Benito Arruñada (Auth.) - The Economics of Audit Quality - Private Incentives and The Regulation of Audit and Non-Audit Services-Springer US (1999)

The Economics of Audit Quality
The Economics of Audit Quality
Private Incentives and the Regulation

of Audit and Non-Audit Services
by
Benito Arrufiada
Pompeu Fabra University, Barcelona
SPRINGER SCIENCE+BUSINESS MEDIA, LLC

Library of Congress Cataloging-in-Publication Data
ISBN 978-1-4419-5082-6 ISBN 978-1-4757-6728-5 (eBook)

DOI 10.1007/978-1-4757-6728-5
Printed on acid-free paper
All Rights Reserved

© 1999 Springer Science+Business Media New York
Originally published by Kluwer Academic Publishers, Boston in 1999
No part of the material protected by this copyright notice may be reproduced or
utilized in any form or by any means, electronic or mechanical,
including photocopying, recording or by any information storage and
retrieval system, without written permission from the copyright owner.
To Marisa with gratitude for her love,
support and patience
Contents
1 Auditing quality 3
1.1 The demand for financial auditing 3

1.1.1 Financial auditing as a contractual facilitator 4
1.1.2 Empirical evidence 7
1.2 The dimensions of quality 8
1.2.1 Technical competence and independence 8

1.2.2 The importance of professional judgement and the danger of
inducing "defensive auditing" 9
1.3 The conflict with third parties and among clients regarding auditor
independence 15
2 The safeguard of quality 19
2.1 General analysis of safeguarding mechanisms 20
2.1.1 Explicit and implicit safeguards 20

2.1.2 The role of quasi-rents in the safeguard of quality 20
2.1.3 Formal analysis of quality assurance 21
2.2 Strategies adopted to create quality-safeguarding incentives 24
2.2.1 Low-cost strategies 25
v
VI Contents
2.2.2 High-cost strategies 32
2.3 The "last period" problem and its consequences for auditing quality 35
2.4 The regulation of quality: Attainable objectives 37

2.4.1 Independence failures in professional practice 37
2.4.2 Regulation as a facilitator of judicial and market controls 38
2.4.3 Fragmentation versus safeguard regulatory strategies 40
3 Auditor independence as an economic decision 43
3.1 Concepts of auditor independence 43
3.2 Economic factors behind auditor independence 44
3.3 The cost of independence: the loss of client-specific assets 49

3.3.1 Volume of quasi-rents to be received by the auditor 49
3.3.2 Specific assets caused by excess capacity 53
3.3.3 Probability of auditor switching 54
3.3.4 Value of client-specific assets when only auditing services
are provided 55
3.4 The costs of dependence 57

3.4.1 The loss of quasi-rents associated with current clients 58
3.4.2 The loss of quasi-rents linked to potential clients 59
3.4.3 Losses related to professional and criminal liability 60
3.4.4 The role of risk aversion 64
4 The supply of non-audit services by auditors 69
4.1 Definition and types of non-audit services: the importance of a

general analysis 69
4.2 Economies of scope 71
4.2.1 Auditing technology and firms' complexity 71

4.2.2 The supply of non-audit services as a natural consequence of
modem auditing 73
4.2.3 Origin and types of economies of scope 74
Contents Vll
4.2.4 The existence of diseconomies of scope 79
5 Effects of non-audit services on auditor independence 81
5.l Effects of non-audit services on the economic incentives for

independence in fact 81
5.1.1 Higher quasi-rents linked to each client make diversified
auditors more independent 82
5.1.2 The increase in firm-specific assets favors independence 85
5.1.3 Other effects 86
5.1.4 Empirical evidence regarding the effects of the supply of
non-audit services on independence in fact 88
5.1.5 The reaction of audit clients 91
5.2 Effects on the appearance of independence 92

5.2.1 Studies based on surveys 92
5.2.2 General limitations of the studies based on surveys 96
5.2.3 The correlation between supply of services and qualified
auditor opinion 98
5.3 Private safeguards of independence with respect to the supply of

non-audit services 98
5.3.l Self-regulatory safeguards 99
5.3.2 Organizational safeguards 101
5.3.3 The federation of firms 104
6 Effects of non-audit services on market competition 109
6.1 The argument regarding auditing as a "loss-leader" 109
6.2 Doubts as to the prevalence of introductory pricing 111
6.3 The efficiency of introductory pricing 113

6.3.1 Introductory pricing as a consequence of the competition for
quasi-rents 114
6.3.2 The implausibility of predatory behavior through cross-
subsidization between different kinds of service 120
Vlll Contents
6.4 Elements for a positive analysis of the provision of non-audit

services and the prohibitionist proposals 122
6.4.1 Consequences for ftrms specialized in auditing 122
6.4.2 Consequences for providers of other services 124
6.4.3 Towards a new equilibrium 127
7 Normative possibilities for non-audit services 131
7.1 The public debate 131

7.1.1 The initial debate in the United States 131
7.1.2 The eclecticism of Cadbury and other reports on corporate
governance 133
7.1.3 The opinion of sector reports 134
7.1.4 The European Union Green Paper 137
7.2 Analysis of the main regulatory possibilities 139
7.2.1 Rules forbidding or restricting the supply of non-audit
services 140
7.2.2 Rules facilitating the work of the market and self-regulation 143
7.2.3 Minimum diversiftcation rules 143
7.2.4 Disclosure rules 148
8 Let knowledge evolve in the market 159
8.1 Private quality safeguards and the role of regulation 160

8.1.1 Dimensions of auditing quality which are crucial for
regulation 160
8.1.2 Market-based quality safeguards and the role of regulation 163
8.2 The provision of non-audit services: Consequences and regulation 165

8.2.1 Consequences of the supply of non-audit services 166
8.2.2. The regulation of non-audit services 171
References 177
Index 191
Foreword
This book focuses on market mechanisms which protect quality in the provision of
audit services. The role of public regulation is thus situated in the context defmed
by the presence of these safeguard mechanisms. The book aims to contribute to a
better understanding of these market mechanisms, which helps in defining the con-
tent of rules and the function of regulatory bodies in facilitating and strengthening
the protective operation of the market.
An analysis at a more general level is provided in the three chapters making up
Part 1. In the four chapters of Part 2, on the other hand, this analysis is applied to a
particular problem to determine how those non-audit services often provided by
auditors to their audit clients should be regulated. Finally, Chapter 8 contains a
summary of the analysis and conclusions of the work. The conclusion with regard
to non-audit services is that their provision generates beneficial effects in terms of
costs, technical competence, professional judgment and competition and, moreover,
need not prejudice auditor independence or the quality of these services. This as-
sessment leads, in the normative sphere, to recommending a legislative policy
aimed at facilitating the development and use of safeguards provided by the free
action of market forces. Regulation should thus aim to enable the parties-audit
firms, self-regulatory bodies and audit clients-to discover through competitive
market interaction both the most efficient mix of services and the corresponding
quality safeguards, adjusting for the costs and benefits of each possibility. Particu-
lar emphasis is placed on the role played by fee income diversification and the
enhancement through disclosure rules of market incentives to diversify.
This book has benefited greatly from the help and comments received from nu-
merous colleagues. In particular the stimulus given by Jiirgen G. Backhaus has
been fundamental in writing and publishing it in its present form. With his usual
perspicacity and enthusiasm, Luis Garicano discussed the mechanisms for safe-
IX
x Foreword
guarding quality and normative proposals, making a substantial contribution to

better understanding and refinement of the arguments involved. Several earlier
collaborations with Candido Paz-Ares to a large extent assisted in comprehension of
the role of the regulatory framework in its constant interaction with the private
activity of economic agents. Neither can I overlook initial steps in relation to the
role played by the existence of specific assets and quasi-rents in audit quality, taken
in collaboration with Jesus Lozano, nor the comments of other colleagues based on
various public presentations of other works in some way connected with this one,
particularly William J. Carney. In addition, numerous participants at conferences
(The University of Barcelona School of Law-CIDDRIM, The Catalonian Associa-
tion of Public Auditors), seminars (The Carlos III and Pompeu Fabra Universities)
and scientific congresses (12 and 14 Annual Conferences of the European Associa-
tion of Law and Economics, the 1996 International Symposium on Audit Research,
the 21 Annual Congress of the European Accounting Association) at which I have
had the opportunity to present and discuss various parts of the study have also sug-
gested and put to the test many preliminary ideas, thus helping to improve them.
Anna Cano, Georgina Folguera, Manuel Gonzalez, Urie1 Gonzalez-Montes and Luis
Vazquez have, at different stages of the work, generously helped me in compiling
and processing the different bibliographical and empirical materials used. I would
like to express my most sincere gratitude to them all. Needless to say, none of these
individuals has any responsibility for the errors and value judgments which the
work surely contains, for which the author alone bears sole responsibility. Finally,
my thanks and appreciation to Bill Harrison, M.A. (Oxon.), Solicitor and translator
from the international Legal Language Services group of lawyer-linguists, for his
work in translating the manuscript into English and, at one stage, pointing out to me
that Japan is not yet a member of the European Union!
Part 1
THE DEMAND FOR QUALITY
AND ITS SAFEGUARD IN
AUDITING
1 AUDITING QUALITY
As with any other goods or services it is appropriate to examine the quality of

auditing services from the point of view of their value to those who make use of
them. These are both direct users-those who purchase them-and indirect but
perhaps the more important users-those for whom audited accounts are destined.
This latter group includes the shareholders in and creditors of organizations whose
accounts are audited as well as their customers, employees and the public bodies
with whom they deal. I In some cases such users employ the accounts to monitor the
conduct or performance of those taking decisions in the organizations being audited.
In other cases, however, they use the accounts in a prospective manner as an in-
strument to assist them in taking decisions, whether direct investment, lending,
purchasing or employment.
1.1 The demand for financial auditing
In both cases, economic agents acquire aUditing services to increase the informative
value of accounts in the hands of those using them. As a starting point in the analy-
sis, it will therefore be assumed that the clients of a fIrm of auditors desire good
quality auditing. In particular they thereby reduce the transaction or agency costs
that they incur in their contractual dealings (mainly those with creditors and share-
holders); in short, they achieve better contractual terms in their exchanges. 2
I In this respect the classification made by the Auditing Practices Board is interesting in establishing
decreasing degrees of auditor liability in relation to three sets of stakeholders: primary (shareholders
and regulators), secondary (creditors and employees) and tertiary (potential investors, intermediaries,
tax authorities, etc. not included as primary stakeholders). See APB (1994, pp. 25-6).
2 Despite the difficulties inherent in observing this type of variable, the effect of audit quality in im-
proving contractual terms has been quantified in some specific situations. For example, Balvers,
McDonald and Miller (1988) and Beatty (1989) found that the discount typically associated with public
share offers granted by companies seeking a stock exchange listing for the first time was lower when
their accounts had been audited by prestigious firms. Balvers et al. (1988) also reveal that the effect is
3
4 The economics of audit quality
1.1.1 Financial auditing as a contractual facilitator
Auditing is generally understood in terms of Agency Theory as a type of bonding

contracted by an "agent" management team which does so in relation to its own
shareholders in large companies with specialized ownership and control, or the
proprietor of a business in relation to his creditors and/or partners. 3 Reference will
be made in this work to the auditor's "clients" in a generic sense. This concept of
client comprises both the currently predominant circumstances in which the client is
an agent who, by means of the audit, wishes to bond his activities, and those of a
principal who wishes to monitor an agent. This duality is the result of the fact that
although auditors are usually contracted by a company itself in legal terms, the
economic relationship is more complex and the contracting party can be considered
as being either the management of the company or its shareholders depending on
the circumstances. In virtually all countries auditors contract with the company
itself, the contract being approved by the shareholders' meeting on the proposal of
the directors. 4 The existence of an audit committee may however modify this allo-
cation of roles, converting the audit into a monitoring rather than bonding
mechanism, in which the contractual initiative comes from the prinCipals, particu-
larly shareholders when they are represented on a relatively independent audit
committee. 5
complementary to that produced by the fact that the operation is handled by a merchant bank of high
standing. In a more recent study, Hogan (1997) shows that the owners of businesses must choose
between accepting greater underpricing-i.e., a higher discount on the share price-and providing
increased guarantees by means of an auditor of higher standing.
3 In order to facilitate the study of contractual relationships Agency Theory makes an abstraction by
considering a single relationship between two individuals in which, moreover, only one of them, the
"agent", is obliged to provide some service to the other, who is usually referred to as the "principal". In
this type of abstract contractual relationship it is appropriate to classify the contractual costs and safe-
guards on the basis of which of the respective parties initiates and pays for them. In an agency
relationship all the costs aimed at reducing a deviation in the agent's performance from the interests of
the principal are referred to as "agency costs". Three categories are usually distinguished: (I) "moni-
toring costs", which are those paid for by the principal, (2) "bonding costs", paid for by the agent, and
(3) a certain "residual loss", which is the result, even with optimum levels of monitoring and bonding,
of the fact that it is usually preferable that a certain deviation persists in the conduct of the agents from
what would be optimum in the hypothetical or ideal case in which there are no contractual costs. For an
introduction to the theory, see the classic work of Jensen and Meckling (1976).
4 Auditors were appointed by a government body in Korea until the 1980 reforms (Park, 1990). In
many countries a similar mechanism is used for appointments by the courts and the Companies' Regis-
try. A lottery mechanism has also been proposed on some occasions (see, for example, Cinco Dias, 17
October, 1995).
5 It has nevertheless been argued that companies may create audit committees purely for image pur-
poses (Bradbury, 1990). The study by Menon and Williams (1994) shows that audit committees
voluntarily created by US companies have in general had a very inactive life, although their activities
increase with the presence of external (non-management) members on the board of directors. This
outcome seems logical; the audit committee can operate as a body which assists the board in its moni-
toring work but this function only makes sense if the board itself also acts as a monitoring body and is
Auditing quality 5
Demand for external auditing is directly related to the fact that it is the directors
or management (when considering the relationship with shareholders in the context
of a company with specialized ownership and control), or debtors (in the context of
credit relationships) themselves who prepare the accounting information. This
leading role of the agent or "party to be monitored" in producing the information
used to monitor his own conduct is based on cost reasons-a large part of such
accounting information is used internally (and this is perhaps its primary function)
in decision-making and internal control by management itself. In this context, the
need arises for an external verification mechanism which to some extent ensures the
reliability of and reinforces confidence in such accounting information. 6 The audi-
tor is a specialist who examines the accounts in order to verify that they meet a
series of requirements and give an assurance thereof to third parties backed up by
the guarantee of his reputation. From this point of view the client can be taken in a
certain way to acquire or hire the reputation of the auditor in order to improve his
contractual opportunities. 7
This argument, which treats demand for auditing as a result of the need to safe-
guard contractual conduct, is usually situated in a context of "moral hazard",8 in
which the audit contributes to making the degree of contractual performance ob-
servable ex post and therefore provides an incentive for it. The audit can also be, on
a complementary basis, a remedy for problems of ex ante contractual asymmetry
to some extent independent of management. (There is another explanation, however, related to the
desire of external directors to be seen to be independent, particularly in relation to any possible liability
suits ).
6 This view of auditing is derived from Jensen and Meckling (1976, pp. 338-9). Auditing is obviously
not the only safeguard nor perhaps the most important. Auditing services can in fact be replaced by
other monitoring mechanisms. Anderson, Francis and Stokes (1993) examined such replacement by
empirically analyzing the relationships which exist in Australian companies between external audit
fees, salaries paid to internal auditors and the overall remuneration of the board of directors. One of
their results shows that as the percentage of the value of a company of a less tangible and therefore
more conflictive nature increases so also does the cost of auditing in relation to board remuneration.
7 One of the principal creators of modern organizational economics views thus the role of auditing as
hiring reputation with the safeguard of the reputation of the auditor himself: "owners win the confi-
dence of investors by renting the reputation of auditors for accuracy and fairness. Auditors earn a
market rate of return on their investment, via the costs of quality audits, in building a reputation among
investors" (Wilson, 1983, p. 305). This rental of reputation is valuable because the client only needs it
occasionally (particularly when having recourse to capital markets) whilst the auditor is permanently
present as a supplier of contractual safeguards. The auditor's role as specialist who stakes his reputation
when backing part of the activities of his clients is similar to that of other specialists, such as merchant
banks (Gilson and Kraakman, 1984), underwriters of security issues (Downes and Heinkel, 1982; Booth
and Smith, 1986), debt rating agencies (Wakeman, 1981) and insurance companies to the extent that
they also monitor certain assets (Mayers and Smith, 1982).
8 In the economics of contracts moral hazard problems arise as a consequence of post-contractual
information asymmetries leading to performance failures.

(ex ante and ex post meaning here before and after contracting, respectively),9 sig-
naling the good quality of the audited business. This is particularly appropriate in
circumstances when the firm undertakes a major reorganization of its contractual
structure, as exemplified by going public on the Stock Exchange.
The existence of a public or regulatory demand for auditing does not diminish
the relevance of an analysis based on voluntary demand. lo This is so for several
reasons. Firstly, it is clear that there is a demand for auditing beyond the regulatory
minimum (in terms of quantity and particularly quality), a minimum which could be
achieved at a much lower cost. II Companies that choose auditors of proven quality
are in the majority-at least in terms of value-despite the fact that the minimum
legal requirements could be covered at a lower cost using auditors with lesser repu-
tations. A substantial demand of a purely voluntary, contractual origin can also be
seen even after the introduction of mandatory auditing.12 A substantial part if not
the majority of audits of large companies obviously also form part of contractual
demand as evidenced by the fact that many of them submitted their accounts to
external auditors before it became mandatory to do so.
rt is true that voluntary demand for auditing has historically been weak in conti-
nental Europe and Japan. Nevertheless, it is doubtful whether it would also be weak
now on the hypothetical basis of no mandatory auditing legislation. This doubt is
9 See, for example, Titman and Trueman (1986).

10 Mandatory auditing has been established by the domestic legislation of EU member states on the
basis of different Community directives (mainly the Fourth Directive, 78/660/EEC, of 25 July 1978,
which has already been implemented in all member states). (See in this respect Buijink, Maijoor,
Meuwissen and Witteloostuijn, 1996, pp. 25-31). Except for some countries in which all companies are
obliged to have their accounts independently audited (Finland, Ireland, Norway [not a member of the
EU) and Sweden), in the remainder the rules only make it obligatory when they reach a certain size.
Mandatory auditing and a more detailed disclosure of financial statements by listed companies was
adopted in the United States by the Securities Act of 1933. The adoption of mandatory auditing in the
US was explicitly supported by the auditing profession and not opposed by firms, with scant discussion
by the legislature (SEC, 1994, pp. 5-6). This has also been the case in many European countries. The
etTects of the rule are not clear, however. For empirical attempts to assess the consequences of the
Securities Act, mostly focusing on stock exchange prices, see Simon (1989), Chow (1983), Jarrell
(1981), Benston (1973) and Stigler (1964).
11 The birth and historical evolution of the profession is testimony to the extent of voluntary demand for
auditing, which in greater part took place outside the framework of legal obligations. This was par-
ticularly the case in Britain where much nineteenth century company legislation did not make it
obligatory to carry out audits or make it necessary for them to be carried out by professionals. At the
end of the nineteenth century, when the 1900 Companies Act made auditing generally mandatory for
companies listed on the stock exchange, virtually all accounts of such companies were already being
audited by professional chartered accountants. For a discussion and several historical references, see
Watts and Zimmerman (1983, pp. 628-9), who demonstrate how the auditing of accounts came about a
long time before any type of legal or regulatory obligation and appears as far back as the thirteenth
century.
12 For instance, after six years of mandatory audits of large and medium-sized companies, voluntary
audits accounted for 23.01 percent of the total audits carried out in Spain in 1995 (Boletin Oficial del
[CAC, no. 25, 1996, p. 38). Moreover, this voluntary demand has been growing since 1992.
Auditing quality 7
based on the notion that circumstances have radically changed. Low historical
demand was probably due to traditional features of continental fmancial systems,
such as weak competition, close links between banking and industry and cross
shareholdings. It is plausible that when these circumstances changed there would
have been growing demand for safeguarding services. Moreover, in the years prior
to mandatory auditing a substantial increase could be seen in demand amongst large
companies for auditing services which at that time were largely voluntary and
probably based more on a desire to provide positive solvency signals than to insure
against future contractual hazards.
Furthermore, it might be argued that for these purposes the position of private
users of accounting information and that of public users are largely comparable.
Certainly, the latter cannot protect themselves by increasing the explicit price of
their contribution to the company. They can however increase their regulatory and
monitoring work in relation to it, both in fiscal terms and in complying with other
types of legislation. This is the case, for example, with inspections of financial
institutions in respect of which this same argument has been used to defend the
conventional deposit insurance system based on a fixed explicit premium independ-
ent of the level of risk.
1.1.2 Empirical evidence
The contractual costs of businesses vary substantially depending on their structure.

