the second half of the 1920s. However, the relevant data,
summarized in Table 3.6, indicates that industrial growth was not uniform in
all sectors over the three decades. The first decade was a particularly good
one for the textile industry. Growth in textiles continued in the following
decade, but the creation of new firms was extensive in many other sectors as
well, especially food products, beverages, tobacco and oil refining. In the 1920s,
the textile industry suffered a severe decline (Ocampo and Montenegro, 1984),
reflecting lack of interest by investors, but the creation of food and beverage
companies accelerated, and companies also sprang up in the cement and some
metallurgical industries.
In regional terms, Antioquia took the lead during the first two decades. The
region's industrial leadership, however, was much less striking and slower to
develop than is often stated.12 On the contrary, one of the fundamental fea-
tures of industrialization in Colombia from the outset has been a marked
regional diversification. Moreover, this characteriztic was reinforced in the
second decade of the century and accentuated in the 1920s, when Antioquia
suffered a relative recession. According to EchavarrõÂa (1992), each region had its
own special advantages: BogotaÂ was located in the major concentration of
population; MedellõÂn was better supplied with electrical power and had a
more educated and trained workforce (partly a legacy of its dynamic mining
sector); Cali had the best transportation infrastructure at the time and access to
the main Pacific port; and Barranquilla was still the country's principal port.
A breakdown of the available data on manufacturing production in the 1920s
by industry tends to confirm the overwhelming weight of the food and tobacco
industries, 63 per cent and 15 per cent of industrial GDP, respectively, in 1925±
29 ± estimated at 1975 prices. Despite the progress in other industrial sectors in
78
Table 3.6 Share of net worth of industrial firms in 1945 by date of establishment (%)
Sector Year of establishment
1880 or From 1881 From 1901 From 1911 From 1921 From 1931 From 1941 Net worth
earlier to 1900 to 1910 to 1920 to 1930 to 1940 to 1945 in 1945
Food products 70.8 1.5 11.0 16.9 29.8 18.8 27.0 19.0
Beverages 0.0 87.3 9.9 10.8 12.7 2.5 3.0 18.1
Tobacco 0.0 0.0 1.2 14.4 1.1 0.4 1.3 3.3
Textiles 0.0 0.0 67.7 27.6 6.1 30.9 13.1 23.3
Clothing and leather goods 0.1 0.2 1.0 7.6 6.4 11.1 11.2 6.9
Wood and furniture 0.6 0.3 0.1 0.9 3.5 2.9 7.2 2.3
Paper 0.0 0.0 0.0 0.0 0.0 1.4 0.7 0.5
Printing 2.7 3.1 3.7 3.0 2.5 2.1 1.3 2.6
Chemicals 6.8 2.3 0.8 3.7 6.6 5.6 10.4 4.9
Rubber 0.0 0.0 0.0 0.0 0.2 2.2 1.4 0.8
Petroleum products 0.0 0.0 0.0 12.7 0.0 0.0 0.0 2.4
Non-metallic minerals 1.0 4.5 3.8 1.5 17.3 10.3 15.0 8.8
Metallurgy and metal working 17.9 0.8 1.0 0.8 13.6 10.7 7.0 6.9
Other 0.0 0.0 0.0 0.0 0.1 1.1 1.6 0.5
Departments
Antioquia 10.4 11.2 70.1 41.8 12.0 30.2 15.0 28.9
Atlantic 0.0 0.4 0.1 14.9 14.9 19.1 15.2 12.4
Cundinamarca 15.0 79.5 9.6 16.8 30.8 17.1 27.1 28.0
Valle del Cauca 68.3 1.4 2.7 4.8 20.4 18.1 17.3 13.0
Rest of country 6.3 7.5 17.5 21.7 21.9 15.4 25.5 17.7
Investment in each period as %
of net worth in 1945 1.7 13.3 10.4 18.7 18.8 28.6 8.4 100.0
Source: 1945 industrial census.

the early decades of the century, especially in textiles, they still lagged far
behind for a country of Colombia's size and level of development. National
production, in fact, supplied only one-third of apparent demand for industrial
goods and less than one-fifth of national consumption of textile products (Chu,
1972; Ocampo and Montenegro, 1984). Moreover, in the period of prosperity
the country enjoyed in the 1920s, industrial production tended to lag behind as
a result of a crisis in the textile industry and other important branches of
manufacturing.
In contrast to some successful nineteenth-century firms founded by foreign
immigrants, the capital of the chief industrial enterprises in the early part of the
century was mainly provided by Colombian nationals. These investments satis-
fied the need of the country's richest families to diversify their fortunes, a
pattern with firm roots in the Colonial and nineteenth-century past. Even in
the case of the principal industrial investors, the percentage of family wealth
invested in industrial ventures appears not to have exceeded 40 per cent. In
many cases, an interest in starting up domestic production had its roots in the
particular investor's previous experience with marketing imported products, a
circumstance that may have helped to reduce the risk entailed by the new
investment. In keeping with a pattern firmly established in some regions of
the country, particularly Antioquia, in the nineteenth century, most of the large
enterprises were initiated as corporations even though they were closely family-
controlled. Moreover, a tendency towards concentration was evident early on
in some sectors ± tobacco, beer, matches, carbonated beverages. The acquisition
of other firms or the establishment of subsidiaries in other regions of the
country had generated a relative concentration in these industries even before
the crisis of the 1930s (EchavarrõÂa, 1992; Ospina VaÂsquez, 1955; Botero, 1984).
From the labour standpoint, the most important trend in manufacturing
during the early decades of the century was the rapid decline in handicraft
employment. Up to at least 1918, handicraft employment remained very
important, particularly among female workers (Table 3.7). Although we cannot
pinpoint the start of the decline, there was a sweeping change between 1918
and the middle of the century during which period the percentage of handicraft
workers in the total workforce dropped from 24 per cent to 7.5 per cent. The
decline particularly affected women's jobs, many of them located in rural
areas or smaller urban centres. The process was most noticeable in the regions
of the country that had had strong handicraft traditions but had been at the
periphery of modern industrial development (BoyacaÂ, Santander, Huila, Cauca
and Narin Äo).
As the handicraft tradition was dying out, manufacturing development was
generating demand for labour on a very limited scale, employing some 20 000
or fewer workers by 1918, that is, a fraction of those employed in the handicraft
sector.13 Up to the 1920s, labour law placed no great restraints on labour
80
Table 3.7 Employment in handicraft and manufacturing activities, 1870 and 1918a
Employment in handicraft and % of total employment (excluding % of total population

manufacturing businessesb domestic service)
1870 1918 1870 1918 1870 1918
M F M F M F M F M F M F
Antioquia 8 435 9 767 22 899 23 015 7.2 44.9 6.6 69.1 4.7 5.3 3.7 3.6
BoyacaÂ 32 135 70 681 41 135 72 890 21.8 74.7 21.7 49.4 13.8 29.6 13.2 21.3
Cauca 15 138 42 110 28 744 115 688 11.3 39.5 9.2 63.0 7.2 18.8 6.3 24.6
Cundinamarca 12 345 17 080 20 562 53 782 10.9 49.4 9.2 43.8 6.3 8.0 5.3 12.6
Santander 12 781 66 922 20 481 53 564 10.3 88.3 10.1 79.5 6.3 30.2 6.2 15.4
Tolima 3 192 20 765 8 079 50 084 5.2 71.9 5.6 49.5 2.9 17.3 3.4 18.1
Subtotal excluding Atlantic

Coast 84 116 227 325 141 900 368 993 12.1 62.7 10.0 56.3 7.4 18.9 6.1 14.7
BolõÂvar 4 641 3 672 6.7 46.0 4.0 2.9
Magdalena 3 073 10 684 13.0 89.3 7.6 24.0
Total 91 830 241 681 11.6 63.2 7.1 17.6
a
The departments shown correspond to those existing in the period 1863±86.
b
Artisans, manufacturers and artists in 1870; arts, crafts, manufacturing and industrial production in 1918.
Source: Population censuses.
mobility. It fostered a situation, at least in Antioquia, in which initially a large

percentage of the workers were women, boys and girls who were paid poorly
despite having relatively high levels of education for the time (Botero, 1984).
While a high level of education was also a distinctive characteristic of the
industrial workforce in other regions and subsequent periods, the dependence
on female and child labour was not as widespread and in fact soon disappeared
in Antioquia. In addition, because industrial workers were not concentrated in
large numbers in a single workplace, they played a secondary role in the early
union movements that emerged during the 15 years prior to the depression of
the 1930s.
Although still in its early stages, it was clear by the end of the 1920s that the
industrialization process had achieved the momentum it needed to become
consolidated in the decades that followed. The slow integration of the domestic
market meant that the advantages of market expansion were to be fully realized
only in the 1930s. As a result of these dynamics, the ùnbalanced' growth
pattern ± to borrow Hirschman's term (1958) ± that characterized the first few
decades of the century in which manufacturing lagged behind, gave way to a
new phase of unbalanced development ± or rather a series of such phases ± in
which industry took the lead in the development process.
Notes
1 On nineteenth-century coffee development, see Palacios (1983), Arango (1977), Deas

(1976) and Ocampo (1984, ch. VII, and 1989). See also the analyses of Nieto Arteta
(1975) and Urrutia (1979, chs 2 and 3) on the links between coffee and overall
economic development.
2 The importance of paper money in the coffee boom which started in the 1880s has
been subject to historical controversy. For recent contributions to this debate, see the
works of Ocampo, Palacios and Urrutia quoted in the previous note and Banco de la
RepuÂblica (1990).
3 Nieto Arteta (1996) and Ospina VaÂsquez (1955) remain the classical accounts of nine-
teenth-century economic history. Melo (1987) provides an excellent synthesis of
economic developments. Bergquist (1978) treats the interaction of economic and
political developments at the end of the century and LeGrand (1988) presents the
best overall view of agrarian history. The recent overall history of Bushnell (1993) is
also the best of its kind and must also be consulted for this period. Among regional
histories, Brew (1977) and Posada-CarboÂ (1996) stand out. See also works quoted
later on.
4 The percentage of the population living in the 10 main urban centres (not necessarily
the same ones in different population censuses) declined considerably between 1835
and 1851, from 8 to 6.3 per cent. In 1870, the figure was still low (6.6 per cent), and
only from then on did it begin a steady rise. By 1912 it had risen to 8.8 per cent.
5 This can be inferred, for example, from the descriptions of the difficulties encountered
at the turn of the century in transporting the machinery for the country's first modern
sugar refinery (Manuelita) to Palmira (Valle del Cauca) and the first modern
equipment for what would become, decades later, Colombia's leading textile firm
(Coltejer) to MedellõÂn (in Antioquia).
6 This would be equivalent to growth in per capita GDP of 1.9 to 2.5 per cent, higher
than that estimated by Bulmer-Thomas (1994, appendix 3) for 1913±28 (1.4 per cent),
which seems too low. This estimate of per capita growth takes as a reference a
population growth of 2 per cent, which is consistent with Colombian population
censuses but lower than that used by Maddison and Hofman.
7 Some recent indirect estimates from the Banco de la RepuÂblica place GDP growth in
1905±29 at 4.9 to 6.7 per cent a year, but these calculations seem high, particularly
the latter.
8 For a more extensive analysis of the coffee economy in the early twentieth century,
see Arango (1977), Machado (1988), Ocampo (1989) and Palacios (1983).
9 The low monetary cost of labour was the result of the fact that peasants also grew
their own food on their land. The importance of monetary cost minimization as a
source of competitiveness in precapitalist economies has been emphasized by Kula
(1974), in a brilliant essay on Polish farms in the sixteenth and seventeenth centuries.
This source of competitiveness for Colombia's coffee industry has been pointed out
by many modern authors, but was mentioned by Mariano Ospina PeÂrez as head of the
National Federation of Coffee Growers as far back as 1934. See Ospina PeÂrez (1934).
10 Alongside different texts quoted in this section, the early history of Colombian
industrialization has also been analysed by Berry (1983), Brew (1977) and Poveda
(1970).
11 According to ECLAC estimates (ECLAC 1957), in 1953 when less than 40 per cent of
the population lived in cities, urban centres accounted for 64 per cent of industrial
consumption. Only in the case of textiles did rural consumption (61 per cent) exceed
urban consumption. Rural consumption of tobacco products was only slightly lower
than urban consumption.
12 This fact is corroborated in the 1916 Ministry of Finance Annual Report, which
presents a brief, incomplete inventory of the industrial firms at that time. According
to that source, less than 30 per cent of the capital invested in industrial activities at
that time was in Antioquen Ä o firms.
13 In reality, this figure could well be too high. The incomplete list drawn up in the
Ministry of Finance Annual Report of 1916 included 121 establishments employing a
total of 12 223 workers. Of those, however, 5000 were employed by the Sincerin sugar
refinery, most of whom must have been agricultural labourers. On the other hand,
the inventory did not provide information on employment in some large enterprises
(Bavaria, Manuelita and two of the biggest textile firms of Antioquia) or in some
categories of industrial activity, especially the more traditional industries (coffee
hulking, distilling, sugar mills and printing).
References
Arango, M. (1977) CafeÂ e industria, 1850±1930, BogotaÂ: Carlos Valencia Editores.

Archila, M. (1989) `La clase obrera colombiana (1860±1930)', in Nueva historia de Colom-
bia, (vol. III, ch. 9), BogotaÂ: Planeta.
Avella, M. (1987) Pensamiento y polõÂtica monetaria en Colombia, 1886±1945, BogotaÂ: Con-
tralorõÂa General de la RepuÂblica.
Banco de la RepuÂblica (1990) El Banco de la RepuÂblica: antecedentes, evolucioÂn y estructura,

BogotaÂ.
Bejarano, J. A. (1979) El reÂgimen agrario de la economõÂa exportadora a la economõÂa industrial,
BogotaÂ: Editorial La Carreta.
Bell, P. L. (1922) `Coffee, the Mainstay of Colombia', in The Tea and Coffee Trade Journal,
February.
Bergquist, C. W. (1978) Coffee and Conflict in Colombia, 1886±1910, Durham: Duke Uni-
versity Press.
Berry, A. (1983) À Descriptive History of Colombian Industrial Development in the
Twentieth Century', in A. Berry (ed.), Essays on Industrialization in Colombia, Tempe:
Arizona State University.
Botero, F. (1984) La industrializacioÂn en Antioquia: geÂnesis y consolidacioÂn, 1900±1930,
MedellõÂn: CIE±Universidad de Antioquia.
Botero, M. M. (1995) Èl desarrollo del sector exportador en Antioquia 1850±1890: Comer-
cializacioÂn y hegemonõÂa', Paper presented to LASA Conference, October.
ÐÐ (1994) Antecedentes del desarrollo exportador en Antioquia, 1850±1890, MA thesis, Uni-
versidad Nacional, BogotaÂ.
Brew, R. (1977) El desarrollo econoÂmico de Antioquia desde la independencia hasta 1920,
BogotaÂ: Banco de la RepuÂblica.
Bushnell, D. (1993) The Making of Modern Colombia: A Nation in Spite of Itself, Berkeley and
Los Angeles: University of California Press.
CEPAL, 1957) AnaÂlisis y proyecciones del desarrollo econoÂmico, Vol. III: El desarrollo econoÂmico
de Colombia, Mexico: ECLAC.
Chu, D. S. C. (1972) The Great Depression and Industrialization in Latin America: Response to
Relative Price Incentives in Argentina and Colombia, 1930±1945, doctoral thesis, Yale
University.
Deas, M. (1976) Ùna hacienda cafetera de Cundinamarca: Santa BaÂrbara (1870±1912)',
Anuario Colombiano de historia social y de la cultura, BogotaÂ: Universidad Nacional,
Departamento de Historia.
EchavarrõÂa, J. J. (1992) External Shocks and Industrialization in Colombia, 1920±1950, doc-
toral thesis, Oxford University.
Eder, P. J. (1959) El Fundador, Santiago M. Eder, BogotaÂ: Antares.
Hirshman, A. O. (1958) The Strategy of Economic Development, New Haven: Yale University
Press.
Hofman, A. (1998) Latin American Economic Development: A Causal Analysis in Historical
Perspective, Groningen Growth and Development Centre, monograph series no. 3.
Junguito, R. (1995) La deuda externa en el siglo XIX: cien an Ä os de incumplimiento, BogotaÂ:
Tercer Mundo±Banco de la RepuÂblica.
Kalmanovitz, S. (1985) EconomõÂa y nacioÂn: una breve historia de Colombia, BogotaÂ: Siglo
XXI±Universidad Nacional±CINEP.
Kula, W. (1974) TeorõÂa econoÂmica del sistema feudal, Buenos Aires: Siglo XXI.
LeGrand, C. (1989) Èl conflicto de las bananeras', in Nueva historia de Colombia (vol. III,
ch. 8), BogotaÂ: Planeta.
ÐÐ (1988) ColonizacioÂn y protesta campesina en Colombia, 1850±1950, BogotaÂ: Universidad
Nacional.
London Äo, J. L. (1989) Àgricultura y transformacioÂn estructural: una comparacioÂ n inter-
nacional', Revista de PlaneacioÂn y Desarrollo, July±December.
LoÂpez Toro, A. (1970) MigracioÂn y cambio social en Antioquia durante el siglo diez y nueve,
BogotaÂ: CEDE±Universidad de los Andes.
Machado, A. (1988) El cafeÂ: de la aparcerõÂa al capitalismo, 2nd edn, BogotaÂ: Tercer Mundo.
Maddison, A. (1995) Monitoring the World Economy, 1820±1992, Paris: OECD Development
Centre.
McGreevey, W. P. (1971) An Economic History of Colombia, 1845±1930, Cambridge: Cam-
bridge University Press.
Melo, J. O. (1987) `Las vicisitudes del modelo liberal (1850±1899)', in J.A. Ocampo (ed.),
Historia econoÂmica de Colombia, BogotaÂ: Siglo XXI±Fedesarrollo.
Nieto Arteta, L. E. (1975) El cafeÂ en la sociedad colombiana, BogotaÂ: Ediciones Tiempo
Presente.
ÐÐ (1973), EconomõÂa y cultura en la historia de Colombia, revd edn 1996, BogotaÂ: Editora
Viento del Pueblo.
Ocampo, J. A. (1994) `RegõÂmenes monetarios variables en una economõÂa preindustrial:
Colombia, 1850±1933', in F. SaÂnchez (ed.), Ensayos de historia monetaria y bancaria de
Colombia, BogotaÂ: Tercer Mundo±Fedesarrollo±Asobancaria.
ÐÐ (1989) `Los orõÂgenes de la industria cafetera, 1830±1929', in Nueva Historia de Colom-
bia (vol. V, ch. 9), BogotaÂ: Planeta.
ÐÐ (1984) Colombia y la economõÂa mundial, 1830±1910, BogotaÂ: Siglo XXI±Fedesarrollo.
Ocampo, J. A. and Montenegro, S. (1984) Crisis mundial, proteccioÂn e industrializacioÂn:
ensayos de historia econoÂmica colombiana, BogotaÂ: CEREC.
Ordon Ä ez Plaja, A. and PolanõÂa, D. (1992) `Cambios de estatura en Colombia durante el
presente siglo', Coyuntura Social, no. 6, June.
Ospina PeÂrez, M. (1934) Informe del Gerente de la FederacioÂn al Sexto Congreso Nacional de
Cafeteros, BogotaÂ: FederacioÂn Nacional de Cafeteros.
Ospina VaÂsquez, L. (1955) Industria y proteccioÂn en Colombia, 1810±1930, BogotaÂ: Editorial
Santa Fe.
Palacios, M. (1983) El cafeÂ en Colombia, 1850±1970: una historia econoÂmica, social y polõÂtica,
2nd edn, Mexico/BogotaÂ: El Colegio de MeÂxico/El Ancora Editores.
Parsons, J. J. (1968) Antioquen Ä o Colonization in Western Colombia, Berkeley and Los Angeles:
University of California Press.
Posada CarboÂ, E. (1996) The Colombian Caribbean: A Regional History 1870±1950, Oxford:
Clarendon Press.
ÐÐ (1987) Una invitacioÂn a la historia de Barranquilla, BogotaÂ: CEREC.
Poveda, G. (1970) `Historia de la industria en Colombia', Revista Trimestral, no. 11, ANDI.
Romero, C. A. (1994) `La banca privada en BogotaÂ, 1870±1922', in F. SaÂnchez (ed.), Ensayos
de historia monetaria y bancaria de Colombia, BogotaÂ: Tercer Mundo±Fedesarrollo±Asocia-
cioÂn Bancaria de Colombia.
Safford, F. (1965) Commerce and Enterprise in Central Colombia, 1821±1870, Doctoral thesis,
Columbia University.
Tovar, B. (1984) La intervencioÂn econoÂmica del Estado en Colombia, 1914±1936, BogotaÂ:
Banco Popular.
Urrutia, M. (1979) Cincuenta an Ä os de desarrollo econoÂmico colombiano, BogotaÂ: La Carreta.
ÐÐ (1969) Historia del sindicalismo en Colombia, BogotaÂ: Universidad de los Andes.
White, J. (1978) Historia de una ignominia: la United Fruit Co. en Colombia, BogotaÂ: Editorial
Presencia.
4
The Economies of Central America,
1860±19401
HeÂctor PeÂrez Brignoli*
Introduction
The size and basic characteristics of the countries of Central America make
them an ideal case study of the development of export economies. They have
relatively simple economic structures, which, in the majority of cases, have
been dominated for a considerable time by one or two export products. In other
words, the effects of internal diversification that were quick to appear in larger
economies have taken several decades to appear here. In particular let us note
that not so long ago, from 1960±75, the Central American food, drinks and
tobacco industries underwent a rate of growth ± measured in respect of the
added value of the whole industrial sector ± similar to that experienced by the
same industries in the Argentine economy from 1900±29 (PeÂrez Brignoli and
Baires MartõÂnez, 1983, p. 380).
This means that they are export economies in which the development of
import substitution was late and incomplete. For this reason they have been
consistently open to the world market since the middle of the nineteenth
century and broadly favourable to foreign investment with only a small inter-
ruption from 1960 to 1985.
Given that the region is dependent upon just a few export products, there is
also a wide discrepancy with regard to income distribution and social structure.
It would seem that the coffee economy, the very one that sustained long-lived
dictatorships such as that of Manuel Estrada Cabrera in Guatemala (1898±
1920), also helped in other instances, as was the case in Costa Rica, to establish
an equally long chain of democratic and stable regimes. In other words, in so
* The author acknowledges with thanks the comments of Victor Hugo Acun Äa, Enrique CaÂrdenas,
Jorge LeoÂn SaÂenz and Mario Samper. The ideas contributed by Rosemary Thorp, contained in the
convocation documents and summary of the seminar held in Panajachel (Guatemala) in December
1996, have been a constant source of inspiration and encouragement. Thanks to her and members of
the export economies team, work on this project has been as enjoyable as it has been demanding.
85
86 Central American Economies, 1860±1940
far as the relationships between economy, state and business are concerned, the
history of the region offers examples as varied and complex as those of any
Latin American country.
Preconditions
In the middle of the nineteenth century the integration of the Central Amer-
ican economies into the world market continued to pose problems (Lindo-
Fuentes, 1993, pp. 141±201). Products like cocoa, tobacco, indigo, cochineal,
sugar and cotton were produced for the local market; others, like balsam,
sarsaparilla and high-quality timber, seemed very promising with regard to
possible international demand. The main difficulty in establishing an effective
relationship with external markets was still exactly the same as in colonial
times: production costs were not competitive. This was due to very expensive
maritime transport, a primitive internal transport network, which was depend-
ent almost entirely upon mule transport, and the cost and quality of labour.
Moreover, the very size of the countries and producing regions limited the scale
of production and discouraged the interest of exporters, at least in the medium
term. The history of Central America has been characterized ± as officials at the
Guatemalan Consulate (Informe del Consulado, 1983, pp. 123±41) never tired of
repeating ± by an almost infinite succession of short cycles in which a promis-
ing product would appear, production and trade would flourish, and then it
would all disappear as quickly as it had begun. Against this background it is easy
to explain the success, though short-lived, of products with a high unit price as
regards volume, like indigo or cochineal, and of those cheap and easy to obtain,
such as timber from the coastal areas. This colonial model, prevalent since the
sixteenth century, was still common in the middle of the nineteenth century.
What were the factors exerting such a disastrous influence on Central Amer-
ica for the situation to continue for three hundred years? Above all, it was the
cost of transport by sea. In the sailing-ship age, not only was access to the
Pacific coast of Central America difficult but the region was also an extremely
long way from European markets. Besides, the isthmus was underpopulated,
with a number of isolated settlements in the central highlands, making com-
munication very difficult. The rugged, forested, mountainous terrain, which
imposed such a barrier to north±south circulation, also made it difficult to
reach the Caribbean coast. These geographical features, and the fact that
there were no precious metals on the isthmus that could be very profitably
exploited, condemned the region to a marginal and peripheral role within the
Spanish colonial empire. In fact, the Central American isthmus' only compet-
itive `value' was its strategic importance for crossing from one ocean to the
other (PeÂrez Brignoli, (1999)). However ± and in spite of the many attempts and
indubitable potential of other routes ± this strategic importance was only really
HeÂctor PeÂrez Brignoli 87
of value in Panama, and for one simple economic reason: the crossing there was
shorter than anywhere else.
It is therefore easier to understand why the countries of Central America,
which together formed the Audiencia de Guatemala in colonial times (1570±
1821) and went on to independence as a Federal Republic (1824±38), remained
marginal and peripheral for so long. At the same time, these enduring circum-
stances make possible an assessment of the revolution sparked off by the
sustained success of the export of coffee from the second half of the nineteenth
century.
The development of coffee
The expansion of the coffee industry followed a clearly defined chronological

sequence as shown in Figure 4.1 (Cardoso, 1975, pp. 9±55; PeÂrez Brignoli and
Samper, 1994; and Samper, 1993). It began early ± in the 1830s ± in the high-
lands of Costa Rica, with exports sent to England by the ValparaõÂso and Cape
Horn route, and in the decades that followed it went on expanding towards the
north of the Central American isthmus. Both the type of coffee ± araÂbigo, also
known as mild aromatic ± and the techniques of cultivation and processing
used in Costa Rica were adopted by the rest of Central America.2 The produc-
tion of coffee required rich soil, situated at an altitude of between 600 and 1200
metres above sea level, with abundant and seasonal rainfall ± preferably in May,
to coincide with the moment of flowering ± and with a dry season long enough
to allow for harvesting, processing and export of the coffee bean ± from October
or November until May of the following year. From Costa Rica to Guatemala,
the hillsides and valleys running down to the Pacific, with their covering of
volcanic soil, were ideal.
International markets considered that Central American coffees were quality
products; they were particularly useful for achieving aromatic blends, and
therefore they normally commanded a higher price than that quoted for
Brazilian coffee. These characteristics were a result not only of the type of coffee
but also of the techniques of cultivation, harvesting and processing. As the
nineteenth century drew to a close, coffee cultivation in Central America was
labour-intensive, resembling gardens rather than large-scale plantations. Mixed
cultivation ± including animal husbandry ± was common, especially on small
and medium-sized farms; trees of a different type were planted alongside the
coffee plants to provide shade and a wind-break; kidney beans and other pulses
were grown and there were also areas given over to natural pastures and woods.
Thus the use of the soil was relatively varied although food production was not
generally intended for the market.
The improvements incorporated into the coffee lands were labour-intensive
and required a continuous effort of maintenance. In order to obtain good
(a) Costa Rica, Guatemala and EI Salvador

100.0
10.0
millions of kg
1.0 Costa Rica

Guatemala
El Salvador
0.1
0.0
1940
1840
1844
1848
1852
1856
1860
1864
1868
1872
1876
1880
1884
1888
1892
1896
1900
1904
1908
1912
1916
1920
1924
1928
1932
1936
(b) Nicaragua and Honduras
100
10
millions of kg
Nicaragua
Honduras
0.1
0.01
1840
1844
1848
1852
1856
1860
1864
1868
1872
1876
1880
1884
1888
1892
1896
1900
1904
1908
1912
1916
1920
1924
1928
1932
1936
1940
Figure 4.1 Coffee exports, 1840±1940

Sources: (a) ANACAFE (1975); FAO (1947): El Salvador and Costa Rica, Anuarios EstadõÂsticos. (b) FAO
(1947); Belli (1975), pp. 2±30; Central Bank of Honduras, 1961.
yields, pruning and weeding were as important as abundant and regular rainfall.
As the loss of fertility of the soil was to a certain extent compensated for by the
presence of leguminous trees used for shade and a moderate use of organic
fertiliser, it took several decades before the yield was affected. The decrease in
yield became apparent in the 1920s and continued until the 1950s. As cultiva-
tion techniques underwent little change it followed that there were few techno-
logical changes in the way coffee was cultivated. As this was the case, the only
effective way of dealing with decreasing yields would have been a system of
gradual crop rotation and a phased replacement of the oldest coffee plants, but
during this period international coffee prices never reached the high levels
recorded before 1897. Therefore the principal incentive to push for these innova-
tions and thereby achieve a new expansion of production was lacking. In the
1930s the Central American coffee industry had two obstacles to overcome:
there was no further supply of good soil and most coffee plants were too old.
Harvesting the coffee was extremely labour-intensive, with no possibility of
mechanization. Furthermore, only the system of `picking out' ± several pickings
selected according to the ripeness of the coffee bean ± guaranteed a product of
the highest quality. There were two methods ± commonly called beneficio ± of
processing the coffee bean: the dry method, which was suitable for small-scale
production, and the labour-saving wet method, which, however, required
costly installations and an abundant supply of water. In Costa Rica the wet
method predominated from the 1840s whilst in the other Central American
countries both methods were used. Once again harvesting and processing are
labour-intensive, as already examined when discussing cultivation techniques.
It would be safe to assert that the secret of the quality of Central American
coffee, and therefore its competitiveness in the world market, had much to do
with the quality and organization of the workforce.
The expansion of the coffee industry was closely associated with the triumph
of liberalism ± the so-called `liberal reforms' ± in the final quarter of the nine-
teenth century. The changes in the organization of the state were as much a
requirement for, as a consequence of, the expansion of the coffee plantations
themselves. In Costa Rica liberal reform took place gradually, from the 1830s to
the 1880s. In Guatemala and El Salvador this reform brought about a much
more rapid and violent change between 1871 and 1884, whilst in Honduras and
Nicaragua the reforms started later and were incomplete. In Honduras, despite
the government's attempt to achieve a systematic transformation, the develop-
ment of agro-exports was limited and did not produce a true entrepreneurial
class. In Nicaragua, the liberal regime of JoseÂ Santos Zelaya (1893±1909) ended
with a long period of civil wars and US intervention (1910±34).
In Costa Rica (Cardoso, 1973; Hall, 1976) the expansion of the coffee industry
did not compete with other products but coexisted with the agricultural colon-
ization of the Central Valley for almost a century. The social structure created
by the coffee industry consisted of an elite of exporters and `beneficiadores'3 on

the one hand, and a great variety of small and medium-sized producers who
used family labour on the other. The endurance of products such as cochineal
in Guatemala and indigo in El Salvador can be explained by high international
prices until the second half of the twentieth century. However, this situation
did not last after the advent of synthetic dyes, thereby providing an excellent
opportunity for coffee. The large-scale expansion of this crop demanded a
reassignment of resources that could only be achieved by drastic social change.
In Guatemala (McCreery, 1994; Cambranes, 1985) there were three factors in
the creation of the land market: (i) the expropriation of the substantial amount
of land owned by the Church; (ii) the sale of national lands, and (iii) the
pressure on indigenous communities to sell their land. This process accelerated
when the liberals gained power in 1871. Privatization was imposed very quickly
over a period of approximately twenty years in all areas suitable for the cultiva-
tion of coffee ± at altitudes of between 600 and 1200 metres ± on land sloping
down to the Pacific and the Caribbean. At an altitude of over 1200 metres, there
was a high plateau, inhabited mainly in the west by indigenous communities
who managed not only to hold onto their communal lands but also to obtain
land titles. The other key factor was the provision of labour. In Guatemala it was
a question of mobilizing the labour force of the indigenous communities,
where most of the Guatemalan population resided ± in the order of 70 to 80
per cent ± to work in industries such as coffee-growing which were particularly
labour-intensive. In order to achieve this the government revived ± or rather
recreated ± colonial systems, like the enganche (debt peonage) or the manda-
miento (another form of coerced labour). At harvest time, the indebted Indians
came down from their settlements to the coffee plantations under close milit-
ary surveillance. A permanent workforce was guaranteed by placing tenant
farmers on the plantations and granting them allotments for growing subsist-
ence crops and the possibility of building a hut. With regard to the use of the
soil ± and this was a model common right across Central America ± every farm
had an area planted with coffee, another with subsistence crops and a sizeable
amount of land for cattle and forestation. These social relations marked a kind
of `social pact' between the Guatemalan coffee-industry elite and the indigen-
ous communities: ethnic divisions were frozen and the system functioned with
a mixture of paternalism and violence, in varying doses. In some ways these
social relations were a recreation of the colonial social structure but in a much
stronger and more dynamic external market than in the past. In Guatemala the
coffee industry really took off between 1880 and 1905.
El Salvador (Browning, 1971; Lauria Santiago, 1992) faced problems similar to
those in Guatemala. The agrarian reorganization fostered by the `liberal reform'
was even more drastic as the communal lands disappeared altogether. However,
where the expropriation of the communities played a crucial role was not so
much in the land market but rather in the labour force. In El Salvador, more
bound to the world market by indigo and more densely populated than any of
the other countries, the development of private property in the period prior to
the expansion of the coffee industry had been greater and had already created a
very unequal distribution of land, with the coining of the well-known binomial
latifundio-minifundio (large estate/smallholding) (Lindo-Fuentes, 1990, pp. 149±
51). Therefore, the expropriation of communal lands not only affected land
suitable for coffee but was responsible for the creation of a landless peasantry
which had no alternative but to work on the new coffee plantations. The coffee
industry in El Salvador underwent a more rapid development between 1880
and 1910 than anywhere else in the isthmus.
Coffee-growing in Nicaragua developed in two quite distinct areas (see
Radell, 1964, Dore and Gould in PeÂrez Brignoli and Samper, 1994). The first
was located in the uplands to the south of Managua, very near the Pacific coast.
Although this was a relatively small area, the low cost of transport had, with
government help, enabled the coffee industry to expand there since the 1860s.
In the second area ± which was far from the coast, in the mountainous region of
Matagalpa ± coffee production expanded at the end of the nineteenth century.
Because of the distance, the cost of transport was very high and although there
were plans to extend the railway system this never happened. With the expan-
sion of the coffee industry came a certain amount of privatization of communal
and common lands. It is important to note that in Nicaragua coffee never
achieved the dominant status that it did in Guatemala, El Salvador and Costa
Rica. Instead it coexisted with other important industries such as animal
husbandry and left room for the development of a sizeable racially-mixed
peasantry involved in the growing of subsistence crops. This was particularity
linked to resource availability. Because of the altitude and composition of the
soil the amount of land naturally suited to coffee-growing was relatively small
and confined to the two above-mentioned areas, which meant that the lack of
labour caused quite a problem. In order to ensure a supply of labour,
the landowners, supported by the government, employed a debt-bondage
system. Thus there was a contrast between the social profiles of both areas:
large estates ± which had already existed before the advent of coffee ± predom-
inated in the region around Managua, with a coerced labour force, whilst in the
mountains of Matagalpa family-based labour in medium and small enterprises
abounded.
In Honduras coffee was introduced systematically but not very successfully
during the 1870s and 1880s. This can be easily explained. In the rugged moun-
tains of the Honduran hinterland there was no large, continuous area with
volcanic soil such as there was in Guatemala, El Salvador, Costa Rica and in part
of Nicaragua ± only to the south of Managua. As a result, coffee could only
be produced profitably in certain valleys of the Honduran hinterland. The
isolation of these areas from one another and the high cost of transport
explains why coffee only expanded in Honduras after 1945, precisely during a
period of very high international prices when the routes of communication
were first being modernized.
With the coffee boom the transport system was totally revolutionized. In
particular it became essential to improve and extend the internal road network.
Until well into the twentieth century the ox-drawn cart was the favourite
means by which to transport the coffee beans from the farms to the processing
centres and from there to the railway station. Suddenly, the construction of the
railway to the ports became essential to reduce transport costs from the central
highlands to the coast. Because of the geographical location of the coffee-
growing plantations the Pacific ports were the most accessible, which meant
that the first railways in Guatemala, El Salvador and Nicaragua converged there.
However, in the absence of the Panama Canal ± completed in 1914 and open to
commercial traffic in 1919 ± the Pacific route did not seem the most economical
option. As soon as they were able, the governments of Costa Rica and
Guatemala built railways to the Caribbean coast, or, as they then preferred to
say, `to the Atlantic'. This work was long and difficult, and the cost was invari-
ably underestimated. However, it was finally completed successfully. Around
1910 both Costa Rica and Guatemala could rely on a rail network that passed
through the central highlands and provided an efficient route towards both
oceans. Years later the railways of El Salvador were still connected to those of
Guatemala.
Coffee entrepreneurs were a heterogeneous group. As well as traditional
families firmly rooted in the colonial era there were recently `risen' creoles
and mestizos, who invariably came from political and military backgrounds.
Moreover, from earliest times many foreign immigrants entered the coffee
sector as exporters. In Guatemala there was a majority of German entrepren-
eurs, whereas in El Salvador the immigrant community spanned many nation-
alities. In Costa Rica they were predominantly English and German, while in
Nicaragua there were fewer foreign immigrants than in the rest of Central
America. The profile of the coffee-growing elite was, in any case, unmistakable.
Their power was derived first and foremost from the export trade and from
their control over processing, as well as handling important credit arrange-
ments for the producers; in Guatemala and El Salvador this dominant position
was reinforced and extended by the existence of large estates. In El Salvador the
coffee-growing elite was not composed of only 14 families as commonly
thought, but comprised a little over 100. In the 1920s it is clear that in Guate-
mala, Costa Rica and Nicaragua there was an elite of 100±200 families, in each
country, at the top of the coffee-growing industry. Their interests were not of
course limited to coffee, but included other agricultural, commercial, financial
and industrial activities.
The banana plantations
The banana export business (Kepner and Soothill, 1957; Posas, 1993; pp. 111±
65; and Ellis, 1983) was established towards the end of the nineteenth century
and soon became as important as coffee (see Figure 4.2). Geographically, ba-
nanas were grown in an èmpty' area, namely some alluvial valleys along the
Caribbean coast.
The introduction of the cultivation and export of bananas in Costa Rica was
closely related to the problems of the construction of the railway, during the
many years that it took to link the central highlands of Costa Rica to the
Caribbean coast. In the 1880s Minor Keith, the North American entrepreneur
responsible for the construction of the railway, also became involved in the
cultivation and export of bananas and managed to obtain a full share of both
businesses. Soon afterwards he extended his plantations to Bocas del Toro in
Panama and Santa Marta in Colombia. The business worked closely with the
railways: plantations sprang up along the riverbanks and the trains could thus
quickly transport the fruit to the port. From there this perishable and relatively
fragile product was transported by steamships, which were soon equipped with
refrigeration, to the consumer markets of New Orleans and the east coast of the
USA.
However, the `banana-railway' diptych did not only refer to the technical
organization of production. Governments signed contracts with the railway
100 000
10 000
Thousands of bunches
1 000
of bananas
100
Costa Rica
10 Honduras
Guatemala
Nicaragua
1
1883
1885
1887
1889
1891
1893
1895
1897
1899
1901
1903
1905
1907
1909
1911
1913
1915
1917
1919
1921
1923
1925
1927
1929
1931
1933
1935
1937
1939
Figure 4.2 Banana exports, 1883±1940

Sources: Bingham (1974); Jones (1966); Belli (1975), pp. 2±30; Posas (1993), pp. 152±3; Central Bank
of Honduras (1957); FMI (1949), p. 21.
and banana companies granting them ownership of large tracts of virgin territ-
ory along the main railway lines and branchlines, and tax exemptions for the
import of rolling stock required by the trains. Thanks to these facilities, within a
few years there were several extraordinarily prosperous companies involved in
the production and export of bananas.
The expansion of the plantations on the Caribbean coast of Central America
began in Costa Rica and Panama in the 1880s, spreading into Honduras and
Nicaragua and reaching Guatemala a little later. The lushness of the new soil
facilitated a rapid expansion with high yields; however, by the beginning of the
twentieth century diseases which attacked the plantations had already
appeared. As agronomic research took so long to find solutions for this menace,
banana glut and famine occurred in close succession, affecting practically all
the areas suitable for banana-growing on the Caribbean coast. This never-
ending search for new territories, and the abandonment of the old, had already
given rise to the introduction of banana plantations on the Pacific coast of
Panama (Puerto Armuelles). The same phenomenon occurred, years later, in
Costa Rica and Guatemala.
It is possible to identify three distinct phases in the organization of the
banana industry. The first, from 1870±98, was characterized by the importance
of small producers and the appearance of several companies involved in trans-
porting and commercializing the product. Towards the end of this period
several commercializing companies merged and the market started to become
more focused. The second phase, 1899±1929, is defined by the domination of a
few companies headed by the United Fruit Company which managed to draw
together the vertical and horizontal strands of the banana industry. Just one
company controlled all the stages, from plantation to embarkation, using boats
with refrigeration chambers. After 1930 the number of banana companies
contracted even further and the onset of economic depression and banana
disease brought about a significant decline in the industry. It was not until
the 1950s that there were very important changes in this industry, such as the
introduction of disease-resistant species and the packing of the fruit in boxes
instead of exporting them in bunches.
Honduras was the only country where integration into the world market was
almost exclusively dependent on bananas. As already mentioned, the liberal
governments at the end of the nineteenth century attempted, unsuccessfully,
to extend the cultivation of coffee. Silver mining still enjoyed only limited
success, so nothing changed until the banana boom at the end of the nine-
teenth century. In the 1920s Honduras became the largest producer of bananas
in the world.
The presence of banana companies struck a novel note in the relationship
between the state and entrepreneurs; very serious situations of corruption and
confrontation were caused by granting new concessions or renewing those
already in existence. The worst example was when the conflicts between the
United Fruit Company and the Cuyamel were reflected in the Honduran civil
wars of the 1910s and early 1920s. Some writers have described these episodes,
marked by the occasional deployment of North American marines, as `banana
wars'. Viewed in perspective this description does not seem exaggerated. In any
case, it is clear that the Honduran civil wars ended once the Cuyamel and the
United Fruit Company merged into a single company in 1930.
In Costa Rica the situation was different (Casey, 1979; Carcanholo, 1978).
There was only one banana company (the United Fruit Company) operating in
the country, but with several national producers engaged in the industry.
Although they could only export through the United Fruit Company, their
political and social influence was considerable. Overall, the Costa Rican state,
which had been shaped and dominated by coffee-growing interests, was less
sensitive to the pressures of the banana company, and this is reflected in the
relatively early renegotiation of the contracts with the company. Moreover, in
the case of Costa Rica the United Fruit Company only managed to control the
railway between San JoseÂ and LimoÂn: the San JoseÂ-Puntarenas line, completed
in 1910, always remained state property and was under its control. In Nicar-
agua, the North American banana companies ± in particular the Standard Fruit
Company ± combined such activities with timber exploitation and mining. But
continual civil wars (1909±34) had a very damaging effect, culminating in 1931
when the Sandinistas set fire to plantations and installations in the Puerto
Cabezas area. Moreover, diseases like the so-called `Panama disease' suddenly
took hold. In this context it is no wonder that the banana companies virtually
went out of business. In Guatemala (Dosal, 1993) the United Fruit Company
negotiated an agreement in 1904 and immediately, in addition to the cultiva-
tion and export of bananas, entered into the railway business, with almost
immediate results: by 1912 the International Railways of Central America
(IRCA), a subsidiary of the United Fruit Company, controlled the railways,
unloading bays and port installations in Guatemala.
As the banana plantations were in areas that had only recently been opened
up to agricultural exploitation, there were very few disputes over land. Labour
on the plantations generally came from Jamaica. Black workers had come to
build the railways and continued to arrive in large numbers during the con-
struction of the Panama Canal (for the French phase as well as the US). Once
this work had been completed, large contingents of black workers moved from
Panama to Honduras and Costa Rica.
The nature of social relations on the banana plantations is noteworthy. On
the one hand, there was a large wage-earning proletariat which soon developed
syndicalist grievances. On the other hand, the officials of the banana compan-
ies practised and spread a type of racism not very different from that which
existed in the USA. Immigration from Jamaica ceased around 1930, and after
that date many peasants of mixed race from the central highlands joined the
workforce on the banana plantations.
Contrasts between coffee and bananas
From a socioeconomic point of view, the main difference between the way in
which the coffee and banana industries were organized was in the introduction
of two or three foreign banana companies with a very high degree of vertical
and horizontal integration. Furthermore, the fact that the plantations were
situated on low-lying land previously virtually uninhabited made the banana
settlements seem disconnected from the rest of their respective countries. Thus,
many writers have described the banana-growing areas as ènclave' economies.
Although there is an undeniable element of truth in this image, it is also true
that there is a large degree of exaggeration. The banana settlements were not
static and the tremendous initial isolation with respect to the rest of the
country was undergoing rapid change.
From the point of view of the social structure the contrasts were also worthy
of note. In the banana plantations, two or three very large foreign companies
competed with a few national producers and with a large number of wage-
earning labourers. Everything was organized around the large foreign company
versus salaried workers relationship. Further, the fact that during the early
decades this workforce was largely black and of Jamaican origin introduced a
significant element of cultural diversity. The relationship between these black
workers and their mixed-race companions ± whose number was steadily
increasing ± was complex and forever swinging between solidarity and conflict.
In addition, companies threatened by strikes and movements for workers'
rights did not hesitate to exploit these ethnic and cultural differences to disem-
power the workers.
In the coffee communities the social structure was much more complex.
There were at least two distinct opposing relationships: (i) the elite made up
of exporters, beneficiadores, and big landowners versus a large number of me-
dium-sized producers; (ii) all the producers, opposed in their turn to the
permanent agricultural workers who had settled on the estates, and also to
the workers who helped with the harvest and processing. Coffee production
required a sizeable number of bosses and workers but left plenty of opportunity
besides for family work in all stages of production and processing. This also
contrasted with the banana plantations where there was room only for single
men and where middle-ranking workers were much fewer in number.
Something common to both activities/industries was the absence of forward
linkages within the national economies, and the requirement of a relatively
large stream of investments. The coffee plantations required constant reinvest-
ment in work both to keep the bushes producing ± in the 1930s many coffee
plantations in Central America were already more than 50 years old ± and to
replace them. In the banana plantations, on the contrary, they normally moved
to other areas once the yield dropped or there was evidence of disease. As was to
be expected, the strategy for dealing with both situations was quite different.
The coffee growers tried, by all means possible, to reduce the monetary cost of
permanent workers and so explored other systems of settlement and payment
in kind with rations of food produced on the estate itself. The banana compan-
ies, for their part, were anxious to obtain additional concessions from public
lands, always leaving a considerable amount of uncultivated land in reserve.
Exports and the national economy
According to the calculations of Bulmer-Thomas (1987, appendix tables),

around 1925 agricultural exports accounted for 20 to 30 per cent of GDP.
However, this proportion, though significant, does not reveal in full the relative
weight of the exports in Central American economies. We shall examine suc-
cessively how the exporting boom permitted the monetary, banking and finan-
cial system to develop, and, finally, how a true national market arose through
imports.
Money and foreign debt

At the end of the nineteenth century and the beginning of the twentieth, the
countries of Central America experienced serious monetary instability. It was
not easy to abandon the colonial monetary system, based on silver coins. Ever
since the international price of silver had begun to fall in 1877, Central Amer-
ican currencies had been losing value against those in gold which ± given the
international preference for payments in gold ± was a serious problem, as much
for the payment of imports as for the interest on foreign debt. On the other
hand, as there was not much minting activity in Central America, silver coins
from South America, Mexico and North America circulated freely along with
the official coins of their respective countries, not to mention, of course,
worthless, adulterated coins containing less precious metal than required by
law. Within each country there were also paper notes in circulation which were
mostly inconvertible ± issued by the government or government-authorized
banks ± and there was no lack of transactions paid for in cocoa beans or salaries
cancelled with tokens issued by the plantation owners themselves. It may be
said that monetary circulation was universal but that there were different types
of currency circulating: metallic coins, paper money and various substitutes for
money. We know little about the connection between these different means of
exchange.
Unification under a national monetary system came about in the twentieth
century, and in some cases, such as Honduras, did not become established until
the 1950s. It is not possible to list the problems of each case here, although it is
worth pointing out some dates and processes.
The government of Costa Rica adopted the gold standard in 1896 (Soley
Gu È ell, 1926), just one year before the beginning of the long coffee crisis caused
by overproduction in Brazil and which continued until 1907. This created
difficulties which meant that the reform did not become effective until 1900.
It enjoyed remarkable success although the government, drained of funds, had
to hand over the Costa Rican railway to the Northern Railways Company (a
subsidiary of the United Fruit Company) in 1905.
The second country to adopt the gold standard was Nicaragua (Cumber-
land, 1928; Hill, 1933) in 1912, as a result of the fall of Zelaya and US
intervention. This reform followed the Dawson Accords (1910) and the
Knox±Castrillo treaty (1911). Nicaragua received an initial loan of US$ 1.5
million guaranteed by customs revenue, and proceeded to implement the
changes. The new currency had parity with the dollar, and the paper notes
issued during the governments of Zelaya, Madriz and Estrada started to be
withdrawn from circulation. During the years that followed, the system
was underpinned by new loans and similar guarantees. As in the case of
Costa Rica, the successful adoption of the gold standard implied an arrange-
ment for paying foreign debt. However, in the case of Nicaragua it implied
North American intervention in the collection of revenue and the presence of a
detachment of marines. In other words, the settlement of the debt and the
adoption of the gold standard took place within the framework of a virtual US
protectorate.
El Salvador (Fonseca, 1924) tried to adopt the gold standard in 1892 but the
plan failed. This reform was eventually successful in 1919 ± a rather ideal
moment because the War led to a rise in the price of silver. For a country that
was still on the silver standard it was the best time to change. There were
difficulties when the price of coffee fell in 1920 and 1921 and the subsequent
suspension of the debt interest payments. The situation was redeemed in 1922±
23 with the approval of a US$21.5 million foreign loan guaranteed by 70 per
cent of customs revenue.
Guatemala adopted the gold standard in 1923, immediately after a prolonged
period ± over 25 years ± of devaluation of the silver peso and with inconvertible
notes in wide circulation. Within three years the new currency had completely
replaced the old notes; the creation of the Central Bank of Guatemala in 1926,
with a 10±year guaranteed monopoly for the issuing of currency, completed the
framework of reforms.
Honduras was the last country to adopt the gold standard and to have, as
previously mentioned, a unified system of monetary circulation. There were
several attempts at monetary reform in the 1920s but the government was not
able to achieve this until 1931. The new monetary unit, called Lempira, began
Table 4.1 Years in which there was no payment of interest on external public debt
Countries Years Years of non- % of years

considered payment of debt of non-payment
Costa Rica (1827±1928) 102 23 23

Guatemala (1827±1928) 102 66 65
Honduras (1827±1928) 102 93 92
Nicaragua (1827±1928) 102 53 52
El Salvador (1827±1928) 102 36 36
Source: Jones (1971) [1931], p. 247 ff.
to circulate, although not without problems. Note that the IMF delegation that
visited Honduras in 1949 included the following in their report:
Honduras has never had a well-developed monetary system. Currently, the

major part of the currency in public circulation consists of US silver coins.
(Central Bank of Honduras, 1957, p. 10)
The adoption of the gold standard and the organization of a unified national
monetary system was, amongst other things, tied up with the settlement of
foreign debt. The countries of Central America began to get into serious debt in
the second half of the nineteenth century when their governments sought
loans from Britain in order to boost railway construction. Both the embezzle-
ment of funds and overly optimistic estimates of the final cost of the works to
be financed were responsible for the spiralling of debt: within a few years, all
the governments found it difficult to pay the interest and fell into arrears. The
record of arrears in repaying foreign public debt in the period 1827±1928 (Table
4.1) more than illustrates this point. Costa Rica shows the best management of
arrears, not paying its debt for 23 years, whilst Honduras at the other extreme
spent 93 years without paying. As all countries always negotiated the terms and
arrangements for payment, arranging in their turn new loans, the above-
mentioned table is a complementary indicator of the relative success of each
country in its relationship with the world market.
Credit and the banking system

Credit was for a long time in private hands. At the beginning of the period
under discussion, export companies and commercial companies advanced
money to the producers, charging interest at a rate more akin to profiteering
than to capitalist credit. The establishment of the first commercial banks, run
by identical interests, did little to improve matters other than bestowing an air
of modernity and progress on the creditors.4 As could be seen in the previous
section, state intervention was limited to the gradual regulation of the issuing
Table 4.2 State intervention in banking development: central bank and rural credit
`De facto' central banka Central banksb Rural and long-term creditc
Costa Rica 1914 International 1950 Central Bank of 1914 Rural boards
Bank of Costa Rica Costa Rica (International Bank)
1937 Banking reforms 1937 Section of Agricultural
and Industrial Credit
Guatemala 1926 Central Bank of 1946 Bank of Guatemala 1933 National long-term
Guatemala Credit
1933 Mutual Savings
El Salvador 1934 Central Reserve 1961 Central Reserve 1935 Long-term Credit
Bank Bank of El Salvador Bank of El Salvador
1940 Rural Savings Banks
Nicaragua 1941 National Bank of 1960 Central Bank of 1930 Long-term Credit
Nicaragua Nicaragua Bank of Nicaragua
Honduras None 1950 Central Bank of 1950 National
Honduras Development Bank
Notes: In 1924 the National Insurance Bank was created in Costa Rica as a state monopoly.
a
Mixed capital institutions controlled by the state.
b
Public institutions (autonomous and decentralized).
c
Mixed capital institutions controlled by the state.
of currency and negotiation of foreign loans to promote works of infrastructure.

As a result, during the first decades of the twentieth century what Bulmer-
Thomas called a `de facto central bank' (Bulmer-Thomas, 1987, p. 74) became
established ± that is, the introduction of an entity of mixed capital controlled
by the state, which acted as a government financial agent and enjoyed a
monopoly in the issuing of currency (see Table 4.2).
State intervention in the sphere of rural and mortgage credit provides
another subject for discussion. In 1874 Guatemalan liberals attempted to
form a bank with funds obtained from the confiscation of all Church wealth;
but the experiment failed and in 1876 the National Bank of Guatemala went
into liquidation. In Nicaragua the liberal regime of JoseÂ Santos Zelaya (1893±
1909) favoured those plantation owners who were supportive of his govern-
ment (Teplitz, 1974, pp. 192±5; Young, 1925, pp. 124±6) with subsidies ± in
paper money ± but the inflationary environment ruined whatever medium-
term benefits were derived from such financing. Indeed, state intervention only
became more effective in this field following the crisis of 1930.
It is generally accepted that bank credit was limited and concentrated in few
hands. This had a detrimental effect on small and medium-sized producers,
who were thus left without any alternative but to use credit offered by profit-
eers. The government was slow and reluctant to act in this area. Examination of
values per capita (Table 4.3) reveals a significant regional contrast. In 1939 the
total monetary supply of Costa Rica was almost three times higher than that of
the rest of Central America. This was clearly indicative of a much more mon-
etized and therefore modern economy. Similar contrasts and conclusions
Table 4.3 Money supply per capita in 1939 (US$)
Dollars per inhabitant
Guatemala 6.8
Honduras 9.0
El Salvador 6.1
Nicaragua 7.1
Costa Rica 19.7
Source: Extracted from IMF (1950).
are apparent if available figures on banking credit per capita in 1950 are exam-
ined: 10 dollars in Guatemala, 7 in Honduras and 51 in Costa Rica.5
Urbanization, migration and attributed effects

Urbanization was one of the most significant effects attributed to the export
boom. With its need for services, the coffee-growing industry encouraged the
development of intermediate towns in two ways. First, the population of the
existing cities and towns increased and, secondly, new pockets of urban devel-
opment appeared filling underpopulated areas. There was a similar urbaniza-
tion process in the banana-growing areas, though less intense and more closely
attached to the railway routes. This was certainly to be expected since it con-
cerned hitherto empty areas. On the other hand, given the high degree of
vertical integration in the banana industry, the demand for services and man-
ufactured goods turned out to be much lower than in the coffee-growing areas.
Urban growth was accompanied by internal migratory movements of various
kinds. The increase in population gradually pushed the unemployed or landless
peasant population towards the fringes of the coffee and banana plantation
areas. With the single exception of El Salvador, where there were no unoccu-
pied areas, in all the countries of Central America there was an internal settle-
ment border which advanced gradually towards the PeteÂn in Guatemala, the
Mosquitia in Nicaragua, and the north and south of the Central Valley in Costa
Rica. In Honduras these population movements were directed towards various
regions of the hinterland, occupying previously unpopulated areas separating
the longest-established areas of settlement. In the 1930s the banana areas began
to attract Salvadorean immigrants to Honduras and Nicaraguans to Costa Rica.
Imports and the national market

With the sustained export boom every country was able to have at its disposal a
permanent and constant abundance of imported goods. Non-durable consumer
goods (Table 4.4) were those in most demand and this changed very little until
the 1950s. The feeble importance of raw materials and intermediate goods is a
clear indication that the effect of ìmport substitution', which can be observed
Table 4.4 El Salvador and Costa Rica: composition of imports, 1946 (%)
Costa Rica El Salvador
Consumer goods 53.3% 52.0%

Intermediate goods 22.3% 33.1%
Capital goods 20.2% 14.9%
a
Fuels 4.1%
a
Included in intermediate goods.
Sources: AlbarracõÂn y Brignoli (1977); Bovedani (date n/a.), Costa Rican External Trade
Statistics (1907±1946).
from earliest times in countries like Argentina, Brazil and Colombia, does not
occur in Central America until after the period under discussion. In other
words, during the early part of the twentieth century there was little economic
diversification in Central America. During the 1930s depression the typical
response was the expansion of agriculture for the domestic market (Bulmer-
Thomas, 1987, pp. 79±82).
The role of imports, which were mainly composed of consumer goods, was
confined to completing the integration of a national market of goods and
services. Integration may seem a word little suited to describing, for example,
a market which as far as the workforce was concerned displayed a great deal of
diversity and had many imperfections ± systems of forced labour, remuneration
in kind ± and that as far as credit was concerned included relatively primitive
systems for paying advances. However, we should not overlook an even more
fundamental fact: the continuing export boom and the scale of production
meant that all other economic activities were subjected in some way to the
dynamics of the external sector. The distribution of imported consumer goods
reached and affected, in one way or another, all sectors of the population. This
was the first time this had occurred in the history of the isthmus.
Investment for development and public expenditure

Until the 1950s economic growth was basically fuelled by the increase in coffee
and banana exports and an increase in population. Once production became
export-oriented, towards the end of the nineteenth century, there were no
significant technological changes, so population increase was largely respons-
ible for economic growth and the economy had to adapt to fluctuations in the
world market. If we consider the population variable as exogenous, growth can
be seen to be wholly determined by the revenue generated by exports. Thus it is
necessary to examine in more detail the use and purpose of this revenue.
As already indicated, exports served largely to finance the import of consumer
goods, and customs taxes were responsible for more than 50 per cent of the
state's resources. Did the entrepreneurs reinvest their profits? Overall, probably
hardly at all. From the First World War to the end of the 1940s, the international
price of coffee and bananas was generally stagnant; thus there was no incentive
to bring about technological change and increase efficiency ± it was enough to
maintain the capacity of the production installations and extend them, if at all
possible, thereby minimizing monetary costs. If reinvestment in the same busi-
ness was not very attractive, neither were there great incentives to invest in other
industries. The tiny size of markets, determined as much by the small population
as by their lack of purchasing power, was a definite obstacle in the rise of import
substitution industry. Under these conditions it is no wonder that a large part of
the income obtained by the exporters should be spent on extravagant consumer
goods or invested abroad. In the case of the banana companies, investment
abroad had always been an integral part of their investment strategies.
In this context, it is not surprising that state expenditure constituted the
main source of investment for future development. Indeed, in the final decades
of the nineteenth century, governments began to encourage the construction
of railways, ports and roads, with recourse to foreign loans guaranteed by the
future income promised by foreign trade. In spite of slipping up on more than
one occasion, it has to be said that this strategy, at the time, produced good
results. The way in which public expenditure is distributed can be examined in
Figures 4.3 and 4.4, in which such information is presented for Guatemala and
Costa Rica.
100%
Debt
War
Public sector
Development
Education
80% Administration
60%
40%
20%
0%
1911
1883
1886
1890
1893
1895
1898
1900
1902
1904
1906
1908
1913
1915
1920
1922
1924
1926
1928
1930
1932
1934
Figure 4.3 Guatemalan government expenditure, 1883±1935 (classified by destination)

Note: Information is missing for the years 1884, 1888, 1889±91, 1897, 1909, 1917±19. These years are
ommitted from the horizontal axis.
Sources: Guatemala (1883±1935), Memorias de la SecretarõÂa de Hacienda y CreÂdito PuÂblico, Adler,
Schlesinger and Olson (1952).
100%
80%
60%
40%
Debt
War
20% Health
Public sector
Development
Education
Administration
0%
1870
1873
1876
1879
1884
1887
1890
1893
1896
1899
1902
1905
1908
1911
1914
1917
1920
1923
1926
1929
1932
1935
1938
Figure 4.4 Costa Rican government expenditure, 1870±1940 (classified by destination)
Note: Information is missing for the years 1880±1. These years are ommitted from the horizontal axis.
Sources: RomaÂn Trigo (1995).
These two cases were selected because they are representative of the contrast-
ing trends which can be observed in the region. In Guatemala ± and the
Guatemalan model is repeated in El Salvador, Honduras and Nicaragua ± the
burden of military spending, debt-service and administration costs conspired
continually against increasing the percentage spent on public works and educa-
tion. In Costa Rica, where to start with there is a very similar spending structure
to the one mentioned above, it is clear how, on the contrary, by about 1920 the
trend is reversed and expenditure on public works and education gain in
importance over those on defence and administration. As we shall see below,
this is what finally and in no uncertain terms differentiated Costa Rica from the
rest of Central America. Why this should be so can be explained: in economies
such as those of Central America the only element in the early part of this
century that could be effectively left aside for the future, apart from those
investments in infrastructure already in existence, was investment in public
works and human capital.
Exports and the nation state
In Central America the social foundations of the liberal state came from coffee-
growing. In its political history this corresponds to the so-called `liberal
reforms'. In Guatemala and El Salvador the process was very rapid and adopted
the style of a `revolution from above'. The new liberal rulers, under the
notorious aegis of the Guatemalan Justo Rufino Barrios, seized power by
force and in a few years developed a full programme of coffee-export promo-
tion, changes in agrarian ownership and construction of railways and ports.
The result was striking. Coffee exports increased rapidly and a powerful
business elite came into being straight away, joined by some foreign nationals.
National wealth was thus identified with coffee and the social structure
provided the vehicle by which that image could also be shared by the lower
classes through a complex combination of paternalism and violence. The
`from above' character of this revolution needs stressing. In practice things
were not very different, in political terms, from the previous conservative
period. Patronage and paternalism went hand in hand in the exercise of
power, which had very little to do with the impeccable framework of liberal
laws and regulations.
With less purity, but with similar results, liberal reform also took place in
Honduras and Nicaragua. In the case of the former, the programme of eco-
nomic reform failed and only more than twenty years later, with the expansion
of the banana plantations, did integration into the world market bear the
expected fruits. In the Nicaraguan case, the fragmentation of elite interests
put a stop to liberal reform, whilst the civil war and US intervention created a
long parenthesis of twenty years (1909±34). At the end of this period it could be
said that Anastasio Somoza completed the work begun by Zelaya. From the
point of view of the exercise of power, the situation in Honduras and Nicaragua
was very similar to that in Guatemala and El Salvador.
If in the rest of Central America it is possible to talk of a liberal revolution `from
above', in Costa Rica, on the other hand, the process of political transformation
adopts the form of a revolution `from below'. This means that liberal reform
took place gradually over almost a century (1821±1905), in a process of con-
tinual interaction between the coffee-growing entrepreneurs, civil society and
military leaders. The extension of the channels of political participation was
possible and even natural in a social context in which the middle-sized and
small agricultural producers enjoyed a strongly established position.
The later development of the banana export business had special implica-
tions of its own. In Costa Rica and Guatemala banana interests developed in a
political arena already marked out by coffee-growing interests. In Honduras on
the other hand ± owing to the failure of liberal reform ± the latter appeared and
expanded in opposition to an outdated landowning oligarchy whose interests
had difficulty transcending the local and regional scale, with a familiar out-
come: a very weak state confronting the powerful ambitions of foreign banana
companies, characterized by a political economy frequently summed up by the
derogatory, caricaturist term `Banana Republic'.
Long-term development
Now that we have described the context and its most striking structural char-
acteristics, we can review the long-term evolution of the economies of Central
America and try to obtain some synthetic and comparable measures of its
successes and failures in the long term. In order to analyse the temporal trend
we shall closely examine the data gathered on the volume and value of the
exports.
The physical quantities of coffee exported (Figure 4.1) increased in value at a
constant average rate of between 0.7 per cent per annum in the case of Costa
Rica and 1.7 per cent in the case of El Salvador.
In the period under discussion, Central America exported all the coffee it
could produce and never stockpiled. The export curve therefore reflects the
internal capacity for growth ± that is, the rates of incorporation of land and
labour in the coffee-growing sector. The differences express, at the same time,
important national contrasts. The slow growth of Costa Rica and Nicaragua can
be easily explained in terms of the market of factors of production. In the first
case, labour shortage is to blame: because of the predominance of small and
medium producers, the rate of incorporation of lands suitable for coffee was very
slow. In this sense it is worth recording that in the main coffee-growing area of
Costa Rica, the Central Valley, with a total area of approximately 2700 sq km, the
ecological limit for this crop was reached in about 1930 ± that is, a century after
the coffee-growing industry started to expand. In Nicaragua, both the availabil-
ity of land (the most extensive areas lacked cheap transport) and the availability
of labour (there was strong competition from other agricultural and cattle-
raising activities) were to blame. The export curves of El Salvador and Guatemala
show much stronger rates of increase than those of Nicaragua and Costa Rica. In
both cases this can also be explained by the behaviour of the market of factors of
production: coercive methods ± Guatemala ± or the relative abundance of land-
less workers ± El Salvador ± assured the labour supply necessary for a much more
accelerated incorporation of land suitable for coffee-growing.
In purely economic terms it seems clear that the Salvadorean entrepreneurs
were much more successful in achieving a higher rate of growth than any of
their Central American counterparts. However, the brilliance of these achieve-
ments fades into insignificance in so far as extraregional comparisons are con-
cerned. Colombia ± a coffee-growing country wholly similar as regards the type
of product, the techniques of cultivation and the characteristics of the natural
environment ± saw its coffee exports grow (in volume) at a rate of 9.7 per cent
per annum between 1887 and 1899, and of 2.7 per cent between 1906 and 1940
(Samper, 1948, pp. 85±9). A systematic comparison of the factors that explain
this more than apparent Colombian `superiority' could throw new light on the
profound nature of the Central American coffee-growing economies.
HeÂctor Perez
Â Brignoli 107
(a) Costa Rica, Guatemala and EI Salvador

40.0
35.0 Guatemala
EL Salvador
30.0 Costa Rica
25.0
20.0
15.0
10.0
5.0
0.0
1864
1867
1870
1873
1876
1879
1882
1885
1888
1891
1894
1897
1900
1903
1906
1909
1912
1915
1918
1921
1924
1927
1930
1933
1936
1939
(b) Nicaragua and Honduras
30.0
Nicaragua
25.0
Honduras
20.0
15.0
10.0
5.0
0.0
1864
1867
1870
1873
1876
1879
1882
1885
1888
1891
1894
1897
1900
1903
1906
1909
1912
1915
1918
1921
1924
1927
1930
1933
1936
1939
Figure 4.5 Exports, US$ millions, 1926 prices

Sources: (a) Anuarios EstadõÂsticos of Guatemala, E1 Salvador and Costa Rica; Schoonover (1989),
Anuario de Estudios Centroamericanos, vol. 15(1), pp. 93±117, table 20. The Wholesale Price Index of
the US Bureau of Labor Statistics (1926 100), L15±25 series in Bureau of the Census (1949), pp. 233±
4. (b) Central Bank of Honduras (1957), FMI p. 31; Belli (1975), pp. 2±30; Vargas (1990), p. 261;
Schoonover (1989), Anuario de Estudios Centroamericanos, vol, 15(1), pp. 93±117, table 20. Wholesale
price index as in Figure 4.5(a).
Table 4.5 Rates of growth of physical volume of coffee exports and variability of annual
fluctuations
ra vb
Costa Rica (1854±1940) 0.7 3.7

Guatemala (1870±1940) 1.1 6.1
El Salvador (1885±1940) 1.7 8.1
Nicaragua (1885±1940) 1.1 3.7
a
Average annual rates of growth, in percentage terms. Calculated from the data in Figure 4.1 using an
exponential function adjustment All rates are statistically significant at a 95 per cent level.
b
Standard deviation of the first differences ± the absolute change from one year to the next. As the
average of the series is approximately zero, comparison of deviations does not present a problem and
can be regarded as an approximation of the annual fluctuations of each series.
Source: Data from Figure 4.5.
Let us now consider the annual fluctuations, using the same coffee export
curves. The variability of the quantities exported from one year to the next was
noteworthy but changed little in the long run. In other words, if the trend of
the series is eliminated its variance remains stable. Under these conditions the
standard deviation of the first differences (Table 4.5, 3rd column) clearly
describes the annual variability of each one of the series. Within this model,
exports from El Salvador and Guatemala show greater fluctuations than Costa
Rica and Nicaragua. In all cases the annual fluctuations can be explained by
pointing to two factors: climatic change (especially rainfall) and random varia-
tions; the limited data prevents, however ± at least for now ± the elaboration of
a strict econometric model to prove this hypothesis.
Examination of the physical volume of banana exports (Figure 4.2) is equally
illustrative as far as the dynamics of this sector are concerned. Costa Rica first
and then Honduras stand out as large producers compared with the much more
modest contributions of Guatemala and Nicaragua. Apart from the export of
coffee, banana exports experienced a rapid cycle of growth followed by another,
no less rapid, fall; annual fluctuations, for their part, are less noteworthy than
in the case of coffee. There is a simple explanation for these cycles as they can
be attributed entirely to diseases affecting the banana plantations ± the so-
called `Panama disease' and the sigatoka. In fact, the appearance of these
diseases caused a rapid drop in production which could only be halted, given
the nature of such pests, with the incorporation of new land. Given the parti-
cular concentration of banana-growing in the hands of a few companies, this
occurred as often within each country as in neighbouring countries. The
national framework is revealed as insufficient when these, as well as with
other conditions, are studied in the banana sector.
We now move on to consider the value of exports (Figure 4.5). It is import-
ant to explain first of all the origin and nature of the information used.
Government data regarding exports and imports have existed since the middle
of the nineteenth century; there are often gaps in these sets of data and they
cannot be considered very reliable until the end of the last century when the
customs register was modernized and statistics improved. However, as we do
not have other sets of data at our disposal we have to make do with the picture ±
at times diffused and blurred ± conveyed to us by these data. The data regarding
value, expressed originally in national currency, were first converted into US
dollars, and then into US dollars of 1926. Market exchange rates were used for
the conversion and ± for want of a better deflationary measure ± the wholesale
price index of the US Bureau of Labor Statistics with 1926 100. This option is
reasonable given that a large part of Central American imports came from the
USA.
Exports from Guatemala, El Salvador, Costa Rica and Nicaragua (Figure 4.5),
based above all on coffee, show a similar sustained upward trend from 1864 to
the end of the nineteenth century; stagnation and contraction until 1921; and
a short rapid rise until 1930, followed by an equally sharp fall from then until
1940. If even longer trends are sought, two periods can be identified: one of
rapid rise until 1897 in which the average value of exports grew by at least a
million US dollars per annum to approximately US$ 12 million, and a second ±
until 1940 ± of relative stagnation. During this last period the average value of
the series remains static but broad movements of rise and fall can be observed
around it.
Annual fluctuations, measured in percentage of change from one year to the
next, are illustrated in Figure 4.6. These variations were relatively strong and
had immediate repercussions on the amount of imports. Although these con-
tractions did not put the agro-export economy in danger owing to the dampen-
ing effect of the non-monetized sectors, they did endanger fiscal income and
public investment. As indicated above, state resources depended strictly on
customs revenue (60 per cent or more).
The performance of international coffee prices (Figure 4.7) may explain
almost point by point the trends just analysed. This was to be expected in
economies where the relative weight of coffee exports was overwhelming and
helps us to better understand a famous phrase by Ricardo JimeÂnez, who
affirmed not without humour that there was no better Chancellor of the
Exchequer than a good coffee harvest. As Ricardo JimeÂnez was President of
Costa Rica three times one has to concede that he knew very well what he
was talking about: the country was a price-taker in the international market so
that the value of exports ± and hence of fiscal revenue ± basically depended on
the size of the coffee harvest.
It is important to underline the following aspect of the evolution of the
international price of coffee in real terms (Figure 4.7): in 1897 a large upward
cycle came to an end, interrupted only briefly by the low prices of 1881±85. In
(a) Guatemala
160
Change %
140
Moving average (5)
120
100
80
60
%
40
20
–20
–40
–60
1891
1921
1864
1867
1870
1873
1876
1879
1882
1885
1888
1894
1897
1900
1903
1906
1909
1912
1915
1918
1924
1927
1930
1933
1936
1939
(b) EI Salvador
160
140
Change %
120 Moving average (5)
100
80
60
%
40
20
–20
–40
–60
1870
1873
1876
1879
1882
1885
1888
1891
1894
1897
1900
1903
1906
1909
1912
1915
1918
1921
1924
1927
1930
1933
1936
1939
1867
1864
Figure 4.6 Annual fluctuations of exports (US$, 1926) (a) Guatemala (b) El Salvador (c)
Costa Rica (d) Nicaragua
Sources: Series as in Figure 4.5.
%
%
0
20
40
60
80
–60
–40
–20
100
120
140
160
40
20
60
80
100
0
120
140
–40
–20
160
–60
1864 1863
1867 1866
1870 1869
Figure 4.6 Contd.