Consequently, they tend to demand auditing of differing quality. This is a demand
derived from that of their contracting parties, who are usually more specialized the
more complicated the contractual structure of the business (exemplified by the sepa-
ration-specialization-of ownership and control inherent in going public on the
Stock Exchange). It should be noted that regulatory demand also has an interest in
audits of differing quality, increasing with the scope of the external effects of the
business, which increase not only with its size but also through involvement in
particular activities (such as banking, for example). It is therefore understandable
that mandatory auditing legislation covering all types of company of a certain size is
usually preceded by mandatory auditing legislation of a specific and/or sector na-
ture. 13
Many empirical studies support the theory according to which audit firms pro-
vide differing quality (increasing with the size of the firm and highest, in particular,
13 This was the case in Spain, for example, where financial institutions were obliged to audit their
accounts a long time before a general obligation was laid down by Act 1911 989.
in the case of the "Big Five" frrms)14, in response to a varying demand for quality
amongst clients. The explanation lies in the fact that different companies have
fmancial and contractual structures with a highly varying degree of potential con-
flict, in short with different levels of agency costs. As a result, they also have a
demand for different degrees of quality in auditing their fmancial statements.
A large number of empirical works have in fact verified the existence of a posi-
tive relationship between different variables which are connected to the intensity of
agency costs and the choice of higher quality auditors. In a retrospective analysis,
Chow (1982) showed that in 1926, prior to legislation in the United States making
auditing obligatory, voluntary demand for auditing amongst American companies
depended on different indicators which it seems reasonable to believe were directly
related to the degree of conflict between management, shareholders and creditors.
Such indicators included the proportion of shares not owned by management, levels
of indebtedness, the size of the business and the number of clauses linked to ac-
counting information in loan agreements. Other empirical studies have
demonstrated the significance of agency cost variables on change of auditor, re-
vealing the expected effect even after taking account of varying client size and
growth, the variables considered to have the greatest influence on change of auditor
decisions. 15 This is also shown by the fact that most companies which go public on
the Stock Exchange (and therefore increase potential contractual conflict) switch
auditor to one of the large frrms. Carpenter and Strawser (1971) documented this in
the United States in a classic study. It has furthermore been shown how companies
going public are more likely to obtain audit services from larger firms the greater
their agency costs (Firth and Smith, 1992).
1.2 The dimensions of quality
1.2.1 Technical competence and independence
As an initial approximation to understanding the meaning of "quality" in auditing

services let us assume that these services concern a particular set of data and that the
information which the auditor needs to use in issuing his opinion is also fixed. 16 In
14 Empirical works in this respect particularly include that of Craswell, Francis and Taylor (1995),
which restates and improves on several preceding works, showing that part of the price differential
previously considered as a "reputation premium" (see Section 2.1.3 in this respect, below) should in
fact be attributed to the investment required to produce specialized aUditing. The interested reader will
find references to these works in Craswell el af. (1995, p. 298, n. 2).
15 See, in particular, Francis and Wilson (1988), as well as Defond (1992).
16 Section 1.2.2 deals with a more complex scenario in which the auditor can use a more or less exten-
sive series of variables on which to base his opinion, in which event the regulatory system can have a
decisive effect by favoring or preventing the use of qualitative indications which are not verifiable by
Auditing quality 9
these conditions, the quality dimensions are essentially technical competence and
independence. Technical competence is defined as the auditor's ability to detect
errors or shortcomings in the fmancial statements being checked. Independence on
the other hand is taken to be the willingness of the auditor to reflect in the audit
report all problems and defects he has detected in the financial statements. In prob-
ability terms, technical competence can thus be defmed as the probability of
detecting defects in financial statements and independence as the probability of
reporting them once they have been detected. I7 Auditing quality is thus defined by
the combined probability that the auditor will detect and report on defects in ac-
counts. I8
These quality dimensions fit into the typology classifying different goods and
services (or rather their different attributes) on the basis of the greater or lesser
capacity of the client to evaluate their quality. From this perspective auditing pres-
ents characteristics of "experience goods" and "credence goods" rather than "search
goods". 19 As with experience goods, the purchaser of auditing services is only
aware of some quality attributes after acquiring them: mainly technical competence.
Similarly, as with credence goods, the independence dimension can take even
longer to ascertain. This is so, in particular, when there are no conflicts between the
auditor and the client or when the auditor behaves independently with respect to
that client but not with respect to others (any lack of independence by the auditor in
relation to one client could easily take time to be ascertained by others).
1.2.2 The importance of professional judgement and the danger

of inducing "defensive auditing"
The view provided in the previous section whereby only the technical competence
and independence dimensions of quality are taken into account, presupposes that the
auditing relates to and in his assessments and judgement the auditor only uses a
third parties or, on the other hand, restrict the auditor to just using indicators which can be verified (in
what can be known as "defensive auditing"), which safeguards him from any potential liability in the
event of future litigation.
17 It should be pointed out that this definition of auditing quality in terms of technical competence and
independence provides a useful breakdown for analysis since it defines two relatively distinct problems
and therefore enables specialized rules and practices to be designed to support each of them. The two
dimensions are not totally independent, however, as emphasized by the fact that a lack of independence
can be shown by decisions which reduce effective technical competence. This would happen when the
auditor decides not to make an effort to discover problems which he does not wish to report on.
18 This definition of quality was developed from the so-called Positive Accounting Theory. See, in
particular, Watts and Zimmerman (1980, p. 8; 1986, pp. 314-5) and DeAngelo (l98Ib, p. 186).
19 The concepts of "search" goods and "experience" goods were formulated by Nelson (1970) and that
of "credence" goods by Darby and Kami (1973).
fixed certain volume of information. The problem becomes more complicated

when taking into account that in his work the auditor may use different types of
information, which vary in terms of third party verifiability.
1.2.2.1 The informative content of auditing
From the point of view of informative content the auditor should use, and his report
therefore convey to the market and end users, as much information as possible. The
safeguards required for using different types of information, however, vary de-
pending on the degree to which the information can be verified. Moreover, the
auditor himself will have reservations in using information which may be of little
use in justifying his opinion in a future dispute.
A superficial understanding of auditing functions and, in particular, overlooking
the need for the auditor's opinion to incorporate his professional judgement, often
means that the problem of auditor independence is treated in a very narrow manner.
This happens when attempts are made to further independence by increasing legal
sanctions since such policies mean that independence in fact becomes limited to
mere compliance with a series of verifiable criteria. This superficial approach is
deceptive and its application in the legislative field may be responsible for reducing
the value of auditing. The reason is that when the legal system makes auditors liable
to a disproportionate extent, there is a risk of trivializing independence. Borrowing
a term coined in the medical field, the result could be catalogued as "defensive
auditing":20 the auditor decides to play safe, reporting any possible doubt which
may be raised by the accounts produced by management on the basis of verifiable
criteria. 21 In this situation, the auditor is constrained to base his decision only on
information which can be judicially verified (the only information which those who
consider independence to be an absolute value mistakenly seem to acknowledge as
valid). He thus dispenses with that information regarding the company which is not
verifiable since such information is useless as a defense in the event that the com-
pany ends up with unforeseen financial problems which generate litigation with the
auditor.
20 "Defensive medicine" is understood to be the administration of all types of treatments with very low
expected benefits but which are administered in any event to avoid possible legal liability.
21 Such consequences of the liability system are far from trivial, as shown by the results of research by
the National Bureau of Economic Research which in 1996 reliably quantified its effects on "defensive
medicine" in the United States (Kessler and McCellan, 1996). Various direct reforms of the professional
liability system (maximum limits, abolition of punitive damages, calculating interest from the date of
judgement only, reducing compensation by sums obtained from other sources) enabled medical ex-
penses to fall by between 5% and 9% without adverse consequences in terms of mortality or medical
complications. Another indication of the importance of the matter is provided by the fact that the
benefits obtainable by reducing liability are estimated at 600 million dollars per year simply in relation
to myocardial infarction costs.
Auditing quality 11
From the company's point of view it is not generally optimum for auditors to be
merely independent. What is required of auditors is that they exercise their profes-
sional judgment independently.22 It would be prejudicial however if, in order to
preserve his independence, the auditor is obliged to refrain from making a profes-
sional judgement since this provides valuable information to those using the
company's accounts. This, broadly speaking, is the argument as to the relative
nature of auditor independence in terms of information economics propounded by
Grout, Jewitt and Whittington (1994). In developing their argument they make a
fundamental distinction in modem information economics literature and in the field
of incomplete contracts in particular between information which is merely observ-
able by the parties (in this case the auditor) and that which is also verifiable by third
parties (here, the legal system).23 Grout et al. consider that the auditor can use two
types of information in his professional judgement: what they call "hard" informa-
tion, which is observable and verifiable, and "soft" information, which is observable
but not verifiable. Specifically, they assume (although similar conclusions can be
reached on more restrictive assumptions) that in the case of auditing hard informa-
tion can be suppressed from the accounts but not equivocally transmitted whilst soft
information can be the subject of equivocation and subjetiveness in transmission.
In order to maximize the value of auditing, the regulatory framework must make
it possible for auditors to use and transmit this soft information regarding the client
to the market and third parties. To this end, he must exercise his professional
judgement based on both categories of information. If the regulatory framework
penalizes him excessively, he will only make use of the hard information. Conse-
quently, the audit will contribute less information and may sometimes even be
misleading. The latter may occur when the auditor is obliged to play safe and base
his opinion on hard information. He would then, for instance, disclose his reserva-
tions even when he considers that the client's situation would not warrant such
reservations after taking into account the soft information available to him regard-
ing the client, information that he is not in fact allowed to transmit. 24
22 This requirement is highlighted in the report on good government of companies prepared by the
Centre for European Policy Studies (CEPS, 1995, pp. 44-5).
23 Grossman and Hart were pioneers in differentiating between observable and verifiable information in
their article on vertical integration (1986).
24 Evidence regarding auditor switching provides some indirect indication that different economic
agents have access to and can verify varying types of information. Krishnan (1994) found that a change
of auditor is more likely when the auditor gives a qualified opinion applying apparently conservative
criteria in relation to the observable financial position of the business. The results of another study
(Krishnan and Stephens, 1995) are consistent with the following explanation: the first auditor has
private information which leads him to be conservative and the second also has access to this informa-
tion and therefore does not modify his opinion compared with that of the former auditor. Obviously, he
also cannot previously commit himself to issuing an unqualified opinion given his lack of knowledge of
Using this type of distinction, Grout et al. argue that negotiation between auditor
and client as to the content of accounts should be positively reconsidered. In this
way, accounts have a more subtle and contextual informative function:
the most informative thing about accounts as currently prepared might not be the lit-
eral interpretation of the information they contain (claimed debt-equity ratios, profits,
etc.), but the fact that an apparently competent group of professional auditors, after
exercising their professional judgement, are prepared to let the company make the
claims that it does, even though the auditors thereby leave themselves open to the
possibility of large law suits (1994, p. 332).
Pursuant to this line of argument,25 the role of auditor decisions must be under-
stood in the context of how recipients of accounts process the information (not
simply shareholders but increasingly by investment analysts in a developed and
therefore specialized economy). For example, assume that an auditor agrees to
suppress detailed hard information in such a way that the accounts only include a
generic and ambiguous statement of the "including adjustment of taxation reserves"
type. It could be argued that this phrase is sufficient for readers to assume the worst
in relation to the amount affected, which is not divulged. 26 The recipient of the
information will positively assess the fact that the auditor has agreed not to detail
the information as a signal of confidence, however. The reader knows that the
auditor is taking a risk by allowing this and that he would only do this if he consid-
ered that it was unlikely that the matter would give rise to future litigation or loss of
reputation. In other words: the suppression of negative hard information from some
audited accounts in itself impliedly involves the dissemination of positive soft in-
formation. "In the signalling equilibrium, statements like 'including a transfer from
taxation reserves' has a conventional meaning: the client fum has soft and hard
information that together exclude the worst possible cases" (Grout et al., 1994, p.
335).
1.2.2.2 General normative consequences: the argument for rules which

optimize the informative content of auditing
This analysis has far-reaching consequences for both the ends and means of legisla-
tive and regulatory policies in the auditing field. In the first case, it justifies
the client. Although the set of information which is verifiable by a judge probably includes all the
information observable by researchers carrying out this type of empirical study, a comparison of the two
situations has at least an indicative value to the effect that those participating in economic activities use
sets of information with different properties in terms of its observability.
25 This contextual interpretation of financial statements also accounts for possible systematic bias, if
any, as modeled by Antle, Nalebuffand Baiman (1991).

26 This was the situation in the Royal Mail case, dealt with by Grout el al. (1994, pp. 333-4).
Auditing quality 13
legislators and regulators adopting the objective of optimizing the value of the in-
formation which the audit provides to users. To this end, they should consider that
the rules define the type of audit which it is possible to carry out. Secondly, with
respect to the means for bringing about this objective, it must be taken into account
that regulations which purport to advance independence to the limit can prevent
auditors from transmitting the soft information available to them. If, consequently,
they base their opinion on hard information only, the audit will be likely to have
less informative content, however. 27
From this point of view, when defining the context in which auditors work, leg-
islators can make as many mistakes in one direction as in the other, laying down the
rules in either too lax or two stringent a manner. If the rules are too stringent,
auditors will tend to carry out "defensive audits" and only use hard information
since this can be used as a defense in litigation. Moreover, they will refuse to base
their decisions on soft information which will not assist in their defense if the client
ends up with problems in the future. As a result, the accounts and audits will con-
tain less information.
It could be thought on the other hand that if the rules are too lax auditors will
bow to the demands of their clients and in tum will tend to be very lenient, giving
their approval to companies which do not warrant it as a result of their high degree
of vulnerability.28 This consequence cannot be generalized, however. This may be
the conduct of some auditors but the market provides incentives for at least part to
develop a reputation for exercising balanced professional judgements. Let us con-
sider an extreme case in which the law does not provide any system of sanctions
against auditors and it is also impracticable to sue them. It is foreseeable that in
such a situation there would be greater development of private enforcement and
sanctioning instruments. These would be based on the auditor's reputation and the
castigation of underperformance through switching decisions.
Firstly, there is not doubt that in this situation auditing firms have incentives to
develop a good professional reputation since there will be clients who demand
auditing with a optimum degree of independence and this would not be the outcome
of a very lax legal system. On the other hand there is no risk that the market will
itself generate an excessive level of sanctions which would be equivalent to a situa-
tion of excessively stringent rules. The reason is that the market is probably more
competent than the legal system when it comes to verifying qualitative information,
on several grounds. Without being exhaustive, it should be taken into account that
27 The report from the Centre for European Policy Studies on the proper government of companies
shared this view in relation to the regulation of accounting information by statutory rules in continental
Western Europe (CEPS, 1995, p. 18).
28 As do Grout et al. for example (1994, p. 336).
(a) the market is not restricted to the use of specific types of evidence; (b) it acts by
accumulating an almost infmite number of individual decisions and therefore indi-
vidual variables are of little importance; (c) such decisions are taken largely by
professionals who can be assumed to be well informed because they have incentives
to be informed; and (d) their negotiating costs in this case are nil (as compared with
their equivalent in the legal process or an out of court agreement). As a result,
much information which is observable by auditors but not verifiable by the legal
system is verifiable by the market and auditors will have no hesitation in using it if
the potential sanction comes from the market rather than the legal system.
There is thus substantial asymmetry between the consequences of the two types
of regulatory error. The risk with laxity can at least in part be corrected by the mar-
ket and, moreover, does not pose obstacles to using all types of information. The
risk with excessive strictness on the other hand cannot be corrected: all auditors,
whatever their reputation, are forced to dispense with soft information when pre-
paring their professional judgements. It should further be noted that if there is
greater variability in the sanctions of a legal type, this adds force to the argument
that increasing them tends to bias the content of the audit, making it less informa-
tive.
1.2.2.3 Particular normative consequences: non-audit services as a source

of information for professional judgement
If the informative value of auditing is taken as an objective and, therefore, there is

some appreciation for the need for auditors to exercise their professional judgement,
understood in the same sense as in the previous sections, the provision of non-audit
services will have a doubly positive effect. Firstly, when such services are provided
to audit clients the auditor can reach a more deeply grounded professional judge-
ment since he will have a greater depth of knowledge of that part of the value of the
business which is rarely reflected in the accounts, such as intangible assets (reputa-
tion, solid organizational structure, management capability, etc.). By carrying out
purely auditing tasks, it is more difficult to gain a better idea of the extent of such
assets. Those using the accounts would like to receive information on the existence
of such assets, however. The audit can at least provide an indication of them and
this indication will be more reliable the greater and better the knowledge serving as
a basis for the professional judgement which the auditor must construct, a knowl-
edge which can be substantially improved by providing non-audit services to the
same client. In addition, the provision of such services will enable the auditing firm
to contract and make efficient use of the experts required to improve and extend its
professional judgement in terms of greater reliability and depth as well as extending
it to highly specialized activities and undertakings.
Auditing quality 15
1.3 The conflict with third parties and among clients

regarding auditor independence
The greater part of the added value of auditing is generated as a contract facilitating
service. As such, at least three parties (client, auditor and user of the accounts) have
an interest in auditing quality. The asymmetries and conflicts that arise are thus
more complex than in the majority of goods and services markets in which only two
parties are involved. From this point of view, conflicts which arise on the basis of
the two dimensions or attributes of auditing quality, i.e. technical competence (a)
and independence (b) can be distinguished.
a) Conflicts in relation to technical competence in the strict sense have fairly or-
dinary characteristics inasmuch as it can be assumed that the client will always want
the auditor to provide quality in this respect as with any other supplies. In such
cases the only relevant informational asymmetry is that which may exist between
client and auditor since the former always has an interest in preventing it from aris-
ing, which will tend to prevent it appearing in relation to third parties. Moreover,
the quality can be perceived at the same time or immediately after the service and is
therefore an attribute which can be categorized as an "experience" rather than a
"credence" attribute.
b) Independence is the more complex qualitative attribute. Firstly, the role
played by auditing in third party contracting means that the client will require inde-
pendence or otherwise depending on his situation. In problematical cases, although
the client in principle usually wants high quality auditing, after his fmancial situa-
tion changes he may prefer a low quality audit in terms of independence. More
precisely, he would prefer a dependent auditor, although one perceived by third
parties as being independent. The reason is that this deceptive audit will enable him
either to contract with third parties on better terms than those available in his situa-
tion if the same is known to such third parties, or to postpone the review of those
contracts which involve corrective action based on financial situation (as normally
occurs, for example, in the case of loan agreements).29 In order to practice this
deception on third parties the latter must obviously attribute the auditor with supe-
rior quality to that which he is actually providing since otherwise they would
discount a favorable report which would lose its value in terms of evidencing the
client's situation.
Third parties who contract with a client of an auditor trusting in an audit which
they believe is independent are not the only economic parties prejudiced by a de-
ceptive audit, however, nor perhaps even the most prejudiced when the legal system
29 See Smith and Warner (1979).

imposes strict professional liability on the auditor. They are not the only ones be-
cause other clients of the auditing ftrm will also be harmed. If it is assumed that,
particularly in the view of third parties, auditors generally provide a uniform level
of quality in their audits (particularly in terms of independence), the clients who
continue wanting an independent auditor to demonstrate their fmancial good health
will see their names associated with that of an auditor tainted by lack of independ-
ence. Moreover, unlike these clients of the auditor, affected third parties can have
recourse to the legal system to obtain compensation for the loss which may be
caused to them by the lack of independence. 3o They also have the active presence
of professional bodies to support them along with regulators of the auditing profes-
sIOn.
As a result, a large part of the conflict regarding independence does not manifest
itself so much between auditors and those using the accounts of a client in a difftcult
situation as between auditors and their other clients who do not wish the independ-
ence of their auditor to be dirninished. 31 The cost to one of these clients of his
auditor reducing quality and this being visibly so (by a scandal, for example) arises
not only in respect of the direct loss to him from the fact that those using his ac-
counts have less conftdence in them, but also because of the deterioration in
"speciftc" assets, those whose value is associated with continuity in his relationship
with the auditor. If, by way of reaction, he decides to change auditor, he will lose
all these assets. Alternatively, if the client decides to continue with his original
auditor, he will suffer a loss in terms of the diminished value of the audit to recipi-
ents and, therefore, in terms in the increased agency costs he will incur in his
relationships with them. It is true that since the scandal affects his negotiating
power with the auditor it can also be expected that if he decides to continue the
relationship he will transfer part of this loss to the auditor by means of a price re-
30 This compensation may also be excessive in those countries in which the auditor is subject to a
system of joint and several liability, including several European countries (Buijink et aI., 1996, p. 96).
This matter will be considered again in Section 3.4.3.
31 Readers familiar with management practices and the economic literature of franchises will recognize
a structure in this conflict very similar to that often arising between franchisees of the same franchisor.
Franchisee establishments tend to reduce the quality of their services to the prejudice of the reputation
of the brand name and the network of establishments of which they form part. For this reason, an
essential function of the franchisor is to safeguard quality and take disciplinary measures against those
who do not fulfil minimum standards. When taking disciplinary measures such as expulsion the fran-
chisor does so to the benefit not only of himself but, and perhaps principally, to the benefit of the other
franchisees. It is believed that the periodic collection of fees ofa variable nature (which incurs a cost in
lessening the incentives of franchisees) is precisely aimed at giving the franchisor an incentive to carry
out this disciplinary function effectively. See, in relation to this aspect, principally the works of Rubin
(1978, p. 227), Brickley and Dark (1987, p. 410) and Lafontaine (1992, p. 279). Moreover, litigation
brought by franchisees against their franchisors as a result of inadequate control of members of the
network is commonplace. A famous case in the United States was Creel Enterprises vs. Mr. Gatti's, in
which the former sued the franchisor as it was harmed by the repeated breach by another member of the
chain of obligations imposed on all the franchisees (Johnson, 1992, p. 18).
Auditing quality 17
duction, as examined in Section 3.3. (From this point of view, the need to
discipline the auditor who compromises his reputation for independence helps ex-
plain why under a system of freedom of contract auditing contracts are entered into
on an annual basis, despite the fact that the relationship usually lasts for a much
longer period, of between 30 and 40 years).32 It will be shown later in this work
that regulation would do better by modifying its current emphasis on the effect that
auditing failure has on third parties, because in so doing it would strengthen market
incentives.
Summarizing, auditing produces three principal types of conflict in relation to
quality: (a) asymmetry between the auditor and client which is directly manifested
basically in relation to technical competence which can be seen as an "experience"
service and which is the least conflictive; (b) the indirect asymmetry between the
auditor and client on the one hand and users of accounts on the other which is mani-
fested in terms of independence and is of the nature of a "credence" service, since
quality is only perceived with the passage of time and when the financial situation
of the client deteriorates; and (c) the indirect asymmetry between the different cli-
ents of the auditor in relation to his independence in respect of other clients. The
latter is probably the most distinctive and conflictive feature of auditing because the
fall in quality is only perceptible on an occasional and delayed basis.
32For European companies, data from Ridyard and de Bolle (1992, pp. 89-91) enable the average
rotation period to be estimated at between 30 and 40 years. Large companies seem to have a lower
change rate. Thus, in a study of 3,500 audits carried out between 1980 and 1988 in England it was
estimated that the average length of each relationship was 40 years ("Auditors Too Cosy with Clients?"
[Accountancy, January 1995, p. II)). Ridyard and de Bolle provide similar data for Great Britain: in a
sample of 137 large companies the rotation rate was lower than I % between 1987 and 1990 (1992, p.
89). The figures are similar in the United States. There, rotation affects a percentage of between I % of
large companies and 6% of small companies each year. See, on this, the studies cited by DeAngelo
(1981 b, pp. 188-9) and Beck, Frecka and Salomon (I 988b, pp. 68-9).
2 THE SAFEGUARD OF
QUALITY
In most economic exchanges resources must be devoted to prevent the parties from
seeking to further their own interests to the prejudice of a successful outcome to the
exchange. If this were not so, not only the achievements but also the very existence
of the exchange would be threatened as happens when, before contracting, the other
participants anticipate the possibility of being subjected to opportunistic conduct.
To ensure that the agreement is fulfilled, in addition to using external mechanisms,
most importantly represented by the legal system, a series of instruments develop
which can be classified as "market" instruments. These essentially consist in intro-
ducing automatic penalties for non-compliance, pursuant to which the party who is
obliged to perform will be castigated if others in the marketplace ascertain that he
has failed to comply with his obligations to another of their number.
These mechanisms usually operate in a very simple manner through the receipt
of a remuneration which exceeds the highest remuneration which would be avail-
able in the event that some non-compliance is observed (this remunerative premium
is usually technically referred to as a "quasi-rent"). The most important example of
this type of mechanism is provided by a good commercial reputation, which enables
a higher price to be charged compared with that which could be charged by those
who contract without a reputation. Nevertheless, continuing receipt of this higher
price is only ensured if a level of quality is provided consistent with the expecta-
tions and remuneration: non-compliance generates a more or less automatic
sanction and, as a result, reputation acts as a safeguard of quality (Section 2.1).
In more general terms, there are a variety of more or less costly possibilities for
placing an economic agent in a position where, faced with the threat that the
market-potential contracting parties-penalize him, he has an interest in fulfilling
his obligations. Sections 2.2 to 2.4 examine how the principal ones can be used to
ensure the quality of auditing and how regulatory measures can facilitate their use.
19
2.1 General analysis of safeguarding mechanisms
2. 1. 1 Explicit and implicit safeguards
In very broad tenus, quality can be safeguarded by both explicit and implicit con-
tracting. With the former, breach by the supplier can result in litigation and a
consequent legal sanction. This method is in general only used for certain types of
breach, without doubt due to its high cost and rigidity and, in particular, the fact that
in order to be effective the breach must be verifiable by third parties. In the case of
auditing, for example, an auditor may be obliged to compensate those affected by
the insolvency of a company whose accounts have been defectively audited.
With implicit contracting, on the other hand, the penalty for breach is decided by
potential contracting parties by withdrawing their confidence from the person in
breach and, as a result, not contracting with or demanding more onerous tenus from
him. The clearest example is the loss of reputation which is usually associated with
the disclosure of any type of breach. It will be shown that with auditing the loss of
clientele which accompanies public awareness of deficient audits is very high. In
general terms, the advantages of implicit over explicit contracting reside, firstly, in
its automatic nature and its scope: more information is processed at a lower or nil
cost since, as this information is a by-product, no administration cost is required.
Moreover, decisions to sanction are decentralized. At root it is each individual
potential contracting party who in part acts as a first instance judge of the auditor's
conduct. This makes it possible to process a greater quantity and variety of infor-
mation in such a way that non-compliance can be penalized in a more precisely
fitting manner. Unlike the case with the legal system which must often deal with
discrete criteria of the "all or nothing" type, the quasi-judicial operation of the mar-
ket thus provides for a whole range of sanctions.
2. 1.2 The role of quasi-rents in the safeguard of quality
For the above reasons, a large proportion of quality safeguards in both auditing and
other economic activities basically arise by means of "implicit contracts". In this
case, compliance with obligations cannot be required legally, but depends on inter-
nal incentives of the parties, such as those which generate reputation in particular.
The mechanism on which such implicit contracts are based is thus the benefit to the
party himself who is obliged to perform them. For example, imagine that the qual-
ity of an auditor's work could only be observed one year after completing it. If he
has a reputation as an auditor of quality, it will be in his own interest to maintain it,
The safeguard of quality 21
thus offsetting a possible incentive to act in an opportunistic manner and reduce

quality.
In economic terms, this type of benefit is a "quasi-rent". A quasi-rent is taken to
mean the difference between the remuneration for any productive resource in its
current use and the maximum remuneration which would be received for its best
alternative use. I In order for implicit contracting to be effective, there must be a
credible commitment that the agent-in this case the auditor-will receive a stream
of quasi-rents giving him the incentive to perform properly.
Several conditions must be fulfilled for this to occur. Firstly, the sale price re-
ceived by such agent must be greater than the marginal or opportunity cost in order
that the quasi-rent is thus generated. In this way he obtains income above his
avoidable costs and the possibility of losing this profit can motivate him to fulfil his
contracts 2• Furthermore, the expected value of the quasi-rents arising from the
difference between the price and this opportunity cost must be sufficient to discour-
age non-compliance. This effectiveness of quasi-rents depends on their amount and
the time scale over which the supplier receives them, aspects which define their
present value. Finally, customers must be aware of the supplier's incentives so that
they can therefore trust him. For this reason, it is desirable that the quasi-rents are
visible and this generally necessitates an explicit communication strategy.
2. 1.3 Formal analysis of quality assurance
In order to show how the reputation mechanism which ensures quality functions in
a context of informational asymmetry, a very simple, but typical, situation will be
formally analyzed in which the producer can select the level of quality and transac-
I More strictly speaking, a quasi-rent is that part of the remuneration of a production factor which
exceeds the minimum necessary to prevent its current activity being abandoned; for example, for a
worker to give up his job and seek work in another business. A "rent", on the other hand, is the excess
remuneration over the minimum necessary to motivate a resource to enter or be devoted to an activity.
The distinction is more relevant for the pUllloses of assessing the situation from the point of view of
competition policy that in understanding the private incentives which they generate. Studies into how
the expectation of obtaining a stream of quasi-rents in the future provides an automatic incentive for
producers to preserve the quality of their products or services were begun by Becker and Stigler (1974)
and developed by Klein and Leffler (1981), Williamson (1983) and Shapiro (1983), with application to
the employment field in modeling the use of remuneration for length of service to motivate perform-
ance, avoiding the "moral hazard", begun with the work of Lazear (1979) and, more recently, in the
literature on "efficient wages" beginning with Shapiro and Stiglitz (1984).
2 Note that the fact that the product is sold at a price exceeding its marginal cost is compatible with the
fact that the company selling it is in competition; it suft\ces, for example, that in the relevant production
range, fixed resources are employed with a nil opportunity cost.
tions are repeated. 3 (The objective is to clarify the analysis, but the reader can skip
this Section, without substantial loss of continuity, and move on directly to Section
2.2. The same is true for all the other mathematical sections contained in this
book).
Let us suppose that each firm, in each period of time, sells a single unit of prod-
uct or service of a certain quality, q, generating a cost which increases with the level
of quality, given by a function c(q) such that c '(q) ~ 0 and c' '(q) > 0, and providing
the purchaser with utility which increases with the quality of the service. In order to
examine what happens when the supplier knows more than the customer regarding
the quality of the service, it will also be assumed that the customers are incapable of
fully identifying the quality of the service before acquiring it. Specifically, it will
be taken that they are capable of identifying the quality of the service up to a certain
level, qa, but before acquiring the service it is impossible for them to know whether
this quality exceeds qa. 4 As a result, they only learn at time t+ 1 about the quality of
the service acquired in the preceding period t.
On this basis, an equilibrium situation can be characterized by a process of cre-
ating expectations amongst purchasers, and a commercial strategy on the part of
firms in terms of the quality of their products. In particular, it will be assumed,
firstly, that customers base their decisions on the reputation of the firms and that
they form their expectations solely by considering the prior quality of each firm. In
this way, they expect to receive a quality from each one during period t equal to that
provided by the same firm in the preceding period, t-e
Secondly, each firm is free
to choose the level of quality of its services and develop a reputation associated
with this quality level. 6 In order to reach an equilibrium, however, the firms must
have an interest in not deceiving their customers. For firms to want to maintain
their reputation, the condition must be fulfilled that the value of the stream of future
quasi-rents associated with their reputation exceeds the benefit obtainable from
J The model provided in this Section was developed by Shapiro (1983) and constitutes a stylized
version of Klein and Leffler's (1981). These models contemplate the situation which is most typical in
commercial relationships: that in which the producer can select the level of quality and it is expected
that transaction will be repeated in the future. See, however, Tirole (1988, pp. 106-26) for a systematic
introduction to the diversity of quality assurance problems, defined on the basis of whether quality is
fixed or can be modified by the seller and whether transactions are repeated or not.
4 This level of quality can be treated as the minimum required by professional standards in such a way
that the model describes a supply of services of a quality exceeding this minimum.
S The introduction of other hypotheses regarding the formation of these expectations does not substan-
tially modify the results. See Shapiro (1983, pp. 671-3).
6 When there is informational asymmetry and it is costly to evaluate quality, it can be expected that
producers will tend to identify and specialize themselves in the production of goods and services of
uniform quality in order thereby to reduce the contractual costs associated with safeguarding quality.
The reason is that it is usually more economical to develop the productive technology and establish a
reputation for producing uniform quality than not to produce and ensure differing qualities.
opportunistic behavior which, given the conduct previously assumed from custom-
ers, will only be beneficial in one period:
l+r
[p(q) - c(q)] -r- ~ p(q) - c(qo) [1 ]
from which it follows that the price of a service of quality q must exceed a certain
level which is given by:
p(q) ~ c(q) + r [c(q) - c(qo)] [2]
Moreover, in the competitive equilibrium, there must be no incentives for entry

into the market at any level of quality, q, and therefore, if Pc represents the entry
price, the following condition must also be fulfilled:
Pc - c(q) +p(q); c(q) s 0 [3]
It seems logical that the price at which a service which is entering, and therefore
lacks reputation, is sold will equal c(qo). The reason is twofold. Firstly, there
would be no point in it being lower, since consumers are prepared to pay a price
c(qo), as it is assumed they are capable of identifying the level of quality qo. On the
other hand, it cannot exceed c(qo), since otherwise this would provide an incentive
for the entry of multiple producers who would earn an abnormal profit providing
minimum quality. This enables us to replace Pc by c(qo) in [3]. On resolving p(q),
it is found that for a certain quality q, the price will never exceed a certain minimum
in order not to encourage entry:
p(q) s c(q) + r [c(q) - c(qo)] [4]
Taking [2] and [4] into account simultaneously, the price of a service of quality q
is at the same time the minimum price motivating suppliers to maintain quality and
the maximum price at which potential market entrants have no interest in entering
the market:
p(q) = c(q) + r[c(q) - c(qo)] [5]
In these conditions, the supply curve as a function of quality will be as shown in

Figure 2.1. The minimum quality qo is sold at its cost, c(qo), whilst all other quali-
ties are sold at a price exceeding cost. This difference between price and cost
constitutes a premium which ensures quality by providing incentives to maintain it.
This quality premium is higher for a higher quality of service (Figure 2.1). (It also
increases as the interest rate rises or the duration of the purchase period is ex-
tended).7
p
p(q) = c(q)+r[c(q)-c(qoJJ
c(q,)
Figure 2.1. Supply of services as a function of their qualityB
This quality premium can be interpreted in terms of the cost of producing a

service of quality q being equal to the production-or, more strictly speaking,
"transformation"-cost, c(q), plus the transaction cost given by r[c(q) - c(qa)}. The
latter is the cost of establishing a reputation of quality q and, in the simplified terms
of this model, is incurred in the initial period in the form of introductory prices.
The quality premium thus represents the normal return on this capital "invested" to
establish the reputation.
2.2 Strategies adopted to create quality-safeguarding

incentives
Different strategies can be used in order to generate a stream of quasi-rents to en-