1873 1872
1876 1875
1879 1878
1881
Change %
1882
1885 1884
1888 1887
Moving average (5)

1891 1890
1894 1893
1897 1896
1900 1899
1903 1902
1906 1905
(d) Nicaragua
(c) Costa Rica
1909 1908
1912 1911
1915 1914
1918 1917
1921 1920
1924 1923
Â
1926
Change %
HeÂctor Perez
1927
1930 1929
1933 1932
Moving average (5)
1936 1935
1939 1938
Brignoli 111
35.0
30.0
25.0
(US cents per pound)
20.0
15.0
10.0
5.0
0.0
1863
1869
1872
1875
1878
1881
1884
1887
1890
1893
1896
1899
1902
1905
1908
1911
1914
1917
1920
1923
1926
1929
1932
1935
1938
1866
Figure 4.7 Coffee price in New York (US$, 1976)

Sources: Commodity Yearbook (1939). Wholesale price index as in Figure 4.5.
the first forty years of the twentieth century coffee prices never again reached
those of that golden epoch.
Let us now consider the case of Honduras as shown in Figure 4.5(b). In the
first place, we do not have many data before 1901; secondly, banana prices
correspond not to market prices but to a fixed value, estimated by the govern-
ment based on that stated by the banana companies; and, thirdly, there were
exports that were not recorded by the Honduran customs. Although this last
factor tends to disappear in the 1920s, the second lasts until 1950. Only after
the creation of the Honduran Central Bank do we have at our disposal a
valuation of the banana exports adjusted in accordance with the FOB quotation
of that fruit. Thus, for example, during the period 1931±50 the declared valua-
tion remained fixed at US$0.50 whilst in 1947±48 the FOB price per bunch was
calculated at US$2.05 (Central Bank of Honduras, 1957, p. 34). In the rest of the
banana-exporting countries there was a similar situation. In Costa Rica the
valuation declared in the period 1895±1920 fluctuated between US$0.4 and
US$0.5 per bunch, rising to US$0.75 during the 1920s. When in 1950 the
Central Bank decided to calculate the real value (FOB) of the banana exports
the same thing happened as in Honduras: the value declared by the banana
companies was multiplied by three (Central Bank of Costa Rica, 1951, p. 88).
This being so, it is obvious that before 1950 we lack an adequate market
valuation of such exports.
Are these sets of data of no use? In fact that is not the case as at least they
reveal the physical volume of the exports and the resources seized by the
government through the tax per bunch exported (one US cent per bunch in
the case of Honduras). On the other hand, whilst the banana companies may
not have been obliged to bring into the country the foreign exchange resources
generated by the exports ± and this only happened after 1950 ± we can assume
that the series of undervalued exports reflect the resources generated by this
sector that were available within the country. In other words, if we had been
able to obtain a set of real values before 1950, what this set would reveal would
probably be company profits that, in general, never reentered the country.
With these observations in mind we can return to Figure 4.5(b), which even
with these problems is simple to analyse: the curve of the value of Honduran
exports shows a sharp rise in the 1920s followed by an equally dramatic fall in
the following decade.
The well-known phrase of Ricardo JimeÂnez quoted earlier is now in its true
context: for state resources and the general well-being of the population ba-
nanas count for much less than coffee. Moreover, it must be remembered that
in Costa Rica the total value of exports was shared almost equally by coffee and
bananas. In Honduras, where the state basically depended on the banana
sector, the situation was much more extreme. An International Monetary
Fund mission that studied the Honduran economy in 1950 was in a position
to assert, amongst other things, that
the operations of the fruit companies constituted such a distinct, self-suffi-

cient sector of the economy and found themselves orientated to such a
degree towards the export markets, that their impact on the economy is
only indirect and relatively unimportant.
(Central Bank of Honduras, 1957, p. 7)
Profile of the backwardness of Central America
In conclusion, we shall analyse some economic and demographic indicators of

the Central American situation around 1925 (Table 4.6). This will allow us to
evaluate, in perspective, the successes and failures of the development of agri-
cultural exports reached just before the world economic depression began,
towards the end of 1929.
Let us consider first of all the product per capita. El Salvador, Nicaragua and
Guatemala show approximately 60 per cent of the product which corresponds
to Costa Rica. The position of Honduras is more favourable, accounting for 79
per cent of the level of Costa Rica. These differences remain relatively constant
until the 1960s (Bulmer-Thomas, 1987, appendix tables). From the 1960s to the
1980s the gulf between Costa Rica and the rest of Central America widens
considerably. The product per capita is shown to match the other indicators per
capita. Costa Rica shows better figures than those of the rest of Central America
as regards the value of exports and imports as well as in kilometres of railways in
operation or government expenditure. The socio-demographic indicators also
reveal similar discrepancies: in Costa Rica it is typical to find less illiteracy,
Table 4.6 Central America around 1925; selected indicators
Indicators Guatemala Honduras El Salvador Nicaragua Costa Rica
Life expectancy at birtha 33.2 n/a 31.4 n/a 39.5

Rate of infant mortalityb 242.0 n/a 254.0 n/a 172.0
c
Total population 1841.0 756.0 1244.0 698.0 450.0
Rate of population growthd 1.7 2.1 1.3 1.1 2.3
% of urban populatione 9.5 10.0 20.0 16.9 22.6
Illiteracy (%)f 87.0 74.0 79.0 72.0 24.0
Indigenous population (%)g 65.0 21.0 30.0 10.0 2.0
Gross domestic product (GDP)
per capitah 193.0 243.0 186.0 172.0 308.0
Agriculture for domestic market
(% of GDP)i 17.0 22.0 25.0 44.0 14.0
Agriculture for export
(% of GDP)i 20.0 30.0 21.0 18.0 25.0
Manufacturing industry
(% of GDP)i 15.0 5.0 12.0 7.0 9.0
Public sector (% of GDP)i 7.0 4.0 5.0 1.0 3.0
Exports per capitaj 39.0 86.0 58.0 43.0 136.0
Imports per capitaj 34.0 39.0 42.0 36.0 79.0
Railways (km)k 801.0 1240.0 214.0 253.0 671.0
Km of railway for each 1000
inhabitants 0.44 1.64 0.17 0.36 1.49
US investmentsl 38.2 12.9 15.3 24.0 35.7
British investmentsm 15.0 6.0 15.0 6.0 25.0
Public expenditure on education
(per capita)n 0.6 0.3 0.6 0.5 1.6
Public expenditure on defence
(per capita)n 1.0 2.3 1.5 0.9 1.3
Public expenditure on developmentn 1.0 0.4 1.0 0.6 2.5
o
Government revenue per capita 5.1 9.2 8.2 6.2 14.3
a
In years, own estimates using birth and death registers, censuses and ìnverse projection' method.
b
Per thousand births, Idem.
c
In thousands, data from Collver (1965), and own estimates.
d
Average annual growth in percentages, data from Collver, op. cit., and Statistical Abstract of Latin America, vols
20 (1980), 24 (1985) and 26 (1986), and own estimates.
e
Population living in towns of more that 5000 inhabitants, census data.
f
As a percentage of the total population, census data
g
As a percentage of the total population, census data and own estimates.
h
Average 1924±26 in 1970 dollars, data from Bulmer-Thomas (1987), statistical appendix. To calculate the per
capita indices we have used population estimates which appear in (c) and not those of Bulmer-Thomas.
i
As a percentage of gross domestic product, average 1924±26, data from Bulmer-Thomas, op. cit.
j
In 1970 dollars, Bulmer-Thomas, op. cit.
k
Railway kilometres in 1921±22: data from Long (1925).
l
In millions of dollars, 1928, data from Winkler, quoted in Jones, op. cit. pp. 293±5.
m
In millions of dollars, data from Jones, op. cit.
n
Current dollars.
o
Current dollars.
greater life expectancy and a higher increase in population. In 1925 the per-
centage of illiteracy in Costa Rica (24 per cent) was similar to that of Chile and
Italy, and substantially less than that of Spain (43 per cent) and Mexico (59 per
cent) (Bunge, 1940, pp. 418±19).
Note that if we aggregate the indicators corresponding to Panama its values
would be more similar to those of Costa Rica than to those of the rest of Central
America. At least from 1920 it is clear that the isthmus is characterized by a
notable north±south divide. To the north of the San Juan river the demographic
and economic indicators reveal less wealth per capita, a large rural population
and high mortality rates. To the south of the San Juan river, in Costa Rica and
Panama, the situation is different. These countries can be seen to be in a much
more favourable situation given their position in the majority of the given
indicators, which is above the Latin American averages although below coun-
tries like Argentina, Chile, Uruguay and Cuba (Bulmer-Thomas, 1994, p. 444).
The data on Honduras merit a special mention. In the 1920s some indicators
show higher values than in subsequent years. I refer to the product and exports
per capita, and to the kilometres of railways per thousand inhabitants. During
the 1930s the Honduran indicators again resemble those of Guatemala owing
to the deep recession caused by the fall in banana exports as shown in Figure
4.5(b). In this sense the Honduran economic boom of the 1920s has all the air
of an exceptional opportunity, with no match before or since.
To sum up, the data in Table 4.6 reveal appreciable discrepancies in the nature
and consequences of economic growth, and the situation did not change much
in the following two decades. The levels of the internal product per capita
varied little. In Honduras, Guatemala, El Salvador and Nicaragua more than
60 per cent of the population were still illiterate. From 1910, when the laying of
the main railway lines was virtually complete, until the time they began to
construct the inter-American highway (1945), the road and rail networks and
transport experienced very little change.
The persistence of profound social inequalities was rather more serious and
much more difficult to change. The boom in agricultural exports expanded and
in a certain way `froze' the ethnic divisions of Guatemala at the same time as it
created new ethnic barriers in the banana areas of Honduras and Costa Rica. In
El Salvador the ethnic conflicts and the polarization of classes led to the great
peasant rebellion of 1932 with its tragic consequence of death and repression
(PeÂrez Brignoli, 1995, pp. 232±61). Political instability and the weakness of the
state in the face of foreign interests was another inheritance of the agro-export
model, present particularly in Honduras and Nicaragua. Behind the caricature
of `Banana Republics' hid perhaps a much deeper reality to which the region
was condemned by geopolitics and the reduced size of its countries.
In brief, it may be affirmed that the development of agricultural exports inte-
grated the Central American economies into the world market, incorporating a
modest number of technological changes. But the initial benefits disappeared

fairly soon and were followed by decades of virtual stagnation.
To what extent did the countries benefit from the export-led growth? It seems
that the level achieved by Costa Rica was the best that could be achieved in
Central America. As the linkages to industrial development were almost non-
existent owing to the reduced size of the economies, only the investments in
human capital and transport infrastructure were in place for the long term.
From this it is clear that Costa Rica achieved a significant advantage, already
apparent at the end of the nineteenth century.
Notes
1 On the Central American economies in general, see Cardoso, Ciro and PeÂrez Brignoli
(1977), Torres Rivas (1971) and Munro (1918), which is still indispensable.
2 Coffee was introduced to Costa Rica from the Caribbean where the earliest techniques
of cultivation and processing also had their origin. However, once commercial cultiva-
tion began there was a marked process of technological adaptation and invention
which encompassed the phases of cultivation, harvesting and processing.
3 In Central America the term refers to the owners of coffee-processing plants.
4 One example among many: an English commercial agent who visited El Salvador
around 1910 found annual interest rates fluctuating between 18 and 40 per cent. See
MartõÂn (1985).
5 Guerra Borges (1994, p. 47); for Honduras the author's calculations are based on the
IMF report, p. 54 in Central Bank of Honduras (1957).
References
AlbarracõÂn, P. and PeÂrez Brignoli, H. (1977) EstadõÂsticas del comercio exterior de Costa Rica
(1907±1946), San JoseÂ: Avances de InvestigacioÂn, Universidad de Costa Rica.
Adler, J., Schlesinger, E.R. and Olson, E.C. (1952) Las finanzas puÂblicas y el desarrollo
econoÂmico de Guatemala, Mexico: Fondo de Cultura EconoÂmica.
ANACAFE (1975) BoletõÂn EstadõÂstico, Guatemala.
Belli, P. (1975) `ProlegoÂmenos para una historia econoÂmica de Nicaragua, 1905±1966',
Revista del Pensamiento Centroamericano, vol. 146 pp. 2±30.
Bingham, J.L. (1974) Guatemalan Agriculture during the Administration of President Manuel
Estrada Cabrera, 1898±1920, masters thesis, Tulane University.
Bovedani, B. (no date) El Salvador's Foreign Trade and Development Problems, San Salvador:
Ministerio de EconomõÂa (no date available).
Browning, D. (1971) El Salvador. Landscape and Society, Oxford: Clarendon Press.
Bulmer-Thomas, V. (1987) The Political Economy of Central America since 1920, Cambridge:
ÐÐ (1994) The Economic History of Latin America since Independence, Cambridge: Cam-
bridge University Press.
Bunge, A. (1940) Una Nueva Argentina, Buenos Aires: Editorial Kraft.
Cambranes, J.C. (1985) CafeÂ y campesinos en Guatemala, 1853±1897, Guatemala: Editorial
Universitaria de Guatemala.
Carcanholo, R. (1978) `Sobre la evolucioÂn de las actividades bananeras en Costa Rica',

Estudios Sociales Centroamericanos.
Cardoso, C.F.S. (1975) `Historia del cafeÂ en CentroameÂrica (siglo XIX): Estudio compar-
ativo', Estudios Sociales Centroamericanos.
ÐÐ (1973) `La formacioÂn de la hacienda cafetalera en Costa Rica (siglo XIX)', Estudios
Sociales Centroamericanos.
Cardoso, C.F.S. and PeÂrez Brignoli, H. (1977) CentroameÂrica y la economõÂa occidental (1520±
1930), San JoseÂ: Editorial Universidad de Costa Rica.
Casey, J.J. (1979) LimoÂn 1880±1940. Un estudio de la industria bananera en Costa Rica, San
JoseÂ: Editorial Universidad de Costa Rica.
Central Bank of Costa Rica, Memoria Anual 1951, San JoseÂ: Imprenta Tormo.
Central Bank of Honduras (1961) El cafeÂ en Honduras, Tegucigalpa: Banco Central.
ÐÐ (1957) Historia financiera de Honduras. Informes de las misiones Young (1920±21), Bern-
stein (1943) y Fondo Monetario Internacional (1949), Tegucigalpa: Banco Central.
Collver, O.A. (1965) Birth Rates in Latin America: New Estimates of Historical Trends and
Fluctuations, Berkeley: University of California, Institute of International Studies.
Commodity Year Book (1939) New York.
Costa Rica (1883±1940) Anuarios EstadõÂsticos, various issues.
Cumberland, W.W. (1928) Nicaragua. An Economic and Financial Survey, Washington, DC:
Government Printing Office.
Dosal, P.J. (1993) Doing Business with the Dictators: A Political History of United Fruit in
Guatemala, 1899±1944, Wilmington, Delaware: SR Books.
El Salvador (1911±40) Anuarios EstadõÂsticos, various issues.
Ellis, F. (1983) Las transnacionales del banano en CentroameÂrica, San JoseÂ: Editorial Univer-
sitaria Centioamericana.
Food and Agricultural Organisation (1947) The World's Coffee, Rome: FAO.
Fonseca, P.S. (1924) La moneda salvadoren Ä a, San Salvador: Imprenta Nacional.
Guatemala (1883±1935) Memorias de la SecretarõÂa de Hacienda y CreÂdito PuÂblico, various
issues, Guatemala.
Guerra Borges, A. (1994) Èl desarrollo econoÂmico', in H. PeÂrez Brignoli (ed.), Historia
general de CentroameÂrica. Tomo V. De la posguerra a la crisis (1945±1979), San JoseÂ:
Facultad Latinoamericana de Ciencias Sociales.
Hall, C. (1976) El cafeÂ y el desarrollo histoÂrico-geograÂfico de Costa Rica, San JoseÂ: Editorial
Universidad de Costa Rica.
Hill, R.H. (1933) Fiscal Intervention in Nicaragua, New York: Columbia University, doctoral
thesis.
Historical Statistics of the United States, 1789±1945, (1949) Washington, DC: Government
Printing Office.
Archivo General de Indias, Guatemala 895, Ìnforme del Consulado de Guatemala sobre
las causas que tienen obstruõÂdo el comercio y los medios de removerlas, 1798', in
Anuario de Estudios Centroamericanos, Universidad de Costa Rica, vol. 9, 1983.
International Monetary Fund (1950) International Financial Statistics, vol. 3, no. 6, June,
Washington, DC: IMF.
Jones, C.L. (1971) [1931] Caribbean Backgrounds and Prospects, New York: Kennikat
Press.
ÐÐ (1966) Guatemala. Past and Present, New York: Russell & Russell.
Kepner, C.D. and Soothill, J.H. (1957) El imperio del banano. Las companÂas Äõ bananeras contra
la soberanõÂa de las naciones del Caribe, Buenos Aires: Editorial TriaÂngulo.
Lauria S. and Aldo A. (1992) An Agrarian Republic: Production, Politics, and the Peasantry in El
Salvador, 1740±1920, doctoral thesis, University of Chicago.
Lindo-Fuentes, H. (1993) ÈconomõÂa y sociedad (1810±1870)', in H. PeÂrez Brignoli, (ed.),

Historia General de CentroameÂrica. Tomo III. De la IlustracioÂn al Liberalismo (1750±1870),
Madrid: Ediciones Siruela.
ÐÐ (1990) Weak Foundations: The Economy of El Salvador in the Nineteenth Century, 1821±
1898, Berkeley: University of California Press.
Long, W. R. (1925) Railways of Central America and the West Indies, Washington, DC:
Martin, P. (1985) El Salvador de siglo XX (1st edn London, 1911), San Salvador: UCA Editores.
McCreery, D.J. (1994), Rural Guatemala. 1760±1940, Stanford: Stanford University Press.
Munro, D.G. (1918) The Five Republics of Central America. Their Political and Economic
Development and their Relations with the United States, New York: Oxford University Press.
PeÂrez Brignoli, H. (1995) Ìndians, Communists, and Peasants: The 1932 Rebellion in El
Salvador', in W. Roseberry, L. Gudmundson and M. Samper (eds), Coffee, Society, and
Power in Latin America, Baltimore and London: Johns Hopkins University Press.
ÐÐ (1999) `Transformaciones del espacio centroamericano' in M. Carmagnani, A. H.
ChaÂvez and R. Romano (eds), Para una historia de AmeÂrica Latina, vol. 2 MeÂxico:
Fondo de Cultura EconoÂmica.
PeÂrez Brignoli, H. and Baires MartõÂnez, Y. (1983) `Growth and Crisis in the Central
American Economies, 1950±1980', Journal of Latin American Studies vol. 15.
PeÂrez Brignoli, H. and Samper, M. (eds) (1994) Tierra, cafeÂ y sociedad. Ensayos sobre la
historia agraria centroamericana, San JoseÂ: Facultad Latinoamericana de Ciencias Sociales
(FLACSO).
Posas, M. (1993) `La plantacioÂ n bananera en CentroameÂrica (1870±1929)', in O. Acun Äa
and V. Hugo, (eds), Historia General de CentroameÂrica (Vol. IV), Las RepuÂblicas agroexpor-
tadoras (1870±1945), Madrid: Ediciones Siruela.
Radell, D.R. (1964) Coffee and Transportation in Nicaragua, Berkeley: Department of Geo-
graphy, University of California.
RomaÂn Trigo, A.C. (1995) Las finanzas puÂblicas de Costa Rica. MetodologõÂa y fuentes (1870±
1948), San JoseÂ: Universidad de Costa Rica, Centro de Investigaciones HistoÂricas de
AmeÂrica Central.
Samper, A. (1948) Importancia del cafeÂ en el comercio exterior de Colombia, BogotaÂ.
Samper, M. (1993) `CafeÂ, trabajo y sociedad en CentroameÂrica (1870±1930): una historia
comuÂ n y divergente', in O. Acun Äa and V. Hugo, (eds), Historia General de CentroameÂrica,
(Vol. IV), Las RepuÂblicas agroexportadoras (1870±1945), Madrid: Ediciones Siruela.
Schoonover, T. and Schoonover, E. (1989) `Statistics for an Understanding of Foreign
Intrusions into Central America from the 1820s to 1930', Anuario de Estudios Centroa-
mericanos, vol. 15(1), Universidad de Costa Rica.
Soley Gu È ell, T. (1926) Historia monetaria de Costa Rica, San JoseÂ: Imprenta Nacional.
Statistical Abstract of Latin America, 1980 (vol. 20), 1985 (vol. 24), 1986 (vol. 26), Los
Angeles: UCLA Latin American Center Publication.
Teplitz, B. (1974) The Political and Economic Foundations of Modernization in Nicaragua. The
Administration of JoseÂ Santos Zelaya, 1893±1909, doctoral thesis, Howard University.
Torres Rivas, E. (1971) InterpretacioÂn del desarrollo social centroamericano, San JoseÂ: Editorial
Universitaria Centioamericana (EDUCA).
US Burean of Labour Statistics (1949) Wholesale Price Index, Historical Statistics of the United
State, 1789±1945.
Vargas, O.R. (1990) La revolucioÂn que inicioÂ el progreso. Nicaragua, 1893±1909, Managua:
Comunicaciones Nicaraguenses S.A. (CONSA).
Young, J.P. (1925) Central American Currency and Finance, Princeton: Princeton University
Press.
5
Export-led Growth in Mexico,
c.1900±30
Alan Knight
Initial thoughts: units and periods of analysis
All the chapters in this book concern individual countries;1 this chapter con-
cerns Mexico. Using countries as units of analysis, however, is neither inevit-
able nor unproblematic, especially for the period c. 1900±30. Latin American
nation-states of that period were usually weak states,2 ruling highly imperfect
nations. Though states were tending to get stronger ± hence the emergence of
Wolf and Hansen's order and progress dictatorships (Wolf and Hansen, 1967,
pp. 168±79) ± their political and (a fortiori) economic powers were quite limited;
they could repress, but not represent; their burgeoning military capacity was
not matched by an equivalent capacity to mould their societies or control their
economies.3 This partly reflected the familiar condition of dependency, which
early twentieth-century governments could do little to mitigate.
But, more important, it reflected the fragmentation of political economies.
Porfirian Mexico was ostensibly more stable and integrated than many Latin
American states, yet its politico-economic cohesion ± hence, its validity as an
analytical unit ± were both limited (Reynolds, 1970, pp. 6±7). Porfirian Mexico
lacked a central bank;4 in the far south, Guatemalan pesos circulated; in some
villages markets were based on barter rather than money transactions.5 The
YucataÂn peninsula was cut off from the rest of Mexico, logistically, economic-
ally and psychologically: there was no railway connection, trade was skewed
towards the USA and (to a lesser extent) the Caribbean, and when a thirsty
customer ordered a `foreign' beer in MeÂrida he got a glass of Dos Equis ± from
Orizaba (Gamio, 1982, p. 12). Northern border ports ± still small, but fast
growing ± were already closely integrated with the US economy (hence, during
the Revolution, wild talk of northern secession flourished); northern company
towns ± such as Cananea ± enjoyed a large measure of both economic and
political autonomy (Sariego, 1988). Government spending in 1910 stood at
4.4 per cent of GDP (table 5.1 below); while the Revolution pushed up this
119
120 Export-led Growth in Mexico, c.1900±30
percentage (1925: 5.9 per cent), the real `rise of the active state' came post-
1930.6 Revenue was also a reflex of trade, especially foreign trade, above all
imports, hence fiscal dependency mirrored trade dependency.7 Although rail-
ways had dramatically affected the economy, helping to forge a more integrated
market (see below), this process chiefly favoured foreign trade (Coatsworth,
1984, ch. 5); many parts of Mexico ± towns bypassed by the railway as well as
more remote rural districts ± remained economically isolated, awaiting the
road-and-truck revolution of the 1930s and 1940s.8 Of 246 villages in 10 states
surveyed in 1922±24, only 5 per cent could be reached by rail, 2 per cent by car,
and 1 per cent by water transport; ox-carts served 27 per cent, while the
remainder (64 per cent) depended on horses, burros or foot transport (Tannen-
baum, 1968, pp. 85±7).
Throughout Mexico, but particularly in the centre, the south, and the sierras,
the ancient tlameme could still be seen footslogging along primitive roads, back
bowed, the tumpline biting into a calloused forehead.9 About 20 per cent of the
Mexican population were reckoned to be literate;10 while this emphatically did
not mean that 80 per cent of Mexicans were ignorant `parochials', scarcely
aware of the existence of the patria,11 it did mean that networks of economic
exchange were patchy; that Mexico was still ± to borrow a Durkheimian meta-
phor ± in a phase of `mechanical' economic solidarity, which perhaps embraced
two-thirds of the population.12 Contrary to crude Marxist analyses, it would
seem, some sort of loose national consciousness preceded the formation of an
integrated national market and a dynamic national bourgeoisie. Certainly the
domestic market was thin and fragmented; in some remote communities, we
have noted, barter prevailed; while even in northern Mexican communities ±
major cities apart ± furniture was scant, spoons were rare, and glass windows a
luxury (McKellar, 1994, pp. 51, 66). Merchants lamented the `damned want-
lessness' of the Mexican people, their chronic failure to consume.13 These were
old and familiar complaints,14 which were paralleled by equally old and famil-
iar complaints about Mexico's feckless working class, its chronic ± even con-
genital ± failure to respect the time and work discipline required by modern
capitalism.15 Consumer and labour markets alike, we may conclude, were
highly imperfect; much of the spadework of state-building, infrastructural
investment and/or `primary accumulation' had yet to be done (Marx, 1957,
vol. 2, ch. 24).
We have to be careful, then, when we generalize about `Mexico'. National
statistics are not only inherently unreliable; they also offer a very partial picture
of the country, since they can better capture ± for example ± foreign trade rather
than subsistence agricultural production; and, of course, they can tell us little
about subjective notions such as nationalism, business confidence, class antag-
onism or political legitimacy ± all of which the `new institutional economics'
would wish us to take into account (see Appendix). The notion of simple
Alan Knight 121
enclaves probably has to be qualified, if not rejected altogether,16 but we should

still note the importance of distinct local and regional economies: YucataÂn was
the clearest case; but, by the 1910s, the oil-producing zone inland from Tam-
pico, while it sucked labour from the interior, was something of a political and
economic enclave ± it imported its cement from Britain rather than central
Mexico, and it brought in black West Indian labourers to work the docks
(Haber, 1989, p. 40; Knight, 1986a, vol. 1, p. 217); the north-west had tenuous
links to the centre; and much of rural Mexico ± Oaxaca, Guerrero, highland
Chiapas, the Sierra Madre Occidental ± consisted of a patchwork of local com-
munities, markets and political fiefs. The Revolution made this starkly clear
and, as a result, provoked an aggressive ± but only partially successful ± project
of state-building during the 1920s.
We turn to the Revolution, and the related questions of periodization and
national specificity. How can `Mexico' ± itself a slightly arbitrary category of
economic analysis ± be fitted within the bigger Latin American picture? All
countries are peculiar; some are more peculiar than others (Knight, 1992a, p.
99ff ). Mexico shared with Brazil and Argentina a certain critical mass, which,
for example, made import substitution industrialization more feasible than in,
say, Guatemala or Paraguay. Unlike Brazil and, a fortiori Argentina, Mexico was
not a nation of immigrants (whether enslaved blacks or free whites); and
demand for labour could be met domestically (although, as already mentioned,
this involved problems: employers regularly complained about feckless, ineffi-
cient labour, which, for them, justified the use of coercion to secure and retain a
labour force). Why did the old dream of European immigration ± the idealized
Argentine solution ± not materialize? On the demand side, Mexico's main
exports remained minerals, and mining was not labour-intensive in the manner
of sugar or coffee; in this respect, Mexico resembled the Andean mining eco-
nomies of Peru and Bolivia.17
Coincidentally, Mexico also resembled Peru and Bolivia ± and northern
Central America ± by virtue of its large indigenous and mestizo population, a
legacy of the old Precolumbian empires;18 this `solved' the labour supply
problem and, by ensuring low wage rates, deterred immigration; but it also
meant that Mexico, especially southern Mexico, would ± like Guatemala,
Peru or Bolivia ± have to `solve' a related problem ± that of ethnic fragmenta-
tion, exploitation and racism which carried important consequences for labour
markets and state-formation. Labour coercion seemed more acceptable when
the victims were brown, illiterate, non-Spanish-speakers; and labour coercion,
of course, favoured the creation of authoritarian states.19 Here, structural pro-
pensities conspired with conjunctural events: like Peru and Bolivia (and Vene-
zuela), Mexico had a stormy passage to independence, failed to achieve political
stability and, throughout the nineteenth century and well into the twentieth,
politics was vulnerable to authoritarian and praetorian tendencies.20 These
tendencies clearly contravened Douglass North's institutional prescriptions for

growth (North, 1990). Finally, Mexico ± `so far from God and so near the United
States' ± enjoyed one geopolitical peculiarity which set it apart from the rest of
Latin America: its long, porous border with the USA, which offered certain
economic benefits but also a constant threat of political pressure and economic
penetration ± as well as `leakage' ± even armed intervention.21
To these assorted peculiarities must be added an obvious but tricky historical
peculiarity: starting in 1910, Mexico experienced a massive social revolution;
our analysis closes ± in theory, at least ± in 1930, at a time when the recon-
structive and state-(re)building tasks of the revolutionary regime were still
underway and far from complete. All Latin America, it is true, suffered a
major economic upheaval ± sometimes coupled with political bouleversement ±
in the second decade of this century as result of the First World War; in some
respects, Mexico's Revolution, while eclipsing the impact of the war, produced
analogous consequences: economic dislocation, inflation, export demand (that
is, for oil) and, perhaps, a degree of import substitution industrialization. But
Mexico's upheaval went far beyond this (and was seen to do so): it was deeper,
more `structural' and, essentially, domestically-induced.22 This raises a couple
of pertinent questions which this chapter must try to address: to what extent
did Mexico's pre-1910 pattern of desarrollo hacia afuera `cause' the Revolution?
And how did the Revolution change the Mexican economy, in both the short
and the long term?
Some answers to the first question will emerge en route; the point to stress at
the outset is that export-led growth, though crucial, cannot be understood in
isolation and must be analysed within the broader context of the Porfirian
political economy. The analysis which follows, therefore, goes beyond the
specific question of external markets, and seeks to relate them to domestic
considerations, both economic and political.23 The second question requires a
more immediate response, since it touches on the key question of periodiza-
tion ± that is, the question of how we slice up the seamless garment of history,
using, if you like, the familiar pinking shears of `continuity' and `rupture'. For
most of Latin America the year 1910 is economically irrelevant; even the
rupture of 1914±18 was followed by strenuous, and partly successful, efforts
to restore the prewar system, both domestically and internationally. With
hindsight we can see that that this was a very imperfect restoration; Humpty-
Dumpty could not be put back together again. But the war did not destroy faith
in desarrollo hacia afuera, the Gold Standard, and the benefits of foreign trade
and investment (in many respects, its chief external consequence was to
enhance US influence at the expense of British); it did not bring sudden,
structural changes to domestic economies ± at best, it accelerated some incre-
mental changes and created additional political strains. None of this was re-
motely revolutionary.
Alan Knight 123
What of Mexico's revolution? It did not, I shall suggest, subvert Mexico's

external economic relations. It boosted US influence, while bringing to power
new elites who admired as much as they feared the coloso del norte (Knight,
1987, pp. 56, 77; CoÂrdova, 1988, pp. 248±61, 379±401). The stage was therefore
set for a series of skirmishes ± especially as regards oil ± which implied a
qualified shift in the external economic relationship. But that shift can easily
be exaggerated. The Revolution did not revolutionize external economic rela-
tions: oil was something of a special case; mining enjoyed an Indian summer
during the 1920s; the old logic of desarrollo hacia afuera was neither seriously
questioned, nor practically subverted (Knight, 1987, pp. 24±5; SolõÂs, 1971, p.
108; Bernstein, 1964).24 The great depression ± whose premonitory symptoms
antedated 1929±30 ± was much more significant. Now, export-led growth col-
lapsed and was seen to collapse. New initiatives were attempted by a state
which, little by little, had begun to acquire powers of economic intervention:
a Central Bank, ready to pump-prime the depressed economy; and a series of
state agencies ± development banks, a labour ministry, an agrarian reform
bureaucracy ± which could, for better or worse, venture dirigiste policies.
This more robust response to the 1930 depression was, in large measure,
possible because of the domestic consequences of the Revolution. For there
can be no doubt that the Revolution's domestic consequences were far greater
and more `structural' than its external impact; and this, of course, set Mexico
apart from every other `non-revolutionary' country in the continent. The
Revolution did not ± pace many analyses of Tocquevillean temper ± swiftly
establish a Leviathan state; but it did substantially transform both the character
of the state and the state's relationship to a society which itself was substantially
transformed ± even `revolutionized' ± by the upheaval of 1910±20, though not
necessarily in ways that were planned, scripted or legislated (Knight, 1986a, vol.
2, pp. 517±27). The Revolution destroyed or weakened the old landed class,
especially the Porfirian oligarchs; it stimulated ± better, it embodied ± extensive
popular mobilization, rural and urban; it set in train an agrarian reform which,
if limited in scope prior to 1930, represented a significant shift in class relations
and subjective perceptions; and it gave the infant labour movement an unusual
degree of leverage, which contrasted with the weakness not only of Mexican
labour pre-1910 but also of labour elsewhere in Latin America. The Revolution
provoked hyperinflation, with all that that entailed in terms of social mobility
and the capricious redistribution of resources (Kemmerer, 1940); it encouraged
unusual spatial mobility as people ± armies, camp-followers, refugees, migrants
± trekked across country or, in increasing numbers, to the USA;25 it eroded old
barriers of deference; it very likely fostered greater entrepreneurialism, job
mobility and a kind of collective work ethic; and, in place of the Pax Porfiriana
± a `Roman peace', it is true ± it left a legacy of both political and personal
violence.
Thus, apart from its well-known social reforms ± which were often slow in
coming ± the Revolution brought significant socioeconomic change, albeit
change that is hard to calibrate, given not only the lack of reliable statistics
but also the ìmpressionistic' nature of many of these changes. We cannot
measure, we can only suggest, an enhanced entrepreneurialism and work
ethic; a decline of deference; class polarization; greater social and spatial mobil-
ity; and perhaps greater regime legitimacy.26 What can be said with some
confidence is that the Revolution was a genuine revolution which had import-
ant economic consequences; but that these consequences affected the domestic
political economy much more than they affected Mexico's external economic
relations ± even though those relations had helped provoke the Revolution in
the first place. `Dependency' (to use a loose catch-all) endured, perhaps it was
even enhanced; but the threads of dependency now wove their way through
both a transformed state (not, as yet, a `Leviathan') and a transformed civil
society. On that level, 1910 marked a decisive break which had no parallel in
the rest of Latin America. Hence the peculiarity of Mexican periodization which
must embody two distinct ruptures: the rupture in the domestic political econ-
omy in 1910, and the rupture in external relations in 1930. The former is
unique, the latter bears comparison with the rest of Latin America. But the
fact of the first rupture helped determine the Mexican reaction to the second.
Thus, as I have argued elsewhere, Cardenismo was the peculiar result of the
intersection of a distinctive domestic (political economy) trajectory and a
shared external (world depression) conjuncture (Knight, 1991a, ch. 5, especially
pp. 245±6).
The Porfirian economy
If, as I have just argued, the domestic political economy of the Porfiriato
entered into crisis in 1910, and its associated external economic relations in
1930, thus giving us two clear terminal points, when did the process of export-
led growth begin? It is obvious that 1900 is, for these purposes, a meaningless
date, another case of irrational hectohistory (Fischer, 1970, p. 145). To under-
stand Mexico's phase of desarrollo hacia afuera, we have to go back into the
nineteenth century, as we do in most other Latin American cases; indeed, one
of the interesting puzzles concerns the timing of this phase, which clearly
varied from country to country.27 In Brazil it was an incremental story, lacking
sudden surges. In Peru it was a more episodic, stop±go saga. In Chile it came
relatively early, in Venezuela relatively late. The question of timing is clearly
linked to impact: incremental growth was less destabilizing than sudden booms
and busts; the latter depended, to a large degree, on the nature of a country's ±
or a region's ± engagement in the new global division of labour. DõÂaz Alejandro's
`commodity lottery' was a key determinant of economic trends long before the
Alan Knight 125
Great Depression. For Mexico, engagement in the global division of labour

came late, quite dramatically, and hence tended to destabilize; however, the
relative diversity of Mexican exports afforded some countervailing security.
Meanwhile, subjectively, destabilization was not immediately evident and,
indeed, was masked by the deceptive stability of the Porfirian belle eÂpoque.
Like other revolutions, that of Mexico sneaked up on an unsuspecting people.
Mexico's engagement in the global division of labour came late because, for
some two generations after Independence (that is, 1821±c.1876), the country
suffered political instability and economic stagnation, the two related in
obvious ways. A chronically bankrupt government could not govern; instability
deterred investment; and lack of investment and exports retarded growth and
political stabilization. This period stood in marked contrast to the last phase of
colonial rule ± c.1760±1810 ± when the economy had burgeoned, stimulated by
the last great mining boom; when population had expanded, depressing real
wages and living standards; and when the Bourbon government had milked
Mexico, above all to pay for its recurrent foreign wars.28 Independence, blood-
ily achieved, sundered the bonds with Spain and left Mexico economically
depressed; however, this transition from a phase of economic `compression'
to `decompression' may have resulted in rising living standards; as taxes fell,
subsistence agriculture flourished and peasant communities acquired a measure
of political leverage.29 Although external and conjunctural factors ± the French
revolutionary and Napoleonic Wars ± provoked this turnaround, there is evid-
ence, too, that late colonial Mexico, for all its success in narrow mercantilist
terms, had reached the limits of growth which the colonial system allowed and,
indeed, was teetering on the brink of Malthusian crises ± evidenced by the
dearths of 1785 and 1809 (Florescano, 1986, chs 3, 4). Thus, the subsequent
phase of `decompression' ± perhaps even ìnvolution' ± may have been on the
cards: Mexico, in other words, was locked into a familiar pre-industrial cycle
whereby bursts of population growth ± unalleviated by gains in productivity ±
eventually led to falling living standards, dearth, disease and demographic
retreat (Leroy Ladurie, 1969).
Independence brought significant political changes, and, as I have said, was
associated with a serious `conjunctural' downturn: mining output slumped,
foreign trade was dislocated, Spanish capital fled, and a fragile republican
regime replaced the durable rule of the colony. If the external relations of
Mexico ± characterized by formal imperial `dependency' ± were subverted,
domestic economic structures remained much the same (thus, we might say,
the 1810 revolution had the opposite effect of that of 1910). In consequence,
economic historians of Mexico and Latin America more generally now tend to
see the period c. 1760±1850 as possessing a certain historical unity (Taylor,
1985, pp. 122±3). Infrastructure remained primitive: Mexico lacked navigable
rivers and canals and, although the first Mexico City±Veracruz railway
concession was given in 1837, the line was not completed until 1873. Indeed,
in 1860 ± when the USA had 30 000 miles of track ± Mexico had 15, three of
which served to carry pilgrims from the capital to the shrine of La Guadalupe.
No national network existed until the 1880s: Mexico lagged behind not only
the USA, but also Cuba, Brazil and Argentina.
If domestic capital was scarce and timid, foreign investment was no less leery.
The foreign loans of the 1820s ran into the sand; in the absence of export
markets, foreign direct investment was scant; and its pioneers ± such as the
British Real del Monte Mining Company ± found the problems of transport and
labour recruitment daunting (Randall, 1972, especially ch. 7). The domestic
market was shallow, hence attempts at industrialization failed: the govern-
ment's bold attempt to foster textile manufacturing in the 1840s failed (Potash,
1959). Mexican entrepreneurs, following the trail blazed by Mexican landlords,
strove to make a profit on a low turnover, capturing local markets, and relying
heavily on political favours: tariffs, credit, government contracts (Haber, 1989,
pp. 5, 62). The economy was therefore highly politicized: governments, though
shifting and unstable, could reward their friends and punish their enemies; and
recurrent civil wars gave ample opportunity for punishment and bred a climate
of business desconfianza.30 This in turn encouraged a suspicious and introverted
business ethic, an entrepreneurial equivalent of Banfield's àmoral familism' ±
or, to take an example closer to home, Foster's `limited good' (Banfield, 1958;
Foster, 1965, pp. 293±315). Business was a zero-sum game, played with fixed
stakes. Technological innovation and expansion were risky, even counterpro-
ductive.
In such a climate, families stuck together; high prices and low turnover were
the norm; and the kind of trust-cum-confidence which favours risk-taking and
joint-stock ventures was lacking. These features survived well into the twenti-
eth century.31 Meantime, although it is impossible to estimate welfare for this
period, impressionistic evidence suggests that popular living standards, though
low, were fairly stable.32 Hence, amid recurrent civil wars, popular mobilization
seems to have been directed against political and clerical abuses, not economic
exploitation (Coatsworth, 1988). The clearest exception was YucataÂn, where
the massive rebellion known as the Caste War derived, in part, from the region's
precipitate entry into world markets and the agrarian commercialization which
this provoked (Reed, 1964; Rugeley, 1996).
In a sense, YucataÂn pointed the way forward. During the last quarter of the
nineteenth century Mexico achieved political stability and experienced rapid
export-led growth. The old vicious circle of instability and desconfianza gave
way to a virtuous circle: trade and investment ! government solvency !
government stability ! confidence ! trade and investment. Foreign and
domestic factors intertwined in ways that are difficult to disentangle analytic-
ally. The onset of the Porfiriato (1876) coincided roughly with the Second
Alan Knight 127
Industrial Revolution and the new Kondratieff cycle associated with it (Car-
doso, 1980, pp. 259±67; on Kondratieff (and other) cycles, see Maddison, 1991,
pp. 89±123.). Demand for new industrial raw materials (rubber, fibres, copper,
zinc, lead) combined with increased consumer demand in Europe and the USA
± for coffee, sugar, tropical fruits and gum (chicle). Surplus capital sought foreign
outlets, with Latin America a favoured ± ìnformally imperial' ± outlet.33
Meanwhile, the newly-established government of Porfirio DõÂaz (1876±1911)
capitalized on Mexico's war-weariness to create a stable, authoritarian regime
which made light of the liberal provisions of the 1857 Constitution and, like
contemporary regimes throughout Latin America, adhered to a positivistic
belief in the primacy of economics. Without economic development, positiv-
ists reasoned, political liberalism was a sham and a recipe for instability; devel-
opment in turn required a strong, centralized government which would favour
foreign trade and investment. Although Porfirian policy-makers ± notably the
celebrated CientõÂficos ± revered the market, they argued, realistically enough,
that in a society where the market was weak the state had to create the appro-
priate conditions for market-led growth.34 Hence, in pursuit of a distant eco-
nomic liberalism ± and a still more distant political liberalism ± the Porfirian
state resorted to a variety of direct interventions designed to promote growth. It
introduced a new Commercial Code in 1884, and a Mining Code in 1892
(Sariego et al., 1988, pp. 28±32); promoted a banking system; suppressed inter-
nal customs (the old colonial alcabalas, 1896) and hiked external tariffs;
balanced the government budget and consolidated the national debt; subsid-
ized railway building and, in the 1900s, consolidated the bulk of the railways in
the Ferrocarriles Nacionales de MeÂxico (CalderoÂn, 1965, pp. 613±22); modernized
ports and promoted telegraph communications; legislated to encourage the
colonization of virgin land and the break-up of corporate village landholdings;
impeded labour organization and repressed social protest; and wooed foreign
business and carefully cultivated Mexico's image in the external world.35 Like
other `late-industrializers', therefore, Mexico pursued a strategy of development
which took the market to be central, but the role of the state to be vitally
supportive (cf. Gerschenkron, 1962).
By its own lights, the Porfirian regime was strikingly successful. It preserved
political stability for 35 years during which time GDP grew some 2.6 per cent;
or, given a population growth rate of 1.4 per cent, at 1.2 per cent per capita per
annum (Reynolds, 1970, pp. 16, 23). Foreign investment rose from minimal
levels to 3.4 billion pesos in 1911 (Rosenzweig, 1965c, vol. 2, pp. 987±90, 1153±
7). Exports, growing at 6.1 per cent per annum, were the motor of growth; and,
within Mexico's diverse export basket, precious metals declined in relative
importance as non-precious metals (copper, lead, zinc) and agricultural raw
materials (coffee, henequen, rubber, cotton) gained, with exporters benefiting
from the long-term depreciation of the Mexican silver peso.36 The new railway
system ± which grew from 400 to 12 500 miles during the period ± made this
export performance possible (freight rates actively favoured exports); it also
tended to accentuate US dominance of Mexico's markets (Coatsworth, 1984,
chs 4, 5). Railways made bulk production and transport feasible: the cost of
shipping a ton of textiles the 130 miles from QuereÂtaro to Mexico City fell from
US$61 to US$3.37 Such savings made possible the dramatic growth of mining
and metal-refining in northern Mexico; and of aggressive cash-crop production
throughout the country: henequen in YucataÂn, sugar in Morelos, coffee in
Veracruz, cotton in the Laguna, cattle and vanilla in the Huasteca, wheat,
chickpeas and vegetables in the north-western river valleys (CossõÂo Silva,
1965; Knight, 1986a, vol. 1, p. 80ff.). In some cases (minerals, henequen)
production was almost entirely exported; in others (coffee, cotton, sugar), the
domestic market was significant ± hence it would be wrong to dichotomize the
two. In addition, Mexican manufacturing ± of textiles, steel, cement, glass,
dynamite, newsprint, soap ± grew rapidly during the 1890s, which decade is
now rightly seen as the first phase of import substitution industrialization
(ISI).38 ISI and desarrollo hacia afuera were not, therefore, incompatible.39
There can be little doubt, however, that exports were the motor and that the
domestic market lacked the dynamism of the export sector: while agricultural
exports boomed, domestic production lagged ± though by how much it is hard
to say.40 The most dynamic sector of domestic cash-crop production was that of
pulque, which could now be rail-freighted from the huge hacienda pulqueras of
the plains of Apam, into the maw of a thirsty capital whose burgeoning popula-
tion had abundant reason to drown its sorrows.41
For, with the renewed `compression' of the Porfiriato, living standards tended
to fall and social tensions to rise.42 The social impact of Porfirian growth can be
plotted both spatially and chronologically. First, the impact varied by regions.43
In general, northern Mexico ± economically dynamic, underpopulated, mes-
tizo, and more closely tied to the US economy ± benefited from Porfirian
growth. The mines of Sonora and Chihuahua flourished, attracting US capital;
and at the same time ± stimulated by the McKinley tariff of 1890 ± a Mexican
smelting industry grew up dominated by the ASARCO (Guggenheim) interests
(Nava Oteo, 1965, p. 241; Sariego et al., 1988, pp. 36±7). Commercial agricul-
ture, benefiting from the new railways though restricted by water shortage,
sprang up in the Laguna, the Sonora river valleys and the lower RõÂo Grande
valley. Land prices rose (CossõÂo Silva, 1965, pp. 73±4). Monterrey became the
preeminent industrial city of the north, producing steel, glass and beer, to the
advantage of its dynamic entrepreneurial elite (Saragoza, 1988, chs 1±3). Long
familiar with free wage labour, the north experienced no revolution in its
relations of production: the northern bourgeoisie could rely on cash incentives
to attract migrant workers from the populous centre/south (French, 1996, pp.
30±1, 43, 120); hence coercion was at a discount, save in those pockets of the
Alan Knight 129
north ± such as the Yaqui River Valley ± where recalcitrant peasant farmers had
to be dispossessed.44 The north thus conformed to the classic Marxist syndrome
whereby market growth generated free wage labour and an entrepreneurial
bourgeoisie.
At the other end of the country ± the trans-Isthmian south-east ± a very
different outcome ensued. Again, population was scant (and dispossessed pro-
letarians were even scantier); landlords who benefited from buoyant markets ±
for coffee, timber, rubber and henequen ± combined coercion and cash incent-
ives in their search for labour.45 The coffee finqueros of Soconusco on the
Chiapas Pacific coast used the enganche system to draw highland Indians
down from the sierras; with time, a regular flow of labour leading to a perman-
ent labour force developed; and the relations of production in coastal Chiapas
thus roughly resembled those of coastal Peru (GarcõÂa de LeoÂn, 1985, vol.1 pp.
191±8; cf. Albert, 1985). Elsewhere in the south, outright coercion was more
apparent: in the harsh monterõÂas of highland Chiapas, in the notorious coffee
fincas of Oaxaca's Valle Nacional, and on YucataÂn's henequen estates where the
`Divine Caste' ± the Yucateco oligarchy ± blended traditional paternalism with a
more coercive commercial exploitation (Knight, 1986a, vol.1, pp. 97±9; Wells,
1985). Though large subsistence Indian populations remained ± in the Yucatan
lowlands and the highlands of Chiapas and Oaxaca ± it did not prove easy to
àrticulate' their subsistence mode of production with commercial farming: not
because Indians were indifferent to market opportunities, but rather because
the available market opportunities ± poor wages for bad jobs ± were unattract-
ive. Thus, in south-eastern Mexico, as in Guatemala, the plantocracy ruled;
labour was coerced and coercion was justified in racist terms: el indio no oye sino
por las nalgas.46
Central Mexico, the most densely populated part of the country, also pre-
sented the most complex economic picture. The rural sector, long dominated
by the hacienda (and rancho),47 eagerly responded to new market opportunities:
the pulque of Hidalgo and the sugar of Morelos were classic cases. The progress-
ive expropriation of peasant village lands ± and the proletarianization of arti-
sans ± helped ensure an adequate supply of land, water and labour; hence ± at
the price of considerable social tension ± landlords increasingly relied on the
swelling free labour market to meet their needs.48 Indeed, some landlords
sought to rationalize their ìnternal' labour relations by replacing costly `pater-
nalist' perks with a straight cash wage ± a rational option in times of inflating
food prices and growing labour surplus (Bellingeri, 1978, pp. 121±36). In
doing so, they found ± like their Peruvian counterparts ± that rationalizing
`traditional' labour patterns could provoke peon protest (cf. MartõÂnez Alier,
1977, pp. 82±90). `Feudal' relations acted as a brake on commercialization and
a focus for subaltern resistance. Manufacturers in central Mexico ± the country's
industrial heartland ± also benefited, individually and as employers, from
falling real wages;49 but, collectively and as sellers, they found their products ±
typically, cheap cotton textiles ± hard to sell in Mexico's shallow, constricted,
domestic market. In general, indeed, firms which catered to the domestic
market ± the trams, light and power companies, and some branch railways ±
experienced a less profitable Porfiriato than their exporting counterparts
(Haber, 1989, ch. 7).
Chronologically, too, there were significant differences. During the first
twenty years of the Porfiriato demand for labour rose, encouraged by the rail-
way boom. After the mid-1890s, however, demand slackened and, as popula-
tion growth conspired with both urban and rural proletarianization, the labour
market came to favour buyers (Reynolds, 1970, p. 25; Simpson, 1937, p. 712).
Given both the weakness of the early sindicatos and the hostile stance of the
state ± evident in the celebrated cases of Cananea and RõÂo Blanco ± the counter-
vailing force of unionization could do little to arrest the erosion of real wages.50
Food prices and food imports also rose during the 1900s, squeezing consump-
tion across the board.51 Poor Mexicans probably ate less, appear to have drunk
and smoked less, and certainly bought less.52 They also procreated less or, more
likely, died earlier: population growth fell from 1.6 per cent per annum in 1877±
1900 to 1.1 per cent in 1900±10 (Reynolds, 1970, pp. 22±3). Demand for cheap
textiles ± perhaps the best barometer of popular living standards53 ± declined,
leading to a crisis in Mexico's chief manufacturing sector (Rosenzweig, 1965a,
pp. 330, 338; Knight, 1986a, vol. 1, pp. 134±5). Finally, in 1905 the government
embarked on an ambitious currency reform designed to place Mexico on the
gold standard and thus avoid the vicissitudes of a silver-based currency (BaÂtiz
VaÂzquez y Canudas Sandoval, 1980, pp. 421±31). The desired monetary stability
proved elusive; and deteriorating terms of trade after 1905, coupled with the
coincidental US depression (1907) and poor harvests in 1909±10,54 aggravated
social tensions at a time when ± largely independently ± a vigorous political
opposition was mounting an unprecedented challenge to the DõÂaz dictator-
ship.55
The Revolution which ensued, starting in 1910, obeyed economic motives ±
among others. It was also intimately associated with the preceding phase of
desarrollo hacia afuera. That is not to say that it was inevitable, òverdeter-
mined', or proof of Mexican exceptionality. Most of the socioeconomic ingre-
dients were to be found elsewhere in Latin America: an authoritarian, racist and
positivistic regime bent on export-led development; a dynamic commercial
agriculture linked to peasant dispossession and a degree of labour coercion; a
growing manufacturing sector ± especially in textiles ± catering to a shallow
domestic market;56 and a falling death rate and hence a growing population
which, coupled with urban and rural proletarianization, eased the labour mar-
ket to the advantage of employers who usually had the ear of the state. Mexico
differed only in degree: its growth came late and was unusually rapid, spurred
Alan Knight 131
by the close US connection; aggressive state-building, coupled with a `success-

ful' project of desarrollo hacia afuera, benefited a minority but increasingly
offended the majority; and this majority ± mostly peasants ± possessed valuable
collective resources which, for example, Brazil's peasants, subject to the `pater-
nalist' control of the coronel often lacked: village solidarity, a degree of local
autonomy, and a historical consciousness of previous popular struggles.57
Meanwhile, the growing middle class like their counterparts in the southern
cone sought responsible, constitutional rule (not, of course, social revolution);
and the peasantry, threatened by the loss of land and political autonomy,58
looked to arrest Porfirian state-building in the name of a nostalgic popular
liberalism. DõÂaz's failing powers and faulty judgement encouraged an anom-
alous alliance between these disaffected groups: Mexican analogues of, on the
one hand, Argentina's middle-class Radicals and, on the other, the peasant
rebels of the Andean highlands who came together in unique, ephemeral
alliance, thus making the Mexican Revolution.
The Revolution and after
As already noted, the Revolution marks a major caesura in Mexican history ±

including economic history. Like the First World War, which coincidentally
affected the rest of ± non-revolutionary ± Latin America, it had both conjunc-
tural and structural effects; in the case of the Revolution, however, the latter
were more significant and durable; hence the label `Revolution' seems to me, at
least, entirely appropriate.59
Initially, the civil war had a limited effect on the economy: not because the
civil war was superficial or `narrowly political', but because the economy could
± like other belligerent economies ± soak up a surprising amount of punishment
before sinking to the canvas.60 The Madero government (1911±13) therefore
presided over a fairly buoyant economy; the chief departure from Porfirian
precedent was the growing fragility of government finances, brought on by
increased military spending which now began a rapid upward spiral.61 Certain
regions were adversely effected: Morelos by Zapatismo, Chihuahua by Oroz-
quismo; but, at the national level, this was offset by the continued stability
and profitability of the plantocratic south and the initial take-off of the oil
industry ± on which Madero slapped the first of many additional taxes (L.
Meyer, 1977, pp. 16, 31). During and after 1913, however, the continuation of
armed conflict ± now on an extensive, conventional scale ± took a severe toll.
Communications were severed, exports (notably of minerals) declined, foreign
and domestic confidence faltered, and, with the profligate printing of paper
currency, the peso went into a tailspin. In 1914, Mexico defaulted on its
national debt; it would not secure foreign loans for another generation.62
Political instability peaked during 1914±15, but the economic decline
continued, reaching its nadir in 1916±17 (Knight, 1986a, vol. 2, p. 406ff.). Now
not only confidence and exports collapsed; the very human fabric of the
country suffered as hyperinflation ruined firms and families, as entire regions
retreated into a barter economy (or dealt in US dollars) (tables 5.1, 5.2), as
hunger, sickness and destitution stalked the land. Aggregate national statistics,
it is true, paint a rosier picture; but that picture is heavily coloured by the
petroleum boom which, impelled by the First World War and facilitated by
the enclave character of the industry, ensured a crucial flow of foreign exchange
and tax revenue well into the 1920s.63
While oil financially bolstered the infant revolutionary government, a grow-
ing war-weariness undermined armed resistance, weakened the radical
elements of the labour movement, and, perhaps, engendered a collective
ethic of `reconstruction' ± a phenomenon witnessed in other traumatized
societies (Knight, 1986a, vol. 2, pp. 524±6). As the regime of the Sonorans
(1920±34) embarked on economic reconstruction and renewed state-building,
they could therefore count on broadly favourable domestic conditions, coupled
with grudging US support.64 The `conjunctural' costs of the Revolution were
fairly quickly covered: oil production peaked in 1921, the same year in which
textile production reached its prerevolutionary level (Meyer, 1977, p. 9; Haber,
1989, p. 125). Electricity consumption ± which had been fast rising in the late
Porfiriato ± recovered as early as 1918; beer consumption by 1920 (Reynolds,
1970, p. 28; Haber, 1989, pp. 126, 138). By 1925, GDP was 44 per cent greater
than in 1910 ± 41 per cent in per capita terms (Reynolds, 1970, p. 16). Even
population made a relatively rapid recovery (table 5.1) helped by incremental
improvements in public health.65 In short, as Reynolds observes, `the opportu-
nity cost of the Mexican Revolution was much less than might have been
supposed' (Reynolds, 1970, p. 326).
In this sense, the Revolution produced few dramatic economic consequences.
At best, prerevolutionary trends were somewhat accelerated, or retarded. Popu-
lation growth briefly halted; but rural±urban migration ± and, even more,
Mexico-US migration ± accelerated.66 Manufacturing grew in importance and
± for reasons which will be mentioned shortly ± domestic consumption appears
to have increased, thus alleviating the under-consumptionist crisis of the later
Porfiriato.67 Petroleum production declined following the providential boom of
1910±21; but mining, benefiting from US investment, technological change
and concentration of ownership, enjoyed a prosperous Indian summer.68 In the
agricultural sector, subsistence farming appears to have stagnated,69 but com-
mercial production was buoyant, particularly in respect of certain crops ±
cotton, henequen, sugar, coffee, bananas, chickpeas ± and certain regions
(notably the north-west: the homeland of the Sonoran dynasty) (Reynolds,
1970, pp. 103±5). In the short term, therefore, the Revolution did not reverse
the subsistence/commercial dualism evident in Porfirian agriculture.
Alan Knight 133
Table 5.1 Mexico, GDP and federal government spending
GDP Federal government

(millions of pesos of 1950) spending as % of GDP
1900 8 250 4.9

1910 11 650 4.4
1921 11 273 6.1
1925 14 816 6.8
1930 15 540 6.4
Source: SolõÂs (1971), pp. 90±1.
Table 5.2 Mexico, federal government spending
Millions of current pesos Millions of pesos of 1950 1950 pesos per capita
1900±01 59.8 407.0 29.5

1910±11 101.2 516.5 33.0
1921 226.4 683.8 47.7
1925 302.2 1 007.2 66.3
1930 279.1 996.9 60.1
Source: Wilkie (1970), pp. 22, 2.
Table 5.3 Population
Population (millions)
1900±01 13.8
1910±11 15.2
1921 14.3
1925 15.2
1930 16.6
Source: Wilkie (1970), p. 24.
No more did the Revolution reverse `dependency'. The structure of Mexican

foreign trade remained much the same; the combination of world war and
Revolution boosted US trade and investment at the expense of European; and
the Sonorans ± despite their long squabble with the Anglo-American oil com-
panies ± were keen to attract and retain US investment (of the right sort) (Calles,
1927, pp. 39±41; CoÂrdova, 1988, pp. 381±401). They were also keen to restore
the country's credit, which involved reaching an agreement with foreign
bankers and bondholders. Twice, during the 1920s, agreements were reached,
but economic downturns prevented a resumption of debt repayments. In the
process the hawkish oil companies were marginalized; and Thomas Lamont,
representing Mexico's creditors, demonstrated that the Sonoran leaders were
men with whom US capital could profitably do business (Smith, 1972; ZebaduÂa,
1994, chs 6, 8, 9). Further proof came with the flow of new US manufacturing
investment to Mexico: Ford, Dupont, Colgate-Palmolive.70 Pace some versions,
this did not represent a betrayal of the Revolution. The Sonorans took a `busi-
nesslike' attitude to government, were themselves successful entrepreneurs,
and had never indulged in militant anti-Americanism. Nor had the Revolution
in general been premised on xenophobic anti-Americanism. Porfirian desarrollo
hacia afuera, involving hefty US investment and closer US±Mexican trade rela-
tions, had made a mighty contribution to social tensions and, eventually, to
revolutionary upheaval; but the causal links were complex and indirect, hence
foreign trade and investment were rarely perceived as inherent evils. Revolu-
tionaries targeted proximate victims ± landlords, old regime officials, the
military, the Church ± and, with some exceptions, allowed the major Anglo-
American companies to survive, even to prosper. These companies were not
much liked, but neither were they hated; and, of course, they paid taxes and
brought in foreign exchange. At worst ± in the case of the oil industry ± the
1920s witnessed sporadic skirmishing between companies and government as
the latter sought to increase tax revenue and curtail the autonomy and arrog-
ance of over-mighty corporate subjects (Meyer, 1977, chs 4±7). But there was no
dramatic deÂnouement; that awaited the depression, the rise of Cardenismo, a
new labour militancy, and a new official commitment to desarrollo hacia adentro
(Knight, 1991a).
Revolutionary change was more apparent in the political realm ± a realm
which, in contrast to the economic, was amenable to brusque and purposive
transformations.71 Here, certain formal changes are well-known: the new 1917
Constitution provided for agrarian and labour reform, affirmed national own-
ership of subsoil deposits ± thus `nationalizing' minerals and hydrocarbons ±
and envisaged a more interventionist, social-engineering state which would
arbitrate class conflict, Mexicanize material resources and modernize Mexican
culture (CoÂrdova, 1988, ch. 5). These, however, were distant often rhetorical
goals which required enabling legislation if they were to be implemented. They
also depended on the shifting balance of political forces, both Mexican and
foreign. As regards land reform, progress was slow. Prior to 1933, 4090 ejidal
communities comprising 754 600 ejidatarios had been established; 22 per cent
of all rural communities had benefited from the programme; and 4 per cent of
the area of the Republic, 10 per cent of actual and potential crop land, were to
be found in ejidos (Simpson, 1937, chs 10±11). Sweeping reform therefore
awaited the CaÂrdenas presidency of 1934±40 (Knight, 1991a, pp. 256±64).
However, three important caveats are in order. First, the Revolution severely
weakened the labour-repressive ± even quasi-servile ± systems which had flour-
ished in the Porfirian south. Coercive debt-peonage was ended, or mutated into
free wage labour (Vos, 1988, pp. 228±9; Joseph, 1988, pp. 103±5; Gruening,
1928, pp. 139±40). Second, the agrarian reform ± however gradual and grudging
Alan Knight 135
± represented a marked change from the Porfiriato; the peasantry, far from
being proletarianized, were now being reconstituted under the aegis of the
state. This was, of course, unique in Latin America in this period. Third, the
bare numbers tell only part of the story. Land distribution varied by region:
some states, such as Morelos, were disproportionately affected.72 Even land-
lords who escaped expropriation faced higher taxation and became nervously
aware that the good old days of the Porfiriato were irrevocably gone; some
prudently moved into trade, industry and banking ± or married their daughters
off to powerful revolutionary generals. The character of Mexico's economic
elite began to change as agrarian oligarchs gave way to entrepreneurial busi-
nessmen.
Meanwhile, behind the impersonal statistics of the reparto lay a host of local
stories: of conflict, mobilization, murder, land invasion, litigation and repres-
sion, complete with a cast of agraristas, rojos, bolcheviques, CROMistas, caciques,
pistoleros, defensas sociales, and guardias blancas.73 This, even more clearly,
represented a departure from the Roman peace of the Porfiriato. It did not, of
course, involve a calm, rational and disinterested division of property rights,
but rather a messy and violent process of class conflict with the state ± in its
several manifestations, national, regional and local ± playing an ambiguous
role: sometimes encouraging reform, sometimes stymying it (Knight, 1991b,
pp. 73±104). The outcome of this complex dialectic was therefore to a large
degree unplanned and unforeseen: between 1930 and 1933 it seemed as if the
agrarian reform had been halted, at least at the national level (Simpson, 1937,
chs 8, 24); after 1934, it forged ahead as never before. Popular pressure and
political rivalries determined the outcome, and the `Leviathan state' was more
end-product than first cause of this decisive and unique process.
Agrarismo therefore had a socioeconomic impact out of proportion to the
mere number of hectares distributed. Labour reform, too, though of less struc-
tural significance, also had implications beyond simple ± or not so simple ±
questions of wages, hours and conditions. Concerning the latter, the evidence is
mixed and imperfect. Article 123 ± the Magna Carta of Mexican labour ± was not
regulated until 1931 (Saragoza, 1988, pp. 155±67; Carr, 1981, p. 175). Through
the 1920s, therefore, labour reform displayed an ìnescapable variation', by both
region and sector (Gruening, 1928, p. 364; note also Carr, 1981, p. 136). Some
points are clear, however. Unionization, feeble during the Porfiriato, forged
ahead; and by 1926 the major labour confederation, the CROM, claimed 2
million members (Gruening, 1928, p. 339; Carr, 1981, p. 158). In some indus-
tries ± printing, railways, oil, textiles, tramways ± unions were now well-
established, and both employers and political authorities had to reckon with
syndical power. The latter depended on political clout, tactical alliances
with major caudillos, a union presence on the new labour tribunals ± Juntas de
ConciliacioÂn y Arbitraje ± and a gradual trend towards collective contracts
embodying the clauÂsula de exclusioÂn (closed shop). Employers deplored the

pretensions and irresponsibility of (many) union leaders; US observers lament-
ed ± and exaggerated ± the àlmost total control of industry by the workers'.74
The textile factories, wholly unionized, suffered from `ruinously undisciplined'
sindicatos; and miners in Chihuahua committed theft of `breathtaking propor-
tions' (Gruening, 1928, p. 350; French, 1996, pp. 174±5). Managerial prerog-
atives, sacrosanct during the Porfiriato, were under threat; `time and work
discipline' was challenged, not by lazy natives, but by insolent Bolsheviki.
Emulating their political masters, post-revolutionary union leaders indulged
in violence, intimidation and corruption. On the other hand, they (sometimes)
got results for their members ± no mean achievement, given the inherent
labour surplus of the Mexican economy. Conditions in the mines improved;
job security was enhanced (at the cost of some featherbedding); and wages on
the railways and in the tobacco industry rose (Gruening, 1928, pp. 345, 351;
Krauze, 1977, pp. 87, 89). Aggregate figures suggest a moderate increase in real
wages; certainly, a reversal of the decline of the late Porfiriato.75 And, taken
together, agrarian and labour reforms helped deepen the domestic market to
the advantage of domestic industry. Employers selling in the domestic market
might individually deplore working-class militancy and Bolshevik agrarianism,
but, collectively, they benefited from enhanced purchasing power in the short
term, and enhanced social stability in the long term. The `relatively autonom-
ous' state of the Sonorans, we may hypothesize, helped promote a sounder
capitalism than had its predecessor, the narrowly oligarchic àgent' state of the
Porfiriato (Knight, 1997).
Not that social reform was dictated from above by a Leviathan state. Some-
times, indeed, reform occurred by virtue of local or regional pressure, in defi-
ance of federal government indifference ± or hostility.76 And the Sonoran state
was still no Leviathan: it failed to crush the Church and had to compromise
with the oil companies; its fiscal take grew, but slowly and modestly.77 Halting
and ambivalent though the process of reform was, it nevertheless carried
important implications for the state, which began to amass power and to assert
its legitimacy in ways that the narrowly personalist and authoritarian Porfiriato
had never attempted ± indeed, had never conceived of. Armed agraristas and
organized labour proved to be pillars of the new regime, not least during the
critical days of the 1923±24 De la Huerta revolt. As praetorian challenges were
faced down ± in 1923±24, 1927 and 1929 ± so the regime cut back its bloated
military expenditure, professionalized the army, and began to devote resources
to social and economic purposes.78 The public health programmes of the
Porfiriato were extended, cutting the death rate; educational expenditure
rose, and the federal government embarked on a programme of rural schooling
that would burgeon during the 1930s.79 The economic role of the federal
government also grew. The state sought not to crowd out private capital, but
Alan Knight 137
rather to facilitate a responsible, productive capitalism: hence, the ambitious

public works of the 1920s (especially roads and irrigation); the recruitment of a
cadre of skilled technocrats (Pani, GoÂmez MorõÂn, Gonzalo Robles); and the
creation of both new regulatory devices (the labour tribunals, the National
Electricity Code, the Mining and Petroleum Laws) and new state agencies ±
the National Irrigation Commission, the National Bank of Ejidal Credit and,
most important of all, the new central bank, the Banco Nacional de MeÂxico.80
The significance of these measures was not immediately apparent. During the
later 1920s, indeed, both economic growth and state-building were checked by
the twin crises of 1928 (the assassination of president-elect ObregoÂn) and 1929±
30 (the onset of the depression). In fact, the latter had been signalled in
advance: like many primary producers, Mexico faced deteriorating terms of
trade after 1926; demand for key exports ± like henequen ± began to decline;
and government revenue fell accordingly (Gruening, 1928, p. 163; ZebaduÂa,
1994, p. 335; Wilkie, 1970, p. 22). The Sonoran leaders, chastened by the war
with the Cristeros and cajoled by the canny US Ambassador, Dwight Morrow,
reacted conservatively. They curtailed the agrarian reform, curbed labour milit-
ancy and adopted deflationary fiscal policies (ZebaduÂa, 1994, p. 324ff.; DõÂaz
Fuentes, 1993, pp. 71±3). Domestic reform would be sacrificed on the altar of
international orthodoxy.
But the internal dynamic of the Revolution was not yet spent. The tensions
engendered by the twin crises of 1928±30 finally broke the Callista coalition,
split and radicalized the labour movement, revived revolutionary agrarianism,
and promoted an intense debate over Mexico's economic project. The result ±
which takes us well beyond our 1930 terminus ± was Cardenismo, the last great
reformist movement/administration of the Revolution. The crisis of 1928±30
made Cardenismo possible. But, equally, without the popular revolution of
1910±17 and the institutional revolution of 1917±30, Cardenismo could not
have emerged. The crisis called into question Mexico's external economic rela-
tions and pointed to a different path of desarrollo hacia adentro, involving
greater state regulation, heightened economic nationalism, and radical social
reform. But that path could be taken only because the revolutionary state now
possessed the institutional power and incipient legitimacy which had been
built up since 1917. They made possible a response to the depression which
was unique within Latin America.
Appendix: tables and a note on statistical series
Economists and economic historians are expected to work with quantitative

data. Those who do not are taken to task by their colleagues (see, for example,
Haber's swipe at Aldo Ferrer who `managed to write an economic history of
Argentina without including a single statistical table': Ìntroduction', p. 8).

Clearly, quantitative data are valuable in enabling us to measure, compare
and model-build; without statistical series we must settle for `soft' conclusions,
built upon ìmpressionistic' evidence (as most historians of most subjects do).
This does not mean, as Haber seems to suggest, that neoclassical economics has
a monopoly of statistical methods and sources, still less that neoclassical eco-
nomics has a privileged claim to scientific method (`the dispassionate analysis
of systematically gathered quantitative data and the specification of testable
hypotheses': Ìntroduction', p. 7). Other branches of the social sciences and
other branches of economics can make as good a claim; indeed, I would go
further and maintain that history ± including economic history ± can claim a
`scientific' credibility even when using non-quantitative data (that is, it can
claim to make a `dispassionate analysis of systematically gathered . . . data and
the specification of testable hypotheses'). Conversely, some quantitative
(including Cliometric) research can be deficient (ùnscientific') in this regard.
So, quantitative data are useful, but not the Holy Grail. Problems arise,
however, when the quest for the Holy Grail becomes a wild-goose chase;
when, in order to find or create statistical series, economists and economic
historians display a distinctly unscientific gullibility, swallowing statistics
which are inherently flawed, dubious or downright wrong (paraphrasing Ste-
phen Decatur's famous toast, it becomes a case of òur statistical series! In its
intercourse with models, theories and testable hypotheses may it always be
right; but our statistical series, right or wrong!'). This is a pertinent considera-
tion, since Mexican statistical series are very unreliable. They are unreliable in
several respects. First, the `cognitive' capacity of the Mexican state has been
quite limited (in which respect Mexico is not, of course, unique). The state has
been unable to collect reliable data either because it has lacked the institutional
powers ± during the Revolution these were at an all-time low, hence the lacunae
in several series ± or because actors in civil society declined to cooperate. Refusal
to comply with state demands for information is an old story which transcends
time, space and class, and which links Balzac's peasants with today's multi-
nationals. So, too, in Mexico. Prior to 1895, when the first official census was
conducted, the registration of births was an exercise in `complete futility' (Cook
and Borah, 1974, vol. 2, p. 317).
The advent of the census did not eliminate all the problems. The 1921 census
was notoriously inadequate, since, in the wake of a major revolution during
which forced recruitment (and other abuses) had been legion, Mexicans ±
particularly poor, rural Mexicans ± were highly suspicious of state agents (see
note 65). However, such suspicions were not wholly new: the 1900 census, too,
had been affected by popular evasion, which included the `200 [inhabitants]
whom neither persuasion nor threat could induce to report at San AndreÂs
[Chihuahua]' (Tannenbaum, 1968, p. 50). Specific census data are also
Alan Knight 139
unreliable. Demographers have noticed a strange clustering of births in parti-

cular years, spaced decennially: the result, not of sudden spurts in natality, but
of informants giving ages, or of enumerators recording ages, in round numbers
± 20, 30, 40 and so on (Cook and Borah, 1974, vol. 2, p. 274 describe this
phenomenon of `heaping'). Literacy figures are also tricky (for discrepancies,
see Tannenbaum, 1968, p. 28, Wilkie, 1970, pp. 207±8); but here the problems
are not only cognitive (what can the state find out?) but also conceptual (what
is literacy? How is it measured?). Similar problems arise with ethnicity.
Counting inanimate objects is not necessarily easier than enumerating sus-
picious people. Companies have a vested interest in avoiding taxation; the oil
companies were notorious for their creative bookkeeping, which supposedly
underestimated production and profits (Meyer, 1977, p. 304, which also notes,
p. 18, that `data on the total amounts of rents and royalties paid are lacking').
As for the profits of Mexican companies, they too remain opaque; to calculate
returns from share performance and dividends is to miss most of the picture,
since the majority of Mexican companies relied on private finance and jeal-
ously preserved company ± even family ± control. Foreign trade, crucial for
government revenue, also presents problems. Contraband had flourished in
nineteenth-century Mexico and, though the Porfirian state certainly estab-
lished tighter border and port surveillance, this again lapsed with the Revolu-
tion. Despite renewed state-building after 1920, Mexico's borders remained
porous; no Mexican administration, for example, has managed to impose
effective exchange controls ± a fact which imposes a marked conditionality
on Mexican economic policy.
Agricultural data, also crucial for understanding the Mexican economy, are
incomplete and often contradictory. A good analysis is provided by Lamartine
Yates whose conclusions, though they relate to post-1930, also shed light on
the pre-1930 period. (I do not subscribe to the Whiggish view that data collec-
tion must improve with time; on the other hand, if we find the relatively large,
bureaucratic state of the 1950s and 1960s getting it wrong, this can hardly
inspire confidence in earlier efforts.) Yates (1981, pp. 271±9), highlights the
problems concerning agricultural data: agricultural censuses only began in
1930; and even thereafter they were seriously deficient. Shifting (slash and
burn) cultivation presented problems: by counting all such land as arable,
enumerators greatly inflated the area of cultivated land; and in 1970, when
this error was corrected, it caused `massive apparent declines' in southern arable
totals. Other overestimates were common, to the extent that Campeche was
credited with more `land in farms' than the surface area of the state. As regards
productivity, sudden jumps (of up to 100 per cent) were common, this being
attributed to arbitrary upward revisions by reporting agencies. So, too, with
livestock: San Luis had 15 asses in 1960, 97 638 five years later (p. 275). Separate
estimates made by the farm census and the Department of Agriculture `diverge
markedly': Yates (p. 275) notes discrepancies to the tune of 2.7 million cattle,
1.5 million horses, 1.4 million mules, 1.1 million hogs and 3 million sheep.
Rival estimates of the increase in total agricultural production for the period
1950±57 range between 130 to 180 per cent. In general, it seems, estimates are
`downwardly sticky'; officials prefer to inflate rather than deflate. Thus, òn one
occasion . . . the Secretary of Agriculture took the undersecretary's figures for the
corn harvest and added 30 per cent before making his announcement' (Yates, p.
274). In light of these post-1930 problems, we can hardly take pre-1930 figures
at face value. It is quite likely, for example, that estimates exaggerate hacienda
production (much of which was marketed) at the expense of peasant produc-
tion (much of which was locally consumed), thus overestimating the negative
impact of the post-1917 agrarian reform.
The state's `cognitive' failure is sometimes deliberate, the result of design
rather than official incapacity or civic evasion. Corruption is the most obvious
and pervasive example: customs officials turn a blind eye; army officers pad
their payrolls (see note 78); and tax-collectors underrate properties (Knight,
1986a, vol.1, p. 93). If, in the 1990s, the black economy constitutes ± it has been
said ± between 20 per cent and 30 per cent of Mexican GDP ± what was its
extent in 1900 or 1930?
Even if data-collection is honest and efficient, our retrospective reading of
the data may be error-prone. Major misapprehensions have arisen from histor-
ians' (mis)understanding of contemporary categories of analysis: for example,
haciendas, ranchos and pueblos (see notes 47, 57). Tannenbaum (and others)
readily equated peon with `debt peon' and `debt peon', with serf, concluding
(Tannebaum, 1968, p. 30) that millions of Mexicans groaned in rural servitude
in 1910 (which they did not). Errors of this kind easily occur when data series
are collected and collated by analysts who are not closely familiar with the
original sources and who may be taken in by deceptive semantic similari-
ties. The higher the level of aggregation, the greater the danger of such false
assimilation.
There is a final problem, of a more grand theoretical kind. By virtue of seeking
quantifiable data, the economist places a premium on what can be measured.
Thus, as I have said, hacienda production may be exaggerated at the expense of
subsistence production; and industrial output may show up more readily than
artisanal output. Monetized market transactions are therefore readily access-
ible, other transactions are obscured. If `the economy' is defined in terms of
market relations, then the problem is tautologically removed. If it is more
broadly defined ± that is, in Polanyi's terms, as the material means-provisioning
system ± the problem remains. For example: despite vicissitudes the Mexican
economy grew substantially during the twentieth century; per capita income
rose. To the extent that, say, Mexicans lived longer, this was a real transforma-
tion. But in many other respects inferring welfare from income is clearly risky.
Alan Knight 141
Urbanization correlated with higher money incomes; but some of those

incomes had to be devoted, for example, to intra-city travel, previously unne-
cessary. So, too, with food: we have seen, in the case of the rationalizing land-
lords of the late Porfiriato, that workers sometimes resisted cash payments and
clung to non-monetary perks ± that is, food rations. To the extent that more
transactions entered the market ± hence could be statistically captured ± `the
economy' seemed to grow. Yet, while some of these transactions represented a
net increment of `welfare', others were simply monetized alternatives to pre-
vious non-monetary transactions. `Welfare' may have remained unaltered; or it
may have declined ± as the protesting peons clearly believed.
Finally, it should be remembered that a large slice of Mexico's enhanced GNP
has derived from greater consumption of non-renewable fossil fuels, particu-
larly oil, which represents current consumption at the expense of future
generations. Again, conventional systems of national accounting neglect this
cross-generational transfer of resources.
In light of these caveats, I have doubts about both the validity and the
usefulness of statistical series; broad trends ± those mentioned in the text of
this chapter ± can be established with some confidence, at least in a broadbrush,
ìmpressionistic' way; but to go beyond this to putatively precise series ± poten-
tial raw material for sophisticated comparative analysis ± is another matter, and
may generate more misapprehension than enlightenment. Sometimes, discre-
tion is the better part of scholarly valour; as Wittgenstein put it, `whereof thou
cannot speak, do not speak'.
Notes
1 With the exception of HeÂctor PeÂrez Brignoli's chapter on `Central America'. However,
Central America (i) has witnessed repeated efforts at political and economic integra-
tion and (ii) is, in its totality, smaller than many individual republics, Mexico
included.
2 For a useful discussion, involving Mexico, see Migdal (1988).
3 In Michael Mann's terminology, such states, for all their `despotic' power, lacked
ìnfrastructural' power ± `the capacity of the state actually to penetrate civil society
and to implement logistically [sic] political decisions throughout the realm' (Mann,
1988, p. 5).
4 The Banco Nacional, established in 1884, was closely tied to the state and was respon-
sible for 46 per cent of the currency in circulation in 1910. It did not, however, enjoy a
monopoly of emission: the Banco de Londres y MeÂxico accounted for 16 per cent ±
down from 29 per cent in 1900 ± and a swathe of provincial banks scattered through-
out the states of the Federation were responsible for a further 38 per cent: Rosenzweig,
(1965b); ZebaduÂa (1994, pp. 44±5).
5 `They make their sales in the primitive form . . . they do not use money, only by means
of exchanging maize, beans, dried chile and tortillas': report from Coyotepec, Puebla,
1926, cited in Tannenbaum (1968, pp. 85±6). The use of Guatemalan pesos ± which
traded at 0.75 Mexican centavos ± represented a `great advantage' to the Soconusco

coffee planters: GarcõÂa de LeoÂn (1985, vol. 1, p. 198 citing Karl Kaerger).
6 Reynolds (1970, p. 16); Wilkie (1970, p. 22, ch. 4). Even after 1930, state expenditure
was not particularly high by Latin American or (a fortiori) Western European stand-
ards; a second `rise of the active state', characterized by the proliferation of state
enterprises, occurred after 1970.
7 In 1910 the Mexican government raised 44 per cent of its revenue from external trade
± almost entirely imports ± and 26 per cent from internal trade: Victoria and VelaÂz-
quez RamõÂrez (1988, pp. 298±9).
8 In 1930 Mexico possessed 14 600 miles of railway, compared to 900 of paved road. By
1950 the respective figures were 14 600 and 13 400: Cedillo VaÂzquez (1964, p. 48 and
(maps) pp. 129, 135).
9 Hassig (1985, pp. 187±92, 197±204); and, for pictorial examples, Gruening (1928,
facing p. 88), and GonzaÂlez (1979, p. 3).
10 Wilkie (1970, pp. 207±9), gives a national figure of 76.9 per cent illiteracy in 1910, but
offers no discussion of the problems of defining `literacy' (see Appendix).
11 Cf. Robert Scott, cited in Turner (1968, p. 10), who asserts that in 1910, 90 per cent of
Mexicans being subsistence peasants lacked a sense of nationality.
12 Haber (1989, p. 27), estimates that in 1895 only some 4.8 million Mexicans ± out of a
population of 12.6 million ± were included within the consumer economy, that is,
had ìncomes large enough to purchase manufactured goods'. It should be added that
this is a pretty rough and arbitrary calculation, based on an estimate of 2 million
wage workers engaged in `mining, manufacturing, petroleum production [sic], com-
merce, transport, and other services', each with 1.5 dependents. Haber's argument is
sound; but I am not sure that his estimate derives from `the dispassionate analysis of
systematically gathered quantitative data', which he ± very reasonably ± urges upon
the remiss economic historians of Latin America: Stephen Haber, Ìntroduction', in
Haber (1997, p. 7).
13 `Verdammte Bedu È rfnislosigkeit', as a disgruntled German merchant put it: Chase
(1931, p. 313).
14 The colonial Maya were seen as similarly wantless: Farris (1984, p. 45).
15 French (1996, pp. 23, 27, 44) is a good case study.
16 Cf. Albert (1985, pp. 231±49); and LeGrand, work in progress on the Santa Marta
banana ènclave' on the Colombian coast. CaÂrdenas (1987, pp. 24±8), offers figures of
enclave ownership and, as an index of linkage, the returned value of exports, which
range from a low of 39 per cent (oil) to a high of 87 per cent (henequen); mining, at
81 per cent, is also relatively high.
17 The Mexican mining labour force in 1910 stood at 100 000 in a population of
15 million; the mining and extractive sector in general, which generated about
9 per cent of GDP and 60 per cent of exports, represented only 2 per cent of
the economically active labour force: Reynolds, 1970, pp. 18, 63, 205. Bolivia's
miners, in 1900, numbered only 13 000 (1 per cent of the economically active labour
force).
18 The correlation between mining and Precolumbian sedentary civilizations may be
random; it may, on the other hand, reflect the fact that ± with the notable exception
of the Maya ± states and civilization tended to form in the highlands, in more
temperate climes, thus coinciding with mineral-rich strata.
19 For an interesting discussion of Barrington Moore's thesis and its applicability to
Latin America, see Huber and Safford (1995). Unfortunately, Mexico does not figure
as a case in this symposium.
Alan Knight 143
20 In passing, I should like to discount analyses of Latin American (including Mexican)