sure quality. Although they are of similar effectiveness, given by the present values
7 Having defined the interest rate, r, for a period and not in annual terms, this rate will increase as the
duration of the re-purchase period increases. (In formal terms, if i is the annual interest rate and T the
duration of each period expressed in years, then r = eiLJ). For this reason, an increase in the duration
of the purchase period, T, has the same effect as an increase to the annual interest rate, i.
8 Source: Shapiro (1983, p. 668).
of the quasi-rents generated, they can give rise, however, to very different costS.9
The reason is that some of these are based on a simple difference in prices over time
or on the use of productive assets, which generate quasi-rents as a result of their
specific nature to the undertaking obliged to perform. lo In both cases the quasi-
rents can be generated without incurring any cost. (More precisely, if there is a
cost, it is incurred only to the extent that the pricing strategy or investment decisions
would have been sub-optimal in the absence of the safeguard value involved in the
generation of quasi-rents). Following other quality-assurance strategies, on the
other hand, actual resources have to be employed solely to generate the quasi-rents.
This occurs with uninformative advertising and also with the different unproductive
forms of competition in the "rent seeking", the aim of which is to enter markets in
which entry is restricted, this latter situation being typical of the more corporatist
professional services.
When analyzing these strategies below and their possible use in auditing, it will
be clear that in auditing activities the investment in learning associated with each
client and the consequent quasi-rents to which it gives rise are a mechanism of the
first type, enabling quality to be safeguarded at low cost. On occasions, naIvely
drawn up regulations try to prevent or obstruct the existence or receipt by the audi-
tor of these quasi-rents. (Various examples of this will be provided when analyzing
current regulations in different countries in Section 6.3.1.4). This is the case, in
particular, with regulations which prohibit non-audit services being supplied to
clients. A consequence of this prohibition and of all those in general which prevent
the receipt of individual quasi-rents is that the parties are obliged to employ a sub-
optimal technology to generate safeguards, and are thus obliged to devote additional
resources to this task with the consequent increase in cost of the services.
2.2. 1 Low-cost strategies
There are several low-cost strategies for generating incentives to perform. They are
mainly based on the prior performance of obligations (contracted for at introductory
prices, whether implicit or explicit), on the ownership of specific assets whose value
is linked to continuity of presence in an activity or market and on extending reputa-
tion across markets or activities. These three generic strategies are all applicable in
9 The costly nature of developing a reputation which safeguards the quality of auditing has been veri-
fied in various empirical studies, the most important being Craswell, Francis and Taylor (1995).
10 The specific assets are resources with a value in their best alternative use which is less than in their
current use. Basic works in this field include those of Klein, Crawford and Alchian (1978) and Wil-
liamson (1975 and 1979). The concept had already been applied in the field of human capital, starting
with the work of Becker (1964).
the auditing field and, as a result of their low cost, are preferable from the social
point of view. Regulators would thus do well to facilitate their development and
enhance their effectiveness. Let us now briefly examine how they function both in
general and in auditing.
2.2.1.1 Introductory pricing
The most elemental form of creating a good reputation is to introduce the product or
service at a price below that corresponding to its quality.1I The provider may even
settle for losses during the take-off stage in the hope of recovering them later if he
manages to establish a good reputation for his services. 12 From the moment when
customers are convinced of his good quality, he will be able to raise the price (or
more precisely the price/quality ratio), and recover the initial investment. At times,
an up-market or quality improvement strategy is used, in which the seller begins by
producing low quality but, as the product is improved, he manages to gain access to
higher quality market segments. To this end, it suffices for him not to increase
prices as the quality of the service improves, thus little by little surpassing the ex-
pectations of his customers.
In the case of auditing the strategy which consists in constructing a reputation by
providing a higher level of quality than that expected and abstaining from oppor-
tunistic conduct is very important and, in its different versions, constitutes an
essential element in the system of safeguards. Allocative inefficiencies may occur,
however, when introductory pricing is used, although these are probably of a lesser
order. In addition, it generally operates slowly. To the extent that it falls into the
category of those goods or services which were earlier categorized as of a "cre-
dence" nature, the quality of auditing takes a long time to be appreciated.
Accelerating this strategy would necessitate using explicit introductory pricing,
selling services substantially below cost, a practice which suffers, as in other sec-
tors, from a high risk of confusion: it is not easy to prevent a name being associated
with low quality.13 To this is added the mistrust amongst many observers of price
competition in the sector. The long life of the main auditing firms provides an em-
pirical indication that this strategy is far more viable in the long term, since it seems
to suggest difficulties in consolidating a good reputation in a short period of time.
II For an introduction to the multiple possibilities involved in the use of pricing as signals of quality,
see Carlton and Perloff (1994, p. 562, n. 4).
12 This in essence is the specific case formerly dealt with in Section 2.1.3 of this Chapter.
J3 This refers to a possible policy of generalized discounting for all clients for the time it takes to estab-
lish a reputation, and not to the common practice of discounting the price only for new clients
("Iowballing"). Evidence on cross-sectional variation in the pricing of initial audits in Australia is
consistent with the discount playing an introductory function for a product attribute whose quality is
known only through experience. There, lowballing is observed only when the client upgrades its audit,
starting to be audited by one of the big firms (Craswell, Francis and Sneddon, 1997).
In the market for non-audit services, several observations indicate that this strat-
egy may be viable more rapidly. Firstly, the rate at which new businesses appear is
higher, especially nowadays. Secondly, there are some circumstantial indications
regarding the use of introductory pricing in this activity for this purpose. Thus, in
the 1990s, for example, Andersen Consulting offered its relatively new strategic
management services at prices 20 percent lower than those of the more established
consultancy fIrms, whilst its prices were higher in those business lines in which it
had traditionally been a world leader (Nanda and Yoshino, 1996, p. 3).
2.2.1.2 Decisions regarding the size of fIrms and the structure of their
assets
When the value of capital goods which businesses use depends on their continuity
and successful progress, the possibility that such goods may lose value also operates
as a quality guarantee. In the auditing case, however, the productive process does
not result in the use of many specifIc tangible assets. Firstly, the processes involved
utilize little physical capital but are intensive, however, in human capital. Moreo-
ver, the physical capital used is mainly of general purpose, largely consisting of real
estate, computer equipment and furniture. Furthermore, neither does a large part of
the human capital serve as a guarantee because it is also of general purpose and
preserves most of its value both outside the fIrm and outside the sector (this, it
should be mentioned in passing, makes it easier to apply up or out policies for
younger professionals). The specifIc assets are thus reduced to the specifIc human
capital built up by the longer-serving employees and which usually take the form of
relevant knowledge in commercializing and rendering services for a series of cli-
ents.14 There is also room to consider that a substantial part of the intangible assets
of an informational nature are not embodied in any of the individuals working in the
14 The existence of this specific human capital helps to explain the particular ownership structure which
prevails in such finns: the fact that these long-serving employees are partners in the finn serves as
automatic protection against possible expropriatory decisions which could occur if the partners them-
selves were not the recipients of the residual rent of the finn. The reason is that the importance of the
specific human capital raises a problem: as with all human capital, its ownership cannot be transferred
to the finn. This can generate perverse incentives since those in the business deciding on quality will
not in the end bear the cost of their decisions. To avoid this problem, the traditional solution has in-
volved structuring businesses as professional companies or partnerships, thereby leading to the
principal proprietors of this human capital also being those entitled to the finn's residual income. It
should be noted that in principle the fact that the professionals are partners has further consequences
which help to explain their dual role as workers and capitalists. On the positive side, it motivates
mutual control of professional activities (Fama and Jensen, 1983a, pp. 315-7; and 1983b, pp. 334-7).
On the negative side, it implies a high concentration of risk in the fate of the finn, which aggravated the
extent of the crisis caused by the step-up in damages paid for professional liability in Anglo-Saxon
countries.
organization but, rather, they lie in both its own organizational configuration and
the interaction of such individuals. Examples include a set of problem-solving
methodologies, decision-making routines and, in short, the organizational technol-
ogy of the firm as a whole.
In analyzing auditor independence, perhaps the most crucial attribute of audit
quality, two types of specific assets should be distinguished, whose character can
present different properties: on the one hand, the assets which are specific to the
firm but of utility in providing services to any of its clients and, on the other hand,
those which only have value in auditing a particular client. The former clearly acts
as a safeguard of quality and, specifically, independence. The effect of the latter, on
the other hand, depends on several variables, particularly the degree of client diver-
sification: (a) if the auditor has a single client, the presence of such assets could
prejudice his independence; (b) on the other hand, if he has several clients, most of
whom demand independence from their auditor, the possibility of losing all these
specific assets will tend to weigh much more heavily on the auditor's decisions than
the potential threat or even certainty of losing a few of them.
As a result of the complexity of the problem, a detailed analysis of the other
variables on which the effect of such assets on auditor independence depends has
been left for Chapter 3. Before examining these other effects, however, the impor-
tance of which is in any event secondary, it should be emphasized that investment
specific to each client can have the function of safeguarding quality. The realiza-
tion of this potential depends on variables which are relatively controllable by
regulation. 15 As such, these specific assets have a substantial advantage: without
losing their directly productive character, they generate a positive external effect in
terms of contractual safeguards. The basic objectives of regulation should, for this
reason, include ensuring that this external effect is genuinely positive. In later
chapters it will shown that client diversification is a necessary condition in this
respect and that this can be achieved either directly through diversification rules or
indirectly by facilitating knowledge by the market of the degree of diversification of
firms.
Since the pioneering work of DeAngelo (1981 b) these specific assets have
tended to be associated with the quality of auditing, and estimated by the size of the
auditing firms. Although this connection is unquestionable in relation to technical
competence, it must be qualified in relation to independence since, as there is a
positive correlation between the size of firms and the size of their clients, client
diversification may determine independence more than the mere size of the firm.
The size of the firm can, however, be associated more directly with independence
on the basis of the guarantee provided, firstly, by those assets which are specific to
IS The idea that the specificity of productive assets economizes on contractual costs was suggested
generally by Klein and Leffler (1981, pp. 627-9) and highlighted in the case of auditing by DeAngelo
(1981 b, pp. 193-4).
the firm but which are not specific to any particular client and, secondly, the wealth
of both the fum and its partners which serves as security for its liabilities. 16
For this reason, it is not surprising that different empirical studies, using different
methods, have consistently confirmed that quality is positively correlated with the
size of the auditing firm. It suffices to refer to a few of these studies in which it has
been observed: (a) that companies that seek listing on the Stock Exchange are then
usually audited by one of the large firms (Carpenter and Strawser, 1971); (b) that
the large firms are subject to a lower litigation rate despite the fact that it is more
remunerative to sue them because of their greater wealth (Palmrose, 1988a); (c) that
those companies audited by large firms are underpriced less when first listed on the
Stock Exchange (Balvers, McDonald and Miller, 1988; Beatty, 1989); (d) that there
is greater Stock Exchange response (and surely, therefore, credibility) from an-
nouncements of increased profits when the auditor is a large fum (Teoh and Wong,
1993); and, finally, (e) it is notable that the number of cases of professional miscon-
duct is usually inversely proportional to the size of the firm (Schaefer and Welker,
1994) and particularly increase amongst individual practitioners in the profession.
2.2.1.3 Diversification coherent with quality
Those producing high quality can easily launch further products and extend their
reputation to them. For example, a pre-existing and recognized brand is usually
used for this purpose amongst consumer products by way of an "umbrella", a phe-
nomenon also known as brand-stretching. 17 In these cases it is essential that the
new products are also of high quality since, otherwise, they would compromise the
reputation not only of the new products but also of the old.
In professional services, this possibility has been and is applicable to facilitating
the contracting of new products. In particular, the provision of advisory services
suffers from a substantial amount of informational asymmetry (it is not easy to
identify quality and the services are intangible and non-repetitive), and is therefore
problematical from the contractual point of view. For this reason it is logical that
use has traditionally been made of a reputation achieved in more or less similar
16 The presence of diverse sources of quasi-rents has led some authors to measure auditing quality in
two ways: on a continuous scale defined by size, and on a discrete scale (specifically large firms versus
other firms), which it is believed distinguishes two groups of firms with different levels of commercial
reputation. This, for example, is the point of view of Francis and Wilson (1988). Although it makes
sense to consider both sources of quasi-rents as quality indicators, since at root they are the cause or
guarantee of it, there is no basis, given that they act in a complementary manner, for the attempt to
convert them into "competing hypotheses as to the quality of auditing firms" (Francis and Wilson,
1988, p. 665), a claim more recently reiterated by Firth and Smith (1992).
17 See Aaker (1991).
products and services to ensure the quality of the advice. In this respect it is nota-
ble, for example, how some large computer manufacturers have been increasing the
percentage of their turnover represented by advisory services and the outsourcing of
all types of information technology services.
In the field of non-audit services, this formula has served to overcome the seri-
ous contractual problems suffered by providing all types of professional services,
from management and fmancial consultancy to legal services. Devoting resources
to developing specific safeguards to protect the quality of these new services is thus
avoided. Any prohibition on taking advantage in this way of economies of scope of
reputational assets would have the consequence of reducing their value and thus
their own capacity for providing safeguards. (Even if this reasoning takes place in
unilateral terms, these contractual economies of scope flow both ways. Thus, not
only new non-audit services benefit from the reputation acquired in providing audit
services but the effect has been and is mutual: reputation and other safeguards are
utilized more intensively by broadening the set of activities developed by the firm.
As a consequence, the quality of audit services is also being ensured at a lower cost
or a better quality is provided when incurring the same level of safeguarding costs).
Figure 2.2. Effect of a production cost reduction
The consequences involved in providing additional services can also be looked

at formally in terms of the model previously developed in Section 2.1.3. To this
end the particular nature of quality in auditing must be abstracted; in particular, the
possibility that the client may want poor quality in specific circumstances. This
abstraction is acceptable, however, since, as shown in Section 3.4, in circumstances
of diversified clientele the pressure which one of these clients may bring to bear
does not affect the auditor's decision.
a) Firstly, if (as argued in Section 4.2.3) there are productive economies of
scope, production costs are reduced and, therefore, the cost function c(q) is shifted
downwards, passing from clq) to clq) (Figure 2.2). This means that it is less
costly to produce any level of quality. It can be seen that each level of quality is
now achieved at lower cost and a lower price paid for it.
Figure 2.3. Effect of the existence of external effects of a contractual nature between
auditing and non-audit services
b) Secondly, it can also be understood, in terms of the model, that contracting

these additional services enables a pre-existing reputational asset to be used
(economies of scope of a contractual nature thus appear). If the auditor provides
both audit and non-audit services, it can be expected that non-performance in
auditing will result in a loss of reputation in regard to the other services, particularly
when the same clients are involved. This effect means that the reputational pre-
mium necessary to motivate provision of the promised level of quality is lower. In
algebraic terms, the supply of auditing services as a function of their quality is now
given, similarly to [5], by the expression: 18
18 It is assumed that reputation in consulting is not of utility in gaining entry to auditing. Moreover, it is
not possible to reduce the quality simultaneously in both markets. The analysis thus contradicts the
argument occasionally used that, when providing additional services, the auditor has less interest in the
quality of the auditing. See, for example, Gonzalo Angulo (1995, p. 618).
p(q'A) = c(q.J + r [c(q.J - C(qAaJJ - (1 +r) [p(qcJ - c(qcJJ [6]
where the subscripts A and C represent the corresponding auditing and service vari-
ables and the final addition is the loss of quasi-rents obtainable from services which
would result from a possible fall in auditing quality (Figure 2.3).19
2.2.2 High-cost strategies
2.2.2.1 Marketing investments
Advertising and all types of expense devoted to developing commercial brands

today play a fundamental role in safeguarding commercial relationships. These
outlays constitute an implicit guarantee to the customer of future continuity of
service and/or reliability of product. In particular, advertising, perhaps the most
conspicuous of them, thus generates an "advertising capital" which serves as a
guarantee, even when it is not directly informative: simply having advertised pro-
vides an incentive to sell a good product of uniform quality which does not let down
expectations. For this reason, advertising impacts, once achieved, constitute an
intangible asset whose rapid volatility as a consequence of defective performance
gives it a substantial advantage in effectively acting as a quality safeguard. In the
case of auditing, there is empirical evidence which is consistent with this argument,
since it points to the fact that implementation of an active commercial policy by
auditing firms improves auditing quality. Specifically, Jeter and Erickson (1995)
observe that defective audits are less frequent in those American states which permit
the unsolicited offering of audit services.
This possibility for developing reputation, however, has one drawback. It is an
investment which in itself, and apart from its informative function, does not create
any other value beyond the safeguard it provides. Moreover, in many countries
advertising is severely restricted or monitored by regulations or professional stan-
dards. Within the European Union, there is only freedom in this field in Austria,
19 It should be indicated in passing that in both this case (b) as well as (a), the effect on prices will only
be decreasing to the extent that actual resources have to be devoted to "produce" the reputational guar-
antee. (The prices in the Figures are the equilibrium prices, those of recurrent services). Otherwise, if
creation of reputation is only based on introductory pricing, the reduction indicated in the previous
Figures would be offset by a rise in the initial price in the period in which investment is made in repu-
tation. Nevertheless, even in this case, such intertemporal differences in prices provide an incentive for
excessive consumption at the outset and sub-optimal consumption in the other periods. Simple reduc-
tion of the quality premium thus gives rise to prices which enable consumption levels to be achieved
which are closer to an ideal situation in which the information costs are nil.
Denmark, Finland, Greece, the Netherlands, Norway, Sweden and Great Britain.
Moreover, even in some of these countries the unsolicited offering of audit services
is prohibited, specifically in Austria, the Netherlands and Britain. 2o Outside the EU,
advertising is permitted, with different qualifications, in Australia, Canada and the
United States. Unsolicited offering of audit services is forbidden in Australia and
Canada. Both practices are forbidden in Japan.
2.2.2.2 Quasi-rents as a result of barriers to entry
Regulation can also place producers in a position of receiving quasi-rents by im-

posing restrictions on entry into the market. Such a restriction, accompanied by
selection of entrants based on potential technical competence, has in fact been the
formula traditionally used to guarantee a minimum level of quality in all types of
professional services, often under a system of self-regulation by the profession
itself, both explicitly and implicitly. The main, if not sole, advantage of this for-
mula is that it can ensure a certain minimum level of quality in situations where-
often because supervision must be of a professional nature and sometimes due to
other regulatory decisions-in practice there are no better methods for safeguarding
quality.
On other hand, this strategy based on restricting competition has at least the fol-
lowing problems: (a) firstly, the welfare loss derived from the price increase and
consequent reduction in quantity, less important when there are legal restrictions
which impose mandatory consumption, as in the case of auditing; (b) secondly, the
cost increase derived from competitive restraints, more serious when organizational
innovation is restricted; (c) costs are also involved in rent seeking and dissipation,
which largely depend on how the entry mechanism is designed; (d) moreover, there
is a considerable risk of regulatory capture in such a way that the entry barriers end
up solely serving to appropriate monopolistic rents without safeguarding quality,
and (e) finally, it is usually more difficult to develop and bring about changes in
contractual technology and therefore the sectors tend to continue to be institution-
alized even after having developed contractual technologies which correct the
information asymmetries in a less costly manner. 21
Despite these limitations, different variants of this formula have and continue to
play an important role in safeguarding quality in those sectors which suffer from
most conflict in this area, such as professional services (medicine, the law, etc.).
The reason why this solution is employed possibly lies in the fact that the profes-
20For a description of these regulations, see Buijink et al. (1996, p. 53).

21These problems have been examined in greater depth in other works, both from a general and sector
point of view (Arrutiada, 1995, 1996, 1998 and 1999).
sional dimension of these sectors involves a large degree of individual discretion in

engaging in the activity, which has made other possibilities nonviable. It is possi-
ble, however, that much of the corporatism seen today is unnecessary and its
continuance is due to the last of the problems mentioned in the previous para-
graph-i.e., obstructing contractual innovation.
This safeguard strategy was also fundamental in the birth of the auditing profes-
sion, which was based at the time on professional institutions. 22 The existence of
alternatives which, as a result of their competitive nature, do not suffer from these
drawbacks makes it inadvisable, however, for regulation of the auditing profession
to be based on this strategy.23 Its use could be considered hypothetically to establish
a minimum level of quality in those service segments in which such alternatives
have few possibilities of functioning or of doing so with adequate effectiveness.
This could be the case, at most, with the segment of individual auditors and very
small firms. It is important to make it clear that, if this possibility is adopted, the
need for separate regulatory patterns for the different types of firm should be very
much taken into account, in order not to generalize the problems inherent in this
possibility. In addition, it should be borne in mind that this strategy would nowa-
days encounter substantial difficulties when applied in many countries as a result of
the structure of professional auditing services. This would be the case where there
is a substantial reserve of certified auditors not employed as such. 24
2.2.2.3 Reinforcement of regulation
The final high-cost strategy to consider consists in reinforcing regulatory supervi-

sion. Costs involved in this possibility are similar to those of the previous one, but
enhanced by those involved in actual management of the regulatory process and the
high risk that the results are not just inefficient but also ineffective: it would not
22 The first professional associations to assume functions of verifying quality in the accounting field
appeared in Scotland in 1854, England in 1870 and in the United States in 1887 (the new term certified
public accountant [CPA 1was coined to prevent damaging the reputation of British chartered account-
ants). The British association arose to safeguard quality in functions relating to bankruptcy and
insolvency (Watts and Zimmerman, 1983, pp. 630-2). In Continental Europe, the creation of these
institutions or associations is a much more recent phenomenon.
23 Donabedian (1993) analyzed how the accountancy profession in the United States seems, since the
I 970s, to have experienced a certain deterioration in the economic factors which contribute to generat-
ing incentives to perform, particularly a reduction in (opportunity) costs in abandoning the profession.
His analysis of the problem seems unsatisfactory, however, for at least two reasons. Firstly, it ignores
the organizational dimension and treats the auditor as an independent professional or individual, thus
not considering the primordial effect of reputation and the internal control mechanisms of firms. Sec-
ondly, it puts forward overall conclusions from an excessively fragmentary analysis of the different
changes.
24 In Spain, for example, over 40 percent of those qualified to work in the profession were not doing so
in 1995 (according to "Situacion de la Auditoria en Espana" BOlCAC, nos. 5, 9, 13, 18,21 and 25).
only provide a poor cost-benefit ratio but there is also even room for doubt that
quality would actually be improved.
The effect that this strategy would ideally have can be shown in the terms of the
model set out in Section 2.1.3. To this end, it suffices to take qo of the model as
representing both a level perceptible by the purchaser as well as a level assured by
third parties. From this point of view, reinforcing regulatory supervision can be
interpreted as raising the minimum level of quality, from qo to q'o (Figure 2.4),
achieved in exchange for introducing regulations which give rise to an increase, R,
in the production cost. Two situations are shown in the Figure. If the costs incurred
by the regulator to guarantee the minimum quality level are nil, production of qual-
ity q will be sold at a price qb, less than the original, qc. In the more realistic case
in which the regulation is costly, the price could be higher or lower than the original
depending on whether the fum produces at quality levels higher or lower than that
of point d where the supply curves associated with qo and q '0 cross.
p p(q) given q'o, R > 0

p(q) given qo
e(q)+R
e(q)
Figure 2.4. Effect of the regulator raising the assured minimum level of quality
2.3 The "last period" problem and its consequences for

auditing quality
It could be thought that the contractual safeguard based on receipt of quasi-rents,