development which place great emphasis on enduring psycho-cultural attributes:
Catholic corporatism, centralism, authoritarianism, and so forth. The latter ± to the
extent that they exist at all ± are products of circumstances, not autonomous deter-
minants. Hence they are of little use when it comes to explaining Latin American
economic development, or the lack of it.
21 Cuba and, to a lesser extent, the republics of Central America also had to live with a
powerful US presence. Mexico was unusual in being a large state/economy which
fell within the US geopolitical orbit. During the period under consideration, the
threat of US armed intervention in Mexico receded during the Porfiriato, peaked
during the armed revolution, then receded again during the 1920s. More permanent
and influential, I would suggest, was US diplomatic recognition ± which affected
business confidence and creditworthiness ± and the growing economic intimacy,
evidenced in trade, investment, border commerce and migration; this intimacy
clearly conditioned Mexican policy, restraining potential reform, but also deterred
US aggression ± for example, in 1925±27 and 1938. I discuss some of the issues in
Knight (1987).
22 It is necessary to stress this point ± which may seem obvious ± since recent `revisio-
nist' scholarship has tended (i) to downplay the significance of the Revolution (some-
times even denying its `revolutionary' status altogether) and (ii) to attribute its
outbreak to external causes. For an outline of the revisionist position(s), see Knight
(1992b, pp. 180±8).
23 The discussion thus parallels the `Brenner debate' concerning the impact of commer-
cialization on the economies of early modern Europe, which began with Brenner
(1975, pp. 30±75).
24 CaÂrdenas (1987, p. 25), shows that in 1926 oil and mining accounted for 33 per cent
and 43 per cent of exports and were 99 per cent and 95 per cent foreign-owned
respectively. In the mines, indeed, both technological change and revolutionary
upheaval favoured concentration of production in fewer, bigger, North American
hands: Sariego, Reygadas, GoÂmez and Farrera (1988, pp. 63±6).
25 Half a million Mexicans legally entered the USA between 1899 and 1928, the great
majority after 1910; by 1930 over 10 per cent of Mexico's population was in the USA;
between 1930 and 1932, at least 2 million returned ± or were returned: Balderrama
and RodrõÂguez (1995, pp. 7, 121). See also French (1996, pp. 42±3).
26 The question of regime legitimacy ± which is often glibly asserted ± is particularly
tricky: Knight (1997).
27 Another interesting puzzle, which also lies beyond our chronological remit, is why,
and to what extent, the economies of Latin America in general and Mexico in
particular increasingly lagged behind those of the USA or Western Europe during
the nineteenth century. For it was then that the gap widened; and twentieth-century
growth has, in general, failed to narrow it: see Haber (1997).
28 Reynolds (1970, appendix A, pp. 311±4) attempts to compare late colonial and
Porfirian national income, concluding that the former was appreciably higher, in
per capita terms, and that èither there was little or no increase in per capita income
over the course of the nineteenth century, or per capita income actually declined over
much of the nineteenth century despite the gains of the Porfiriato'. This strikes me as
an exaggeration, based on an overvaluation of bullion production ± much of which
went to Spain ± and an underestimate of, for example, petty commerce and subsis-
tence agricultural production. It is worth noting that the dearths of the late colonial
period ± notably 1785 ± were far worse than those of the Porfiriato. CaÂrdenas (1997,
p. 65) broadly concurs in respect of a fall in per capita GDP of 37 per cent between
1800 and 1860.
29 Tutino (1986) offers a good overview of these successive phases of `compression' and
`decompression'.
30 Walker (1986); Haber (1997, pp. 17±18) compares the `repressive regulatory' ± and
chronically irresponsible ± financial policies of nineteenth-century Mexican govern-
ments with the more sound and business-friendly policies of their Brazilian counter-
parts, which fostered a more developed capital market.
31 Haber (1989, ch. 6; and pp. 68±83), which discusses industrial finance, noting the
close links between merchant-financiers and industrialists; Haber possibly exagge-
rates the role of foreign finance and entrepreneurship, on which, note Camp (1985,
p. 56 and passim).
32 Travellers accounts of Mexico in the period c. 1820±50 frequently comment on its
rustic primitiveness; but few or none describe poverty, inequality and destitution in
the bleak terms of Humboldt (1988).
33 Knight (1999).
34 The CientõÂficos thus anticipated the òrthodox paradox' of the neoliberal 1990s ± that
is, a `state-led transformation in favor of the market . . . the use of the state to reduce
and diminish the economic influence of politics': Centeno, 1994, p. 34 (the phrase is
Miles Kahler's). Of course, we might better say the neoliberals emulated the CientõÂficos
(mutatis mutandis).
35 Cardoso (1980), second part, offers a good overview based on CosõÂo Villegas' monu-
mental Historia Moderna de MeÂxico: El Porfiriato, La Vida EconoÂmica.
36 CossõÂo Silva (1965) `La agricultura' and Nava Oteo (1965) `La minerõÂa', in CosõÂo
Villegas (1965, vol. 1, chs 1, 3); CaÂrdenas (1997, pp. 80, 83), on the effect of exchange
rate depreciation.
37 Haber (1989, p. 16). `Social savings resulting from the advent of the railroad may have
been as high as 38.5 per cent of GDP for Mexico': Haber, Ìntroduction', p. 16, citing
William Summerhill, `Transport Improvements and Economic Growth in Brazil and
Mexico', in Haber (1997).
38 Haber (1989, pp. 7±8, 30ff); Rosenzweig (1965a, pp. 329, 342). Manufacturing
industries, like exporters, were helped by exchange rate depreciation (CaÂrdenas,
1997, pp. 83±4); meanwhile, both federal and state government concessions
helped offset the higher cost of imported capital goods: Rosenzweig (1965a, pp.
467±72).
39 As Haber generalizes for Latin America as a whole: Haber (1997, p. 13).
40 Coatsworth (1976, pp. 167±87) rebuts the old myth of a massive decline in per capita
food production during the Porfiriato; but even his own evidence shows a deteriora-
tion in the final years of the regime, which required sizeable food imports to cover
Mexico's shortfall at a time when food prices were rising significantly. See also CossõÂo
Silva (1965, pp. 21±2).
41 Leal and Huacuja Rountree (1982, chs 2 and 3) offers good analysis of the pulque
economy.
42 Knight (1986, vol. 1, ch. 3); Simpson (1937, p. 712), offers a (rare) index of real
agricultural wages for the late Porfiriato, showing a decline from 1903100 to
191068.
43 Katz (1980, pp. 25±48) offers a good, conventional, breakdown (north, centre, south),
which I will follow; Reynolds (1970, pp. 7±8, 94±5) prefers a five-region model.
44 The dispossession of the Yaquis provoked intense conflict, leading to mass deporta-
tion of Yaqui families which some landlords ± dependent on Yaqui labour ± sought to
Alan Knight 145
avert; but racism and Realpolitik prevailed over economic interest: Knight (1986, vol.
1, pp. 111±12).
45 Vos (1988); Joseph (1988, chs 1, 2). The ratio of coerced to voluntary labour is a
matter of debate: see Knight (1986b, pp. 41±74).
46 Ìndians only hear through their backsides': Joseph (1988, p. 70).
47 Recent research has rehabilitated the rancho ± roughly, the small farm ± as an
important element in Mexican rural history, thus qualifying the old dichotomy of
hacienda versus village; it has also confirmed ± rather than discovered de novo ± that
categories like hacienda and rancho were often used arbitrarily or inconsistently: see
Tannenbaum (1968, pp. 30±1) and Guerra (1985, vol. 1, pp. 325±9 and vol. 2, annexe
5). A general overview such as this does not permit a full discussion; I shall focus on
the hacienda ± roughly, the large profit-seeking private farm ± although much of what
is said also applies, to some degree, to the rancho ± roughly, the small profit-seeking
private farm. Apart from blurring this (èmic') distinction, I shall avoid making any
strict (ètic') distinction between hacienda and `plantation'; what for some are `plan-
tations', are for me simply haciendas producing `tropical' crops, usually for export.
48 Womack Jr. (1968, ch. 2) is the classic study. On artisanal decline: Knight (1986, vol. 1,
pp. 131±2); Haber (1989, p. 58); Rosenzweig (1965a, pp. 320±2, 340±1), which charts
an 83 per cent decline in artisan consumption of raw cotton during the Porfiriato.
49 Labour may have been cheap, but it was also reckoned to be feckless and inefficient:
Haber (1989, p. 35); Gruening (1928, pp. 350, 385).
50 In fact, the most common form of prerevolutionary working-class organization was
not the trade union but the mutualist society; self-help seemed more realistic than
syndical militancy: Knight (1986, vol. 1, pp. 133±50); French (1996, pp. 133±7).
51 Reynolds (1970, p. 105); Rosenzweig (1965b, pp. 882±3). Cf. Yates (1981, p. 3), which
presents Angulo's optimistic and contentious figures for the 1900s.
52 Cigarette packets sold: 1906±07: 524 m; 1908±09: 505 m; CervecerõÂa CuauhteÂmoc
production: 1900: 4.87 m litres; 1907: 14 m; 1909: 11.58 m. Haber (1989, pp. 49, 53).
53 Rawski (1996, p. 51).
54 Between 1905 and 1910 Mexico's net barter terms of trade fell from 124 to 93
(1937100): Reynolds (1970, p. 43 n. 24). The US recession, of course, hit the north-
ern mining economy hardest: French (1996, pp. 30±2).
55 There is considerable debate concerning the relationship between the economic
downturn of the late Porfiriato and the growth of political opposition. I would
question any neat causal link, partly on theoretical grounds, partly because the
opposition (i) had deep roots, antedating 1907, and (ii) followed a politico-electoral
± rather than a business cycle ± timetable: Knight (1986, vol. 1, pp. 64±5).
56 Shallow in comparison with, say, Argentina, where, though the population was half
that of Mexico, manufacturing output in 1913 was 67 per cent greater: Victor Bulmer-
Thomas (1994, p. 137).
57 I am referring chiefly to Mexico's èxternal peasantry' ± that is, those who remained
in villages and had not been absorbed into hacienda (ìnternal peasant') communities.
Contrary to some received wisdom, they were still a sizeable majority in 1910: Jean
Meyer (1986, pp. 477±509).
58 By no means all peasant rebels were agraristas, seeking land reform; some were
motivated by the desire for local political autonomy, which Porfirian state-building
± partly funded and facilitated by export-led growth ± had severely undermined: I
have referred to such rebels (loosely) as serranos: Knight (1986, vol. 1, pp. 115±27).
59 Again, a necessary caveat, given that revisionist scholarship would question the
`revolutionary' impact of the Revolution (see note 22 above).
60 European economies during the Second World War are a case in point; within our
own field, the case of Colombia ± both now and during the Violencia ± springs to
mind.
61 Military spending as a percentage of total government spending: 1910, 20 per cent;
1913, 26 per cent; 1917, 70 per cent; 1925, 31 per cent; 1935, 21 per cent: Wilkie
(1970, p. 102).
62 Marichal (1989, pp. 178, 213). Mexico's default helped accelerate the flow of US funds
to South America: Stallings (1987, pp. 125±6).
63 Womack (1978, pp. 80±123). Oil production rose from 3.6 m barrels in 1910 to 26.2 m
in 1914, 63.8 m 1918, and 193.4 m in 1921 (the peak); petroleum taxes contributed
11 per cent of federal government receipts in 1918, 21 per cent in 1920 and (a freak)
34 per cent in 1922, after which decline set in; during this period oil production
represented between 6 and 7 per cent of GNP: L. Meyer (1977, pp. 8±9, 15, 16).
64 US policy towards the institutional revolution ± which combined a desire for stability
with antipathy to radical reform ± is well-described by Smith (1972 ) and ZebuduÂa
(1994, chs 5, 8, 9). Erfani (1995, pp. 5±6, 10ff.), argues that the revolutionary state
fortuitously benefited from the emergence of `new norms about state sovereignty'
(that is, respect for national self-determination) which came to govern international
± especially US-Mexican ± relations in the 1920s (or before: if the inspiration is
Wilsonian, we should backdate the shift to c. 1913). There is some truth in this
argument, though not as much as the author believes.
65 The 1921 census gives a total population of 14.33 m, compared to 15.16 m in 1910;
natural increase ± in the absence of revolution ± would have resulted in a population
of around 16.91 m in 1921. However, the 1921 census undercounted; Reynolds
estimates (1970, p. 18, 26) to the tune of nearly half a million, which does not strike
me as excessive. Thus, the absolute population loss was between 300 000 and 400 000;
the demographic deficit ± taking into account natural increase that did not occur ±
nearer 2 millions. Since birth rates do not appear to have changed much over time
(see Reynolds, 1970, p. 18 and Cook and Borah, 1974, vol. 2, p. 319), the chief
determinant was the death rate which, having gradually declined during the Porfir-
iato (reaching 3.29 per cent in 1910), shot up during the Revolution, then, in the
1920s, resumed its gentle descent, initially due to the elimination of smallpox and
typhus: 1921, 2.84 per cent; 1930, 2.67 per cent. Revolutionary mortality was very
largely a matter not of fighting but of dearth and disease ± in particular, the Spanish
influenza which ravaged a weary world in 1918±19 (Knight, 1986, vol. 2, pp. 420±2);
thus, younger males did not suffer disproportionately. Arguably, disease killed off the
older, weaker members of society, leaving a younger, more productive postrevolu-
tionary population.
66 The rural population was 71.2 per cent of the population in 1910, 68.8 per cent in
1921, and 66.5 per cent in 1930 (Reynolds, 1970, p. 93). During the decade of
revolution, Mexico City grew by a quarter; Ciudad Juarez by 140 per cent (Knight,
1986, p. 523).
67 For example, cotton textile production: 1910, 34 700 tons (cotton consumed); 1918,
20 300; 1921, 31 700; 1925, 43 700: Haber (1989, p. 125).
68 Meyer (1977, pp. 8±9); Bernstein (1964); Sariego et al. (1988, pp. 63±74). Brown (1985,
pp. 362±85) convincingly argues that subsoil depletion was more important than
radical reform in obliging US interests to shift from Mexico to Venezuela.
69 Again, this is tricky question. According to Reynolds (1970, p. 105) production of
corn and beans fell by 40 per cent and 31 per cent respectively between 1907 and
1929, while population grew 9 per cent. For graphic confirmation, see Simpson
Alan Knight 147
(1937, p. 499). While there was no doubt some switch of consumption away from
these staples, the fall seems excessively large and may reflect the old problem of
adequately measuring subsistence production ± especially during a period when a
peasant revolution had recently concluded and endemic rural violence, including the
Cristero rebellion, was continuing.
70 Haber (1989, p. 143); Krauze (1977, pp. 289±93). The development of the Mexican
consumer market, which was obviously linked to processes of import substitution,
urbanization and US direct investment in manufacturing, has been little studied; as a
glance at newspaper adverts suggests, consumerism ± incipient during the later
Porfiriato ± forged ahead during the 1920s and (even) the 1930s.
71 Socialist revolutions ± for example, Russia 1917, Cuba 1959 ± undertook rapid and
radical economic transformation; but Mexico's Revolution was not socialist and ± in
the manner of other `bourgeois' revolutions, perhaps ± it placed more emphasis on
political and cultural change, especially in the 1920s.
72 Simpson (1937, p. 622), shows that ejidos controlled as much as 87 per cent of
Morelos crop land in 1933: compare a national average of 10 per cent. Not that
reform guaranteed well-being: Gruening (1928, pp. 162±3).
73 Good studies are: Friedrich, (1977) and Craig (1983).
74 Carr (1981, p. 134), quoting a US embassy report of April 1924.
75 Simpson (1937, p. 712): real agricultural wages (1903 100): 1910: 68; 1920: 120;
1925: 135; 1930: 118.
76 For example, Carrillo Puerto's brief radical experiment in YucataÂn, or Adalberto
Tejeda's Veracruzano agrarianism: Joseph (1988, part 3); Fowler Salamini (1977).
77 Government spending in relation to GDP: 1910: 4.4 per cent; 1925: 5.9 per cent;
1930: 6.7 per cent; 1935: 6.8 per cent; 1940: 8.2 per cent. From Wilkie (1970, p. 22)
and Reynolds (1970, p. 16). Comparable, though slightly different, figures are given
in Table 5.1.
78 Wilkie (1970, pp. 58, 160±1). The process of demilitarization not only incurred
praetorian protest and revolt, but also involved considerable resort to graft and
peculation by way of buying off dissident officers: for a good example, see ZebaduÂa
(1994, p. 343). Gruening (1928, 319±31) presents a sustained indictment of `scanda-
lous conduct and public debauchery. . . in the army'; the economic costs of graft are,
of course, hard to quantify.
79 Cook and Borah (1974, pp. 432±5); Gruening (1928, pp. 544±50), which shows the
federal health budget rising from around half a million pesos p.a. in the late Porfiriato
to 8.4 m in 1927.
80 ZebaduÂa (1994, pp. 242, 255ff.), which notes (p. 258) the parallel with the Kemmerer
missions to South America: Kemmerer was not called upon to reform Mexico's finan-
cial system ± although he did play an advisory part in the currency stabilization of
1916±17, from which flowed his valuable study of Mexico's revolutionary inflation:
Inflation and Revolution: Mexico's Experience of 1912±17 (Princeton, 1940); however, the
Mexican banking reforms of the 1920s were distinctly Kemmerer-ish in character. See
also Krauze (1977, pp. 35±52, 134±45) (irrigation); Wionczek (1967, pp. 64±5, 191).
References
Albert, B. (1985) Èxternal Forces and the Transformation of Peruvian Coastal Agriculture',
in C. Abel and C. Lewis (eds), Latin America, Economic Imperialism and the State, London:
Athlone Press.
Balderrama, F.E. and RodrõÂguez, R. (1995) Decade of Betrayal. Mexican Repatriation in the
1930s, Albuquerque: University of New Mexico Press.
Banfield, E. (1958) The Moral Basis of a Backward Society, Illinois: Glencoe.
BaÂtiz VaÂzquez, J.A. and Canudas Sandoval, E. (1980) Àspectos financieros y monetarios',
in C. Cardoso (ed.), MeÂxico en el siglo XIX, Mexico: Nueva Imagen.
Bellingeri, M. (1978) `Del peonaje al salario: el caso de San Antonio Tochatlaco de 1880 a
1920', Revista Mexicana de Ciencias PolõÂticas y Sociales, vol. 24, no. 91.
Bernstein, M. (1964) The Mexican Mining Industry, 1890±1950, Albany: State University of
New York.
Brenner, R. (1975) Àgrarian Class Structure and Economic Development in Pre-Industrial
Europe', Past and Present, vol. 70.
Brown, J. (1985) `Why Foreign Oil Companies Shifted their Production from Mexico to
Venezuela during the 1920s', American Historical Review, vol. 90.
CalderoÂn, F.R. (1965) `Los ferrocarriles', in C. Villegas (ed.), Historia Moderna de MeÂxico: El
Porfiriato, La Vida EconoÂmica (vol. 1), Mexico: Editorial Hermes.
Calles, P.E. (1927) Mexico Before the World, New York: Academy Press.
Camp, R.A. (1985) Entrepreneurs and Politics in Twentieth-century Mexico, New York: Oxford
University Press.
CaÂrdenas, E. (1997) À Macroeconomic Interpretation of Nineteenth-Century Mexico', in
S.H. Haber (ed.), How Latin America Fell Behind, Stanford: Stanford University Press.
ÐÐ (1987) La industrializacioÂn de MeÂxico durante la Gran DepresioÂn, Mexico: El Colegio de
MeÂxico.
Cardoso, C. (1980) MeÂxico en el siglo XIX (1821±1910): Historia EconoÂmica y de la Estructura
Social, Mexico: Nueva Imagen.
Carr, B. (1981) El movimiento obrero y la polõÂtica en MeÂxico, Mexico: Ediciones Era.
Cedillo VaÂzquez, V. (1964) Principales factores del desarrollo agrõÂcola en MeÂxico (1925±60),
Mexico: Universidad AutoÂnoma de MeÂxico.
Centeno, M.A. (1994) Democracy Within Reason. Technocratic Revolution in Mexico, Univer-
sity Park, Penn.: The Pennsylvania State University, Press.
Chase, S. (1931) Mexico: A Study of Two Americas, New York: The Literary Guild.
Coatsworth, J.H. (1988) `Patterns of Rural Rebellion in Latin America: Mexico in Com-
parative Perspective', in F. Katz (ed.), Riot, Rebellion and Revolution: Rural Social Conflict in
Mexico, Princeton: Princeton University Press.
ÐÐ (1984) El impacto econoÂmico de los ferrocarriles en el porfiriato, Mexico: Ediciones Era.
ÐÐ (1976) Ànotaciones sobre la produccioÂn de alimentos durante el Porfiriato', Historia
Mexicana, vol. 26(2).
Cook, S.F. and Borah, W. (1974) Essays in Population History. Mexico and the Caribbean, 2
vols, Berkeley: University of California Press.
CoÂrdova, A. (1988) La ideologõÂa de la RevolucioÂn Mexicana, first pub. 1973, Mexico: Edi-
ciones Era.
CosõÂo Villegas, D. (ed.) (1965) Historia Moderna de MeÂxico: El Porfiriato, La Vida EconoÂmica,
2 vols, Mexico: Editorial Hermes.
CossõÂo Silva, L. (1965) `La agricultura', in CosõÂo Villegas (ed.), Historia Moderna de MeÂxico.
El Porfiriato, La vida econoÂmica, Vol. 1, Mexico: Editorial Hermes.
Craig, A.L. (1983) The First Agraristas, Berkeley: University of California Press.
DõÂaz Fuentes, D. (1993) Las polõÂticas fiscales latinoamericanas frente a la gran depresioÂn,
Madrid: Ministerio de EconomõÂa y Hacienda.
Alan Knight 149
Erfani, J.A. (1995) The Paradox of the Mexican State. Rereading Sovereignty from Independence
to NAFTA, Boulder: Lynne Rienner.
Farris, N.M. (1984) Maya Society under Colonial Rule, Princeton: Princeton University Press.
Fischer, D.H. (1970) Historians' Fallacies, New York: Harper.
Florescano, E. (1986) Precios de maõÂz y crisis agrõÂcolas en MeÂxico, 1708±1810, Mexico:
Ediciones Era.
Foster, G. (1965) `Peasant Society and the Image of the Limited Good', American Anthro-
pologist, vol. 67.
Fowler Salamini, H. (1977) Agrarian Radicalism in Veracruz, 1920±38, Lincoln: University of
Nebraska Press.
French, W.E. (1996) A Peaceful and Working People. Manners, Morals and Class Formation in
Northern Mexico, Albuquerque: University of New Mexico Press.
Friedrich, P. (1977) Agrarian Revolt in a Mexican Village, Chicago: University of Chicago
Press.
Gamio, M. (1982) Forjando Patria, first pub. 1916, Mexico: Editorial PorruÂa.
GarcõÂa de LeoÂn, A. (1985) Resistencia y utopõÂa, MeÂxico: Ediciones Era.
Gerschenkron, A. (1962) Economic Backwardness in Historical Perspective, New York: Praeger.
GonzaÂlez, L. (1979) Historia de la RevolucioÂn Mexicana. PerõÂodo 1934±40. Los artõÂfices del
Cardenismo, Mexico: El Colegio de MeÂxico.
Gruening, E. (1928) Mexico and its Heritage, London: Stanley Paul.
Guerra, F.X. (1985) Le Mexique. De l'Ancien ReÂgime aÁ la ReÂvolution, 2 vols, Paris: L'Harmat-
tan.
Haber, S.H. (ed.) (1997) How Latin America Fell Behind: Essays on the Economic Histories of
Brazil and Mexico, 1800±1914, Stanford: Stanford University Press.
ÐÐ (1989) Industry and Development: The Industrialisation of Mexico, 1890±1940, Stanford:
Stanford University Press.
Hassig, R. (1985) Trade, Tribute and Transportation. The Sixteenth-Century Political Economy
of the Valley of Mexico, Norman: University of Oklahoma Press.
Huber, E. and Safford, F. (eds) (1995) Agrarian Structure and Political Power. Landlord and
Peasant in the Making of Latin America, Pittsburgh: University Pittsburgh Press.
Humboldt, A. von, (1988) Political Essay on the Kingdom of New Spain, first pub. 1811,
Norman: University of Oklahoma Press.
Joseph, G.M. (1988) Revolution from Without. YucataÂn, Mexico and the United States, 1880±
1924, Durham: Duke University Press.
Katz, F. (1980) La servidumbre agraria en MeÂxico en la eÂpoca porfiriana, Mexico: Editorial Era.
Kemmerer, E.W. (1940) Inflation and Revolution: Mexico's Experience of 1912±17, Princeton:
Princeton University Press.
Knight, A. (1999), `Britain and Latin America in the Nineteenth Century', in W. R. Louis,
(ed.), The Oxford History of the British Empire, Vol. 3, Oxford: Oxford University Press.
ÐÐ (1997), `The Modern Mexican State: Theory and Practice', paper presented at the
LASA congress, Guadalajara, April.
ÐÐ (1992a) `The Peculiarities of Mexican History: Mexico Compared to Latin America,
1821±1992', Journal of Latin American Studies, quincentenary edition.
ÐÐ (1992b) `Revisionism and Revolution: Mexico Compared to England and France',
Past and Present, vol. 134, pp. 180±8.
ÐÐ (1991a) `The Rise and Fall of Cardenismo', in L. Bethell (ed.), Mexico since Indepen-
dence, Cambridge: Cambridge University Press.
ÐÐ (1991b) `Land and Society in Revolutionary Mexico: The Destruction of the Great
Haciendas', Mexican Studies/Estudios Mexicanos, vol. 7(1), pp. 73±104.
ÐÐ (1987) US±Mexican Relations, 1910±40: An Interpretation, San Diego: Center for US-
Mexican Studies.
ÐÐ (1986a) The Mexican Revolution, 2 vols, Cambridge: Cambridge University Press.
ÐÐ (1986b) `Mexican Peonage: What Was It and Why Was It?', Journal of Latin American
Studies, vol. 18(1).
Krauze, E. (1977) La reconstruccioÂn econoÂmica, Mexico: Colegio de MeÂxico.
Leal, J. F. and Huacuja Rountree, M. (1982) EconomõÂa y sistema de haciendas en MeÂxico,
Mexico: Editorial Era.
Leroy Ladurie, E. (1969) Les paysans de Languedoc, Paris: Flamarion.
Maddison, A. (1991) Dynamic Forces in Capitalist Development, Oxford: Oxford University
Press.
Mann, M. (1988) `The Autonomous Power of the State: Its Origins, Mechanisms and
Results', in Mann, M., States, War, and Capitalism, Oxford: Blackwell.
Marichal, C. (1989) A Century of Debt Crises in Latin America, Princeton: Princeton Uni-
versity Press.
MartõÂnez Alier, J. (1977) Haciendas, Plantations and Collective Farms, London: Frank Cass.
Marx, K. (1957) Capital, 2 vols, London: Dent.
McKellar, M.M. (1994) Life on a Mexican Ranche [sic], Cranbury, NJ: Associated University
Presses.
Meyer, J. (1986) `Haciendas y ranchos, peones y campesinos en el Porfiriato: algunas
falacias estadõÂsticas', Historia Mexicana, vol. 35(3).
Meyer, L. (1977) Mexico and the United States in the Oil Controversy, 1917±42, Austin:
University of Texas Press.
Migdal, J.S. (1988) Strong Societies and Weak States. State±Society Relations and State Cap-
abilities in the Third World, Princeton: Princeton University Press.
Nava Oteo, G. (1965) `La minerõÂa', in CosõÂo Villegas (ed.), Historia Moderna de MeÂxico. El
Porfiriato, La vida econoÂmica, Vol. 1, Mexico: Editorial Hermes
North, D. (1990) Institutions, Institutional Change, and Economic Performance, Cambridge:
Potash, R.A. (1959) El banco de avõÂo en MeÂxico. El fomento de la industria, 1821±46, Mexico:
Randall, R.W. (1972) Real del Monte: A British Mining Venture in Mexico, Austin: University
of Texas Press.
Rawski, T.G. (1996) Ìssues in the Study of Economic Tends', in Rawski et al., Economics and
the Historian, Berkeley: University of California Press.
Reed, N. (1964) The Caste War of YucataÂn, Stanford: Stanford University Press.
Reynolds, C.W. (1970) The Mexican Economy: Twentieth-Century Structure and Growth, New
Haven: Yale University Press.
Rosenzweig, F. (1965a) `La industria', in CosõÂo Villegas (ed.), Historia Moderna de MeÂxico. El
Porfiriato. La Vida EconoÂmica, 2 vols, Mexico: Editorial Hermes.
ÐÐ (1965b) `Moneda y bancos', in D. CosõÂo Villegas (ed.), Historia Moderna de MeÂxico. El
Porfiriato. La Vida EconoÂmica, 2 vols, Mexico: Editorial Hermes.
ÐÐ (1965c) `Las inversiones extranjeras', in D. CosõÂo Villegas (ed.), Historia Moderna de
MeÂxico. El Porfiriato. La Vida EconoÂmica, 2 vols, Mexico: Editorial Hermes.
Rugeley, T. (1996) YucataÂn's Maya Peasants and the Origins of the Caste War, Austin: Uni-
versity of Texas Press.
Saragoza, A.M. (1988) The Monterrey Elite and the Mexican State, Austin: University of Texas
Press.
Sariego, J.L. (1988) Enclaves y minerales en el norte de MeÂxico, Mexico: Editorial de la Casa
Chata.
Alan Knight 151
Sariego, J.L. Reygadas, L., GoÂmez, M.A. and Farrera, J. (1988) El Estado y la minerõÂa
mexicana. PolõÂtica, trabajo y sociedad durante el siglo XX, Mexico: Fondo de Cultura
EconoÂmica.
Simpson, E.N. (1937) The Ejido: Mexico's Way Out, Chapel Hill: University of North
Carolina Press.
Smith, R.F. (1972) The United States and Revolutionary Nationalism in Mexico, 1916±32,
Chicago: The University of Chicago Press.
SolõÂs, L. (1971) La realidad econoÂmica mexicana: retrovisioÂn y perspectivas, Mexico: Siglo XXI.
Stallings, B. (1987) Banker to the Third World: US Portfolio Investment in Latin America, 1900±
Summerhill, W. (1997) `Transport Improvements and Economic Growth in Brazil and
Mexico', in S.H. Haber (ed), How Latin America Fell Behind, Stanford, Stanford University
Press.
Tannenbaum, F. (1968) The Mexican Agrarian Revolution, first pub. 1929, Hamden, Conn.:
Archon Books.
Taylor, W.B. (1985) `Between Global Process and Local Knowledge: An Inquiry into
Early Latin American Social History, 1500±1900', in O. Zunz (ed.), Reliving the Past.
The Worlds of Social History, Chapel Hill: University of North Carolina Press.
Turner, F.C. (1968) The Dynamic of Mexican Nationalism, Chapel Hill: University of North
Carolina Press.
Tutino, J. (1986) From Insurrection to Revolution in Mexico: Social Bases of Agrarian Violence,
1750±1940, Princeton: Princeton University Press.
Victoria, C.S.J. and VelaÂzquez RamõÂrez, S. (1988) Èl estado y las polõÂticas econoÂmicas en el
porfiriato', in C. Cardoso, (ed.), MeÂxico en el siglo XIX. Historia econoÂmica y de la estructura
social, Mexico: Nueva Imagen.
Vos, J. de, (1988) Oro verde. La conquista de la Selva Lacandona por los madereros tabasquen Ä os,
1822±1949, Mexico: Fondo de Cultura EconoÂmica.
Walker, D.W. (1986) Kinship, Business and Politics: The MartõÂnez del RõÂo Family in Mexico,
1823±67, Austin: University of Texas Press.
Wells, A. (1985) YucataÂn's Gilded Age: Haciendas, Henequen, and International Harvester,
1860±1915, Albuquerque: University of New Mexico Press.
Wilkie, J.W. (1970) The Mexican Revolution: Federal Expenditure and Social Change since
Wionczek, M. (1967) El nacionalismo mexicano y la inversioÂn extranjera, Mexico: Siglo
Veintiuno Editores.
Wolf, E.R. and Hansen, E.C. (1967) `Caudillo Politics: A Structural Analysis', Comparative
Studies in Society and History, vol. 9(1), January.
Womack, J. (1978) `The Mexican Economy During the Revolution, 1910±40: Historio-
graphy and Analysis', Marxist Perspectives, vol. 1(4).
Womack, J. Jr, (1968) Zapata and the Mexican Revolution, New York: Knopf.
Yates, P. L. (1981) Mexico's Agricultural Dilemma, Tucson: University of Arizona Press.
ZebaduÂa, E. (1994) Banqueros y revolucionarios: la soberanõÂa financiera de MeÂxico, 1914±29,
Mexico: El Colegio de MeÂxico.
6
Peru, 1884±1930: A Beggar Sitting on a
Bench of Gold?
Paulo Drinot
In the period 1884±1930 Peru was a multi-export economy. It thus differed

from other Latin American countries like Brazil, Colombia or the Central
American republics, whose insertion into the world economy in this period
occurred overwhelmingly through a single export commodity, coffee. Peru, like
Mexico or Argentina, could count on a series of export products: wool, rubber,
oil, sugar, cotton and metals.1 However, and perhaps to a greater extent than
Mexico or Argentina, these products were intimately linked to regional econ-
omies. Of course this had much to do with Peru's varied geography. Wool was
produced both in the central and southern sierra; however, most exported wool
came from the south. Although some sugar was grown in the ceja de selva,
overwhelmingly both sugar and cotton were coastal products. Rubber, natur-
ally, was limited to Peru's selva regions while oil was extracted from a few pozos
in the northern tip of the country. Although metal mining was a feature of the
northern, central and southern sierra, it became increasingly concentrated in
and dominated by production in the central sierra. The regionalized character
of Peru's natural resource, or more specifically, export commodity endowment,
inevitably translated into the development of distinct regional economies. In
short, Peru's natural regionalization results from its complex geographical char-
acteristics. However, the country's historical development has enhanced these
internal geographical borders. Indeed, thickly traced social, political and
economic borders have superimposed themselves on thinly traced physical
borders, thus making the physically surmountable often psychologically insur-
mountable. As such, the regionalization of Peru's export commodity endow-
ments inevitably led the period of export orientation to be inscribed within the
regionalized social and political structure of the country.
* I am grateful to all the participants at the conference held at Lake AtitlaÂn, Guatemala, for their
comments and suggestions on an earlier version of this chapter. Special thanks go to Shane Hunt,
Nelson Manrique, Rosemary Thorp and the editors who have commented extensively. The usual
caveats apply.
152
Paulo Drinot 153
After a brief glance at the background to the period under review, I begin by
studying the period from a macro or national perspective. This first section
examines both the exceptional circumstances arising in the 1890s, which
resulted in import substitution and industrialization and the potential for
autonomous development, and the failure to sustain this `model' of develop-
ment, linked to both endogenous and exogenous factors including interna-
tional price fluctuations and the penetration of foreign, especially US, capital.
The second section of the paper disaggregates the national economy into its
regional components. Here emphasis is placed on the effect of the export
economy at the regional level and its interaction with and impact on pre-
existing conditions. The end of the War of the Pacific, 1884, has been chosen
as the starting date, although as will become evident the multiple processes
under study both predate and postdate this particular year. The end year, 1930,
is self-explanatory, although again, many of the processes or themes that arise
in this period, impervious to academic periodization, naturally flow through to
the following period. Unlike Brazil or Argentina, in Peru the Depression did not
result in substantial structural change, or, perhaps more to the point, in desar-
rollo hacia adentro.
Peru's geography makes three regional spaces immediately obvious: the coast,
the sierra and the selva. However, the country's history adds further complexity
to this neat geographical triptych. For example, it is worth noting the coin-
cidence between the regions under study and distinct pre-Inca cultural areas
(Lumbreras, 1980). Similarly, the historical trajectory of colonial rule varied
substantially from one region to the next, leaving markedly different legacies
(Lockhart, 1974, ch. 2). The contrast between the central sierra, where the
colonial legacy was one of strong peasant communities and weak landlords,
and the south, where the opposite held, is instructive. The early republican
period saw further evidence of disarticulation. For the first two to three decades
of republican rule, a succession of civil wars were fought to determine the
country's political centre of gravity. In the end, coastal forces and interests
overcame southern interests. The coast's victory and the subsequent pacifica-
tion and consolidation of a Lima-based state rule under Castilla was in no small
way aided by the discovery of undreamed-of wealth piled up on islands off the
mainland (Gootenberg, 1989).
Between 1840 and 1879 Peru exported 11 to 12 million tons of guano, worth
some 750 million pesos (Bonilla, 1980). Hunt has argued that guano created a
rentier economy that, in combination with poor state policy, resulted in a lost
opportunity (Hunt, 1985). However, as Hunt himself shows, the guano prosper-
ity through its varied linkages to the national economy, some of greater benefit
than others, was central to Peru's mid-century revival. Indeed, guano revenues
contributed to the pacification of the country, to investment in ± largely
unneeded ± transport infrastructure, and to the creation of a modern, albeit
154 Peru, 1884±1930
highly inefficient, state (in nineteenth-century terms). Moreover, though alter-

native visions of development abounded in this period (Gootenberg, 1993),
guano set Peru on a firm path to free trade and export-led growth. Indeed,
arguably, guano revenues `made' the sugar and cotton export sectors; the
revenues were, for example used to indemnify coastal planters following the
abolition of slavery. As early as the 1860s, Peruvian cotton producers, bolstered
by new credit facilities ± essentially recycled guano revenues ± provided by both
local and foreign merchants and newly established banks, responded to the
opportunities created by the US Civil War by importing new technology and
increasing areas under cultivation (Bell, 1985, pp. 16±31). Meanwhile, sugar
producers invested heavily in their industry to meet the rising Chilean demand:
between 1866 and 1879, sugar exports rose from 5000 tons to 83500 tons. These
transformations, which according to some authors corresponded to a `national
project', were orchestrated by a new elite, organized around the Civilista Party
(Mc Evoy, 1994).
The War of the Pacific marked the end of the guano prosperity. The war's
effect on an already severely weakened economy was devastating: nitrates,
whose exports represented 26 per cent of total exports in 1878, were completely
lost to Chile (Hunt, 1985, p. 258); while guano ceased to play any major role in
the export quantum. The war's impact varied from one region to the next. It
was particularly destructive in the central sierra, occupied by the Chilean army
on three occasions and the backdrop to CaÂceres' war of resistance. The fighting
severely weakened the strongly collaborationist regional elite, who saw its
economic basis, livestock and mining, devastated by the war, and whose
haciendas were occupied by CaÂceres' peasant guerrillas (Manrique, 1988, p. 24).
By contrast, the south was virtually untouched by the war, and trade through
the port of Mollendo was only briefly altered.2 The disruption encouraged a
degree of import substitution. More important, and contrary to the experience
in the central sierra, the traditional power structure, gamonalismo, was left
untouched and arguably reinforced as the war left military power in the
hands of local gamonales. A similar scenario is found in Cajamarca and Chacha-
poyas in the northern Peruvian Andes, where the end of the war resulted in the
development of clan-based private armies which would dominate the political
arena until the 1920s (Taylor, 1986a; Nugent, 1997).
National overview, 1884±1930
Following the war, the CaÂceres government undertook the difficult task of
rebuilding the country's economic foundations. In 1889, the CaÂceres regime
signed the Grace contract which handed over Peru's railways and other
assets to British bondholders organized around the Peruvian Corporation,
created to run the railways, in exchange for the settlement of Peru's foreign
Paulo Drinot 155
debt and the restoration of its international creditworthiness (Miller, 1976 and
1983). Incipient institutions were created to integrate a country whose disunity
was singled out in contemporary debates as a major factor in Peru's wartime
defeat. In addition to identifying and mapping Peru's disputed international
boundaries, the members of the Sociedad GeograÂfica de Lima, founded in 1884,
set out to map the country's natural resource endowment, braving unexplored
Amazon rivers and the highest peaks of the Andes (LoÂpez-OcoÂn Cabrera, 1995).
With Pierola's defeat of CaÂceres in the Civil War of 1894±95, the Civilista party,
in government when the War of the Pacific broke out, returned to power in
alliance with an old foe. This event marked the beginning of a new era in
Peruvian history, known as the Aristocratic Republic. Although not as homo-
geneous or hegemonic as much of the literature indicates (see Miller, 1982), the
Civilistas promoted an export-oriented economy with minimal state interven-
tion and which, in practice, favoured coastal sugar and cotton planters. Produ-
cer associations like the Sociedad Nacional de Agricultura and the Sociedad
Nacional de Industrias, were created to fulfil a parastatal coordinating role and
to promote development.
It was in this context that an exceptional export boom occurred. As Tables 6.1
and 6.2 show, between 1895 and 1900 almost all export products experienced
strong increases in volume and value. In all, exports by value rose by 137 per
cent in this period. As Thorp and Bertram (1978) have shown, export growth
went hand in hand with economic diversification and industrialization. Several
factors combined to create such an outcome, including a highly favourable
depreciating exchange rate ± between 1890 and 1897 the depreciation was
nearly 40 per cent ± and the high reinvestment propensity of certain export
sectors, particularly coastal agriculture and mining. To the natural protection
afforded by the exchange rate was added a high import tariff policy, adopted
Table 6.1 Peru, growth rates of exports by volume, 1884±1930 (Five-year averages)
Sugar Cotton Wool Alpaca Silver Copper Rubber Oil
1884±1890 14 159 50 54 30 115

1890±1895 50 21 13 20 18 200 5
1895±1900 91 44 21 16 113 1 729 56
1900±1905 20 32 65 52 72 24 47 400
1905±1910 8 87 7 3 327 124 4 960
1910±1915 79 50 46 13 50 28 184
1915±1920 13 65 38 47 16 35 9
1920±1925 17 20 33 49 15 36 113
1925±1930 62 30 61 10 39 61 41
Source: Up to and including 1910 the estimations are based on Thorp and Bertram (1978), p. 330.
From 1911 to 1930, the estimations are based on Portocarrero, BeltraÂn and Romero (1992), pp. 131,
135, 142. All oil figures are taken from the latter source and are expressed in thousands of barrels.
156 Peru, 1884±1930
Table 6.2 Peru, growth rates of exports by value, 1884±1930 (five-year averages)
Sugar Cotton Wool Alpaca Silver Copper Rubber Oil Total
1884±1890 12 160 58 61 22 69 2
1890±1895 16 24 11 8 26 133 0.4 7
1895±1900 115 8 8 36 99 3 028 115 137
1900±1905 17 22 61 48 72 17 72 6 479 7
1905±1910 8 132 18 283 282 84 41 864 53
1910±1915 139 50 46 17 134 47 988 68
1915±1920 26 530 53 12 5 208 11 73
1920±1925 21 18 6 5 39 87 88 19
1925±1930 8 30 50 55 14 64 35 4
Source: Up to and including 1910 the estimations are based on Thorp and Bertram (1978), pp. 334±5.
From 1911 to 1930, the estimations are based on Portocarrero, BeltraÂn and Romero (1992), p. 133, 137,
142. All oil figures are taken from the latter source.
not for protection but as the only source of revenue available for a government
too weak to impose export duties on its export-oriented oligarchy.3 Indeed, in
1900 customs duties represented around 60 per cent of all tax income (Porto-
carrero, BeltraÂn and Romero, 1992, p. 120).
Elite investment groups fuelled diversification through a dynamic realloca-
tion of export revenues. Coastal planters, for example, strongly diversified their
business interests. In 1927, RamoÂn Aspillaga, owner of the CayaltõÂ plantation,
held 60 per cent of his investments in banks and insurance companies; both the
Pardo and Prado families established partnerships with Italian immigrant
industrialists to invest in the emerging textile industry;4 and in addition the
Prado family invested in electric and transport infrastructure (the Empresas
EleÂctricas Asociadas, which ran the Lima tramway system) and finance (the
Banco Popular, as well an insurance company) (Gilbert, 1982, pp. 127, 158±60;
Portocarrero, 1995; Quiroz, 1988; Portocarrero and TorrejoÂn, 1992).
As a result, the share of manufactured consumer goods in total imports fell
from 58 per cent in 1891±92 to 39 per cent in 1907. Further evidence of import
substitution is reflected in the rise of local production of cotton textiles as a
percentage of total supply from 5 per cent in 1891 to 47 per cent in 1908. By
1905, one survey claimed there were 173 urban manufacturing firms in Lima.
Similarly, Lima saw the development of large utilities companies and a financial
and banking sector. Moreover, the export boom had important indirect
demand effects on the local capital goods industry: until the early 1900s most
sugar mills as well as the smelters used in mining were produced by local
foundries (Thorp and Bertram, 1978, pp. 28±36).
Eventually, the diversification process was reversed with the adoption of the
gold standard in 1897, and the later introduction of export duties. The industrial
impulse of the 1890s was not sustained. Much of the industrial investment had
Paulo Drinot 157
been undertaken by agro-exporters; and as such their allegiance to an industrial-

ization project was fickle. In the context of favourable prices for exports and an
appreciating exchange rate, these entrepreneurs lost interest in their industrial
ventures. The limitations of the financial system soon became apparent. Banks
were anything but akin to strong Gershenkronian institutions; rather they were
cliquey and tended to restrict lending to their own shareholders. Broadly speak-
ing, the financial structure, composed of public, private and foreign finance,
proved inchoate and contributed little to growth (Quiroz, 1993).
Meanwhile, as a result of shifts in relative international prices, both copper
and oil increasingly came to dominate Peru's export quantum. Conversely,
cotton and in particular, sugar gradually lost importance within the export
quantum. As Table 6.3 shows, in 1920 coastal agriculture, sugar and cotton,
accounted for 72 per cent of the export quantum, with, by contrast, copper and
oil accounting for only 7 and 5 per cent of exports by value respectively. By
1925, however, copper and oil's share had grown to 32 per cent, and by 1930
this figure had risen to 40 per cent. Meanwhile, coastal agriculture's share had
fallen to 29 per cent. The relative price fluctuation experienced by these sectors
was central to this process. The 1920s were gloom years for coastal agriculture,
as prices experienced their well-documented precipitous declines. As Table 6.2
indicates, sugar exports by value grew by only 21 per cent between 1920 and
1925, and 8 per cent between 1925 and 1930. Cotton's fate was far more
dramatic. Despite positive growth rates in exports by volume, cotton exports
by value actually fell by 18 and 30 per cent in each period. By contrast, both
copper and oil exports exhibited strong growth rates: between 1920 and 1925,
copper and oil exports by value grew by 39 and 88 per cent respectively, while
in the period 1925 to 1930 these rates were 14 and 35 per cent. As foreign-
owned sectors, the overall spread effect of the export economy was reduced, a
process accentuated by inappropriate government policy as pointed out below.
Table 6.3 Peru, composition of exports by value, 1890±1930 (percentage shares)
sugar cotton wool silver copper rubber oil
1890 28 9 15 33 1 13 ±
1895 35 7 15 26 1 14 ±
1900 32 7 7 22 18 13 ±
1905 32 7 8 6 10 16 ±
1910 20 14 7 10 18 18 2
1915 26 11 5 5 17 5 10
1920 42 30 2 5 7 1 5
1925 11 32 4 10 8 1 24
1930 11 18 3 4 10 ± 30
Source: Thorp and Bertram (1978), p. 40.

158 Peru, 1884±1930
At the same time, Peru underwent a shift in external dependence as the USA
became its principal trading partner, a role played by Britain since the early
republican period (Portocarrero, BeltraÂn and Romero, 1992, pp. 152, 161).
Needless to say, both the copper and oil sectors were overwhelmingly con-
trolled by US capital.
Until the 1920s, the state played a marginal part in the processes outlined
above. The Peruvian state was small and weak, and its income depended
almost entirely on import taxes and taxes on tobacco, alcohol, salt and sugar.
For much of the period, export or income taxes were out of the question. By
the time export taxes were introduced, in the mid-late 1910s, international
commodity prices had begun their downward spiral. State expenditure was
limited: in 1914 only 210 000 and 100 000 Peruvian pounds were spent on
education and development respectively, whilst by comparison expenditure on
the army and the police amounted to 477 000 and 314 000 Peruvian pounds
respectively (Portocarrero, 1983, p. 25). The state was particularly weak in
physically outlying areas, such as the highlands of the Puno department,
Cajamarca and Chachapoyas in the north, or the selva, where power was priva-
tized in the hands of gamonales, local political bosses or rubber barons. Fiscal and
monetary policy and the rules of national accounting fit uneasily in these cases.
Taxation on rubber production was difficult if not impossible and the taxation of
exports highly inefficient. In the altiplano, most economic transactions were
carried out in Bolivian currency (Mitre, 1986, p. 51).
In the 1920s, government expenditure, and the general presence of the state,
increased significantly. Revenues increased nearly fourfold between 1919 and
1927, and the number of public employees increased over sixfold between
1920 and 1930 (Collier and Collier, 1991, p. 139). Most of the expenditure went
to finance irrigation works, road building, railways and the modernization and
expansion of Lima. For once, education received some attention: investment
doubled and enrolment rose significantly (Portocarrero and Oliart, 1989, pp.
58±9). These changes resulted in a significant increase in GDP (Seminario and
BeltraÂn, 1998). However, the expansion of the LeguõÂa state was largely financed
by foreign borrowing and subject to high degrees of corruption and fraud;5 local
investment was hardly encouraged, and credit systems remained inadequate. The
creation of the Peruvian Reserve Bank in 1922 did little to alter the bleak financial
situation, but, nevertheless, infrastructural investment, particularly in roads, led
to greater national economic integration and in particular to the demographic
expansion of Lima ± by 1931, nearly 40 per cent of the capital's population
originated in the provinces (Stein, 1980, pp. 56±7). Meanwhile, the growth in
the commercial sector and the state bureaucracy contributed to the emergence of
an urban middle class (Parker, 1998), which, along with the incipient working
class, constituted the basis for the emergence of new political alternatives, and
especially of the Alianza Popular Revolucionaria Americana (APRA) (Stein, 1980).
Paulo Drinot 159
The export economy in regional perspective
So far, I have presented an overview of the export economy in the period

and its interrelated effects on social and political processes. I now turn to
examine the export economies more closely, placing them within a regional
context.
The south and the wool trade

It could be argued that the wool trade played a negligible role in Peru's export
quantum, and more generally in its insertion into the world economy. As Table
6.3 shows, Peruvian wool never accounted for more than 15 per cent of the
export quantum. Peru produced low quality wool and in the period 1900±25 its
wool exports amounted to less than 1 per cent of world production (Jacobsen,
1993, p. 152). Similarly, the impact of wool at the regional level was limited;
although wool constituted the south's major export, it never monopolized the
regional export quantum. Moreover, wool production and its impact on the
local economies varied considerably within the southern area. In some places,
trade in alcohol products, either wine or aguardiente, outweighed the wool trade
(Manrique, 1997). The bulk of wool production originated in the department of
Puno and the highlands of the Cuzco and Arequipa departments. Although
much wool was oriented to the export market, a vast proportion remained
within indigenous commercial circuits. The wool economy created few formal
linkages, yet wool profoundly transformed the southern economy.
Wool production entailed control of both grazing land and sheep or alpacas.
There was little room for technological improvement, although some attempts
were made to improve flocks and rationalize production by importing better
breeds and through enclosure. There were essentially two types of wool produ-
cers: the hacienda and the independent Andean peasantry. Typically, haciendas
specialized in sheep wool production while peasants concentrated on alpaca
fibre, in which Peru held a virtual monopoly (Jacobsen, 1993, p. 161). In the
haciendas, labour tenants, known as colonos, undertook production; they
tended to their own flocks, known as huaccho, as well as the landowner's flocks.
As such, both within and outside haciendas, peasant production dominated
(Burga and Flores Galindo, 1994, pp. 68±9).
The wool economy was articulated around a complex regional commercial
system, which filtered the wool economy's important indirect effects on the
regional economy. This system linked the highland wool producers and the
Arequipa merchant houses which exported the wool through the port of Mol-
lendo (Miller, 1982c, p. 301; Burga and ReaÂtegui, 1981, p. 118). Between these
two actors there existed a series of intermediaries, either directly or indirectly
linked to the wool trade. The opening of the Southern Railway in 1873 altered
the structure of trade to a considerable extent.6 Commercial centres like Puno,
160 Peru, 1884±1930
Juliaca, Ayaviri and Sicuani developed around the railway line,7 and Arequipa
merchant houses were thus able to penetrate into the highlands through their
agents, undermining the role previously played by itinerant merchants and
muleteers. The replacement of annual fairs by Sunday fairs further intensified
the pace of the wool economy.
Merchants played a central role, as credit providers for both producers and
traders, in the development of the wool trade and the regional economy.
Indeed, the commercial houses' ventures into wool contributed to the monet-
ization of important sectors of the altiplano and generated a strong demand for
a variety of goods: by 1914±19, the Ricketts house's rural sales as a percentage of
total sales amounted to 50 per cent. Merchants helped articulate a regional
economy that extended beyond Peru's borders. Ricketts bought foodstuffs,
coca, cocoa, coffee and palillo in Cuzco and sold them in Iquique, Antofagasta
and the wool regions of Puno and even Bolivia. These regional products could
account for more than 20 per cent of income. The development of trading
centres along the railway line enhanced this process (Burga and ReaÂtegui, 1981,
pp. 32, 37; Jacobsen, 1988, p. 150).
The wool economy had important linkage effects at the regional level. Fol-
lowing the arrival of the Southern Railway in 1908, the city of Cuzco experi-
enced marked growth, including a degree of industrialization.8 Both the city's
growth and higher demand from wool producers increasingly drawn into the
market stimulated regional agriculture.9 Arequipa merchants were largely
responsible for much of the new investment; as well as providing imports of
all descriptions to the population of Cuzco, from basic tools to foie gras and
champagne, they played a central role in the commercialization of Cuzco's
products. Cuzco's growth, therefore, was a part of Arequipa's growing domin-
ance of the southern economy.10
However, at one level the boom had negative consequences. As wool prices
soared, demand for Cuzco products followed suit. The city of Cuzco was soon
facing drastic food shortages and became strongly demonetized. The sale of
metal currency, indispensable for wool transactions in the highlands, became
highly profitable. `Turco' merchants, minoristas of Arab origin, speculated in the
sale of coins to wool traders. Wool transactions in the highlands could only be
carried out in low denomination coins, usually Bolivian coins; and as a result
the high denomination Peruvian currency was strongly devalued locally.
Attempts first to limit and then stop the exodus of both currency and foodstuffs
failed. On 13 January 1918 Cuzco was shaken by popular protest. With the end
of the war in Europe prices began to drop (GuilleÂn MarroquõÂn, 1989, p. 53;
Burga and ReaÂtegui, 1981, p. 39).
Arequipa's main function in the wool trade was that of an entrepoÃt city. As the
obligatory stopping point between the highlands and the port of Mollendo, the
city inevitably grew. However, devoid of Cuzco's geographically determined
Paulo Drinot 161
natural protection, Arequipa did not industrialize to the same extent as the old
Inca capital. Nevertheless, some factories were erected,11 and Arequipa's growth
stimulated local agricultural production in the city's hinterland.12
The effect of the wool trade on the regional economy was unevenly distrib-
uted; whole departments appear to have been left behind. Ayacucho, stuck
between a south stimulated by wool exports and a central sierra under the
influence of copper, did not participate much in the regional trade flows and
therefore did not benefit from the indirect effects of the south's insertion into
the world economy. The city of Ayacucho exhibited the lowest population
growth rate of all sierra cities between 1876 and 1940. The result appears to
have been the strengthening of landlord power and an increased exploitation
of the peasantry (Gamarra, 1992, pp. 104±6; Degregori, 1990, pp. 29±31; Galdo
Gutierrez, 1992, pp. 107±8).
Perhaps the most important direct social consequence of the wool trade was
the usurpation of peasant land. This process has received the attention of many
scholars and need only be briefly sketched out here. Underlying this process
was gamonalismo.13 The gamonales, often intermediaries in the wool trade, used
local political bosses, or indeed became themselves political bosses, encroach-
ing on community lands, and thus helping to form the avalanche of hacienda
expansion that characterized this period. The number of haciendas, many of
which were created by newcomers, rose from 705 in 1876 to 3219 in 1915.
Correspondingly, the percentage of land held by peasant communities fell from
half during the early republican period to a third by the 1920s ( Jacobsen, 1988,
p. 152).
Pressures internal to the peasantry contributed to the encroachment process.
Jacobsen estimates that the population of communities may have increased by
as much as 50 to 60 per cent between 1876 and 1940; social differentiation
within the peasantry gradually undermined preexisting solidarity mechanisms;
and, finally, the peasantry's increasing insertion into regional commercial cir-
cuits resulted in greater vulnerability ± the possibility of withdrawing into self-
sufficiency in times of crisis was partially eroded. All three factors combined to
facilitate the transfers of land. Independent peasants were incorporated, as
colonos, into expanding haciendas.14 In the early 1920s, however, the avalanche
of hacienda expansion met a solid barrier: peasant rebellion.15
The end of the war in Europe precipitated a major drop in world commodity
prices. The price of Peruvian wool fell by 50 per cent in England and by 80 per
cent in Arequipa between its peak in 1918 and its nadir in 1921. Merchant
houses in Arequipa withdrew from the market, and at the same time, these
merchants acquired land as bankrupt landowners were forced to sell. The
region was plunged into crisis.16 Burga and ReaÂtegui (1981, pp. 34, 48±9) have
estimated that the peasantry's purchasing power fell by as much as 80 per
cent.17 As landowners' incomes from wool fell, the impetus to overexploit the
162 Peru, 1884±1930
haciendas's colonos grew and attempts to extract greater free labour services from
the peasantry broke the delicate balance which the patron±client relationship
was built on and created the conditions for peasant uprisings. By 1923, early
state tolerance and veiled support of peasant claims was replaced by active
repression, either through direct intervention or by allowing landowners to
respond with predictable violence (Kapsoli and ReaÂtegui, 1972; Flores Galindo,
1993b, pp. 365±9; Burga and Flores Galindo, 1994, pp. 177±200).
Some landowners tried to counteract the effect of falling prices. Encouraged
by the success of the sociedades ganaderas in the central sierra, some haciendas in
the south moved towards the introduction of higher yielding sheep varieties,
improvements in grazing and enclosures. However, attempts like the experi-
mental farm at Chuquibambilla were thwarted by the frequent and sometimes
systematic opposition from both the peasantry, for whom rationalization of
production meant the elimination of their huaccho flocks, and Arequipa mer-
chants. Similarly, attempts by the latter to enter production, as in the case of
the Gibson merchant house and the creation of the Sociedad Ganadera del Sur, in
association with Puno landowners, failed as a result of conflicting interests. The
merchants wanted to increase control over alpaca flocks for commercialization,
while the landowners were eager to venture into more capital-intensive sheep
wool production (Bertram, 1977, pp. 8±15).
In summary, the insertion of the south into the world economy produced a
commercial dynamic in the region. Merchant networks spread out of Arequipa
into wool-producing highlands and beyond into the fertile valleys and the ceja
de selva regions of Cuzco and the coastal valleys of Arequipa. Moreover, the
wool economy stimulated urban and industrial growth. These processes
resulted in increasing conflicts over land, aggravated by demographic growth,
and which, despite resistance, typically reduced the peasantry's share. More-
over, the growing dependence on a single export product increased the region's
vulnerability to shocks. Thus, when the price of wool collapsed in the 1920s,
the regional economy was plunged into crisis. Isolated attempts to rationalize
production and increase productivity were resisted by, on the one hand, a
peasantry for whom these changes represented further reductions in income
and, on the other, a regional elite uneasy about the prospects of increased
competition and change in general.
The decentralist movement of the early 1930s became the political expres-
sion of the changes undergone by the south. An alliance between the urban
middle class and progressive landowners, the decentralist movement was com-
posed of those groups most affected by falling wool prices in the 1920s, and for
whom the rising centralism of the LeguõÂa period was an obstacle to regional
autonomous development (Jacobsen, 1988, p. 167; Deustua and ReÂnique, 1984,
pp. 97±113). The decentralists' political programme was full of good intentions
vis-aÁ-vis the emerging proletariats of Cuzco and Arequipa and towards the
Paulo Drinot 163
Indian masses, however, unlike APRA the decentralists did not present them-
selves as an àlliance of intellectual and manual workers'. Their political base
was small and restricted to specific southern interests with little hope of expan-
sion, and although the movement actually held office during the brief Samanez
Ocampo government in 1931, it never really took power.
Coastal agriculture
The Peruvian coast is a long and narrow desert, and agriculture can only take
place in narrow valleys irrigated by rivers flowing from the Andes. Thus,
although both sugar and cotton in Peru benefit from particularly favourable
natural conditions, the successful exploitation of both products has tradition-
ally been conditional on overcoming the constraints imposed by water and
labour shortages.18
In Peru sugar is subject to substantial economies of scale and requires vast
quantities of water. Until the 1920s sugar planters showed little interest in
modern irrigation works, and irrigation laws were generally in their favour
(OreÂ, 1990, pp. 18±20). The modus operandi was to gain access to water rights
through land acquisition, thus unleashing a process of hacienda expansion
beginning in the 1850s. In the nineteenth century sugarcane had been grown
in most valleys in the coast. In our period, however, sugar production became
increasingly concentrated on the north coast, which by 1932 accounted for 83
per cent of production. It was on these plantations, concentrated in the valleys
of La Libertad and Lambayeque departments, that newcomers to the sugar
industry initiated a process of modernization in the 1860s and 1870s. These
men, and some women, often widows, had usually made their fortunes in the
guano and nitrates booms, or were merchants and financiers. Some of these
new planters were recent migrants, like the Italian Larco family, but overall
foreign ownership in the sector was limited.19 Steam ploughs and railways were
imported, as was sugar-making machinery which increased processing capacity
and raised productivity. For the most part, however, milling remained anti-
quated (Albert, 1976, p. 67a). Nevertheless, installed capacity increased just in
time to meet growing international demand.
In particular, two foreign events favoured the development of the industry. In
1893, Chile gave preferential treatment to imports of unrefined raw sugar to
protect its refining industry, and at the same time Peruvian sugar exports did
particularly well during the Cuban Revolution of 1895 and the subsequent
Spanish±American War, following which the USA became a major market for
the first time.20 As Table 6.2 indicates, sugar exports by value rose by 115 per
cent between 1895 and 1900. In the 1900s, however, the adoption of the gold
standard and changing relative prices spelled a downturn for the sugar indus-
try.21 By contrast, the First World War years marked a new boom period for
sugar. As a result of significant investments from 1908, Peru's sugar sector was
164 Peru, 1884±1930
able to respond to the rise in international prices,22 and export growth reached
139 per cent in the period 1910±15.23 The 1920s, again, was a decade of
stagnation. Peru's competitors had also responded to the rising prices in the
mid±late 1910s. By the 1920s, the beet sugar producers had recovered. In the
central and southern valleys, many planters switched to cotton, which until
1925 continued to benefit from relatively good prices as a result of a boll-weevil
plague in the USA in 1921. In the north, where switching to cotton was not
economically feasible, planters attempted to reduce costs to a minimum and
boost production to compensate for falling prices. Smaller mills closed. One
major plantation, Roma, went bankrupt and was bought by another, Casa
Grande. Wage cuts led to industrial unrest: in 1921 a strike at the Roma
plantation resulted in violent repression from a state still shaken by the coun-
try's first general strike two years earlier (Blanchard, 1982, pp. 126±35; Burga
and Flores Galindo, 1994, pp. 235±50).
Indeed, labour played a critical in role in shaping the Peruvian sugar industry.
The abolition of slavery did not itself create labour shortages: sugar planters
complained of `lack of hands' well before emancipation. It is true that many
slaves moved to the cities, particularly Lima, and by 1876 blacks accounted for
9.3 per cent of Lima's population (Cuche, 1975, p. 34). But an emancipated
slave population could hardly have provided the labour to fuel, a sugar export
boom in the second half of the nineteenth century. Independent peasant
communities on the coast could have provided the labour and some were
indeed incorporated via the process of hacienda expansion. However, on the
whole, the coastal peasantry was reluctant and more often than not could not
be forced to leave its land to work in the sugar or cotton plantations (Bell, 1985,
pp. 41±2; GoÂmez and Bazan, 1981, pp. 53, 55). Eventually, the devastation of
the War of the Pacific, deteriorating economic conditions, and particularly the
growth of the internal market and a growing internal differentiation would
draw the peasantry into the export industries. In the 1850s and 1860s, however,
labour was drawn from an unexpected source: China. Between 1849 and 1874,
when British pressure at Macao put an end to the trade, some 90 000 coolies
were brought to Peru. Some were put to work on the guano islands, but the
bulk were destined for the sugar and cotton plantations on the Peruvian coast.
Yet, as with slavery (to which coolie labour bore a close resemblance in terms of
work conditions), the ending of the trade effectively sounded the death knell of
coolie labour. Indentured Chinese labourers had to work off debts over a period
of eight years. Though planters tried a number of strategies to keep the Chinese
on the plantations after the eight years, typically the Chinese left. During the
War of the Pacific, many Chinese fled the plantations, and in some cases
collaborated with the Chilean army (RodrõÂguez Pastor, pp. 95±111; Peloso, pp.
26, 39, Stewart, 1951). In the 1890s, Japanese workers were contracted. Again,
however, this type of labour proved unsuitable to the planters' needs. Like the
Paulo Drinot 165
Chinese, the Japanese left the plantations as soon as they fulfilled their con-
tracts. Moreover, planters complained that the Japanese could only perform
light work. Unaccustomed, or perhaps unwilling to accept the harsh treatment
meted out by plantation overseers, the regular protests by Japanese workers
sometimes escalated into full-blown strikes (Gardiner, 1975; Morimoto, 1979,
Fukumoto, 1997, ch. 3 and 4).
From the 1880s onwards workers from the northern sierra, particularly from
the departments of Cajamarca and the sierra of La Libertad provided the bulk of
labour on sugar estates. A form of labour contracting, known as enganche, was
adopted as the mechanism to draw labour to the coast. Contratistas (often local
merchants or muleteers), hired by the estates, provided peasants with cash
advances in exchange for labour on the coast. Though enganche is often por-
trayed as a demoniac system of oppression, it represented a response to the
needs of both planters and the sierra peasantry. For the peasants, enganche
labour on the coastal sugar plantations represented a seasonal income adjust-
ment. Wages were high by comparison; and the seasonal work could be carried
out between harvests. Moreover, many used enganche to escape from political
turmoil and banditry, conscription into the army, debts or the law. From the
planters' point of view, enganche had a number of advantages, too, not least of
which was the low militancy among workers whose primary interests remained
elsewhere (Gonzales, 1985, pp. 84±95, 118±23; GoÂmez and Bazan, 1981, pp. 53,
59±60; Albert, 1984, pp. 102±11; Blanchard, 1979; KlareÂn, 1973, ch. 2).
Enganche is best understood as a byproduct of the long-established commer-
cial circuits linking the north coast and the northern sierra. From the 1860s
onwards, the coastal economy had had a strong demand effect on two types of
sierra goods: pastoral (cattle) and human (labour). Sierra landowners responded
to the first demand vigorously. Despite poor transport infrastructure, a strong
commercial nexus was established, articulated around a network of merchants
and muleteers. Cattle exports were soon on the rise, and attempts were made to
expand properties and capitalize production, especially through the introduc-
tion of better breeds (Taylor, 1983, pp. 13±17). The local merchants were often
representatives of coastal commercial houses, and typically would export local
foodstuffs, cattle products and some artisan goods to the coast, and import salt,
rice, sugar, agricultural tools and imported textiles. Moreover, some merchants
were also involved in the commercialization of Cajamarca's ceja de selva pro-
ducts, including coca, cocoa, coffee, cotton, rice and aguardiente (RodrõÂguez
Doig, 1986, pp. 24±5; Puga de Losada, 1893, p. 126).
Between the late nineteenth century and the 1940s, haciendas held around 60
to 70 per cent of the land in Cajamarca. However, over 75 per cent of the rural
population lived from subsistence agriculture on its own land. The peasantry
was highly mobile and strongly inserted in the market through artisan produc-
tion. Landowner attempts to access the labour of this free peasantry, either
166 Peru, 1884±1930
through force or otherwise, often resulted in failure (Taylor, 1986b, p. 18;

Taylor, 1983, pp. 11±12; RodrõÂguez Doig, 1986, p. 21; Puga de Lozada, 1893,
p. 127). However, the peasantry was subject to severe demographic pressures:
between 1876 and 1940, average density rose from 11.4 to 23.1 inhabitants/sq
km. With less land to go round, and therefore less capacity to engage in
subsistence agriculture, the peasantry was forced to enter the market to an
even greater extent than before. Coastal plantation work became necessary to
supplement the income of the peasant household. Initially, non-economic
coercion by enganchadores was necessary to establish the practice of seasonal
migration, but as the process of alienation from the land accelerated, peasants
came to rely more and more on coastal work. By the 1920s, despite a high
susceptibility to coastal diseases (Cueto, 1992), growing numbers of workers
were migrating of their own volition without an advance. Indeed by 1924 there
were more `free', proletarianized workers than enganchados in the CayaltõÂ plan-
tation in Lambayeque. Many had severed their ties to their place of origin in the
sierra (Gonzales, 1985, p. 135; Deere, 1990, p. 46). As such, the sugar sector
gradually acquired, both through technological investment and labour arrange-
ments, a distinctly `modern' and industrial character.
The regional impact of sugar production varied from one place to the next. By
1927, the Gildemeister family's Casa Grande plantation controlled almost all
sugar production in the Chicama Valley in La Libertad department. Here, both
medium-sized haciendas and peasant subsistence plots were eaten up by the
sugar plantation. Casa Grande's high degree of vertical integration reduced and
eventually eliminated potential regional linkages.24 In addition, in 1915 the
Gildemeisters were given a concession on the port of Malabrigo by a govern-
ment hungry for cash. Casa Grande was thus able to import both consumer and
capital goods without paying import duties. Moreover, it soon established a
general bazaar on the plantation, which sold imported goods far below market
prices to the plantation population and later to local merchants and villagers
from the highlands. In time, the bazaar undermined traditional commercial
circuits and as a result the regional economy of the Chicama Valley and the city
of Trujillo went into decline (KlareÂn, 1973, ch. 4; Beals, 1934, pp. 200±1).
Arguably, these processes, including the emergence of a proletariat in the
valley's sugar estates, constituted a fertile ground on which radical political
projects and more specifically APRA could develop (KlareÂn, 1973, ch. 8).
By contrast, in Lambayeque these processes were less pronounced, and
indeed opposite. Unlike Trujillo, Chiclayo, whose population nearly doubled
between 1876 and 1931, underwent something of a commercial boom, until
the 1920s at least. Artisan industry increased significantly (GoÂmez and Bazan,
1989, p. 140), and, moreover, local domestic use agriculture benefited from the
changed economic climate. The greater availability of credit lowered the entry
barrier into rice production, which in the nineteenth century had been limited
Paulo Drinot 167
to haciendas. In the twentieth century, the local peasantry, with credit from
local mill-owners, was able to engage in production for both regional and
national markets. By 1918 rice had become the principal peasant crop, benefit-
ing from a high import tariff (GoÂmez and BazaÂn, 1981, pp. 36±45). A similar
process occurred in the Jequetepeque Valley (Burga, 1976, pp. 189±95).
On several counts, cotton production varied substantially from sugar. First,
cotton was less circumscribed to a single area, with production taking place in
Piura, the Sur Chico ± Chincha, Ica and Pisco ± and from the 1900s onwards the
ex-sugar producing central valleys of Can Äete and Chancay (Thorp and Bertram,
1978, pp. 44, 57; Albert, 1976, p. 138a; Peloso, 1999, ch. 3; Arroyo, 1981, p. 112;
Pachas Castilla, 1976). Moreover, land tenure in cotton agriculture, which
requires less water and is not subject to economies of scale, has been more
varied, especially during the period under review. Finally, although enganche
labour was a feature of cotton agriculture, it did not predominate nor did it
lead, like in the sugar sector, to the formation of a rural proletariat. Rather,
cotton planters turned to sharecropping, or yanaconaje, in the 1870s. As enter-
ing a sharecropping agreement did not substantially alter the living and work-
ing conditions of the coastal smallholding peasants, many were easily drawn to
cotton cultivation. For the planter, yanaconaje, combined typically with sea-
sonal wage labour, provided a series of benefits including the sharing of risks
and costs of production, and low militancy (Matos Mar, 1976, pp. 40±51; Bell,
1985, p. 45; Peloso, 1999, ch. 6).
In the 1890s and particularly the 1900s, the cotton export boom had less to
do with a disruption to a major world producer, as it had in the 1860s, and more
with the particular qualities of Peruvian cotton, better suited to the fine spin-
ning techniques developed in the 1890s (Bell, 1985, p. 36). In the 1900s,
production and exports rose dramatically as area under cotton increased and
Peru's markets diversified. Much of the increase in production resulted from a
switch to cotton, either from sugar in the central valleys, or vine in Pisco and
maize in Chancay (Bell, 1985, pp. 37±9). As Table 6.1 indicates, exports by
volume rose by 32 per cent and 87 per cent in 1900±05 and 1905±10 respect-
ively. Favourable prices translated these increases into a 132 per cent rise in
exports by value in the period 1905±10. Irrigation projects, on the whole
limited to Piura, did not greatly increase production. On the other hand, the
installation of modern cotton gins in the 1890s appears to have made a more
substantial contribution. Although Britain remained Peru's top market, by
1912±14 Peru had become the USA's fourth major source of supply. In addition,
in the 1890s cotton benefited from the development of a Lima textile industry
and the emergence of markets for its byproducts like cotton seed oil, used for
soap and candle manufacturing, and cotton seed cake used for cattle-feed.
Unlike the sugar sector, cotton offered few avenues for mechanization,
although some innovation did occur. During the First World War, for example,
168 Peru, 1884±1930
some planters switched to Mitafifi cotton which was used for heavy tyres
and aeroplane fabric. More important was the development of a new variety,
introduced in 1912 by a planter from Pisco called FermõÂn TanguõÈs, a Spaniard
born in Puerto Rico, who revolutionized the Peruvian cotton industry. TanguõÈs
cotton was both wilt-resistant and yielded 25 per cent more lint than other
varieties. In 1918 it accounted for eight per cent of Peru's total cotton produc-
tion. By 1931, the figure stood at 91 per cent. In Piura, where TanguõÈs did not
adapt, the introduction of the Pima variety from Arizona had similar con-
sequences. Pima accounted for 5 per cent of Piura's cotton exports in 1925
and 98 per cent in 1930, thus completely displacing the indigenous Aspero
cotton (Cueto and Lossio, 1999; Low, 1979, pp. 146, 148; Thorp and Bertram,
1978, p. 58).
Immigrants like TanguõÈs played an important role in the success of the cotton
sector. Calixto Romero, a Spaniard, reinvested much of the wealth accumulated
in the production and commercialization of Panama hats in the Catacaos
area of Piura into cotton cultivation (Rean Äo and VaÂsquez, 1988, ch. 2 and 3).
In the area around Lima, Italian merchants and petty traders diversified into
cotton, often entering the sector through reclamation of debts (Bonfiglio, 1995,
pp. 65±6). In Chancay, the Japanese came to dominate cultivation both as
landowners and yanacones (Matos Mar, 1976, pp. 82±102; Fukumoto, 1997,
pp. 203±7).
In contrast to the diversification in production, foreign merchant houses like
Duncan Fox, Graham Rowe and Locketts were firmly in control of commercial-
ization. Moreover, the merchants exercised significant influence over the pro-
duction process through their control of the credit system.25 Ginning, too, was
increasingly dominated by these merchants, although until the 1930s most
large haciendas continued to gin their own cotton (Albert, 1976, p. 159a;
Thorp and Bertram, 1978, p. 52).
Like sugar, cotton experienced the boom and bust effect of the First World
War. As Tables 6.2 and 6.3 indicate, cotton exports by value rose by 530 per cent
between 1915 and 1920, while cotton's share within the export quantum rose
from 11 to 30 per cent in the same period. Acreage under cotton grew, as did
productive units and employment in the sector (Albert, 1985, p. 239; Low,
1979, p. 120). Following the introduction of a sliding-scale cotton export tax
in 1915, the cotton sector became a major source of government revenue: by
1919 the tax yielded 21.9 per cent of total government revenue. In the 1920s,
the bust was delayed by the effect on world prices of a boll-weevil plague that
wiped out much of the US crop. Between 1920 and 1925 exports by value only
fell from some US$18 million to US$15 million; but by 1930 they had fallen to
US$ 10 million. Indeed, the recovery of the US cotton crop in 1925 coincided
with floods which wiped out the Peruvian crop. With rock-bottom world prices,
the cotton sector fell into depression, although it would recover in the 1930s.
Paulo Drinot 169
Oil
In the 1900 to 1930 period, oil production was limited to a small area on the far
northern coast.26 Although developed initially in the 1860s, it was not until the
first decade of the twentieth century that the industry really took off. In 1904,
internal demand suddenly rose as both the Central Railway and the Peruvian
Steamship Company switched from coal to oil. At the same time international
demand was rising; the industry expanded rapidly and by the late 1920s oil had
become Peru's leading export.
Oil was produced on three oilfields, Negritos, Lobitos and Zorritos. However,
the bulk of production originated in Negritos, which was owned from 1913
onwards by the International Petroleum Company (IPC), a subsidiary of
New Jersey Standard Oil. Indeed, Negritos' output as a percentage of total
output rose from 59 per cent in 1890 to 63 per cent in the 1900s, 80 per cent
in the 1920s and 83 per cent in the 1930s. Both Lobitos and Zorritos were
forced to enter into working alliances with the IPC. Lobitos, owned by a
British company, exported all its oil through the IPC's refinery at Talara. By
the 1920s exports from Zorritos, owned and exploited by the immigrant
Faustino Piaggio, were insignificant. Rather, it concentrated on satisfying a
growing local demand made highly profitable by a price fixing agreement
with the IPC.
Oil created almost no linkages to the national economy. Indeed,
for the IPC the profits on its sales of products within Peru were sufficient to
meet the entire local-currency cost of production, with the result that the
IPC, despite its role as the single leading export enterprise of the 1920s,
played virtually no part in the actual supply of foreign exchange to the
market.
(Thorp and Bertram, 1978, p. 103)
Moreover, the lack of linkages was a direct result of state policies. In 1922 the
new Petroleum Law imposed a 10 per cent royalty on crude oil output. How-
ever, following negotiations the IPC was exempted from payment of royalties
and a series of land and export taxes in exchange for US$1 million in cash and
assistance in the placing of US$2.5 million in petroleum bonds in the USA. In
other words, the LeguõÂa government exchanged the potential benefits of devel-
oping an oil industry for short-term financial support.27 Such short-termism
was not irrational but corresponded to the time-horizon of Peruvian decision-
makers (Thorp and Bertram, 1978, p. 110).
Given the lack of empirical studies, it is difficult to establish the oil industry's
impact at the local and regional level. In many respects the oilfields were
typical enclaves; workers drawn from nearby villages are likely to have spent
most of their wages within the oil camps. Their typical demands during strikes
170 Peru, 1884±1930
put as much stress on improving living conditions as working conditions

(Davila Apolo, 1976). Indeed, as the mayor of the nearby port of Paita com-
plained in 1935, the oil fields had become commercial and industrial centres
that drained resources, especially labour, from nearby towns and villages and
contributed to their demise (Concejo Provincial de Paita, 1935).
Thorp and Bertram have estimated that up to 83 per cent of the IPC's export
earnings in the period 1916±34 represented clear profit. They calculate that, in
this same period, an identical national firm
would have yielded returned value of US$350 million compared to IPC's

actual US$63 million; or, subtracting the income from local market sales,
US$285 million compared to IPC's zero. In either case, the net gain would
have been US$285 million approximately.
Not surprisingly, they conclude: Òn any reasonable assumptions, Peru would
have been better off without the IPC' (Thorp and Bertram, 1978, p. 107).
Rubber and the selva

For most of Peru's colonial and republican history the selva had remained an
unknown quantity. Following the rebellion of Santos Atahualpa in the mid-
eighteenth century, the selva was sealed to outside intervention for a century or
more. The republican period saw a renewed interest in the selva. Adventurers in
search of the elusive El Dorado or Paititi had fuelled the colonial experience.
Overwhelmingly, however, it had been an evangelical mission. Republican Peru
had other concerns. In 1848, President Castilla sent a mission to the Chancha-
mayo region to `recruit' workers for the coast. The mission was turned back by
Indians of the PereneÂ. However, it was only a matter of time before the selva's
hermetism was pierced by nineteenth-century frontierism. The selva now
became a region to be colonized and made productive through the introduc-
tion of haciendas. In 1873, President Prado passed a new immigration law to
encourage colonization, and the legislation was preceded by several `pacifying'
military missions to the selva. Thus precedents were set for the violence which
resulted in the extermination of entire Indian groups and which became an
integral part of the rationale of rubber exploitation.28
Prior to the 1880s the economy of the selva had been dominated by hunting
and gathering activities. Commercial activity was limited to products like
sarsaparilla, a vegetable product used for skin and blood diseases, and other
medicinal plants. In Moyobamba there existed a thriving straw-hat industry
which fed the Brazilian market, as did most products. However, these activities,
along with fishing, amounted to a very small regional economy (Roux, 1994,
pp. 97±102). Most trade was conducted as barter or through the use of pseudo-
currencies like tocuyos, a rough cloth. Moyobamba, the only significant
Paulo Drinot 171
commercial centre, channelled trade with the Loreto region and Brazil as well
the ceja de selva regions of Huanuco, Huaraz, Pataz and Chachapoyas. However,
trade was limited by prohibitive transport costs (Roux, 1994, p. 157).
Rubber was first used in the manufacture of raincoats and rubber shoes in the
1830s. The economic prospects of the material were transformed in 1839 by the
discovery of the vulcanization process, which made rubber resistant to tem-
perature changes. From this date onwards, world demand for rubber increased
steadily with consumption in Europe and the USA rising from 400 tons in
1850 to between 50 000 and 60 000 tons in 1900 (Thorp and Bertram, 1978,
p. 67). The Amazon area was soon producing around 50 per cent of world
rubber; and the rapid rise in prices generated a rubber-rush, not unlike its
more famous (but less infamous) gold-rush cousins (Roux, 1994, p. 104; Pen-
nano, 1988, p. 110).
As prices rose the logic of rubber changed. Before the 1890s, local and foreign
entrepreneurs had exploited rubber in an artisan fashion. Increasingly, rubber
production came to be dominated by a handful of rubber barons like the
notorious Fitzcarrald and Arana (Pennano, 1988, pp. 157±68). By the 1890s
foreign merchant houses, often with the backing of international banks, had
taken control of commercialization. Finally, the 1900s saw the arrival of large
foreign companies, which attempted to take over at the level of production
(Roux, 1994, p. 114). As Table 6.2 indicates, the changes in the production
process resulted in substantial increases in exports: rubber exports by value rose
115 and 72 per cent in the periods 1900±05 and 1905±10, respectively. In 1898,
in an attempt to regulate and, more important, to tap the benefits from the
boom, the government passed a law granting land concessions to rubber tap-
pers in exchange for a small lease and a royalty on the rubber produced (Thorp
and Bertram, 1978, p. 69). In 1902 the tax amounted to two soles per hectare
(Roux, 1994, p. 115). Taxation at the level of production was made necessary by
the difficulties in taxing the export of rubber. Pierola had introduced export
duties on rubber in 1896, but the region's isolation as well as local officials' high
propensity for corruption made the enforcement of these taxes extremely
difficult and contraband highly enticing: in 1899 only one-third of total rubber
exports were taxed (Roux, 1994, p. 107). A few companies ± like the Inca Rubber
Company and the Peruvian Amazon Company ± took advantage of the new
law. However, these enterprises proved incapable, as Weinstein has suggested,
of introducing a capitalist mode of production in the selva, and breaking the
patron-client relations on which rubber exploitation was based (Weinstein,
1983, pp. 131±2).
The rubber boom gave birth to a highly cosmopolitan urban centre in the
middle of the jungle, Iquitos (San RomaÂn, 1994, p. 150), whose population in
the 1900s stood at around 20 000 (Thorp and Bertram, 1978, p. 68). Iquitos's
high import propensity and opulence is reflected in the level of import tax
172 Peru, 1884±1930
revenues which in 1910 stood at 217 600 Peruvian pounds, up from 167 373 the
previous year (Pennano, 1988, p. 204).
With the rubber boom, a complex commercial circuit was quickly put in
place. The commercialization of rubber occurred at three separate levels. The
first or highest level was that which linked the merchant houses in Iquitos,
Manaus and Belem and the European and US markets. In 1906, 59 per cent of
the rubber exported through Iquitos went to Britain, 37 per cent to France and 3
per cent to the USA. By 1914, however, exports to the USA had overtaken
exports to the European continent (Bonilla, 1977, p. 130; Pennano, 1988,
p. 188). The second level linked the merchant houses and the rubber entrepre-
neurs through the system of aviamiento, whereby the merchant provided the
entrepreneur tools and foodstuffs in exchange for rubber. Finally, the third
linked the rubber patron and the peon, either mestizo enganchados or semi-
enslaved Indians, through the stores owned and run by the rubber entrepren-
eurs, where rubber and goods were exchanged (Roux, 1994, p. 110; Weinstein,
1983, p. 134). At all these junctures, exorbitant profits were made. One final
agent was the itinerant trader, often of Armenian or Jewish extraction, who
undercut the rubber patron's monopoly sales to the peons. For most of the
period trade was funnelled through Iquitos and then the two Brazilian ports,
Manaus and Belem. Peru's failure to develop a merchant fleet in the Amazon
handed the control of river transport to Brazil. The latter's monopoly was only
broken when rubber production expanded to the south-eastern regions of
Ucayali and Madre de Dios, where articulation occurred with the neighbouring
regions of Cuzco and Puno through the Southern Railway.29
Although the bulk of rubber production was undertaken in the Loreto region,
a proportion was also harvested in the ceja de selva regions of the northern,
central and southern sierra and in what became, in 1913, the Madre de Dios
department.30 In 1908 the government ceded selva lands to two foreign com-
panies interested in exploiting the rubber in the jungle regions of the Cuzco
department. The demand for foodstuffs from these and other `plantations'
stimulated foodstuff production in the north of La ConvencioÂn province
(GuilleÂn MarroquõÂn, 1989, pp. 32±3). This coincided with Cuzco's òrientalist'
period when, frustrated by its isolation from the rest of the country, the old Inca
capital turned its attention to the east and the possibility of an Amazonian link
to the Atlantic. The construction of a railway to a navigable point on the
Urubamba River was subject to pressure from the aguardiente producers of
La ConvencioÂn province. The narrow-gauge railway reached Macchu Picchu
in 1930 and La ConvencioÂn's capital, Quillabamba, only recently (Tamayo
Herrera, 1981, pp. 100±1).
The rubber boom ended almost as abruptly as it had started. In the 1880s
Britain established rubber plantations in its Asian protectorates of Ceylon,
Malaysia and Indonesia. Stimulated by rising prices, Asian plantation rubber
Paulo Drinot 173
production rose sharply. By 1910 these plantations were exporting over 8000
tons annually (Thorp and Bertram, 1978, p. 68). Prices plummeted (Pennano,
1998, p. 110), and the higher costs involved in wild-rubber production had
made much of Peruvian rubber uncompetitive by 1912. Plantation rubber came
to dominate the world market: thus, while at the beginning of the 1910s the
Amazon produced over half of the world's supply, by the end of the decade it
produced less than one-tenth (Weinstein, 1983, p. 132). Peru's share of world
rubber production fell from 2.4 per cent in 1910 to 0.05 per cent in 1925
(Pennano, 1988, pp. 122±30). The First World War generated a short-lived
boom, but the industry's days were numbered. As Table 6.2 indicates, rubber
exports by value fell by 87 per cent in 1920±25 and 64 per cent in 1925±30. A
few alternate industries were developed; wood exports tripled between 1923
and 1935, and coffee, cotton and tobacco began to be grown by colonos, or
settlers (San RomaÂn, 1994, pp. 180±2). Politically, the decline of the region
found expression in a failed separatist rebellion in 1922 in Iquitos (Pennano,
1988, p. 168; Basadre, 1983, vol IX, pp. 274±6).
The central sierra and copper

By its geography and history, the central sierra has traditionally been one of the
most integrated regions in Peru. The seventeenth and eighteenth centuries had
seen the development of a strong regional economy flanked by two mining
poles, Huancavelica and Cerro de Pasco, which together exerted a magnetic
attraction on the region's resources drawing both foodstuffs and labour. With
the decline of Huancavelica in the late eighteenth century, the economy of the
central sierra became more closely linked to the Cerro de Pasco and HuarochirõÂ
silver mines (Manrique, 1987, p. 252).
Independence extracted a strong toll on the region. The mining sector
went into prolonged decline and came to be dominated by small-scale opera-
tions typically conducted by village entrepreneurs. Attention switched to regio-
nal agriculture. Over time, the central sierra had evolved an original system of
land tenure that had resulted in the fertile valleys being dominated by inde-
pendent peasant communities while the haciendas were relegated to the sur-
rounding highland plateaux. Thus, in 1876, 27 per cent of the rural population
lived in haciendas whereas in the central sierra the figure was 13 per cent. The
independent communities achieved a high level of commercialization; the vast
majority of goods consumed by the mining industry, for example, were region-
ally produced. The only exceptions were dynamite, which came from Lima,
some mercury, from the USA and Spain, and certain luxury goods for the local
elites which came from Lima or were imported from abroad. Contreras esti-
mates that regional trade accounted for up to 50 per cent of all Cerro de Pasco
trade. Moreover the production of regional goods was subject to local special-
ization.31 All this made for a highly commercial region,32 where, perhaps
174 Peru, 1884±1930
surprisingly, was a regional society where the balance of power between

haciendas and Indian communities was not completely stacked against the
latter. As argued above, this situation was heightened by the War of the Pacific.
Rising international prices resulting from the development of the electricity
industry created a growing interest in copper production. The massive cut in
total production costs following the construction of the Central Railway link-
ing Callao and La Oroya, completed in 1893, catapulted local interest in the
metal. Copper, unlike silver, could be profitable only if mined in very large
quantities. While silver production was carried out in an essentially artisan
manner, copper could only be produced industrially and as such required
large infrastructural investment. Although silver and gold enjoyed brief
booms in the 1880s and early 1890s, by 1897 the central sierra had undergone
a general shift to copper production. Significant local investment followed,
stimulated by a 1890 mining law which exempted the industry from taxation
for a quarter century, and the establishment of a National Corps of Mining
Engineers which undertook surveys of the subsoil wealth of the country. In
1900 a syndicate of Lima capitalists, the Empresa Socavonera, obtained a con-
cession from the government to build a drainage tunnel in Cerro de Pasco to
allow access to higher ores lying below the water table. By 1903, copper produc-
tion was just under 10 000 tons (Werlich, 1978, p. 121; Thorp and Bertram,
1978, pp. 73±9; Yepes del Castillo, 1972, pp. 147±9).
The Cerro de Pasco Corporation (Cerro) born out of the USA mining syndi-
cate which, in 1902 bought 80 per cent of mining claims in the Cerro de Pasco
area, undertook further investment. Cerro closed down the mines for five years
and carried out a massive infrastructural investment programme totalling some
US$25 million.33 It bought 2000 acres of coal properties and erected two hydro-
electric plants close to La Oroya. In 1908, Cerro bought out the Empresa
Socavonera. In 1905±07 most of the mines at Morococha were bought, and in
1919 the company took over the Backus and Johnson smelter and other inter-
ests at Casapalca. As well as investing in integrating copper production, Cerro
invested heavily at the level of production itself (O'Brien, 1996, p. 115; Mallon,
1983, p. 171). Some authors have characterized Cerro as a quintessential
enclave economy (Kruijt and Vellinga, 1979, p. 31); whilst by contrast, Thorp
and Bertram estimate that Cerro's returned value was as high as 55 per cent. At
the regional level, the impact of Cerro is undeniable. The livestock sector,
virtually destroyed by the War of the Pacific, remained depressed until the
1900s. However, in that decade it was revived and transformed in response to
the new demand created by Cerro. Between 1905 and 1910, local landowners,
in association with Lima capitalists, formed livestock associations, joint-stock
companies aimed at rationalizing and expanding production by bringing separ-
ate properties under a single administrative regime (Mallon, 1983, pp. 174±5;
Manrique, 1987, pp. 245±61). Meanwhile, colonization of the ceja de selva
Paulo Drinot 175
region brought new lands into production that specialized in sub-tropical

products, particularly coffee (Mallon, 1983, pp. 137±8; Barclay, 1989).
At the village level, the impact of Cerro resulted in significant social differ-
entiation within the local peasantry, already squeezed by population growth,
and to the emergence of a kulak class which doubled up as regional merchants
and sub-enganchadores to Cerro, thus filtering the half million soles tied up in
advances to enganchados by 1910. These enganchados, fast on the way to land-
lessness, were soon to make up the ranks of the new stable workforce in the
mines (Mallon, 1983, pp. 139±56, 187±205; Roberts, p. 51).34 As merchants, the
kulaks formed a secondary commercial system to supply the virtually mono-
psonistic market for foodstuffs constituted by Cerro. As Wilson has shown in
her study of the Tarma region, the central sierra saw the development of two
separate commercial circuits: one immigrant and one indigenous. Italians and
Chinese primarily ran the immigrant system, using the railway to import food-
stuffs and luxury goods from the coast and specializing in the export of Tarma's
principal product, coffee, from the Chanchamayo ceja de selva. The indigenous
merchants, who also acted as muleteers, specialised in providing foodstuffs to
the mining camps and the ceja de selva haciendas, as well as aguardiente to the
sierra. As long as supplying food to the mining camps was profitable, there was
little incentive for these merchants to attempt to sell their foodstuffs in the
Lima market, especially since that meant switching from mule to more expens-
ive rail transport. These systems simply coexisted. They did not mix, but both
had been created by the expansion and deepening of the cash economy in the
region since the 1880s (Wilson, 1987, pp. 130±7).
Wilson's findings square well with Miller's study of the impact of the Central
Railway on the regional economy. Indeed, according to Miller only those
pastoral haciendas closest to the railway appear to have derived any benefit at
all from the railway. Arable agriculture on the other hand received no stimulus,
and few agricultural products from the central sierra actually found their way to
Lima. Indeed, more foodstuffs came up than down (Miller, pp. 5±41, 47).
However, this was not simply a result of the structure of agricultural produc-
tion, it derived from the existence of a near monopsonistic market and a
dependent commercial system which in conjunction precluded, through lack
of incentives, the diversification of markets for foodstuff production in the
central sierra.
At the regional level the impact of the indigenous commercial circuit was
uneven. Tarma town was faced with supply shortages as the production of her
hinterland was siphoned away to the mines. Other towns fared rather better;
many served as entrepoÃts. Huancayo, in particular, was able to profit from its
geographical location and specific social composition to achieve a dominant
commercial position in the region (Long and Roberts, 1984, pp. 35±42; Roberts,
1976).
176 Peru, 1884±1930
By the late 1910s, Cerro had begun to reassess its system of production, forced
to do so, by several factors. Higher copper exports from Chile and Mexico were
making the copper industry more highly competitive; the mineral content of
the ores from Cerro de Pasco had fallen off sharply; and finally, the continued
acquisition of mining properties had made it necessary to integrate operations
and expand smelting capacity (O'Brien, 1996, pp. 121±2). The company's
response was to build a gigantic smelter at La Oroya, completed in 1922. The
smelter could produce 2500 tons of copper a day, as much as the two smelters
that were closed. The Oroya smelter profoundly changed the rhythm and
rationale of production, and its increased smelting capacity multiplied the
demand for ores. Moreover, the new technology made possible the commercial
production of new metals like lead and zinc.
The transformations, which corresponded to a transition from a manufactur-
ing to an industrial rationale (Mallon, 1983, ch. 6), had significant con-
sequences for the mining workforce. Prior to the construction of the smelter,
Cerro's production regime was based on contractors who employed enganchado
labourers. In the general changes of 1920s, however, contract mining was
curtailed and relegated to peripheral activities. Enganche labour, an expensive
and irregular source of labour, was removed from most operations. The new
production rationale required a stable and skilled labour force, which enganche
could not provide. Typically, enganchados resisted regularization of work-time
and preferred to work for very long shifts in order to pay back their advance and
return to their fields as quickly as possible (Mallon, 1983, p. 210; Flores
Galindo, 1993a, ch. 3). Gradually the company moved to the direct hiring of
labour, which in the 1920s came to represent just below 50 per cent of the
company's labour force.
The ecological disaster created by the fumes of the Oroya smelter pro-
foundly altered Cerro's links with the regional economy. The company omitted
to install filters, and as a result of the ensuing pollution vast amounts of land
were left in an unproductive state and many livestock were killed. Faced with
protests from both communities and landowners, Cerro promptly undertook
the purchase of contaminated land. In the process it accumulated over 200 000
hectares, and in the long run with the installation of filters Cerro benefited
greatly from this. It invested heavily in what became the DivisioÂn Ganadera
(cattle-raising section). It imported cattle and sheep and was able to raise
yields and achieve high levels of meat and wool production for local, national
and international consumption. Moreover, the DivisioÂn Ganadera produced
meat for the mining work force thus permitting the company to keep
wages low.
In conjunction with the gradual phasing out of contract mining and
enganche, the DivisioÂn Ganadera undermined the role of the local petty and
middle commercial class in the articulation of the mining centres with the
Paulo Drinot 177
regional economy. Of course, some merchants survived, especially those

who had diversified their interests. However, the majority proved unable to
compete in the changed situation. The introduction of company stores, for
example, made it difficult for small merchants and itinerant peddlers to com-
pete in the large market represented by the Cerro workforce. As wages were
consumed in situ and the company underwent a greater degree of vertical
integration, linkages to the regional economy probably declined. In 1931, the
delinking of Cerro and the regional economy was officially recognized through
the creation of the Pasco department (Espinosa Bravo, 1936, p. 117; Manrique,
1987, p. 268).
At the same time, the ecological disaster contributed directly to the continu-
ing proletarianization of the regional peasantry, as peasants were forced to sell
their now worthless land. Meanwhile, the road construction programmes of the
early 1920s played a major role too in the peasantry's transition to proletarian
status by facilitating migration to the mines, as well as increasing the peasan-
try's ties to the growing urban markets. By the early 1930s, like their comrades
on the coastal sugar plantation factories and in the oilfields, the Cerro workers
had developed working-class demands and methods of struggle and organiza-
tion. No longer did they seek to work endless shifts as quickly as possible so as
to return to their plots of land in time for the harvest. Now they demanded an
eight-hour working day and better working conditions (Mallon, 1983, pp. 234±
43; Flores Galindo, 1993a, ch. 5).
Thorp and Bertram (1978, p. 95) have argued that despite the relatively high
returned value of Cerro, the denationalization of mining was on the whole
detrimental to the national economy. The regional story adds further complex-
ity to the macro picture. Indeed, it is clear from the narrative that the develop-
ment of copper mining both before and after the arrival of Cerro created
strong regional linkages as both demand for labour and foodstuffs increased
and trade and transport networks were established. Meanwhile, a process of
social differentiation was unleashed as a peasant bourgeoisie emerged com-
posed of kulak peasants and a new group of merchant farmers, at the expense
of a pauperized and increasingly landless subsistence peasantry increasingly
dependent on wage-labour as enganchados working in the mines in a transi-
tional phase towards proletarian status. In the cities, and particularly in
Huancayo, a predominantly immigrant merchant class developed around the
new commercial circuits. In the 1920s, as a result of both exogenous and
endogenous factors, Cerro broke the delicate balance established with the
regional economy. The ecological disaster of La Oroya merely precipitated
these changes, as Cerro was able to increase its vertical integration and embark
on the formation of a stable workforce. As such, many of the regional linkages
were broken, a process that was both to Cerro's advantage and detrimental to
regional development.
178 Peru, 1884±1930
Conclusion
From many perspectives in the 1884±1930 period Peru was indeed, as the
Italian explorer Antonio Raimondi first noted more than a century ago, a
beggar sitting on a bench of gold. Indeed, Peru's natural resource, or export
commodity, endowment was unashamedly large. The loss of guano and nitrate
revenues following the War of the Pacific was mitigated by the rapid recovery of
other export sectors. In a sense, Peru's loss of nitrates was not as economically
important as Chile's gain of them. Thereafter, a strongly diversified export
quantum, with relatively high growth rates, came to characterize Peru's eco-
nomic development until the 1930s. However, as shown above, the relative
importance of each component within the export quantum changed over the
period. As a result of international price variations, locally-owned sectors with
high returned values lost their pole position to foreign-owned sectors, which in
the case of copper over time strongly reduced its returned value, or in the case
of oil contributed next to nothing to the national economy. These changes
were inscribed within a context of underdeveloped financial systems and
inadequate state management, which contributed to the general reduction in
the export sector's contribution to the national economy. It is therefore quite
natural that, as Portocarrero and Oliart (1989, pp. 108±12) have argued, the
non-establishment view of Peruvian history attributes the inconsistency
between the existence of natural riches and the existence of poverty and under-
development to a combination of bad government and imperialism. However,
this general view of Peru's export experience in the period 1884±1930 is insuffi-
cient, while not entirely false. Peru's export economy was strongly regionally
circumscribed, creating regional and local export economy histories character-
ized by their diversity and complexity.
It is clear from the preceding discussion that Peru's insertion into the world
economy profoundly transformed its regional economies. Each region exhib-
ited specific social, economic and political ± even cultural ± conditions prior to
the export economy period. At the same time, each region was inserted into the
world economy through different export products. Two questions arise from
this. First, to what extent was the nature of change determined by preexisting
regional conditions? And second, to what extent were the changes determined
by the export product itself?
With the exception of oil, all the export products created linkages in their
respective regional economies. In some cases, these linkages were articulated by
a preexisting commercial network, as in the northern sierra. In others, new
networks were created. These either competed with the traditional networks, as
in the south, or coexisted with them, as in the central sierra. The main railways
were built ahead of demand. The Central Railway only became viable through
the increased bulk needs of the copper trade; the Southern Railway played a
Paulo Drinot 179
minor role in the wool trade; but the smaller often private railways in the sugar
sector played a major role in cutting costs and releasing labour for other
activities. However, none of these railways contributed substantially to the
formation of an internal market.
The commercial networks served to stimulate urban growth and incipient
industrialization, oiled by merchant credit systems. Both the wool trade and
cotton production worked on cash advances in exchange for future crops. Non-
economic ties were often created alongside more strictly financial ones. On the
coast, foreign merchants established links with coastal planters through mar-
riage with the latter's daughters.
The implications for the indigenous population varied from region to region.
Changes at the level of the peasant economy were strongly linked to the ra-
tionale imposed by the export economy in each region. In the central sierra and
the coast, copper and sugar contributed to the proletarianization of the peasan-
try, and demographic growth played a central role in this process. The wool
sector, on the other hand, did not see any changes in the relations of production.
Cotton, paradoxically, adopted a precapitalist mode of production, although
sharecroppers became a rural proletariat in every other sense. The transition to
proletarian status resulted in new grievances and new forms of struggle. How-
ever, overall, this transition was largely incomplete at the end of our period.
In physical and capital terms the size of the export sectors and units varied
from region to region. All regions shared a common vulnerability to interna-
tional price fluctuations. The selva's rubber boom ended suddenly when prices
fell. The south, perhaps, remained the least vulnerable sector; its export quan-
tum was somewhat more diversified than that of other regions. On the other
hand, the power structure in the south meant that the effects of cyclical crises
were transmitted most directly to the most vulnerable elements in society. On
the coast, vulnerability was conditional on the degree to which sugar and
cotton dominated in each valley; in those valleys where export crops could be
substituted with domestic-use agriculture, it was possible to weather the worst
effects of the crises. By contrast, in the Chicama Valley the downturn in the
sugar process in the 1920s had far-reaching consequences, contributing to the
emergence of APRA.
To conclude, both preexisting conditions and the export product shaped and
fashioned the insertion of Peru's regions into the world economy. In all regions,
the interplay between the two factors had substantially different results.
Although all regions shared the intrinsic vulnerability represented by depend-
ence on an export product, the effects of this vulnerability were themselves
dependent on the prevailing economic, social and political conditions. The
prevailing conditions set the parameters of each region's insertion into the
world economy and, more important, of its consequences for regional and
national development.
180 Peru, 1884±1930
Notes
1 For comparative export commodity concentration ratios in Latin America see table 3.2
in Bulmer-Thomas (1994, p. 59).
2 Merchants stocked wool during the war period. At the end of the war, total wool
exports rose from 1460 metric tons in 1883 to 5256 metric tons in 1884.
3 By 1891, import items were subject to a 39 per cent duty on average. In 1893, an 8
per cent surcharge on all import duties was added (Thorp and Bertram, 1978, p. 30).
4 The Italian community played a key role in the development of industry. When the
Sociedad Nacional de Industrias was founded in 1896, 25 of the 76 founding members
were Italians (Bonfiglio, 1994, p. 179). On the contribution of Italian immigration to
the Peruvian economy, see also Worrall (1990), Chiaramonte (1983) and Alfaro Valle-
jos and Chueca Posadas (1975).
5 Almost all of the 77 million soles spent on public works were financed by loans
from the USA. The proportion of state revenues deriving from loans rose from around
5 per cent in 1919 to around 45 per cent in 1926±28. By 1931, the total public debt
stood at 500 million soles of which 360 million corresponded to the external debt
(Burga and Flores Galindo, 1994, p. 212; Collier and Collier, 1991, p. 139; Portocarrero,
1983, pp. 23±6).
6 The railway's effect on the structure of trade was greater than its impact on the volume
of trade. Transports costs accounted for a relatively small percentage of the total FOB
value of wool ± perhaps no more than 15 per cent ± and therefore the switch
from llama and mule transport to rail transport represented only a marginal saving
(Jacobsen, 1993, p. 185).
7 A similar commercial circuit was established along the shores of Lake Titicaca follow-
ing the introduction of steam navigation (Tamayo Herrera, 1982, p. 83).
8 By 1905 the city boasted four beer factories and four cocaine factories. In 1913, Italian
immigrants built a hydroelectric plant. The same industrial group took advantage of
the new energy supply to build two textile factories in the city. A flourmill was built in
1921. `Modernization' was a priority for a city that considered itself ± and in many
ways physically remained ± isolated from the rest of the country. In the late nineteenth
century the telegraph was installed, followed by the tramway in the 1900s. The
telephone arrived in 1921. The first cars for public service arrived in 1919. More
significantly for a city separated from the coast by some of the highest mountains in
the world, the first aeroplane arrived in 1921 (Flores Galindo, 1993b, p. 353; Tamayo
Herrera, 1981, pp. 108±9, 130, 144).
9 Three types of agriculture existed in the Cuzco area. The provinces of Cuzco, Quispi-
canchis, Calca and Urubamba were characterized by small haciendas producing corn,
wheat, barley, oats, rye, quinua as well as vegetables, potatoes, fruits and meat, milk and
dairy products. Significantly, these haciendas were for the most part directly exploited
by their owners. The principal markets were the highland region and to a lesser extent
the city of Cuzco. In this way, these haciendas were intimately linked to price fluctua-
tions in the wool markets. The highland areas of Cuzco were characterized by wool-
producing haciendas, exhibiting many of the same features of the Puno haciendas.
Finally, the ceja de selva region, in particular the province of La ConvencioÂn, has been
characterized by Hobsbawm as neo-feudal. These haciendas specialized in the produc-
tion of sugar ( for aguardiente ( and coca, palillo, annatto, woods, fruits, coffee and other
tropical products for the highland markets. Throughout the period, demand for
these products, especially alcohol and coca, remained high (Tamayo Herrera, 1981,
pp. 102±6; GuilleÂn MarroquõÂn, 1989, pp. 127, 130±7; Hobsbawm, 1970).
Paulo Drinot 181
10 However, at least in one respect, Arequipa's hegemony was challenged. Local land-
owners undertook the commercialization of their haciendas' production. By 1910,
three textile factories had been built in the Cuzco hinterland; and cocoa-growing
landowners from the ceja de selva region set up chocolate factories in Cuzco. One
local hacendada, Maria Romainville widow of Latorre, had perhaps the most diversi-
fied `concern' of all. As well as a textile mill in her hacienda in Urcos, she owned
another two haciendas in La ConvencioÂn Valley and one in Quispicanchis. Her
almaceÂn in Cuzco sold textiles, aguardiente, sugar, cocoa, coffee, corn and wheat
(Mo È rner, 1979, p. 28).
11 By 1920, a biscuit factory, a sweet factory, a beer factory, a soap and candle factory, a
shoe factory, as well as a textile factory, dotted the Arequipa landscape (Flores
Galindo, 1993b, pp. 351±2).
12 Unlike Puno or Cuzco, land tenure in most of Arequipa has traditionally been of the
minifundio kind. These minifundios produced foodstuffs, but there appears to have
been a growing amount of land planted with corn for animal fodder. At the same
time, sugar plantations were built in the Tambo Valley on Arequipa's coast and in the
Camana and Majes Valleys viniculture gave over to cotton cultivation (Flores
Galindo, 1993b, pp. 348±51).
13 Manrique (1997, p. 355) defines gamonalismo as à type of exercise of power and social
structure of domination whose crystallization played a critical role in the accumula-
tion of wool'.
14 According to Jacobsen (1993, pp. 259±90), the transition to colono status did not
significantly alter the peasant family's living and working conditions: the labour
services performed for the landowner replaced those previously performed for local
political authorities. Thus, although occasionally peasants sold land as a result of
pressures of a coercive and sometimes violent kind, more often peasants sold their
land willingly, if reluctantly. Although this is likely to have been the case for the
poorest peasants, for the average peasant the transition to colono status represented a
decline in both economic and social status. Moreover, land was not merely a source of
economic well-being for the Peruvian highland peasant; the links with land were also
spiritual. These elements combine to explain why peasants resisted the expansion of
haciendas.
15 According to Burga and Flores Galindo (1994, ch. 4), to contrast in nineteenth-
century rebellions, the Gran SublevacioÂn del Sur acquired a political millenarian, rather
than merely fiscal context. The `rebels' called for the reimposition of the Inca Empire.
However, recent research indicates that these uprisings, which were far less isolated
and spontaneous than previously assumed, had little to do with millenarianism.
Rather, they were tagged millenarian by local landowners intent on showing that
the Indians were seeking to subvert the national order through calls for the restora-
tion of the Tawantinsuyo and by allying with Indians from Bolivia to establish an
alternative anti-national state (Manrique, 1996, p. 227).
16 One exemption appears to have been domestic-use agriculture, which benefited from
the demand generated by a growing urban population. This situation allowed some
Arequipa merchants to weather the crisis ( Jacobsen, 1988, pp. 161±3).
17 Of course, this is not to imply an oversimplified Pavlovian correlation between
falling incomes and peasant rebellion. As Burga and ReaÂtegui have noted, the Rumi
Maqui rebellion of 1915 occurred during a period of rising wool prices, with corre-
sponding increases in the peasantry's living standards. However, there is little doubt
that by the early 1920s, the altiplano peasantry was undergoing a process of acute
pauperization.
182 Peru, 1884±1930
18 Sugar, which in Peru is particularly resistant to pests, can be cut and processed
throughout the year because its production is independent of rainfall. Moreover, the
indigenous varieties of cotton grown in Peru are high-yielding: cotton growers can
expect to produce two crops a year. In addition, Peruvian cotton is typically of a long
staple variety, a characteristic that commands a premium in international markets.
19 In 1912, foreign ownership accounted for no more than 25 per cent of output. By
1930, this percentage had fallen (Albert, 1985, pp. 240±1; Thorp and Bertram, 1978,
pp. 42±3).
20 Exports to the USA rose from 1299 tons in 1897 to a peak of 58 731 tons in 1901. This
represented 1.2 per cent and 51.2 per cent respectively of Peruvian exports (Albert,
1976, p. 22a).
21 The percentage increase in sugar production was 83 per cent in the period 1889±1900,
and 38 per cent in 1900±10 (Thorp and Bertram, 1978, p. 43).
22 Albert (1976, p. 138a) estimates that some one million pounds sterling were invested
in 1912±14.
23 The downturn in the 1915±20 period is accounted largely by the fall from US$16
million in 1919 to US$5 million in 1920 in sugar exports by value.
24 During the expansionary period of the war years, Casa Grande, like other major
coastal plantations, purchased two highland haciendas in Cajamarca, and later a
rice hacienda in the Jequetepeque Valley. In Cajamarca, cattle rearing was combined
with foodstuff production on a sharecropping basis. Casa Grande was thus able to
produce cheap foodstuffs for its growing plantation labour force, and thus maintain
low wages. Moreover, by requiring that highland hacienda tenants pay off their labour
services on the sugar plantation, it was able to secure cheap peasant labour (Taylor,
1983, p. 34±42; Deere, 1990, pp. 48).
25 The situation changed somewhat in 1918 with the entry of the Banco Italiano into
the market for short-term loans to cotton growers. However, at first the Banco Italiano
restricted its credit facilities to the Italian community (Quiroz, 1993, p. 66).
26 A small oilfield in Puno operated in the first decade of this century (Thorp and
Bertram, 1978, p. 96).
27 On the other hand, Lobitos was not able to secure such concessions from the
government. Its contribution to the Peruvian exchequer was therefore significantly
higher than the IPC's (Miller, 1982b, pp. 421±3).
28 According to Michael Taussig (1987, p. 100), commenting on the well-documented
Putumayo events: `Torture and terror were not simply utilitarian means of produc-
tion; they were a form of life, a mode of production, and in many ways, for many
people, not least of whom were the Indians themselves, its main and consuming
product'. See also Varese (1973, pp. 169±209, 235, 247).
29 In 1912, 112 tons of rubber from the Ucayali River were commercialized through
Iquitos, while 127 tons were sent to the Pacific port of Mollendo (Roux, 1994, pp.
158±68).
30 See Flores Marin (1987, ch. 3) for an account of the workings of the Tambopata
Rubber Syndicate Ltd in the Madre de Dios region.
31 The Mantaro Valley provided flour or unprocessed cereals, as well as meat. The
Huaylas region specialized in sugar and sweets. Huanuco was a provider of fruits,
coca and aguardiente. Wood was brought from Paucartambo, mercury from Huanca-
velica and Chonta, coal from Yanahuanca and ChayacaÂn, and salt from the Salinas de
Huaura and later from San Blas, on Lake JunõÂn (Contreras, 1988, pp. 40±3).
32 In the 1870s, the Sunday fair at Huancayo brought together between 25 000 and
30 000 traders (Long and Roberts, 1984, pp. 32±5; Manrique, 1987, p. 255).
Paulo Drinot 183
33 The company built a 125km railway from Cerro de Pasco to La Oroya and a new
smelter at Tinyahuarco, the largest in Peru. By 1907 the Tinyahuarco smelter was
capable of processing a total of 1000 tons of ore a day, more than five times the
amount processed by the next three largest smelters (Mallon, 1983, p. 171).
34 For a more detailed treatment of these issues see Samaniego (1974).
References
Albert, B. (1985) Èxternal Forces and the Transformation of Peruvian Coastal Agriculture,
1880±1930', in C. Abel and C.M. Lewis (eds), Latin America, Economic Imperialism and the
State: The Political Economy of the External Connection from Independence to the Present,
London/Dover, New Hampshire: The Athlone Press, pp. 231±49.
ÐÐ (1984) `The Creation of a Proletariat on Peru's Coastal Sugar Plantations, 1880±1920'
in B. Munslow and H. Finch (eds), Proletarianisation in the Third World, London: Croom
Helm, pp. 99±120.
ÐÐ (1976) An Essay on the Peruvian Sugar Industry, 1880±1920 and The Letters of Ronald
Gordon, Administrator of The British Sugar Company in Can Ä ete, 1914±1920, Norfold: Uni-
versity of East Anglia.
Alfaro Vallejos, J. and Chueca Posadas, S. (1975) El proceso de hacer la AmeÂrica: Una familia
italiana en el PeruÂ, Lima: Pontificia Universidad CatoÂlica del PeruÂ.
Arroyo, E. (1981) La hacienda costen Äa en el PeruÂ: Mala-Can Äete, 1532±1968, Lima.
Barclay, F. (1989) La Colonia del PereneÂ: Capital ingleÂs y economõÂa capitalista en la configuracioÂn
de la regioÂn de Chanchamayo, Iquitos: Centro de Estudios Teologicos de la Amazonia.
Basadre, J. (1983) Historia de la RepuÂblica del PeruÂ, 1822±1933, 10 vols, Lima: Editorial
Universitaria.
Beals, C. (1934) Fire in the Andes, Philadelphia/London: J.B. Lippincott.
Bell, W.S. (1985) An Essay on the Peruvian Cotton Industry, 1825±1920, working paper no. 6,
Centre for Latin American Studies, Liverpool: University of Liverpool.
Bertram, G.I. (1977) `ModernizacioÂn y cambio en la industria lanera del PeruÂ, 1919±1930:
un caso frustrado de desarollo', Apuntes 6, pp. 3±22.
Blanchard, P. (1982) The Origins of the Peruvian Labor Movement, 1883±1919. Pittsburgh:
University of Pittsburgh Press.
Blanchard, P. (1979) `The Recruitment of Workers in the Peruvian Sierra at the Turn of the
Century: The Enganche System', Inter-American Economic Affairs, vol. 33, pp. 63±83.
Bonfiglio, G. (1995) `Los italianos en Lima', in A. Panfichi and F. Portocarrero (eds),
Mundos Interiores: Lima, 1850±1950, Lima: Universidad del PacõÂfico, pp. 43±73.
ÐÐ (1994) Los italianos en la sociedad peruana: Una visioÂn histoÂrica, Lima: SAYWA edi-
ciones.
Bonilla, H. (1980) `Guano y crisis en el PeruÂ del XIX' in Nueva Historia General del PeruÂ,
Lima: Mosca azul editores.
ÐÐ (1977) Èl caucho y la economõÂa del oriente peruano', in Los mecanismos de un control
econoÂmico, (Vol. V), Instituto de Estudios Peruanos/Fondo del Libro del Banco Industrial
del PeruÂ, pp. 123±34.
ÐÐ (1974) Guano y burguesõÂa en el PeruÂ, Lima: Instituto de Estudios Peruanos.
Burga, M. (1976) De la encomienda a la hacienda capitalista: El valle del Jequetepeque del siglo
XVI al XX , Lima: Instituto de Estudios Peruanos.
184 Peru, 1884±1930
Burga, M. and ReaÂtegui, W. (1981) Lanas y capital mercantil en el sur: La Casa Ricketts,
1895±1935, Lima: Instituto de Estudios Peruanos.
Burga, M. and Flores Galindo, A. (1994) Apogeo y crisis de la RepuÂblica AristocraÂtica, in
A. Flores Galindo, Obras Completas Vol. II, Lima: FundacioÂn Andina/SUR Casa de
Estudios del Socialismo.
Chiaramonte, G. (1983) `La migracioÂn italiana en AmeÂrica Latina. El caso peruano',
Apuntes, vol. 13, pp. 15±36.
Collier, D. and Collier, R.B. (1991) Shaping the Political Arena, Princeton: Princeton Uni-
versity Press.
Consejo Provincial de Paita (1935) Memoria leõÂda por el Alcalde Titular Sen Ä or Rodolfo Alfa-
geme en la sesioÂn de Concejo celebrada el 2 de enero de 1935, dando cuenta de la marcha de la
InstitucioÂn durante el an
Ä o anterior, Paita: Imprenta Mercantil.
Contreras, C. (1988) Mineros y campesinos en los Andes: Mercado laboral y economõÂa cam-
pesina en la sierra central, siglo XIX, Lima: Instituto de Estudios Peruanos.
Cuche, D. (1975) Poder blanco y resistencia negra en el PeruÂ: un estudio de la condicioÂn social
del negro en el PeruÂ despueÂs de la abolicioÂn de la esclavitud, Lima: Instituto Nacional de
Cultura.
Cueto, M. (1992) `Sanitation from Above: Yellow Fever and Foreign Intervention in Peru,
1919±1922', Hispamic American Historical Review HAHR, vol. 72(1), pp. 1±22.
Cueto, M. and Lossio, J. (1999) InovacioÂn en la agricultura: FermõÂn Tanguis y el algodoÂn en el
PeruÂ, Lima: Universidad del PacõÂfico/Cosapi OrganizacioÂn Empresarial.
Davila Apolo, D. (1976) Talara, los petroleros y la huelga de 1931, Memoria para optar al
grado de Bachiller en Ciencias Sociales con mencioÂn en SociologõÂa, Lima: Pontificia
Universidad CatoÂlica del PeruÂ.
Deere, C.D. (1990) Household and Class Relations: Peasants and Landlords in Northern Peru,
Berkeley: University of California Press.
Degregori, C.I. (1990) El surgimiento de Sendero Luminoso: Del movimiento por la gratuidad de
la ensenÄ anza al inicio de la lucha armada, Ayacucho 1969±1979, Lima: Instituto de Estu-
dios Peruanos.
Deustua, J. and ReÂnique, J.L. (1984) Intelectuales, indigenismo y descentralismo en PeruÂ,
1897±1931, Cuzco: BartolomeÂ de Las Casas.
Espinosa Bravo, C.A. (1936) Facetas de Jauja, Jauja: C.E. Sanguinetti.
Flores Galindo, A., (1993a), `Los mineros de Cerro de Pasco, 1900±1930 (Un intento de
caracterizacioÂn social)', in A. Flores Galindo, Obras Completas, (Vol. I), Lima: FundacioÂn
Andina/SUR Casa de Estudios del Socialismo.
ÐÐ (1993b) Àrequipa y el sur andino: Ensayo de historia regional, Siglos XVIII±XX', in
A. Flores Galindo, Obras Completas, (Vol. I), Lima: FundacioÂn Andina/SUR Casa de
Estudios del Socialismo.
Flores Marin, J.A. (1987) La explotacioÂn del caucho en el PeruÂ, Lima: Consejo de Ciencia y
Technoloqia del PeruÂ CONCYTEC.
Fukumoto, M. (1997) Hacia un nuevo sol: Japoneses y sus descendientes en el PeruÂ, Lima:
AsociacioÂn Peruano Japonesa del PeruÂ.
Galdo Gutierrez, V. (1992) Ayacucho: Conflictos y Pobreza ± Historia Regional (siglos
XVI±XIX), Ayacucho: Universidad Nacional San CristoÂbal de Huamanga.
Gamarra, J. (1992) Èstado, modernidad y sociedad regional: Ayacucho, 1920±1940',
Apuntes, vol. 31, pp. 103±14.
Gardiner, C.H. (1975) The Japanese and Peru, 1873±1973, Albuquerque, University of New
Mexico Press.
Gilbert, D.L. (1982) La oligarquõÂa peruana: Historia de tres familias, Lima: Editorial Hori-
zonte.
Paulo Drinot 185
GoÂmez, J. and BazaÂn, I. (1989) Capitalismo y formacioÂn regional: Chiclayo entre los siglos XIX
y XX, Chiclayo: PoblacioÂ n y Desarollo/Instituto de InvestigacioÂ n y CapacitacioÂn.
ÐÐ (1981) Capitalismo y regioÂn en Lambayeque, 1879±1930, Chiclayo: Gaviota Roja
Editores.
Gonzales, M. (1985) Plantation Agriculture and Social Control in Northern Peru, 1875±1933,
Austin: University of Texas Press.
Gootenberg, P. (1993) Imagining Development: Economic Ideas in Peru's `Fictitious Prosperity'
of Guano, 1840±1880, Berkeley: University of California Press.
Gootenberg, P. 1989, Between Silver and Guano: Commercial Policy and the State in Postinde-
pendence Peru, Princeton: Princeton University Press.
GuilleÂn MarroquõÂn, J. (1989) La economõÂa agraria del Cusco, 1900±1980, Cusco: Centro de
Estudios Rurales Andinos BartolomeÂ de las Casas.
Hobsbawm, E. (1970) À Case of Neo-Feudalism: La ConvencioÂn, Peru', Journal of Latin
American Studies, vol. 1(1), pp. 31±50.
Hunt, S. (1985) `Growth and Guano in Nineteenth Century Peru', in R. CorteÂs Conde, and
S. Hunt (eds), The Latin American Economies: Growth and the Export Sector, 1880±1930,
New York/London: Holmes and Meier, pp. 255±318.
Jacobsen, N. (1993) Mirages of Transition: The Peruvian Altiplano, 1780±1930, Berkeley and
Los Angeles: University of California Press.
ÐÐ (1988) `Free Trade, Regional Elites, and the Internal Market in Southern Peru,
1895±1932', in J. Love and N. Jacobsen (eds), Guiding the Invisible Hand: Economic
Liberalism and the State in Latin American History, Westport: Praeger, pp. 145±76.
Kapsoli, W. and ReaÂtegui, W. (1972) El campesinado peruano: 1919±1930, Lima: Universi-
dad Nacional Mayor de San Marcos.
KlareÂn, P. (1973) Modernization, Dislocation and Aprismo: Origins of the Peruvian Aprista
Party, 1870±1932, Austin/London: University of Texas Press.
Kruijt, D. and Vellinga, M. (1979) Labor Relations and Multinational Corporations: The Cerro
de Pasco Corporation in Peru, 1902±1974, Assen: Van Gorcum.
Lockhart, J. (1974) Spanish Peru, 1532±1560: A Colonial Society, Madison/London: Univer-
sity of Wisconsin Press.
Long, N. and Roberts, B. (1984) Miners, Peasants and Entrepreneurs: Regional Development in
the Central Highlands of Peru, Cambridge: Cambridge University Press.
LoÂpez-Ocon Cabrera (1995) Èl nacionalismo y los orõÂgenes de la Sociedad GeograÂfica de
Lima', in M. Cueto (ed.), Saberes Andinos: Ciencia y tecnologõÂa en Bolivia, Ecuador y PeruÂ,
Lima: Instituto de Estudio Peruanos.
Low, A. (1979) Agro-Exporters as Entrepreneurs: Peruvian Sugar and Cotton Producers,
1890±1945, doctoral thesis, University of Oxford.
Lumbreras, L. (1980) Èl PeruÂ prehispaÂnico', in Nueva Historia General del PeruÂ, 2nd edn,
Lima: Mosca Azul.
Mallon, F. (1983) The Defense of Community In Peru's Central Highlands: Peasant Struggle and
Capitalist Transition, 1860±1940, Princeton: Princeton University Press.
Manrique, N. (1997) Circuitos mercantiles y acumulacioÂn en escala regional: Las provincias
altas de Arequipa, siglos XVIII-XX, manuscript.
ÐÐ (1996) Historia de la RepuÂblica, Lima: Cofide.
ÐÐ (1988) Yawar Mayu: Sociedades terratenientes serranas, 1879±1910, Lima: Instituto
FranceÂs de Estudios Andinos/DESCO.
ÐÐ (1987, Mercado interno y reÂgion: La sierra central, 1820±1930, Lima: Centro de Estudios
Teologicos de la Amazonia.
Matos Mar, J. (1976) Yanaconaje y reforma agraria en el PeruÂ: El caso del valle de Chanca,
Lima: Instituto de Estudios Peruanos.
186 Peru, 1884±1930
Mc Evoy, C. (1994) Un proyecto nacional en el siglo XIX: Manuel Pardo y su visioÂn del PeruÂ,
Lima: Pontificia Universidad CatoÂ lica del PeruÂ .
Miller, R. (1983) `The Grace Contract, The Peruvian Corporation and Peruvian History'.
Ibero-Amerikanisches Archiv vol. 9(3/4), pp. 319±48.
ÐÐ (1982a) `The Coastal Elite and Peruvian Politics, 1895±1919', Journal of Latin American
Studies, vol. 14(1), pp. 97±120.
ÐÐ (1982b) `Small Business in the Peruvian Oil Industry: Lobitos Oilfields Limited Before
1934', Business History Review, vol. LVI(3), pp. 400±23.
ÐÐ (1982c) `The Wool Trade of Southern Peru, 1850±1915', Ibero-Amerikanisches Archiv,
vol. 8(3), pp. 297±311.
ÐÐ (1976) `The Making of the Grace Contract: British Bondholders and the Peruvian
Government, 1885±1890', Journal of Latin American Studies, vol. 8(1), pp. 73±100.
ÐÐ (c. 1976) `Railways and Economic Development in Central Peru, 1890±1930', in
R. Miller, C.T. Smith and J. Fisher (eds), Social and Economic Change in Modern Peru,
Liverpool: University of Liverpool, pp. 27±52.
Mitre, A. (1986) El monedero de los Andes. RegioÂn econoÂmica y moneda boliviana en el siglo
XIX, La Paz: Hisbol.
Morimoto, A. (1979) Los inmigrantes japoneses en el PeruÂ, Lima: Universidad Nacional
Agraria.
MoÈ rner, M. (1979) Notas sobre el comercio y los comerciantes del Cuzco desde fines de la colonia
hasta 1930, Lima: Instituto de Estudios Peruanos.
Nugent, D. (1997) Modernity at the Edge of Empire: State, Individual, and Nation in the
Northern Peruvian Andes, 1885±1935, Stanford: Stanford University Press.
O'Brien, T.F. (1996) The Revolutionary Mission: American Enterprise in Latin America,
1900±1945, Cambridge: Cambridge University Press.
OreÂ, T. (1990) Riego y organizacioÂn: EvolucioÂn histoÂrica y experiencias actuales en el PeruÂ, Lima:
Intermediate technology Development Group.
Pachas Castilla, R. (1976) EconomõÂa y sociedad en el valle de Chincha: 1860±1918, Lima:
Universidad Nacional Mayor de San Marcos.
Parker, D. (1998) The Idea of the Middle Class: White-Collar Workers and Peruvian Society,
1900±1950, University Park, Penn.: The Pennsylvania State University Press.
Peloso, V. (1999) Peasants on Plantations: Subaltern Strategies of Labor and Resistance in the
Pisco Valley, Peru, Durham/London: Duke University Press.
Pennano, G. (1988) La economõÂa del caucho, Iquitos: Centro de Estudios Teologicos de la
Amazonia.
Portocarrero, F. (1995) El Imperio Prado: 1890±1970, Lima: Universidad del PacõÂfico.
Portocarrero, F. and TorrejoÂn, L. (1992) Las inversiones en valores nacionales de la eÂlite
econoÂmica, PeruÂ: 1916±1932, Lima: Universidad del PacõÂfico/Consorcio de InvestigacioÂn
EconoÂmica.
Portocarrero, F, BeltraÂn, A. and Romero, M.E. (1992) Compendio estadõÂstico del PeruÂ,
1900±1990, Lima: Universidad del PacõÂfico/Consorcio de InvestigacioÂn EconoÂmica.
Portocarrero, G. (1983) De Bustamante a OdrõÂa: El fracaso del Frente DemocraÂtico Nacional,
1945±1950, Lima: Mosca Azul Editores.
Portocarrero, G. and Oliart, P. (1989) El PeruÂ desde la escuela, Lima: Insituto de Apoyo
Agrario.
Puga de Losada, A., (1893), Ensayos literarios, Lima: Imp. de la IlustracioÂn Sud-Americana.
Quiroz, A. W. (1993) Domestic and Foreign Finance in Modern Peru, 1850±1950: Financing
Visions of Development, London: St Antony's/Macmillan.
Quiroz, A.W. (1988) Financial Leadership and the Formation of Peruvian Elite Groups,
1884±1930, Journal of Latin American Studies, vol. 20, pp. 49±81.
Paulo Drinot 187
ReanÄ o, G. and VaÂsquez, E. (1988) El Grupo Romero: Del algodoÂn a la banca, Lima: Universi-
dad del PacõÂfico/Centro de InvestigacioÂ n y PromocioÂ n del Campesinado.
Roberts, B. (c. 1976) `The Social History of a Provincial Town: Huancayo, 1890±1972', in
R. Miller, C.T. Smith, and J. Fisher (eds), Social and Economic Change in Modern Peru,
Liverpool: University of Liverpool, pp. 130±97.
RodrõÂguez Doig, E. (1986) El Camino de los Enganchados, Lima: AsociacioÂn Peruana para el
Fomento de las Ciencias Sociales.
RodrõÂguez Pastor, H. (1989) Hijos del Celeste Imperio en el PeruÂ (1850±1900): MigracioÂn,
agricultura, mentalidad y explotacioÂn, Lima: Instituto de Apoyo Agrario.
Roux, J.C. (1994) L'Amazonie peÂruvienne: Un Eldorado deÂvoreÂ par la foreÃt, 1821±1910, Paris:
L'Harmattan.
Samaniego, C. (1974) Location, Social Differentiation and Peasant Movements in the Central
Sierra of Peru, doctoral dissertation, University of Manchester.
San RomaÂn, J. (1994) Perfiles histoÂricos de la AmazonõÂa peruana: Iquitos, Centro de Estudios
TeoloÂqicos de la Amazonia, Centro AmazoÂnico de Antropologia e AplicacioÂn Pinctica,
Instituto de Investigacioes de la Amazonia Peruana.
Seminario, B. and Beltran, A. (1998) Crecimiento econoÂmico en el PeruÂ: nuevas evidencias
estadõÂsticas, Lima: Universidad del PacõÂfico.
Stein, S. (1980) Populism in Peru, The Emergence of the Masses and the Politics of Social
Control, Madison: The University of Wisconsin Press.
Stewart, W. (1951) Chinese Bondage in Peru, Durham: Duke University Press.
Tamayo Herrera, J. 1981, Historia social del Cuzco republicano, Lima: Editorial Universo.
Taussig, M. (1987) Shamanism, Colonialism, and the Wild Man: A Study in Terror and Healing,
Chicago/London: The University of Chicago Press.
Taylor, L. (1986a) Bandits and Politics in Peru: Landlord and Peasant Violence in Hualgayoc,
1900±1930, Cambridge: Centre of Latin American Studies.
ÐÐ (1986b) Estates, Freeholders and Peasant Communities in Cajamarca, 1876±1972, work-
ing paper no. 42, Centre of Latin American Studies, Cambridge: University of Cam-
bridge.
ÐÐ (1983) Cambios capitalistas en las haciendas cajamarquinas, 1900±1935, working paper
no. 39, Centre of Latin American Studies, Cambridge: University of Cambridge.
Thorp, R. and Bertram, G. (1978) Peru, 1890±1977: Growth and Policy in an Open Economy,
London: Macmillan.
Varese, S. (1973) La sal de los cerros: Una aproximacioÂn al mundo Campa, Lima: Retablo de
Papel Ediciones.
Weinstein, B. (1983) `Capital Penetration and Problems of Labour Control in the Amazon
Rubber Trade', Radical History Review, vol. 27, pp. 121±43.
Werlich, D. (1978) Peru, A Short History, Illinois: Southern Illinois Press/ Carbondale and
Edwardsville.
Wilson, F. (1987) `The Conflict between Indigenous and Immigrant Commercial Systems
in the Peruvian Central Sierra, 1900±1940', in R. Miller (ed.), Region and Class in Modern
Peruvian History, Liverpool: University of Liverpool, pp. 125±62.
Worrall, J.E. (1990) La inmigracioÂn italiana en el PeruÂ, 1860±1914, Lima: Instituto Nacional
de Cultura.
Yepes del Castillo, E. (1972) Peru, 1820±1920: Un siglo de desarrollo capitalista, Lima:
Instituto de Estudios Peruanos/CampodoÂnico ediciones.
7
Bolivia, 1900±39: Mining, Railways and
Education
Manuel E. Contreras
Introduction
During the first 40 years of this century, the Bolivian economy was a classic
example of rapid export-led growth, based on tin exports. This new export
expansion consolidated Bolivia's entrance into the international market in
the last third of the nineteenth century with the export of silver. However,
Bolivia's export base at the onset of the twentieth century was more diversified
than it was to be by the end of the 1930s. In 1900, silver represented 36 per cent
of Bolivia's total exports, rubber represented 26 per cent, while the tin industry
was responsible for the remaining 24 per cent. The relative importance of tin
surpassed the importance of silver and rubber by 1903. Silver exports decreased
significantly from 1905 onwards. Rubber exports remained stable until 1910,
when they accounted for 37 per cent of total exports, but thereafter decreased,
reaching 5 per cent in 1920.1
Yet it is tin that has caught the attention of scholars both within and outside
Bolivia. According to Almaraz (1976): `The twentieth century [in Bolivia]
arrived on the shoulders of the tin industry', and the destiny of this mineral
has marked Bolivia's development until recently. Since 1910, tin has repres-
ented more than 50 per cent of the value of total exports of the country, and
since 1930 it has represented 70 per cent. Of the remaining 30 per cent, two-
thirds has been composed of other minerals including silver, tungsten, anti-
mony, copper, lead, zinc, bismuth and sulphur, while agricultural and farming
products such as rubber, coca and cattle have represented only 8 per cent
(GutieÂrrez Guerra, 1940, pp. 6, 23, 26±7). Tin formed the economic base for
Bolivia's growth and modernization, and it was the exploitation of this mineral
that both attracted the greatest amount of investment and paid the highest
taxes. In order to maximize the exploitation of tin, an extensive infrastructure
including railways and roads was constructed. At the same time, public services
such as drinking water, sewage systems, electricity, health and public education
188
Manuel E. Contreras 189
were created. Finally, new financial institutions were established, including the
Banco Nacional de Bolivia in 1911, the Banco Central, and the General Comp-
troller of the Republic in 1929 and 1930, respectively. The military was profes-
sionalized and an important process of urbanization and modernization took
place.
In 1900 Bolivia had a population of 1.8 million. By 1905 this had been
reduced to 1.7 million as a result of the territories recently conceded to Brazil
and Chile. At the time the country had 1.19 persons per square kilometre,
placing Bolivia last in population density in South America, while occupying
third place in territorial expansion and seventh place in total population (Ofi-
cina Nacional de InmigracioÂn, EstadõÂsticas y Propaganda GeograÂfica, 1905, pp.
235±7). Due to its mountainous topography, Bolivia was, in addition, a land
with little internal communication and where access to many of the country's
areas was quite difficult. The construction of communication systems, there-
fore, was vital for the growth of the export sector and its general development
and modernization.
We will study the evolution of the Bolivian economy between 1900 and 1939
by examining the development of the tin mining industry. We will attempt to
weave together the political economy of the export sector and the association
between the mining industry and the other productive sectors. In addition, this
chapter will study the effects of the Great Depression (1929±33) and the Chaco
War (1932±35) on the tin market as well as on the relation between the Bolivian
state and the export sector through fiscal and exchange policy. The evolution of
the tax structure and its significance in the Bolivian economy represents a vital
aspect in understanding the development of the state (Gallo, 1990).
The chapter also briefly analyses the construction of railways and road sys-
tems and the establishment of the national education system. These develop-
ments reflect the process of state consolidation that was carried out during this
period of èxport-led growth'. Railways and education were part of the liberal
legacy ± the Liberals governed Bolivia between 1899 and 1929 ± and were
effectively financed by the export sector. Railways were not only essential
for mining, they also formed part of the liberals' modernizing project.
Construction of roads replaced railways and constituted, at first glance, the
government's response to the 1930s Depression. Although the connection of
education with the export sector was less direct, it reflected the positivist ideals
of the Liberals and the need to establish a national education system, a goal
unfulfilled since the nineteenth century. As such, the export sector represented
perhaps the most important element in the modernization of the country as it
established the foundations of the Bolivian education system.
Indeed, the main argument of this chapter is that the export sector's
resources allowed investment in both physical infrastructure and human capi-
tal, factors now recognized as central to development. The èxport-led growth'
190 Bolivia, 1900±39
model, in its expansive phase and in its moments of crisis, had a significant
internal impact: it effectively established the foundations on which future
development would be built. The chapter maintains that the various govern-
ments' capacity to extract the surplus coming from the mining sector was
greater than is traditionally accepted. This chapter also analyses the variations
in the mechanisms of financial extraction by the state and argues that at those
times when considerable resource accumulation took place, either at the gov-
ernment level or at the level of companies, other factors helped determine its
impact on the country's development. In the case of the government, military
expenditure, investment in non-productive projects, and corruption played a
dominant role in minimizing the potential impact of such capital accumula-
tion. However, in the case of the mining companies, the ineffective use of
resources was primarily due to the absence of opportunities and conditions
for productive investment.
The chapter is divided into six parts. Following this introduction we analyse
the chronological development of the tin mining industry. We then focus on
the influence of the mining industry on the national economy through fiscal
policies, followed by a summary and synthesis of the development of railways
and roads. After this we describe the evolution of the education system and the
efforts towards modernization during this period, and in the final part review
the development of the export economy, emphasising the role of the state in
relation to the interests of miners and their efforts towards modernization
during this period.
Development of the mining sector

Beginnings of tin mining
During the first decades of the twentieth century, the growth of tin produc-
tion in Bolivia was spectacular. As a direct response to prices and demand,
production ± measured in terms of total exports (since production statistics
do not exist) ± more than doubled between 1900 and 1910 (from 9740 fine
metric tons (FMT) to 22 885) and increased by a further 30 per cent during
the 1920s, reaching 28 230 FMT in 1920 (Contreras, 1994, table 1, p. 9).2
This rise was reflected in the importance of Bolivian production as a percent-
age of world supply: production rose from 11 per cent of world production
in 1900 to 20 per cent in 1920 (LoÂpez Rivas, 1955, pp. 11, 23). Similarly,
the importance of tin for the Bolivian economy increased significantly since,
as we have seen, the value of tin exports came to represent 70 per cent of
exports.
The rapid expansion in Bolivian production in the 1900s ± with an annual
growth rate of 10.1 per cent (see Table 7.1) ± was the result of the development
of tin mining during the nineteenth century and the wealth of ores. The
Table 7.1 Bolivia: tin exports in four decades; annual averages, maximum, minimum,
and annual rates of growth, 1900±39
Period Annual Maximum Year Minimum Year Annual rate

average FMT FMT of growth
(FMT) (per cent)
1900±09 14 909 21 342 1909 9 739 1900 10.1

1910±19 24 710 29 280 1918 21 324 1916 2.6
1920±29 33 216 47 191 1929 19 086 1921 5.9
1930±39 24 844 38 723 1930 14 957 1933 3.6
Source: Calculated from export data in Walter GoÂmez-D'Angelo (1978), pp. 218±20.
expansion benefited from the industry's capacity to resolve the problems posed
by internal transportation, the high cost of fuel, and an insufficiently large
technical and managerial workforce (Contreras, 1994).
The methods of production during this first period (1900±15) were primitive
and not unlike those utilized during the eighteenth and nineteenth centuries.
With the exception of certain mining companies, exploitation was not sys-
tematic: there were no geological maps, and production was undertaken with-
out prior planning, exploiting only the richest and most accessible veins while
using as little capital as possible. This unsystematic process often resulted in the
deterioration of the ore deposits' economic value, as a great percentage of
the mines' potential was ignored. The rise in prices, the advent of railroads,
and the increasing availability of capital led to large-scale imports of machinery
and, from 1908 onwards, concentration mills (Contreras, 1980, pp. 27±9).
At first, merchants were the main source of capital, mostly advances and
credits. However, these sources were able to satisfy the demand for capital only
in the short term. Given that the financial participation of local banks was
marginal, investment capital was sought internationally. After 1906 ± following
the Peace Treaty with Chile, which solved the territorial problem ± there was a
significant flow of Chilean capital into the industry, thereby increasing the
number of companies that were floated in the Santiago Stock Exchange. The
main Chilean purchase, the Llallagua mine (the richest in Bolivia), was made in
1906. In 1912, investment from Chile reached £2.6 million. By 1913, foreign
capital exceeded £10 million, composed primarily of Chilean, British and, to a
lesser extent, US capital (Contreras 1980, pp. 33±7).3
The capitalization of mining made possible the concentration of production
by a group of `modern' companies characterized by a high level of mechaniza-
tion in the exploitation and concentration of tin, modern administration,
greater access to working capital, and the direct selling of minerals. This
group accounted for 55 per cent of total tin production in 1912, and 65 per
cent by 1917 (Contreras, 1984, p. 14).
192 Bolivia, 1900±39
First World War

With the exception of the 1907 crisis, which caused the closure of various small
mines and increased unemployment in various sectors of tin mining, the pre-
war period was favourable to Bolivian mining, both in terms of prices and the
continuous increase in demand. However, following the outbreak of the First
World War and the closure of the Metal Stock Exchange in London (August
1914), mining was subject to falling prices and demand until about 1915,
resulting in the closure of many mines and the reduction of wages ± which
until then had been increasing. Moreover, following the interruption in railway
construction and the decrease in production of Chilean nitrate mines, unem-
ployment increased. The mining depression had repercussions on commercial
activity, which soon faced severe credit restrictions. Reduction of production
and exports in Bolivia resulted in a scarcity of foreign exchange and hence a
considerable decrease in imports. This had a negative impact on fiscal income,
since taxes on imports (as a percentage of taxable value) were five times greater
than the tax rate on mining ± the government was forced to emit customhouse
bonds amounting to Bs.10.25 million. In addition, for the first time the govern-
ment forced mining companies to hand over 10 per cent of the value of their
exports in letters of exchange to the National Bank (Contreras, 1984, p. 23).
The impact of the First World War on the modern tin companies was differ-
ent. They had already constructed a solid financial base which allowed them to
suspend or decrease production and to direct their resources to development
tasks and work their tailings so as to provide employment for their manual and
technical workforces. Once the markets reopened, these companies were able to
continue exporting and selling their accumulated stocks and their new produc-
tion, taking advantage of rising prices (Contreras, 1984, pp. 24±31).
One result of the First World War was the change in the direction and
emphasis of Bolivian foreign trade. Before the war (1913), most trade was
carried out with Europe; Britain was the destination of most exports, while
Germany was the main source of imports (see Table 7.2). The War made trade
with Europe more difficult, and in addition encouraged the USA to install its
own smelters and to buy tin from Bolivia directly, regardless of British opposi-
tion. Three smelters were constructed in the USA, the most important of which
was the 8000 short tons capacity smelter built by the American Smelting &
Refining Co. (ASR). The ASR actively participated in insuring the provision of
Bolivian minerals, offering attractive advance payments to mining companies,
advances that contributed to the reactivation of mining in 1916. Furthermore,
American interests also attempted to control the production of minerals by
offering to buy various established companies. Rubber, antimony, tungsten and
other minerals found markets in the USA. Furthermore, the provision of the
local market with letters of exchange from New York ± foreign exchange in
dollars ± rather than London (as the result of restrictions on trade with Europe)
encouraged US imports. This association was strengthened to the point that in

1923, when the US smelters stopped processing Bolivian minerals, the trade
links were already established with such firmness that suspension did not affect
them.4 As demonstrated in Table 7.2, in the period when the USA smelted
Bolivian minerals (1920), the USA became one of the main destinations for
Bolivian exports. However, the importance of the US in the Bolivian export
industry decreased in the 1930s.
Äo
The First World War also had an effect on the large mining companies. Patin
was forced to abandon his interests in Germany (where he sold minerals) and to
develop them in England, where he invested in the Williams Harvey smelter.
On the other hand, the Aramayo Francke Mines ± originally registered in
London in 1906 ± were forced to change residence and move to Switzerland
under the name Compagnie Aramayo des Mines en Bolivie, due to the
increased taxes implemented in Great Britain (Contreras, 1980, p. 64).
Table 7.2 Bolivia: exports and imports by country of destination (% of total value)
(a) Exports
Country 1900 1913 1920 1930 1939
Great Britain 17.4 80.8 44.9 76.5 64.8

United States 0.3 0.6 46.1 13.0 9.1
Germany 11.1 8.5 0.1 2.4 0.5
France 3.8 4.9 0.6 0.4
Brazil 9.0 1.8 1.2
Belgium 2.5 16.9
Argentina 2.2 2.7 1.6
Subtotal 41.6 94.8 95.7 98.7 92.9
Others 58.4 5.2 4.3 1.3 7.1
Total 100.0 100.0 100.0 100.0 100.0
(b) Imports
1900 1913 1920 1930
Great Britain 17.0 20.3 21.3 16.7

United States 6.1 7.4 30.2 27.8
Germany 23.3 36.7 3.1 13.6
France 5.8 3.8 2.7 3.4
Peru 12.8 10.4 5.5
Argentina 7.7 6.0 11.3
Subtotal 72.7 68.2 73.7 78.3
Others 27.3 31.8 26.3 21.7
Total 100.0 100.0 100.0 100.0
Sources: 1900: Oficina Nacional de InmigracioÂn, EstadõÂsticas y Propaganda GeograÂfica (1905), pp.
457, 460±1; 1913: Victor Bulmer-Thomas (1994), pp. 74, 76; 1920: LeoÂn E. Bieber (1984), p. 96; 1930:
DireccioÂn General de EstadõÂstica (1936); and 1939: Antonio Mitre (1996), p. 119.
194 Bolivia, 1900±39
Following the War, the saturation of tin stocks in the US led the government
to prohibit imports until 1919, significantly affecting Bolivian mining. Once
this problem was overcome, tin mining continued to expand, as the short post-
war boom materialised.
The 1920s
After reaching record levels in 1920, tin prices began to fall in step with the
world recession and growing tin stocks (Fox, 1974, p. 115). The depression hit
both the mining industry and the economy at large: in 1921 tin exports
plummeted to 19.086 FMT, down from the previous high of 29 280 FMT in
1918 (see Table 7.1).
Once this crisis was overcome, from 1924 to 1928 tin experienced substantial
growth, with both consumption and production booming. Production rose
throughout the period: in 1929 a record 47 191 FMT were exported, and during
the 1920s Bolivian production came to represent 18 per cent (1921) and 24 per
cent (1929) of world production (LoÂpez Rivas, 1955, p. 115). Between 1920 and
1929, exports increased at an annual rate of 5.9 per cent, surpassing the rate
achieved in the previous period (1910±19, see Table 7.1).
This growth was directly linked to huge investment flows. Between 1920 and
1926, investment in new operations and in the capitalization of existing mines
amounted to US$27 million. Of this total sum, US$12 million stemmed from
the investments of the Guggenheim Brothers in the Caracoles Tin Company in
1920 (Ness, 1938, table XIV, p. 130, and appendix IV, p. 274). In 1925 this
company was one of the most important producers in Bolivia (5960 gross tons),
second only to the Patin Äo Mines (18 827 gross tons) (Mitre, 1993, table 19, p.
116). The Guggenheim Brothers were equally interested in acquiring other
companies, including those of Aramayo in which their investment reached
almost one million dollars invested in 70 000 shares of Aramayo Company
stocks.5 Finally, in 1928, the Guggenheim Brothers reached an agreement with
Aramayo in exchange for a share in the profits. This contract lasted only until
1933, when Aramayo took advantage of the tin crisis and the heavy losses in
Caracoles to buy the mine from the Guggenheim Brothers (Crespo, 1981, pp.
247±8).
The continued growth of the tin mining industry increased its importance in
the national economy, leading to disputes about the contrast between the
relative boom of this sector and the state of crisis of the country. Between
1918 and 1925 there were many criticisms in Congress over the outflow of
capital from the mining sector, criticisms which went as far as suggesting the
need to `nationalize Bolivian mining' and recommending the direct participa-
tion of the state in the production companies (Prado, 1988, pp. 70±1).
In 1925, the mining industry formed the Bolivian Association of Industrial
Miners, which grouped together the principal mining companies along with tin
retailers, in order to promote their interests (Lofstrom, 1972). The Association

was in constant opposition to the Permanent Fiscal Commission ± created to
oversee the loan of 1922. The Commission went out of its way to revise the
accounting records of the mining companies and charged significant sums to
many of them as back taxes. In 1926, the Commission went as far as to appeal
to the Supreme Court in Sucre, resulting in a favourable verdict issuing two
debit notes to Patin Äo and a total of 1.4 million Bolivianos to the then Chilean-
owned Llallagua Tin Company, equivalent to half a million dollars (Contreras,
1990a, p. 269).
The Association was also possibly a response to the increasing labour organ-
ization in the mines and the introduction of the first socially oriented laws in
the country. The eight-hour working day and a series of measures regulating
working conditions were introduced, including requirements for the provision
of medical services, and steps were taken regarding safety conditions in the
mines (Contreras, 1985, pp. 115±20). The Mining Code of 1925 included sig-
nificant advances in labour safety legislation and gave the state greater control
over mining concessions. The more precise definition given to the concessions
helped reduce one of the main causes of litigation in the industry. The founding
of the Association demonstrates both the level of institutional development of
the mining industry and its interest in responding in a corporate manner to
governmental intervention in mining activities.
In 1929, the main mining companies accounted for 77 per cent of total
exports. Patin Äo Mines represented 46 per cent of total exports:6 following
their creation in 1924, production increased significantly as a result of new
investments in machinery for the mills and for the mines, and the introduction
of a new administration (Beard, 1939). These improvements were to be put to
the test during the Great Depression.
The Great Depression and the control of tin