and on repetition in general, ceases to function when the time horizon is shortened.
To understand the problem, it suffices to wonder what incentive the auditor would
have to provide high quality in the last period, t. If he lacks incentives to provide
good quality, the client will anticipate this possibility and will not contract him in
year t, and therefore year t- J becomes the fmal period and, as a result, the incentive
for quality in t- J disappears, and therefore he will not contract him in this year and
so on successively down to the fIrst year. This problem will be avoided at root if
the relationship is of infmite duration. In practice this is obviously not possible.
Nevertheless, this effect of infmite duration is approached by different routes con-
sisting in making the duration indefmite (a); the contracting is repeated with third
parties and the latter are informed of the degree of performance of the prior transac-
tion (b); making the survival of the business independent from that of its creators
(c); and, fInally, on the basis of the information asymmetries regarding the cost and
propensity to perform (d).
a) Indefinite duration. When it is the same parties who repeat, they have more
interest in performing if they do not know the date on which their relationship will
end. In the case of auditing, relationships are contracted annually (probably in
order to encourage auditor independence, as indicated in Section 1.3), but the dura-
tion of the relationship is indefmite except when a legal rule dictates otherwise, as
occurs under mandatory auditor rotation systems which have been introduced in
some countries. 25
b) Repercussion on contracting with third parties. When non-performance af-
fects the possibilities of contracting with third parties, the effectiveness of
reputation becomes independent of continuity of a specifIc relationship, it sufficing
that there is continuity of presence in the market of the contracting party obliged to
perform. In the auditing fIeld, this solution again raises the importance of client
diversifIcation for proper functioning of reputational safeguards.
c) Time scale horizon. When the contracting parties are individuals, their ex-
pectations depend on their life expectancy and this is subject to unavoidable
biological limitations, which tend to heighten conflict as the time scale horizon
shortens. Ifbusinesses are legal entities, this problem is solved and continuity is not
subject to restrictions of this type. The structure of auditing is also affected by
using this solution since it provides a substantial comparative advantage to fIrms
with several auditors compared with those based on an individual auditor or a few
individuals. It is in fact fairly normal for small fIrms which were once successful
(while their founders continued to have a sufficiently long future) to overcome their
generation crises by being taken over by larger fIrms.
25 For an economic analysis of the largely negative consequences of the mandatory auditor rotation rule
in terms of cost, quality and competence, see Arruiiada and Paz-Ares (1997). In Spain, mandatory
rotation, which was introduced in 1988, was repealed in 1995 before the first obligatory rotation oc-
curred. In italy, it is still in force at the time of writing, although there are moves to repeal it.
d) Informational asymmetries. Finally, it has been shown that, even in a context

of a finite time scale, reputation functions effectively as a safeguard if purchasers
are unaware of the seller's objectives or technology (the party generally obliged to
perform), and attribute a certain positive probability that the seller will provide them
with high quality, whether because it is just as costly for the latter as lower quality
or because he has an honorable nature leading him to fulfil his promises. 26 In the
case of auditing, the technology is relatively well known, but not the amount of
quasi-rents. Moreover, there will generally be doubts regarding the advantages and
drawbacks of providing different levels of quality, which reduces the seriousness of
the final period problem.
2.4 The regulation of quality: Attainable objectives
The aim of this section is to provide some guidance in relation to the general defini-
tion of regulatory objectives in terms of auditing quality. As a first step we will
pause for a moment to examine the direction in which the most experienced regu-
latory body is moving (Section 2.4.1). This will highlight the most important
problems which the regulator must help to solve in this field. He has two pertinent
courses of action, consisting in exerting a facilitating function and adopting a safe-
guard strategy. On the one hand, the regulator should carry out his work indirectly,
facilitating sanctioning, and therefore motivating activities of the market and judges
(Section 2.4.2). On the other hand, the regulator must take account of the superior-
ity in terms of quality of a guarantee based on developing safeguards which enables
advantage to be taken of economies of scope of a contractual nature rather than on
fragmentation, which suffers from diseconomies of both scale and scope (Section
2.4.3).
2.4. 1 Independence failures in professional practice
In order to understand where the problems of auditor quality and, in particular,

independence lie, it is worthwhile reviewing the enforcement actions brought by the
oldest regulator in this field, the United States Securities and Exchange Commission
(SEC). To this end, in a comprehensive study, Campbell and Parker (1992) exam-
ined 132 measures taken by the SEC in relation to auditors between 1972 and 1989
from its Accounting Series Releases and Accounting and Auditing Enforcement
Releases.
26See the works of Kreps and Wilson (1982) and Milgrom and Roberts (1982a and 1982b) or, for an
introduction, Tirole (1988, pp. 124-6).
In this study, the data regarding the location and nature of bad professional prac-
tice stand out particularly. Firstly, in 25 of the 26 cases in which fraudulent
behavior was shown on the part of auditors, the guilty fInns were local or regional.
Furthermore, the fraudulent behavior generally amounted to recklessness in con-
ducting the audit, without knowledge of the fraud itself (constructive fraud). On the
other hand, fraudulent clients seemed capable of deceiving their auditors, even in
the case of large firms. In a different set of 20 cases the SEC did not fmd auditor
fraud, but a breach of generally accepted accounting principles occurred in connec-
tion with fraudulent behavior on the part of the client. In these cases the auditors
had been deceived by their client. They were guilty of negligence, however, since,
in the view of the regulator, they would have detected the deceit if they had exer-
cised due care in the conduct of the audit. Significantly, in 19 of these 20 cases the
auditor belonged to a large firm (this data clearly seems open to interpretation in
several ways). The SEC attributed the cause of the remaining 106 non-fraudulent
failures to the fact that the auditor erred in applying professional standards, often
basing his judgement on inadequate information and trusting excessively in that
provided by the client.
2.4.2 Regulation as a facilitator of judicial and market controls
Reflection on the role of regulation and the action of regulators in the auditing field
must take into account that two types of control operate on the auditing profession:
legal, via civil and criminal liability, and the market, mainly via reputation. Regu-
lators can opt for an indirect intervention strategy consisting in facilitating the work
of these two controls, or by a strategy of direct intervention. Most probably, the
indirect strategy is more effective, flexible and adaptive, and therefore should playa
preferential role. In any case, it is clear that the regulator must concentrate his ef-
forts in areas where he has a comparative advantage. On examining the different
possibilities in this respect, the functions of regulators (and, to the desired extent,
those of self-regulatory bodies) of auditing can, amongst others, cover the following
stages or tasks:
a) Regulation as a protector of reputation. The effectiveness of the reputational
guarantee derives from its high degree of sensitivity. This characteristic involves a
danger when firms with reputation may be victims of opportunistic or arbitrary
action. For this reason, one essential role of regulation is to maintain proportion
between breach of professional duties and reputational sanctions. Above all, firms
should be prevented from suffering sanctions of this type when they have not failed
in their professional duties. Those elements of the institutional framework which
occasionally allow both competitors and in particular supposedly independent third
parties from causing unwarranted hard to reputation by means of frivolous actions
or taking precipitate precautionary measures in some cases, should be reviewed. In
a more general way, judges and regulators in this field should be very much aware
that a good commercial reputation is surely the most valuable, and of course the
most sensitive asset of firms of the highest quality.
b) Precaution in the definition of quality standards. In the case of auditing, the
existence of minimum quality standards seems essential for the explicit safeguard
mechanisms to function effectively, principally that providing for professional li-
ability. A standard is in fact needed to establish which particular conduct
constitutes diligent action and which constitutes professional malpractice. 27 Stan-
dards are also necessary in order for professionals to know what constitutes an
adequate level of quality. In this respect, the fundamental caution arising out of the
analysis developed in Section 1.2.2 should be borne in mind, however: the legisla-
tive framework must avoid the temptation of defining overly stringent standards
which lead to defensive audits of little informational value. Moreover, the greater
the relative importance assigned to implicit safeguards, the less will be the role of
legislative standards and the greater the role of firms' internal ones. Furthermore,
the function of the latter is different, insofar as the market's judgement of profes-
sional conduct is not based on compliance with standards but on the results.
c) The production of information relevant to market functioning. Regulation can
considerably facilitate the stream of information useful to the functioning of the
control and sanction mechanisms which operate in the open market. One possibility
consists in making it obligatory to disclose information which private parties are
often reluctant to provide because voluntary disclosure generates external effects.
By providing more information to the market the latter thereby becomes more
transparent and adjustments and thus the sanctions which the market itself imposes
on operators who reduce quality are more rapid. Likewise, the regulatory body is in
a good position to act as a central recipient of information regarding the sector, both
of a statistical and monitoring nature. In particular, it has the advantage in acting as
a verifier and depository of information relating to compliance with legislation, the
disclosure of which to the market could compromise competition between firms,
and which is easily controllable both directly by the regulatory authority and indi-
rectly (in which case, the reliability of the information is only verified in the case of
inspection or litigation).
d) Direct monitoring of compliance with quality standards by the regulator or
self-regulatory bodies has a major advantage-understanding of a discipline of a
professional and specialized nature, which favors control being carried out by the
27The results of the empirical study by Carcello and Palmrose (1994) suggest. for example. that the
supervisory activity of the SEC facilitates litigation against auditors.
experts themselves. Nevertheless, the risks are also considerable. Control by a

public regulator is capable of being ineffective and very costly, given the charac-
teristics of previous administration and experience in other sectors in many
countries. On the other hand, both procedures, particularly self-regulation, can
easily succumb to the temptation of encouraging anti-competitive practices. For
these reasons, priority should be given to facilitating the functioning of the other
control mechanisms of both a legal and market nature.
2.4.3 Fragmentation versus safeguard regulatory strategies
There are considerable problems in contracting professional services as a result of

the substantial informational asymmetries between supplier and client. This infor-
mational inequality usually generates conflicts of interest of all types. To overcome
these problems, two possibilities can be put forward, at a theoretical and highly
general level: which will be referred to as fragmentation and safeguard strategies:
a) The fragmentation strategy functions by assigning the different stages or
components of contractual processes to different professionals, trusting that by
acting in opposition, conflicts of interest will be reduced. It thus seeks to contain
the conflict of interest which gives rise to the problem, and it can therefore be ex-
pected that it will ideally lead to lower safeguard costs. On the other hand, the
economies of scale and, in particular, economies of joint production (or "scope")
which are usually generated when the same professional is involved in several
stages, or plays several roles in the same contractual process, must be sacrificed.
Joint audits, which are compulsory in various European countries,2s is an extreme
example of fragmentation strategy: it aims at improving independence by superim-
posing two parallel checks. Mandatory rotation also involves something of this
strategy, although the aim thereby is sequential rather than parallel verification. An
important problem of both types of regulation is obviously in the cost: rotation
duplicates the start-up costs for each rotation and pure joint auditing duplicates
virtually all costs.
b) The alternative safeguard strategy seeks to eliminate the consequences or
manifestations of conflicts of interest rather than their origin, to which end the in-
centives of professionals need to be reinforced in order that they respond correctly
to all eventualities. This therefore involves greater safeguard costs but there is the
benefit of the advantages of economies of scale and, particularly, economies of
scope. In this area, for example, very different solutions are adopted in different
countries in real-estate transactions. The fragmentation strategy dominates in the
28 Specitically, different variations of joint auditing are required in Denmark, Finland, France, Italy
(distributing the work in part) and Sweden, according to Buijink et al. (1996, pp. 41-2). Joint audits are
also required for specific firms in some other countries (e.g., for banks in Canada).
Anglo-Saxon world, in which each party is represented by his own lawyer. In Con-
tinental legal systems, on the other hand, the figure of the Latin or Roman-
Germanic notary to a large extent plays the role of lawyer simultaneously and
independently for the two or three parties usually involved in transactions. 29 Even
more clearly, the Spanish registration system embodies certification functions in the
Registrar which in other neighboring systems are carried out by a costly process of
notification and defense by the parties affected themselves (Germany). Also, in
those countries which, like the United States, maintain the old "recording" system
of simple deposit of documents to give effect to transactions, even longer and more
costly sequences of intermediaries and title insurance need to be employed. 30
In summary, a guarantee of quality in transactional services can be sought in

fragmentation and confrontation or in the safeguard of integrated action, which is
not only generally more economical but which can also provide higher quality, for
two reasons. Firstly, it enjoys economies of joint production, of both a technologi-
cal and contractual nature. 3 J On the other hand, the reduction in conflicts of interest
which fragmentation aims to achieve is not always implemented. Application of
these ideas in the auditing field is closely connected with the provision of non-audit
services. A prohibitive rule would place us in the costly area of fragmentation. On
the other hand, later chapters argue that a simultaneous provision of auditing and
non-audit services will enable higher quality to be provided at a lower cost.
29 See Arruiiada (1995 and 1996).

30 See Arruiiada (1998 and 1999).
31 Taking advantage of economies of scope is the basis of the so-called "gatekeepers", taken as guardi-
ans of the law, developed by Kraakman (1986), who defines them as private agents "who are able to
disrupt misconduct by withholding their cooperation from wrongdoers [.... This cooperation or] sup-
port-usually a specialized good, service, or form of certification that is essential for the wrongdoing to
succeed-is the 'gate' that the gatekeeper keeps" (pp. 53-4). Whether the auditor's work corresponds
faithfully with this figure essentially depends on the importance given to his role as producer of "exter-
nalities", those economic effects which are external to the parties involved in the transaction.
3 AUDITOR INDEPENDENCE
AS AN ECONOMIC DECISION
This Chapter examines the economic factors which weigh on auditor independence.
When issuing his report, an auditor's decision is influenced by a wide range of
considerations, amongst which his professional and personal ethics usually play a
fundamental role. A positive analysis of these personal and professional ethical
reasons is particularly difficult, however. Moreover, it is in any event desirable that
there should be incentives which lead in the right direction even in the absence of
such ethical principles. For these reasons, attention will focus here on analyzing the
economic incentives weighing on this decision to ascertain the circumstances in
which such incentives will ensure correct action.
3.1 Concepts of auditor independence
Independence was defined earlier as the auditor's willingness to reflect defects and
problems which he has detected in the financial statements reviewed in the audit
report. 1 This concept of independence corresponds to what is usually called inde-
pendence in fact, taken as an absence of interest or influences which could prejudice
the auditor's objectivity and which, given their mental or psychological nature, are
not directly observable. In professional and regulatory circles, considerable impor-
tance is given to the distinction between this independence in fact and independence
in appearance, defined by signs, signals or indicators which are in fact observable?
I A large part of the debate on auditing has centered around independence, perhaps because it is "diffi-
cult to prove and easy to challenge" (Mednick, 1990, p. 86). It is thus understandable that defining
independence has also attracted considerable attention in both the academic literature and works of
professional associations and legislators. See, for example, Mautz and Sharaf (1961), Nichols and Price
(1976), Rittenberg (1977) and, for a more formal treatment, Antle (1984).
2 To the Federation of European Accountancy Experts (known as FEE, the Federation des Experts
Camptables Europeens), independence includes "independence of mind", understood as "the state of
43
The difference between "objectivity" and "independence" should also be empha-

sized: objectivity is an attribute which relates to results whilst independence takes
the form of a prior condition. For example, professionals other than auditors are
usually required to be objective rather than independent. When this occurs, their
requirements are, as a consequence, less strict than those applied to auditors. The
effects of providing non-audit services on this perceived independence will be dealt
with in Section 5.2, concentrating here on independence in fact. (It should be made
clear that, for the sake of brevity, the latter will be referred to simply as "independ-
ence" in this Chapter).
3.2 Economic factors behind auditor independence
In analyzing auditor independence a situation will be taken as a reference in which

it is assumed that there is disparity between auditor and client regarding whether the
auditor should or should not include a particular matter in his audit report, which
could lead to either a qualified report or the impossibility of expressing an opinion.
In a case like this the client may threaten to change auditor and the latter may be
motivated to accede in order not to endanger the continuity of his relationship with
the client.
The aim is to examine the forces on which the auditor's decision depends when
facing a delicate dilemma in terms of independence. It will be assumed that, when
weighing the pros and cons of his decision, the auditor will, along with the effects
which his decision could have on his reputation and liability, consider the cost of
possible loss of the client to him, facing both an explicit or implicit threat of break-
ing the relationship. In order to proceed in an orderly fashion and make the
fundamental assumptions explicit, a certain minimal amount of formalization is
necessary, although more for explanatory rather than analytical purposes. As a first
step, the value of the audit firm can be divided into several components in such a
way that its assets and liabilities are represented by the following equality:
v+Q+P=L+E [7]
in which each letter represents an element of assets and liabilities defined with a
view to subsequent analysis: v is the value of the quasi-rents (remuneration in ex-
cess of opportunity cost) associated with the client whose audit is subject to
scrutiny, which will be referred to as individual quasi-rents; Q represents the value
mind which has regard to all considerations relevant to the task in hand but no other" and "independ-
ence in appearance", consisting of "the avoidance of facts and circumstances which are so significant
that an informed third party would question the statutory auditor's objectivity. When independence is
dealt with in laws and professional rules, it is only independence in appearance which is addressed"
(FEE, 1996, p. 24).
Auditor independence as an economic decision 45
of the assets which are specific to one or more other clients ("specific" here means
that they would lose their value if the relationship is terminated, and are thus equal
to the present value of the corresponding quasi-rents); P is the value of the assets
specific to the firm (their value will disappear if the fum becomes insolvent); L is
the cost in terms of present value of all possible professional liability contingencies;
and, finally, E is the net value of the firm to its partners (this item obviously in-
cludes the net value of all other assets and liabilities not detailed in the other
variables).
Several qualifications should be mentioned at this point: (a) Firstly, in computing
the value of assets specific to current clients, Q, the cost of possible liabilities is not
included, and they are included separately in the variable L. (b) Both Q and P rep-
resent the economic values of separate sets of quasi-rents and not the gross turnover
associated with the corresponding sets of clients. (c) The value of the assets specific
to the firm, P, can be taken as the net economic value of all potential contracts of
the firm, including those which could replace current contracts. The auditor,
through contracts currently in force, receives quasi-rents which remunerate his
investment in assets specific to both a particular client and to the firm as a whole.
Examples of the first type are start-up costs and learning costs of client characteris-
tics. Examples of the second type are the investment necessary to develop effective
auditing methods and auditor training. What is relevant is that assets of the second
type are not specific to any client/ but to the firm. As a result, P includes the value
of contracts which could replace current contracts.
Let us denote ;r as the probability that the client's fmancial situation deteriorates
in such a way that the auditor would be consequently involved in a cause celebre,
with negative consequences in terms of both reputation and liability. Similarly, 7]
denotes the probability that the client will change auditor as a response to the lat-
ter's refusal to accede to his desire to compromise his independence.
Furthermore, kiJ (i = 1,2,3,4; J = v, Q, P, L) denominates the impact which dif-
ferent decisions and events have on the value of the four non-residual components
of the audit firm's balance sheet. These impacts will be considered as a proportion
of the value of the corresponding asset which is lost as a result of degeneration or
otherwise of the situation of the client in respect of which the auditor has or has not
reported correctly.4
3 They may occasionally be so when losing a client results in excess capacity, as mentioned in Section
3.3.2. This problem will be all the more serious the more that possibilities for moving resources be-
tween audit activities and providing non-audit services are limited.
4 It should be recalled that this. formulation is merely for explanatory purposes, and thus does not rule
out the possible existence of non-linear effects despite using linear expressions; similarly, it is assumed
that there are no interdependencies between the asset and liability blocks.
Auditor
X
does not k3vv+k3,.Q+ k 3'P P + k31L
'
report /
The situation of the

client degenerates
with probability tr
]-T[
~ k4v V + k4!1 Q + k4p P + k41 L
Figure 3.1. The auditor's decision on the desirability of reporting or not an anomaly
detected in a client's financial statements
The auditor's decision can now be fonnally represented by means of a decision

tree as shown in Figure 3.1, in which the boxes represent decisions and the circles
random events (changing auditor is in fact a client's decision which can at the same
time be considered a random phenomenon from the auditor's point of view). It is
reasonable to assume that a rational auditor who is risk-neutral (this assumption is
relaxed and its importance explained in Section 3.4.4) wishes to maximize the value
of his net worth, E, which is equal to:
E=v+Q+P-L [8]
To this end, he will choose the course of action which minimizes the total antici-
pated cost. In general terms (Figure 3.1), if the auditor reports the anomaly, he runs
a certain risk that the client will change auditor. This occurs with probability 17 and
involves the loss of assets specific to that client and a certain positive effect on the
auditor's reputation. If the client does not terminate the relationship, the conse-
quence of reporting is simply a positive effect on reputation, which will probably be

less than if the termination occurs.
In algebraic terms, it can thus be expected that klv = 1, klq < 0 (having been de-
fined as costs, the negative sign transforms this effect into a benefit), kip < 0, kJl =
k2v = k21 = 0, k1q < 0 and k1p < O. In addition, it is also foreseeable that k lq < k2q and
kiP < k2p, since if the termination in the relationship occurs it will be more informa-
tive of the auditor's independence. The consequences of the auditor not reporting
depend on how the fmancial situation of the client evolves. 5 If it worsens, the
auditor will suffer losses in all elements of his balance sheet: it can be assumed that
assets specific to the client (k3vV) will be devalued or disappear and that the value of
his assets specific to other clients (k2q Q) and to the firm (k3j') will be diminished.
Moreover, his professional liability (k4IL) will increase. On the other hand, it can be
expected that if the client does not have a fmancial crisis, the effect on all these
asset and liability elements will be minimal (k4j = 0).6
In the following analysis it will be assumed that a large part of these effects are
relatively secondary and can be discounted. Specifically, only those represented in
Figure 3.2 are considered sufficiently relevant to justify specific analysis. In this
context the rational auditor will opt for reporting or not reporting depending on
which of the two possibilities maximizes the net value of his firm which, assuming
neutrality towards risk, leads him to choose the possibility that presents the lowest
anticipated cost. On the one hand, the expected cost of reporting, which will be
called the "independence" cost, is equal to the probability that the client will termi-
nate the relationship multiplied by the expected value of the individual quasi-rents:
C(report) = 1'/ v [9]