By 1928 overproduction of ores had precipitated a fall in prices. In response, in
July 1929, the Association of Tin Producers was set up in London to voluntarily
restrict production. Tin felt the impact of the New York Stock Exchange crash in
1929 and the ensuing depression in the USA; consumption fell by 11 per cent in
1930. As a result, and given the failed voluntary restriction of tin production,
producers were forced to seek national production quotas through an intergo-
vernmental agreement: the First International Tin Agreement, lasting from
1931 to 1933. It was followed by the Second Agreement (1934±36) and the
Third Agreement (1937±41), administered by the International Tin Council
(ITC) (Fox, 1974, ch. 4).
In Bolivia, the Depression caused various mine closures, a rise in unemploy-
ment and great difficulties for the government whose main source of foreign
exchange ± mining ± had felt the full impact of the economic crisis.7 Between
196 Bolivia, 1900±39
1930 and 1939, tin exports decreased at 3.6 per cent a year, reaching in 1933 a
level of exports similar to 1901±03 (see Table 7.1). Bolivian production fell to 12
per cent of world production in 1937 ± its lowest level since 1904 ± to finish the
decade with no more than 17 per cent (Lopez Rivas, 1955, p. 107).
Towards the end of 1929, mining interests demanded a reduction in railroad
cargo freights and the freedom to readjust salaries. In 1930, the AsociacioÂn de
Industriales Mineros formally requested the end of state intervention in wage
levels, a reduction in railroad freights for mineral exports and imports of inputs,
the creation of the Caja de Seguros e Indemnizaciones in order to alleviate the
burden of social benefits on companies, and the devaluation of the national
currency. The government of Hernando Siles (1926±30) had pledged to support
these measures, but was prevented from carrying them out by its overthrow by
the Military Union in 1930 (Lofstrom, 1972, pp. 63±4).
The Depression led to important changes in monetary and fiscal policy. In
1931, after Great Britain had abandoned the gold standard, Bolivia followed
suit and in addition suspended external debt payments. It was now necessary
for the government to fix the exchange rate. At the same time, given the lack of
foreign loans, it had to find a way of solving its constant budget deficits.
Moreover, as we will see below, the need to finance the Chaco War was an
additional stimulus for an increased participation by the state in the national
economy. In both cases the tin mining industry played an important role and
its behaviour was subject to public scrutiny. The rate at which the exchange rate
would be fixed was a source of conflict between the miners and both importers
and industrialists. Importers were keen on sticking to a parity exchange rate
and lobbied for an adequate provision of foreign exchange. Industrialists,
naturally, would have benefited from a devaluation. However, they also wanted
adequate provision of foreign exchange in order to cover their import needs.
The mining industry, on the other hand, sought a currency devaluation, thus
reducing local costs, and opposed foreign exchange controls. This confronta-
tion effectively alienated the mining companies ± especially the larger corpora-
tions such as Patin Äo, Hochschild and Aramayo ± from the importing industries
and from the manufacturing sector (Whitehead, 1972, pp. 78±9).
Perhaps one of the most important effects of the Depression was that it
created a greater role for government in the mining industry. The government
fixed and enforced export quotas, and the small mining companies and later
the medium-sized mining companies joined forces in the CaÂmaras Departamen-
tales de MinerõÂa and the AsociacioÂn Nacional de Mineros Medianos founded in 1939
in order to carry out the distribution of quotas. Quotas became a source of
conflict both between the larger companies (Patin Äo, Hochschild and Aramayo)
and the medium and small mining companies. Through exposure in the
national press, these disputes helped to destroy the illusion of a cohesive
mining sector in the eyes of both the government and the public. At the
same time, they contributed to a greater awareness within the urban middle
classes of the problems facing the mining sector.
The internal distribution of export quotas was based on the actual tonnage
exported in 1929 by each group or company: the quota was set by the ITC to
coincide with this percentage. During the 1938±39 period, the government
used the 1937 production as the base. These internal quotas were distributed
by the government to each mining group in the case of the large mining
groups; to each company in the case of the medium mining groups ± as defined
by those which produced more than 60 FMT per year, but less than Patin Äo,
Hochschild and Aramayo ± and to the Camaras in the case of small mining
groups (Hallowell, 1948, chs I, II).
The large mining groups closed their most expensive mines and concentrated
their quotas in the highest yielding mines. Despite these measures the restric-
tions placed on production by the export quotas had severe effects on them.
PatinÄo Mines' production fell from 20 800 FMT in 1929 to 8000 FMT in 1932.
According to its general manager, more than half of their mines were closed.
They returned to manual drilling in order to employ the greatest number of
workers, and the lowest levels of mines were allowed to flood in order to avoid
the costs of pumping. The mills worked only one eight-hour shift, five days a
week. Salaries were reduced in step with the reductions in production (Quere-
jazu Calvo, 1978, p. 174).
After 1933, the government took control of 50 per cent of any increase in the
Bolivian quota. This allowed the government to distribute its quota between
the different mining sectors, thereby compensating for any distortions result-
ing from the use of the base year (1929) in the distribution, and/or directly
exporting any increase in order to generate foreign exchange (Hallowell, 1948,
p. 55). This last measure was taken only when foreign exchange was scarce and
as the need to cover the costs of the Chaco War with Paraguay (1932±35)
emerged. The distribution of quotas was also subject to financing contracts
between the mining industry and the state. Between 1931 and 1935, the
mining industry's financing of the state increased to 29.5 per cent of the
national budget.8 In many cases the contracts stipulated that the export
quota of the mining group would increase or would remain the same.
The internal distribution of export quotas was only necessary between 1931
and 1933 when Bolivia exceeded its production quota as dictated by the ITC,
giving way to overexporting resulting from the increase in small and medium
mining production starting in 1932, when prices started to improve. This
occurred to such an extent that in April 1933, for example, the Patin Äo and
Aramayo groups voluntarily conceded part of their quotas in favour of the small
companies. By contrast, between 1934 and 1937 Bolivia was unable to fulfil its
quota and several causes can account for that failure. First, the quota assigned
by the ITC was too large in relation to the long-term production capacity of
198 Bolivia, 1900±39
Bolivia. The percentage of foreign exchange turned over to the government by

mining producers increased from 42 per cent in 1932 to 53.6 per cent in 1934,
and was later reduced to 47 per cent after the Chaco War (Contreras, 1990a, p.
283). In order to stimulate production, the principal source of foreign
exchange, the government saw the need for production contracts with the
mining companies under which the handover of foreign exchange would be
reduced in proportion to the increase in production.
However, the problems of production in tin mining were also caused by the
impoverishment of the ore deposits and the fact that investment had not been
undertaken in new concentration processes since the end of the 1920s (Hallo-
well, 1948, table 8, p. 86). Similarly, labour scarcity, especially of qualified
workers drafted into the Chaco War, was a major drawback (Querejazu Calvo,
1978, p. 175). Only in 1933 were the first qualified individuals released from
mandatory participation in the war; in 1934 the release was expanded to all
individuals employed in the mining industry (Lofstrom, 1972, p. 71). The
scarcity of labour pushed the mining companies to import workers from Chile
and Peru, but these efforts did not have the desired result: foreign miners were
not as productive and, in addition, they caused problems with the domestic
miners because of differences in salaries. Women were employed inside the
mine for the first time in mining history.9 The scarcity of labour was responsible
for decreasing development projects ± that is, exploration and preparation of
mines. Indeed, between 1932 and 1935 all available labour was used in exploita-
tion. As a result, when the market improved and export restrictions were lifted,
mines like Llallagua faced insufficient reserves (Hallowell, 1948, p. 92).
The restrictions imposed on production by the quota system limited new
investment in mining. The only important investments made during this
period were undertaken by Mauricio Hochschild. He spent more than one
million dollars in acquiring and developing the Colquiri mine and in the
construction of a mill and a hydroelectric plant.10 Generally, however, new
investment was also restricted by a reduction in foreign exchange availability ±
caused by the increase in the amount of foreign exchange required to be turned
over to the state at the official exchange rate ± and the uncertainty provoked by
the nationalist measures of the postwar period. These measures included the
nationalization of Standard Oil in 1938, the decree of 100 per cent control of
foreign exchange in 1939, and the incursion of the state into the commercia-
lization of minerals through the creation of the Banco Minero in 1936.
By the end of the 1930s, the tin industry was in a difficult situation. More-
over, as far as the industry was concerned this situation had arisen largely
because of the government's fiscal policies and exchange rate controls. Indeed,
according to ReneÂ GutieÂrrez Guerra, Secretary General of the AsociacioÂn de
Industriales Mineros, tin production had dwindled despite favourable prices,
because of `causes external to the mining industry itself'. These included
financial burdens, taxes, and obligations on tin exploitation [which]

increased to such an extent, aggravated by foreign exchange controls by
the state, that production ceased to be a business even mildly interesting
or resting on secure ground for capitalists.
(GutieÂrrez Guerra, 1940, pp. 17±18)
We analyse this situation and the more extensive implications of government

policies in the following section.
The contribution of mining to development and to fiscal policy
Generally, taxes have been used as the main indicator of mining's contribution
to the national economy. However, we should not lose sight of the fact that
mining activity also had a regional impact: it created a demand for national
goods and for imports that, upon being introduced into the country, paid
import taxes and created wealth in the importing sectors. An example of the
boom promoted by tin mining was the urban, commercial and financial devel-
opment of Oruro, described in 1912 by a French observer as the `home of
industrial and commercial progress', with nine banks, two agencies and various
factories and commercial centres (Walle, 1914, pp. 204±5). Moreover, the role of
the mining industry in the introduction of a cash economy to rural areas has
yet to be fully studied. There is no doubt, however, that tin mining was not the
great driving force for development in Bolivia that it could have been, but
detailed explanation is beyond the scope of this work. However, two of the
main causes were the government's failure to extract greater taxes from
the mining industry, in addition to the inadequate use of these funds by the
government and the main mining companies' failure to reinvest domestically
(see GoÂmez-D'Angelo, 1978, pp. 186±7; Contreras, 1990a, pp. 265±87).11 Some
of the reasons for this second phenomenon are examined later in the chapter.
The value of export taxes as a per cent of total mining exports has been used
as an indicator of mining's contribution to the economy. This percentage
increased from a ten-year average of 2.7 per cent in the 1900s to 4 per cent in
the 1910s, reaching 9.3 per cent in the 1920s.12 However, if we calculate the
percentage as a ratio of taxes to profits, the percentage for 1925 rises from 9.1 to
17.7 per cent (Ali Ayub and Hashimoto, 1985, table 3, p. 12). Similarly, the
value of export taxes as a percentage of total fiscal revenues increased from an
average of 6 per cent between 1901 and 1904 to an average 22 per cent for the
1916±20 period (Contreras 1994, p. 8).13 However, there was debate around
mining's contribution to the economy and on how to address this issue.
Until the early 1920s, export taxes, on a sliding scale in relation to price
fluctuations, accounted for the bulk of taxation levied from the mining indus-
try. In 1920, faced with financial difficulties, the government introduced the
200 Bolivia, 1900±39
first tax on profits. But it was not until 1923 that the government of Bautista
Saavedra (1921±25) introduced a new profit tax to cover the financial obliga-
tions of the `Nicolaus' loan, taken out in 1922 to help refinance the public debt.
This tax clearly established the amount that could be deducted in order to
determine profits. The introduction of this law and the existence of a capable
entity for tax collection (the Permanent Fiscal Commission), significantly
increased taxation on the mining industry (the percentage of total taxes ± of
exports and profits ± in relation to the value of exports doubled: from 7.4 per
cent in 1923 to 15.6 in 1924).14 According to GoÂmez-D'Angelo, the 1920s stand
out as the decade during which the tax system was able to extract an ever-larger
proportion of increases in production (GoÂmez-D'Angelo, 1978, pp. 142±6).
However, as the government turned to taxes on profits in the 1920s, and
foreign exchange controls in the 1930s and thereafter, as the main sources of
resource extraction from the tin industry, the relative importance of the export
tax on tin decreased gradually.
The increased taxes on the mining industry during the 1920s and 1930s
reflected the shift in the government's fiscal policy away from import tariffs, a
feature of the 1900±10 period, towards a greater reliance on export taxes.
Although this policy was due, in part, to the state's greater financial needs, it
is still significant that there were no attempts by the government to increase
taxes on agriculture which was ± and still is ± the principal economic activity of
the country (Gallo, 1990, pp. 102±4).
Table 7.3 illustrates the relative importance of the taxes paid by the mining
industry and the greater absolute amount èxtracted' from the sector by the
state during the 1930s. Based on data from the Patin Äo Mines, Table 7.3 demon-
strates that taxes on exports represented 50 per cent of total taxes paid during
the 1920s and 30 per cent of the total taxes over profits. The relative weight of
both taxes decreased significantly in the second half of the 1930s, when taxes
Äo mines, 1924±39
Table 7.3 Bolivia: detail of average taxes for selected periods in Patin
(£ 000s)
Period Differential Import and

Export tax exchange-rate tax machinery taxa Other taxes Total
£ % £ % £ % £ % £ %
1924±29 205.1 50 119.8 30 73.5 16 16.8 4 415.1 100

1930±34 69.8 42 7.5 4 55.2 32 40.3 23 172.8 100
1935±39 74.6 19 19.3 5 38.5 10 46.0 12 394.8 100
a
Includes import taxes on machinery and production goods, on edible goods, and also on urban
properties.
Äo Mines Archives, COMIBOL (1942±43), `Cuadro
Sources: Calculated from information in the Patin
demostrativo de impuestos fiscales', file 1323.
on the exchange rate represented more than half of total taxes collected. This
reinforces our assertion that export tax revenue ± whose importance decreased
significantly after the 1920s ± is an inadequate proxy for estimating the true
contribution of the mining industry to fiscal resources.
Not surprisingly, the mining industry was opposed to all of these measures.
Precisely when it was thought taxes would be increased, Patin Äo, taking advant-
age of the shortage of funds and the regional pressures to complete the railway
between Sucre and PotosõÂ, obtained a five ±year moratorium on tax increases in
exchange for a £600 000 loan (Contreras, 1990a, pp. 270±2).
Äo offered the loan to the government for the first time on 6 August 1924,
Patin
after purchasing the Llallagua Tin Company from Chilean interests and regis-
tering PatinÄo Mines and Enterprises Consolidated, Inc. in Delaware, USA. This
acquisition completed the purchase of Llallagua Company stocks, a process
begun in 1914, through the British firm Duncan Fox and later through the
Anglo South American Bank. Indeed, the purchase of the Llallagua Tin Com-
pany ± the principal tin producer of the country ± and the later establishment of
PatinÄo Mines in the USA are two landmarks in the development of the tin
mining industry and of Bolivian capitalism. Patin Äo was the only Bolivian
capitalist (and one of the few from the periphery) who was able to build a
true transnational company ± with interests in four continents and more than
five countries ± starting from his own country.15
Bolivia was a country without a significant external debt before 1908. From
1909 to 1930, government indebtedness rose to more than US$80 million
(Ness, 1938, table 8, p. 104). More than 80 per cent was acquired in the
1920s, as the public debt tripled during this period and external loans replaced
internal loans. From 1930 to 1935 ± and especially after Bolivia defaulted in
1931 ± internal debt in Bolivia replaced the external debt, and the total debt
tripled. The final destination of the externally acquired funds cannot be deter-
mined with exactitude: Ness's estimation is that between 1909±30, approx-
imately 40 per cent of the net financial resources was utilized for the
construction of railways and 16 per cent was directed to what he terms `general
purposes of the state' ± which included everything from the Vickers contract for
military armaments, to the nation's fiscal deficit, to the payment of public
loans: 8 per cent went to public services, 5 per cent to `financing purposes' ±
in order to establish the Banco Nacional Boliviano in 1911, which was converted
into the Banco Central in 1929 by recommendation of the Kemmerer mission,
and the rest to refinance the internal debt (Ness, 1938, pp. 104±9). In contrast,
during the 1920s, railways were responsible for 80 per cent of acquired loans.
The mining industry was opposed to the external debt acquired by the
government, since it associated loans with increased taxes (US National
Archives, File 824.51/307, 1 April 1925). In 1924, Congress was contemplating
imposing an integral tax on exports, thereby eliminating the recently created
202 Bolivia, 1900±39
profit tax. Congress argued that by increasing their declared capital, mining
companies were paying between one-third and one-half of the taxes they
owed.16 In order to defend themselves from future increases in taxes ± particu-
larly now that mechanisms were in place to forcefully impose and collect them
± mining companies created lobby groups in Congress that opposed tax legisla-
tion through the AsociacioÂn de Industriales Mineros. The Association hired pro-
minent members of the Senate as lawyers (US National Archives, 28 November
1924, p. 23). While the 1920s experienced the development of an efficient
system to tax the mining industry, in the 1930s the state was able to substan-
tially decrease the export of profits through the mandatory payment of foreign
exchange, and, in paying the mining industry for its foreign exchange at the
official (overvalued) rate, managed to impose an additional tax on them. This
implicit tax rose from 12 per cent of the value of mining exports in 1936 to 17
per cent in 1938, decreasing to 14 per cent in 1939, which meant that the
percentage of taxes coming from mining during this decade increased to 14
from 9 per cent in the previous decade.17 In Patin Äo Mines, the tax resulting
from exchange controls during the second half of the 1930s exceeded the
export and profit taxes, and represented 59 per cent of total taxes paid (see
Table 7.3). In addition, the mining industry had to take a loss over its accumul-
ated balances due to the devaluation of the Boliviano, from Bs.19.6 per pound
sterling in 1933 to Bs.141 in 1939.
However, the actual impact of these measures for the country's development
was not as expected. Large quantities of money in Bolivianos were accumulated
in Bolivia without apparently finding productive investment. For example, in
1933 the Banco Mercantil had accumulated £650 000 in current accounts and
had to suspend interest payments on deposits `regardless of their time period or
condition' (Banco MercantõÂl, 1934, p. 2). The increase in foreign exchange
payments allowed Patin Äo Mines, for example, to accumulate Bs.19 million in
Bolivia in 1935 ± equivalent to 34 per cent of income and 10 per cent of
national expenditure ± which it could not put to productive use (Patin Äo
Mines and Enterprises Consolidated, 1935; Banco Central de Bolivia, 1937, p.
104). PatinÄo Mines tried to invest these accumulated balances in several areas:
in mining, in diversification and in loans. The investments made in the mining
sector were concentrated in the exploration of new mining properties and new
minerals, especially in gold, and in the purchase of stocks in existing compan-
ies. The principal efforts at diversification consisted of investment in the colon-
ization of the Chapare which, it was hoped, would yield valuable woods for
mining struts and a cattle industry to provide meat for the mining industry ±
meat had until then been imported mainly from Argentina. Unfortunately,
these projects met bureaucratic obstacles and on the whole proved to be fail-
ures. PatinÄo Mines also attempted to invest in the cement factory of Viacha, but
was not able to achieve a satisfactory agreement with its American owner, the
Foundation Company, on what type of exchange rate would be applied to the

amount that Patin Äo was to invest. Patin
Äo Mines invested in the CreÂdito Hipote-
cario and in the Banco Minero, which had been recently created in 1936. Finally,
PatinÄo Mines made loans to banks and to other mining and agro-industrial
companies in which it had interests.18
Further study is necessary to fully understand this exceptional `saturation of
capital'. It demonstrates the Bolivian economy's small capacity for absorption
of capital and the aversion to risk of both national and international investors.
As such, it undermines the axiom ± implicit in many studies on the pre-
nationalization mining industry ± that greater taxes on the mining industry
would have resulted in greater development. It also questions the belief that the
mining industry did not fulfil that èntrepreneurial function' that is required of
the export sector and, according to Gallo, in particular of the Bolivian tin
industrialists (Gallo, 1990, p. 91). This experience suggests that when the
industry wanted to fulfil this function, it was constrained by structural factors.
At the same time, it is worth asking whether the attempt to diversify was not a
consequence of control having been in the hands of a Bolivian capitalist, as in
Äo Mines. We will return to this point in the conclusion.
the case of Patin
The monetary and fiscal policies that the Depression and the war pushed the
government to adopt, resulted in a long inflationary process during the 1930s in
addition to the effects on the mining industry. In the first half of the 1930s the
main problem was the need to finance the Chaco War through the Banco
Central ± the internal debt caused by the war reached Bs.400 million. As a
mechanism for reducing inflation, subsidies were given via exchange rate con-
trols to the producers of basic goods and for imports of essential articles (United
Nations, 1958, ch. IV). The establishment of differential exchange rates resulted
in the creation of an Exchange Control Board that assigned foreign exchange to
the various sectors. As expected, this had a perverse effect: commerce grew
disproportionately and merchants multiplied as if by magic, fronting artificial
businesses that had, as their principal activity, foreign currency speculation. As
a result, many young professionals with university degrees became merchants
(Mendoza LoÂpez, 1940, p. 160).
In addition to the commercial boom, industry also grew. Direct fiscal protec-
tion, monetary policy and the provision of foreign exchange at a preferential
rate were responsible. In 1939, 51 per cent of the industrial sector was made up
by the textile and clothing industry, while beverages accounted for 18 per cent,
foodstuffs 9 per cent and electricity another 6 per cent (GutieÂrrez Guerra, 1940,
pp. 29±30).
In the second half of the 1930s, inflation increased. The annual average cost
of living increase in La Paz for this period was 51 per cent, while for the
period 1932±35 it was 17 per cent. This was due to the fact that the govern-
ment resorted to increasing the money supply to cover the costs of the
204 Bolivia, 1900±39
demobilization after the war and, principally, for public works. Likewise, the
state's entry into the productive sector às the result of the political need and
the ideas for renovation of the country that had vigorously surfaced as a
reaction to the structural faults that the war had emphasised' (United Nations,
1958, ch. IV, p. 62) was the cause of greater public expenditure. In the following
section we analyse the investments made by the state.
Railways and roads
The lack of a communications infrastructure has been a constant of Bolivian

history. Although infrastructural development began in the nineteenth cen-
tury, the establishment of a railway system and the construction of roads and
highways were primarily features of the twentieth century. While the motives
and circumstances under which both projects were undertaken differed, in the
following paragraphs we outline the development of the Bolivian railway
system in order to later summarize the boom of road and highway construction
in the 1930s.
Railways constituted the first and principal link to the Pacific Ocean. The
railroad from Ollague (on the border with Chile) to Uyuni, completed in 1889,
which united the country with the port of Antofagasta, and the Uyuni±Oruro
railway, completed in 1892, had a positive effect on the development of silver
mining; they cut transport costs and contributed to production increases,
thereby consolidating the economic boom that took place towards the end of
the century despite an unfavourable international market.19 In addition, they
established the foundations for the era of tin.
There was great popular demand for the construction of railways in the
twentieth century. Such construction was necessary for the political survival
of the liberals and the republicans: railways had an aura of modernity and
development that made them indispensable in all government programmes.20
During the first third of the twentieth century, we identify two distinct
periods. Between 1900 and 1915, railways were constructed to link various
regions of the country with the Pacific. These railways were oriented to the
external market, and were strongly associated with mining. In the second
period, starting after 1915, the principal objective was to establish communica-
tion between the principal cities and their regions. The interests that promoted
such development during the second period were not linked directly to the
mining sector.
In the first 15 years of this century, under the impulse of the liberals, approx-
imately 1000 additional kilometres of railway were constructed. Unlike animal
transport, rail transport was not subject to volume restrictions, it reduced
general transport costs (both in terms of quality of transport and time, as well
as freight) and facilitated the import of machinery.
During the second period, after 1915, more railroad systems were initiated
than were finished. The railway between La Paz and Yungas ± with Beni as final
destination ± begun in 1915, was abandoned in the 1950s after less than 80
kilometres of track had been laid. The Sucre±PotosõÂ railway, initiated in 1916,
was only finished in 1943. Finally, the Cochabamba±Santa Cruz railway, begun
in 1928, was also left incomplete. The last two railways were the source of
intense regional and national debate since they represented important regional
aspirations (RodrõÂguez, 1992). It is evident that these railways are markedly
different from those built during the first period, since the former had the
underlying motivation of linking domestic areas.
Railroads completed in this period included the Atocha±VillazoÂn railway,
begun in 1915 and finished in 1925, and the Oruro±Cochabamba railway ±
although initiated during the first period (1906), it marked a turning point in
railway construction between the phase of externally-oriented railways and the
phase of internal integration. Indeed, this railway was the first effort to estab-
lish a line of communication between two important cities.
Finally, there were several frustrated attempts to construct railways of ìnteg-
ration' (in contrast to èxport' railways), primarily financed by private capital;
the notable exception being Patin Äo's Cochabamba±ChimoreÂ railway project
aimed at colonizing the Chapare. After almost two years of discussion, the
Bolivian Congress decided to deny Patin Äo the railway concession (Geddes,
1972, pp. 152±67). We underline this fact because it illustrates the private
sector's interest (in this case the mining sector's) in developing railways away
from Bolivia's borders and focused, instead, on promoting regional develop-
ment, an aspect that is ignored by historiography.
Another interesting example is the Sucre±PotosõÂ railway. Begun with domes-
tic funds, in the 1920s it became partly responsible for growing foreign indebt-
edness. It is singled out because it a good example of a `political' railway. It was
not built as a response to economic criteria (it was not very profitable); rather,
construction was carried out in order to satisfy political interests: important
regional interests demanded that a railway be built, and it was initiated in 1916
following many debates. By 1922 it was evident that the volume of traffic was
not sufficient to pay for the interest on the loan to cover construction costs
(McQueen, 1925, p. 108).
It is perhaps not surprising, once we take into account the political and
economic necessity of railway construction, that railways were responsible for
approximately 40 per cent of Bolivia's external debt between 1900±30 (Ness,
1938, table 8, pp.107). However, the railways were not only financed with
external debt: compensations received by Bolivia following the wars in the
Pacific and Acre were an additional source of finance. As a product of the
PetroÂpolis Treaty (1903) the country received £2 million, which was utilized
by the Bolivian Railway (formed in 1906) to construct the main railways of the
206 Bolivia, 1900±39
country. Similarly, through the Peace Treaty with Chile in 1904 the Chilean
government pledged to finance the railway between Arica and La Paz, which
was finished in 1913. Finally, as we mentioned previously, private capital also
invested in railroads.
Railroad construction stopped following the suspension of external credit to
Bolivia and the advent of the 1930s economic crisis. Roads took over. Bolivia
had no road system at the beginning of the century, and the few roads that did
exist had been constructed towards the end of the nineteenth century around
the country's principal cities and the adjacent rural populations. These were the
municipal roads. The national roads connected the principal cities and the
mining centres. The majority were dirt roads. By 1941, however, there were
11 000 kilometres of roads, although only a fraction of these could be used
throughout the year (Oficina Nacional de InmigracioÂn, EstadõÂsticas y Propa-
ganda GeograÂfica, 1905, pp. 464±7; Contreras, 1990b, p. 193).
In the 1930s, the Central Bank served as the principal driving force behind
road construction. In 1933 it presented a plan of road and public works to the
legislative body to `mobilise capital that would strengthen businesses in gen-
eral, giving work to thousands and increasing the possibility for wealth in the
diverse regions of the motherland'. It was hoped that the initiation of the
public works would be à powerful stimulus for the productive districts, that
would intensify interdepartmental commerce of products, thereby invigorating
the activities of businesses in domestic markets' (Banco Central de Bolivia,
1934, p. 14). This initiative lasted until 1939, by which time the Central Bank
had handed out more than Bs.100 million in municipal and departmental loans
(Banco Central de Bolivia, 1940, p. 74).
Education
Political instability and slow economic growth were responsible for the limited
development of Bolivian education in the nineteenth century. By the twentieth
century there were only a few hundred schools and less than 50 000 students in
a country of 1.7 million people. The rate of illiteracy was 83 per cent; and the
education budget amounted to 5.9 per cent of the total budget. It was in this
context that the Liberals came to power, armed with an education project.
Ismael Montes, who governed the country from 1904 to 1909 and then from
1913 to 1917, is perhaps the clearest example of the new thinking on education
in this period. With the support of various Ministers of Education who pledged
to change the Bolivian education system, Montes achieved the first education
reform of the century. The principal aspects of the reform were,
1 improvement of the administrative system of education

2 initiation of teacher-training programmes in Bolivia
3 giving priority to primary education in urban areas

4 development of a curriculum for primary and secondary education
5 strengthening indigenous education
6 promotion and provision of education for women
7 commercial and technical education
A characteristic of this reform process was the search for educational models in
Europe, going as far as hiring Belgian pedagogues to direct the first teacher-
training schools.
The administration of the education system was strengthened with the crea-
tion of the Direcciones de EducacioÂn, and with the state's gradual takeover of
educational responsibilities from both the municipalitiesand the Catholic
Church. Until then, primary schools were under the sole guidance of the
municipalities, and there was no national coordination of these schools.21 For
the directors of the reform it was not simply a question of increasing infra-
structure, but also of organizing the little that existed. In some cases the state
was able to recover the municipal schools with municipal funds, but in
the majority of cases the state had to dedicate new funds to education. During
this period the education budget increased from Bs.382 724 in 1900 to
Bs.4.083 million in 1924. The number of public schools rose from 84 in
1900 to 614 in 1924 (Contreras, 1990b, p. 316). Meanwhile, the reformers did
succeed in developing standard curricula at the national level.
Bolivia had no teacher-training system until 1909. There were few teachers,
and most had been educated outside the country and/or were foreigners ±
usually Chilean or European. In order to increase the number of teachers, the
liberal government first sent Bolivians on scholarships to train abroad, until in
1906 it formed the Normal School for Teachers in Primary Education in Sucre,
in 1915 the first Rural Normal School in Umala, La Paz, and in 1916 the first
institutes to train teachers in secondary education. In effect, the `Normal
School of Sucre was the base of the entire Bolivian education reform' (Minis-
terio de EducacioÂn, 1917, pp. 58±59). Not only did it help develop teacher-
training in Bolivia, but also, in hiring European pedagogues to do so, a new
teaching system was developed that sought to instil in pupils a `solid and
extensive' knowledge, to form their `scientific spirits' and to prepare `men
and women of initiative' (Ministerio de EducacioÂn, 1917, pp. 58±59).
From another perspective, the education of indigenous individuals contained
a clear `civilizing' concept and sought to `Castilianize' them in order to incor-
porate them into national life. The first action by the liberal government was to
establish a system of travelling schools with specially chosen teachers to visit
various ayllus, or indigenous communities. In order to give greater continuity
to this process, permanent schools were established; indeed, many were created
and developed by the indigenous populations' own initiative.
208 Bolivia, 1900±39
The reform established mixed education in the standard schools, and under
liberal rule the first lyceums for girls were set up in La Paz in 1917. The reform
also established technical education in the cities, and commerce schools
were set up to train accountants and skilled office workers. In PotosõÂ and
Oruro, the first surveying and mining schools were created which would later
serve as the foundation for mining engineering schools.22
The liberal reforms addressed the principal aspects of Bolivian education.
It was a top-down reform, in which European influence was significant,
and arose out of decisions taken by the governing elite and visionary education
ministers. The liberals established the foundations of the education system
that was later subjected to the fluctuations of the Bolivian national economy.
Although the budget for the education system increased in the 1920s, in
the 1930s it was tightly restricted by the economic crisis. However, in 1930
an administrative reorganization of the education sector was attempted
with the creation of the National Council of Education. This council was
composed of the Director General of Education and three general inspectors,
and presided over by the Education Minister. Its mandate was to direct public
education, supervise private education, appoint and promote teachers and
manage the education budget. This attempt gave considerable autonomy to
the education sector by placing policy in the hands of individuals with
educational or administrative qualifications, instead of in the hands of a
politically nominated minister. However, the main weakness of the proposal
was that the Council never achieved financial autonomy from the Ministry.
Nevertheless, the attempt is a clear indication of pro-education initiatives at
this time. In 1939, the education budget reached levels similar to those at the
end of the 1920s, representing 8.4 per cent of the total budget (Nelson, 1949,
pp. 8, 11).
In the 1930s, indigenous education also grew. In contrast to the 1920s, in
the first half of the 1930s this expansion was due more to the indigenous
population's own efforts and to the efforts of ecclesiastical organizations
than to efforts by the state. Schools were part of `the struggle to brake the
expropriations of communal lands and to restore previously usurped com-
munities' (Soria Choque, in Choque et al. 1992 pp.41±78). The Warisata ayllu
school developed in this context. Founded by Indian communities in 1931,
it reflected the Indian communities' needs by providing practical training in
agriculture and crafts. It became the model of Indian education had a profound
impact on Bolivian education and is widely regarded as a genuine effort at
reform `from below'. The Chaco War interrupted this effort, as it recruited
teachers and Indians, and children left school to take care of the fields
(Contreras, 1999).
After the war, the Toro (1936±38) and Busch (1939) administrations supported
indigenous education vigorously. Despite this support, Warisata was closed
down because landowners felf threatened by its empowering effect. Indigenous

education stopped being perceived as a `civilizing' element and became an
instrument of national integration, and as such became integrated with the
rise of the unionized agrarian movement (Choque, 1992, p. 76) in the 1940s.
Conclusions
Bolivia responded to the new trends in international demand. The develop-

ment of the Bolivian economy during the first four decades of this century
rested on the growth of the tin industry. A modern sector developed and
structured itself in such a manner that it was able to come out gracefully from
the problems precipitated by the First World War, the Great Depression and the
Chaco War. Despite the pessimistic commentaries at the end of the third
decade, the sector entered the 1940s with great possibilities. Even though the
grades of mineral deposits had diminished, labour had become more expensive,
and the tax situation was more exacting than at the beginning of the century,
the tin mining industry believed that a more favourable exchange rate would
improve its situation. Contrary to popular belief, significant investment had
taken place during this time, and if there had been greater security over the
destiny of such investments and if international demand had allowed it, there
would have been greater investment, as later demonstrated during the Second
World War.
An important characteristic of the development of the Bolivian export sector
during this period is that, after 1924, the main company was controlled by a
Bolivian capitalist, SimoÂn I. PatinÄo. In addition, the third most important
company, the Aramayo Mines, also featured significant Bolivian participation.
What implications did this have for the development of the sector and for its
relation to the state? In terms of development of the sector, the alliance with
foreign capital and the dependence on external technology and foreign engi-
neers meant that the implications were not great. Probably, the participation of
Bolivian capitalists promoted the formation of an association that would
defend its interests in a corporate manner.
With respect to its relation with the state, it was similar to the relation that
would have existed had the interests been completely external. Where we may
speculate that there was a difference was in its interest in diversifying its
investment and its attempts to invest in other activities. The investments in
electric companies, railways, commercial companies and banks would not have
been undertaken by external investors. It is important to emphasise that these
activities were developed almost exclusively by Patin Äo. Other companies
behaved more like external investors.
Before the Great Depression, Bolivia benefited from the availability of inter-
national credit and loans and achieved a significant process of modernization.
210 Bolivia, 1900±39
Although this process was of greatest advantage to urban centres, rural areas
benefited from the expansion of education and the development of infrastruc-
ture in railways and roads. The different governments during this period were
able to increase the resources extracted from the mining sector, despite the
increasing organization of this industry. Once the possibility of depending on
the external sector had gone and confronted with the need to finance the
Chaco War, the Bolivian government was able to shift its taxation policies.
Through the mandatory payment of foreign exchange and through differential
exchange rates, the government was able to extract greater resources from the
mining sector.
Contrary to what we might expect, the reduction in the outflow of dividends
and the accumulation of resources in Bolivianos in the country did not result in
a greater level of development. The true barriers that opposed growth and
development were the lack of viable investment opportunities for the private
sector in the 1930s, and the squandering of public funds caused by corruption
and investment in nonproductive, although prestigious, projects in the 1920s.
Both cases point to the structural problems fundamental to the Bolivian eco-
nomy, which to a greater or lesser extent persist even today.
In changing its fiscal policies, the government demonstrated its capacity to
redirect its actions and adapt to new situations, both in the formulation of new
policies and in their implementation. Nevertheless, it is fitting to ask if there
were alternatives to the policies carried out during this period. According to
Gallo (1990), confronted by the external debt the government had little alter-
native but to increase taxes on the mining industry. Clearly this is correct, but
in order to understand the manner in which it was done ± through a single
differential exchange rate and the mandatory payment of foreign exchange ± it
is necessary to understand the influence that the Chaco War had in the case of
Bolivia. This conflict was a powerful stimulus for the state to take on a more
active role, and once exchange controls were imposed during the war they were
not rescinded after the end of the conflict, as they were perceived to be effect-
ive. In addition, there existed an institutional mechanism to carry out such a
policy ± the Banco Central ± while the apparatus for collecting taxes remained
deficient.
Moreover, there is little doubt that the state found it easier to use this
mechanism than to create new taxes for the mining sector. From this perspect-
ive, the Bolivian case seems to confirm Hirschman's dictum that where the
export sector is powerful and influential, it is politically difficult to increase
taxes and, in this context, the great advantage of an ìnflation-overvaluation'
arrangement is not only that it has an indirect influence on politically and
socially influential groups, but that the mechanism also has greater legitimacy
given that it has the respectable aim of `defending the national currency against
depreciation' (Hirschman, 1971, p. 118).
In attempting to resolve the conflicts over distribution of resources generated

by the export sector, the state stimulated imports and supported the growth of
the industry. However, inflation was a negative factor for development due to
its effects on income distribution, distortions in investment and the erosion of
savings.
After the Chaco War we observe the emergence of a more active state. In
addition to monetary policies, this new role is illustrated in the attitude of the
Banco Central in financing public works, and also in the entrance of the state
into the productive sector ± through the formation of the YPFB (Yacimientos
Petroliferos Fiscales Bolivianos) the establishment of the Mining Bank to com-
mercialize minerals and to provide support to the small mining companies. In
the first case, the development of the oil sector cost the Banco Central Bs.11
million, and the purchase of the stocks of the Mining Bank Bs.46 million
(United Nations, 1958, p. 62).
As stated at the beginning of this chapter, one of our objectives has been to
demonstrate that the period under review was not only, as export economies
are frequently stereotyped, one of export-oriented growth. Bolivia developed a
transport infrastructure. While it is undeniable that this was subservient to the
export model, part of the railway and later road expansion helped establish an
economic backbone. Moreover, during the first 20 years of this century the
foundations for the Bolivian education system were established. The state and
the Central Government recovered their control over pre-university education,
established standardized schools in urban and rural areas, and assigned greater
resources to education. This system undoubtedly set the base on which the
development of citizenship would later be constructed. The case of Bolivia
confirms what is observed in countries such as England, France, and the USA,
where the education of the masses emerged neither by popular demand nor by
the action of market forces, but instead from the characteristics of the state and
their development. Similar to other experiences in Latin America, the role of
the state was important in the constitution of the national education system
(Green, 1992; Ratinoff, 1994, pp. 22±35; and Newland, 1994).
Bolivia entered the 1940s with a stronger state that had resolved problems of
territoriality, developed institutions and extended its administrative capacities,
and undoubtedly had greater control over its resources. These are the three
characteristics proposed by Whitehead for identifying the modern state (White-
head, 1994). Without pretending to suggest that the Bolivian state was at this
time a modern one, the foundations for this modernization were established
during this period. The proof of the level of development of the state is to be
seen in the following decades, where it was to resolve problems such as
exchange rates, internal debt, access to external credit, and the redefinition of
its policies towards the mining sector.
212 Bolivia, 1900±39
Notes
1 Data taken from Bieber (1984, tables II.5 and II.10, pp. 20 and 32); on the develop-
ment of rubber production see Gamarra (1993, pp. 23±76).
2 Fine metric tons (FMT) is the actual metal content of the ore, as opposed to gross
tons.
3 Mitre (1993, table 11, p. 212) estimates that for 1928, national capital represented 41
per cent, US capital 39 per cent, and Chilean capital only 20 per cent, of total mining
investment.
4 In effect, `bringing Bolivian minerals to the US not only meant cheaper tin and
employment for thousands of American workers, it also meant the creation of a
market for products from the US that . . . could reach $ 50 million annually' Daily
Commerce Reports, vol. 1, no. 4 (6 January 1916), p. 53, cited in Contreras (1984, p. 34).
5 See the correspondence between Carlos VõÂctor Aramayo and Carlos Navarro in the
Navarro Papers in the La Paz Archives. These papers contain many letters covering the
period from 1919 to 1928.
6 Calculation from Bolivia, 1936, El estanÄ o en Bolivia 1935, La Paz, table 7.
7 Concerning the effects of the Great Depression on the Bolivian economy, see White-
head (1972), Pacheco (1990) and Contreras (1990a).
8 Calculated from data from the General Budget from the Banco Central de Bolivia
(1937, p. 104).
9 Archivo Patin Äo Mines-CorporacioÂn Minera de Bolivia, `Turnover of labor'. Concern-
ing the role of women in Bolivian mining during the nineteenth and twentieth
centuries, see Gustavo RodrõÂguez Ostria (1991, ch. IV).
10 See Altshuler (1939, pp. 41±7) for a technical description of the mine. Letter from the
vice-president of Mauricio Hochschild S.A.M.I to the president of the Commission of
Nationalization of Mines dated 8 August 1952, in Informe de la ComisioÂn de Naciona-
lizacioÂn de Minas.
11 During the 1920s and 1930s, Contreras notes that `the real barriers to the growth
were the absence of opportunities for investment for private capital in the 1930s, and
the waste of public resourses due to corruption and the investment in prestigious yet
non-productive projects during the 1920s' (1990a, p. 287).
12 Calculation from United Nations (1958, p. 10). The value of exports is an estimate
from customs and includes part of the added value of transport and smelting. This is a
proxy indicator.
13 The later decrease is explained by the fact that after the 1920s, export taxes were no
longer the only taxes levied on the tin mining industry. First a tax on profits on the
industry was imposed and later an ìmplicit tax' was introduced by establishing the
mandatory payment of foreign exchange using an overvalued rate of exchange, as
detailed in the following paragraphs.
14 Concerning the `Nicolaus' loan and the ComisioÂn Fiscal Permanente, see Contreras
(1990a); data from GoÂmez-D'Angelo (1978), table 13, p. 204.
15 See Contreras (1994, pp. 26±8), for the emergence of the Patin Äo èmpire'.
16 The tax rate, according to the Law of 1923, was fixed by establishing the relation
between profits and the declared capital. The greater the relation, the greater the tax
rate. Therefore, by increasing their declared capital the mining companies could
effectively reduce their tax rates.
17 GoÂmez-D'Angelo (1978) implicit tax in table 18, p. 217, averages calculated from
table 13, pp. 203.
Äo Mines Archives ± COMIBOL: Informes sobre propiedades mineras recibidas del

18 Patin
Departamento de Exploraciones.
19 See Antonio Mitre (1981, p. 178), who concludes ` . . . the railroad, more than any
other change in the sphere of production, was responsible for the slow growth
observed in the production of silver after 1889 . . . prolonged the yields of mining
activity in moments in which the crisis of the international price of silver conspired
against it.'
20 See, for example, the declarations of Ismael Montes in 1904 cited in Juan AlbarracõÂn
Millan (1972, pp. 90). Similarly, according to Margaret Marsh (1928): ` . . . no govern-
ment that does not attempt to promote the development of railroads could hope to
retain the control of public issues in Bolivia' (1970, p. 97). Herbert S. Klein (1971),
with the advantage of hindsight, confirms that railroads were ` . . . a concrete land-
mark of ``progress'' in Bolivian politics since the days of Arce (1888±92)' (1971, p. 84).
21 At the beginning of the century more than 90 per cent of the educational establish-
ments were under the power of the municipalities (RauÂl CalderoÂn Jemio, 1994, pp.
53±83).
22 Concerning the development of the engineering of mines during the first half of this
century, see Contreras (1990b).
References
AlbarracõÂn Millan, J. (1972) El poder minero en la administracioÂn Liberal, La Paz: Editorial

Akapana.
Almaraz, S. (1976) El poder y la caõÂda. El estan Ä o en la historia de Bolivia, series Historia
EconoÂmica y Social, La Paz & Cochabamba; Los Amigos del Libro.
Ali Ayub, M. and Hashimoto, H. (1985) The Economics of Tin Mining in Bolivia, Washington
D.C: The World Bank.
Altshuler, H. I. (1939) CompanÂaÄõ de Minas de Colquiri, Ministry of Mines and Oil, BoletõÂn, 1,
Ä o Mines Archives, La Paz: COMIBOL,
Patin
Aramayo, C.V. and Navarro, C. (1919±28) correspondence in the Navarro Papers, 1942±43
Cuadro demostrativo de impuestos fiscales, file 1323 in the Archive of La Paz.
Banco Central de Bolivia (1940) UndeÂcima Memoria Annual presentada al Ministerio de
Hacienda y Estadistica, 1939, La Paz: Imprenta Artistica, sucursal A.H. Otero.
ÐÐ (1937) Octava Memoria Annual presentada a los Accionistas y a la Superintendencia de
Bancos, 1936, La Paz: Imprenta Renacimiento.
ÐÐ (1934) Quinta Memoria Anual presentada a los Accionistas y a la Superintendencia de
Bancos, 1933, La Paz: Imprenta Renacimiento.
Banco MercantõÂl (1934) Novena memoria que presenta el directorio a la junta semestral de
accionistas, La Paz: Escuela Salesiana de Artes GraÂficas.
Beard, R.R. (1939) `Property and Operations of Patin Äo Mines and Enterprises at Llallagua,
Bolivia', reprinted from data from the Engineering and Mining Journal.
Bieber, LeoÂn E. (1984) Las relaciones econoÂmicas de Bolivia con Alemania 1880±1920, Berlin:
Colloquium Verlag.
Bolivia (1936) El estanÄ o en Bolivia, 1935, La Paz: Imprenta Renacimiento.
CalderoÂn Jemio, R. (1994) `La ``deuda social'' de los liberales de principios de siglo: una
aproximacioÂn a la educacioÂn elemental entre 1900 y 1910', DATA, Revista del Instituto de
Estudios Andinos y AmazoÂnicas, vol. 5.
214 Bolivia, 1900±39
Choque, R. et al. (1992) EducacioÂn indõÂgena: ¿ciudadanõÂa o colonizacioÂn?, La Paz: Aruwiyiri.

ComisioÂn de NacionalizacioÂn de Minas (1952) Ìnforme sobre la Cia. Minera Unificada del
Cerro de PotosõÂ', unpublished manuscript, La Paz.
Contreras, M.E. (1999) `Reformas y desafios de la educacioÂn' in Campero F. (ed), Bolivia en
el sglo XX. La formacioÂn de la Bolivia contemporaÂnea, La Paz: Harvard Club de Bolivia.
ÐÐ (1994) TecnologõÂa moderna en los Andes. MinerõÂa e ingenierõÂa en Bolivia en el siglo XX, La
Paz: Biblioteca Minera boliviana/ILDIS.
ÐÐ (1990a) `Debt, Taxes and War: The Political Economy of Bolivia, c. 1920±35', Journal
of Latin American Studies, vol. 22.
ÐÐ (1990b) `The Formation of a Technical Elite in Latin America: Mining Engineering
and the Engineering Profession, 1900±54', unpublished PhD thesis, Columbia Univer-
sity.
ÐÐ (1985) `La mano de obra en la minerõÂa estanÂfera Äõ de principios de siglo, 1900±25',
Historia y Cultura, vol. 8.
ÐÐ (1984) `La minerõÂa estanÂfera
Äõ boliviana en la Primera Guerra Mundial', MinerõÂa y
economõÂa en Bolivia, La Paz: Biblioteca Minera Boliviana.
ÐÐ (1980) `Tin Mining in Bolivia, 1900±25', unpublished MA thesis, University of
London.
Crespo, A. (1981) Los Aramayo de Chichas: Tres Generaciones de Mineros Bolivianos, Barce-
lona: Editorial Blume.
DireccioÂn General de EstadõÂstica (1936) Anuario de Comercio Exterior, 1932±33, La Paz:
Imprenta y Literatura Uraria.
Fox, W. (1974) Tin: The Working of a Commodity Agreement, London: Mining Journal Books.
Gallo, C. (1990) Taxes and State Power. Political Instability in Bolivia, 1900±50, Philadelphia:
Temple University Press.
Gamarra, M.P. (1993) `La participacioÂn estatal en la industria de la goma elaÂstica. Legisla-
cioÂn fiscal y economõÂa gomera', DATA. Revista del Instituto de Estudios Andinos y AmazoÂ-
nicos, vol. 4.
Geddes, C.F. (1972) Patin Ä o. The Tin King, London: Robert Hale.
GoÂmez-D'Angelo, W. (1978) La minerõÂa en el desarrollo econoÂmico de Bolivia, 1900±70, La
Paz: Editorial los Amigos del Libro.
GonzaÂlez, F. (1925) `Ferrocarriles en Bolivia', BoletõÂn de la Sociedad de Ingenieros de Bolivia,
vol. 1, nos 6,7.
Green, A. (1992) Education and State Formation. The Rise of Education Systems in England,
France and the USA, London: Macmillan.
GutieÂrrez Guerra, R. (1940) SituacioÂn econoÂmica y financiera de Bolivia, La Paz: Editorial
Universo.
Hallowell, B.C. (1948) `Some Aspects of Tin Control as Applied to Bolivia', unpublished
doctoral thesis, Princeton University.
Hirschman, A. (1971) `The Political Economy of Import-Substituting Industrialization in
Latin America', in Hirschman, A., A Bias for Hope. Essays on Development and Latin
America, New Haven and London: Yale University Press.
Klein, H.S. (1991) Historia de Bolivia, 3rd edn, La Paz: Juventud.
ÐÐ (1971) Parties and Political Change in Bolivia, 1880±1952, Cambridge: Cambridge
University Press.
Lofstrom, W. (1972) Àttitudes of an Industrial Pressure Group in Latin America, the
AsociacioÂn de Industriales Mineros de Bolivia, 1925±1935, unpublished MA thesis, Cornell
University.
Long, R.W. (1927) Railways of South America, Part II, Washington, DC: US Department of
Commerce, Government Printing Office.
LoÂpez Rivas, E. (1955) Esquema de la historia econoÂmica de Bolivia, Oruro: Universidad

TeÂcnica de Oruro.
LoÂpez Videla, J. (1922) Èl saneamiento de la ciudad de La Paz y la inspeccioÂn fiscal del
alcantarillado', BoletõÂn de la Sociedad de Ingenieros de Bolivia, vol. 1, December.
Mamani Capchiri, H. (1992) `La educacioÂn india en la visioÂn de la sociedad criolla:
1920±43', in R. Choque et al., op. cit.
Marsh, M. (1928) The Bankers in Bolivia, a Study in American Foreign Investment, New York:
AMS Press.
McQueen, C.A. (1925) Bolivian Public Finance, Washington, DC: US Government Printing
Office.
Mendoza LoÂpez, V. (1940) Las finanzas en Bolivia y la estrategia capitalista, La Paz: Escuela
de TipografõÂa Saleciana.
Ministerio de EducacioÂn (1917) La reforma educacional en Bolivia, La Paz.
Mitre, A. (1996) Los hilos de la memoria: AscensioÂn y crisis de las casas comerciales alemanas en
Bolivia, La Paz: CENDES.
ÐÐ (1993) Bajo un cielo de estan Ä o. Fulgor y ocaso del metal en Bolivia, La Paz: Biblioteca
Minera Boliviana/ILDIS.
ÐÐ (1981) Los patriarcas de la plata: Estructura SocioeconoÂmica de la Mineria Boliviana en el
Siglo XIX, Lima: Instituto de Estudios Peruanos.
Nelson, R.H. (1949) Education in Bolivia, Washington, DC: US Government Printing Office.
Ness, N.T. (1938) `The Movement of Capital into Bolivia: A Backward Country', unpub-
lished PhD thesis, Harvard University.
Newland, C. (1994) `The Estado Docente and its Expansion: Spanish American Elementary
Education, 1900±50', Journal of Latin American Studies, vol. 26.
Oficina Nacional de InmigracioÂn, EstadõÂsticas y Propaganda GeograÂfica (1905) GeografõÂa
de la RepuÂblica de Bolivia, La Paz: TipografõÂa Comercial-Ismael Argote.
Pacheco, N. (1990) ÀproximacioÂn a una visioÂn histoÂrica de las crisis econoÂmicas en
Bolivia', DinaÂmica EconoÂmica, vol. 10.
Patin Äo, Mines and Enterprises Consolidated (1936) Annual Report to the Stockholders,
Delaware and New York.
Patin Äo Mines Archives ± COMIBOL: Informes Sobre propiedades mineras recibidas del Depar-
tamento de Exploraciones, 15 Dec. 1936, file 1052; vice presidente to SimoÂn I. Patin Äo, 20
Jan. 1937, file 837; vice presidente to Patin Äo Mines (Llallagua), 29 Oct. 1937, file 912;
`Memorandum sobre la peticioÂn Santa Albina de tierras del Chapare', 10 Nov. 1937, file
1156; Secretary General to Patin Ä o Mines (Oruro), 26 Jan. 1938, file 975; vice presidente
to PatinÄo Mines (Llallagua), 17 June 1938, file 982; and General to Patin Äo Mines (La
Paz), 8 June 1943, file 1307.
Prado R. G. (1988) `La apertura de la economõÂa boliviana al comercio mundial (1825±25)',
mimeo, Santa Cruz de la Sierra.
Querejazu Calvo, R. (1978) Llallagua. Historia de una montan Ä a, 2nd edn, La Paz: Editorial
los Amigos del Libro.
Ratinoff, L. (1994) `Las retoÂricas educativas en AmeÂrica Latina', BoletõÂn del Proyecto Prin-
cipal de EducacioÂn, vol. 35.
RodrõÂguez Ostria, G. (1992) `La periferia central: Elites, mercado y cuestioÂn regional en
Cochabamba, 1885±1932', unpublished MA thesis, Facultad Latinoamericana de Cien-
cias Sociales, Quito.
ÐÐ (1991) El socavoÂn y el sindicato. Ensayos histoÂricos sobre los trabajadores mineros, siglos
XIX±XX, La Paz. ILDIS.
Schurz, W. L. (1921) Bolivia. A Commercial and Industrial Handbook, Washington, DC: US
216 Bolivia, 1900±39
Soria Choque, V. (1992) `Los caciques apoderados y la lucha por la escuela (1900±52)', in
R. Choque et al., op. cit.
United Nations (1958) AnaÂlisis y proyecciones del desarrollo econoÂmico, Vol. IV: El desarrollo
econoÂmico de Bolivia, Mexico City.
US National Archives (1925) File 824.51/307 `Roswell to State Department', Washington
DC, 1 April.
ÐÐ (1924) Roswell Barker to State Department, 28 November.
Walle, P. (1914) Bolivia: Its People and its Resources, its Railways, Mines and Rubber-Forests,
London: T. Fisher Unwin.
Whitehead, L. (1994) `State Organization in Latin America since 1930', in L. Bethell (ed.),
The Cambridge History of Latin America, Vol. IV, Part 2, 1930 to the Present, Cambridge:
ÐÐ (1972) Èl impacto de la Gran DepresioÂn en Bolivia', Desarrollo EconoÂmico vol. 12,
no. 45.
8
Trying to `Tax and Spend' Oneself out
of the `Dutch Disease': The Chilean
Economy from the War of the Pacific to
the Great Depression
Gabriel Palma*
Introduction: the context
It might be argued that in general one should guard against the tendency to
overestimate the importance of economic factors behind any war; however, in
the particular case of the War of the Pacific perhaps the danger lies in under-
estimating these factors. Firstly, the Peruvian government nationalized the
nitrate deposits in the mid-1870s, affecting a number of important interests,
both Chilean and British.1 Among the former were those of the past President
of the Republic, Manuel Montt, and the future President JoseÂ Manuel Balma-
ceda, as well as those of a number of leading national entrepreneurs.
Second, Chile had been experiencing economic crisis since 1873; and
towards the end of the decade this crisis was fast becoming unmanageable.
Between 1872 and the outbreak of the War of the Pacific the export price index
fell to the lowest level recorded since foreign trade statistics were first collected
(1844). The collapse of export prices was a consequence of the international
economic crisis which followed the end of the `Great Victorian Boom' ± a cycle
of rapid economic growth which lasted from about 1850 to 1873. Largely
because of the high levels of external public debt ± which in real terms had
trebled in the decade leading up to 1873 ± the decline in Chile's exports was
accompanied by an even greater retraction in imports. These fell by half
between 1875 and 1879, producing a chain reaction throughout the national
economy.2 Public revenues were also seriously affected, mainly as a result of the
* I am grateful for comments received at the conference of AtitlaÂn and at various seminars, especially
in Santiago, CancuÂn, Bilbao, Cambridge, London and Sussex. Edna ArmedaÂriz, Enzo Faletto, Daniel
Hahn, Francisco LeoÂn, HernaÂn SaÂez, IgneÃs SodreÂ, the editors of this book and particularly Geoff
Harcourt and Luis Ortega have also made valuable contributions. The collaboration of Margaret
Hancox has been invaluable. The responsibility for the final result is, of course, mine.
217
218 The Chilean Economy and `Dutch Disease'
high proportion drawn from foreign trade activities, falling by nearly 40 per
cent in 1874. By 1878 revenues were still 28 per cent below the 1873 level.3
Contemporary commentators were unanimous in affirming that the eco-
nomic crisis was the worst that the country had experienced in the whole of
its independent political life.4 In 1878, in ValparaõÂso alone, seven of the leading
commercial houses went bankrupt leaving debts in excess of £ 500 000;5
importing and trading concerns, some as important as that of A. Gibbs, closed
down a significant proportion of their operations; some of the country's leading
exporters, such as JoseÂ TomaÂs Urmeneta and Maximiano ErraÂzuriz, were unable
to meet their obligations; while the government itself was hard pressed to pay
public employees. In 1877 the public deficit reached a level equivalent to one-
third of public expenditure.
Given the critical nature of the financial situation, on 22 July 1878 the
Treasury Minister presented the Chamber of Deputies with a bill temporarily
suspending the convertibility of the peso; so evident was the crisis that Parlia-
ment approved the law in less than 24 hours.6 The outlook for the following
year was gloomier still; for example, on 31 January ± that is, only a fortnight
before the outbreak of hostilities ± the newspaper El Ferrocarril predicted in an
editorial that 1879 `would be a year of mass business liquidation'.
Perhaps the most telling account of the situation facing the country is found
in President Pinto's personal diary. Just three months before the beginning of
the war he wrote in a way that left no room for doubt:
The economic situation of the country is very bad and the forecast is of a
deterioration rather than an improvement. The harvest has been abomin-
able and the price of copper in Europe is falling as never before. Another bad
year with a situation as delicate as the one we have, could have nothing but
bad consequences . . . If a new mining discovery or other news in the same
style does not come to improve our position, the crisis that has been affect-
ing us for years will get worse.
(Personal Diary of President Pinto, 18 November 1878, p. 309; quoted in
Ortega, 1979, p. 455)
As it happened, although 1879 was not a year of mining discoveries, it was

nevertheless a year in which a chain of events culminated in the War of the
Pacific and the annexing of Bolivian and Peruvian nitrate deposits. The spark
which set off this war was the approval by the Bolivian Congress of a tax on
nitrate exports which affected Chilean companies operating between parallels
23 and 24. According to the Chilean government, this tax was in violation of a
treaty signed between the two governments in 1874, which explicitly prohib-
ited such taxes for 25 years on this part of Bolivian territory. The Bolivian
government, meanwhile, argued that this treaty had not yet been approved
Gabriel Palma 219
by its Parliament and was therefore not in force.7 Chilean companies which
were affected refused to pay the new tax on the grounds that it violated the
treaty, and the Bolivian government ordered an auction of their assets. As a
result the Chilean government occupied Antofagasta and thus initiated the
bloody War of the Pacific. Chilean troops disembarked and occupied Antofa-
gasta on 14 February 1879, the very day fixed for the auction of the assets of the
CompanÂaÄõ de Salitres y Ferrocarriles de Antofagasta, a company in which three of
the four civil ministers in President Pinto's first war cabinet were important
shareholders.
It is obviously impossible to guess how the Chilean government would have
reacted to these events had there not been an internal economic crisis of such
magnitude, and if the economic interests of such important politicians had not
been affected so directly. In any case, from an economic point of view Bolivia's
suicidal violation of the treaty could not have occurred at a more critical time
for Chile and, from a political point of view, it could not have attacked more
influential Chilean interests.
There were also other political problems, dating from earlier years. In the
forefront of these was the long struggle between Chile and Peru for hegemony
along the Pacific coast. This was stated clearly by JoseÂ Manuel Balmaceda when
he was a deputy:
On the Pacific coastline in South America there are only two centres of
attraction and progress: Lima and Callao, and Santiago and ValparaõÂso.
One of those centres must succumb so that the other can arise. For our
part we need TarapacaÂ as a source of wealth and Arica as a frontier outpost
on our coastline.
(Chile, Congreso Nacional, Chamber of Deputies, Extraordinary Session 18,
9 August 1881; author's own translation)
However, as Balmaceda himself was later to affirm in a letter to Chilean diplo-

mats as Foreign Minister in the government of Domingo Santa MarõÂa:
The territories of Antofagasta and the nitrate territory of TarapacaÂ were the
real and direct cause of the war. . . [and therefore] to return to the enemy
dominion over the very cause of the conflict, after our triumphs and the
occupation of those territories, would show unjustifiable negligence, and an
absolute lack of the knowledge which affairs of state presume.
(Chile, Memoria del Ministerio de Relaciones Exteriores, 1881, docto. 16;
author's own translation)
Finally in the AncoÂn Peace Treaty between Chile and Peru, signed in October
1883, which brought the War of the Pacific to an end, nine of the 14 clauses
made reference to nitrates and to guano; this, together with the facts described
above, has caused many historians to refer to this war as `The Fertiliser War'
(Loveman, 1988).
A Hirschmanian `generalized linkages' and Keynesian `multiplier'

approach
In the economic history literature there is consensus that in Chile in the period
between 1830 and the First World War the èngine' of economic growth was
located in the export sector. However, this consensus does not extend to the
type of economic, political and institutional dynamics which emerged from this
export-led growth; there are strong divergences in the literature with respect to
the way in which the ènergy' or `stimulus' generated by this export engine was
transmitted to the rest of the economy, in particular to manufactures.
The engine of growth was located in the export sector because of both the
quantity and diversity of exportable natural resources, and the fact that these
were the only activities not constrained by the reduced size of the internal
market. Constraints on exports were given rather by the capacity to mobilize
productive resources ± national as well as foreign ± in order to prospect, extract,
transport and sell these natural resources.
The controversy in the literature is centred on two areas. The first concerns
the kind of productive structure and political institutions that resulted from
this export-led growth. The second relates to the effectiveness with which the
potential for economic growth generated by the export engine was used; in
particular, how effectively the injection of aggregate demand from the export
sector was oriented towards the internal market. The first area concerned the
type of problems analysed by the theory of linkages proposed by Hirschman,
the second those of the Keynesian multiplier.
In the case of nitrates, the basic economic question raised by the first con-
troversy is whether, through input±output relationships, the nitrate industry
put into motion significant investment activities in the rest of the economic
structure, or whether it became an ènclave' with little productive impact on
the country's economy. Concerning the second controversy, the main question
is whether the rapid expansion of exports tended to promote or restrict the
increase in the economy's non-export tradable activities, particularly in manu-
facturing industry; that is, whether the resulting structure of relative prices and
expenditure was favourable or unfavourable for the development of these
activities. In the post-oil-crisis terminology of the 1970s this controversy ques-
tions whether the nitrates expansion caused an outbreak of `Dutch Disease'.
In the analysis which follows we will examine various aspects of both con-
troversies by combining the model of Hirschmanian linkages with that of the
Keynesian multiplier in a generalized linkages approach; that is, one of direct
Gabriel Palma 221
and indirect linkages generated by the export sector with the rest of the eco-
nomy.8 From this point of view, the dynamics of the export sector stimulate
direct linkages with the rest of the economy, be they backwards or forwards,
through purchases of inputs and machinery, the development of infrastructural
work needed for their operation, and the demand for agricultural products for
those working in the export sector, or through the processing required by the
primary product prior to its being exported.
The indirect linkages between the export sector and the rest of the economy,
in turn, have two sides; one is related to the generation of foreign exchange, and
the other to the injection of aggregate demand from the export sector. As for the
foreign exchange generated by the export sector, equivalent to the returned
value of exports (total value of exports minus profit repatriation by the
foreign firms which controlled nitrates and later copper), such revenues lift
the foreign exchange constraints on the rest of the economy. As far as the
injection of aggregate demand from the export sector is concerned, to the direct
component ± related to the production process of the exported goods ± must be
added the indirect component, linked mainly to the public expenditure
financed by export taxes and to the local consumption and investment expend-
iture of the exporting entrepreneurs who returned their surplus to the country.
The dynamics of the export sector depend, on the demand side, on the
volume, composition and stability of international demand for the country's
exports. On the supply side they depend on the export sector's capacity to
respond to the demand stimuli it receives. This capacity to react is a function
of the availability of natural resources and of the possibility of mobilizing
national and foreign factors of production. The direct linkages depend on
such specific characteristics of the export sector as its geographical location,
ease of transport and convenience of processing prior to export, on the tech-
nology used (which affects, for example, the labour intensity of the activity),
and on the input-purchasing policy of the companies involved in the export
sector.
The indirect linkages depend on the way in which the export sector is linked
to the rest of the economy. On the demand side, three factors stand out. The
first is the volume of the aggregate demand generated by this sector. The second
is the composition of this aggregate demand, particularly the proportion of the
surplus retained by the government in the form of taxes and the structure of
public expenditure ± in particular the proportion destined for investment ± and
the propensity to save, invest and consume of those controlling the export
sector. The third is the effectiveness with which the injection of aggregate
demand from the export sector is oriented to the internal market, which is a
function of relative prices ± in particular of the structure of import taxes, the
rate of exchange, the difference between internal inflation and that of the
international market and the behaviour of transport costs.
On the supply side, the most important factors enabling the indirect linkages
are given by the proportion of the export surplus invested in the country's non-
export activities, their technological complexity, entrepreneurial initiative and
the availability of skilled labour.
The dynamic of the (new) export sector and its direct linkages
The volume, composition and stability of international demand for Chile's
exports
After 1873 the rapid fall in international demand for copper, silver, wheat and
flour ± Chile's main exports before the War of the Pacific ± was quickly com-
pensated by a large international demand for nitrates, an essential input for the
explosives industry and for agriculture in more developed countries (at a time
when, furthermore, the guano industry was in rapid decline). Chile's nitrate
exports from occupied territories grew so quickly that by the time Chilean
troops took Lima in 1881, these exports had already reached a level higher
than that for any of the other prewar export products mentioned. After the War
of the Pacific, Chile, in possession of the provinces of TarapacaÂ and Antofagasta
± annexed from Peru and Bolivia respectively ± became the only country in the
world with economically exploitable reserves of natural nitrate salts.
Nevertheless, the concentration of international demand for Chile's exports
on a single commodity brought an important element of instability. It also led
to a decline in the degree of manufacturing value-added in Chile's exports, as
nitrates, unlike copper and silver, did not require significant local processing in
view of the simplicity of the process of extraction and the small gains possible
in time and cost of transport.
As regards the stability of international demand for this new leading export
product, we find an accentuation of the cycle which characterized export
products in the preceding period. An initial boom in international demand ±
with local production responding rapidly ± was followed by a decline towards
the end of the 1890s, a subsequent recovery followed by another rapid decline
after the First World War, and then by a brief increase in the 1920s, and finally a
fall after the 1929 crisis. The first downturn in demand towards the end of the
1890s was due to the problems in the world economy up to 1896±97.9 The
second took place after the end of the First World War and was caused as much
by the development of both natural and synthetic substitutes for nitrates ±
which exposed the low level of productivity of Chilean exports (Figure 8.1) ± as
by the state in which the war left the world economy.10 The third drop, from
1929, was one from which the natural nitrate industry would never recover.
The industrial production of synthetic nitrates began in Germany at the
beginning of the twentieth century, but as the cost was initially significantly
higher than that of Chile's natural nitrates, the first major threat to the
Gabriel Palma 223
monopoly which Chile enjoyed came with the discovery of the fertilizing
potential of another mineral, ammonium sulphate. This was found in Europe,
and although the cost of production was higher than that of Chile's nitrates, it
enjoyed a number of technical advantages. Thus, while Chile's production
of nitrates grew by 73 per cent between 1903 and 1913, global production of
ammonium sulphate grew by 157 per cent. Nevertheless, the final blow to
Chile's nitrate producers came with the subsequent introduction of a number
of technical improvements in the production of synthetic nitrates. Whether or
not the nitrate industry in Chile was ever capable of transforming its methods
of production, or whether this was simply left too late, the fact is that ammo-
nium sulphate and synthetic nitrates swiftly displaced natural nitrates from the
market.11
However, in Chile's history not only were errors repeated, but also strokes of
fortune. If some leading export products collapsed, others appeared with equal
celerity to replace them; in this particular case, paradoxically, there were vari-
ous elements common to the technological changes which made the produc-
tion of synthetic nitrates economically viable, and those which enabled the
large-scale exploitation of copper minerals of low concentration, of which
Chile had one of the largest reserves in the world. Thus, with the opening of
the mines of El Teniente and Chuquicamata in 1912 and 1915, and of Potrerillos
in 1927, large-scale copper-mining, the Gran MinerõÂa del Cobre, was to become,
with the collapse in demand for Chilean nitrates, what nitrates had been to
copper, wheat, flour and silver at the end of the 1870s.
Nevertheless, the crisis of the 1930s was to change this export cycle entirely;
it would not be the problem of productivity which in the final analysis would
reveal the vulnerability of the new sector; Chilean exports were not now to be
replaced in the international market by new and more efficient producers; new
export products would no longer be relied upon to rescue the national eco-
nomy from successive crises. The drop in international demand for Chile's
exports would be of such magnitude and so long-lasting,12 and its effects on
the national economy so profound, that in the four following decades the
remedy was to be sought in a change in the `style of development', rather
than in that of the export product of the moment ± a process that had begun
around the time of the First World War. This new and profound èxport pessim-
ism' ± which began with the nitrate problems and worsened with the 1929 crisis
± was to be the common denominator of the economic policies of the eight
governments following the crisis, in spite of the fact that ideologically they
ranged from conservatives to socialists.
The response capacity of the export sector and its direct linkages
In the year before the outbreak of hostilities ± 1878 ± Chilean exports were at
their lowest level since 1866; in 1882, while hostilities still continued, they had
reached the highest level ever recorded in the country's history. Nitrates' share
of total exports was to grow rapidly, to peak at 81 per cent a year before the First
World War broke out. Nitrates' relative decline thereafter was swift. By 1929 the
share was on a par with smelted and refined copper, and fell sharply through
the 1930s and into the 1940s so that by 1940 nitrates accounted for only 20 per
cent of the total.
In terms of the quantum of production, nitrate exports from Bolivian territ-
ory before the war ± a significant part of which was carried out by companies
with Chilean capital ± had risen continuously during the 1870s, while those
from Peruvian territory had declined by 18 per cent between 1875 and 1878,
mainly as a result of the Peruvian nationalization of the deposits. The war
barely affected production in the province of Antofagasta, which achieved a
record level only one year into the conflict. The recovery of production in the
province of TarapacaÂ began in 1880, and in 1882 also reached its highest level
of production ± in spite of the fact that fighting was still going on in Peruvian
territory. In 1890 the combined volume of production in TarapacaÂ and Antofa-
gasta was already 3.3 times higher than in 1878, but almost all the growth had
taken place in the province of TarapacaÂ. If we compare the level of production
in 1890 with 1875, the highest prewar level, we find that the former was 3.2
times higher. Later, as a result of the rapid drop of the price of nitrates, the main
exporters ± in particular the British speculator John Thomas North who by then
had become the so-called `Nitrate King' ± pursued a policy constraining produc-
tion which led to relative stagnation of exports between 1890 and 1897. How-
ever, when the nitrate price began to rise again, towards the end of the decade,
production responded rapidly causing exports of this mineral to increase 2.6
times between 1897 and 1913. Production fell rapidly in 1914 and 1915, but
recovered quickly, reaching over three million metric tons in 1917, its highest
level, never since exceeded. Production fell again in 1919, but by 1929 had
almost reached its record level of 1917. The 1930s crisis brought about the
industry's final collapse.
The downturns of the nitrates cycles were not a product of the exhaustion of
natural resources or of an increase of production costs brought about by a
significant reduction in the quality of resources. They were a product of turbu-
lence in the international economy, the producers' reaction to this, and of the
development of alternative sources of supply which exposed the local produ-
cers' inability to maintain competitiveness ± when ammonium sulphate and
synthetic nitrates brought about this situation, around the time of the First
World War, the Chilean nitrate industry was still employing the same technol-
ogy used in the 1880s. In fact, as Figure 8.1 shows, productivity in the nitrate
industry had been declining rapidly since the mid-1880s.
It is indisputable that the combined effect of the lack of substitutes for
nitrates before the First World War and of the fact that Chile owned the only
Gabriel Palma 225
Metric tons per worker

120 120
100 100
80 80
60 60
40 40
20 20
0 0
1880 1890 1900 1910 1920 1930
Figure 8.1 Chile: productivity in the nitrate industry, 1880±1930 (output per worker, 3-
year moving average)
Source: Statistical appendices in Palma (1979). Unless otherwise stated, this is source for all graphs in
this work.
exploitable natural reserve of sodium nitrates in the world, was a powerful

disincentive to technological change, to investment and to productivity increase
in this industry. Such a disincentive was heightened by the fact that interna-
tional demand for nitrates turned out to have a low price elasticity. In fact, it was
this that incentivated the producers' output restriction policy between 1890 and
1897. All this, together with the exploitation of deposits of more difficult extrac-
tion, culminated in a progressive decline in productivity in spite of repeated
warnings from numerous scientists and politicians against the danger of this
situation, given that the development of natural and synthetic substitutes were
soon to transform international markets of fertilisers and explosives.
Concerning copper exports, these rose nine-fold between 1911 (the year
before the opening of the El Teniente mine) and 1929, demonstrating an annual
average growth of 13 per cent during these 18 years.13
Figure 8.2 shows Chile's overall quantum of export performance during this
period. The quantum of all exports showed a nine-fold increase between 1878
and 1929. Its subsequent fall was spectacular: according to the Economic
Commission for Latin America and the Caribbean (ECLAC), the quantum of
exports declined over only three years to less than one-quarter ± the 1932 level
was just 23.7 per cent that of 1929 ± making Chile (from this perspective) the
Latin American economy most affected by the crisis, followed closely by Cuba
1878=100
1000 1000
800 800
600 600
400 400
200 200
0 0
1865 1875 1885 1895 1905 1915 1925 1935
Figure 8.2 Chile: quantum of exports, 1865±1935
(Palma, 1984). Its subsequent recovery was slow and difficult. The export
volume at the end of the 1920s was not achieved again for three decades14 ±
and in per capita terms, not for half a century!
As regards the direct backward linkages of the nitrate export sector with the
rest of the economy, this reveals a panorama similar (if not worse) to that
shown by wheat production and extractive mining in the previous period;
despite its enormous importance to the national economy, its direct capacity
to stimulate local manufacturing via purchasing inputs was limited.15 However,
if demand for food from the nitrate area is included among the direct backward
linkages, this was not the case for agriculture.16 It has been argued that this
demand provided a strong stimulus for agricultural production (Kirsch, 1973
and Cariola and Sunkel, 1982), in particular because it was linked to successive
governments' policies for supplying the nitrate area with domestic production
and not with imports. It should be taken into account, however, that the war
obliged a significant proportion of the Bolivian and Peruvian population of the
occupied territories to emigrate to their respective countries. Therefore, an
important part of the demand for agricultural products from the nitrate areas
was in order to feed workers transferred from other parts of Chilean territory,
which indicates a reallocation of demand rather than a net increase. Further-
more, in the case of products such as wheat and flour, part of the additional
demand from the nitrate areas simply absorbed production which had
Gabriel Palma 227
previously been exported. In other words, the rapid increase in demand for
food in the nitrate area was probably more a stimulus for activities associated
with the transport, storage and commercialization of agricultural products,
than for agricultural production itself.
As for the direct forward linkages, nitrates did not require significant process-
ing before export. However, because of its crystalline structure nitrate has a low
density, so its export required a large fleet since each metric ton exported
needed a cargo space of approximately 1.5 cubic metres.
As far as manufacturing industry is concerned, as nitrate exporting did not
call for local processing activities it did not generate connected local manufac-
turing activities; nitrates also did not require large amounts of manufacturing
inputs or machinery. Nevertheless, we should not pass over the purchase of
some locally manufactured inputs, or, with greater reason, the investment in a
railway network to facilitate nitrate exports, and the port works that the export
traffic required. Although the largest oficinas imported a substantial proportion
of their capital equipment, there is good evidence that the nitrate installations
came to be among the major customers of many of the foundries, workshops
and establishments producing transport materials in the period, and the
immediate causes of the creation of a number of similar establishments.
Among the latter, we may pick out as an example the creation in Antofagasta
of the FundicioÂn Orchard Hermanos, which produced a wide range of metal goods
for the nitrate works, and also installed a steel converter of the Bessemer type.
Again, in the first years of this century Siemens-Schuckert Ltd ± the world's
largest conglomerate producing machinery and electrical equipment ± opened
a subsidiary in Antofagasta exclusively for the purpose of supplying the oficinas
with the products they required (Kirsch, 1973; Palma, 1979; and Pinto and
Ortega, 1991).
In relation to the development of railways in the nitrate zone, an idea of the
transport requirements of the nitrate industry can be derived from the fact that
the deposits were found discontinuously along a strip of land more than 4000
sq kms in extension. Railways were essential if production was to be increased,
as otherwise it would only have been possible to extract deposits near the ports
of exit.17 Between 1868 and 1875 a Peruvian company built two important
railway lines, those of Iquique and Pisagua; these later became part of the
Empresa Nacional de Ferrocarriles del PeruÂ.18
After the War of the Pacific, with the rapid expansion of the nitrate business a
number of new railway companies were formed, most of them British-owned.
The total length of private railway line in the country, which in 1875 had
amounted to 650 kms, exceeded 1000 kms within a decade, to reach 1254 kms
in 1887 and 1614 kms in 1890; by this time, however, the private rail network
was not confined to the mining areas of the north but included the coal fields of
the south. By the time the 1930s crisis struck, Chile had 9000 kms of track.19
Finally, attention must be paid to the important copper industry which

emerged at the end of the period. Although a smelting and refining capacity
of considerable proportions was created in Chile, the combination of the
technological complexity of the equipment and machinery required and the
policies followed by the foreign companies involved in the operations meant
that the level of purchases of local inputs was limited (see, for example, Reyn-
olds, 1965). Furthermore, as activities were concentrated in only three loca-
tions, the infrastructural projects required were proportionately much less
important than those which earlier export activities had stimulated. This was
to be one further step, and a definitive one, in a phenomenon which had been
developing since the middle of the nineteenth century; in relative terms, a shift
was taking place from direct towards indirect linkages. In other words, the
export sectors which successively developed were to be progressively trans-
formed ± as a result of their characteristics, the changes in the international
economy, the growing technological complexity of their operations, the pol-
icies followed by the foreign companies involved, and the lack of coherent and
adequate trade and industrial policies from successive governments ± into
virtual ènclaves' within the local economy, as direct linkages were rapidly
reduced relative to both their potential and the indirect linkages.
The articulation of the export sector with the rest of the economy
via indirect linkages
The volume of the demand injection from the export sector
From the point of view of the real value of exports, the export profile is even
more erratic than that of the volume exported, as the export price index
followed a cyclical pattern which tended to magnify changes in the volume
exported.
Figure 8.3 shows that the real value of Chilean exports tripled between 1878
and 1888. Exports then barely rose until 1900, almost doubled in just 8 years
and continued to grow until the First World War. As a result, the real value of
exports was 6.6 times higher in 1913 than in 1878 ± representing an annual
average growth rate of 5.6 per cent. The collapses after 1913 and 1919 are also
worth noting, as is the rapid recovery until 1929 ± now led by the Gran MinerõÂa
copper exports. Exports later fell even more spectacularly, and in 1932 only
achieved a sixth of their 1929 level.20
This cyclical behaviour was even more striking in the case of nitrates. The real
value of nitrate exports grew initially at an annual average rate of 15 per cent in
the 1880s; this dropped to 4 per cent in the 1890s, to be revitalized between
1899 and 1913 at 10 per cent. The real value of exports of these nitrates was 13
times higher in 1913 than in 1880. In 1919, on the other hand, the value
exported was only a seventh of that of 1913, but recovered in the 1920s.
Gabriel Palma 229
million £ (1900)
35 35
30 30
25 25
20 20
15 15
10 10
5 5
0 0
1865 1875 1885 1895 1905 1915 1925 1935
Figure 8.3 Chile: exports, 1865±1935 (at constant 1900 prices)
Nevertheless, the real value of nitrate exports for 1929 was only three-quarters
of the 1913 value. With the recession and the long-term effect of the develop-
ment of substitutes, in 1935 the value of nitrate exports went back to a level
one-seventh that of 1913, and this time did not make a significant recovery.
The injection of aggregate demand into the rest of the economy did not
depend on the total value of exports but on the returned value of exports and
the structure of its expenditure. A series of attempts has been made to try to
quantify the loss of surplus from the export sector due to the repatriation of
profits by the foreign companies which controlled nitrates and, later, copper. It
is agreed that on the subject of nitrates gross value of production was divided
into three broadly similar parts: production costs, export taxes and profits
(Mamalakis, 1976; FernaÂndez, 1978). As nitrate entrepreneurs were predomin-
antly foreigners, a significant part of profits ended up outside Chile. Further-
more, one must bear in mind that an important part of the cost of nitrate
production corresponded to rail transport and the provision of water to the
oficinas, either for drinking or for leaching the caliche, and that these activities
were also mainly British-controlled and hence an additional source of remit-
tance abroad. This enormous loss of surplus from the export sector ± estimated
by Mamalakis at 7 per cent of GDP (1976, p. 56) ± undoubtedly substantially
diminished the aggregate demand injection into the local economy.
There has been enormous speculation as to what would have happened to the
Chilean economy if this syphoning off to the exterior of the surplus from the
export sector had not existed; in particular, regarding what would have hap-
pened if this surplus had been returned to the country and invested. Mamala-
kis, for example, hypothesises that `had these resources been used for
investment, the investment coefficient, which ranged from 15 to 20 per cent
of GDP, could have increased by the possible critical amount necessary to
achieve a self-propelling transformation' (1976, p. 60).
However, this type of counterfactual analysis ignores the institutional frame-
work of the Chilean economy at the time. It is one thing to argue (rightly in my
opinion) that if nitrates had been in the hands of national entrepreneurs, these
might well have been able to respond to international demand in the same
degree as foreign entrepreneurs ± at least, there do not appear to be technical or
financial reasons to suggest the contrary.21 It is another, very different, thing
not to realise that given the historical propensities to consume, save and invest
of the national exporting entrepreneurs, it would have been highly unlikely
that the additional portion of the surplus remaining in the country would in
fact have been directed towards investment (rather than consumption).
Furthermore, some doubt must arise as to whether in such a hypothetical
situation the government would have imposed the taxes on nitrate exports
that were in fact introduced, particularly if we take into account what hap-
pened in the previous period, when export activities were mainly in the hands
of national entrepreneurs.22 Thus, public investment would never have reached
the high levels actually achieved during this period (see later, p. 244 ff ).
Finally, we should not forget the difficulties that attended the marketing of
copper exports before the War of the Pacific, when they were not in the hands
of British companies ± a similar problem could have been faced by nitrate
exports if these had been in Chilean hands.23
All this leads to three conclusions regarding the loss of a substantial part of
the nitrate surplus: first, that there is no convincing evidence that the presence
of British entrepreneurs in the nitrate industry was a `necessary' factor for the
development of this industry; that is, other than in relation to commercializa-
tion, the nitrate industry does not seem to have required North's oligopolistic
control, either for its technological development, the financing of its invest-
ment, or the development of its infrastructural works. Second, given all this,
the large repatriation of profits by British entrepreneurs could hardly be justi-
fied as payment for a net contribution made by these entrepreneurs. Third,
although there can be little doubt that any use to which the lost surplus
might have been put would have been more positive for the local economy ±
including its consumption by the local oligarchy ± than its direct loss,24 it is not
particularly productive to speculate nostalgically (as much of the literature
does) upon the hypothetical situation which would have prevailed if all had
Gabriel Palma 231
remained unchanged bar the assumption of national control, and if all the
surplus had been invested locally.
Returning to the indirect linkages of the nitrate industry, what matters from
the point of view of the analysis of the injection of aggregate demand is not
only the size of the surplus that did return to the country, but also the way it
was spent; in particular, whether this returned surplus was spent in its entirety ±
either in consumption or in investment ± and whether the private or the public
sector had a larger propensity to save than to invest. One of the causes of the
`Dutch Disease' is that it is often difficult to absorb all the additional income
generated by a rapidly expanding export sector, especially its surplus compon-
ent; this leads to an excess level of savings ± either in the private or the public
sector. The counterpart of this excess savings is a surplus in the balance of
payments which puts pressure on the rate of exchange to revalue.
The available evidence shows that the part of the nitrate surplus that was
appropriated by the government was spent in its entirety. In fact ± as we will
discuss in detail later (p. 234 ff) ± in 31 of the 37 years between the War of the
Pacific and the First World War the public sector was actually in deficit. We do
not have sufficient information, however, regarding the expenditure of the
surplus returned and retained by the private sector, although we do know
that the oligarchy's propensity to consume was particularly high, and that
though the level of investment made by nitrate entrepreneurs in the nitrate
industry was low, their investment in other economic activities ± especially
agriculture and manufacturing industry ± was significant.
Thus, even though the returned value in the nitrate industry was signific-
antly lower than the total value of its exports, at least the level of the injection
of aggregate demand made by this industry and the expenditures associated
with it in the rest of the economy ± including the government ± seems to have
been similar to that returned value.
The degree of stability of aggregate demand coming from the export sector
As discussed above, exports during the nitrate era followed a strong cyclical
pattern. Part of the reason for this phenomenon lies in the growing instability
of export prices, and the resulting terms of trade and its growing cyclical
pattern are shown in Figure 8.4.
This shows that from the War of the Pacific to the end of the difficult period
through which the world economy passed before 1895±96, the export price
index fell at a similar rate to the estimated import price index; afterwards,
however, the import price index began to fluctuate more. The growing degree
of sensitivity of the prices of Chile's export products to fluctuations in the
international economy (which were already significant in themselves) meant
that even if the volume of the demand injection from the export sector reached
particularly high levels, the growing instability produced an equally significant
1878=100
150 150
125 125
100 100
75 75
50 50
1865 1875 1885 1895 1905 1915 1925 1935
Figure 8.4 Chile: terms of trade, 1865±1935 (three-year moving average)
degree of uncertainty in the local economy. For example, from the point of
view of the instability of the terms of trade, the average annual percentage
variation ± in absolute terms ± in this relative price index more than doubled
between 1865±95 and 1896±1929.
Figure 8.5 shows the increase in the annual percentage variation of the terms
of trade from an average of 4.3 per cent per annum between 1865 and 1895 to
one of 10.2 per cent between 1886 and 1929. As Chile exported more than one-
third of its output, the increased instability resulting from terms of trade varia-
tions ± assuming everything else constant ± was equivalent to an average
random shock in Chile's national income of about 1.5 per cent per annum in
the first period and 3.5 per cent in the second.
Moreover, as the stability of the injection of aggregate demand coming from
the export sector depends on the stability of the returned value of exports ± and
not on its total value ± the situation was made even more uncertain by the
changing share of foreign ownership of the nitrate industry. For example,
British ownership of nitrate production in TarapacaÂ and Antofagasta increased
from 14 per cent the year before the War of the Pacific to 60 per cent in 1895, to
decrease again to 39 per cent before the First World War (Cariola and Sunkel,
1982, table 16).
Gabriel Palma 233
40% 40%
20% 20%
0% 0%
–20% –20%
–40% –40%
1865 1875 1885 1895 1905 1915 1925 1935
Figure 8.5 Chile: annual percentage variation of terms of trade, 1865±1935
The growing instability of Chile's external sector greatly affected import

capacity. Figure 8.9 shows how the behaviour of imports was even more
unstable than that of exports. This situation was deemed unsustainable after
the First World War, when both exports and imports fell by an amount unpre-
cedented in the country's history. To this fall was added the instability of the
international economy in the early 1920s and the development of substitutes
for nitrate. Also, within the country the dominant economic ideology was
being rapidly transformed, and from the beginning of the century substantial
economic, institutional and social changes were taking place which culminated
in the election of Arturo Alessandri Palma to the presidency in 1920. It is
principally for these reasons that the changes in economic policy and in the
role played by the state in the economy, which occurred in most Latin Amer-
ican countries only in the wake of the 1930s Depression, occurred in Chile two
decades earlier (Palma, 1984).
In short, the cornerstone of the transformation in economic thinking was the
realization that if exports are basically an indirect way of producing imports, it
became apparent from the First World War onwards that it was more sensible to
produce at least some of those imports directly at home, rather than indirectly
through increasingly unpredictable exports. This would have the added benefit
of fostering the development of manufacturing, always seen as the key to a
rapid and sustained process of economic development.25 Thus the increase in

import tariffs in 1916, together with three successive general tariff revisions
in the 1920s, point the way for policy in this area for the next half century. In
turn, the political constitution of the country was changed, and a fairly radical
labour legislation was approved; public expenditure also increased rapidly,
especially on public works and on finance for productive activities. The solid
foundation of these transformations was made evident when the rapid ± but
temporary ± recovery of exports towards the end of the 1920s did not alter the
political climate favouring these changes.
The greater participation of the state in economic activities during the 1920s
was reflected not only ± as in other Latin American countries ± in the creation of
the Central Bank (in an attempt to control monetary anarchy), but also in the
state progressively taking on the role of a direct agent of production in
the economy. Thus 1925 saw the creation of the Servicio de Minas del Estado,
1926 that of the Caja de CreÂdito AgrõÂcola, 1927 that of the Caja de CreÂdito Minero,
and 1928 that of the Caja de CreÂdito CarbonõÂfero, and of the Instituto de CreÂdito
Industrial. These institutions were not restricted to financing the activities in
question, but gave technical assistance and progressively took into their hands
a significant degree of control over the commercialization of the products.
In consequence, the 1930s crisis found Chile with an economic policy
already directed towards import substituting industrialization and a state
already playing a role both quantitatively and qualitatively different from
that which it had played in the first century of political independence.
The important role of the 1930s crisis was to accelerate and consolidate
these political, economic and institutional transformations, but not to initiate
them.
The composition of the demand coming from the export sector