On the other hand, the expected cost of not reporting, the "dependency cost", is
given by the probability that the client's situation will degenerate multiplied by the
sum of the losses which the auditor would suffer in this case. These losses are asso-
ciated with both the client himself and other current clients, as well as other
potential or future clients, as well as the increase in civil liability. Eliminating the
first subscripts of the coefficients kij, since they have become unnecessary, the ex-
pected cost of not reporting is given by the following expression:
C(not reporting) = 7r (v + kqQ + kj' + kl-) [10]
5 The studies by Palmrose (I 988a, 1991) support consideration of this situation as reference since they
show that low audit quality is normally associated with client insolvency.
6 The diagrammatic representation in Figures 3.1 and 3.2 should not be interpreted as indicating that
the auditor change and client evolution nodes are simultaneous. On the contrary, it could in general be
expected that client continuity will be decided in a shorter period than the latter'S financial evolution.
As a result, the auditor will report and show himself to be independent provided
that the independence cost is less than the dependence cost:
[11]
and will not report otherwise. By moving all elements relating to the client under
scrutiny to one side of the previous expression, the following independence condi-
tion is also found:
[12]
Let us next examine, in general tenns, each of the addends of the previous ex-
pression. In Section 5.1 the effect of providing non-audit services on each of them
will also be examined.
Loss of quasi-rents
A
~~V_____'_S_~C_i_mOO
__w_it_ht_h_eC_li_ffi_t__~
I]
C!:::~::~es
probability ~
I-I]
Audi tor ""

reports ""
Less of quasi-rents associatedwith the client

Lo~ of q.aasi-rents associated
wi1h current dients
LClis of finn-specific assets
Aud i tor I LCfiiS fran civil

Iiabil ity
X
does not / I
report Ie" v + kq Q + Ie" P + k, L
The silll.tDn oflbe

client degffieratt:s
with probability"
'"
I-;r
Figure 3.2. Simplified version of the auditor's decision as to the desirability of

reporting or otherwise an anomaly detected in a client's financial statements
3.3 The cost of independence: the loss of client-specific

assets
The cost to the auditor of the independence option basically originates in the possi-
bility of losing the specific assets associated with the particular client. This expected
cost can be evaluated as the product of the value of these assets, v, multiplied by the
probability of terminating the relationship, n.
3.3. 1 Volume of quasi-rents to be received by the auditor
The value of the assets specific to a client equals the present value of the quasi-rents
which the auditor will receive in the relationship (v), which in tum depend on the
value of all the quasi-rents generated in the relationship and how these are shared
between auditor and client. Let us examine each of these two elements separately.
3.3.1.1 Total volume of quasi-rents to share between auditor and client
In an audit relationship, there is a considerable stream of quasi-rents associated with

the use of assets specific to the relationship, which have a value in their context but
are useless out of it. This is the case with all types of knowledge which the parties
have accumulated over time, beginning with the substantial investment necessary to
carry out the first audit of any business. 7 (The model developed in Section 3.3.4
will show that the total value of quasi-rents is equal to the difference between the
cost of the initial audit, represented by C, and of recurring audits, c).
3.3.1.2 Parties' bargaining power with respect to the division of quasi-

rents
In general, the appropriation of quasi-rents associated with specific assets by one or

other party in a commercial relationship depends on their ex post bargaining power,
independently of which of them may have previously incurred cost or expenditure
in producing the specific asset (in the model in Section 3.3.4 the bargaining power
of the parties will be reflected by the parameter a, which is the proportion of quasi-
rents received by the auditor). If the parties correctly anticipate the future situation,
they will simply fix the prices and other terms and conditions so that, when consid-
ering the consequences of this bargaining power and the subsequent attribution of
quasi-rents, the exchange is profitable to both.
7 See, on this aspect, Arruftada and Paz Ares (1997, pp. 32-5).
Given a certain stream of total quasi-rents or, which is the same thing, a certain
volume of specific assets, their appropriation by each party-auditor and client-
depends on their bargaining power and the mechanism used to fix the price. In
these cases economic models normally introduce an assumption of equal bargaining
power, according to which the quasi-rents are distributed in equal parts between the
two contracting parties. s Curiously, on this point there has been a certain confusion
in accountancy literature, because some authors appear to have taken what at the
time was no more than a working hypothesis as a real pattern. The problem arose
with introduction of an extreme assumption, attributing all quasi-rents to the audi-
tor, in the original work of DeAngelo (1981a). The author was clear, however, in
stating that "for simplicity only, assuming that the incumbent auditor sets F [the
initial price] in order to extract the maximum entry-preventing quasi-rent" (l981a,
p. 121, italics added). The reason for her assumption was that in that work attrib-
uting them wholly to the auditor was of no importance, because DeAngelo was
trying to model the initial price discount, or "lowballing" and not how this discount
might affect independence. The pattern followed in dividing quasi-rents, on the
other hand, can be critical from the point of view of independence. To this end, an
assumption which in DeAngelo did not give rise to a loss of generality is more im-
portant in models aimed at analyzing independence and its inappropriate use can
even invalidate analyzes such as those of Beck, Frecka and Solomon (l988a), as
will be mentioned in Section 5.l.3.b.
Because of the importance of this point this problem of allocating quasi-rents
will be now examined in some depth, analyzing it from both the theoretical (a) and
empirical (b) points of view and then assess the relative role of the individual quasi-
rents (c).
a) Theoretical view. The point in question has been brought to light since the
works of Magee and Tseng (1990), Dye (1991) and Kanodia and Mukherji (1994):
• Magee and Tseng (1990), on the one hand, show that if the auditor possesses all
the bargaining power and there is no disagreement between auditors on how to
interpret generally accepted accounting principles, clients would have nothing to
gain by threatening to cancel their relationship with their auditor and, conse-
quently, the quasi-rents would not compromise the latter's independence.
• Dye (1991) also rightly pointed out that the assumption regarding ex post rela-
tive bargaining power is crucial in determining the extent of ex ante lowballing
(ex ante and ex post meaning here initial and recurrent audits). His model tries
to endogenize the ex post bargaining relationship between auditor and client by
considering the effect of auditor quasi-rents on the perception of auditor inde-
pendence by external observers. The attempt is incomplete, however, because it
8 As in Grout et al. when dealing with this same problem: "For simplicity, assume that the client firm
and auditor have equal bargaining power and split the surplus" (1994, p. 329).
considers only one audit relationship. As a consequence, observable auditor

quasi-rents exert an unambiguously negative effect on the perception of mde-
pendence. The allegedly negative effects of lowballmg are then reconciled with
its existence by claiming that quasi-rents are not observable by outsiders and
their only purpose is to reduce mdependence and so offset the negative expecta-
tions of mtelligent outsiders. The effects that quasi-rent appropriation by the
auditor exert on the value of the clientele of the fmn as a whole are also impor-
tant, however. Its consideration opens the possibility that auditor quasi-rents
might favor mdependence as discussed m Section 3.4.1. Applymg Dye's rea-
sonmg, lowballmg could then even serve just the opposite purpose: to raise real
auditor mdependence to the level expected by outsiders (the verifiable predic-
tions of Dye's model also seem mconsistent with the apparent mcrease m
discounts as a consequence of fee disclosure).
• Lastly, Kanodia and Mukherji (1994) argue that for the mitial price discount to
compromise mdependence, it is necessary for the client to have a certam ex post
bargammg power and be able to remove the auditor, m which circumstances it is
not clear that lowballmg needs to exist m the mitial audit. In other words, low-
ballmg ex ante can only be generated if the auditors receive quasi-rents ex post
and these are incompatible with the client havmg all bargammg power. Kanodia
and Mukherji (1994) show that quasi-rents and lowballmg may even exist, how-
ever, when clients have bargammg power superior to auditors. To this end, they
model a situation m which it is the client who specifies the pricmg mechanism
(by means of an offer that can be interpreted as of the "take it or leave it" type)
and both nonetheless enjoy quasi-rents on the basis of recurrent audits, which
generate lowballmg m the initial audit. The auditor can only accept or reject the
client's pricmg offer. Nevertheless, m the model the auditor has an mforma-
tional advantage over the client smce it is assumed that the latter is not accurately
aware of the true cost of recurrent audits and that, moreover, the client will sus-
tain rotation costs each time there is a change m auditor. Under these conditions,
some auditor turnover results as the client tries to exploit the competition in the
market for auditors, but it is less likely over time.
b) Empirical view. It would be useful to have solid empirical evidence on this

pomt. Unfortunately the few existmg indications are either very weak or simply
questionable:
• First, circumstantial evidence, such as the desire of clients to continue with their
auditor, illustrated for example by their opposition to mandatory rotation rules or
by the very long duration of the relationship, could be mterpreted as a symptom
that they are receiving quasi-rents from the relationship.
• Along the same lines, the fact that both parties, auditor and client, have an in-
formational advantage regarding some elements of the transaction leads us to
think, in the light of the economic theory of information, that both have a certain
bargaining power. Firstly, the auditor knows his own costs better than the client,
even the start-up costs of the client with the new auditor, whilst the client is bet-
ter aware than the auditor of the price which can be obtained from a new auditor
and the value to him of his audit being carried out by one or other firm.
• The extent of initiallowballing could also provide indirect evidence of how it is
anticipated that quasi-rents will be divided ex post in recurrent audits. Nonethe-
less, empirical evidence in this respect is of little help. Firstly, different
institutional and market environments appear to provide different incentives.
Thus, in the United States the existence of introductory pricing in the initial
years of the relationship has been empirically verified (Simon and Francis, 1988;
Ettredge and Greenberg, 1990).9 However, in Australia, an initial study found
prices in excess of the cost of the initial audit (Francis, 1984), perhaps associated
with a greater risk (Roberts, Glezen and Jones, 1990), and more recent studies
detect lowballing, but only when the client upgrades the quality of its audit and
begins to be audited by one of the big firms (Craswell, Francis and Sneddon,
1997). On the other hand, the existence of lowballing provides very limited in-
formation since it only means that the auditor enjoys quasi-rents ex post, but not
that he enjoys them to the point where he appropriates all the quasi-rents from
the relationship, a point at which the client would be indifferent to changing
auditor or not.
• Finally, attempts to throw light on the question from the point of view of ex-
perimental economics have been fruitless. It is known, in relation to similar
problems, that the outcome depends-as indeed predicted by the theoretical
models-on how the procedure followed for fixing prices is simulated (who
makes the offer and the existence or otherwise of counter-offers, etc.) and on the
information revealed to the parties regarding the cost and value of the transaction
to themselves and others. As a result of this dependency, and because unreal
conditions were assumed, the results obtained by Schatzberg (1990) in his at-
tempt to test the DeAngelo lowballing model are not applicable to these effects.
They cannot be used to elucidate the problem under discussion here because the
experiment is aimed at demonstrating lowballing and not the pattern used in di-
viding the quasi-rents. To this end, he simulates the market with the
characteristics assumed by DeAngelo in terms of competition and cost structure
under different learning and switching cost conditions associated with the first
audit of a client (classified as "transaction costs" in the article). His results show
9 On the other hand, many studies did not find systematic discounts in initial audits. This result is of
little interest, however, because they were unable to reject the hypothesis of no fee cutting.
that the auditor receives all the quasi-rents from recurrent audits but discounts
the initial price to the point where he obtains a nil benefit in terms of present
value. The experiment does not tell us, therefore, anything about distribution of
the quasi-rents because it in fact implicitly assumes their total allocation to the
client. It does so because of the manner in which it simulates bargaining as to
the price of the recurrent audit since in the experiment it is the auditor who pro-
poses the price, whilst the client can only accept or reject it without making a
counter offer. It is known that this type of price fixing forces a solution which
attributes all bargaining power to the person making the offer. The reason is
that, since counter-offers are not possible, the problem of credibility or self-
commitment which the person formulating the first offer otherwise suffers from
is resolved. In the way he simulates price fixing, the result is therefore implicit
that the quasi-rents will be appropriated by the person making the offer who, al-
though in this experiment is the auditor, could have been the client, with the
result that it would be then expected that the very opposite outcome would
emerge in terms of appropriation of quasi-rents. 10
c) The excessive emphasis on individual quasi-rents. When considering that the

existence of specific assets affects auditor independence only as a function of pro-
viding the auditor with quasi-rents associated with the client, their actual effect is
being unduly exaggerated for two reasons. Firstly, as we have just examined, the
compensatory power of the fact that the specific assets also provide the client with
quasi-rents must be taken into account, which reinforces interest in not changing
auditor. Secondly, the compensatory power of the quasi-rents themselves should
also be restated here, even though this was already mentioned when discussing the
Dye (1991) model and will be examined in more detail in Section 3.4.1: a higher
volume of individual quasi-rents associated with one client also means a higher
volume of quasi-rents associated with other clients of the firm. For this reason, the
receipt of quasi-rents in itself has an ambiguous consequence in terms of independ-
ence. This effect is only clearly negative if the auditor has a single client, but it is
potentially positive if he has many and these, because they demand independent
audits, require their auditor to be independent with all his clients.
3.3.2 Specific assets caused by excess capacity
To the auditor, terminating his relationship with a client involves losing the assets
specific to that client. This always includes those assets which in no event have
10 See, for example, Roth (1995) for a description and analysis of the available evidence on bargaining
experi mentation.
value outside the relationship, such as all knowledge relative to the client. In addi-
tion, when the audit firm has excess capacity, other assets which in principle are not
client-specific can also become inactive or be under-utilized as a result of the client
terminating the relationship. 11 This would be the case with human capital not spe-
cific to the client but which is specialized and is therefore specific in relation to the
sector or activity.
The second possibility suggests two reflections with consequences of regulatory
interest: firstly, since it is more likely that it will arise with clients representing a
large proportion of the auditor's workload, it again reaffIrms the importance of
diversification rules as a quality safeguard. Secondly, and due to the fact that the
diffIculties in relocating resources are greater the narrower the range of services
offered by firms, the more specialist firms will be more affected in terms of both
services and sectors.
Analysis of the effect of the presence of client-specific assets on independence
also suggests that independence may be subject to cyclical variations connected
with general economic evolution. It can be expected that in times of economic
recession two phenomena will apply which in principle prejudice independence.
Firstly, a large number of clients will pass through fmancial diffIculties. It may
therefore be thought that a large number of clients will also ask their auditors to
relax standards (this effect is probably offset, at least in part, by the expectations of
users of financial statements who, aware that they are passing through a recession,
will be more likely to tolerate poor performance). Secondly, demand for audit firms
will tend to decrease when economic growth ceases. As a result, the cost of inde-
pendence increases as it is more diffIcult to fmd a use for resources which are
surplus as a result of possible client losses.
3.3.3 Probability of auditor switching
The anticipated probability that the client will switch auditor (1]) or, which is the
same, the credibility of the threat of terminating the relationship, depends on many
factors and is perhaps the element subject to more variation. One essential element
is clearly the volume of quasi-rents received by the client, highly dependent on his
bargaining power as mentioned in Section 3.3.1.2, but other factors can in many
cases be very important, particularly the effect on the client's other contracts of the
information which the auditor must divulge, the market reaction to the change of
auditor and the availability of "softer" or more ductile auditors. With respect to the
volume of quasi-rents received by the client, the benefit to the client of auditor
11 Although in a difTerent context, since a prior decision to increase capacity is not necessarily present,
these assets would have a similar nature to those Williamson calls "dedicated assets" (1985, p. 96) or
those which other authors have attributed with "temporal specificity" (Masten, Meehan and Snyder,
1991).
continuity is complementary to that obtained by the auditor himself and therefore

depends on the volume of quasi-rents to be distributed between auditor and client
and the client's bargaining power in appropriating a larger part of the total quasi-
rents associated with continuity in their relationship. The analysis in Section 3.3.1.2
makes it unnecessary to repeat the arguments here.
A large part of the rotation cost depends on the regulatory design. Almost all le-
gal systems grant statutory auditors who are threatened with dismissal the right to
attend and defend their position before the general shareholders meeting and to
make a submission either orally or in writing to the shareholders. Moreover, in
Australia, a change of auditors of publicly listed companies has to be approved by
the Australian Securities Commission and the reason known to the Commission. In
the United States, upon resignation or dismissal the auditor is required to publicly
disclose the cessation of the relationship and the nature of the disagreements and
other reportable events, as defined by SEC rules. Also, in Denmark and Portugal
the auditor can call an extraordinary general meeting. In Austria and Italy, the
auditor can defend himself before the court which must decide on the matter. In
Germany, Portugal and the United States, the auditor's rights only apply in the
event of effective dismissal. 12 The position in Spain is unique in this respect, since
the decision not to renew the auditor is at the discretion of the general meeting and
the auditor has no right to defend his position when threatened with non-renewal of
his contract.
3.3.4 Value of client-specific assets when only auditing services

are provided
In order to provide a formal analysis of how the amount of quasi-rents associated

with each client varies and on what it depends, this Section adapts the model devel-
oped by Grout, Jewitt and Whittington (1994, pp. 349-50) who, in turn,
reformulated in a more elegant and comprehensive manner that put forward by
DeAngelo (1981a) to analyze introductory pricing, or lowballing. In developing the
model a large number of assumptions, most of which have become standard in this
type of study, are introduced.13 Essentially, it assumes: competitive markets, uni-
form products, a particular inter-temporal cost structure and technological and
informational symmetry. The assumptions appear restrictive. 14 This are not so,
12 See Buijink et at. (1996, pp. 69-70).