As mentioned above, it has been estimated that broadly speaking the value of
nitrate exports could be divided into three components of similar size: produc-
tion costs, taxes and profits. This section concentrates on the study of export
taxes.26
The national budget and the export sector

If the most significant characteristic of public revenues in the period before the
War of the Pacific had been the small size of the contribution made by exports,
the most striking aspect of this period is precisely the opposite. While in the
1860s and 1870s taxes on exports had reached a maximum of 10 per cent of the
total of public revenues and 2.9 per cent of the total value of exports ± with a
minimum of 1.5 per cent and 0.7 per cent, respectively in 1875 ± in the 1890s
they reached 50 per cent of all public revenues, and 30 per cent of the total
value of exports. The reason for this is that the Chilean state appropriated in
Gabriel Palma 235
taxation approximately half the surplus generated by this activity (or one-third
of the gross value of production: Palma, 1979).
In constant prices of 1900, while export taxes before the War of the Pacific
never reached £100 000, in 1890 they provided more than £2.2 million, to
reach £5.5 million in 1910. After the First World War the revenue from these
taxes began to decline, reaching a level half that of the beginning of the
century. This was not only the result of the fall in nitrate exports, but also
because taxes on the Gran MinerõÂa del Cobre were proportionately four or five
times lower than those levied on nitrate exports.27
Thus the relative contribution of the export sector to public revenues
increased almost ten-fold between 1878 and 1885 ± from 2.7 per cent to 25.4
per cent of òrdinary' public revenues ± while that of all the external sector
(taxes on exports and imports), grew from 44 per cent to 68 per cent. It
continued to grow, coming in 1895 to represent more than 80 per cent of
Chile's ordinary public revenues, then levelled off to around 75 per cent until
the First World War. Thereafter, with the reduction in the volume of foreign
trade, as well as in the level of export taxes, its contribution fell steadily to reach
48 per cent in 1925 and 41 per cent in 1930. As a tesult, internal taxes had to be
increased rapidly from their insignificant levels during the period of booming
foreign trade, when they fell to less than 1 per cent of public revenues.
Public expenditure and the export sector

Public expenditure, which had grown disproportionately during the War of the
Pacific, continued to rise after the war reaching a particularly high level during
Balmaceda's government: by 1891, public revenues had grown nearly four-fold
in real terms since 1878 (see Figure 8.10). As a result, and despite the enormous
revenues derived from nitrate taxes, a budget deficit was recorded in almost
every year during this period, equal, on average, to 18 per cent of public
spending and 11 per cent of exports.
The new government that replaced Balmaceda first implemented a drastic
reduction in public expenditure, halving it between 1891 and 1893. However, a
cut of this magnitude produced such economic and political turmoil that after
1893 public expenditure began to grow rapidly again, increasing nearly three-
fold between 1893 and 1907: the new àustere' post-Balmacedista fiscal policy
lasted for just two years! In all, during the nitrate cycle (1879±1914) the public
deficit amounted on average to a level equivalent to 12 per cent of public
expenditure and 8 per cent of exports. As for the external public debt, it
continued to grow throughout the period, but debt service remained fairly
constant at around 10 per cent of the value of exports.
Given the nature of our investigation we are concerned not only with the level
of public spending and the contribution made to it by the export sector, but also
with two related issues. One is whether the increase in public expenditure (and
its deficit financing) was fast enough to help counteract the possible `Dutch
Disease' effect of the rapidly growing exports: that is, whether public expendi-
ture increased at the pace needed to absorb the foreign exchange supplied by the
returned value of exports, so as to avoid revaluation. The other is related to the
quality of investment: whether enough of this public expenditure went into
investment in physical and human capital ± that is, whether it also helped to
increase the country's productive capacity, particularly of non-tradables, so that
domestic output could increase at the pace of expenditure.
The first issue is discussed later (p. 244 ff). As far as the second is concerned, as
a result of the economic crisis toward the end of the 1870s, public expenditure
devoted to investment in physical capital and education was falling in both
absolute terms and in relation to total public expenditure. After the war ±
particularly as a result of the vast plan for public works and education put
into effect by President Balmaceda ± public expenditure on investment in
physical capital quadrupled in real terms in only a decade, while that on educa-
tion increased eight-fold. As a result, the total devoted to investment grew five-
fold in these years ± or at a rate of 17 per cent per year. In consequence, almost
half of public expenditure went on investment in these two areas, and they
reached, in all probability, their highest relative levels in the history of the
country (Palma, 1979). As Blakemore correctly observes,
Balmaceda's programme for Chile's development was simple in conception:

the proceeds of the export duty on nitrates, now the chief source of govern-
ment income, were to be invested in public works and education, so that
when that source declined as the nitrate deposits were diminished, Chile
would have derived permanent benefit, and would possess other productive
assets to take the place of nitrate.
(1974, p. 71)
For this purpose, in 1887 Balmaceda created the Ministry of Public Works and
Industry, and endowed it with a rapidly expanding budget. Education also
received a substantial boost, for while ± according to Blakemore's estimates ±
there were 79 000 pupils in the country at primary and secondary level when
Balmaceda came to power in 1886, there were over 150 000 by the time he was
brought down (1974, p. 87). In the five years following his defeat public
investment fell by 25 per cent, partially recovering thereafter ± in a cyclical
fashion ± to reach about one-third of public expenditure towards the end of the
1920s, with the public works programme of President Iban Äez.
Transmission mechanisms
The effectiveness with which demand from the export sector was oriented to-
wards the internal market depended in essence upon fiscal policy and on the
Gabriel Palma 237
structure of relative prices; the latter, in turn, was a function of the structure of im-
port tariffs, the exchange rate, the relationship between the internal and interna-
tional rates of inflation, and the movement of transport costs. As fiscal policy was
discussed above, this section will study the determinants of relative prices.
The structure of the import tariff

In an economy with abundance of foreign exchange and in which taxes on
imports grew nearly fifty-fold in real terms in just 12 years, one could expect
strong pressures from exporters, from producers of non-tradable goods, from
producers of tradable goods in relation to their inputs, and from consumers in
general for a drastic reduction in import tariffs. Thus, the most striking character-
istic of the structure of tariffs in the period ± and in particular during the early
export boom ± is that the rapid increase in public revenues provided by taxes on
nitrate exports not only failed to produce a reduction in the average level of
protection afforded by import tariffs to local agricultural and manufacturing
activities, but were accompanied by systematic increases in tariffs ± another policy
helping to avoid possible `Dutch Disease' effects of booming exports. This was true
of all the general revisions of the tariff structure between the War of the Pacific and
the 1930s crisis, the most important being in 1897, 1916, 1925 and 1928.
The point requires further investigation, for it appears a priori that a sudden
increase in revenues from export taxes, by reducing the need for taxes on
imports to provide for public expenditure, might be expected to induce the
opposite effect. Nevertheless, the political structure of the country underwent a
radical shift, for while manufacturing entrepreneurs had already shown that
they had become a group with some influence at national level, it was now the
big landowners who were to seek protectionist policies and thus give rise to a
new powerful political alliance.
With the relative stagnation of agricultural and livestock exports, and of
wheat and flour in particular, farmers too found that the internal market ±
and in particular that of the nitrate area ± had become the major focus of their
operations. They would therefore benefit (at least in the short term) from
restrictions on competing imports, particularly in a period in which a number
of countries were rapidly increasing their agricultural productivity, and when
transport costs were declining very rapidly (mostly rail costs from Argentina).
Thus it was that the recently created association of manufacturing entrepren-
eurs, the Sociedad de Fomento Fabril,28 almost immediately presented to Con-
gress a proposal for the reform of the tariff, at the same time that the association
representing agricultural entrepreneurs, the Sociedad Nacional de Agricultura,
presented one seeking to establish an import tax on cattle to protect themselves
from competition, in particular from Argentina.29 Although Congress refused
to approve either project, at least the pressure may have made a move in the
other direction more difficult.
The growing political unity of the two powerful economic groups took con-
crete form in 1897 with the creation of the Centro Industrial y AgrõÂcola. This
association immediately declared that `the flag of the Centro Industrial y AgrõÂcola
is the protectionist flag' (quoted in Wright, 1975, p. 52), and began a vigorous
campaign to raise the tariffs on a wide range of products.30
The campaign met with immediate success, for on 23 December of the same
year Law 980 was approved raising the import tariff on a number of agricultural,
livestock and manufactured products competing with local production. There
were four economic developments which helped to shape this event. The first
relates to the return to the gold standard in 1895. This had placed the producers
of tradable goods in a difficult situation because it had been implemented with
an overvalued nominal exchange rate, and with an internal inflation higher
than the international. Between 1894 and 1897 the peso had revalued with
respect to the pound sterling by 22 per cent in real terms. The second relates to
the fast fall in import prices, which that year had reached their lowest level
since statistics on foreign trade were first elaborated (1844). As imports had a
high price-elasticity of demand, both the revaluation of the real rate of
exchange and falling import prices had led imports to grow in real terms by
44 per cent in just two years ± producing in the process one of the few trade
deficits during the nitrate cycle, and a large deficit in the current account.
The third development relates to the fact that the return to the gold standard
with an overvalued nominal exchange rate and relatively high internal inflation
had also strained the internal macroeconomic equilibrium. The government
was obliged to reduce its deficit ± by 1896 30 per cent of public expenditure ± by
reducing its expenditure in 1897 by 20 per cent. In these circumstances, the
government obviously saw as positive the possibility of raising its revenues
through an increase in import tariffs. Finally, the protectionist lobby found an
unexpected ally in the oreros (those in favour of the gold standard); the political
accord which had brought the gold standard back after 20 years of inconvert-
ibility was very fragile, and those in favour of this regime were forced to support
the change in tariffs as a way to consolidate the political support for the gold
standard ± even though, given their monetarist view of the balance of pay-
ments, they would normally have been against a tariff increase of this nature.
The law placed ad valorem taxes on a wide range of agricultural and manu-
facturing products and a fixed tax on 28 products, including cattle.31 The
increase in tariffs was so drastic and demand for imports was so price-elastic
that the real value of imports fell by 30 per cent in just one year. Furthermore, as
exports continued to grow, the trade deficit of 1897 was reversed the following
year into the largest trade surplus the country had seen.32 Afterwards imports
did recover, but in real terms it took them a whole decade to return to their
1896 value; meanwhile, exports grew by 80 per cent. This asymmetric beha-
viour generated a sustained trade surplus.
Gabriel Palma 239
It is important to note in this context that the structure of the new import
tariffs ± in particular the low level of the tariff, or the exemption from it, for a
wide range of intermediate inputs and machinery ± and the fall in imports, led
to a relatively small increase in public revenues. What was substantial was the
increase in the ratio between revenues from taxes on imports and the value of
imports. All this is another indication that the main aim of the 1897 law was
not so much an increase in public revenues ± welcome though this might have
been ± but an increase in effective protection for agriculture and existing
branches of manufacturing. Subsequent changes in tariffs confirm the exist-
ence of an industrial policy and a trade policy of selective protection.
For example, one of the additional changes in tariffs, which took place the
very next year, 1898, increased the fixed duty on refined sugar. The higher
duty permitted local producers to raise the internal price, but the ferocity of
the resulting popular reaction forced the government to reverse its original
measure ± reducing this time the tariff on imports of refined sugar by
32 per cent. However, as the government also reduced the import tariff
on unrefined sugar by 50 per cent, it was able to achieve two goals simul-
taneously: to please consumers by forcing a reduction in the internal price
of sugar, and to appease the refiners by increasing their effective protection ±
as the tariff on unrefined sugar fell by significantly more than that on the
final product. Similar changes took place in regard to textiles and cotton
thread.
Later, in 1916, a further general revision of import tariffs took place, which,
according to a report from the Superintendent of Customs, `raised the import
tariff on `èssential'' goods by 50 per cent, other manufactured foods by 86 per
cent, and various manufactured goods by an even greater proportion' (quoted
in Wright, 1975, p. 56). The US Federal Trade Commission calculated that
taking into account import duties and other customs charges, American-
preserved peaches, for example, were now paying a tariff equivalent to 250 per
cent of the CIF value of this product (US Federal Trade Commission, 1916, p. 52;
quoted in Wright, 1975, pp. 56±7).33
The tariff structure was revised again in 1921, 1925 and 1928, but as a
result of the already mentioned fall in export taxes after the First World War,
these industrial and trade tax changes had fiscal considerations as well as
forming part of a policy of giving priority to local production of inputs and
machinery, while the tariff changes of 1897 and subsequent years had favoured
the producers of final consumer goods.
However, the Sociedad Nacional de Agricultura was still not satisfied despite all
the increases in the tariff on the importing of competing agricultural and
livestock products; it also pressed for more èffective' protection from foreign
competition, by preventing the construction of the Chilean stretch of the
railway from Salta to Antofagasta, in order to keep Argentine products out of
the nitrate area. As a result, the Chilean government did not complete the
portion of track on its territory until 1940 ± by which time work on the line
had been suspended for two decades.
Despite the nature of the 1897 law, that of the subsequent revisions of the
import tariff and the widespread political support for increased protection ± in
particular the agro-industrial alliance that brought about the 1897 law ± most
writers on this period persist in the claim that the nitrate-led expansion only
consolidated the commitment to free trade, frustrating ever-more decisively the
development of manufacturing in the country. For example, Claudio VeÂliz's
famous `three-legged table' hypothesis insists on the validity of the theory of
the permanent conflict between the economic interests of landowners and
manufacturers: the former in favour of free trade, the latter for protectionism
(VeÂliz, 1963, p. 104). Other authors have also argued that the level of import
tariffs before the 1930s crisis was very low and that its structure was not
designed to stimulate specific productive activities. In their opinion, the rapid
expansion of nitrate exports, and the related growth of public revenues, gave a
final blow to those who tried to develop local manufacturing industry through
protection. Thus, for example, AnõÂbal Pinto dismisses the protectionist inten-
tions of the 1897 law as à timid and temporary fit of protectionism' (Pinto,
1959, p. 40). Teresa Janneret, in turn, wrongly states that the tariff policy in
Chile before 1930 `did not discriminate between different sectors of produc-
tion' (1972, p. 64), and that `protection in Chile [before 1930] was limited to
general and relatively modest tariffs on imports (25 to 30 per cent)' (ibid., p. 70;
see also Hinkelammert, 1970, p. 62). Finally, two historians of very different
ideologies ± Encina, a conservative, and RamõÂrez Necochea, a communist ±
make the same claim (1942±50 and 1958, respectively).
Although it is obvious that the tariff structure after the crisis of the 1930s
increased the degree of discrimination and selectivity between different eco-
nomic activities, it is equally certain that it did not initiate the process. The
heterogeneity of the tariff structure brought about by the 1897 law and sub-
sequent revisions are a clear indication of the more selective approach adopted;
and the levels of the tariffs on many imported products imposed by the 1897
law (60 per cent, or even more in the case of many specific tariffs), can hardly be
called `timid' or `modest'. Obviously, this is not the right way of labelling a law
which resulted in huge increases in effective protection, a drop of almost 30 per
cent in imports ± a fall that took a decade to be reversed ± and which produced a
record surplus in the balance of trade. Also, subsequent changes in the tariff
structure clearly show that the law was not a `temporary fit of protectionism'.34
The real rate of exchange

In this period we find that the relationship between the rate of exchange and
the balance of trade is apparently paradoxical, at least as far as the nominal rate
Gabriel Palma 241
of exchange is concerned. In these five decades, sustained periods of surplus in

the balance of trade were accompanied by systematic nominal devaluations of
the peso: while there were only six years in which trade deficits were recorded ±
and in four of them they were small deficits at that ± the peso fell, in nominal
terms, by almost six-sevenths of its initial value against the pound sterling.
There were two reasons for this phenomenon: the differential behaviour of the
trade and current accounts of the balance of payments (mainly due to the
repatriation of nitrate profits, and the servicing of the large foreign debt), and
the rapid domestic inflation ± the cumulative inflation between 1878 and 1929
was 714 per cent, equivalent to an annual average of 4.2 per cent.35
The 1870s witnessed the beginnings of an increasing divergence not simply
in the rate but even in the direction of change in the levels of prices internally
and internationally. In the period studied here these phenomena became pro-
gressively more significant. From the War of the Pacific to the First World War
the discrepancy between the rate of internal and external inflation in Chile was
quite extraordinary. According to our price index, from the War of the Pacific to
1897, while in Chile the general level of internal prices increased by 66 per cent,
wholesale and export prices in Great Britain and in the USA were falling steadily
± with Britain's export price index falling by 25 per cent. Again, even though
from then until 1913 all the indices show increases, the rate of inflation in
Chile was so much in excess of those elsewhere that the Chilean index in 1913
(relative to 1880) was four times the British export index. Between 1913 and
1925, the date of the return to the gold standard, Chile's index had doubled
again.36
However, as Figure 8.6 shows, at least until the First World War the differ-
ential between internal and external inflation seems to have moved in line with
the nominal devaluation of the peso. There was, therefore, little change in the
real rate of exchange ± defined as pesos per pound sterling, at constant prices ±
between 1880 and 1913. Thereafter, the real rate devalued rapidly until the
return to the gold standard in 1925.
If, in turn, we measure the real rate of exchange as the relative price between
tradables and non-tradables (Figures 8.7 and 8.8), the nitrate cycle also seems to
have been characterized by its relative stability. That is to say, despite the rapid
increase of exports during this cycle, the real rate of exchange did not revalue;
as a result, relative prices did not turn against tradable producers. The mechan-
ism that avoided this revaluation, and the associated `Dutch Disease', will be
discussed in the next section.
The effect of reduced transport costs on relative prices

The reduction in transport costs continued to be significant throughout the
period, especially sea transport to Europe and the USA and rail to Argentina.
Even though this helped the competitiveness of Chilean exports, it reduced the
`natural' rate of protection of domestic agricultural, livestock and manufactur-

ing production. It is also important to note that this reduction in transport
costs of competing imports took place mainly while the FOB price of imports
was also falling; therefore, it significantly hurt domestic producers of tradable
goods ± and the indirect linkages of the export sector.
It should also be noted that while traditional estimates for the reduction in
transport costs show that the most rapid falls took place in the 1860s and 1870s,
the fact that maritime routes in the South Pacific experienced these reductions
some time after they had appeared on other major routes, means that it is
probable that the bulk of the impact, as far as Chile is concerned, came after
the War of the Pacific and made itself felt in the first decades of the period
studied here (Girard, 1965). This was also the period when the Trans-Andean
Railway through the Uspallata pass began to operate, increasing access to the
Chilean market for Argentinean exports.
In sum: in this section we have seen how the degree of efficiency with which
aggregate demand from the export sector was oriented towards the internal
market was negatively affected by the nature of the productive activity in the
export sector. With the exception of railways and infrastructure in general, and
food agriculture, there were relatively few direct linkages between the nitrate
industry and the rest of the country's productive apparatus. Indirect linkages,
meanwhile, were also negatively influenced by the high internal inflation and
the significant reduction in transport costs for imports. However, fiscal, indus-
trial and commercial policies ± especially those entailing high taxes on nitrate
surpluses, rapid growth in public expenditure and its composition (in particular
the large components of investment and education) and those involving
import tariffs and the nominal rate of exchange ± made a significant effort to
reverse the movement of relative prices in favour of the national production of
tradable goods. In this way it was possible to make the rapid growth of exports
compatible with the development of tradable activities like agriculture and
manufacturing industry, without the export boom causing a `Dutch Disease'
effect. In manufacturing industry, principally due to the structure of import
tariffs, during the nitrate cycle there were incentives for the production of final
consumer goods. This tendency was reversed after the First World War,
providing a strong stimulus to production of many intermediate and some
capital goods.
How Chile managed to avoid `Dutch Disease'

One of the most striking characteristics of the Chilean economy during this
period is that there is no evidence of the export boom bringing about Dutch
Disease. This is obviously a phenomenon which requires further analysis, as
this syndrome is one which economies experiencing a primary-export boom of
this magnitude normally have in common.37 As is well known, the core of the
Gabriel Palma 243
`Dutch Disease' phenomenon is rapid export growth which tends to revalue the
real rate of exchange and, through this, to move relative prices against tradable
production. As Figures 8.6, 8.7 and 8.8 show, available evidence suggests that
this effect did not occur during the nitrate cycle.
The average real exchange rate during the nitrate cycle (1878±1914) was 13.9
pesos to the pound sterling, with a standard deviation of only 1.17 pesos; this
stability is surprising, since this period includes not only the initial export
boom but also one foreign and one civil war. The stability of the rate of
exchange can also be seen in the fact that in 1913±14, at the end of the nitrate
cycle, the real rate (12.7 pesos to the pound sterling) was almost the same as
before the War of the Pacific, and this in spite of the fact that in real terms
exports had increased seven-fold. Even in the first decade of this cycle, when
exports tripled, the real rate of exchange remained stable.
Furthermore, the only years where there was some variation correspond to
specific periods of short duration, such as the two wars and the unsuccessful
return to the gold standard between 1895 and 1898, years which included the
1897 tariff reforms and the resulting sharp drop in imports. Other ways of
measuring the real rate of exchange, such as the ratio between tradable and
non-tradable prices, show a similar picture.
If the price indices for tradable and non-tradable goods used in Figure 8.7 are
an accurate reflection of what happened to prices of these goods and services,
the real rate of exchange (defined in this way) demonstrates, first, a revaluation
50 50
40 40
30 30
20 20
10 10
0 0
1865 1875 1885 1895 1905 1915 1925 1935
Figure 8.6 Chile: real rate of exchange, 1865±1935 (pesos per £ at constant 1900 prices)
Source: Rate of exchange, Palma (1979).
1880=100
200 200
150 150
100 100
50 50
0 0
1865 1870 1875 1880 1885 1890 1895 1900
Figure 8.7 Chile: real rate of exchange, 1865±1900 (prices of tradables divided by those of
non-tradables)
Note: A fall in the index indicates a revaluation. The prices of tradable goods are measured as the
average of the indices of `British parity' (Wagner) and of wholesale agricultural products (Latorre,
recalculated by Wagner); the prices of non-tradable goods are measured by the index of Clavel and
Ferraro (which is available only until 1900). In the figure, 1880 is shown as the base year (and 1889,
the beginning of the civil war, as a reference point) because relative prices in both 1879 and 1891 were
badly affected by the two wars which the country experienced during this period.
Source: Wagner (1992) and Clavel and Ferraro (1990); see also end-note 35.
1880=100
200 200
150 150
100 100
50 50
0 0
1865 1870 1875 1880 1885 1890 1895 1900
Figure 8.8 Chile: real rate of exchange, 1865±1900 (prices of tradable divided by nominal
wages)
Note: A fall in the index indicates a revaluation.
Source: Prices of tradable goods are measured as in Figure 8.7; wages, according to Rojas' index (1982).
Gabriel Palma 245
during the five year-period of Santa MarõÂa (when the index dropped by 25 per
cent), only to recover almost entirely during Balmaceda's government (until
the beginning of the civil war) ± in 1889 this relative price index reached a value
95 per cent that of 1880. That is to say, Balmaceda's fiscal, commercial and
industrial policies seem to have managed to reverse almost all of the earlier
drop in relative prices between tradables and non-tradables.
Another way of measuring the real rate of exchange involves using nominal
wages as an indicator of prices of non-tradable products. The resulting real rate
of exchange is shown in Figure 8.8. According to this definition, the real rate of
exchange did not revalue during the decade of high initial export growth; it did
not even experience the initial revaluation during the five years of Santa Maria
shown in the previous index. Furthermore, towards the end of the century this
real rate of exchange was devalued significantly.
The evidence available thus far shows that by the end of the first decade
of the nitrate cycle, when exports tripled, the real rate of exchange ± by any of
its definitions ± had a value similar to that for the beginning of the cycle.
This relative stability of the real rate of exchange requires more detailed
analysis, as export growth of such magnitude would normally produce a
significant surplus in the balance of payments which would put pressure
on the real rate of exchange to revalue. This is due to the fact that although
growth in exports brings about an increase in imports (through its effects
on income), this increase tends to be lower than that in exports ± unless there
is a revaluation of the real rate of exchange or a compensating increase in
expenditure.
For example, given the usual assumptions of the theory of the Keynesian
multiplier (including constant coefficients) in an economy without a public
sector and an autonomous private investment, a growth in exports increases
income by an amount equivalent to this growth in export multiplied by the
multiplier [1/(s m)] (where s is the average propensity to save and m the
average propensity to import). The increase in income, in turn, raises imports
by an amount equivalent to this growth in income multiplied by the propensity
to import; this amount is equivalent to the initial growth in export multiplied
by [m=s m]. As this last multiplier is necessarily less than one, the resulting
increase in imports is always lower than the growth in exports.
If the public sector is introduced into the analysis through autonomous
public expenditure and income equivalent to a proportion t of income, the
difference between export growth and import growth is even more marked ± as
direct taxation on income introduces a new `leakage' to the circuit of income
and expenditure. In this case, the multipliers become [1=s m1 t t] and
[m=s m1 t t] respectively; the second multiplier is not only less than
one, but lower than in the earlier case ± so the difference between the growth in
exports and imports is even larger.
That is, if private and public investment are autonomous, the mechanism
which adjusts income when faced with an export expansion is not able to
generate equilibrium in the balance of payments ± the ìncome effect' generated
by the increase in exports is not sufficient to maintain the internal and external
equilibria (that is, between the increase in income and that of absorption and
between imports and exports).
In turn, if investment ± private and public ± is not autonomous but rather
induced, the final equilibrium would depend on the value of the accelerator.
One apparently paradoxical result of this multiplier±accelerator equilibrium is
that even if the accelerator induces an increase in investment large enough as
to absorb all the savings generated by the growth in exports, the resulting
increase in expenditure would still not be sufficient to achieve macroeconomic
equilibria ± and thus avoid a revaluation. In order to achieve this equilibrium
(without revaluing) it would be necessary to have an additional (compensatory)
increase in expenditure (over and above the one induced by the previous
accelerator), and so, at least initially, a deficit expenditure either in the public
or in the private sector.38 This deficit expenditure was precisely what fiscal
policy provided. Without this, a primary-export boom would have tended to
generate an insufficient increase in expenditure and thus a surplus in the
balance of payments, which would have put pressure for a revaluation of the
real rate of exchange. The role of this revaluation would have been to recover
internal and external equilibria through an inducement to increase both
imports and expenditure ± that is, it would have added a `price effect' to the
insufficient (export-led) ìncome effect' on imports.39
This phenomenon can be seen more clearly through the macroeconomic
relation between the surplus ± or deficit ± in the current account and that
between savings and investment. The income and expenditure accounts yield
the following relationship between both surpluses (or deficits):
X M S Ipr S Ipu ], where X is total exports, M is total imports
(including the repatriation of profits and the servicing of the foreign debt),
S Ipr is the difference between savings and investment in the private sector,
and S Ipu is the same for the public sector ± equivalent also to the difference
between income and expenditure in each sector. If exports grow more rapidly
than imports (including factor payments), it is because investment is growing
less rapidly than savings (or expenditure less rapidly than income) whether in
the private or the public sector. One way of reducing both surpluses is to make
imports increase more rapidly than the level induced by the growth in export
(via the ìncome effect' of the latter). This can be achieved either by a compens-
atory increase in expenditure, or through the `price effect' of a revaluation, or
through a combination of the two.40
What is unusual about the case of Chile in this period, as shown in the earlier
graphs and in Figure 8.9, is that the solution to the above-mentioned problem
Gabriel Palma 247
1878=100
1000 1000
800 800
600 600
400 400
200 1 200
2
0 0
1865 1875 1885 1895 1905 1915 1925 1935
Figure 8.9 Chile: exports and imports, 1865±1935 (at constant 1900 prices)
[1] Imports [2] Exports
was found through a compensatory increase in expenditure ± an active fiscal

policy ± and not through a revaluation. Figure 8.9 illustrates this point, show-
ing that without a revaluation imports grew at the same rate as exports.
Taking the year before the War of the Pacific as a base, one can see that
imports not only followed the growth in exports step by step, but that in fact
during certain periods they grew even faster. As the nitrate cycle began with a
surplus in the trade account, equivalent in 1878 to 18 per cent of exports,
the similar growth of imports and exports implied that, in relative terms, the
trade surplus existing before the War of the Pacific did not on average change
significantly during this period.41 This trade surplus was not only maintained
in relative terms, but also, as before the war, was totally absorbed by the
financial requirements of the current account, this time, after the war, not
only for the servicing of foreign debt but also for the financing of the repatria-
tion of profits of foreign companies involved in the nitrate industry. In fact, in
aggregate terms, in the three and a half decades of the nitrate cycle (1879±1914)
± albeit with noticeable subcycles ± the current account of the balance of
payments actually recorded a deficit. Even during the first decade of the nitrate
boom the current account was on average in relative equilibrium. This shows
that even during the period of greatest export growth, absorption managed to
grow at a rate similar to that of income. In the following decade, and up
until 1897 due to the lower export growth, absorption increased more
quickly than income, generating a deficit in the current account of the
balance of payments. After the tariff reforms of 1897, this tendency was
reversed for another decade ± both due to a slower growth of imports and
a recovery in exports. In the following decade, however, the opposite pheno-
menon appeared.
Adding together these cycles one can see that what really characterized the
nitrate cycle was the fact that instead of producing a surplus in the current
account and pressure for a revaluation of currency, and domestically an excess
of savings, the cycle was associated, though unstably, with a deficit in the
current account, a shortage of internal savings, an increase in absorption larger
than that of income, and relative stability in the real value of the currency.
Figure 8.10 looks at one of the key aspects of this phenomenon, comparing the
growth in public expenditure with that of exports.
Figure 8.10 shows how fiscal policy caused public expenditure to grow,
though unstably, at a comparable rate to that of exports between the War of
the Pacific and the First World War. In fact, this expenditure grew somewhat
faster than exports. Furthermore, at the crucial moment when the nitrate
industry was incorporated into Chile's economy ± that is, when the danger of
the `Dutch Disease' was at its greatest ± between 1878 and 1891 public expen-
diture increased 42 per cent faster than exports. As public expenditure in 1878
was equivalent to three-fifths of exports, this faster growth rate of public
expenditure implies that, in terms of absolute value, this expenditure grew by
1878=100
1200 1200
1000 1000
800 800
600 600
400 400
200 1 200
2
0 0
1865 1875 1885 1895 1905 1915 1925
Figure 8.10 Chile: public expenditure and exports, 1865±1930 (at constant 1900 prices)
[1] Exports [2] Public expenditure
Gabriel Palma 249
a similar amount to that of exports. This huge injection of public expenditure ±

with a significant deficit component ± was the main reason for the fact that
during this time there was no surplus of internal savings, that absorption grew
at a rate similar to income, and that imports grew at a rate similar to exports.
This expenditure, as we have already seen, contained a significant compon-
ent of investment in physical and human capital, particularly during the
Balmaceda period when this component made up almost half of public expend-
iture. The growth in public expenditure not only stimulated an increase in
absorption to a rate similar to income, and a growth in imports similar to
that of exports, but also rapidly expanded the country's productive capacity,
in particular in non-tradables.
In order for an increase in expenditure to fulfil the objectives proposed here,
it is essential that the supply of non-tradables be elastic. Without this, the
increase in expenditure would provoke a rise in the price of non-tradables,
and thus, the opposite result to that required: if the increase in expenditure is
followed by an increase in prices of non-tradables at a rate faster than that of
tradables ± whose supply is elastic during an export boom, given the availability
of foreign exchange ± relative prices will change against tradable products,
synonymous with a revaluation of the real exchange rate. However this did
not occur in Chile during the nitrate boom ± as Figures 8.7 and 8.8 show. This
indicates that from the point of view of supply, perhaps the most striking aspect
of Chile's economy during this period was the high elasticity of production of
non-tradables. It is this ability to respond by the supply of these products which
prevented the overheating of the economy, and explains the success of fiscal
policy in avoiding Dutch Disease.
Other economic indicators also show that the economy did not overheat. For
example, as Figure 8.8 shows, nominal wages only rose by a proportion similar
to that of the prices of tradable goods; also, the interest rate in the 1880s was
lower than in the 1870s. One essential factor in ensuring that the supply of
non-tradables could respond to the increase in expenditure was the practically
unlimited supply of labour ± with very limited opportunity cost ± in the rural
sector.
However, as the experiences of other countries with export booms have
shown, avoiding `Dutch Disease' is a more complicated process than has been
suggested so far. This is because export booms ± particularly of primary com-
modities ± tend to generate strong institutional and political tensions which
make it difficult to design and efficiently implement a package of coherent
economic policies.
One of the main problems has its origin in the fact that many primary
commodities ± including nitrates ± are characterized by the generation of a
large surplus, equivalent to a high proportion of the value of production. This
situation often results in fierce political struggles over the distribution of this
surplus, which make it particularly difficult to coordinate economic policy. This

distributive conflict not only emerges between exporters and the government ±
over the level of taxation on the export sector ± but also between different
interest groups over the distribution of the surplus appropriated by the govern-
ment. In Chile both distributive conflicts ± especially that between the govern-
ment and British exporters and between the government and parliament (still
dominated by large landowners) ± contributed to the first ten years of the
nitrate cycle ending in the first civil war the country had suffered in sixty
years.42
Thus, one of the key economic policy dilemmas of the period was what to do
with the large profits from nitrates ± approximately equivalent to 15 per cent of
GDP. There were three main elements to this problem. The first concerned the
loss of national income due to the fact that foreign entrepreneurs, especially
the British who controlled about 60 per cent of production after the War of the
Pacific, tended to repatriate practically all their profit ± potentially a loss of up
to 10 per cent of national income.
The second problem concerned the absorption of the share of the nitrate
profit left in the hands of Chilean entrepreneurs ± and some non-British for-
eigners who became resident in Chile ± who voluntarily tended to return their
profits to the country. Given the magnitude of this surplus one could hardly
expect that even members of the Chilean oligarchy could have spent it all,
whether in consumption or investment. In other words, even if these entre-
preneurs did contribute to national income by returning their profits, it was
unlikely that they could contribute to absorption by the same amount as these
profits were left in the hands of a very small group of people. Unless part of
these nitrate profits was transferred to other economic agents (public or private)
with a greater propensity to spend, a surplus of private savings would be
generated and, through this, an internal and external disequilibrium which
could lead to a revaluation. The third problem concerned the absorptive capa-
city of the public sector itself; high taxes on nitrate profits could also create a
savings surplus, as the government might not want or be able ± because of
exogenous constraints ± to spend all the income received from the export
sector.
Fiscal policy at the time ± especially during the Santa MarõÂa and Balmaceda
periods ± managed to find an intelligent solution to these problems through a
policy of `tax, transfer and spend'. Firstly, through taxation, the government
obliged foreign companies to return about half of their profits to the country.43
Secondly, the government also taxed those nitrate entrepreneurs who kept their
profits in the country. The first measure increased national income by about 5
per cent, while the second shifted approximately 3 per cent of national income
to the public sector. The next problem for the government consisted in how to
absorb these resources. As Figure 8.10 showed, this did not seem to have been
Gabriel Palma 251
much of a problem for the government; also, a high proportion of public

expenditure was allocated to investment.
One of the main reasons for the government's high capacity to absorb the
resources received was that it did not just stick to a radical fiscal policy of `tax
and spend'; it also added a third element: `transfer'. This policy consisted in
shifting a significant share of the nitrate profits appropriated by the govern-
ment through taxation back into the rest of the private sector, by reducing the
tax burden on that sector almost to the point of elimination. Thus, internal
taxes, which had contributed 14 per cent of ordinary public revenues in 1878,
contributed only 9 per cent in 1885, 4 per cent in 1890 and 0.1 per cent in
1895.44 The government further transferred to the private non-nitrate sectors
part of the revenues it received from nitrate exports by reducing tariffs on some
public services; this reduced the relative contribution of these public revenues
to the national exchequer to less than half, from 39 per cent of the total in 1878
to 18 per cent in 1885 and 17 per cent in 1890 and 1895. Since the rest of the
private sector had a high marginal propensity to spend, this income redistribu-
tion significantly increased absorptive capacity. Finally, Figure 8.11 shows that
the government counterbalanced any excess savings left in the private sector by
running, cyclically, a significant fiscal deficit.
As Figure 8.11 shows, between 1883 and 1891 a fiscal deficit was created
equivalent, on average, to 20 per cent of public expenditure (14 per cent of
50% 50%
25% 25%
0% 0%
–25% –25%
–50% –50%
1870 1880 1890 1900 1910 1920 1930
Figure 8.11 Chile: public deficit, 1870±1930 (three-year moving average)

[1] Public deficit as a percentage of public expenditure
[2] Public deficit as a percentage of exports
exports). As the current account, on average, ended up in balance during the

first nitrate decade (and there was no overheating in the non-tradable sector),
this deficit seems to have been the right amount to compensate a surplus of
private savings brought about due to the growth of income both from exports
and from the increase in public spending itself. Contrary to the view in the
literature up to this point, what is paradoxical is that the public deficit was
excessive not during the Balmaceda years but after his overthrow, as it was only
then ± partially due to lower export growth ± that it was associated with a deficit
in the current account and a shortage of domestic savings.
One of the peculiarities of the economic historiography of this period to date
is that it has almost unanimously criticized the various aspects of the fiscal
policy of the period, which we are calling `tax, transfer and spend'. From a neo-
liberal perspective, criticism has centred on the high levels of taxation and
spending. From this perspective ± and almost by definition ± the nitrate
resources taxed and spent by the state would have been put to use more
efficiently by the private sector (Chilean or foreign). Critics have also pointed
to the fact that the high level of taxes on nitrate exports would have resulted in
a strong disincentive to that industry's technological development.
From the dependentista perspective, and those of similar views, criticism has
centred on the first part of this fiscal policy, the level of taxation of nitrate
profits, but only in relation to foreign entrepreneurs, North in particular. As
these entrepreneurs barely provided new capital, did not bring new technology
or new forms of management ± and, furthermore, obtained possession of their
nitrate mines through manoeuvres of doubtful legality and even less morality ±
from the national economic development point of view there could be little
justification for their repatriating such a significant amount of their earnings.
Worse still, there is ample evidence that part of the earnings which they did not
repatriate were used by them to corrupt and pervert the country's political
institutions. So, looking at their real contribution to the country's economic
and political development, the optimal rate at which their profits should have
been taxed was close to 100 per cent. In other words the Santa MarõÂa govern-
ment should not have yielded the bulk of the country's new-found export
wealth to these entrepreneurs; and if, due to British diplomatic and financial
pressures, this government had little option but to do so, they should have
imposed even higher taxes on these entrepreneurs' profits ± particularly on
profits not returned to the country.
Those who analyse this part of Chilean history from this perspective have
also criticized the government for what we have called the third aspect of their
fiscal policy, the `transfer' of part of the nitrate surplus through the reduction of
domestic taxation. The state would have made a greater contribution to the
development of the country, it is argued, had it spent these resources directly,
by, for example, providing a stimulus to industrial development instead of
Gabriel Palma 253
transfering them to non-nitrate high-income groups. This was particularly the

case as much of the transfer benefited the inefficient (and `pre-capitalist') land-
owning oligarchy, either directly through the reduction in their taxes and
tariffs paid for public services, or indirectly since a significant share of public
expenditure on infrastructure and other services was oriented towards agricul-
ture.45
Finally, criticism from a `structuralist' and `neo-structuralist' point of view is
aimed at another aspect of the fiscal policy of `tax, transfer and spend'. This
criticism comes from what is now called an institutionalist perspective: this was
an èasy' policy with little fiscal discipline, generating long-term negative
effects on the development of the public sector ± and the economy in general.
Meller, for example, says that:
In developing countries, the growth of the public sector illustrates an impor-

tant impact of a boom in the exports of a primary commodity, which has
often been ignored by literature on the `Dutch Disease'. The abundance of
taxation revenues generated by the export boom erodes the government's
fiscal discipline; financial constraints are no longer necessary. The govern-
ment obtains funds through the taxation of foreigners (who control exports)
and uses them to increase expenditure. This was what happened in Chile:
Chilean society became used to low levels of taxation, at the same time that
fiscal expenditure was being increased.
(1990, p. 56; author's own translation)
It is obvious that the fiscal policy of this period generated some of the
negative effects which Meller identifies ± which can be seen, for example, in
the public sector's difficulty in adjusting its expenditure after Balmaceda when,
as a result of lower growth of exports, the high levels of public expenditure
created an excess of absorption and a deficit in the current account. In turn,
public revenues became overly dependent on a very unstable external sector.
However, this fiscal policy also had the positive effects already analysed. A
further effect, ignored by critics, was that by eliminating internal taxes it helped
to increase the value of the multiplier; and, by doing so, helped to maximize
the growth of income arising from an increase in exports ± as from other
demand injections such as public expenditure and private investment. We
have already mentioned that the multiplier of an autonomous growth of
exports in such an economy is [1=s m1 t t]; however, in an economy
without a public sector, this rises to [1=s m], as one of the income leakages
(t) is eliminated. What seems paradoxical, however, is that this higher multi-
plier for an economy without a public sector is also relevant for one with a
public sector, but in which its expenditure is financed (in Meller's words)
`through the taxation of foreigners (who control exports)'; that is, through
taxes which are outside the income-expenditure circuit. So although this fiscal
policy did generate institutional problems, it also had its successes. It increased
national income ± forcing foreign entrepreneurs to return about half of their
earnings to the country, increasing public investment in physical and human
capital, and increasing the value of the multiplier. It also increased absorption
(an essential factor in avoiding `Dutch Disease').46
Thus approximately a third of nitrate exports ± equivalent to the cost of
production ± was absorbed through the productive process, and another third
(taxes on exports) through fiscal policy (public expenditure and transfers to the
private non-export sector). As for the remaining third, profits after taxes, the
share that went to the national nitrate entrepreneurs ± or foreigners who had
become resident in the country ± was absorbed in consumption or investment;
while the share taken by foreign entrepreneurs who repatriated their earnings
was absorbed, from a balance-of-payments point of view, by this process of
repatriation.
All of the above helps to explain how in this period in Chile it was possible to
combine a particularly rapid increase in primary commodity exports with
development in manufacturing. It is well-known that the income elasticity of
expenditure on manufactures is greater than that of agricultural and non-
tradable products. However, from a `Dutch Disease' point of view, it is not a
question of how the aggregate demand for manufactures increases, but of how
the demand for nationally-produced manufactured goods increases. Apart from
the quality of these products, this depends mainly on relative prices and the
level of expenditure and its structure. We already know that on average absorp-
tion grew more rapidly than income, that there was no revaluation of the real
rate of exchange and that import tariffs increased for a wide range of manufac-
tured goods. As a result, in spite of domestic inflation and the drop in transport
costs, the evidence shows that effective demand for nationally-produced man-
ufactures grew at least at a rate more or less similar to that of income. Studies of
manufacturing production for this period tend to confirm this fact (Kirsch,
1973; Palma, 1979 and 1984). In sum, the demand-side evidence, as well as
the supply-side, shows that national manufacturing production was stimulated
by the nitrate boom: exactly the opposite of what one would expect if this had
been a case of `Dutch Disease'.
This also helps us to understand why, contrary to what is usually argued (see
for example Behrman, 1976) the extraordinary expansion of exports did not
cause a shrinking of traditional exports, at least for the first decade of the nitrate
cycle. Although wheat exports fell during the following decade, this was not the
case for wheat production. Production of other crops also grew significantly, as
in the case of barley, beans, potatoes, peas and chickpeas.47 In the next decade,
wheat production practically doubled and for a few years exports again reached
the level of the end of the 1880s.
Gabriel Palma 255
Copper production after the end of the first decade of the nitrate cycle fell
gradually at the beginning of the twentieth century, to recover slowly after-
wards. So the beginning of the Gran MinerõÂa of copper in 1912 found this
industry with an export level similar to the figure for 1888 ± and the second
half of the 1870s. However, studies show that this stagnation was also not
directly due to the nitrate boom ± that is, to problems of relative internal prices
± but to a growing technological backwardness, because in Chile the copper
industry had been unable to follow the particularly rapid technological devel-
opment which characterized this industry internationally from the end of the
nineteenth century (Reynolds, 1965; Culver, 1982; and Mayo, 1982).
To sum up this section: the growth in nitrate exports was linked to a rate of
growth of absorption at least similar to, if not higher than, that of income. It
did not generate surpluses in the current account or in national savings, or a
real revaluation of the peso. It also saw the growth of effective demand for
nationally-produced manufactures at a rate similar to that of income, and did
not cause a mechanical shrinking of traditional exports. The axis for this
phenomenon was fiscal policy, which taxed exports heavily and injected a
high level of public spending ± with a strong investment component ± into
an economy characterized by an elastic supply of labour and non-tradable
goods.
Conclusions
We have seen how the nature of production in the export sector, fiscal policy
and the relative price structure were essential factors determining how effect-
ively demand from the export sector was oriented towards the domestic mar-
ket. Concerning the first of these factors, the production process of nitrates
and, later, that of copper, had a low level of direct linkages ± forward as well as
backward. For this reason, neither nitrates nor copper greatly helped the pro-
cess of absorption, nor did they provide a significant stimulus to other product-
ive activities. The exceptions to this rule are the railway network for the mining
area, the development of infrastructure related to both products and, in the
case of nitrates, agriculture (and related services). As far as the second factor is
concerned, the fiscal policy of `tax, transfer and spend', did help to generate the
kind of external and internal macroeconomic equilibria necessary for avoiding
Dutch Disease, especially during the first decade of the nitrate boom. Regarding
the third factor, the relative price structure ± the main factor which determines
the export sector's indirect linkages ± we know that, of its four principal
components, two operated in one direction and two in the other. The nominal
rate of exchange and the level and structure of import tariffs provided an
incentive for the orientation of aggregate demand towards the domestic
markets. Meanwhile, the other two factors ± the large gap between internal
and external inflation and the reduction of transport costs ± operated in the
opposite direction. Furthermore, it seems that until 1913 the difference
between the internal and external rate of inflation was similar to that of the
depreciation of the peso. The final effect depended on the balance between the
reduction in transport costs and the increase in the level of import tariffs and
the change in their structure. Although the available empirical evidence is
insufficient, we can conclude that during the nitrate cycle producers of tradable
goods did not have to face either the economic policy or the changes in relative
prices that one would expect in a rapidly expanding export economy. Therefore
they did not suffer Dutch Disease. This conclusion is confirmed by the estim-
ates for manufacturing production in this period.
Later, between the First World War and the return to the gold standard in
1925, producers of tradable goods benefited from the rapid real devaluation of
the peso, only to be harmed by the rapid revaluation which took place during
the short period of convertibility ± 1925±31. They were also harmed by having
to face two very different exchange rate polices in a short period.
One striking aspect of the policies of this period was that the rapid growth in
public revenues from high taxes on the export sector did not induce a reduction
in tariffs on those imports in competition with domestically produced pro-
ducts. On the contrary, the abundance of public revenues enabled the imple-
mentation of a tariff structure which increased effective protection for many
tradable goods. Together with an increase in tariffs on these products, it allowed
a reduction in tariffs on the inputs and machinery necessary for their domestic
production, in spite of the drop in public revenues that this implied. This
orientation of tariff policy gradually produced a horizontal rather than a ver-
tical structure of manufacturing. The conditions which favoured producers of
consumer goods worked against producers of manufacturing inputs, and until
the First World War fiscal, commercial and industrial policies favoured the
former at the expense of the latter. Later, after the crisis in the external sector
and the country's political and ideological changes, these policies changed,
increasingly favouring inputs.
As far as agriculture is concerned, the cost of the tariff policy was the reinfor-
cing of attitudes which were anti-export and anti-specialization. However,
indiscriminate protection of agriculture seems to have been the necessary
price for the support of rural producers for a policy of industrial protection.
Finally, since the policies protecting both sectors, manufacturing and agri-
culture, were not conditioned to increases in productivity, or limited in time,
their effect, given the reduced size of the domestic market, was a significant
reduction in the level of competition faced by producers. Protection thus
eliminated perhaps the most important factor necessary for the efficiency of
the market: the discipline which only competition can provide. So Chile
escaped Dutch Disease but not without a cost.
Gabriel Palma 257
Notes
1 The fact that the nationalization of Peruvian nitrates affected British investment and
that political relations between Bolivia and Great Britain went through a period of
tension, has led to a general tendency to overestimate the role played by Britain in
the war, and to pay insufficient attention to the political and economic problems of
Chile itself which lay behind it. This view is expressed very bluntly by the US
Secretary of State James G. Blaine, who stated that the War of the Pacific was àn
English war on Peru, with Chili [sic] as the instrument . . . Chili [sic] would never have
gone into this war one inch but for the backing of English capital' (quoted in Sater,
1990, p. 39). Among the numerous accounts of British involvement in the War of the
Pacific, see Collier (1993), Dennis (1931), FernaÂndez (1978), Ortega (1979) and Sater
(1990).
2 Unless otherwise stated, all data in this chapter have as a source the statistical
appendices in Palma (1979). Figures in sterling are at 1900 prices, and all rates of
growth are expressed in real terms.
3 When the crisis began in 1873 import taxes made up some 51 per cent of what was
called òrdinary' public revenues, and export taxes contributed another 4 per cent; the
rest was made up of the revenues from public enterprises (32 per cent), internal taxes
(11 per cent), and other smaller taxes (2 per cent). Èxtraordinary' public revenues were
made up of external and internal loans which financed the fiscal deficit.
4 For an analysis of the literature of the period, see Ortega (1979, ch. 6).
5 See Archives of A. Gibbs, ValparaõÂso to London, 28 December 1878.
6 The day following the passing of this law, President Pinto wrote in his diary: `the
banks would have gone broke that same day if inconvertibility had not been
declared', quoted in Ortega (1979, p. 453).
7 See Dennis (1931, especially ch. 3). The Bolivian government needed this tax because
its finances were going through an extremely difficult period since the effects of the
international economic crisis were exacerbated by the disastrous effects of the recent
civil war. It has been argued that the main reason for the dictator Daza deciding to
increase public revenues in violation of the treaty was as a result of his perception
that Chile was in no position to go to war both because of its own serious internal
economic problems and because of its border tensions with Argentina ± which had
recently obliged them to give Patagonia up to Argentina and to agree to joint control
of the Magellan Strait.
8 Corden and Neary called these two types of linkages between the export sector and
the rest of the economy the `movement of resources' effect and the èxpenditure'
effect (1982).
9 The price of nitrates fell by 44 per cent just between 1884 and 1888. The end of the
`Victorian Boom' in 1873 was followed by a generalized drop in prices of primary
commodities and manufacturing products in the world economy until 1896±97.
According to S. B. Saul (1976), this generalized drop in prices over 14 consecutive
years was the most significant economic phenomenon of the nineteenth century.
10 1919 was a particularly difficult year for the Chilean economy. The value of exports,
in real terms, fell to about one-third the 1918 levels, with the attendant effect on
public revenues from this source. This fall in export revenues was owed to the fact
that the price of nitrates fell by 17 per cent that year and the quantum exported by
more than two-thirds. The main reason for this decline was the fact that European
countries and the USA had accumulated large stocks of nitrates during the war, and
these stocks were being used in 1919. Exports to the USA in 1919, for example,
reached only one-fifth of the 1918 level ± 370 000 tons and 1.7 million, respectively
± while those to Britain were equivalent to just 5 per cent of the 1918 volume (25 000
tons and 535 000 tons, respectively).
11 Between 1913 and 1923 the share of Chilean nitrates in world production of nitro-
genous fertilizers fell by almost half ± from 55 per cent to under one-third. This rapid
drop and the problems with the price of nitrates caused the Chilean nitrate industry
to negotiate the formation of a cartel with the world's two largest producers of
synthetic nitrates, I. G. Farbenindustrie and Imperial Chemical Industries. These
three producers controlled 70 per cent of world nitrogen production. US producers,
which controlled around 10 per cent of production, were unable to become a part
of this cartel because of their country's anti-trust legislation (see Keynes, 1985, pp.
565±7).
12 For many economic historians what was exceptional about the crisis in the 1930s was
not so much its initial magnitude but its long duration. The world economy had
already got through various abrupt falls in prices relatively quickly ± one of which
took place shortly before the 1929 crisis (1921); what was unusual about the 1929
crisis was that prices not only failed to recover, but they actually continued to fall
throughout the world for many years. For an analysis of this phenomenon and of its
possible causes see specially Kindleberger (1984).
13 In the case of copper its initial increase was also followed by a period of extreme
instability and eventual relative stagnation after 1929. Thus in 1954, 25 years later,
exports from the Gran MinerõÂa ± after having being through a cycle of abrupt swings ±
were just 12 per cent higher than in 1929. This aspect of the copper export cycle
mirrored the characteristics of early export cycles, except that in this case it was
necessary to wait four decades for the appearance of new export products to revitalize
the export sector ± in this case agricultural, wood and fish products.
14 In the period from 1929 to the second decade of the 1950s, only 1942 shows the
quantum of exports higher than those of 1929. This is not to overlook, of course, the
problems inherent in quantum indices (Laspeyres) when used over periods as long as
those studied here.
15 The nitrate industry was even practically self-sufficient for one of its principal inputs,
explosives; these were necessary for the removal of the caliche and were produced
from the nitrates. The industry was also self-sufficient for one of its fuels, wood from
the tamarugo tree, although indiscriminate felling of these trees led to massive
ecological devastation.
16 The majority of nitrate mines paid with tokens instead of money. These could only be
exchanged for food and other essentials in the mines' own pulperõÂa. For this reason,
labour conflict of this period ± some of which were repressed with particular brutality,
like the massacre in the Santa MarõÂa de Iquique School ± were characterized not only
by the amount of tokens paid as salary, but also by the workforce's desire to improve
the `rate of exchange' between tokens and food.
17 In 1835, after visiting the nitrate deposits in TarapacaÂ, Darwin wrote: `The chief
expense is its transport to the sea coast. The mine consists of a hard stratum, between
two and three feet thick, of the nitrate mingled with a little of sulphate of soda and a
good deal of common salt. It is close beneath the surface and follows for a length of
150 miles the margin of a grand basin or plain' (1845, p. 41).
18 These ironically proved very useful in the Chilean advance into TarapacaÂ. After the
war, Colonel North, the `Nitrate King', incorporated them into his nitrate empire,
when in April 1887 he purchased 58 per cent of their shares.
Gabriel Palma 259
19 To get an idea ± in relative terms ± of the quantity of resources which were allocated
to this work, it may be noted that relative to population the Chilean network was 2.8
times larger than Great Britain's at this time.
20 The fall of exports was so extraordinary that, according to a League of Nations study,
the 1929 Depression affected the Chilean economy not only more than any other in
Latin America (as was the case for volume exported), but more than any other in the
world. See Ellsworth (1945, p. 23).
21 Even the financial bases upon which North constructed his nitrate èmpire' were of
local origin; it was funds provided by the Iquique branch of the Bank of ValparaõÂso
which allowed North to purchase the `keys to his kingdom', the famous nitrate
certificates which gave him control of the oficinas Primitiva, Peruana, RamõÂrez,
Buen Retiro, Jazpampa and Virginia, among others. For a detailed account of North's
manoeuvres, see Blakemore, (1974).
22 From 1840 onwards agricultural and livestock exporters paid no export taxes at all,
and of their connected processing activities only leather paid a 6 per cent tax, while
mining exports also came to pay a maximum 6 per cent ± this being the rate paid by
processed copper. This was the reason why during the 1870s, for example, export
taxes were, on average, equivalent to just 1.4 per cent of the value of exports, and that
they only contributed 3.3 per cent of òrdinary' public revenues. See Palma (1979).
23 For the problems faced previously by Chilean exporters of copper, particularly in the
British market, see Palma (1979).
24 To be spent financing North's rapid upward social mobility, his purchase of honorific
titles, his sumptuous gifts ± such as a famous piano for Queen Victoria ± or his lavish
entertaining of the British oligarchy. Blakemore (1974) provides a vivid account of
this side of North's life.
25 For an analysis of economic ideology in Chile since Independence, particularly with
respect to the role of manufacturing in economic development, see Palma (1979).
26 The characteristics of the production costs of nitrates can be found in RamõÂrez
Necochea (1958), Blakemore (1974), FernaÂndez (1978), Palma (1979), Cariola and
Sunkel (1982), and Pinto and Ortega (1991).
27 The taxes applied to exports from the Gran MinerõÂa in this period were only equival-
ent to approximately 7 per cent of the gross value of production. See Reynolds (1965,
pp. 363, 365).
28 Created in 1883, this association of manufacturing entrepreneurs was the successor of
the Junta de Fabricantes.
29 This was a direct response to the opening of the Trans-Andean Railway through the
Uspallata pass. As a result of this new railway, import of cattle had increased by 75 per
cent in just one year (to 74 000 head). For a justification of the need for this tax, see
Sociedad Nacional de Agricultura, 1886, p. xvii. For an analysis of the situation, see
Wright (1975. pp. 50±1).
30 For an interesting study of the influence of Litzen's ideas on tariff policy in the
period, and on the economic thinking of some politicians, MalaquõÂas Concha in
particular, see GarcõÂa-Huidobro (1972).
31 On 59 of these agricultural and manufactured products, this law placed an ad valorem
tax of 60 per cent; it also placed a tax of 35 per cent on manufactured products which
mainly served as inputs for other products of the same sector. It left a standard tariff
of 25 per cent for products not otherwise specified, and divided raw materials and
machinery into three categories, placing 54 at 15 per cent, 66 at 5 per cent, and
leaving 118 free of all import duties. Finally, specific taxes were placed upon 28
products; the taxes in these cases, as Kirsch reminds us, `were usually much higher
than the 60 per cent ad valorem category' (1973, p. 221). For the text of the law see
Anguita, 1912 (vol. III), pp. 400±7. For an analysis of the political situation which led
to the passing of this law, see Eyzaguirre (1957). For an analysis of the law, see
Sociedad de Fomento Fabril (1901) and Ho È rmann (1901).
32 Imports of cattle, for example, which had reached an average of over 100 000 head in
1895±96, dropped to just 43 000 in 1898 and less than 28 000 in 1899.
33 The protection of the preserved food industry seems to have stimulated local produc-
tion to such an extent that Chile became an exporter of these products. For example,
NicolaÂs Rubio's FaÂbrica Nacional de Conservas Alimenticias (located in his hacienda El
Estanque) exported a significant proportion of its output of two million tins of
preserved foods both to Europe and Latin America; see Palma (1979, p. 297).
34 One criticism of the tariff policy before the First World War that could be made is that
in protecting manufacturing activities producing mainly final consumption goods, it
discriminated against the production of inputs and machinery; but this would be a
criticism against the specific type of protectionism followed at the time and not
against the existence of a protectionist policy in itself.
35 There are several estimates of the movement in the level of prices of the period (for an
analysis of them, see Wagner, 1992). From these indices we have constructed a new
one which is the average of three indices: that of wholesale agricultural prices of
Latorre (expanded by Wagner); Wagner's `British parity' index (which measures the
changes in the nominal exchange rate and prices in Great Britain); and Clavel and
Ferraro's 1990 index (which measures consumer prices, mainly of food, household
items and clothing). In our opinion our index is a better approximation of the general
level of domestic inflation than the individual indices, as it includes a wider variety of
products and services, and because it averages different biases in previous indices (the
first two indices mentioned have a wholesale bias and the third one a consumer-
price-index bias).
36 There is an extensive literature on some aspects of this inflationary phenomenon
in Chile, in particular as regards the monetary anarchy which prevailed through-
out (see, for example, Hirschman, 1963). The growing need to impose some order on
the monetary chaos of the country led to the recourse in 1924 to the famous
Kemmerer Mission; this created the Central Bank and engineered the return
to the gold standard in 1925 (see Fetter, 1931). For the legendary controversy between
oreros (the proponents of the system of convertibility of the peso) and papeleros (their
opponents), see GarcõÂa-Huidobro (1972), Bustos (1988) and Drake (1989).
37 As already mentioned, as a result of the War of the Pacific, Chilean exports trebled in
real terms in just a few years. To put this into perspective, this expansion of exports is
similar to the one oil-exporting countries had in 1973±74. In the debate of the effect
of the discovery of oil on the North Sea in the British economy, an important
distinction was made between the possible outbreak of a `Dutch Disease' resulting
from a primary-export expansion based on a product already exported by a country
and that, as in Great Britain, based on a new export product. Although in Chile the
export boom was not caused by the `discovery' of a new product, the Chilean case
resembles the British in that it was based on the incorporation of one into the
national economy. See, for example, Worswick (1980), Byatt et al. (1982) and Pesaran
(1984). For studies of export booms contemporary to nitrates in other Latin American
countries, see Baham and Coomes (1995).
38 For a proof of these points, see Palma (1998).
39 The tendency to revalue the real exchange rate could be made worse by speculative
pressures on the currency based on good prospects for exports and the balance of
Gabriel Palma 261
payments (see Corden, 1977). However, the fact that national and international
financial markets at the time were not properly developed reduced the possibility
of this situation taking place.
40 Added to this, of course, is the fact that throughout this period the Chilean economy,
as with most economies in Latin America, was increasing its degree of integration into
the world economy. This process was mostly the result of technological and institu-
tional changes. This long-term process was increasing its income elasticity to import, a
process which also helped to absorb additional foreign exchange produced by nitrates.
41 As already mentioned, following the international financial crisis of 1873 it was not
possible to get new foreign loans; in fact, the only one which was possible to acquire
between 1873 and the War of the Pacific in 1875, was due to the fact that it was
necessary for the amortization of an already existing loan ± secured in 1867. As a
result, despite the crisis, the country had to generate a surplus in its trade account in
order to accumulate the foreign exchange needed to service the existing foreign debt.
42 Yet another aspect of public expenditure which proved to be very conflictive, parti-
cularly with the rural oligarchy, was its effect on the labour market in the country-
side. Until then, wages in the agricultural sector were not only extremely low but
were paid in kind ± as part of a system that tied the peoÂn to the hacienda. As a result of
the rapid growth of public expenditure, particularly in infrastructure, demand for
labour in the rural sector increased so fast that landowners not only had to increase
wages to retain their labour, but also had to begin to pay at least part of wages in
money. This rise in agricultural wages, their partial monetarization, and the increased
labour mobility in the countryside marked the beginning of capitalist relations of
production in the agricultural sector; but these were `modernizations' that big land-
owners never forgave Balmaceda for.
43 In the special case of nitrates, this measure did not act as a disincentive to foreign
investment for three reasons. Firstly, as already mentioned, foreign investment
proper was not significant in nitrates anyway ± North financed the purchase of his
oficinas with a loan secured in Chile. Secondly, even after this level of taxation, profits
of foreign entrepreneurs were still very large ± equivalent to approximately one-third
of the value of production. Thirdly, these entrepreneurs, in any case, could not
produce nitrates in another part of the world as Chile was the only country with
economically exploitable nitrate deposits.
44 Among the internal taxes abolished by Balmaceda were excise and mining licences.
When he took over the presidency, these two taxes contributed almost half of internal
taxes. Yet another source of conflict between Balmaceda and big landowners was the
fact that Balmaceda (despite being himself a large landowner) did not abolish (or even
reduce) the other major internal tax, that on agriculture ± in 1885 this contributed
about 40 per cent of internal taxes. Given the nature of the political coalition that
deposed Balmaceda, it should not be surprising to find out that one of the first
measures taken by the new government was to abolish this tax on agriculture.
45 For a detailed study of these arguments, see Palma (1979).
46 The main difference between Meller's and our understanding of the role of fiscal
policy in this period is that, in our opinion, whatever might have been its subjective
motivation, its actual effect was positive for the Chilean economy. In other words,
what really matters is not whether the reduction in internal taxes ± and in the tariffs
of some public services ± took place due to `crowding out', `populism' or by the
conscious design of fiscal policy (in our opinion, at least in Santa MarõÂa's and
Balmaceda's governments it was due fundamentally to the last), or if the rapid
increase in public expenditure created a culture of lax fiscal `discipline'. What matters
is that in the short and medium term the effect of this fiscal policy was clearly
positive ± and in the long term, as Keynes used to say, the protagonists of this fiscal
policy would not be around . . .
47 Between the middle of the 1880s and 1900, for example, production of potato doubled
in volume and that of barley and beans grew by 12 and 17 times, respectively.
References
Albert, B. (1976) Àn Essay on the Peruvian Sugar Industry, 1880±1920', mimeo, Norwich,
University of East Anglia.
Anguita, R. (1912) Las Leyes Promulgadas en Chile desde 1810 hasta el 1 de junio de 1912, 4
vols, Santiago: Imprenta Barcelona.
Barham, B.L. and Coomes, O.T. (1995) `Reinterpreting the Amazon Rubber Boom: Invest-
ment, the State, and the Dutch Disease', Latin American Research Review, vol. 29, no. 2,
pp. 73±109.
Behrman J.R. (1976) Foreign Trade Regimes and Economic Development, Chile/New York:
Columbia University Press.
Bethell, L. (ed.) (1993) Chile since Independence, Cambridge: Cambridge University Press.
Blakemore, H. (1974) British Nitrates and Chilean Politics, 1886±1896: Balmaceda and North,
London/Santiago: The Athlone Press.
Blomstrom, M. and Meller, P. (eds) (1990) Trayectorias Divergentes: comparacioÂn de un siglo
de desarrollo econoÂmico Latinoamericano y Escandinavo, Santiago: CIEPLAN-Hachette.
Bustos, C. (1988) Monetary Problems of an Export Economy. Seventy Years of Monetary Debate
in Chile, 1844±1914, M.Phil dissertation, University of Hull.
Byatt, I. et al. (1982) `North Sea Oil and Structural Adjustment', working paper no. 54,
London: Government Economic Service.
Cariola, C. and Sunkel, O. (1982) ÈxpansioÂ n salitrera y transformaciones socioecono- Â
micas en Chile, 1860±1930', mimeo, University of Sussex.
Chile, EstadõÂstica Comercial (various years), ValparaõÂso: Superintendencia de Aduanas.
Chile, Congreso Nacional (various years), Santiago: Imprenta del Ferrocarril.
Chile, Memoria del Ministerio de Relaciones Exteriores (various years), Santiago: Imprenta
Nacional.
Clavel, F. and Ferraro, L. (1990) Ùna nueva versioÂn de los precios en el siglo pasado',
Encuentro Annual de Economistas de Chile, 1990, Santiago: University of Santiago.
Collier, S. (1993) `From Independence to the War of the Pacific', in L. Bethell (ed.), Chile
since Independence, Cambridge: Cambridge University Press.
Corden, W.M. (1977) Inflation, Exchange Rates and the World Economy, Oxford: Clarendon
Press.
Corden, W.M. and Neary J.P. (1982) `Booming Sector and de-Industrialization in a Small
Open Economy', Economic Journal, vol. 92, pp. 825±48.
Culver, W. (1982) `The Rise and Decline of Mining Regions and Mining Policy: Chile's
Role in the Nineteenth Century World Copper Trade', mimeo, New York.
Darwin, C. (1845) The Voyage of the Beagle, Journal of Researches into the Natural History and
Geology of the Countries Visited during the Voyage Round the World on HMS `Beagle',
London: Amalgamated Press Ltd.
Dennis, W.J. (1931) Tacna and Arica, New Haven: Yale University Press.
Drake, P. (1989) The Money Doctor in the Andes: The Kemmerer Missions, 1923±1933,
Durham, NC/London: Duke University Press.
Ellsworth, P.T. (1945) Chile, an Economy in Transition, New York: Macmillan.
Gabriel Palma 263
Encina, F.A. (1940±52) Historia de Chile desde la Prehistoria hasta 1891, Vols. 1±20, Santiago:
Editorial Nacimiento.
Eyzaguirre, J. (1957) Chile durante el Gobierno de ErraÂzuriz EchaÂurren, 1896±1901, Santiago:
Zig-Zag.
FernaÂndez, M.A. (1978) The Development of the Chilean Economy and its British Connections,
1895±1914, doctoral Dissertation, University of Glasgow.
Fetter, F.W. (1931) Monetary Inflation in Chile, Princeton: Princeton University Press.
Garcia-Huidobro, G. (1972) El Desarrollo EconoÂmico Chileno durante el Siglo XIX: Una
Historia CrõÂtica, BA dissertation, University of Chile.
Gibbs, A. Archives of Messers Antony Gibbs and Son Ltd., Guildhall Library, London.
Girard, L. (1965) `Transport', The Cambridge Economic History of Europe, Vol. VI, London:
Hinkelammert, F. (1970) El Subdesarrollo Latinoamericano: Un caso de Desarrollo Capitalista,
Santiago: Ediciones Nueva Universidad.
Hirshman, A.O. (1963) Journeys Towards Progress: Studies of Economic Policy Making in Latin
America, New York: Twentieth Century Fund.
HoÈ rmann, J. (1901) Èfecto de la ley de impuestos del 23 de diciembre de 1897; el progreso
de la industria nacional', El Mercurio, ValparaõÂso, 30 August.
Janneret, T. (1972) Èl sistema de proteccioÂn a la industria chilena', in O. Mun Äoz, Proceso a
la IndustrializacioÂn Chilena, Santiago: Ediciones Nueva Universidad.
Keynes, J.M. (1985) Collected Works, Vol. 15, London: Macmillan.
Kindleberger C. (1984) `The 1929 World Depression in Latin America ± From the Outside',
in R. Thorp (ed.), Latin America in the 1930s: The Role of the Periphery in World Crisis,
London: Macmillan.
Kirsch, H. (1973) The Industrialisation of Chile, 1880±1930, doctoral dissertation, Univer-
sity of Florida.
Loveman, B. (1988) Chile, the Legacy of Hispanic Capitalism, 2nd edn, New York: Oxford
University Press.
Mamalakis, M. (1976) The Growth and Structure of the Chilean Economy: From Independence
to Allende, New Haven: Yale University Press.
Meller, P. (1990) Ùna perspectiva de largo plazo del desarrollo econoÂmico chileno, 1880±
1990', in M. Blomstrom, and P. Meller, Trayectorias Divergentes: comparacioÂn de un siglo de
desarrollo econoÂmico Latinoamericano y Escandinavo, Santiago: CIEPLAN-Hachette.
Mayo, J. (1982) `Commerce, Credit and Control in Chilean Copper Mining before 1880',
mimeo, Barbados.
Mun Äoz, O. (1972) Proceso a la IndustrializacioÂn Chilena, Santiago: Ediciones Nueva Uni-
versidad.
Ortega, L. (1979) Change and Crisis in Chile's Economy and Society, 1865±1879, doctoral
dissertation, University of London.
Palma, J.G. (1998) À Keynesian-Macroeconomics View of the Dutch Disease', mimeo,
Cambridge. Department of Applied Economics.
ÐÐ (1984) `From an Export-Led to an Import-Substituting Economy: Chile 1914±1939',
in R. Thorp (ed.), Latin America in the 1930s: The Role of the Periphery in World Crisis,
London: Macmillan.
ÐÐ (1979) Growth and Structure of Chilean Manufacturing Industry from 1830 to 1935:
Origins and Development of a Process of Industrialisation in an Export Economy, doctoral
dissertation, Oxford University.
ÐÐ (1978) `Dependency: A Formal Theory of Underdevelopment, or a Methodology for
the Analysis of Concrete Situations of Underdevelopment?' World Development, vol. 6,
nos 7/8, July/August.
Pesaran, M.H. (1984) `Macroeconomic Policy in an Oil-Exporting Economy with Foreign

Exchange Controls', Economica, vol. 51, pp. 253±70.
Pinto, A. (1959) Chile, un Caso de Desarrollo Frustrado, Santiago: Editorial Universitaria.
Pinto, J. and Ortega, L. (1991) ExpansioÂn Minera y Desarrollo Industrial: un caso de creci-
miento asociado, (Chile 1850±1914), Santiago: University of Santiago.
RamõÂrez H.N. (1958) Balmaceda y la ContrarrevolucioÂn de 1891, Santiago: Editorial Univer-
sitaria.
Reynolds, C.W. (1965) Development Problems of an Export Economy: The Case of Chile and
Copper, Illinois: Richard D. Irwin.
Rojas, M.E. (1982) Sueldos y Salarios, 1860±1940, BA dissertation, University of Chile.
Sater, W.F. (1990) Chile and the United States: Empires in Conflict, London: The University of
Georgia Press.
Saul, S.B. (1976) The Myth of the Great Depression, 1873±1896, London: Macmillan.
Sociedad de Fomento Fabril (various years) BoletõÂn de la Sociedad de Fomento Fabril,
Santiago.
Sociedad Nacional de Agricultura (various years) BoletõÂn de la Sociedad Nacional de Agricul-
tura, Santiago.
Thorp, R. (ed.) (1984) Latin America in the 1930s: The Role of the Periphery in World Crisis,
London: Macmillan.
VeÂliz, C. (1963) `La mesa de tres patas', Desarrollo EconoÂmico, vol. 3, April±September.
Wagner, G. (1992) Trabajo, ProduccioÂn y Crecimiento. La EconomõÂa Chilena, 1860±1930,
Santiago: Universidad CatoÂlica de Chile.
Worswick, G.D.N. (1980) `North Sea Oil and the Decline of Manufacturing'; National
Institute Economic Review, vol. 94, pp. 22±6.
Wright, C.T. (1975) Àgriculture and Protectionism in Chile, 1880±1930', Journal of Latin
American Studies, vol. 7, May, pp. 45±58.
9
The Vicissitudes of an Exporting
Economy: Argentina (1875±1930)
Roberto CorteÂs Conde
Introduction: growth and decline
At the beginning of the First World War, Argentina had gone through several
decades of continuous growth ± only briefly interrupted during the two years of
the 1890s crisis. At the time, Argentina's 7 per cent overall growth rate ± 3.8 per
cent per capita ± was higher than that of the USA and Australia and the majority
of European countries.
It is common currency that the availability of fertile land and the incorpora-
tion of capital and labour were the bases for growth. However, in explaining it,
institutional factors, preconditions for the arrival of capital and labour, have
not received equal attention. Political stability and judicial security, the respect
of property rights ± which guaranteed that earnings and profits would not be
touched ± all came into effect some time between the inauguration of the
Constitution and the country's definite organization in 1880. It would seem
that the explanation given by DõÂaz Alejandro, in a classic book, continues to be
valid:
Pre-1930 Argentine experience fits the broad outlines of the staple theory of
trade and growth . . . it appears that growth then was related closely to succes-
sive booms in the export of land-intensive commodities, with land having a
very low opportunity cost. The economic usefulness of pampean land was not
discovered overnight, as in the case of an oil deposit, but arose as a result of a
combination of growing European needs for primary goods, technological
progress in transport and an increasing interest by Argentine policy makers in
promoting exports, foreign investments and immigration.
(DõÂaz Alejandro, 1970, p. 9)
[and moreover]: Pre-1930 growth can be said to have been èxport led'
not because exports and their associated capital inflows provided growing
265
266 Argentina 1875±1930
aggregate demand (in a Keynesian fashion) but because, more fundament-

ally, exports and capital inflows led to an allocation of resources far more
efficient than the one which would have resulted from autarkic policies. In
particular, the domestic cost of capital goods, which would have been astro-
nomical under autarky, say in 1880, was reduced to a low level by exports of
commodities produced by the generous use of an input ± land ± whose
economic value under autarky would have been quite small.
(Ibid., p. 11)
While growth seems to have corresponded to the neo-classical notion of in-

corporating land, capital and labour, this version does not give a complete view
of the phenomenon. It is true that growth in agriculture was due to the
incorporation of land, but the expansion of the cattle industry in the 1900s
was derived not from more land but from changes in land use and an increase
in the quality and value of capital inputs ± cattle.1 The same land produced
goods of better quality and greater value. Moreover, growth did not simply
result from the incorporation of labour: human capital increased, the result of
an important investment in education.
In 1914 there was an interruption of this growth until 1917, followed by a
recovery which continued up to the world crisis in 1929±30. Analysing this
phenomenon, several authors, including Bunge (1920) and Di Tella (1970),
have argued that extensive growth had reached a ceiling by the First World
War and that a long process of decline had begun at that point. DõÂaz Alejandro,
in the book mentioned above, refutes Di Tella's arguments, pointing out that
what happened in the 1920s was not due to factors which exclusively affected
Argentina, or to policies adopted. Rather, the factors in question were a world-
wide phenomenon, and were particularly strong in the countries relatively
open to international trade. Instead, DõÂaz Alejandro argued, the decline was
caused by the policies adopted from 1930 onwards, which had a negative
influence on the subsequent economic evolution of the country.
In an earlier article, we agreed with DõÂaz Alejandro: the postwar slowdown
was indeed a generalized phenomenon (Maddison, 1991). Further, in the inter-
war period the slow down of growth in the Argentine economy coincided with
a slowdown in the developed economies, whereas following the Second World
War the divergence between local growth trends and those of the rest of the
world is well-known (CorteÂs Conde, 1997). In this chapter, we shall concentrate
on finding explanations for this early change in the trend in the middle of a
successful period of growth.
What caused, then, the strong fall and the feeble recovery which composed
the change in the trend? Was supply, demand or both to blame? On the supply
side, it has been said that growth declined when the frontier was reached: the
bulk of fertile land had been exploited (Bunge, 1920 and Di Tella, 1970). It has
Roberto CorteÂs Conde 267
also been alleged that investment declined. Two reasons are usually given. First,
the end of the frontier meant that the possibilities for exceptional profits
ended. Second, the downturn in the international flow of savings was not
compensated by a greater capacity for domestic savings, due to the high
dependency ratio of a young population (Taylor, 1992). On the side of demand,
it is argued that the new trends of world trade and their negative effect on
countries that relied on exports of primary products are largely to blame ±
although this factor was more influential in the period after 1930.
Other arguments point to rigidity-inducing institutional factors, which ham-
pered the adaptation to new market conditions and to changed resource avail-
ability. Argentine social structure and patterns of land ownership, particularly
the absence of family farming, have been generally blamed (Solberg, 1987).
Moreover, it has been claimed that the sectors traditionally linked to agricul-
ture had opposed protectionist policies which favoured industrialization. In
this chapter we will take up some of these arguments.
Growth trends
The estimates we have constructed (CorteÂs Conde, 1994) allow growth to be

measured from 1875. From this date until the First World War, Argentina
witnessed an unusually long period of high growth, only briefly interrupted
by the 1890 crisis. Between 1875 and 1913 the economy grew by 6.7 per cent
per annum and 3.7 per cent per capita ± both based on three-year averages ± a
difference accounted for by the large increase in population. Growth was
interrupted during the war years but soon recovered. However, the rupture
had produced a change in trend. In the following period (1913±29), growth
slowed down and between peaks ± three-year averages ± was practically non-
existent (0.46 per cent in per capita terms).
Argentina was a small open economy, with an initial resource endowment
characterized by land abundance and scarce population and capital. This
endowment was modified as land was settled and population and capital
from overseas grew. If the process is observed only as a linear trend, that is to
say if the growth rates are compared between peaks, the observations concern-
ing the intermediate years and their fluctuations are lost. We need this informa-
tion to explore the circumstances which influenced each of the cyclical
fluctuations. Let us now look at each of these cycles and phases.
The rate of increase of per capita GDP between 1875 and 1889 was 6.2 per
cent per annum. It fell during the crisis years of 1889±92 to 3:2 per cent per
annum. From 1892 until 1912 growth averaged 3.5 per cent per annum. The
upward trend was interrupted in 1913 by a fall, which lasted until 1917 (with
6:6 per cent per annum), which represented a particularly harsh recession.
From 1917 activity recovered and grew at 2.6 per cent per annum, until the
268 Argentina 1875±1930
world crisis in 1929. Growth in the 1920s was only slightly slower than between
1892 and 1912. We now turn to analyse in detail both the period of accelera-
tion, 1875±1913, and the slowdown, 1913±27.
The period of acceleration, 1875±1913

In the period from 1875 to 1912 agriculture and industry accounted for the
bulk of growth, followed by construction. Agriculture was subject to sudden
fluctuations, due to weather conditions, and construction fluctuated more
widely. Until 1913 it grew by 6.4 per cent per annum and transport by 10 per
cent. A very important part of this growth was investment-led. In new coun-
tries, where land is exploited in relatively short periods of time, the initial
requirement of investment in infrastructure and transport is very large, but it
is not renewed for a long time ± until capital goods become obsolete. Most
infrastructure development, including public and private construction, took
place in the last decades of the nineteenth century and the early twentieth
century. The rate of growth of commerce, on the other hand, came from the
increase in activity in the rest of the economy.
Cattle-raising experienced a 2.9 per cent growth rate, and in the most sig-
nificant area ± meat production ± growth was not just a consequence of an
increase in the rate of slaughter, but of greater value added. In 1895, the value
added by cattle represented 11 per cent of the total, whilst in 1914 it reached 15
per cent. In agriculture, by the end of the 1900s, the rate of land incorporation
had begun to fall prompting Bunge to argue that from 1908 the growth rates of
cultivated land and the population started to diverge, a sign of the coming
decline (Bunge, 1920). However, Bunge failed to acknowledge that land could
have different uses: the frontier could continue expanding based on other crops
± with greater value per hectare and different yields. Alternatively, fields could
be turned into pasture land, as had happened with cattle-raising.
Deceleration, 1913±27
In 1913, the upward phase was interrupted. In 1914, GDP fell 13 per cent. The
following year witnessed a recovery, GDP growing 7.6 per cent. However, in
1916 and 1917 GDP experienced falls of 5.5 per cent and 15 per cent respect-
ively, but from that year onwards GDP grew at a rate of 4.5 per cent per annum ±
measured as a three-year average ± until 1929 when, due to the world crisis, this
trend was again interrupted.
What factors brought about the fall of total GDP by 2.3 per cent per annum
between 1913 and 1917? What factors contributed to the subsequent recovery?
Was it a recovery, or a quite independent growth spurt? If so, why was it
interrupted? Could it be because the long phase of Argentine decline had
already begun in 1913? Was it due to an ephemeral recovery in the 1920s,
unsustainable in the context of the existing economic structure? Were there
domestic causes, changes in factor endowments or other causes which slowed

down savings, investment and growth? Were there market or institutional fail-
ures which were in any case halting the export boom?
In our opinion, the dramatic downturn was due to unique circumstances
which would not necessarily have affected long-term growth. One such circum-
stance was the supply-side effect the World War had on an open economy, that
is, a shortage of inputs, and not, as is often claimed, the demand-side effect:
Argentine products were still in demand in the markets of the allied countries.
Indeed, during the 1900s there had been a change in the direction of trade
towards the countries of Central Europe, which, with the war, had become
isolated. The supply constraints were felt in all the countries which produced
for the war and affected, directly or indirectly, items such as transport, trade,
government ± because of the impact of imports on fiscal income ± and con-
struction. The initiation of hostilities produced a shortage of inputs, affecting
the industries which used the imports, but also the exporters of foodstuffs
owing to the lack of ships. Moreover, the cost of freight increased; and after
the end of the war the higher cost was maintained. This harmed Argentina,
whose nineteenth-century expansion was linked to the drastic reduction of
maritime transport costs. The fall in transport and in trade was due to the
generalized fall in economic activity. Construction suffered not only because
of this, but also because of pessimistic expectations facing a complex panor-
ama, in which an unusual repeated crop failure ± two of wheat and one of corn
and linseed in three years ± were linked to the world conflagration and a lower
expansion in trade. Because of this agricultural GDP fell by 12 per cent in 1914
and 55 per cent in 1917.
Adjustment brought important changes following the war. Growth slowed
down in the sectors responsible for the earlier expansion ± cereals and railways.
Construction continued to reflect pessimistic expectations. Industry recovered
by substituting inputs; investment in railways, automobiles and telephones
increased; and as far as agriculture was concerned, the expansion of cotton
cultivation is noteworthy.
There is little in the evolution of the economic activity of the 1920s to
indicate that growth would suddenly halt in 1929. Indeed, while the crisis of
1929±30 can be seen as a rupture, the growth which marked the period after
1932 can just as easily be interpreted as a continuation of the earlier growth.
Industrial investment, cotton cultivation and the expansion of the textile
industry all bear witness to this continuity. In effect, the industrial develop-
ment of the 1930s corresponded to a substitution of imported inputs rather
than final goods. For example, the lower cost of cotton stimulated textile
manufacturing. In other words, changes in the 1920s ± investment in industrial
machinery and cotton ± were responsible in one way or another for the expan-
sion of the later period.
270 Argentina 1875±1930
Terms of trade
A further element we need in order to study the evolution of GDP and of
income ± for its effect upon welfare ± is the effect of the terms of trade. In
Figure 9.1 we adjust GDP, deflating export values (which are not at constant
prices) by their purchasing power (the ratio of export prices to import prices).
From 1915 to 1927 the terms of trade were unfavourable to exports. Therefore
income ± defined here as GDP adjusted by the terms of trade ± lagged behind
GDP in the postwar period. At the point of greatest divergence, 1920, income
was 14 per cent below GDP: a consequence of the contraction brought about by
the increase in USA interest rates. It is perhaps important to point out that
Argentina responded to the price fall by producing in greater volume: thus,
neither output nor income fell. From 1922 onwards, the terms of trade
improved, returning to prewar levels in 1928.
Factors determining growth

Supply
As we have pointed out, the accepted view holds that growth resulted from the
accumulation of factors, particularly the abundant supply of land. We have
seen that, while true, this analysis is insufficient: a change in land use rather
than greater amounts of land accounts for the growth in cattle-raising. It has
also been said that this growth continued as long as the frontier could continue
expanding, but that when all the fertile land was eventually exploited the
9500
GDP in millions of pesos
GDP adjusted
8500
7500
Million pesos
6500
5500
4500
3500
1913 1915 1917 1919 1921 1923 1925 1927 1929 1931 1933 1935 1937 1939
Figure 9.1 Argentina: GDP adjusted by terms of trade, 1913±39

Source: CorteÂs Conde (1997).
Table 9.1 Argentina: production of cereals, fodder and cotton, 1900±29

(annual rate of growth)
Product % %
Wheat 1900±13 14 1913±29 3.9

Corn 1900±13 14 1912±27 0.6
Linseed 1900±13 12 1913±28 4.8
Alfalfa 1900±13 11.7 1913±27 13
Cotton 1900±13 6 1913±30 23
Source: CorteÂs Conde (1994).
decline began. This argument simplifies a more complex process. Let us explore
whether there existed supply-side limitations to growth and the validity of the
ènd of the agricultural frontier' argument.
Between 1875 and 1900, wheat and corn, followed by alfalfa and linseed,
were responsible for agricultural growth. In the years 1914, 1916 and 1917,
three crops failed. In the following year, production recovered owing to better
weather conditions, but the area under cultivation did not increase as fast as in
the first decade. Production of the principal crops in the pre- and postwar years
developed as shown in Table 9.1.
According to Bunge, the peak in extensive growth was reached in 1908. His
argument was based on a comparison of the divergence in the rates of cultiv-
ated land expansion and population (Bunge in Llach 1985). In fact, however,
the slower incorporation of land for cereal farming coincided with the expan-
sion of other more intensive crops ± and of greater value per ton ± like cotton.
Data on cultivated land do indeed point to the end of the frontier. A similar
conclusion is reached if we compare output growth between 1913 and 1929.
Nonetheless, if the peak of agricultural expansion is taken into account ± 1924,
not 1929 ± the situation is different. Compare the evolution of wheat ± a cereal
± and cotton, a new industrial crop, in Table 9.2.
Between 1913 and 1924, the value of the combined production of wheat and
cotton grew by 10.6 per cent per annum. Moreover, the extension of cotton
Table 9.2 Argentina: wheat and cotton
Wheat Cotton Total
production price in value in production price in value in value in

in thousands pesos thousands in thousands pesos thousands thousands
of tons per ton of pesos of tons per ton of pesos of pesos
1913 2 850 88.3 251 655 1.8 145 261 251 916
1924 5 202 143.3 745 447 51.1 335 17 119 762 565
1929 4 425 87.9 388 958 115.4 215 24 811 413 769
Source: Author's elaboration using data from the Anuario GeograÂfico (1941).
272 Argentina 1875±1930
Table 9.3 Argentina: cultivated areas of cotton and textile production
Cultivated area cotton 1921 100 Textile production 1921 100
1921 100.00 100.00

1922 65.44 113.72
1923 93.73 114.34
1924 262.61 99.54
1925 438.03 119.43
1926 461.27 121.86
1927 300.70 147.98
1928 356.74 256.23
Source: Banco de la NacioÂn Argentina, Report and Balance Sheet (1923) for `Cultivated area'; author's
own elaboration of textile production data.
farming had a considerable influence on the development of the textile indus-

try in this decade and the two following (Table 9.3).
The better use of land and the improvement in the quality of the cattle stock,
rather than the incorporation of land ± as had been the case in 1880 ± were
responsible for the increase of livestock output. Greater value added in meat
was obtained: the capital stock ± cattle ± had increased owing to cross- breeding
with British breeds. The 87 per cent increase in the price of steers between 1900
and 1914 was a result of the change in the quality of the product, from the meat
of criollo cattle to that of crossbred. Between 1888 and 1914 there was no
increase in cattle stocks, and meanwhile, sheep stocks decreased. Cattle output
rose by 2.9 per cent a year between 1875 and 1913; in 1895 the added value for
cattle was 11 per cent, and by 1914 it had risen to 15 per cent.
Part of the increase in capital ± pastures ± was not registered in the conven-
tional capital measurements,2 or in GDP estimates at 1950 prices. The studies
concerning capital in livestock in the second and third national censuses do
take it into account, although not in disaggregated form. It is estimated at
between 1.3 billion pesos (1914 prices) in 1895, and 3.7 billion in 1914 (a rate
of increase of 7.2 per cent per annum). Meat was also the raw material of the
meat-packing plants which led the growth of the foodstuffs industry in
the 1900s. In the postwar years, changes in international markets affected the
industry, initiating a prolonged crisis.
To follow the evolution of cattle-raising, we need to take into account the
characteristics of the cattle cycle. An increase in production is not the con-
sequence of expansion and better prices; on the contrary, when prices rise,
production falls and stocks increase: an expansionary phase occurs. Indeed,
faced with higher prices, producers will retain stocks and cut production. In the
long run, however, an increase in stock will be reflected in an increase in
production. Cattle-raising was affected by the abnormal wartime market con-
ditions and from the mid-1920s by the protectionist measures in the USA. The
future of the industry was also to a degree under threat from Britain's imperial
preference (O'Connell, 1984). When external markets closed, domestic avail-
ability increased, prices fell and domestic consumption increased because the
supply of cattle from natural pastures is relatively inelastic ± it is not possible to
fatten the cattle forever, because the grass runs out. When stocks diminished
and external demand rose ± as at the end of the war ± there was a retention of
cattle, supply diminished, prices went on increasing and domestic consump-
tion decreased. Nonetheless, the contribution of cattle to the increase in GDP
did not result from an increase in production, but from the value of its product.
This phenomenon occurred in the 1900s and was over by the First World War. A
further change of this nature did not happen after the war.
Turning to capital endowments in the two subperiods, these depended upon
investment and both domestic and foreign savings. In Argentina, from the last
two decades of the nineteenth century to 1913, the negative current-account
balances coincided with positive balances in the capital account produced by
massive foreign investment inflows. In 1910, the Argentine Minister of Eco-
nomy ± JoseÂ MarõÂa Rosa ± ordered the preparation of an inventory of foreign
capital (MartõÂnez and Lewandowski, 1912). The inventory estimated that some
5059 million pesos had entered the country: 1572 million corresponded to
investments in Argentine bonds and 1827 million to railways. Banks, ports,
tramways and gas and electricity companies followed far behind. A further 364
million pesos were invested in land companies and mortgages. In 1917, Bunge
estimated that foreign capital amounted to 8813 million pesos of which 3055
million were in railways and 1493 million in bonds (Table 9.4). The bulk of
Table 9.4 Foreign investment in Argentina, 1917
Thousands of paper pesos
Loans and diverse Argentine bonds 1 493 871

Railways 3 055 287
Banks 117 934
Ports 50 372
Tramways 248 854
Meat packing plants 82 991
Services (gas, electricity, water etc.) 178 120
Land and mortgage companies 181 094
Mortgages 1 136 399
Insurance 8 832
Industrial establishments 1 154 000
Communications 48 500
Commerce and finance 1 057 202
Total 8 813 456
Source: Bunge (1928).

274 Argentina 1875±1930
railway investment took place before the war. In 1914, the national census,
estimated private ± foreign ± capital in railways as 3051 millions pesos and the
property of the state as 320 million (see Table 9.5 below).
After the war, technology was changing and new forms of transport and
communication were appearing; the railway and the telegraph began to be
replaced by automobiles, trucks and telephones. The number of inhabitants
per automobile in 1910±14 was 680, by 1939 it had dropped to 25. Moreover,
there had been changes in communications. Telegraph wires stagnated: from
31 000 kilometres in 1910±14, the figure had increased to 41 000 in 1915±19 and
45 000 in 1928, in contrast to the 293 per cent increase from 1910±14 to 1931.
Prior to 1935, the estimates of the balance of payments are not reliable
enough to deduce the movement of capital. However, it does appear that up
to 1914 there was a wide gap in the current account which was covered by
capital inflows.3 However, after the war, when foreign, mainly British, savings
ceased to perform the same role, was there not domestic saving? Was there a
decline in postwar domestic saving? Bank saving deposits, public bonds, mort-
gage bonds ± the income bond most used at that time ± and the volume of
shares in the stock exchange (Figure 9.2) can all be used as indicators of the
development of internal saving. We do not take into account here the saving
which stems from reinvested profits.
Domestic saving is subject to great fluctuations, influenced by special circum-
stances such as the financial deepening in 1903 after convertibility or the
withdrawals which coincided with the 1907 crisis. However, it cannot be said
that domestic saving diminished in the postwar period compared to prewar ± in
1,200
1,000
800
Million pesos
600
400
200
0
1900 1902 1904 1906 1908 1910 1912 1914 1916 1918 1920 1922 1924 1926 1928 1930
Figure 9.2 Argentina: saving trends in public bonds, mortgage bonds, shares and deposits
in savings accounts (three year average ± at 1914 prices)
Source: Author's own calculations using data from the Anuario GeograÂfico (1941).
this case we refer to institutional savings. On the contrary, there seems to have
been an important increase between 1915 and 1923.
Of course, the magnitude of domestic saving is not comparable to that
required for the installation of the railway network financed by foreign
savings in the 1900s. These investments had a long-term amortization, which
meant that the initial rhythm could not be maintained. This is a factor to take
into account when the investment of the first decades of the century is com-
pared to that of the postwar period. On the other hand, in the case of the
railways there was a problem of indivisibility. This explains why, at the begin-
ning, the effect of railway productivity on total capital productivity was low.
Here, the capital/output relation was 9.7 in 1885±89, 11.4 in 1900±04, 9.7 in
1910±14 and 6.2 in 1925±29.4 The paradoxical result was that as the quality of
capital increased, its marginal productivity did not diminish, but rather
increased. At the beginning installed capacity was underutilized, but later
increased as land entered production. The levels of monetization of the eco-
nomy were very high; around 50 to 60 per cent, with deposits at about 30 to 40
per cent.
The greatest investment was in infrastructure and principally in railways: in
1895 investment reached 1.2 billion pesos ± 1.3 billion pesos in 1914 prices ±
and in 1914 increased to 3 billion pesos. No less surprising were the changes in
the quality of stock and cattle installations whose value increased from 1.1
billion to 4.7 billion pesos between 1895 and 1914; in this case the investment
was local. The results were lower than the exceptional results obtained when
the quality was improved. In 1914 the process had been completed. There were
transport networks available which covered the production areas and an
amount of cattle stock capable of supplying the world market demand. The
bulk of investment had been carried out, and what remained required less
effort. Nonetheless, the investment carried out after the war in other sectors ±
industrial machinery, textiles, automobiles, telephones ± was just as important,
albeit of a different magnitude. Postwar Argentina still had to fully utilize its
enormous installed capacity.
Without doubt, the variations in the capital stock point to a slowdown in the
rate of investment. It must be taken into account, however, that it has not been
possible to compare the prewar period with the period which ended in 1930,
due to problems of information. In Table 9.5 the levels of stock in 1914 are
compared with those of 1935, when the country was still suffering the effects of
the 1930s crisis. Our estimations cover only some of the items shown in the
1914 census (Table 9.5). However, the slowdown does not necessarily imply
that domestic saving shortages are to blame. Instead, at least in this case,
savings were greater than investment. We conclude that the positive commer-
cial balances in the decade were a consequence of previous of trade deficits
financed by capital inflows (see Table 9.6).
276 Argentina 1875±1930
Table 9.5 Argentina: variation in the capital stock, 1914±35 (millions of pesos)
Annual
1914 1935 variation (%)
Livestock 3 202 3 288 0.24

Fixed installations 1 037 1 238 1.62
Private railways 3 051 3 190 0.41
State railways 320 931 10.19
Communications 21 350 29.14
Mach. and industrial raw mat. 1 788
Machinery and tools 405
Machinery, plant, accessories, etc. 2 639
Vehicles, moveable plant and tools 111
Total machinery 2 193 2 750 2.08
Total 9 824 11 747 1.64
Source: Author's own compilation based upon publications in the Revista de EconomõÂa Argentina and
the censuses of 1914 and 1935.
Turning to the issue of labour, the rate of growth of population according to

census data was 34.8 per thousand between 1895 and 1914, dropping to 20.4
per thousand between 1914 and 1947. More noteworthy was the decline
in immigration's contribution to this growth which fell from 17.2 per thousand
during the first period to only three per thousand in the second. Up to the
1900s, labour supply was very flexible (CorteÂs Conde, 1979) due to the succes-
sive waves of immigration. Between 1875 and the First World War, more than
three million people took up residence in the country (Table 9.7). During the
war the migration balance was negative; it began to recover very slightly
between 1918 and 1920 and in 1924 again reached prewar levels; in 1925 and
1926 it fell again, recovered in 1927, but since then has declined. The war
generated scarcities which opened new areas for exploitation. The need to
replace coal led to the felling of the Chaco forest; this required many workers,
provided by the northern provinces but also from neighbouring countries ±
mainly Bolivia. It is likely that the labour supply was less flexible during the
postwar years. Moreover, as the immigrants were mostly adults, their weight in
the labour supply gradually decreased. The reduced flexibility must have influ-
enced the negotiating power of the unions.
From 1869, as a result of the impact of immigration, the dependency ratio
declined. It was 45 per cent in 1969, fell to 42 per cent in 1895 and to 40 per
cent in 1914, reaching 31 per cent in 1939 (National Geographic Committee,
1941). The population matured because the majority of immigrants were adults
and male. This reflected positively on the increase of the proportion of the
active population and therefore on GDP. Moreover, it diminished the effect that
the increase in immigrants had on the fertility rate.
Table 9.6 Argentina: some items of the balance of payments, 1920±29 (millions of gold pesos)a
1920 1921 1922 1923 1924 1925 1926 1927 1928 1929
Exports 716.31 677.45 767.00 947.74 903.56 792.50 960.95 1 048.90 1 001.05 661.30
Imports 712.02 567.74 761.00 822.59 850.84 744.70 816.60 843.50 881.70 778.90
Balance 4.29 109.71 6.00 125.15 52.72 47.80 144.35 205.40 119.35 117.60
Servicing of debt 145.50 115.58 157.50 169.33 165.27 153.35 177.55 187.75 197.10 180.00
Total of services 201.50 161.08 201.50 218.33 202.27 190.35 236.55 241.75 248.10 223.00
Current account 197.21 51.37 195.50 93.18 149.55 142.55 92.20 36.35 128.75 340.60
Increase in debt 179.21 49.92 193.60 53.18 158.55 144.20 122.00 186.85 4.24 276.60
Var. in Reserves 18.00 1.45 1.90 40.00 9.00 1.65 29.80 150.50 124.51 64.00
GDP current prices 9 388 8 646 8 073 8 871 9 806 9 197 9 301 10 097 10 084 10 352
Var. reserves as a % of GDP 0.44 0.04 0.05 1.02 0.21 0.04 0.73 3.39 2.81 1.41
Net foreign outflow 22.29 111.16 7.90 165.15 43.72 46.15 114.55 54.90 243.86 53.60
Net external balance (% of GDP) 0.54 2.92 0.22 4.23 1.01 1.14 2.80 1.24 5.50 1.18
a
Five gold pesos equalled one pound sterling.
Source: Based on Phelps (1938).
277
278 Argentina 1875±1930
Table 9.7 Argentina: international migration
Immigration Emigration Balance Cumulative

balance
1857±60 20 000 8 900 11 100 11 100

1861±70 159 570 82 976 76 594 87 694
1871±80 260 885 175 763 85 122 172 816
1881±90 841 122 203 455 637 667 810 483
1891±00 648 326 328 444 319 882 1 130 365
1901±10 1 764 103 643 881 1 120 222 2 250 587
1911±20 1 204 919 935 825 269 094 2 519 681
1921±25 707 717 255 356 452 361 2 972 042
1926±30 689 698 286 058 403 640 3 375 682
1931±35 331 075 204 092 126 983 3 502 665
Source: Anuario GeograÂfico (1941).
How did market conditions affect the price of labour? Real wages were mainly
affected by exogenous factors. First, at the end of the 1880s and beginning of
the 1890s inflation had a negative effect ± even though mitigated by the
international decrease in food prices (CorteÂs Conde, 1979). By contrast, during
the second half of the 1890s, deflation resulted in an important increase in real
wages. In the 1900s, in spite of exchange rate stability, wages were subject to the
negative impact of the international increase in prices. From 1880 to the First
World War they increased by 30 per cent, 1 per cent a year in real terms (CorteÂs
Conde, 1979). These figures undermine the widespread thesis which holds that
wages fell during the period of massive immigration. During the same period
the economy was growing by 3.7 per cent per capita a year, probably owing to
the great flexibility in labour supply from abroad and to the fact that, comparat-
ively, Argentine wages were still higher than those in Italy and in Spain. This, as
well as much else, changed during the postwar period.
The wartime recession affected wages catastrophically. Industrial recession,
construction, the government and international inflation were all to blame.
Between 1914 and 1918, real wages fell by 11 per cent a year. It is likely that
employment was not as harshly hit, perhaps because of wage flexibility.
From 1918, the postwar recovery contributed to an increase in real wages, not
just as a result of greater activity but because prices fell and in general nominal
wages remained stable. From 1918 to 1930 wages rose by an incredible 7.5 per
cent a year. This is notable because GDP growth was around 3 per cent and per
capita growth was slightly negative. Partly because the immigration flow had
decreased, workers acquired more negotiating power and were able to improve
their share of income distribution. However, once again, the decrease in prices
was largely responsible for these changes. The cost-of-living index in 1920 was
186 1914 100; in 1930 it had fallen to 133 ± falling on average 3.7 per cent a
year. As had happened in the 1890s, in the 1920s the decrease in prices was the
determinant factor. As a result, during the second half of the 1920s the condi-
tions of Argentine workers improved much more than those of British workers
in relation to the immediate postwar period (Figure 9.3) while the gap vis-aÁ-vis
Italian workers was even wider, as real salaries were reduced by Mussolini's
policies (Figure 9.4).
180
Real wages (Argentina)

160
Real wages (UK)
140
120
Index (1914 = 100)
100
80
60
40
20
0
1885
1887
1889
1891
1893
1895
1897
1899
1901
1903
1905
1907
1909
1911
1913
1915
1917
1919
1921
1923
1925
1927
1929
1931
1933
1935
1937
Figure 9.3 Real Wages in UK and Argentina, 1885±1938
Source: CorteÂs conde: Ei Progreso Argentino.
180
Real wages (Argentina)

160
Real wages (Italy)
140
120
Index (1914 = 100)
100
80
60
40
20
0
1911
1931
1901
1921
1885
1887
1889
1891
1893
1895
1897
1899
1903
1905
1907
1909
1913
1915
1917
1919
1923
1925
1927
1929
1933
1935
1937
Figure 9.4 Real Wages in Argentina and Italy, 1885±1938

Source: Mitchell: Abstract of British Statistics.
280 Argentina 1875±1930
Higher wages could have undermined the competitiveness of industries

already affected by the high costs of transportation and fuel. The industrial
lobby which had as leading figures Bunge, Tornquist and Colombo ± President
of the UIA ± used this reasoning to request protectionist measures (Bunge, 1920).
However, the devaluations of the 1930s again reversed the situation: they under-
mined labour at the same time that they were welcomed by industrialists and
initially by the farming sectors. There is no doubt that the devaluation, with
differential rates of exchange, made the fall in food prices possible and was an
instrument which reduced the international cost of labour in Argentina, a
country whose working population of European origin had migrated with
high expectations and had a developed social conscience. However, after
many decades' experience, it is difficult to come to the conclusion that devalua-
tion was an efficient instrument from the point of view of long-term growth.
The returns from labour can of course vary with the level of education of the
workforce. In Argentina the progress of primary education was very important,
and must have contributed to the increase in human capital. Between 1903 and
1910, the number of pupils in primary schools increased by 7.9 per cent a year;
the pupils represented 8.4 per cent of the total population in 1903 and 11.5 per
cent in 1910. Between 1910 and 1920, the number of schoolchildren increased
by 6.1 per cent a year, representing 12.7 per cent of the population.
We should note that from the end of the war important changes in the
production structure were taking place. One of them was in agriculture,
where the incorporation of cereal crops now occurred at a lower rate. By
contrast, cotton experienced spectacular increases (see Tables 9.1 and 9.2). In
construction, between 1900 and 1913 public works grew 12 per cent a year, but
from 1913 to 1919 investment in public works suffered a drastic fall, 40 per cent
a year (86 per cent in total), later recovering up to 1929 at 21 per cent a year.
However, it must be understood that the previous rates of growth could not
have been maintained, because of the long period of infrastructure obsoles-
cence. Between 1900 and 1913 private construction grew by 9.9 per cent a year,
it fell by 57 per cent in 1914, and between 1913 and 1919 by 25 per cent. From
1919 to 1929 it recovered and increased by 10.7 per cent a year, even more than
during the prewar period. Private construction was financed with domestic
savings, helped by the success of the mortgage bond which absorbed an import-
ant part of local savings. It does not appear that Argentina was facing a declin-
ing trend in investment in construction.
If the level of railway investment diminished, continuing investment
remained a feature of industry such as in automobiles and telephones. The
importance of capital goods for industry had increased in respect to the prewar
decade by 254 per cent.
Between 1900 and 1913 the physical volume of cargo transported by the
railways had increased by 10 per cent; between 1913±17 it fell by 1.4 per cent.
Between 1917±29 it went up again by 7.5 per cent, and even though that
growth was lower than the prewar rate, it was nonetheless important. But the
difference was that the cargo transported increased more than the amount of
railway lines constructed. This allowed a greater utilization of installed facil-
ities. The network had increased by 5.2 per cent annually between 1900 and
1913 but slowed down to 0.5 per cent between 1913±27. Capital invested
amounted to 6.3 per cent and 1.6 per cent respectively. In 1900±14, the cargo
transported was 1.18 tons/km, and in 1925±29 it increased to 1.36 tons/km.
From the final decades of the nineteenth century onwards, textiles and food-
stuffs had been the leading sectors of industrial growth. However, the evidence
on production alone does not reflect the depth of their transformation.
The food industry began with modern factories and high capital intensity.
The textile sector, on the other hand, included thousands of dressmakers ±
the putting-out system, which was counted as manufacturing in the GDP
estimates. The change towards factory activity took place during the 1920s,
simultaneously with the extension of cotton cultivation which provided low-
cost inputs.
Concerning the evolution of industrial production there are data available
from the censuses of 1895, 1914 and 1935. This information presents a variety
of problems, which led us to carry out annual estimates of production and
industrial product.5 Also, the census only takes into account factory activity
whilst the GDP estimates take into account everything which implies trans-
formation of raw materials. According to the census, between 1895 and 1914
the capital stock in industry grew by 9 per cent annually ± our estimates ± while
GDP growth was below 7 per cent. Between 1914 and 1935 industrial produc-
tion increased by 2 per cent a year. Between 1895 and 1914, capital stock in the
industry increased by an annual 5.8 per cent and from 1914 until 1935 by 2.4
per cent. This would contribute to maintaining the picture of a declining trend.
However, some qualifications are in order. In the food sector the slowing-down
process was more pronounced. The 10 per cent yearly growth before the war fell
to 0.3 per cent a year. In textiles the rate fell from 11 per cent in 1895±1914 to
4.4 per cent in 1913±35. However, if we take the peak trend of 1913 and that of
1928, industrial output grew 1.4 per cent a year according to our estimates.
Nonetheless, food increased 3 per cent a year and textiles 3.5 per cent. We
obtain these results because the production of the whole industrial sector was
adjusted by the census data, which includes less dynamic sectors which do not
appear in our sample.
The textile industry suffered an important fall in 1914. However, it recovered
very rapidly because of the availability of domestic inputs, and during the
recovery phase the growth rates were higher. From 1917 to 1927 industry
increased by 3.6 per cent a year. Between 1914 and 1926 food production
grew by 3.3 per cent, whilst textiles grew at an incredible 7.2 per cent a year.
282 Argentina 1875±1930
The problem was not demand but supply, and the recovery was based on a
partial substitution of local production for raw materials which prior to the war
had been imported. This process continued after the war: cloth was imported
whilst cotton was exported. It should be noted that the import substitution
which took place in the textile industry consisted of a substitution of inputs
rather than final products. The postwar growth, as we have seen, was notable
and took place in the factory sector.
Between 1895 and 1914, capital productivity increased ± the K=Y relationship
fell from 1.51 to 0.76 ± a result of the greater use of installed capacity in food,
where at the outset there had been substantial investment in industrial plants,
meat-packing plants, sugar refineries and mills. Capital productivity went
down slightly again by 1935, probably because of investments in important
factories in the textile industry which was witnessing the transition from put-
ting-out to the factory system. As a result of the transformations, the textile
industry became more organized, and production increased from 13 000 pesos
per establishment in 1895 to 80 000 in 1935.
Demand
Between 1910±14 and 1920±24, exports doubled. Nonetheless, the five-year
growth rates 1900 to 1929 show a notable deceleration, as in Table 9.8.
Until 1929, however, foreign demand remained stable, despite a strong
slowdown in world trade in the postwar period which particularly affected
primary products. World cereal demand no longer had the dynamizing effect
of the first decade, and the meat trade began to face protectionist restrictions
(see Figure 9.5).
The biggest difficulties created by foreign demand were felt in the structures
of the cereals and meat trades. There were countless agricultural and cattle
producers, and no control over supply. Between them and international
demand stood grain-trading firms and meat-packing plants, in a position to
reach collusive arrangements and control demand. In periods of expansion
nobody complained of the cost of middlemen, but during periods of decline
Table 9.8 Argentina: exports
Millions of gold Percentage change

pesos (average)
1900±04 147
1910±14 431 118
1915±19 707 64
1920±24 834 18
1925±29 935 12

50 350
% 45 Index
300
40
35 250
30 200
25
20 150
15 100
10
50
5
0 0
1910 1913 1916 1919 1922 1925 1928 1931 1934 1937 1940
Rate of Protection (Tariff × Value of Aforo)

Total Rate of Production (Rate of protection × Nominal Exchange Rate)
Nominal Exchange Rate
Import prices
Figure 9.5 Argentina: the rate of protection 1910±1940

Note: Rate of protection and total rate of protection are measured in percent against left axis; nominal
exchange rate and import prices are measured as indices agains right axis.
they passed on the effects of lower prices to the producers, provoking protests
and requests for state intervention.6
We turn here to some indirect indicators of the variations of internal
demand. We have already shown that population growth slowed down, how-
ever slightly, influenced by the decline in migration. From 1914 to 1917, there
was also a strong fall in employment. In the Federal Capital, from a base of 100
in 1914 employment fell to 85 in 1917, recovering in the 1920s and rising to
139 in 1929 (National Geographic Committee, 1941). Finally, wages tended to
fall in years of depreciation and rose when the exchange rate revalued.
Investment also experienced important fluctuations, for two reasons. First,
Argentina was a new country and needed to create a production infrastructure.
This is done only once, and has a very long period of amortization. Second,
strong disturbances were felt from the international capital market during the
war and postwar periods. Foreign investment ± in which the railways played the
greatest role ± increased until the war, fell during the years of the conflict and
recovered afterwards. However, the sectors it was directed to had changed,
which also meant that it had a lower magnitude. Public and private construc-
tion slowed down. In part, the war recession and the ensuing slow recovery
were to blame, and in addition the construction of the infrastructure of public
buildings and houses had now been completed. Government expenditure
diminished between 1915 and 1917, though afterwards, during the 1920s, it
grew at slightly above 10 per cent of GDP.
284 Argentina 1875±1930
Policy
Trade policy
DõÂaz Alejandro has rebutted the view that following the First World War Argen-
tina failed to adapt and left domestic industries unprotected, facing unstable
foreign markets and a changed resource endowment. He showed that negative
net protection did not occur and that the tariffs had not only sought a fiscal
objective between 1906 and 1940, but had in some way been protectionist (DõÂaz
Alejandro, 1970, p. 294).
Not many explicit economic policy decisions are documented in this
period. However, even though they may not have been visible, this does not
mean that they did not exist. It is also true that policies did not correspond to a
single objective. Most of the time they responded to different interests, situa-
tions and even pressure from sectors which sometimes coincided and some-
times did not. The government's preoccupation was to avoid increases in the
prices of manufactured goods in the war years ± raw materials or finished
products ± affecting consumers and producers. The cost of imported goods
depended on several factors: the international price, the exchange rate and
tariffs. The effective tariff was a combination of the nominal tariff and the
values which the government fixed on imported merchandise for the purpose
of paying the duty ± the official or aforo value.7 In general, each year up-to-date
lists of aforo values were published which might or might not coincide with
market values; however from 1906 to 1920 the aforo values were not revized. As
the prices of imported goods had increased significantly during the war the
divergence between both values increased, resulting in a decline in protection
(see Figure 9.5 and Jorge, 1971). The exchange rate did not compensate
because, in spite of the suspension of convertibility, the peso appreciated dur-
ing the war years. Instead, protection came from the increase of imported prices
and the strong supply restrictions in the economies dedicated to wartime
production.
The government's aim in reducing the aforos and with that the revenue from
the tariff was to try to relieve the effect of the rising price of goods on con-
sumers and producers. While prices increased up to 1919, the peso appreciated
and the aforos fell ± both reducing protection. Between 1919 and 1920 prices
began to go down and the peso depreciated. In 1920 prices continued to fall; in
1921 prices went up in step with a continued depreciation of the currency. The
depreciation increased the combined effect of the tariff and aforos with protec-
tionist effects.
From 1924, the local currency underwent a revaluation as a result of the effort
to return to convertibility at the prewar parity. At the same time the prices of
imported products fell, which meant that the difference between the real values
and the aforo values was reduced, and eventually eliminated, with a consequent
reduction in protection. From 1920 to 1932, on balance, total protection

increased as the combined effect of all the intervening factors.
In effect, the state used the aforos as an instrument to increase or lower the
effective tariff and compensate for changes in the prices of imported goods ± or
movements of the peso. It seems that the government's greatest concern was
the effect of price increases of imported goods on the prices of consumer goods
and raw materials and, second, fiscal income and protectionist pressures. Refer-
ring to the effect of policy in the 1920s, Villanueva (1972, p. 465) argued that
not only was there important investment in industry in the 1920s, a period in
which the sector began to pick up, but also with the increase of custom aforos in
1923 `. . . the protection the industry was hoping for was achieved', although he
added that ìt is likely that the strong growth and stability of the Argentine
economy during those years was more important than tariff protection in the
decision-making processes of USA firms'.
Monetary and exchange rate policy

Adhering to a rigid regime such as that of the gold standard could have had two
negative consequences. The first was that of maintaining the country's costs
high when a depreciation of the currency could have reduced them (`Dutch' or
`Spanish' disease). In effect, in the 1930s, devaluation meant a reduction of
international wage costs; but even though in the short term its effects might be
favourable, the same could not be said in the long term. However, the begin-
ning of the expansion of the textile industry was not due to the depreciation
but to low-cost inputs. But, again, it seems that given the need to attract
immigrants this artifice could not be permanent.
The second consequence was the negative effect of external shocks on the
money supply and credit in a gold-standard regime. From the empirical in-
formation we present it does not appear that there was a situation of this nature
in the 1920s.
There is no doubt that from the 1920s onwards exogenous factors affected
Argentina, mainly in the meat industry which had been the most dynamic
sector and where the decline was most noted. The restriction placed on
Argentine meat imports by the USA in 1925, due to a foot-and-mouth disease
epidemic, and the earlier British policy of colonial preference, combined to
stop the expansion. In the 1930s, the Roca Runciman Agreement was a
policy response to these problems. We cannot justifiably conclude that govern-
ments, faced with diminishing markets, took far-reaching decisions, which
only much later became evident. A further two factors, which in the previous
period had been favourable, were influential in the postwar period: the inter-
national increases in the costs of transportation and fuel. For a country with
expensive labour at the extreme south of the world, it was something else to
grapple with.
286 Argentina 1875±1930
50
Argentina prices
40
UK prices
30
20
Variation %
10
–10
–20
–30
–40
1886
1888
1890
1892
1894
1896
1898
1900
1902
1904
1906
1908
1910
1912
1914
1916
1918
1920
1922
1924
1926
1928
1930
1932
1934
1936
1938
Figure 9.6 Prices in the UK and Argentina
Sources: CorteÂs Conde (1979); Anuario GeograÂfico (1941); and Mitchell (1975).
From 1885 to 1899 Argentina was under a regime of floating exchange ± and
forced circulation. A system of national and provincial issuing banks led to a
disproportionate creation of money and debt and the crisis of 1890. Liquida-
tion was followed by a period of deflation with a strong contraction of spending
and money, during which the exchange rate was stabilized and domestic prices
made to converge with international ones (Figure 9.6).
When in 1899 the monetary reform was implemented with the creation of
the Currency Board, the country formally joined the gold standard. Thereafter,
and until 1913, the convertibility regime was applied successfully and became a
positive factor for the development of the private financial system and of a
domestic debt market, which grew rapidly in the first decade of the century.8 In
this new atmosphere it was perceived that property rights would be respected, a
change from the repeated confiscations of the nineteenth century.
However, it cannot be said that the convertibility system was the only de-
cisive factor in attracting foreign capital; there was also a fall in the exchange
rate. The 1880s had witnessed massive capital inflows, but this capital was
subscribed in European stockmarkets and had special guarantees or was issued
in pounds or gold. This meant it was not affected by the volatility of the
exchange rate since the legal jurisdiction which protected it was not that of
Argentina but that of the country where it was subscribed. The fixed exchange
rate offered stability for direct investments in domestic currency ± pesos. The
interest rates had fallen sharply at the beginning of the century, had then risen
until 1914, but then fell again and hovered around 5 per cent in real terms.
Also, interest rates converged with international rates, the remaining difference
reflecting the country risk factor.
As in the rest of the world, Argentina suspended convertibility at the begin-
ning of the First World War and reestablished it in 1927. It should be pointed
out that whilst the suspension lasted it respected the rules of the gold standard,
issuing only against gold inflows.9 At the beginning the suspension had a
slightly unfavourable effect on the peso, but later, owing to capital inflows
from Europe, not only was the prewar parity maintained but up to 1920 the
peso, appreciated in relation to the pound. The depreciation took place
between 1921 and 1924, and from then on the trend reversed. In 1927 it
returned to the prewar parity, unlike several European countries such as France
and Italy.
In Argentina the money supply and credit under the gold-standard regime
was conditioned by the seasonal gold inflows which took place during the
summer months. Indeed, the so-called Argentine cycle consisted of expansion
of money and credit in the summers and contraction in the winters. This led
to the official bank ± NacioÂn ± partially sterilizing the entrance of foreign
exchange in the summer and expanding credit in the winter, based upon
surplus reserves. To do so it used the bank's own gold reserves which did not
appear in the assets of the Currency Board.
During the war years, in which Argentina suffered international inflation
until 1916, the increase in prices was quite small. From 1913 to 1916 they
increased 4.5 per cent a year, and an increase of 12.7 per cent a year took
place from 1916 to 1929. Thereafter, a strong effort of stabilization took place,
leading to a fall of 4.9 per cent a year.
Public finances
Under the Currency Board regime, the government could not monetize its
deficit. Once Argentina concluded the 1880s debt negotiations it regained
access to foreign credit, but this financing was interrupted during the war
years. In an atmosphere of stability, the domestic market was an important
substitute. Nonetheless, when the government incurred a deficit it used short-
term instruments and created a floating debt which was periodically consolid-
ated. The financing problems stemmed from the reduction of resources in the
war years and the weight of the previous debt. It could be thought that the use
of domestic savings by the state had a crowding-out effect on private invest-
ment; inferred from the relatively low volume of shares traded. However, the
bulk of the debt market was channelled towards construction through mort-
gage bonds. The success of this bond, the enormous population increase in
Argentina's main cities and a regime of joint-stock companies which even
today provides little security to the minority shareholders, explains the lower
preference for private debt.
288 Argentina 1875±1930
In Argentina, the state had by now taken on the functions which were then
considered part of its responsibility, as probably in other new countries also.
Nevertheless, its importance should not be underestimated. It had a major role
in public works, in promoting agriculture and in social policy from 1904
onwards, when the Labour Department was formed and quite advanced social
legislation was implemented.
Until 1930, the tax regime was based on the import tax. Despite its regressive
nature, because it applied to consumption independently of consumers'
income, in Argentina it was widely accepted. This was due to the low cost of
foodstuffs and imported manufactured goods as a result of low tariffs. Because
of that, as long as surplus demand for labour prevailed ± except during the years
of the war recession ± workers were in favour of free trade and monetary
stability, reflected in high real wages. For several decades this maintained an
ample social consensus ± not extended to the political area ± which ended in
the years after the First World War.
The institutional and political framework

The international scenario
Argentina did not participate in the First World War, and therefore did not
suffer the loss of millions of men and goods as did European countries. None-
theless, the war had a negative effect similar to that which befell the western
countries which up until then formed part of a growing globalized Atlantic
economy. The war caused notable political and economic changes, among
which the most important were:
1 very wide fluctuations of prices which produced considerable imbalances

once postwar controls were abandoned. These imbalances were due to the
different sources of finance used by the belligerent countries. During the
conflict, the maintenance of nominal parities, exchange and price controls
concealed this disparity and prevented the market from readjusting, but
such controls could not be maintained. This was probably the most serious
problem of the conflict. Above all, however, it was a new problem, as paper
money had never until then been so widely used. In a metallic system typical
of the nineteenth century, this problem did not exist.
2 The flotation of the exchange rate and the readjustment of parities had a
negative effect on international trade and capital movements, the expansion
of which had been facilitated under a multilateral system of payments in a
fixed exchange regime.
3 The inflationary effect of financing the war was unexpectedly large and
along with the subsequent deflation produced distributional conflicts of a
previously unknown magnitude. These distributive shocks were complicated
by the increasing weight of labour organizations. Until then the rule had
been that the market resolved the issue of payment for goods and services,
but now incomes were negotiated through organizations and interest
groups. The new situation created a problem of collective action.
4 Even though the state in the nineteenth century did not have an insignifi-
cant role ± much greater in the Continent and in the central powers ± the
new conception of war, which required the subordination of all resources to
its objectives, extended its field of influence. Exchange and price controls
were set up, the government intervened in the purchasing of raw materials
and in war production. This produced an institutional change unthinkable
in any other set of circumstances. The civil administration now experienced
the possibility of exercising greater control over society. Even though at the
end of the conflict many of the controls were given up, the lessons learned in
this period would be used repeatedly in the following decades.
5 Along with these circumstances came others which had an effect upon
exporting countries. These included a decrease of the growth rate of trade,
not just due to the restrictive actions of those who sought to secure a regular
supply of foodstuffs and basic raw materials, but also because the prewar
expansion had reached high production levels and in some cases resulted in
inventory accumulation. Also, the countries which were not involved in the
war went through a brief postwar boom which came to an end with a
dramatic crisis of overproduction in 1921, made more serious by policies
in the industrialized countries.
Argentina: the economic policy of the slowdown

Over the years the effects of the war were uneven in Argentina, which had
already suffered a first shock with the Balkan crisis in 1912, producing a run on
reserves. The news of the conflict caused a gold outflow because everyone
wanted liquidity, and because governments wanted to build their war chests.
Following the example of the European countries, including Great Britain, in
August 1914 Argentina decreed the suspension of convertibility, a measure
criticized by the socialist party in Congress which argued that it would produce
a decrease in real wages, which indeed happened. Similarly, the government
decreed an embargo on the shipment of gold. The pressure was short-lived
because reserves increased as a result of positive trade balances following the
drastic reduction in imports, and the fact that savers in countries in conflict
sent their funds to countries not affected by the war.
The British Royal Navy imposed a blockade upon the central powers, inter-
rupting Argentina's trade with an important supplier of industrial goods, Ger-
many. Meanwhile, the allies moved to wartime production provoking a supply
shock in regard to industrial inputs and coal. Argentine industry had come to
depend upon these inputs, and there was a fiscal impact as lower imports
290 Argentina 1875±1930
translated into lower collection. Moreover, already in 1912 foreign investment

had been interrupted. These circumstances were aggravated by unfavourable
weather conditions: the start of the war coincided with the failure of two
harvests, which began a long and deep recession. However, exports were not
undermined during the following years even though they were affected by the
shortage of storage and the increased cost of transportation. The controls com-
mon in the belligerent countries were not applied in Argentina, nor did the
government intervene so extensively in the economy. Nevertheless, the experi-
ence of these years and those of the postwar period caused profound transforma-
tions in society and the economy. Writing in the 1930s Vernon Phelps argued:
The outbreak of the World War in August 1914 found Argentina particularly
unprepared to withstand the impact of the economic forces which that
catastrophe loosed. Especially susceptible to international economic influ-
ences, because of its close and financial relations with Europe, Argentina was
bound to be severely affected by the outbreak of such a struggle, even under
a normal domestic situation. But in 1914 Argentina was already in the
depths of a depression . . . The immediate cause of this reaction was traceable
to the Balkan War, which tended to check the flow of European capital into
Argentina that had characterized the preceding decade.
(Phelps, 1938, pp. 22±3)
The postwar problem consisted of creating a new institutional framework, as

well as adjustments in the economy. Some of these changes began in the years
of the conflict or in the immediate postwar period. The change foreshadowed
those that would characterize the period between the 1930 crisis and the end of
the Second World War and which would lead to a very different Argentina. It
was the need for these institutional changes, in a country which had known
how to adapt itself to the internationalized world of free trade, which caused
the majority of the conflicts and the greatest costs ± as it did in Europe.
The need for these changes came from the suspension of convertibility, the
administered float which followed and the deflationary measures aimed at
returning to the prewar parity in 1927. The war inflation as well as the sub-
sequent deflation caused distributive shocks. During the war years real wages
fell, but by contrast in the deflationary phase, when the peso was revalued in
real terms, real wages achieved important gains. It is true that Argentina at the
end of the 1880s and 1890s had gone through a similar experience: when the
peso depreciated real wages fell, and when the peso appreciated real wages rose.
But at that time the country had fewer factories and incipient union organiza-
tions. At the end of the war there was a large number of workers and the labour
movement not only had an organizational maturity unequalled in Latin Amer-
ica, but workers who ± because of their recent European origins ± were familiar
with the tumultuous, if not revolutionary, atmosphere being experienced in

the Old World.
The public began to realise that these wide fluctuations of income were not
caused by natural factors. The government could affect working conditions and
consequently the life of different sectors of society through the decisions it took
concerning the exchange rate or the currency. It was not enough to be more
efficient or productive, it was necessary to put pressure on the government to
obtain greater incomes. People began to see, without fully realizing it, that
rather than nominal values what was really important was real income and
that the government had a decisive role in this. Thus began the long period of
conflicts and demonstrations which characterized Argentina during the second
half of the decade. A transition from a framework of liberal institutions to
another of corporatist growth was in motion.
There were also other circumstances which required institutional changes. As
long as the country participated in expanding world trade nobody thought
about the need for trade policies. The slowdown of world trade, but above all
the restrictive practises of industrialized countries, were an incentive to seek
preferential treatment. The preferences sought included a policy of `buy from
those who buy from us' and the search for protection through a variety of
instruments, above all the value of the aforo or depreciation. At the beginning
of the century the cattle producers were successful in adapting to the British
demand for meat; after the War, caught out by less competitive markets, they
sought refuge in the government's defence of the foreign placement of their
products.
It is difficult to imagine the alternative route of a return to prewar parity. Is an
alternative imaginable for a moderately stable country, closely linked to Great
Britain, and still avoiding the hyperinflation of Germany or the pronounced
depreciation of France? Could other trade policies have been thought of? Or is
it that, confronted with a situation in which trade became scarce, the `buy from
those who buy from us' policy and the agreements planned by the D'Aberon
Mission could not have foreseen the bilateralism of the 1930s?
During the period of acceleration of growth, 1880±1912, the rapid expansion
created a consensus that everyone was benefiting, even if it was acknowledged
that distribution was uneven. The agricultural exporters were the main bene-
ficiaries, but labour too enjoyed sustained demand and higher real wages.
Industrial capitalists benefited from a flexible labour supply and cheap raw
materials in a sector which at that time used natural resources intensively.
This changed during the postwar period. Affected by the new conditions in
the meat trade, livestock farmers called for state protection. In this period, as we
have seen, the relative proportions of factors of production changed. With the
expansion of the textile industry, production became more labour-intensive. In
response, entrepreneurs demanded a reduction in labour costs and limits to
292 Argentina 1875±1930
advances in social legislation. Interest groups which sought a better share for
their sector, rather than production increases or general welfare, began to
appear, and the earlier consensus ended. The institutions for the new situation
were not designed until the 1930s and 1940s, but the problem they were
responding to arose during the first postwar period.
To sum up, the country went through strong income fluctuations because of
the depreciation of the peso immediately after the war and its later appreciation
until the return to prewar parity. This led to distributive conflicts which, while
they did not have the magnitude of those of Germany or France, were greater
than that of any other Latin American country. The resolution of these new
conflicts required agreements concerning new institutional rules which were
very difficult to reach. There was little experience and the game itself was to be
resolved in the negotiations. Thus began a long period of uncertainty which
had an effect on investment and, consequently, growth.
Argentina went through important political changes which affected the
governability of the country and, as a consequence, the investment rate and
growth. As a result of the long pressure to widen political participation, in 1910
the electoral system was reformed establishing secret and obligatory universal
suffrage. Limited participation was extended in a rapidly changing society.
With the implementation of the new regime, the opposition party took over
government ending a long period of stability and hegemony of a coalition of
conservative parties which had governed the country since 1880. The general-
ized political apathy of the native population and the fact that an enormous
number of immigrants (a third of the population in 1914) showed no signs of
wanting to become naturalized citizens to obtain political rights ± because to do
so meant having to do military service ± resulted in very low transition costs for
governments whose electoral procedures were constantly under public scrutiny
but which had, nevertheless, an ample consensus on fiscal and economic
aspects.
As participation was extended, the postwar period introduced economic
problems to the realm of policy, thus emphasizing differences and aggravating
conflicts. Argentina initiated an experiment which was soon criticized by those
who left the government, and which in an atmosphere of mutual intolerance in
1930 led to the first military coup of the twentieth century. By 1930 Argentina
had acquired the conflictive elements which would characterize it for the next
half century.
Conclusions
The information presented throughout this chapter gives us a different and

more complex view of the circumstances which led to the interruption of the
strong expansion which characterized the period before the First World War.
It has commonly been claimed that the end of the frontier was the cause
of the slowdown in growth, with emphasis on the factors which influenced
the supply side. It is true that the rate at which areas were sown with
cereals declined, but this explanation is insufficient: the development of
other crops and the changes in livestock-raising ± which led to the growth in
the first decade ± did not depend upon the increase in the amount of land
exploited but on its different use. Indeed, in the case of meat the problem was
one of demand rather than supply. There was no reason, as was seen in the
immediate postwar period, for the cattle stock and production to fail to
increase.
As far as investment is concerned, it is true that the data show a slowdown in
the rate of capital accumulation, but there are two caveats, one concerning the
statistical effect caused by starting from very low levels, as happened at the
beginning of the century; and a second concerning the magnitude and indivis-
ibility of the investments in infrastructure in a new country with a long
maturing period. Some exceptional circumstances should also be taken into
account, including the overexpansion and overinvestment in the first decade;
the financial deepening which followed the move to convertibility in 1899
during which there was an exceptional increase in savings; and the strong
investment in the railway network in the 1900s.
Another line of explanation of the slowdown comes from the institutional
factors which influenced factor markets once there was a change in the avail-
ability of resources. The industrial sector witnessed a modification in the
intensity of factor use, ranging from more intensive land use in the production
of foodstuffs to more intensive labour use in textiles. However, despite popula-
tion increases as a consequence of immigration, the labour supply was never as
flexible as during the prewar period. This was seen in the different behaviour of
wages in the two periods. In this, two circumstances played a part. One was the
greater organizational power of workers, and the other the halt in immigration
flows during the war years.
In the prewar period, in a context of monetary stability and zero inflation,
the relative factor prices stemmed from the combined action of supply and
demand and from the organizational ability of the workers or entrepreneurs
when negotiating them. By contrast, the war witnessed a phenomenon that,
although not totally new, became decisive. The workers' real incomes were now
affected by governmental regulations on currency, exchange rates and tariffs
more than by negotiations over their nominal values. As a result, it became
necessary to influence the government in order to improve a sector's position.
Thus began a long period of distributive conflicts, which continued into the
following decades. This coincided with significant changes in the political
regime: popular participation was increased with the passing of the law of
universal suffrage, secret and obligatory.10
294 Argentina 1875±1930
With respect to demand-side changes, whilst the initial fall of GDP was
caused by a supply shock ± the decrease in imports ± the weakness of the
following recovery was due to the changing conditions of world demand, a
consequence of protectionist policies in the industrialized countries which
affected the prices of cereals and the meat trade. The displacement of Germany
during the war years and the years that followed until the stabilization of 1924
closed a market which had become important for Argentina during the prewar
period.
How did the country respond to the new demand conditions? Were adequate
policies implemented? The alternative policy and the one most debated in the
literature would have involved abandoning the gold standard instead of trying
to deflate and return to parity, either allowing the exchange rate flexibility or
adopting, like France, an undervalued exchange rate (Temin, 1990). It would
seem that because of the previous success, upon receiving an enormous flow of
capital it was reasonable to focus on the advantages ± from the point of view of
credibility ± of maintaining the country within the rules of the gold standard
and returning to the prewar parity.
In a period of decline of prices, however, the pressures on policy had to do
with the patterns of internal trade. While there were countless agricultural
producers, the commercialization of grains and meats were in the hands of
few firms and, in the phases of falling prices, they were able to pass on the losses
to the producers. Because of this producers demanded state intervention in
commerce, pressing for policies which in the following decades favoured the
internal market (Smith, 1969).
Turning to trade policies, first of all it must be said that, unlike Brazil,
Argentina's market share was not large enough to be able to pass cost increases
on to foreign consumers. Also, in cattle production there was no possibility of
accumulating stocks over long periods. As far as imports are concerned we have
seen that there was de facto protection during the war years when prices of
imported products and the cost of transport increased. The policies tended to
be more protectionist when in the postwar period prices went down. Finally,
Argentina's range of instruments included the nominal tariff, the aforo and the
exchange rate. These policies had as their principal aim the avoidance of price
increases in raw materials and basic products.
In this period external circumstances were largely responsible for slow
growth. They included a supply shock due to the fall in imports which pro-
duced the first fall between 1914 and 1917 and later on the changing condi-
tions in the cereal and meat markets which no longer stimulated growth as
before. In response, a new option emerged: the expansion of intensive crops
such as cotton which provided a cheap raw material for the rising textile
industry. It can be said that the successful initial adjustments of the 1930s to
the conditions of world markets were in some way prepared in the initial
postwar years. In this sense, it would seem that there was a certain unity in the
interwar period (1914±45). What also seems clear is that the institutional
changes which caused these adjustments were not positive for the subsequent
evolution of the country (DõÂaz Alejandro, 1970). While Argentina seemed at the
end of the 1940s to have successfully isolated itself from external negative
shocks, several decades later it seems that it was isolated from positive shocks,
if we take into account the way in which world trends moved between 1950
and 1970 (Maddison, 1991). Australia, which also suffered the same external
negative shocks between the 1920s and 1940s, experienced notable growth
from 1948 to 1970, in response to world economic expansion, of a very differ-
ent nature from that of Argentina.
Appendix: estimates of the capital stock in Argentina
The most serious works on investment and capital formation are those of Ganz
and Balboa and Fracchia, which served as a base for the ECLA study on Argen-
tina (Goldsmith and Saunders, 1959). Ganz's estimations were criticized by
DõÂaz Alejandro, because they used 1950 prices. DõÂaz argued that the domestic
price estimates were exaggerated; he converted the estimates into 1935 prices,
closer to the price structure of the first two decades, recognizing that none-
theless they might still be overestimated. Whilst for Ganz (prices of 1950)
investment represented 39 per cent of GDP in 1900±09, 42 per cent in 1910±
19 and 13 per cent in 1915±19 for the same period, DõÂaz's estimates were 17.2
per cent, 18.7 per cent and 6 per cent (DõÂaz Alejandro, 1970). To Ganz, in 1920±
24 the fixed investment represented 26 per cent of GDP and to DõÂaz only 12 per
cent, while the estimates for 1925 also differed, 33 per cent and 15 per cent.
More recently Robert Gallman carried out another revision, this time based
upon 1914 prices, obtaining a result in which investment represented 16.9 per
cent of GDP in 1900±09 and 14.6 per cent in 1910±14 (Davis, and Gallman,
1996). Using Bunge's date (1916) he estimated a capital stock in 1916 (at prices
of 1914) of 12.5 billion pesos while Ganz estimated it at 23.5 billion (at prices of
1914).
ECLA's work on investment (1959) is based on 1950 prices and does not
indicate that deflation has been carried out. Before 1935 there were no surveys
concerning wholesale prices; those published referred to the cost of living. The
Anuario GeograÂfico of 1941 quotes as a source the National Statistics Board for a
series of wholesale prices it publishes, which differ substantially from consumer
prices which, with all its defects, is the only systematic series.
The method used by ECLA is based on the annual flow of domestic produc-
tion and on capital-goods imports. With many difficulties the import data can
be re-done, but there are no annual series of capital-goods production. Our
296 Argentina 1875±1930
production estimate comes from a sample which is quite representative of the

main trends of production, fundamentally the sectors of foods and textiles, but
does not cover capital-goods production. Because of this we prefer to follow the
variations of capital stock in the studies which appear (with varying reliability)
in the 1895 and 1914 censuses.
The most complete study ± with all its drawbacks ± is the 1914 census (which
Bunge also used) which has the advantage that it uses 1914 prices, which are
those used to reach the values for the production of raw materials in our GDP
estimates.
Using criteria similar to those of Gallman, we estimate the capital stock of
1914 as lower than the estimate of Ganz, although higher than that of Gallman.
The capital stock in 1914 would have been around 18.9 billion pesos of 1914. As
we said, 1914 was a year of strong contradictions: in this year the relation K=Y
was 4/1, distorted by the fall of GDP. If the stock is related to the previous year's
(1913) GDP, the relation would be 3.2 and the proportion of the investment of
GDP would be 16 per cent ± the difference with Gallman lies fundamentally in
the price deflator.
There is also a study published in the census of 1895 which allows us to
see the change between both dates; however, we have serious reservations
because the criteria of both are not totally homogenous and the coverage is
not the same. We believe that it is likely that the stock is overestimated in 1895
(cattle and capital in commerce), but underestimated in 1914 because the
coverage was less ample. Excluding certain items, and trying to homogenize,
we reach an estimate for 1895 (at 1914 prices) of 7 billion pesos, which means
that the annual increase between 1895 and 1914 would be 5.3 per cent. These
estimates exclude the inventories included by Bunge.
Notes
1 Similarly, the displacement of the USA by Argentina in the British meat market, rather
than an important growth in demand, was responsible for expansion in the case of
cereals (CorteÂs Conde, 1979).
2 Davis and Gallman (1996). For further details, see Appendix.
3 Derived from information concerning foreign capital from diverse sources. It confirms
that external savings, until the First World War, compensated for the lack of domestic
saving in the magnitude required by the exploitation of virgin land.
4 For the different estimates of the capital stock, see Appendix.
5 The 1895 census has data on the workforce, industrial establishments and capital, but
not on production. The 1914 census is the most complete; however, that was a year of
deep recession. In 1935, again, the 1930s depression was in full swing. Both these years
are too atypical to be used as markers for establishing trends.
6 Gravil (1985) and Smith (1969). One of the characteristics of the Argentine export
business was precisely that on the one hand the supply was in the hands of many
producers. On the other, demand was controlled by the trading houses: `the four big
ones', Bunge, Born, Dreyfus and Weil. On several occasions, in connivance with the
railways and thanks to the impossibility of storing the products and the segmented
nature of the credit market, the trading firms made the conditions for producers less
than competitive. In the meat market, at the beginning the price war between British
and USA meat-packing plants favoured the producers. Later, agreements concerning
participation and prices were reached to the detriment of producers. Strong reactions
from the producers ensued, including growing anti-foreign feelings which in the
1920s resulted in demands for state intervention (Smith, 1969).
7 It is assumed that the official value, aforo, should correspond to the market even
though it was not always so because it was easier to correct the tariff by modifying the
aforo values. Thus, if the tariff was 30 per cent and the valuation coincided with the
market's ± called the `real' value ± for example of 100 pesos, 30 pesos were paid in
taxes, but if the market value was 200 pesos and the aforo continued to be 100 pesos,
the 30 pesos which were paid implied an effective rate of only 15 per cent.
8 Until 1890 the issuing, deposit and discount banks had been mainly state-owned.
The private commercial banks (discount) had fewer volumes of deposits.
9 Because of the war, emergency measures came into effect, and even though they were
not used they were useful during the 1930s to make the currency board's regime more
flexible. It consisted of the authorization to issue, by the currency board, based on the
re-discount.
10 The workers' position would change with the devaluations of the 1930s and improve
with Peronism.
References
Alesina, A. (1988) `The End of Large Public Debts' in F. Giavazzi and L. Spaventa (eds),
High Public Debt: The Italian Experience, New York: Cambridge University Press.
Balboa, M. and Fracchia, A. (1959) `Fixed Reproducible Capital in Argentina', in R. Gold-
smith and C. Saunders (eds), The Measurement of National Wealth, Chicago: Quadrangle
Books.
Banco de la NacioÂn Argentina (1923) Report and Balance Sheet.
Bunge, A. (1928) `Los capitales extranjeros en la RepuÂblica Argentina' in Revista de Eco-
nomõÂa Argentina, no. 131.
ÐÐ (1920) Los problemas econoÂmicos del Presente, Buenos Aires. ed. A.E. Bunge.
ÐÐ (1916) Riqueza y Renta en la Argentina, Buenos Aires: Agencia General de Librerias y
Publicaciones.
ComiteÂ Nacional de GeografõÂa (1941) Anuario GeograÂfico Argentino, Buenos Aires: ComiteÂ
Nacional de GeografõÂa.
CorteÂs Conde, R. (1997) La economõÂa Argentina en el largo plazo: 1875±1975, Buenos Aires:
Sudamericana.
ÐÐ (1994) Èstimaciones del PBI en Argentina 1875±1935', in Documentos de Trabajo,
Buenos Aires: Departamento de EconomõÂa, Universidad de San AndreÂs.
ÐÐ (1979) El Progreso Argentino, Buenos Aires: Editorial Sudamericana.
Davis, L. and Gallman, R. (1996) Àrgentine Savings and Investments, Late XIX Century
and early 20th Century', mimeo.
DõÂaz A.C. (1970) Essays on the Economic History of the Argentine Republic, New Haven/
London: Yale University Press.
298 Argentina 1875±1930
Di Tella, G. (1970) `Criterios para una polõÂtica de desarrollo industrial', in M. Brodersohn

(ed.), Estrategias de industrializacioÂn para la Argentina, Buenos Aires: Editorial del Insti-
tuto.
Di Tella, G., Zymelman, M. and Petrecolla, A. (1967) Las etapas del desarrollo econoÂmico
argentino, Buenos Aires: EUDEBA.
Dyster, B. and Meredith, D. (1990) Australia in the International Economy, in the Twentieth
Century, Cambridge: Cambridge University Press.
ECLA, United Nations Economic Commission for Latin America (1959) El Desarrollo
econoÂmico de la Argentina, 3 vols (mimeographed Statistical Appendix), Mexico.
Eichengreen, B. (1995) Golden Fetters: The Gold Standard and the Great Depression, 1919±
1939, New York: Oxford University Press.
Ganz, A. (1959) `Problems and Uses of National Wealth Estimates in Latin America', in R.
Goldsmith and C. Saunders (eds), The Measurement of National Wealth, Chicago: Quad-
rangle Books.
Gravil, R. (1985) `The Anglo-Argentine Connection, 1900±1939', in Dellplain Latin Amer-
ican Studies, no. 16, Boulder: Westview Press.
Jorge, E.F. (1971) Industria y ConcentracioÂn econoÂmicas Aires, Buenos Aires: Siglo XX Argen-
tina Editores S.A.
Latzina, F. (1889) L'Agriculture et l'eÂlevage dans la ReÂpublique Argentine, Paris: P. Mouillot.
Llach, J.J. (1985) La Argentina que no fue: Fragilidades de la Argentina agroexportadora (1918±
1930), Vol. I, Buenos Aires: Ediciones Ides.
Maddison, A. (1991) Dynamic Forces in Capitalist Development, Oxford/New York: Oxford
University Press.
MartõÂnez, A. and Lewandowski, M. (1912) L'Argentine au XXe siecle, ParõÂs: Colin.
Mitchell, R.H. (1975) European Historical Statistics 1750±1970, New York: Columbia Uni-
versity Press.
Naciones Unidas, ComisioÂn EconoÂmica para AmeÂrica Latina (1958) AnaÂlisis y Proyecciones
del Desarrollo EconoÂmico, El Desarrollo EconoÂmico de la Argentina, Santiago de Chile.
O'Connell, A. (1984) `La Argentina en la DepresioÂn: los problemas de una economõÂa
abierta', in Desarrollo EconoÂmico, vol. 23, no. 92, Jan.±Mar.
Phelps, V.L. (1938) The International Economic Position of Argentina, Philadelphia: Univer-
sity of Pennsylvania Press.
Smith, P.H. (1969) Politics and Beef in Argentina: Patterns of Conflict and Change, New York,
Columbia University Press.
Solberg, C.E. (1987) The Prairies and the Pampas: Agrarian Policy in Canada and Argentine,
1880±1930, Stanford: Stanford University Press.
Taylor, A. (1992) Argentina and Economic Growth in Comparative Perspectives, Cambridge
Mass.: Harvard University Press.
Temin, P. (1990) Lessons from the Great Depression: The Lionel Robbins Lectures for 1989,
Cambrige Mass.: MIT Press
Villanueva, J. (1972) Èl origen de la industrializacioÂn Argentina', in Desarrollo EconoÂmico,
Revista de Ciencias Sociales, vol. 12, no. 47, Oct.±Dec.
10
Alteration, Crisis and Adjustment in
the Cuban Export Economy,
1898±1939*
Antonio SantamarõÂa GarcõÂa
`Without industry there is no nation.