13 See DeAngelo (1981 a, pp. 119-20), Beck et al. (I 988a, pp. 52-4) and Grout et al. (1994, pp. 325-6).
14 Let us briefly examine the meaning and reality of these assumptions: (a) Assuming that the audit
market is competitive is justified by the nature of the market in terms of its entry conditions and func-
tioning, as well as the empirical evidence. (The pioneering work in this respect is that of Simunic
however, in relation to the argument developed because the key lies in discussing
the distribution of quasi-rents and, in particular, the relative weight of the quasi-
rents associated with the client in question and other clients (client diversification,
in short) and not their volume or amount.
Specifically, let C and Po be the cost and price of the first audit of a business, c
and p the cost and price of a recurrent audit, 1'/ the probability of an auditor removal,
s the probability that an unemployed auditor gains a new client, and v and V,I the net
value of being an auditor with and without the client (the value of the quasi-rents
and the rents associated with an audit contract, respectively).
In these conditions, the variables are interrelated in accordance with the follow-
ing series of three equations with four unknowns (the two prices and the values v
and V,I):
_/ .\ (1-1'/) v + 1'/V
-,p - C, + [13]
,l
V (1+r)
f (1-1'/) v + 1'/V II } V,I

vll=sdpo-C)+ (1+r) + (1-s) (l+r) [14]
v - VII = a (v - VII + Po - p) [15]
The first of the above equations provides the value v of the position of the cur-
rent auditor, which equals the profit he obtains in the current year (p - c) plus the
anticipated value of future profits, discounted one year at a discount rate r. This
expectation depends on whether or not he will continue auditing the client next year
(which happens with probabilities [1-1'/] and 1'/, respectively). If he retains the cli-
ent, he again achieves a value v the following year. Otherwise, the auditor only
achieves v," the value of the position involving not being the auditor of that client,
given by the second equation and which in turn depends on whether or not he gains
another client, which occurs with probabilities sand (l - s) in each period. When
he gains one, he obtains a benefit equal to C - Po (a benefit which, in a competitive
[1984]). This evidence can be translated to other geographical spheres thanks to the uniformity of the
service and the presence of operators of different sizes organized in networks of international scope.
However, for a model which initially assumes imperfect competition, see Gigler and Penno (1995). (b)
Auditing is also treated as a uniform product amongst potential providers. This is equivalent to keeping
the analysis restricted to a set of potential providers of comparable quality. (c) Likewise, a particular
intertemporal cost structure is assumed, characterized by the presence of start-up or learning costs. (d)
It is also assumed that all costs, both start-up and recurrent, are the same for all provider firms, as well
as the external effects or cost savings generated by the joint provision of services. (This cost equality
eliminates any possibility of contractual friction and, in particular, guarantees that the present value of
the future quasi-rents are fully offset by the discount in the initial period). Fina\1y, (e), all income and
expenses related to the services are assumed to arise at the beginning of each period, a simplification
which does not mean a loss of generality, at least in the qualitative plane.
environment, will in reality be negative due to lowballing or initial price discount-

ing). If he gains one, he will also gain v in the following period with a probability
[1-1]] of continuity, or VII with probability 1]. In addition, the non-auditor has a
probability of (1- s) of continuing without clients until at least the following period.
The third equation defines the division of total quasi-rents (given by v - VII + Po - p)
between auditor and client, attributing them to the auditor in proportion given by
the parameter a.
The model is closed by introducing the condition, imposed by the assumption of
ex ante (i.e. initial period) competition, by which the value of not being auditor
must be nil (i.e. VII = 0). As a result, the above series of equations can be resolved,
resulting with the price of the initial audit being less than the cost of the first audit
(the initial competition for the quasi-rents to be received ex post results in lowball-
ing) and is given by:
_ C a (1 - 1]) (C - c)
Po - - 1+r [16]
whilst the price of the recurrent audit is higher than the cost of the non-initial audits
(the auditor receives quasi-rents), and is equal to:
a (r + 1]) (C - c)
P =c + (1 + r) [17]
In addition, the quasi-rents of the incumbent auditor, v, are worth:
V = a (C - c) [18]
and the quasi-rents which the client appropriates, Vc> are:
Vc = Po. - P = (1 - a) (C - c) [19]
and by summing the last two expressions, it can be seen that the total quasi-rents are
equal to the difference between start-up costs and recurrent costs:
V + Vc = C-c [20]
3.4 The costs of dependence
Faced with the hypothetical dilemma shown in Figure 3.2, the auditor who decides
not to report an anomaly will have to confront substantial costs if his inaction is
detected. In analyzing these costs, we will follow the same order of effects on dif-
ferent assets and liabilities described above. Section 3.4.1 is concerned with the
effect on the value of client-specific assets (kqQ). Those which follow deal with the
effects on firm-specific assets (kPP) and the expected cost of civil liability (kIL), as
well as that derived from the auditor's risk aversion.
3.4.1 The loss of quasi-rents associated with cu"ent clients
It is foreseeable that public disclosure that an audit firm has carried out audits of a
quality lower than expected of it, because it has acted either negligently or depen-
dently, will cause harm to its current clients. As a result, the latter will take
corrective measures which are costly to the audit firm,15 of which there is also some
empirical evidence.
3.4.1.1 Sanctions imposed by disappointed clients
In principle, it seems likely that the amount of these costs will have a positive cor-
relation with the volume of assets specific to the auditor-client relationship (Q). It
can also be assumed that the effect of scandals on current contracts (k,,) increases
less than in proportion to the size of the client. Moreover, it will certainly vary with
the type of demand depending on whether it is voluntary or obligatory (in other
words, if the client does not demand an independent audit, he will treat the fact that
the auditor has not been shown to be independent in relation to another client as of
little importance). Furthermore, it can be expected that the average volume of
quasi-rents associated with each client which the auditor receives will substantially
decrease as a result of a scandal, for reasons of at least two types:
a) Firstly, when the auditor is affected by the scandal, the bargaining power of
clients considerably increases in terms of sharing the individual quasi-rents associ-
ated with each relationship, enabling clients to appropriate a larger portion of them
(a will tend to decrease). A main reason is that the cost of cancellation will be
lower since the public will attribute it to the wish to be audited by an auditor of
better reputation and not by a softer auditor.
b) In addition, the cost of client reaction increases because the foreseeable cor-
relation of clients' responses tends to give rise to excess capacity. When a finn
suffers reputational loss as a result of baving failed in its duty of independence, it
can be expected that the response of current and potential clients occurs simultane-
IS The effect of poor audit quality on relationships with other clients was formally studied by
Balachandran and Ramakrishnan (1987, pp. 117-9) in an agency context, treating the reputational
repercussions as contingent contracts, entered into by a firm consisting of several auditors. They show
that under certain conditions the price necessary to motivate auditors is less than when the audit is
contracted separately, without the audit quality affecting the compensation which the firm receives
through other audits carried out on other clients.
ously, whether by switching or not commencing relationships with the auditor, or in

negotiating lower prices. In this situation it is likely that the audit flrm will experi-
ence a demand for its services way below its capacity, in such a way that its losses
are multiplied.
3.4.1.2 Empirical evidence of clients' ability to sanction auditor failure
It can be expected that the harm caused to clients will depend on the seriousness of
the failure or breach or, more accurately, the gap between observed and expected
quality. In the most common cases, those where the incident generates low-intensity
information as to auditor quality (for example, an administrative penalty or even a
court judgement or an out-of-court settlement for damages), only indirect indica-
tions as to the cost are available, seen in client reactions (mainly change of auditor
and also price decreases).
The direct effect on the value of a client's stock is probably not of suffIcient
scale to be distinguishable from all the other effects experienced by stock exchange
prices. The effect on a client's value has only been verifled on one occasion, spe-
ciflcally the November 1990 bankruptcy of Laventhol & Horwath, which was then
the seventh biggest audit firm in the United States, resulting in a substantial fall in
the listed value of its clients (between 1.7% for older shares and 9.51 % for those
which had been more recently listed, with an overall average of 2,48%).16 Cer-
tainly, these figures probably include most of the loss of the "insurance" implicitly
provided by the auditor, the value of which became nil after the bankruptcy. Since
the bankruptcy was unexpected, it could be presumed that most of this loss had not
been reflected in the prices of the stock. This is not the case, however, with the
effect of possible differences in the quality of past audits. On the contrary, it could
be expected that information on it would have been taken into account when the
continuing litigation occurred which finally led to the bankruptcy and which there-
fore would have been for the most part discounted when it flnally occurred.
3.4.2 The loss of quasi-rents linked to potential clients
This cost includes the quasi-rents associated with the investment necessary to gain
entry to the sector. It thus takes in the outlay in human capital specific to the pro-
fession or sector, including that invested by individuals, to the professional
16 See Menon and Williams (1994, p. 341). In this case the clients lost all the auditor-specific quasi-
rents, which may not have occurred if the finn had continued to exist. The authors consider that this
effect is minor, but they might underestimate it by concentrating only on the start-up costs.
reputation and other specific investments made by large firms, the value of which
depends on their growth and continuity.
The empirical study by Davis and Simon (1992) highlights part of the cost to the
auditor of an incident in which his reputation is harmed in terms of loss of potential
clientele. In their work they consider the effects of disciplinary measures adopted
by the Securities and Exchange Commission, the body responsible for watching
over proper functioning of securities markets in the USA. Based on both surveys
and public information, they compare the fees received from new commissions by
auditors who have been subject to disciplinary action with those received by audi-
tors not affected by such action. They thus observe that auditors who have been
penalized receive significantly lower prices, even after eliminating the effect of
other possible explanatory variables. They conclude that firms have economic
incentives to provide a level of quality corresponding to their reputation. Moreover,
it would not seem risky to extrapolate this conclusion to other incidents which harm
reputation, such as bankruptcy of a client or being sued for professional liability.
3.4.3 Losses related to professional and criminal liability
An auditor's conduct is also conditioned by the possibility that, in the event of a

lack of independence, he may suffer a loss by having to compensate those who have
suffered harm in connection with his negligence or even be subjected to criminal
charges. Concentrating on civil liability, this potential loss is very substantially
affected by the liability rules governing the activity. The essential function of these
rules is most probably that of providing an incentive for the professional to act cor-
rectly.17 It has also been suggested that liability can also have a multiplying effect
since the detection of a single poor quality audit can unleash a flood of litigation. In
this respect it is noteworthy that, at least at some points, a trend seems to have been
observed towards litigation being concentrated in a single audit firm. IS
17 Liability rules can also have a purely compensatory function, instead of generating incentives (see,
for an introduction, Shapiro [1991] and other works in the monographic issue of Journal of Economic
Perspectives of summer 1991). Auditors, however, are in a poor position to fulfil this function, for at
least two reasons. Firstly, the shareholders of client companies can diversify risks in the capital market
at less cost than auditors are able to with their client portfolio. Secondly, many countries require, and
the traditional organizational pattern of this activity seems to make it advisable, that finns and their
partners be subject to unlimited liability. Nevertheless, the American courts have based certain judge-
ments in cases against audit finns on considering them be financial insurers, which makes little
economic sense. The phenomenon has thus at times been explained as the consequence of a failure in
the political process (for example, Lys and Watts, 1994, p. 66, n. I). For an analysis of the role of the
auditor as insurer, as opposed to that of providing contractual assurance or safeguard used here, see for
example, Hill, Metzger and Schatzberg (1993). These authors also provide clear indications that prices
tend to rise as the client risk rises, but cannot identify whether this is due to greater audit intensity or to
the fact that the price incorporates the higher cost because of the increase in expected litigation.
18 See, for example, Balachandran and Ramakrishnan (1987, p. 118).
Different research confIrms, moreover, the impression that a tougher liability

system tends to encourage better quality audits,19 even if the measures of quality
and, in particular, the long term effect on professional judgement are debatable. It
has also been shown that certain client characteristics (particularly their fmancial
position and to some extent the weight of receivables and inventories in their assets)
which are assumed to be associated with a greater or lesser risk of future litigation,
affect the auditor's work plan and, therefore, the volume of indications that he
needs to reduce the risk, obviously also affecting the price of the audit. Along these
lines, a very signifIcant effect of fInancial strength has been observed. It seems as if
those clients in a worse position give rise to a greater risk of litigation and the
auditor responds by increasing the intensity of the audit and, therefore, his technical
competence (Pratt and Stice, 1994). An increase in price was also observed which
includes not only the higher cost caused by a more intensive audit but also a pre-
mium to cover the additional cost caused by the greater anticipated litigation. 20
As is well known, much of the recent discussion of audit regulation has focused
on different attributes of the liability system,21 essentially: the optimum amount of
compensation (a), the possibility that auditor liability be restricted or be propor-
tional to the loss (b), and whether or not the increase in damages represents an
actual cost (c).
a) One fundamental parameter of the liability system is the expected amount of
compensation, an increase in which seems to result in an increase in average qual-
ity, at least when regarded statistically. Nevertheless, there is room for serious
doubt as to its effect on the type of information used by the auditor and, therefore,
long-term evolution of the actual informational value of audits. In this respect, it is
pertinent to reiterate the risk involved in excessively stringent regulation: exagger-
ated penalties give auditors an incentive to be excessively prudent and base their
opinions solely on information which is verifIable by courts, thereby being reluctant
to exercise their professional judgement. As a result, the audit loses informational
value. This point was dealt with in Section 1.2.2.
19 See, in particular, Dejong (1985), Palrnrose (I 988b), Melumad and Thoman (1990) and Balachandran
and Nagarajan (1991). There are also some less solid indications to the contrary, however, (e.g. Shi-
bano, 1994).
20 See Pratt and Stice (1994, pp. 641-2) for a summary and commentary on earlier works in this field.
21 In fact, liability has ceased to be seen as a complementary but has become a fundamental aspect of
the audit service itself. For example, Dye (1993), when formally analyzing the consequences of differ-
ent liability rules on auditor conduct, considers that the audit price depends on two factors: its
informational value and the value of the option of those using financial statements in terms of the
wealth of the auditor in the event that it is shown that the audit has been defective.
b) Furthermore, the effects of liability also vary with other aspects of the legal
system, which in principle may not affect the expected level of compensation pay-
able by the auditor when shown to be negligent. The most discussed matter,
particularly in Anglo-Saxon countries, has been retention of a system of joint and
severalliability,22 in which each partner in an audit firm may be liable for the whole
of the loss suffered by third parties, even though he was not involved in the audit
and negligence by other parties unconnected with the firm is also proven, and not
just for the proportion of loss reasonably attributable to the defective action of his
firm, but for the totalloss. 23 To the extent that a joint and several liability system
has been retained for most situations,24 a solution adopted by many firms has con-
sisted in limiting the liability of partners, reincorporating such firms as limited
liability companies, when possible, thus rendering the personal assets of partners
safe from litigation as a result of audits in which they have not been personally
involved.
c) It has also been argued that since auditors can under certain conditions trans-
fer an increase in the expected cost of liability to their clients by means of charging
higher fees, the liability system does not really affect either auditors or shareholders
(the latter would only be insured to a greater extent by the auditors). This result
radically changes, however, when considering the fact that when faced with any
problem, actual resources have to be devoted to both litigation and defense. As
shown by Narayanan (1994), the proportionate liability system can thus provide,
ceteris paribus, a higher level of quality than the joint and several liability system.
The reason is that under the proportionate liability system the auditor's litigation
costs depend to a greater extent on his own efforts. On the other hand, under joint
and several liability the connection between greater effort and less anticipated cost
22 The meaning of joint and several liability can be clarified with a couple of examples, the first anec-
dotal but real. In 1987, Walt Disney World was sued by Ms. Wood. The plaintiff had suffered an
accident on a Disney World bumper car track and sued Disney and her own boyfriend. The jury attrib-
uted 14% of the blame for the accident to her, 85% to the boyfriend and the remaining 1% to Disney.
However, as the boyfriend had no resources, Disney had to pay 86% of the damages (Narayanan, 1994,
p. 40). Translated into the auditing field, in a typical example taken from Hanson and Baker (1996) a
client overstated his inventory and his auditor did not detect the problem. Based on the financial state-
ments, a bank gave credit to the client and the latter became insolvent and could not repay it. The judge
or the jury attributed 90% of the fault to the client and 10% to the auditor and evaluated the total loss at
400,000 dollars. Under joint and several liability, the bank could make up its total losses from the
auditor. Under proportionate liability, however, the auditor would only be liable to pay 40,000 dollars.
23 Liability insurance premiums have reached 8% and 14.3% of turnover of the main audit firms in
Great Britain (\994) and the United States (1992) despite which in some cases they have not been able
to obtain it even at these prices. (Data taken on Great Britain from "British Accounting Liability: Big
Six PLC" [The Economist, 7 October 1995, pp. 109-12] and for the US from Mednick and Peck [1994,
p. 891 D. Between 1985 and 1992, in the US, the insurance premiums of the then Big Six firms multi-
plied by three and the deductibles by six (Narayanan, 1994, p. 40, n. 2).
24 During the I 990s different legal and legislative decisions have been reducing the liability of auditors
in the United States. In this respect, see Hanson and Baker (1996).
is attenuated due to the presence of insolvent defendants for which the auditor is
liable. As a result, those really benefiting from the system of joint and several li-
ability end up being the lawyers to the extent that this system increases the demand
for their services.
Table 3.1. Legal framework of statutory auditor liability in the main OEeD countries
Non-contract third parties who Is it possible
Is there a legal
can undertake legal actions to arrange a
Liability against the statutory auditor
liability cap? contractual
Country al/ocation liability cap
Between the Between the
regime Other third contract between the
Shareholders contract contract
parties parties and
parties third parties parties?
Australia Joint and several Yes Yes No No Yes
Austria Proportionate No No Yes No Yes
Belgium Proporti onate Yes Yes No No No
Canada Joint and several Yes Yes No No Yes
Denmark Joint and several Yes Yes No No No
Finland Proportionate Yes Yes No No Yes
France Proportionate Yes Yes No No No
Germany Joint and several No No Yes No Yes
Greece No clear regime Yes Yes Yes Yes No
Ireland Joint and several Yes Yes No No No
Italy Joint and several Yes Yes No No No
Japan Joint and several Yes Yes No No
Luxembourg Proportionate Yes Yes No No Yes
Netherlands Not available Yes Yes No No Yes
Norway Proportionate Yes Yes No No No
Portugal Not available Yes Yes No No No
Spain Proportionate Yes Yes No No Yes
Sweden Proportionate Yes Yes No No Yes
Great Britain Joint and several No No No No No
United States Joint and several* Yes Yes No No No
Note: • Proportionate liability was introduced by the 1996 Securities Reform Act for auditors of SEC
registrants.
Sources: For European countries, adapted from Buijink, Maijoor, Meuwissen and Witteloostuijn (1996,
pp. 95-6, 99); for other countries, specific survey carried out by the author, who forwarded the relevant
questions from the original questionnaire used in the study by Buijink et al. (1996) to audit firms in
each of the four countries (Australia, Canada, United States and Japan), information from which has
been added to that provided by Buijink et al. (1996) in various descriptive tables in this and the fol-
lowing Chapters.
A wide variety of rules govern auditor liability in different countries. As shown

in Table 3.1, the system of proportionate liability for the loss caused dominates,
which possibly provides an incentive for higher quality and reduced legal costs.
Nevertheless, the uncertainties surrounding the possible liability of auditors leads to
the big fIrms fmding it virtually impossible to obtain civil liability insurance. Per-
haps as a consequence, in those countries where it is legal to do so they have begun
to limit their liability contractually. 25
3.4.4 The role of risk aversion
The preferences of economic agents regarding risk also encourage auditor prudence,
provided that the latter is averse to risk and the probability of detection is less than
that of being removed. Let us consider, for example, an auditor confronting a
situation in which, if he reports an anomaly, the probability of losing a client ('1) is
very high, whilst if he does not report, the likelihood of being detected (tr) is very
low. Let us also assume that the amount of the expected cost of reporting ('1v) and
not reporting (tr [v + kq Q + kp P + k/ L]) is the same. In this situation, one would
expect that most auditors would not be indifferent to the two possibilities and would
choose to report because their risk aversion leads them to prefer the less risky op-
tion which consists in suffering a small loss with greater certainty, rather than that
of having a low probability of suffering a much greater loss. In this fIeld, the legal
restrictions on limiting auditor liability are relevant, which often provide that all
members of an audit fIrm or company are vicariously liable on a joint and several
basis for the losses arising out of breach of obligation by another member.
25See, for example, "La ley debe delimitar mejor la responsabilidad del auditor" (Expansion, 21-2
September 1996).
Part 2
THE PROVISION OF NON-
AUDIT SERVICES BY
AUDITORS
Situation, effects and regulation