Without sugar there is no country.'
(Cuban saying)
Introduction
This chapter focuses on the analysis of the period we have called the years of
àlteration, crisis and adjustment' in the Cuban economy. Paradoxically, the
interwar period is one of the least-studied of the island's history and perhaps
the most decisive to understand the country's later evolution. Cuba was one of
the countries most affected by the 1930s crisis because of its dependency on the
production and export of sugar, mainly to the USA. Nevertheless, this depend-
ency was maintained long after the depression had passed. Authors such as
Maddison (1988), DõÂaz-Alejandro (1989) and Zanetti (1983) point out that the
island did not undergo a process of diversification as occurred elsewhere.
The link with the US market since independence in 1898, and reinforced during
the Great War, was, they argued, largely to blame. We take a different line. We
will try to prove that the adjustment of the sugar sector to post-First World War
market changes caused a structural crisis in the economic and sociopolitical
system of the island. At the same time, it provided the means to confront this
crisis, a thesis which involves approaching the problem from the supply side
and from a comparative perspective.
* This work was financed by the DGICYT and DGES programmes nos. PD 94-0373 and PB 96-0868,
and by a postdoctoral grant at St Antony's College, Oxford, from the Ministerio de EducacioÂn y Cultura
(Spain). I am grateful for the comments and suggestions made by my thesis director, C.D. Malamud,
by A. GarcõÂa Alvarez, J.A. Ocampo, R. Thorp and M. Lillo.
299
300 The Cuban Export Economy, 1898±1939
Our thesis offers elements to explain the peculiarity of the Cuban economy
and its comparative evolution. Before the interwar period, it presented a level of
development (per capita GDP, alphabetization or infant mortality rates) similar
to that of the most diversified economies in Latin America and, at the same
time, degrees of dependency on foreign trade, on a single product (sugar) and a
single buyer (USA) which only could be compared with the less-developed
countries in the region. This distinctiveness was maintained after the 1930s
crisis (see Bulmer-Thomas, 1994, pp. 59 and 441), in our opinion because it was
the best adjustment option to the changes in the world economic system.
The Great War resulted in a big increase in Cuban sugar production. This
increase continued in the 1920s when world supply recovered. We think the
increase was a direct consequence of the way in which the sugar industry
responded to the reduction in the old continent production during the war,
during what we are calling the period of alteration of the island's export
economy. The market was rapidly saturated and prices began to fall prior to
the 1930s crisis; consequently, the crisis occurred at a time when the Cuban
economy was very dependent on the external sector. However, this was the
necessary but not sufficient condition for the absence of structural change after
the depression. The sufficient condition was the sugar industry's adjustment,
allowing Cuba to reenter the world market as an exporting economy in a better
situation than that of other competitors, to regain a growth rate similar to that
of the most developed countries in Latin America and to restore the internal
sociopolitical order, which had been destabilized by conflicts since the begin-
ning of the 1920s, culminating in a revolutionary uprising in 1933.
Growth: from independence to the First World War, 1898±1913
No one would question the existence of a direct link between Cuba's independ-
ence and the increase in the supply of sugar, or that access to the US market was
an underlying issue. Although it is difficult to place precise dates on the econ-
omy's evolution, it is fair to say that by the end of the 1830s, when the first
railway was inaugurated, the island had become specialized in exporting sugar.1
In 1847, 50 per cent of Cuban sugar was sold to North America and after 1877
very rarely less than 80 per cent. US demand was concentrated on raw sugar to
feed its own refineries. From the 1860s onwards, island ingenios concentrated on
this product. In SantamarõÂa (1997b) we demonstrated that the commercial
relationship with the USA was accompanied by a strong price convergence
that was altered by the crisis of 1882±83 which depressed the price of sugar
and reduced its buying power by 69 per cent in the next two decades. We also
showed that Spanish policy was consistent with these circumstances. Indeed,
two agreements were signed to facilitate Cuban±US trade. However, the pro-
blem remained unsolved, and in the middle of the 1890s a very high zafra
Antonio SantamarõÂa GarcõÂa 301
coincided with renewed deflation and an increase in North American tariff

duties, especially after the last of the agreements was concluded in 1894.2
In the nineteenth century, and especially in its second half, the moderniza-
tion of the Cuban sugar industry resulted in its total mechanization. This was
relatively slow and was carried out in order to maintain competitiveness in the
face of the increase in world supply, while optimizing the allocation of
resources in response to external dmand. Oversupply, caused by the develop-
ment and protection of European sugar beet sector, which in the final decades
of the last century was produced in greater volumes than sugarcane, resulted in
falling prices. This led to the gradual concentration of Cuban exports to the
USA, due to the closure of former European continental markets, and the
consequent specialization in raw sugar production. The allocation of resources
also determined certain distinctive features in the island ingenios in comparison
to those of other producers. Until the last third of the nineteenth century, most
sugar factories in the West Indies employed slave labour. When international
pressure to abolish the slave trade, and finally slavery itself, prevented the
continued use of this form of exploitation, the solution adopted in Cuba was
to initiate a process of vertical decentralization which did not occur in other
places. Indeed, labour was the scarcest factor of production on the island and
alternatives such as the importation of temporary workers from YucataÂn or
China had not been overly successful.3 There was a break with the traditional
agro-industrial integration of the ingenio and cultivation was left in the hands
of colonos (more or less independent cane growers), who rapidly satisfied most
of the demand for raw material in the industry.4 In contrast, the relative
abundance of cane, owing to the ideal adaptation of this plant to Cuban soil,
resulted in the introduction of high-yielding large-scale production technol-
ogy, saving work and energy, and causing a parallel process of horizontal
concentration known as centralizacion.5
Sugar manufacture requires continuous-process technologies to achieve
economies of scale, and reducing costs is essentially a question of organization:
of an optima coordination between the various parts of the productive chain,
both agrarian and industrial, to ensure that there is a supply of cane in suffi-
cient quantities and of the highest quality, and an uninterrupted flow of inputs
within the mill. The colono system at first caused difficulties with this coordina-
tion, since it involved giving up, or at least sharing, control of the raw material
with the farmer. This was solved by allocating to the centrales vast tracts of lands
and private rail networks which allowed them a monopsony situation vis-aÂ-vis
the colonos. Finally, Dye (1998) has shown that the coordination of this com-
plex organizational system had adjustment costs which translated into the
underutilization of the optimum capacity of the mills for several years after
their construction, and in SantamarõÂa (1995b) we proved the same occurred
after a remodelling to amplify the productive apparatus of a factory.6
We have discussed the workings of the process of sugar production at length

because it is essential to understanding the supply-side story of the periods of
alteration, crisis and adjustment in the Cuban exporting economy. Further, this
discussion allows us to understand the reason why after 1890, with the aboli-
tion of slavery (1886), when many ingenios were transformed into centrales and
colonos grew most of the cane, production increased slightly and was more than
30 per cent higher than the averages for the 1880s.7 During a deflationary
period, and in a context of cost reduction through economies of scale, the
existence of the USA as a ready market, and the possibility of establishing
trade agreements which would lead to reductions in that country's tariffs, the
benefits of independence were obvious. In 1902, following North American
intervention in the Spanish±Cuban War, the Republic of Cuba came into being.
One of its first sovereign acts was the signing of the Reciprocal Trade Treaty,
which reduced the tariff on island sugar exported to the USA by 20 per cent in
exchange for a reduction in its import duties by between 20 and 40 per cent.
This treaty, along with the Platt Amendment which allowed armed US inter-
vention in the country, constituted the pillars on which the economic and
sociopolitical system rested in Cuba.8
After 1898 a new legal framework was also established to facilitate the growth
of sugar supply. Because land and rail were the key elements in the organization
of the productive system, it is not a coincidence that the US Government of
Occupation's most important laws should be to regulate the demarcation and
division of communal lands and to liberalize the laying of railtracks ± to open a
railway all that was needed was to register with the National Commission of
Railways. The new institutional framework provided the incentives for a territ-
orial widening of the industry. In the nineteenth century, ingenios were con-
centrated in the Western half of Cuba, whereas the Eastern half was isolated,
underpopulated and underdeveloped, and the Spanish government failed in its
efforts to lay a railtrack which would cross the length of the island and over-
come the problem of isolation. An economic reason was probably needed; that
reason was sugar, but in order to open new areas for cultivation a market was
necessary and this did not happen until independence. After 1898, the track
was built in only two years ± 1900±1902 ± and modern centrales were founded in
the East where virgin lands meant that in a short time the production of sugar
there was greater than in the West.9
With a rail line which crossed the length of the Western half of Cuba, with
branches extending to the ports in the North and South, together with the large
North American investments on the island before the beginning of the First
World War, mills stayed almost entirely in the hands of local capital.10 Apart
from this significant investment, another factor which was essential in the
territorial integration of the country and in the commercial exploitation of
the cane resources in the Eastern provinces was the great contribution to Cuban
economic growth made by both independence and the establishment of special

trade relations between Havana and Washington. Without taking into consid-
eration the opportunities for increasing sugar production given the availability
of the US market and the effect that the reduction in their tariff duty had on the
price, with the new tariff alone the island earned around US$115 million more
for its zafra between 1903 and 1913 than would have been the case without the
reduction. This sum amounted to 3.6 per cent of the total value of nominal
income in that same period of time. Real income per capita, in turn, increased
at nearly 4 per cent a year.11 The debate over whether the Reciprocal Trade
Treaty caused or increased the structural deformity of the Cuban economy has
until now been conducted in imprecise terms and has generally been ideolo-
gically prejudiced. Until we have studies which take the question to a more
scientific level we will limit ourselves to highlighting some obvious issues. First,
it is clear that the treaty consolidated the pre-existing mono-producer and
mono-exporter structure. However, this was not so much the cause of indepen-
dence but its economic reason. Secondly, the agreement jeopardized opportu-
nities for diversification. But at the beginning of the century economic
specialization was probably the best option for the development of countries
such as Cuba and, surely, the only way to attract necessary investment. Finally,
we have no basis for arguing that the lack of specialization would have led to
more balanced growth. On the contrary, studies show that the diversification of
Latin American economies was the result of the multiplier effects of the exter-
nal sector during the period of upward movement in the cycle of primary-
exporting production, and also of the response to the crisis in that cycle in
each case.
Alteration (1914±25): from the First World War to the end of the
sugar boom
With an expanding market and all of its productive capacity now free, Cuban
sugar supply grew by an annual rate of 18 per cent between 1900 and 1913,
rising out of step with North American demand (see Figure 10.1). The alter-
native in a situation such as this would have been to reduce the production or
to look for new customers; however, the First World War eliminated the prob-
lem and replaced it with an urgent need to increase export due to the contrac-
tion of European supply. The island became the Allies' supplier of sugar,
especially after the USA joined the war in 1917 and began to regulate trade.
The way in which the sector responded to the new incentive of demand was
similar to that observed in previous stages: by building 35 factories and, pri-
marily, by increasing output per mill from 147 to 205 tons between 1913 and
1919. This was achieved by acquiring more land and laying more railways ± in
1913 there were 170 000 hectares and 4500 km. respectively; by 1919, 270 000
7000
Cuban output
Exports to US
6000 Price
Tons and cts US$/lib × 1000
5000
(thousands)
4000
3000
2000
1000
0
1850 70 90 30
Figure 10.1 Cuban sugar: output, exports to the USA and price of raw sugar, 1850±1939a
a
The price corresponding to 1920 ± 11.95 cts/lb ± is not shown due to its extreme value.
Sources: SACT (1903±36) and CEF (1937±40).
and 11 200. Finally, the endemic problem of labour shortage at harvest time was
solved by importing the workforce from other West Indian islands.12
The speed of sugar expansion during the First World War altered three basic
elements of the Cuban economic and export system, as well as its productive
organization. First, the growth in supply became totally detached from US
demand; second, the process of renewal of capital (vintage-capital) was inter-
rupted.13 Third, the sector had remained in the hands of local owners and the
relative slowness with which it was modernized allowed it to be self-financing,
but the urgent need to produce more sugar during the war resulted in a rapid
concentration and modernization of the structure of ownership in order to
facilitate entry of the necessary investment. Control moved away from industry
towards financial capital, especially that of US origin.14 The alteration of these
elements continued on an even greater scale during the postwar period when
the market situation varied radically. In 1919 Washington stopped regulating
sugar trade, and the price increased from 5.96 to 11.95 cts/lb average between
1919 and 1920. Inflation, however, was caused by a mistaken perception of a
supply shortage. Inevitably, a sudden deflation followed, lowering the price to
3.1 cts/lb in 1921, with serious consequences for the sugar sector and the Cuban
economy. When prices rose, many producers obtained loans on futures pawned
at an average of 10 cts/lb. They were unable to pay them back when they had to
sell at 3.1 cts, and they lost their properties. Generally, these producers had only
started to invest in the business in the war years or were simple speculators.
Banks and financial groups who acquired their industrial and/or agricultural
companies by means of mortgage auctions also supported many enterprises

experiencing difficulties with new injections of capital. Amongst these institu-
tions the National City Bank of New York stood out. The banks and financial
groups' intervention was motivated by an attempt to rescue the loans and
investments they had made during the conflict, although they were mistaken
in their perception of the market trend.15
The intervention of financial capital avoided the bankruptcy of many mills
which would otherwise have stopped grinding, so that the productive capacity
of the sector was not reduced. This is important, since what caused the crisis of
1920±21 was a process of worldwide economic adjustment due to the excess of
sugar induced by the start of recovery in the belligerent countries and the
protectionism now faced by the producers who had supplied their markets
during the war, with a consequent fall in prices. For Cuba the most serious
aspect was the increase in the US tariff from 1 to 1.6 cts/lb. These factors
undermined the pillars on which the island economy and sociopolitical system
rested in 1898 ± namely, sugar production for export to the USA with a prefer-
ential tariff (see SantamarõÂa and Naranjo, 1999, p. 243).
In the crisis of 1920±21 we can also find the precedents of the depression of
1930 and of the measures which would allow Cuba to withstand it. The key
element here was state intervention in the regulation of the industry, to which
end the Financial Sugar Commission (FSC) was established, with the respons-
ibilities for selling the zafra if 70 per cent of the mill's owners were agreed, and
the signing of an agreement with the other US suppliers. North American beet
producers, who were most affected by the deflation since their production costs
were higher, suggested to Cubans that they should exert pressure on Congress
to reduce the tariff in exchange for a decrease of their exports to the USA to 2.5
million tons. Meanwhile, several mill-owning US firms on the island founded
the Sugar Export Co. They rejected the aforementioned proposal and also the
idea that the FSC should be in charge of selling the zafra. Instead, they bought
500 000 tons from the 1920 stock, which they initially intended to send to
Europe but eventually placed in New York. Other Cuban sugar producers were
at first in favour of limiting supply, but later they too opted for the same
strategy. As a result, exports grew between 1921 and 1922 to over 2 million
tons, of which 1.6 were destined for the USA. It is also noteworthy that this
increase did not correspond to production, which had remained stable, rising
in 1925 to more than one million tons in respect of 1924.
In SantamarõÂa (1995b) we showed that Cuban mills profited from a very
elastic section in the demand curve for sugar, which allowed them to increase
their exports in 1922. Significantly, however, the rise in the zafra in 1925 is not
explained from the demand side. US consumption and world imports were at
this time at levels similar to those of 1922; in fact, market growth had reached
structural limits. Furthermore, comparatively, the norm for other producers was
a sustained increase in production between 1920 and 1925.16 The sudden rise in
the island's production must therefore be explained from the supply side, and
by analysing the postwar period in the light of the logic of the industry's
growth. Indeed, as we will see below, such an analysis is essential to under-
standing the preservation of the economic system based on sugar exports after
the 1930 crisis. Nevertheless, and even though it may seem contradictory,
before explaining this phenomenon, we must study its outcome.
In the introduction we stated that Cuba did not undergo a process of diversi-
fication similar to that which occurred in other Latin American countries as a
result of the Great War. In general, that event gave rise to increases in exports,
but reduced imports, favouring diversification. The stimulus to this on the
island, where import supply was guaranteed by USA, did not happen in the
same way, and after 1919, because sugar production did not decrease so as to
leave resources for other activities, again the situation did not change.17 The
GDP estimates are constructed with figures heavily dependent on the evolution
of the external sector, and are therefore not reliable as a measure of diversifica-
tion. In 1913 exports appear to have represented 44 per cent of income; 66 per
cent in 1919, and 55 per cent by 1924. In countries such as Chile or Argentina
the per centage was lower ± 11±15 per cent. Meanwhile, the proportion of
industrial output in Argentina's GDP increased from 13 per cent to 17 and 18
per cent in this same period; and in Chile it surpassed 20 per cent by the mid-
1920s exceeding exports in value.
For Cuba, indications such as the employment trend in each sector and the
conclusions of MarqueÂs (1998) show that the lack of structural change indic-
ated in the income estimates reflect the reality. The war and the postwar period
therefore reinforced the dependency of the island's economy on sugar and also
on the US market,18 in spite of the fact that in 1921, due to this double
dependency, the Cuban economic system had shown signs of a structural crisis
caused by the excess of supply of international sugar and the increase in the
North American tariff.
Rising sugar exports from 1922 onwards and production in 1925 had counter-
productive effects. In the medium and long term it worsened the structural
problems of the Cuban economy, and in the short term the opportunity cost of
this strategy was also negative. The following regression equation shows that
the main factor determining the price between 1920 and 1924 was the tariff:19
P 3:49T 0:58X 0:20A 1:78R 56:29D20

2:0 3:6 4:2 1:9 2:7
N 24; R2 adjusted 0:71; SE 19:23; DW 2:52; F 12:59; PV 0:0000
where P is the percentage increase in the average year-on-year value of Cuban

raw sugar in New York; X and R, the island sugar sold and the per capita income
in the USA; the percentage which the North American tariff represented in this
value, and T and D20, dummy variables to capture the effect of time ± calcu-
lated via the rate of increase in price ± and the alterations which the market
underwent in 1920 due to other factors. According to the equation, ceteris
paribus, and except in 1920, when other factors were of primary importance,
the 500 000 tons sold in New York by the Sugar Export Co. depressed the price
by 13 per cent. In response, the USA raised the tariff from 1.6 to 1.76 cts/lb; in
other words, an increase of 0.76 cts in respect of the 1920 level, which alone
caused a deflation of 50 per cent.
It was to be expected that an increase in Cuban sugar exports would have a
negative opportunity cost in the short term, since the end of this course of
action was to carry out dumping to eliminate competition and maintain the
market achieved during the war. The financial backing from the USA for the
island's industry in 1921 allowed it to happen, and further advantage was
gained from a temporary cutback in world supply and the following price rise
after 1922, due to the crisis in the Rhine which curbed the recuperation of
European production, and to a mosaico plague which devastated many sugar-
cane plantations in Puerto Rico and the North American South.20
However, as can be deduced from the previous equation, it was foreseeable
that protectionism would rule out the success of dumping. Moreover, it would
seem that this situation had a harmful effect on the banks' calculations of the
market trend: as such, they prevented Cuban mills in financial difficulties from
stopping grinding in 1921. Cleveland and Huertas et al. (1985) prove that this
was what happened in the most important case, the National City Bank ± it
eventually controlled 37 factories, which produced 25 per cent of the zafra in
1927. In effect, the island sugar producers knew in 1924 that with the end of
the exceptional conditions of the last triennium, European and US industry
would again grow and depress the price. The curious thing is that they then
decided to increase their own supply by 1 million tons, further saturating the
market and worsening the deflation to the point that historiography considers
1925 as the date marking the end of the upward cycle in sugar production.21 It
was not necessary to continue dumping, since the maintenance or raising of
the level of exports did not require producing more sugar ± there was a large
stock accumulated from earlier years. But, furthermore, that was not the aim,
demonstrated by the fact that immediately after the sharp rise in the zafra, the
same producers asked the Cuban state for the regulation which they had
rejected in 1921.22
Crisis and adjustment, 1925±37
A 1926 law stipulated that the mills should reduce their supply by 10 per cent.
There was also an attempt to sign an agreement with European producers and
Washington was asked to revise the Reciprocal Trade Treaty. This policy failed
with respect to its external objectives. Cuba was unable to obtain a reduction in
US tariffs, since exports to this country were not regulated, or to reach a
consensus on a compromise in the old continent, to which Java ± a Dutch
colony ± was opposed. Following the establishment of protectionist measures in
Japan and India, both its traditional customers, Java had begun to send sugar to
Europe, and consequently it was not possible to improve the price. In brief, in
1928 it was obvious that the island was capable of swamping the market and
depressing the value of sugar, but incapable of promoting on its own an
improvement in the latter or of reaching the necessary international agree-
ment.23 In the face of failure, regulation was eliminated and in 1929 the mills
were allowed to produce freely again, worsening the saturation of the market.
This occurred just before the beginning of the 1930 crisis which caused a new
increase in North American tariff. Internally, the restrictive legislation was
successful, since by limiting supply it prevented the larger and more efficient
factories ± especially those which were the property of refining companies in
USA ± from monopolizing the demand of these refineries, the main buyers of
Cuban output and payers of better prices.
That the government should consider preventing the largest factories from
monopolizing the market is explained by the changes in the sociopolitical
situation in Cuba in the mid-1920s. We have pointed out above that the
1920±21 crisis undermined the pillars on which the system of domination
rested and the later events contributed to the destabilization of the sociopoli-
tical situation. Faced with growing conflict, the ruling class grouped together
around a project led by G. Machado, who went on to win the elections in 1925.
The project proposed revisions to the Reciprocal Trade Treaty, the regulation of
sugar production to protect the smallest producers, the abolition of the Platt
Amendment, a tariff reform and a public works programme. We have already
discussed the first two objectives and the third did not materialize, but tariffs
and expenditure did change. On average, budgets grew from US$65.3 million in
1919±24 to US$84.4 million in 1925±29 (from 8.2 to 13.6 per cent of GDP); in
1927 a new customs regulation was published and in 1925 the construction of
the central highway began.24
Machado's project was coherent given the difficult situation in which the
country found itself. Paradoxically, however, that very situation prevented
its success in the short term. The alterations which the World War and the
early postwar years had caused in the growth of sugar supply brought about a
structural crisis in the Cuban economic and sociopolitical system. At the same
time, the changes increased dependency on this product and on the North
American market, reducing the options for adjustment. This was the necessary
condition for the maintenance of the system, even after the 1930 crash, but
the sufficient condition was that sugar also afforded the means for dealing
with the situation. The key was to renew the trade agreements with the USA,
hence the revised bill of the Reciprocal Trade Treaty and the threat to
abolish the Platt Amendment and reform the tariffs to modify the concessions
granted to its imports into Cuba in 1902. This also led to the restriction on
sugar supply, thereby recognizing the responsibility of the island producers in
the strengthening of North American protectionism. However, so that his
policy would work, Machado had to make concessions which thwarted the
achievement of that objective, since what prevented the legislation of 1927±
28 from regulating exports to USA was the pressure exerted by its refiners.25
Machado's bill laid the foundations of what later became the policy to con-
front the depression. However, it also contributed to worsening its effect. The
creation of the oligarchic block and the deferral of the presidential mandate in
1928, in violation of the Constitution, deinstitutionalized the channels for
expressing opposition, especially that of the new social sectors ± the workers'
movement and the urban middle classes. These groups were most affected by
the economic crisis, and they played a leading role in protests which led to the
constitution of the oligarchic block in 1925. Civil confrontation increased as a
result. Furthermore, the failure of short-term economic policy forced a resort to
external credit to finance the rise in expenditure. Between 1900 and 1926 Cuba
obtained US$103 million in loans; between 1926 and 1929, US$189 million and
in 1932 another US$20 million, so that the crisis of the 1930s found the
country in debt and also heavily dependent on the flow of foreign capital,
which broke down in 1929.26
We can now see the significance of the sharp rise in the zafra in 1925, since
basically it caused the adjustment which provided the means to face the crisis
and reinsert Cuba into the reorganisation of the post-depression world market
as an exporting economy. Dye (1998) has shown that the costs of adjustment
inherent in the coordination of the complex sugar production system meant
that mill capacity was underutilized for a few years from its foundation or
technological renewal, and to optimize costs it was necessary to develop this
capacity later. In SantamarõÂa (1995b) we should that the Great War gave rise to
an accelerated increased in industrial infrastructure which had not become
profitable by the end of the conflict. Fifty three percent of the centrales were
in this situation in 1919. They were also, on average, bigger than the rest and
almost all were acquired by new companies founded by the banks, which
averted their closure in 1920±21, or belonged to firms backed by these banks.
From these mills, 63 per cent stepped up their production in 1925 by more than
26 per cent ± a per centage equal to the increase in the island's aggregate supply
with respect to 1924 ± and the rest had in the main developed their optimum
productive capacity before that date. In short, therefore, it can be said that in
1920 Cuba was producing more sugar than the market could absorb, but less
than existing infrastructure allowed.
The development of infrastructure had a harmful effect on the island's eco-

nomic and sociopolitical system, but it also meant that Cuban ingenios were once
again the most competitive in the world. The only question remaining unan-
swered is why 1925 was chosen for the increase in the zafra. We do not know
whether the mosaico plague prevented the increase at the earlier date at which it
ocurred in other islands of the Caribbean. However, the timing of that increase
and the producers' knowledge in 1924 that the period of static European and US
supply had ended could not be simple coincidence. With the failure of dumping,
the alternative became a restrictive policy calculated on a pro rata basis accord-
ing to mill production. In order to avoid penalty, those centrales still operating
below capacity had to step up output. This thesis is also supported by the fact that
the zafra in 1925 seems to have been carried out with unusual haste, since
quantity was given priority over yields which were lower than those of later
years when similar tons of sugar were supplied (1929 and 1930).27
Centrales developed all their productive capacity at the same time as starting a
new stage of intensive growth, known as intensivismo. Both extensive and
intensive growth coexisted in some years, and it is not a contradiction, since
if the aim of the first one was to make the infrastructure incorporated during
the war profitable, the second one was effectively a process of adjustment to
confront the postwar market situation, process coherent with the logic of
sectoral development. In Cuba sugarcane can only be ground for a few months
a year (the rainy season ± June±October ± prevents grinding), and it is best done
between January and May when most of the canes are at optimum ripeness.28 It
is important because intensivismo was characterized by a reduction in harvest
days to save work and energy and further optimize the use of the raw material,
cutting the link between obtaining economies of scale and the increase in total
production per factory. Furthermore, it required little new technology and
involved only a small investment cost.29 Authors such as Chantez and
FernaÂndez (1985) analysed the phenomenon as something typical of the
1940s, and Zanetti (1996) dated its origin to 1926, coinciding with the restrict-
ive legislation. However, Figure 10.2 leaves no room for doubt: intensification
began in 1919, just at the end of the First World War, a time when the growth
curves of output per mill and per day separate.
Several reasons explain the mistake of authores like Chantez and FernaÂndez
or Zanetti. Chronologically, their studies do not go back far enough. Indeed, in
accordance with the logic of continuous-process technologies, because of the
costs of adjustment described by Dye (1991) it is necessary to search for the
beginning of a process of technical change prior to when the results materialize.
The data in Table 10.1 support this thesis. The sharp increase in the sugar/mill/
day quotient between 1913 and 1925 is not reflected in the yield, since for this
to be effective coordination was required with similar advances in other parts of
the productive chain, which did not happen until the second half of the 1920s
1400.0
Output/refinery
1200.0
Output/refinery/day
1000.0
800.0
Index
600.0
400.0
200.0
0.0
1901 1905 1915 1925 1935 1939
Figure 10.2 Index of sugar output per active mill and effective zafra day, 1901±39
(1903=100)
Sources: SACT (1903±36) and CEF (1937±40).
(until 1928 the yield did not rise above 12 per cent). Between 1913 and 1925,
the cane cut/colony/day rose spectacularly, but not the cane transported by rail
per day, whose growth came in 1925±30. Finally, Zanetti (1996) studied how the
mills adapted their machinery to those changes and points out that in the
1920s most had adopted procedures which allowed them to absorb a greater
inflow of raw material, solving what had been an endemic problem until then
in the sector: intensifying the grinding damaged the yield.
Another reason why the intensivismo of the 1920s went unnoticed is that it
slowed down in the 1930s, although this was due to causes of an institutional
nature. After 1930, state regulation was widened to cover all the elements
involved in the sugar industry. It fixed the dates for the start and end of the
grinding, stipulated workers' wages and cane prices, and asigned colonos and
centrales quotas of cane and sugar prduction and sugar export. This caused a
standstill and factor price rises, which nevertheless were compensated by the
intensification process. After an initial deterioration in the indicators in Table
10.1, even with lower zafras due to the crisis in exports after 1929, the sector
regained efficiency levels greater than those of other competitors. Despite
deflation, it operated only with losses in 1931±32, and after the renewal of
the trade agreements with the USA (1934) it once again produced profits, made
possible by a new reduction in the number of grinding days, the adoption of
small technical innovations, more appropriate selection of raw material,
improvements in the coordination of all the activities to avoid interruptions
in the operation and an increase in the efficiency of work ± these two last
aspects had made little progress during the 1920s.30
Table 10.1 Basic indicators of the intensification of sugar production, 1913±39a
Cane/colono/day Cane/km. rail/day Sugar/mill/day Industrial yield
1913 0.70 2.80 147.00 11.27

1925 1.20 2.90 284.00 11.61
1930 1.50 3.30 331.00 12.39
1933 0.90 2.50 281.00 11.70
1935 1.00 2.60 315.00 12.33
1939 1.20 3.10 324.00 12.48
a
1000 arrobas of cane cut/colono and arrobas of cane transported; sugar in tons; yield: 100 arrobas of
cane/arrobas of sugar.
Sources: SantamarõÂa (1995b, chs. IV and VII); SACT (1913±35); and CEF (1939).
Apart from intensivismo, other elements contributed to the successful adjust-

ment of the industry during the crisis. State regulation maintained the colonos'
incentive to guarantee the mills' optimum yield, handing over enough raw
material and of the highest quality, since they continued to receive a per centage
of the sugar obtained from their cane. Furthermore, it allowed the centrales to
give up their quotas to other factories, which meant that the most efficient mills
produced more sugar than the amount assigned to them during the years when
the lowest prices were being paid. In addition, the regulation altered the relative
cost of factors of production, cheapening labour relative to cane. Thus sectorial
employment levels were maintained, despite the reduction in supply and in the
number of harvest days, a key element in an economy which depended so
heavily on the sugar industry and which was also often paralysed by labour
unrest in the 1930s. Finally, the sector was reorganized to solve the problems
of overcapitalization caused by the intervention by the banks during the postwar
period, and it complemented its income with product diversification.
Diversification was a response to the change in the relative price of refined
compared to raw sugar caused by the step up in the US tariff and modifications
in consumer patterns which increased the demand for sugar substitutes. Cuban
sale of refino rose from 3000 tons in 1925 to 300 000 tons in 1929, and this
growth continued afterwards.31 The most important, however, was simply that
the island industry had begun to refine. This changed North American refiners'
attitude. Indeed, in 1927 they had helped prevent the regulation of exports to
their country, but now, the threat posed by this new competition led them to
lobby for a reduction in the tariff and the establishment of a quota system (see
Ballinger, 1971, p. 34). There was also an effective generation of external
economies. Already in the 1920s many mills rendered a public service through
their private railways, and in the 1930s they began to use their estates in a more
productive fashion, encouraging crops other than cane, and livestock, and
giving land to the workers ± the law allowed for part of the wages to be paid
in food or in use of the land.
Table 10.2 Annual variation of GDP and its causes, 1930±34a
Effect of tariffs and Exchange Variation Effect of paralysis

export reduction rate effect in GDP (%) In flow of capital
1930 5.8 5.2 2.6 0.2

1931 16.0 5.4 3.1 0.2
1932 19.7 7.4 4.0 1.2
1933 8.1 10.8 3.6 0.3
1934 17.5 5.6 3.1 0.4
a
Effects measured as per centages of GDP for each year. The tariff and export effects are calculated
estimating the loss of income resulting from failure to supply 50 per cent of US consumption,
following the 1929 tariff; the paralysis in the flow of capital was estimated on the assumption that
Cuba in the 1930s had received the same amount of loans as in the 1920s. The exchange rate was
assumed constant, at the level of 1929.
Source: SantamarõÂa (1995b, ch. VII, and 1998a).
Studying this adjustment of the Cuban sugar industry is the key to under-
standing what happened in its economy in the 1930s. The data in Table 10.2
show that the main depressive factor in this decade and also the most important
obstacle to recuperation was the US sugar tariff, which together with the decrease
of exports reduced the GDP more than the paralysis of the flow of capital or the
fall of exchange rates.32 This is contrary to what occurred in other Latin Amer-
ican countries (see Maddison, 1988; Thorp, ed 1989; and DõÂaz Fuentes, 1994).
This explains why the policy to confront the crisis was also different.
It was rational to expect that the island economy could reinsert itself in the
new post-Depression world order as a sugar exporter. To this end, and in the light
of this new order becoming increasingly determined by bilateral relations, it was
even more necessary to renew the trade treaty with the USA and sign an inter-
national sugar agreement. In 1930 an attempt was made in Brussels, but it failed
because the impact of the crisis on demand was underestimated and importers
were not included. To the USA Cuba proposed a separate pact ± the Chadbourne
Plan ± which was unsuccessful as it did not establish any binding obligations.
Both agreements limited the signatories to supply and export quotas. While
other countries did not comply with them, the island not only complied but
produced and sold below the assigned levels for some years due to the under-
estimation of demand, and maintained this policy after the Brussels Agreement
ended in 1935.33 This form of behaviour, as we said, corresponded to rational
expectations regarding the possibility of renewing the trade treaty with the USA.
To achieve it, Cuban producers recognized their responsibility in the issue:
The fault is ours, the market can carry 3 million tons yet we have sent them 4
million, causing a fall in the price and a defensive move on the part of the
sugar beet producers . . . . [We put forward] as a solution that a commission be
sent offering a pact limiting the production to be sold to the US, and
legislation to prevent the producers from exporting above a previously
established quota.
(ANHCC, 1930, pp. 11 and 21; author's translation)
Producers also knew that they had to be united. In 1927, when US refineries
succeeded in preventing the regulation of sugar exports, there had been no
consensus in Cuba. We have already seen that when the island began to refine
they changed their attitude.
The US attitude was also crucial: apart from considerations over the respons-
ibility it acquired in Cuba in 1898, it had to protect many of its citizens'
interests there, while at the same time its exports lost ground in the island
market, one of the most important in Latin America yet one whose purchasing
power had been severely affected by the crisis. Furthermore, a Pacific war was
considered a likely scenario for the 1930s, in which case Cuban sugar would be
necessary since this war would affect two of North America's suppliers ± Hawaii
and the Philippines.
All of these arguments, however, were of little use if US sugar beet producers
did not change their stance. To this end, Cuba limited its zafra and exports in
the hope that, since their increase had been the reason for raising the North
American tariff in the 1920s, their decrease would now serve to reduce it.
Indeed, in this sense there was one further advantage. Beet producers deter-
mined US trade policy, but they were not the ones who benefited most.
Between 1919 and 1930, US sugar production was on average 2.8 million
tons/year, a figure which increased to 4.3 million tons in 1931±9 (55 per cent)
due to the drop in Cuban sales, but sugar beet production only increased from
990 000 to 1.13 million tons (14 per cent). The Philippines, Hawaii and Puerto
Rico were the greatest beneficiaries; they operated with lower costs and even
threatened to push sugar beet out of the market, especially given that the tariff's
depressive effect on the price was forcing Cubans to accept increasingly smaller
payments for their product in order to avoid the customs tariff.
Besides agreeing to stabilize the market, Cuba still needed to fulfil one more
requirement to succeed in renewing the treaty with the USA: the restoration of
the internal sociopolitical order. With the Depression, conflicts increased and
often ended in violence largely because of the deinstitutionalization of the
forms of expression of opposition to the regime. Unable to resolve them,
Machado lost the backing of those who had supported him in 1925, and in
1933 a strike caused his resignation. However, the oligarchy no longer had
control of the state and the revolutionary government which took over
proved equally unable to restore order. Soto (1985) demonstrates that the revo-
lutionaries lacked an alternative to Machado's economic programme. They did
not go far beyond reinforcing its social justice component to meet the demands
of the middle classes and the labour movement. As a result, the need for agree-
ments with the USA, which would help finance the new social expenditure,
became even more pressing. The key, therefore, was the North American atti-
tude. A speech by Roosevelt before becoming president defined it with precision:
Due to the exceptional relations of our peoples the recognition of a govern-

ment in Cuba implies, more than an ordinary measure, rather a material and
moral support. We desire to start negotiations for a revision of the trade
relations and the modification of the Reciprocal Treaty, but no progress will
be made in these areas if there does not exist in Cuba a government which
has popular support and which can count on the general co-operation,
clearly demonstrating genuine stability.
(Pichardo, 1973, vol. VI, p. 103, author's translation)
Meanwhile, the social conflict in Cuba had reached extremes which Thomas
(1973) described as an armed stalemate, and the solution arose from that stale-
mate. For Tabares (1973) there were two forces:
Irreconcilable, revolution and reaction, they remained in impotent balance.

Neither one could overwhelm the other; there was no other option but to
coexist and make mutual concessions.
(Author's translation)
The pact was provided by the only institution which was unscathed in the crisis
of legitimacy in the political system: the army; namely its middle ranks who
took control in the so-called RevolucioÂn of the Sergeants (1933). From amongst
them there also came the new strong man of Cuba: Batista, who managed to
reunite the conflicting interests into a common programme:
Now the Republic is born on an irrefutable basis because it will take the form
which the country freely shows it wants. It will not be fascist, or socialist, or
communist, it will take the form that the majority wants to give it.
(Batista, 1933, p. 8, author's translation)
With this programme, post-revolutionary governments were able to guarantee

the stability required by the USA. Cuba obtained a preferential quota in the
North American sugar market and a new Reciprocal Trade Treaty was signed.
With this treaty and the international agreement the island reinserted itself in
the international market better than its competitors (for example Java), and it
could restore the sociopolitical order since the strategy yielded enough profits
to maintain a policy of more equal distribution of income which was the key to
achieving a certain consensus in Cuban society.34
Conclusion: a comparative history perspective
The specific nature of the Cuban economy in the Latin American context at the
beginning of the twentieth century ± a strong dependency on sugar production
and export to the USA, and at the same time levels of development similar to
that of the most diversified economies ± was reinforced during the First World
War and the 1920s, despite the reorganization of the international market
during the postwar period. These changes caused a structural crisis and a
destabilization of the internal sociopolitical order, heightened by the Depres-
sion of 1930. However, the island reinserted itself into the market during the
1930s, maintaining its specialization in the production and sale of sugar, espe-
cially to North America. The lack of diversification was the necessary, but not
the sufficient condition that explains this fact. The sufficient condition was the
adjustment in the sugar industry to confront the structural crisis, and an
economic policy concordant with the specificity of the Cuban economy and
with this adjustment, which allowed the country to reenter the world market in
a better situation than that of other competitors, and to regain a growth rate
similar to that of the most developed countries in Latin America. This allowed
the restoration of the sociopolitical order, thereby promoting a more equitable
income distribution.35
This chapter has thus taken issue with the standard view of Cuba, which
views the significance of the interwar years too simplistically in terms of
continued and extreme dependence on the US sugar market. We have told a
more complex story, which shows how the supply-side response to market
developments in the period from the First World War to the 1930s resulted in
an economic structure in the sugar sector which made expansion of production
an imperative, while this fact and the oligarchy consolidated behind Machado
created sociopolitical tensions only partially resolved after the Revolution of
1933, in a way which allowed political developments to reinforce the tight
relationship with the USA. The successful continuation of the preferential
relation with this country was a result of Cuba's competitiveness, of a stable
political situation, and energetic negotiations. The outcome in the 1930s at
least was a relatively good economic performance.
Notes
1 See Sanz et al. (1998, pp. 373±90), for the relation between the integration of Latin
American economies in the world market and the construction of the first railways;
Zanetti and GarcõÂa (1987, pp. 1±20) and SantamarõÂa (1995a and 1998b) for the specific
case of Cuba.
2 In the 1868±90 period, production stagnated somewhere between 579 000 and 775 000
tons due to the reorganization of the sector in those years. Between 1891
and 1897 it was always above 800 000 tons, reaching a record of 1.1 millions. The FOB
price in the USA fell from 3.40 cts/lb to 2.07 between 1890 and 1897 and the tariff
rose from 0 cts in 1890, to 40 per cent ad valorem in 1894 and to 1.685 cts/lb in 1897.
See Moreno (1978) Vol. III. The last general published work on the years prior to
Cuban independence is that by GarcõÂa (1996). Ingenio is the word for a Cuban sugar
factory; central is a completely mechanized ingenio; zafra is the word for the cane
harvest and it is utilized as a synonym of sugar production.
3 See PeÂrez de la Riva (1975) and GarcõÂa (1994b) regarding imports of Chinese coolies
and YucataÂn Indians.
4 See Moreno (1978), Vol. I and Deerr (1950) for an account of the technological
evolution of the industry.
5 In Cuba cane takes between eight and 12 months to grow, whilst in other places in
the world it takes 18. Furthermore, the shoots of each plant are normally used for six
or seven years. See MartõÂn et al. (1987).
6 The natural division between agriculture and manufacturing, explains the need for
building a transport system which would communicate between the canefield and
the central ± a private railway. The industrial sugarcane process can be divided into
three parts ± grinding, evaporation and filtering ± which must be perfectly coordin-
ated. A technical improvement which, for example, increases the flow of cane juice ±
guarapo ± from the mill to the evaporators is not efficient and causes a bottleneck if it
is not used in conjunction with innovations to these evaporators to absorb the
increase. For more on this subject see Dye (1998), pp. 102±49.
7 The construction of centrales meant a reduction in the number of factories and an
increase in production per unit ± 1365 ingenios and 328 tons in 1860; 850 and 445 in
1890, and 205 (all of them centrales) and 1464 in 1900. See Le Riverend, (1985: 490).
8 See Le Riverend (1973), Perez (1986), GarcõÂa, (1998) and Zanetti (1989 and 1998), for
more on this subject.
9 For more on the new legal framework, see Jenks, (1928) and Le Riverend (1973). With
regard to the expansion of the sugar industry in the Eastern half of Cuba, Hoernel
(1976) and LuzoÂn (1989). For more on the extension of the railway, Zanetti and
GarcõÂa, 1998 and SantamarõÂa, 1995a and 1998b.
10 The investment of foreign capital mainly in rail rather than in the sugar industry was
not a phenomenon specific to the twentieth century, and nor was it confined to US
investments. In SantamarõÂa (1998b) p. 303, we explain that this was probably due to
opportunities for self-financing which enabled a slow modernization of the mills.
This did not happen in the rail sector, where large injections of capital were necessary,
either to build a complete network of transport ± Eastern provinces ± or to consolidate
in a single company a dozen lines which had emerged in the Western half of Cuba to
satisfy local needs, and which over time had stopped making a profit due to competi-
tion amongst themselves. US investment did not rise above US$ 50 million in 1896
and of that amount 30 was placed in the sugar industry. Around 1911 the first figure
had grown to approximately 200 million, whilst the second only rose to 50. Rail
construction accounted for 40. See Pino, 1984: 342, Jenks, 1928, chs. I-VI, and LoÂpez
Segrera, 1973: 173.
11 Data from Brundenius (1984)
12 According to Zanetti and GarcõÂa, 1976: 212, the first official licence to import West
Indian casual-workers was obtained by the United Fruit Co. in 1913. From that date
and until the mid 1920s more than 500 000 entered Cuba in order to work in the
sugarcane harvest, coming mainly from the other Caribbean islands and from Spain.
See PeÂrez de la Riva (1975).
13 The excepcional increase of prices allowed inefficient mills to go on producing. See

SantamarõÂa, 1995b, fig. III.1 and Dye, 1998, ch. 4.
14 In 1913, 46 per cent of the mills were public limited companies and 36 per cent were
the property of firms which owned more than one factory; by 1919 these per centages
had grown to 69 and 46 respectively. A single enterprise, the Cuba Cane Sugar Co.,
founded in 1915, invested US$ 50 million in the sector ± a sum similar to the total
North American capital placed in the island until this date. The firm bought 19
centrales and in 1919 produced 14 per cent of Cuban sugar. Bank and financial groups
representatives from US made up 63 per cent of the board of directors. See Santa-
marõÂa, 1996: 246, Pino, 1984: 390, and GarcõÂa, 1994a: 45.
15 For more on the process known as the Danza de los millones see Pino, 1984: 60, Le
Riverend, 1973: 165, and Collazo, 1994. Collazo explains the deflation and the
bankruptcy in 1920±1921 of the main Cuban banks as a result of a confrontation
between the island producers of raw sugar and US refineries; however, in SantamarõÂa,
1994, we proved that the process was more complex, and formed part of a general
downturn which did not affect Cuba alone, and therefore is difficult to describe in
such manichaean terms.
16 In 1922 per capita consumption in the US rose to 103 kg./year, then fell, to rise to 109
in 1926 and then begin a decline which lasted for more than 15 years. See Santa-
marõÂa, 1995b, appendix XII.4.
17 On the reduction of imports during the war and the relationship with economic
diversification, see Thorp, ed., 1989. In Chile, for example, the value of imports fell
by 60 per cent to constant price levels between 1914 and 1919, whilst in Cuba it rose
by 88 per cent between 1913 and 1919 (Palma, 1989: 58, and SantamarõÂa, 1995b, fig.
III.1). Jenks, 1928, explains that the US not only guaranteed the supply of provisions
to the island, but also used this as a weapon so that producers accepted their control
of the sugar market and the prices imposed from 1917 onwards.
18 For the calculation of GDP, see Alienes, 1950, and Brundenius, 1984. The figures for
Chile and Argentina are from Palma, 1989, Bulmer-Thomas, 1994, DõÂaz Fuentes,
1994, and SantamarõÂa, 1995b and 1998a; those on sector employment are from the
DGC, 1907 and 1919, and Memorias ineÂditas del Censo de 1931, 1978. The doctoral
thesis by MarqueÂs, 1998 indicates that even though there was industrial development
apart from the export industry in Cuba, due to the multiplier effects of the external
sector, this was of a complementary nature with regard to sugar. This characteristic
was reinforced during the First World War and the post-war period.
19 Data from SantamarõÂa, 1995b, appendices.
20 On the financial backing given to the sugar industry, according to Wallich, 1953: 106,
bank loans during the inflation of 1920 rose to US$ 80 million; a figure which was
maintained in 1921 (79 million) to finance the losses caused by deflation and the rise
in the US tariff. Zanetti and GarcõÂa, 1976: 133, calculate furthermore that another 23
million should be added to that amount in direct investments.
21 On the producers' forecasts for 1925, see `Balance econoÂmico', Cuba ContemporaÂnea,
138, 1924: 162. On the end of the upward cycle in sugar production, Alienes, 1950.
22 Pino, 1984: 455.
23 On the tariff, from 1924 the US Tariff Commission was recommending to the US
President that it should be reduced due to the damage which was caused to cus-
tomers, USTC, 1926. According to Bernhardt, 1948: 76, the pressure of the sugarbeet
producers, arguing that Cuba had not diminished its exports, prevented putting the
recommendation into effect. For more on the refusal of Java to sign an agreement in
Europe, see Ballinger, 1971: 31.
24 See LoÂpez Segrera, 1980, and Thomas, 1973, vol. II. Pollitt, 1989, is the study which
best explains the relation between the crisis of sugar production and the arrival to
power of Machado. See also SantamarõÂa, 1994, 1995b and 1998a, and IHC, 1998, vol.
III: 240.
25 With regard to the tariff, MarqueÂs, 1989 proves that it was a reaction to US trade
policy and its objective was to alleviate the reduction in state income, heavily
dependent on customs earnings which were depressed by the fall in the price of
sugar, rather than a measure to encourage economic diversification.
26 On the effect of the depression on external credit see Kindleberger, 1985, Galbraith,
1989 and Fishlow, 1985; with reference to loans granted to Cuba, IHC, 1998, vol. III:
405.
27 For more on how in the second half of the 1920s the island industry became again the
most competitive in the world, see Dye, 1993: 586, and USTC, 1926; about the sugar
yield in 1925, SantamarõÂa, 1994: 136.
28 MartõÂn et al., 1987: 567.
29 The effective length of the zafra ± the average time taken by factories to do it ± was
reduced from 132 days in 1919, to 107 in 1925, 90 in 1930, 61 in 1935 and 62 in 1939.
See CEF, 1940.
30 MartõÂn et al., 1987: 576 and Dye, 1991 show that the interruptions in the operation
are the worst possible drawbacks for an effective production process.
31 Apart from sugar refined, the Cuban sugar industry began to produce rich molasses
and syrups, and increase the distillation of rum, demanded by US consumers. See
SantamarõÂa, 1995b, ch. VI.
32 Dye and Sicotte, 1999, calculate that nearly 35 per cent of the losses in real export
earnings between 1930 and 1934 was caused by the US sugar tariff.
33 On bilateralism in trade in the 1930s, see Kenwood and Loughed, 1972: 337±340. On
the Brussels Agreement and the Chadbourne Plan, Zanetti, 1989: 133±135 and Bal-
linger, 1971: 37.
34 For more on all these aspects see SantamarõÂa, 1998a and 1995b, ch. VI. In these
studies we analyse the results for the Cuban economy of the international agreement
and the new treaty with the US in a comparative perspective. This analysis confirms
our conclusions about the relative success of the adjustment in this economy to the
post-Depression market situation. Another aspect also analysed is why some mea-
sures thought to be to confront the crisis were maintained subsequently and what
problems this meant.
35 We have not enough space in this chapter to explain in depth the adjustment in the
Cuban economy and its results in comparative terms. In SantamarõÂa, 1998a and
1995b, ch. VI, the reader may find the data about the more equitable income
distribution and the island's macroeconomic statistics compared to those of other
Latin American countries which prove our affirmations.
References
Alienes, J. (1950) CaracterõÂsticas fundamentales de la economõÂa cubana, Havana: Banco

Nacional de Cuba.
Albert, B. and Graves, A. (eds) (1988) The World Sugar Economy in War and Depression,
London: Routledge.
ANHCC (AsociacioÂn Nacional de Hacendados y Colonos de Cuba) (1930) Estudio sobre el
problema azucarero y sus consecuencias en la economõÂa cubana, Havana: ANHCC.
Ballinger, R.A. (1971) A History of Sugar Marketing, Agricultural Report no. 197, Washing-
ton, DC: US Department of Agriculture.
Batista, F. (1933) Al pueblo de Cuba, Havana: Lex.
Bernhart, J. (1948) The Sugar Industry and the Federal Government, a Thirty Years Record,
Washington, DC: Government Printing Office.
Brundenius, C. (1984) Revolutionary Cuba: The Challege of Economic Growth of Equity,
Boulder: Westview Press.
Cayuela, J.G. (ed.) (1998) Un siglo de historia de Espan Ä a: centenario, 1898±1998, Cuenca:
Universidad de Castilla-La Mancha.
Chantez, S. and FernaÂndez, J.M. (1985) Èl fenoÂmeno econoÂmico del intensivismo en las
zafras azucareras de la deÂcada de 1940', Islas, no. 80, pp. 3±8.
CEF (Cuba EconoÂmica y Financiera) (1937±58) Anuario Azucarero de Cuba, Havana: CEF.
CII (Cuba Importadora e Industrial) (1932±36) Havana: CII.
Cleveland, H.B., Huertas, T.F. et al. (1985) Citibank, 1812±1970, London: Harvard Uni-
versity Press.
Collazo, E. (1994) Una pelea cubana contra los monopolios (un estudio del crac bancario de
1920), GijoÂn: Universidad de Oviedo.
Cuba ContemporaÂnea (1913±26) Havana: Cuba ContemporaÂnea.
Deerr, N. (1950) The History of Sugar, 2 vols, London: Chapman & Hall.
DõÂaz-Alejandro, C. (1989) ÀmeÂrica Latina en los an Äos treinta', in R. Thorp (ed.), (1989)
pp. 13±35.
DõÂaz Fuentes, D. (1994) Crisis y cambios estructurales en AmeÂrica Latina. Argentina, Brasil y
MeÂxico en el perõÂodo de entre-guerras, Mexico: Fondo de Cultura EconoÂmica.
DGC (DireccioÂn General del Censo) (1907, 1919 and 1943) Censo de la RepuÂblica de Cuba,
Havana: DGC.
Dye, A.D. (1998) Cuban Sugar in the Age of Mass Production. Technology and the Economies of
the Sugar Central, 1899±1929, New York: Stanford University Press.
ÐÐ (1993) `ProduccioÂn en masa de azuÂcar cubano, 1899±1929', Revista de Historia Eco-
noÂmica, vol. 11 (1), pp. 563±93.
Dye, A. and Sicotte, R. (1999) `The U.S. Tariff and the Cuban Revolution', umpublished,
New York: Barnad College, and Boston: University of Harvard.
Fishlow, A. (1985) `Lessons from the Past: Capital Markets during the 19th Century and
the Interwar Period', International Organization, vol. 39 (3), pp. 73±87.
Galbraith, J.K. (1989) El crac de 1929, Barcelona: Ariel.
GarcõÂa, A. (1998) `La consolidacioÂn del dominio imperialista', in Instituto de Historia de
Cuba, Historia de Cuba, IHC, (1998) Vol. III, pp. 90±141.
ÐÐ (1996) Èstructuras de una economõÂa colonial en transicioÂn', in C. Naranjo et al.
(eds.), 1996, pp. 195±210.
ÐÐ (1994a) Ùna saga azucarera en torno a dos siglos', in J. UrõÂa (ed.), 1994., pp. 43±56.
ÐÐ (1994b) `Los traficantes del Golfo', Historia Social, vol. 18, pp. 33±46.
ÐÐ and GarcõÂa Mora, L.M. (eds) (1998) Textos claÂsicos de la historia de Cuba (CD-Rom),
ColeccioÂn ClaÂsicos Tavera, Madrid, FundacioÂ n HistoÂ rica Tavera y Digibis.
Hoernel, R.B. (1976) `Social and Social Change in Oriente, Cuba, 1898±1946', Journal of
Latin American Studies, vol. 8(2), pp. 215±49.
IHC (Instituto de Historia de Cuba) (1994±) Historia de Cuba 5 vols, Havana: PolõÂtica.
Jenks, L.H. (1928) Our Cuban Colony, New York: Vanguard Press.
Kenwood, A.G. and Loughed, A.L. (1972) Historia del desarrollo econoÂmico mundial, Madrid:
Istmo.
Kindleberger, C.P. (1985) La crisis econoÂmica, 1929±1939, Barcelona: CrõÂtica.

Le Riverend, J. (1985) Historia econoÂmica de Cuba, Havana: Pueblo y EducacioÂn.
ÐÐ (1973) La RepuÂblica. Dependencia y revolucioÂn, Havana: Ciencias Sociales.
LoÂpez Segrera, F. (1973) Cuba: capitalismo dependiente y subdesarrollo (1510±1959), Mexico:
DioÂgenes.
LuzoÂn, J.L. (1989) EconomõÂa, poblacioÂn y territorio en Cuba, Madrid: Cultura HispaÂnica.
Maddison, A. (1988) Dos crisis: AmeÂrica Latina y Asia, 1929±1938 y 1973±1983, Mexico:
MarqueÂs, M.A. (1998) Empresas y Empresarios en las entidades industriales menores en Cuba,
1870±1920, PhD, Madrid: Universidad AutoÂnoma de Madrid.
ÐÐ (1989) Ìntereses y contradicciones de clase en torno al problema arancelario
cubano', Santiago, no. 74, pp. 119±42.
MartõÂn, J.R. et al. (1987) La can Ä a de azuÂcar en Cuba, Havana: Ed. CientõÂfico TeÂcnica.
MartõÂnez-Alier, J. and V. (1972) Cuba, economõÂa y sociedad, ParõÂs: Ruedo IbeÂrico.
Memorias ineÂditas del censo de 1931 (1978) Havana: Ciencias Sociales.
Moreno, M. (1978) El ingenio. Complejo econoÂmico-social del azuÂcar cubano, 3 vols, Havana:
Ciencias Sociales.
Naranjo, C. (1984) ÀnaÂlisis histoÂ rico de la inmigracioÂn espan Äola a Cuba', Revista de Indias,
no. 174, pp. 216±55.
ÐÐ et al. (eds) (1996) La nacioÂn son Ä ada. Cuba, Puerto Rico y Filipinas ante el 98, Aranjuez:
Doce Calles.
Palma, G. (1989) `De una economõÂa de exportacioÂn a una economõÂa de sustitucioÂn de
importaciones: Chile, 1914±1939', in R. Thorp (ed.), (1989), pp. 223±47.
PeÂrez, L.A. (1986) Cuba Under the Platt Amendment, 1902±1934, Pittsburgh: University of
Pittsburgh Press.
PeÂrez de la Riva, J. (1987) El barracoÂn. Esclavitud y capitalismo en Cuba, Barcelona: Grijalbo.
ÐÐ (1979) `Cuba y la inmigracioÂn antillana, 1900±1931', Anuario de Estudios Cubanos,
vol. II, pp. 5±75.
ÐÐ (1975) Para la historia de las gentes sin historia, Barcelona: Ariel.
Pichardo, H. (1973) Documentos para la historia de Cuba, 7 vols, Havana: Ciencias Sociales.
Pino, O. (1984) Cuba, historia y economõÂa, Havana: Ciencias Sociales.
Pollitt, B.H. (1988) `The Cuban Sugar Economy in the 1930s', in B. Albert and A. Graves
(eds), (1988), pp. 97±108.
SACT (SecretarõÂa de Agricultura, Comercio y Trabajo) (1912±13) Portfolio Azucarero. La
industria azucarera en Cuba, Havana: LibrerõÂa e Imprenta La Moderna PoesõÂa.
ÐÐ (1903±36), Memoria de la Zafra Azucarera, Havana: SACT.
Salter, W.E.G. (1966) Productivity and Technical Change, Cambridge: Cambridge University
Press.
SantamarõÂa, A (2000) `Precios y salarios reales en Cuba, 1972-1914', Revista de Historia
EconoÂmica, Vol 18 (2).
ÐÐ (1998a) `La economõÂa y la polõÂtica econoÂmica cubana en el perõÂodo de entreguerras,
1919±39. Una perspectiva comparada', unpublished, Madrid.
ÐÐ (1998b) Èl ferrocarril y la economõÂa en las Antillas espan Ä olas, Cuba Puerto Rico y La
RepuÂblica Dominicana, 1837±1995', in J. Sanz (ed.) et al. (1998), pp. 289±334.
ÐÐ (1997) ÀnaÂlisis de los costes de produccioÂ n del azuÂ car en Cuba, 1804±1913', unpub-
lished, Oxford: University of Oxford.
ÐÐ (1996) `Can Äa de azuÂ car y produccioÂn de azuÂ car en Cuba. Crecimiento y organizacioÂ n
de la industria azucarera desde mediados del siglo XIX hasta la finalizacioÂn de la
Primera Guerra Mundial', in C. Naranjo et al. (eds.), (1996), pp. 225±50.
ÐÐ (1995a) `Los ferrocarriles de servicio puÂblico cubanos, 1837±1959. La doble natura-

leza de la dependencia azucarera', Revista de Indias, no. 204, pp. 485±515.
ÐÐ (1995b) La industria azucarera y la economõÂa cubana durante los an Ä os viente y treinta, PhD
dissertation, Madrid, Universidad Complutense de Madrid (Instituto Universitario
Ortega y Gasset).
ÐÐ (1994) `La crisis de 1920±21 y el ajuste al alza de la industria azucarera cubana',
Revista de Historia Industrial, no. 5, pp. 121±48.
SantamarõÂa, A. and GarcõÂa Mora, L.M. (1998a) `Colonos. Agricultores can Äeros, ¿clase
media rural en Cuba? 1880±1898', Revista de Indias, no. 212, pp. 131±62.
ÐÐ (1998b) À propoÂsito de la industria azucarera en Cuba, 1860±77. Mano de obra y
tecnologõÂa', in J.G. Cayuela, (ed.), (1998), 283±98.
SantamarõÂa, A. and Naranjo, C. (1999) Èl `98 en AmeÂrica. U Â ltimos resultados y tendencias
recientes de la investigacioÂn', Revista de Indias, no. 215, pp. 203±74.
Sanz, J. (ed.) et al. (1998) Historia de los ferrocarriles en IberoameÂrica, 1837±1995 (book), and
GuõÂa histoÂrica de los ferrocarriles en IberoameÂrica, 1837±1995 (CD-Rom), Madrid: Minis-
terio de Fomento.
Tabares, J.A. (1973) La RevolucioÂn del 30: sus dos uÂltimos an Ä os, Havana: Ciencias Sociales.
Thomas, H. (1973) Cuba, la lucha por la libertad, 3 vols, Barcelona: Grijalbo.
Thorp, R. (ed.) (1989) AmeÂrica Latina en los an Ä os treinta. El papel de la periferia en la crisis
mundial, Mexico: Fondo de Cultura EconoÂmica.
UrõÂa, J. (ed.) (1994) Asturias y Cuba en torno a 1898, Barcelona: Labor.
USCCA (US Commission on Cuban Affairs) (1935) Problems of New Cuba, New York:
Foreign Policy Association.
USTC (US Tariff Commission) (1926) Sugar. Report of the United States President, Washing-
ton, DC: Government Printing Office.
Wallich, H.C. (1953) Problemas monetarios de una economõÂa de exportacioÂn. La experiencia
cubana, 1914±1947, Havana: Banco Nacional de Cuba.
Zanetti, O. (1998) Comercio y poder. Relaciones cubano±hispano±norteamericanas en torno a
1898, Havana: Editorial Casa de las AmeÂricas.
ÐÐ (1996) La dinaÂmica del estancamiento. El cambio tecnoloÂgico en la industria azucarera
cubana entre 1926 y 1958, Havana: Instituto de Historia de Cuba.
ÐÐ (1989) Los cautivos de la reciprocidad, Havana: EMPES.
ÐÐ (1983) `1929: la crisis mundial y la crisis cubana', Santiago, no. 49, pp. 173±94.
ÐÐ (1975) Èl comercio exterior de la RepuÂblica Neocolonial', Anuario de Estudios Cuba-
nos, vol. I, pp. 45±183.
Zanetti, O. and GarcõÂa, A. (1998) Sugar and Railroads, Chapel Hill: North Carolina Uni-
versity Press; Spanish version (1987) Caminos para el azuÂcar, Havana: Ciencias Sociales.
ÐÐ (1976) United Fruit Co.: un caso de dominio imperialista en Cuba, Havana: Ciencias
Sociales.
Index
agrarian reform (Mexico) 135 role of Central Bank in road building

Agrarismo 135 (Bolivia) 206
Antioquia, leader in Colombian Baring crisis 1
development 77 Batista, Fulgencio 315
APRA 158, 163, 166 Bautista Saavedra 200
Argentina Bolivia
agriculture 268, 269, 271±2, 280, communication systems 189
293 economic relations with US 192±4
capital productivity 282 education 206±9
capital stock 295±6 effects of Chaco War 189, 209±11
cattle raising 268, 272±3, 293 effects of Great Depression 189, 209±11
construction 268 export base 188
convertibility 284, 286, 289±90 liberal reforms 189
economic relations with Britain 291 linkages 188±9, 189±90
effect of First World War 269, 288±9 population 189
factors of growth 265±7, 293, 294 railways and roads 204±6
growth trends 267±70 role of state 189, 190, 209±11
industrialization 268, 281±2, 293 tin 188, 190±204, 209 (see also tin)
investment 268, 273, 274±5, 283, 293, Brazil
296 n 3 coffee commercialization 39±40
monetary and exchange rate coffee exports 34±7
policy 285±7 cotton exports 37
political conflict 292 economic regions in 32±3
population 267±8, 276, 283 geography of 33
post-war policy 289±92 labour 39
protectionism 267, 280, 291, 294 market power of (price maker) 41,
public expenditure 280 43±5, 50
public finances 287±8 population 33
railways 269, 274, 280 rubber exports 37±8
role of state 293 state policies towards coffee 40±1,
savings 274±5 46±7, 48±50
social conflict 293 standard of living 42±3
social legislation 292, 297 n 10 valorization 41±2
tariffs 284±5, 297 n 7 Britain 2, 3, 241
terms of trade 270 Brussels Agreement 313
trade policy 284±5 Bunge, A. 266, 268, 280
wages 278±80 Busch, GermaÂn 208
ASARCO 128
CaÂceres, AndreÂs Avelino 154
Balmaceda, JoseÂ Manuel 217, 219, 244, capital
249, 250, 252 domestic: role in Bolivia 209; role in
bananas, 93±6, 108, 112 Colombia 79
banking, 3 foreign: role in Brazil 39; role in
Bank Charter Act (UK) 3 Colombia 69; cuban dependency
deposits (Colombia) 69 on 309
early banking (Central America) 99 estimates of stock of (Argentina) 295±6
324 Index
capital (contd.) import tariffs 237±40

increase in productivity of industrialization 227, 256
(Argentina) 282 inflation 241, 260 n 35, 260 n 36
saturation of (Bolivia) 202±3 linkages 220±2, 226±7, 231
slowdown in accumulation of nitrates 218±20, 222±3, 224±5, 228±9,
(Argentina) 293 247±8, 254, 257 n 9, 257 n 10, 258
structure of (Argentina) 273±5 n 11, 258 n 15, 258 n 16, 258 n 17, 258
Cardenismo 124, 137 n 18, 261 n 43 (see also nitrates)
cattle-raising 268, 272±3 political conflict 250
Central America protectionism 237, 239±40, 242, 260
adoption of gold standard 98±99 n 33
bananas 93±6, 108 public expenditure 235±6, 238, 245,
coffee 87±93, 106±8 247±9, 261 n 42
credit systems 99±101 public revenues 218, 234±5, 237, 239,
entrepreneurship 92 256, 257 n 3
geography 86 railways 227, 259 n 29
imports 101±2 rate of exchange 240±1, 243±5
income distribution 85 relations with Britain 257 n 1
internal migration 101 returned value 229±31
Jamaican workers 95 role of state 234
lack of import substitution 85 terms of trade 232
lack of private investment 102±3 church 7
lack of technological innovation 89 coffee
liberal reforms 89, 104±5 Brazil as price-maker 41, 43±5
monetary instability 97±8 credit for (Colombia) 73
population 86 entrepreneurship in (Central
social inequality 113±15 America) 92
state expenditure 103±4 expansion of (Colombia) 59±62, 71±3
strategic importance 86±7 exports of (Central America) 106±13
transportation 86, 92 impact on standard of living of
urbanization 101 (Brazil) 42±3
Cerro de Pasco Corporation 174±8 innovation in (Colombia) 73±4
Chaco War 189, 197, 198 marketing of (Colombia) 62, 73±4
Chadbourne Plan 313 policies towards (Brazil) 48
Chapare 202, 205 production in Brazil 38
Chile production in Central America 87±92
agriculture 256, 262 n 1 role in Colombian
avoidance of `Dutch Disease' 242±55 industrialization 75±81
convertibility 218 valorisation of 41±2, 47
copper industry 223, 225, 228, 255, Colombia
258 n 13 banking 63, 69
effect of the War of the Pacific 217±20, capital accumulation 55±6
247 coffee 55, 59±62, 67, 71±5
exports 217, 220, 222, 224, 225±6, 228, fiscal instability 66
231, 237±9, 245, 247, 258 n 14, 259 geographic fragmentation 55±6
n 20, 260 n 37 industrialization 56, 65±6, 75±81, 82
fall in transport costs 241±2 n 10, 82 n 12, 82 n 13
fiscal policy 233±4, 248, 250±3, 255, inflation 66
259 n 22, 259 n 31, 260 n 34, 261 n 44, labour 71, 82 n 9
261 n 46 metals 57±9
foreign investment 229±31, 232 National Federation of Coffee
imports 233, 238±9, 244, 247, 260 n 32 Growers 75
Index 325
population 63, 81 n 4 sugar 300±16

quinine 56 US protectionism 307, 309
state 70 US tariff on sugar 306±7, 308, 312, 313,
technological innovation 63, 73±4 318 n 23, 319 n 25, 319 n 32
terms of trade 67 currency
trade unions 71 in Colombia 81 n 2
transportation 62, 63, 69±70, 81 n 5 mixed circulation 9, 119, 160
commercial systems customs 9
in Central Peruvian Andes 175 Cuyamel 95
in Peruvian selva 172
in southern Peru 159±60 D'Aberon Mission 291
commodity lottery 10, 20 Dawson Accords 98
convertibility regime (Argentina) 286, decentralisation (Brazil) 40±1
290 DõÂaz, Porfirio 131
copper DõÂaz Alejandro, Carlos 124, 265, 284
exports of (Chile) 226 diversification
industry (Chile) 228 in Cuba, 306, 312
industry (Peru) 173±4 Di Tella, Guido 266
taxes on (Chile) 235 domestic market integration 24
Costa Rica in Mexico 120
bananas 93, 95 Dutch (or Spanish) disease 6, 20, 22, 25,
coffee 89±90 220
demography 114±15 in Argentina 285
role of immigrants 92 lack of (Chile) 242±55
liberal reforms 105
railways 92, 93±4 education
state expenditure 104 in Argentina 280
country risk factor 287 in Bolivia 206±9
cotton in Chile 236
sector (Argentina) 271±2 in Peru 158
sector (Brazil) 37 see also literacy
sector (Peru) 167±8 elites 7
credit systems (Central America) 99±100 Brazilian coffee oligarchy 41
Cuba changing character of Mexican elite after
diversification 306, 312, 316, 319 Revolution 135
n 31 control of export coffee economy
effect of Great Depression, 299, 319 (Central America) 92
n 26 relations to the state 29
inflation 304 role in import substitution in
labour 304, 317 n 31 Peru 156
modernization of sugar industry 301, El Salvador, coffee 90±1
303, 310 Encina, F. A. 240
organisation of sugar production 301, Enclave 220
302, 317 n 7, 318 n 14 environmental disaster (Peru) 176
railways 302, 303, 312, 317 n 6, 317 ErraÂzurriz, Maximiano 218
n9 Estrada Cabrera, Manuel 85
relations with USA 299±300, 302, exchange rate policy
303±4, 305±6, 311, 313±15, 316, 319 debates in Bolivia 196
n33, 319 n 34 differential rates (Bolivia) 203
social and political conflict 308, 309, in Argentina 284±5, 285±7
314, 315, 316, 319 n 24 in Brazil 48±50
state regulation 305, 307±9, 312 in Chile 240±1
326 Index
financial system 7 Indians 10, 90±1, 121, 170, 179, 181 n 14,
weakness of Peruvian 157 181 n 15, 181 n 17
First World War 1, 2, 11, 74, 77, 122, 224, industrial revolution 3
265 industrialization 56, 65±6, 75±81, 82
effect on Argentine growth 269, 288±90 n 10, 82 n 12, 82 n 13, 77±9, 126, 128,
effect on Bolivian tin 192±4 132, 145 n 56, 156, 180 n 8, 181
effect on Cuban sugar production 300, n 11, 203, 227, 256, 267, 269, 281±2,
303, 306 293
effect on Peruvian cotton 168 inflation
effect on Peruvian wool trade 161±2 66, 132, 203±4, 241, 260 n 35, 260 n 36
foreign debt institutions 3, 5, 6
in Bolivia 201, 205 rigidity in (Argentina) 267
in Central America 99 changes in (Argentina) 290±2
in Chile 217, 235 instability 20±1
in Peru 158 fiscal in Colombia 66
foreign investment monetary in Central America 97
in Argentina 283 political in Central America 115
in Bolivia 194 of Chilean export sector 232±3
in Cuba 304±5 internal migration
in Mexico 126 in Central America 101
in Peru 163, 168, 174 in Colombia 63
frontier 266±7, 271, 293 in Peru 165
GarcõÂa MaÂrquez, Gabriel 71 Janneret, Teresa 240

geography 33, 55±6, 86 JimeÂnez, Ricardo 109, 113
Grace contract 154 Joint Stock Companies 3
Great Depression, 1, 11, 189
gold standard 5, 98±9, 156, 196, 238, Keith, Minor 93
285±7 Kemmerer Mission (Col.) 70
guano 153±4, 222 Keynesian multiplier 220, 245
Guatemala
bananas 94 labour
coffee 90 decline of handicraft employment
role of immigrants 92 (Colombia) 79
liberal reforms 104±5 elasticity of supply (Brazil) 38±9
railways 92 free labour market 6
state expenditure 104 growth of movement
GutieÂrrez Guerra, ReneÂ 198 (Argentina) 290±1
growth of workforce in Argentina 276
haciendas 10, 129, 145 n 47 impact of Mexican Revolution on
Hirschman, Albert 210, 220 coercive labour systems 134
Hochschild, Mauricio 198 impact of rationalization on mining
Honduras workforce (Peru) 176±7
bananas 94±5, 112±13 improvements in Mexican working
coffee 91±2 conditions 136
liberal Reforms 105 in favour of free trade (Argentina) 288
legislation in tin industry 195
Ä ez, President 236
Iban peasant uprisings 18±19
immigration 39, 52 n 15, 92, 121, 164±5, proletarianisation in the Peruvian sugar
168, 276 sector 164
imports 101±2, 217, 233, 238±9, 245, protests (Cuba) 309
247, 260 n 32, 282 real wages (Argentina) 278±80
Index 327
role of women and children in labour living standards 126, 128, 130
force (Colombia) 80±1 markets 120
scarcity in tin mining during Chaco oil boom 132
War 198 political instability 121±2, 125, 131
scarcity of (Cuba) 301 population 132
sharecropping (Peru) 167 Porfirian regime 126±30
use of coercion in Central railways 120, 125±6, 127±8, 142 n 8
America 90±1 regional fragmentation 119, 121,
use of Jamaican workers in Central 128±9
America 95±6 relations of production 128±9, 134±5
Lamont, Thomas 133 role of state 136±7, 147 n 77
land similarity to Andean economies 121
changes in use of (Argentina) 270±1 size of state 119±20
ownership of 7 social conflict 135
usurpation of peasant (Peru) 161 social reform 136
liberalism 89, 105, 127, 189, 291 structure of foreign trade 133
linkages, 20, 22±3 technological innovation 126
absence of forward linkages in coffee and timing of export boom 124±5
banana production in Central unionization 135±6, 145 n 50
America 96±7 US influence 123, 134, 143 n 21
coffee in Brazil 46 wages 130
coffee in Colombia 76 mining 9, 57±9, 173±4, 223±4
from nitrates (Chile) 226±7 Montes, Ismael 206
in Bolivia 209±10 Montt, Manuel 217
lack of in Peruvian oil 169±70 Murtinho Plan 42
rise and fall in copper (Peru) 177 Mussolini, Benito 279
shift from direct to indirect (Chile) 228
theory of 220±2 National Federation of Coffee-Growers
literacy (Col.) 73, 74, 75
in Bolivia 206 Nicaragua
in Central America 115 bananas 94
in Mexico 120 coffee 91
liberal reforms 105
Machado, President 308, 309, 314, 316 role of immigrants 92
manufacturing 23 nitrates
Meller, P. 253 annexation of 218
Mexico demand for Chilean 222
Agrarismo 135 exports of 224±5
Cardenismo 137 production of synthetic 222±3
banks 141 n 4 threat from ammonium sulphate
constitutional reforms 134±5 to 223
domestic agriculture 128 North, John Thomas 224, 230, 252
ethnic makeup 121 Nun Äez, Rafael 65
exports 127±8, 131±2
foreign investment 126, 127, 133±4 oil
hyperinflation 132 in Mexico 123, 132, 33
impact of Revolution 122±4, 130±1 in Peru 169±70
industrialization 126, 128, 132, 145 Orozquismo 131
n 56
labour market 130, 142 n 17 Panama 87, 94
limits of quantitative data 137±41 path dependence 6
literacy 120 Äo, SimoÂn I. 193, 201±3, 205, 209
Patin
328 Index
Peru in Brazil 39
agriculture 180 n 9 in Chile 227
Cerro de Pasco Corporation 174±5 in Cuba 302
commercial systems 159, 168, 172, in Mexico 127±8
175, 179 in Peru 160, 163, 174, 175
cotton production 167±8 RamõÂrez Necochea, H. 240
demonetization 160 Reciprocal Trade Treaty (Cuba) 302±3,
domestic investment 156 308
export boom 1890s 155±6 returned value 20, 113, 177, 229±31
impact of guano 153±4 Reyes, Rafael 55, 67, 70, 76, 77
impact of the War of the Pacific 154 Roca Runciman Agreement 285
industrialization 156, 180 n 8, 181 n 11 Roosevelt, President 315
labour 164±6, 177, 179 Rosa, JoseÂ MarõÂa 273
merchants 160 rubber
oil production 169±70 in Brazil 37
Peruvian Corporation 154 in Peru 170±3
politics 161±2, 179
railways 158, 159, 180 n 6, 186 n 33 Santa MarõÂa, Domingo 219, 245, 250, 252
regional linkages of exports 160, 161, Santos Zelaya, JoseÂ 89, 100
166±7, 169, 175±6, 178±9 Sherman Act 42
regional fragmentation 152, 153 slavery 6, 7, 9, 18, 39, 52 n 15, 57, 301
role of immigrants 163, 168 social indicators 113±16
role of state 158 Somoza, Anastasio 105
rubber production 170±3 Standard Fruit Company 95
shifts in export quantum 157 staple theory 6, 20, 22, 265, 266
social conflict 177, 161, 181 n 17 state, role of 27, 48, 70, 76, 103±4,
sugar production 163±4, 182 n 18 119, 123, 127, 136±7, 158, 180 n 5,
US influence 158, 182 n 20 196±7, 199, 207, 211, 217±18, 231,
urbanization 160, 161, 171 231, 234, 235±6, 248±55, 283, 287±8,
usurpation of peasant land 161, 181 289, 311
n 14 statistics, problems with, 137±41
wool production 159 steamships 9
Phelps, Vernon 290 stock exchange 3
PieÂrola, NicolaÂs 155 sugar
Pinto, President 218 in Brazil 37
Pinto, AnõÂbal 240 in Chile 239
Polanyi, Karl 140 in Cuba, 300±16
Platt Amendment 302, 308, 309 in Peru 163±7
population 33, 63, 71, 81 n 4, 86, 132,
189, 267±8, 276, 283 tariffs 9, 23, 46±7, 65, 76, 155±6, 200,
protectionism, 5, 8, 13, 19, 20, 24 234, 237±40, 284±5, 302, 313, 314
in US 272, 307, 309 taxes 9, 27, 171, 199±202, 234±5, 250,
in Colombia 63±4, 76 251
in Chile 237±8 technological innovation 3, 4, 16±17, 66,
in Argentina 267, 280, 285, 291, 294 103, 162, 163, 167±8, 176, 226, 274,
301, 310, 311
quinine 56 terms of trade 13, 34, 69, 231±2, 270
Thousand Days War 60, 62, 65, 66, 67,
racism 121 70, 71
railroads/railways 9 tin, 188, 190±204
in Argentina 274±4, 280±1 crisis and growth in the 1920s 194±5
in Bolivia 204±6 effect of Chaco War 197±8
Index 329
effect of First World War 192±4 urbanization

effect of Great Depression 195±9 in Brazil 39±40
export quotas 196±9 in Bolivia 199
foreign investment in 191 in Colombia 69
growth in production 190 in Central America 101
impoverishment of deposits 198 in Peru 160, 161, 171
legislation 195 Urmeneta, JoseÂ TomaÂs 218
methods of production 191 United States of America, 3, 13
use of merchant capital 191 relations with Bolivia 192±4
Toro, David 208 relations with Cuba 299±300, 302,
trade unions 303±4, 305±6, 311, 313±15, 316, 319
in Argentina 276, 289, 290±1 n 33, 319 n 34
in Bolivia 209 relations with Mexico 122±3, 134, 143
in Colombia 71 n 21
in Mexico 130, 135±6 relations with Peru 158
transportation 16 .
in Argentina 275, 285 vulnerability 8
in Central America 92
in Chile 241±2 War of the Pacific, 154, 217
in Colombia 62, 63, 69±70, 75 Wittgenstein, Ludwig 141
in Mexico 125±6, 128 wool 163
world trade 11, 14
United Fruit Company 94, 95 Zapatismo 131

[St Antony’s Series] Enrique Cárdenas, José Antonio Ocampo, Rosemary Thorp (eds.) - An Economic History of Twentieth-Century Latin America_ Volume 1 The Export Age_ The Latin American Economies in the Late Nineteen.pdf

Uploaded by

Document Information

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

[St Antony’s Series] Enrique Cárdenas, José Antonio Ocampo, Rosemary Thorp (eds.) - An Economic History of Twentieth-Century Latin America_ Volume 1 The Export Age_ The Latin American Economies in the Late Nineteen.pdf

Uploaded by

Copyright:

Available Formats

St Antony’s Series

General Editor: Richard Clogg (1999– ), Fellow of St Antony’s College, Oxford

Recent titles include:

Customer Services Department, Macmillan Distribution Ltd, Houndmills, Basingstoke,

José Antonio Ocampo

All rights reserved. No reproduction, copy or transmission of

First published 2000 by

List of Tables and Figures viii

Foreword to An Economic History of Twentieth-Century Latin America xi

Notes on the Contributors xii

2 Brazil as an Export Economy, 1880±1930 32

3 Coffee and the Origins of Modern Economic Development in

4 The Economies of Central America, 1860±1940 85

5 Export-led Growth in Mexico, c. 1900±30 119

6 Peru, 1884±1930: A Beggar Sitting on a Bench of Gold? 152

7 Bolivia, 1900±39: Mining, Railways and Education 188

8 Trying to `Tax and Spend' Oneself out of the `Dutch Disease':

9 The Vicissitudes of an Exporting Economy: Argentina, 1875±1930 265

10 Alteration, Crisis and Adjustment in the Cuban Export Economy,

1.1 World economic growth, 1820±1929 3

6.3 Peru: composition of exports by value, 1890±1930 157

1.1 Net barter terms of trade between primary products and

8.7 Chile: real rate of exchange, 1865±1900 244

Marcelo de Paiva Abreu is Professor of Economics at the Department of

Afonso S. Bevilaqua is a member of the Department of Economics, Catholic

Enrique Ca Ârdenas is Professor of Economics and Rector at Universidad de las

Roberto CorteÂs Conde is Professor, Department of Economics, Universidad

Manuel E. Contreras is Dean of the Graduate School of the Universidad

Paulo Drinot is a doctoral candidate in Modern History at St Antony's Col-

Alan Knight is Professor of the History of Latin America at the University of

JoseÂ Antonio Ocampo is currently the Executive Secretary of the United

Gabriel Palma is Lecturer in Economics at the University of Cambridge.

HeÂctor PeÂrez Brignoli is Professor of History at the University of Costa Rica.

Rosemary Thorp is Reader in the Economics of Latin America at the Uni-

foreign investments in the region. International migration was intensive and

The development of an international economy

Table 1.1 World economic growth, 1820±1929 (%)

1820±70 1870±1913 1913±29

Trade and factor movements also needed their institutional support. As

More important than that, national unity remained problematic in at least

A third group of preconditions relates to preexisting rural structures. Wage

1859±61 1899±1901 1911±13 1927±29

Table 1.3 Commodity composition of Latin American exportsa (percentage shares of

Products 1859±61 1899±1901 1911±13 1927±29

Meat 1.1 4.1 5.0 6.7

Total 100.0 100.0 100.0 100.0

facilitated by protectionism, affected all sugar-cane economies in the last dec-

130 Grilli and Yang (1988)

reported in Table 1.2. Individual commodities often experienced quite different

The productive structures of export sectors

a slash-and-burn agriculture was itself a major technological innovation, as it

macroeconomic consequences were the instability of fiscal revenues, given the

shifts which export expansion generated. Two analytical frameworks compete

The development of a modern transport infrastructure was one of the most

Exports GDP Rail ways % of labour Literacy Life-

Argentina 67 188 4.4 34.2 63 46

closely associated with differences in export performance. Argentina, Chile,

The role of the state and the political economy of export

economies, but fiscal dependence on such taxation became common only in

Conventionally, the year 1930 is taken as a watershed in Brazilian history. It

Table 2.1 Brazil: GDP shares, 1910±93 (%)a

Year Agriculture Industry Services

1910 35.8 14.0 50.2

then spreading in the Paulista plateau; Minas Gerais became an important