Up to now the analysis has largely remained in the general plane. It is now time to
apply it to examination of a particular case. The chosen problem is that of deter-
mining the most suitable rules covering the provision by auditors of non-audit
services, often to their own audit clients. The study of this problem begins in
Chapter 4 by examining its general features and the cost advantages resulting from
the joint provision of audit and non-audit services. In subsequent chapters the con-
sequences of such services on independence (Chapter 5), and competition (Chapter
6) are considered. It will be argued that the provision of such services reduces
overall costs, raises the technical quality of auditing, enhances competition and need
not prejudice auditor independence or the quality of these non-audit services. At
the end of Part II the regulatory possibilities will be examined. The basic recom-
mendation will be to leave market forces free to find an efficient equilibrium in
terms of both the mix of services and the right combination of quality safeguards,
with regulation intervening to strengthen the safeguards provided by the market.
Specifically, audit firms should be obliged, at most, to disclose the maximum extent
to which their income is concentrated in a single client (Chapter 7).
67
4 THESERVICES
SUPPLY OF NON-AUDIT
BY AUDITORS
Since their appearance in the XIX century, audit firms have provided a wide variety
of services to the majority of their clients. In most countries, many firms currently
provide a broad series of non-audit services. On average, these account for around
50 percent of their total fees and have being increasing for many years, although
there are substantial differences between firms and countries (Table 4.1 ).1 This
Chapter explains how this pattern of joint provision of audit and non-audit services
is based on the greater efficiency deriving from economies of scope, or joint pro-
duction, of both a productive and contractual origin. It is claimed that they are
considerably greater than any diseconomies or negative effects, the latter relating in
particular to an appearance of diminished independence in the mind of users of
audited financial statements and the public generally.
4.1 Definition and types of non-audit services: the

importance of a general analysis
Throughout this work non-audit services are taken as all those services provided by
audit firms which do not involve auditing, including advice and consultancy, assis-
tance in applying methods and support in implementing transactions and supplying
staff for specific tasks. 2 Some of these services are closely allied to accountancy
and auditing, such as advising on information systems, whilst others are more pe-
ripheral, such as market research or optimizing factory layout, for example.
1 Figures on this matter are only provided as an indication, given that they are not very reliable. Apart
from a lack of data for some periods and firms, there are substantial discrepancies in the criteria used to
split the total fees. It also difficult to assess the importance of joint fees. Furthermore, no distinction is
made between non-audit services provided to audit and non-audit clients.
2 For more detail, see AICPA (1996, item 5).
69
Table 4.1. Evolution of audit firm fee splits in different markets during the period
1993-1997
Auditing &
Tax services All other services
Accounting services
1997 1993 1997 1993 1997 1993
World, all firms 50.3% 52.2% 20.6% 19.5% 29.1% 28.3%
World, Big Five firms 41.8% 48.8% 19.5% 19.3% 38.8% 32.0%
World, non Big Five 53.5% 53.6% 21.0% 19.5% 25.5% 26.9%
USA, all firms 41.0% 47.4% 28.3% 30.7% 30.7% 21.9%
USA, Big Five firms 34.5% 48.8% 20.7% 22.7% 44.8% 28.5%
USA, non Big Five 43.0% 47.0% 30.7% 33.1% 26.4% 19.9%
Europe, Big Five" 47.3% 51.4% 22.3% 21.9% 30.4% 26.7%
Europe, non Big Five 62.6% 59.6% 18.6% 19.0% 18.7% 21.4%
ASIA, non Big Five 58.3% 59.7% 18.6% 22.0% 23.1% 18.3%
Notes: Means for each market were calculated from samples of firms with available data. " Average of
Big Five firms in three main European markets with freedom to provide non-audit services (Germany,
Spain and the United Kingdom).
Source of data: International Accounting Bulletin, several issues and years.
It has often been argued that the provision of these non-audit services to audit
clients could give rise to various problems affecting both types of service. This
second part of the book formulates a general framework for analysis of this ques-
tion. It will therefore concentrate less on the problems posed by specific types of
service, which are used more to help illustrate the arguments put forward,3 and
focus, on the contrary, on the more generic and common problems. In favor of this
more general approach it should be considered that the problems of such services
reveal characteristics which will enable regulators to generalize their analysis and
treatment, adopting a more facilitating rather than controlling strategy over private
incentives. Rules should therefore be of a minimalist nature because the market
itself provides incentives to control possible problems which may arise in this field.
Regulation should thus play an essential role in facilitating this market function,
ensuring that the market receives relevant information to enable such incentives to
operate effectively.
This does not mean that it is unnecessary to analyze the particular problems of
each type of service and formulate specific solutions, but that both public regulators
and external analysts are in a poor position to tackle these tasks successfully. The
3 Reports for the United States provide several analyses of the problems of specific services. See, for
example, Metcalf (1977, pp. 34-5 and 50-2), on taxation and management consultancy; Cohen (1978,
pp. 98-102), on tax, accounting and recruitment services; POB (1979, pp. 46-54), on marketing, pro-
duction and other management consultancy, insurance and executive recruitment.
The supply of non-audit services by auditors 71
problem is that from outside audit fIrms or the profession itself it is unlikely that the
type of information required to assess the problems posed by specifIc services can
be observed or gathered with adequate speed and accuracy. As a result, detailed
regulation of such services should fall to the self-regulation of professional associa-
tions and, in particular, the organizational decisions of fIrms themselves, motivated
by a market as transparent as possible. Regulatory differentiation, on the other
hand, makes little sense because to be useful it would require overly specifIc infor-
mation which will rarely be available, which thus renders it advisable for this to be
established in the context of fIrms themselves.
Let us look at an example. It has often been recommended that auditors should
only provide services closely allied to the accounting function.4 The proposal ap-
pears attractive because such services are supposedly those in which cost savings
will be maximized due to economies of scope. Moreover, it would harm a lesser
number of fIrms. This appearance is deceptive, however. Firstly, the existence of
economies is accompanied by a higher potential risk that independence may be
endangered since the auditor approaches a position in which he must review his
own work or the outcome of it. In addition, attention is only paid to the tangible
economies but not to those of a contractual nature, which are often highly relevant.
Finally, what is most important is that it is not the physical nature of the services
which determines the potential for compromising independence, but earning sub-
stantial quasi-rents (compensation exceeding the maximum obtainable with the best
alternative use of resources) associated with their provision. Section 3.3 illustrated
how these quasi-rents depend on specific conditions (for example, the degree of
utilization of fixed assets) which trivializes all attempts to classify the risk to inde-
pendence in fact on the basis of the physical nature of the service (although not,
perhaps, technical competence or independence in appearance).
4.2 Economies of scope
4.2.1 Auditing technology and firms' complexity
As indicated in Chapter I, auditing basically consists of an examination by a third

party, the auditor, of the accounts of an organization for the purpose of making
them more trustworthy for those who need to make use of them. It is clear that the
effectiveness of this examination requires, both now and in the past, a certain under-
standing by the auditor of the activity carried out by the undertaking or body being
4 See, for example, the Metcalf report (1977, recommendation no. 8) and some of the opinions set out
in the EC Green Paper (1996, item 4.13).
audited, including a knowledge of both its physical technology and its contractual
patterns. For this reason, the degree of business variety and complexity is one of
the basic determinants of the technology which must be used in auditing activity.
A large part of the misunderstanding surrounding discussion of auditing is con-
nected with a simplistic and outdated view of the technology which the auditor
needs to use in his work, sometimes still seen as little more than someone who rakes
through supporting documents and checks arithmetic. In reality, it is impliedly
assumed that the subject matter of auditing is a certain type of business which is
increasingly less important-one devoted to industrial activities, and therefore with
tangible assets and products, of a manageable size, with little contractual complex-
ity, which uses physical technology with relatively well known possibilities, has a
simple fmancial structure, and, fmally, is relatively unaffected by public regulation.
This archetypal undertaking continues, perhaps, to be the model transmitted by
many textbooks. It is increasingly less representative of economic reality, however,
where the varied types of undertaking and the weight of the service and intangible
sectors has increased considerably. In the same way, auditing technology has been
changing since the growing complexity of undertakings has necessarily made it
interdisciplinary. To carry out an effective examination, the auditor of today must
understand the nature of increasingly specialized and complex economic activities,
often regulated in a virtually inextricable manner, with innovative physical and
contractual technology, using emerging financial products and a large part of whose
value takes the form of intangible assets such as trade marks, reputation and human
capital. If one wants the auditor to produce useful information for decision-making
and contracting, he must make use of information input which will help him to
evaluate these and many other questions,S input which requires the use of all sorts of
experts-legal, tax, financial, engineering, information technology, etc.
The intention in the following pages is to demonstrate that there is no good rea-
son why the auditor, instead of acquiring these services from third parties, should
not contract the necessary experts to provide them within his firm, and make effi-
cient use of these resources by employing them in part in providing services other
than auditing. Moreover, in order better to safeguard the quality of his basic final
product (auditing), it seems logical that the auditor should be able to exercise tight
control over the quality of the inputs involved in his audit and such tight control is
more credible if it can be exercised with more solid ties than those obtainable from
a purely commercial relationship. Similarly, non-audit services can be provided at
less cost thanks to the existence of economies of scope.
; One alternative is impliedly ruled out, consisting of settling for a minimalist audit in the sense re-
ferred to in Section 1.2.2: a defensive audit which does not produce a professional opinion but mere
documentary verification. Its effectiveness would certainly not be nil, but would be very close to the
level achieved by a purely formal examination.
4.2.2 The supply of non-audit services as a natural consequence

of modem auditing
In this context, an efficient use of the resources required to carry out quality audit-
ing can only be achieved if these resources are also used to advise clients. It has
been traditional to take advantage of auditing to improve a client's situation in areas
closest to the traditional auditing configuration, such as the installation of mainly
accounting and information systems,6 and in the field of tax advice. The novelty in
recent decades has meant that as audit firms needed to contract experts to help them
understand the position of undertakings which were changing and becoming in-
creasingly complex, they found themselves in the position of being able to satisfy
an increasingly wide range of demand in the field of advice and, in general, the
provision of many professional services. 7 For this purpose they also had a com-
parative advantage compared with the experts specializing in providing such
services: their reputation as reliable suppliers under a regime of absolute confidenti-
ality eliminated the reservations which some clients may have had, particularly with
respect to new services.
Restrictions on the use of these resources could impose substantial costs, both in
terms of lower quality and more costly and expensive audits. To maintain the same
level of quality, additional costs would arise because firms could not use these ex-
perts and would have to contract them externally (which is inefficient to the extent
that they are not already doing so under an umestricted system). Moreover, under
this external contracting regime there would be fewer incentives to produce knowl-
edge during the audit relevant to improving the client's situation and the knowledge
eventually produced would be much more difficult to apply.
It should also be indicated, although the matter will also be treated later on in
Section 7.104, that the trend towards external imposition of an obligation to extend
the scope of auditing reinforces the need for experts. Thus, for example, the Euro-
6 A historical curiosity: the first major business information technology application was developed in
1952 by the audit firm Arthur Andersen for the General Electric company and consisted of an automatic
system for processing payrolls (Nanda and Yoshino, 1996, p. 2).
7 The financial press is constantly announcing new areas in which audit firms are effectively applying
their human capital. Particularly in those Anglo-Saxon countries most advanced in the use of auditing,
clients increasingly demand audit services with greater added value. Some firms, such as the US arm of
KPMG, for example, have introduced comparative analysis practices, or benchmarking, which they
carry out in parallel with the current audit (Berton, 1996). In some cases, the demand necessitates
collaboration with other companies, to which end cooperation agreements have also been reached with
service supply firms, for example information technology firms, in order to offer combined services to
common clients (co-contracting). In this respect Microsoft's interest in allying itself with accounting
firms is interesting (Scott, 1997).
pean Commission Green Paper (1996) includes amongst the expectations weighing
on the auditor's work those of providing
reassurances concerning the accuracy of financial statements, the going concern

status/solvency of the company, the existence of fraud, compliance by the company
with its legal obligations, the responsible behaviour of the company with regard to
environmental and societal matters [(items 3.7 and following). Moreover, with re-
spect to the latter aspect it indicates that] ... although it is unreasonable to expect the
statutory auditor to make judgements on matters outside his competence and exper-
tise, it can be argued that auditors should accept that their responsibilities will tend to
increase in line with public expectations. The auditor can, given time and a suffi-
ciently clear consensus on what is expected, avail himself of the necessary expertise
in areas which go beyond strict financial audit (item 3.35).
4.2.3 Origin and types of economies of scope
4.2.3.1 History
Both accountants at first, and then independent auditors later, have provided a wide
range of services to their clients from the beginnings of their profession. In fact,
auditing appeared later than the provision of other services. In reality, it is only
around the tum of the century when it accounted for 50 percent of turnover, and
then only amongst the largest firms. Moreover, accountants and auditors provided
advisory services both before and after the appearance of auditing. In the case of
both Great Britain and the United States, the provision of non-audit services has
been a constant feature of the profession, always including the areas of accounting,
tax and management consultancy services and with the growing importance during
the XIX and first half of the XX century of trustee and executorship (see Table 4.2
for evolution of breakdown of fee income for one of the main British firms). It is
believed that the main external factors in this evolution were the institution of taxes
on corporate income, which provided the auditor with an undeniable advantage in
the field of tax services, and information technology which extended the market for
services connected with information systems. 8
8 For the background to the current situation, see in particular Previts (1985), who analyses the evolu-
tion of the profession in Great Britain and the United States, paying particular attention to the scope of
services provided at each point in history.
Table 4.2. Breakdown of fee income of the firm of Whinney, Smith and Whinney
between 1848 and 1960
Trustee &
Special
Period Insolvency Accounting Auditing Taxation Executor- Total
Works
ship
1848-1859 77.93% 12.29% 7.13% 2.65% 100.00%
1860-1870 89.68% 6.03% 1.93% 2.16% 0.20% 100.00%
1870-1880 93.63% 2.38% 2.18% 1.74% 0.07% 100.00%
1880-1890 61.00% 9.63% 23.38% 4.03% 1.98% 100.00%
1890-1900 26.90% 16.53% 48.00% 4.00% 4.57% 100.00%
1900-1910 30.03% 11.27% 48.87% 0.35% 4.80% 4.70% 100.00%
1910-1920 40.05% 7.58% 41.60% 1.63% 3.35% 5.80% 100.00%
1920-1930 26.03% 9.50% 51.40% 4.27% 4.00% 4.80% 100.00%
1930-1940 3.60% 12.40% 69.37% 6.07% 2.33% 6.23% 100.00%
1940-1950 3.48% 8.75% 68.45% 7.80% 1.73% 9.80% 100.00%
1950-1960 1.73% 6.80% 64.30% 12.17% 1.87% 13.13% 100.00%
1960 0.20% 8.50% 59.70% 11.10% 2.40% 18.10% 100.00%
Note: Unweighted averages are shown for each period to avoid large fluctuations in annual figures.
Sources: Own preparation with data from Jones (1981, pp. 47 and 99), taken from Previts (1985, pp. 23-4).
4.2.3.2 Types of economies of scope
The reason why accountants and auditors provide services which complement their
principal task is connected, now and in the past, with the considerable economies of
scope, or joint production, involved-these meaning cost savings obtained when
both types of service are provided by the same person or firm. A distinction should
be made within these economies of scope between those which originate in the
transformation process directed towards the production of information and knowl-
edge, often known in accounting literature as knowledge spillovers, from those
arising from making better use of assets or advantages of a contractual nature.
a) Knowledge spillovers. These usually arise from the fact that both types of
service need to utilize the same set of information and/or the same professional
qualifications. For example, in order for the auditor to evaluate the reliability of
financial statements, he must examine the quality of the accounting and internal
control processes, useful information when it comes to detecting shortcomings and
recommending possible improvements. Similarly, an audit necessitates evaluating
the adequacy of provisions for paying taxes, which requires substantial competence
on the part of the auditor in the tax field as well as in many other areas. Conversely,
qualification in all these areas facilitates audit work and the provision of these
services enables the auditor to make a better-founded judgement regarding the cli-
ent.
b) Contractual economies ofscope. The existence of economies ofa contractual
nature is connected with the fact that the exchange of professional services involves
high transaction costs due to the informational asymmetry existing between supplier
of and client for such services. For this reason, the ability to use the same contrac-
tual resources (brand-name, reputation, conduct rules, control systems amongst
professionals, client confidence) is particularly valuable in safeguarding or protect-
ing the provision of a variety of services, even in the absence of economies of scope
of a technological nature in the strict sense. 9 Frequently, these contractual advan-
tages are referred to under the label "one-stop shopping", which might be slightly
deceptive, as it leads to thinking in terms of the cost of merely searching for provid-
ers, which probably is not the most important when compared with the cost of
ensuring contractual performance. It is also important to notice that these econo-
mies 0 scope may and will in general flow both ways, from audit to non-audit
services and vice versa. Thus, the contractual safeguard of auditing will be easier
for firms providing non-audit services than those fully specialized in auditing.
The first type of economy of scope, associated with the joint use of information
to provide different services to the same client, was very important in the past and
remains so amongst smaller firms. Amongst large firms, on the other hand, its im-
portance is perhaps less and decreasing (at least in the case of those economies of
scope which are specific to a client), as shown by the fact that nowadays different
teams or even divisions and companies are responsible for providing each type of
service. In the large multidisciplinary firms, the advantage of joint provision of a
9 The contractual economies have merited little attention in the literature and, when dealt with, usually
involve placing emphasis on excessively specific situations. This is the case, for example, with the
contractual efficiencies analyzed by Antle and Demski (1991) in terms of contractual "frictions", con-
structing a sophisticated informational asymmetry model which represents a series of situations, some
of which affect demand for advisory services, and then studying theoretically how it affects the joint
provision of audit and non-audit services. The informational externality which they consider is of the
contractual type and is related to knowledge regarding the cost of providing a service for a client prior
to its provision. They find that the determining factor is the ex ante relationship between the costs of
both types of service, after examining the possible benefits of providing similar services when there are
informational externalities connected with the existence of contractual frictions (principally informa-
tional asymmetry as to the cost of supplying services to a specific client). The model takes into account
the effect of regulatory restrictions in terms of the minimum quantity of audit services to be acquired (as
an approximation of the obligatory use of certain procedures) and/or their price (restrictions on con-
tracts between client and auditor, as an approximation of prohibiting contingency fees). Given that the
model assumes constant quality, it is closer to considering those regulatory provisions which mandate
use of a procedure of the type which seeks to raise quality (for example, peer review). Moreover, it is
not concerned with the effect on independence.
wide range of services seems increasingly to reside, therefore, in all types of con-
tractual advantages, and perhaps also in scale and network economies in the
production of knowledge, which is not specific to a single client, as in the case of
the knowledge spillovers previously referred to, but which can be utilized on a gen-
eral basis. This latter is particularly the case with the investment necessary for the
functioning of a global network of offices providing uniform quality, in terms of
training centers and programs as well as data bases, quality control systems and
management and organizational systems generally.
4.2.3.3 Empirical evidence of the existence of economies of scope
The existence of economies of scope between audit and non-audit services is clearly
apparent when observing the circumstances and evolution of the sector. This ob-
servation reveals, firstly, that since the birth of the profession there has been a
mutual interest amongst clients and auditors in using the knowledge and work of the
latter to carry out broader tasks beyond mere accounting and/or auditing. The best
evidence of such economies is thus the very success of audit firms in diversifying
their activities. As Antle and Demski indicate, "the prominence of management
consulting services in the largest CPA firms leaves little doubt that there are
economies of scope between auditing and at least some types of consulting" (1991,
p. 1). A second empirical indication, of an indirect nature, is provided by the results
obtained by DeBerg, Kaplan and Pany (1991), who observed how the volume of
recurrent services acquired from a new auditor fell significantly after a change of
auditor compared with those acquired from the previous auditor. This decrease is
consistent with the existence of informational external effects, which necessitate in-
depth knowledge of the client which is only acquired after some years of
relationship with his auditor. A third indication, of a qualitative nature, no less
valid as a result, regarding the economies achieved when auditors provide services
in similar fields, is the fact that most internal auditors in businesses (employees
carrying out internal audit work) also provide advice to their organizations and,
moreover, expect to provide this type of service to a growing extent in the future. 10
Finally, the use of auditing as a "gateway" to attract service business (not confirmed
empirically, as indicated in Section 6.2) would also indicate the existence of
economies of scope since, in their absence, such a pricing policy would make no
sense in a market of these characteristics. Such a policy could only be justified as a
predatory practice which, if scarcely credible on a general basis, is all the less credi-
ble in this case since the market enjoys few restrictions on entry and many firms of
all types are present in it, as examined in more detail in Section 6.3.2.
10 "Internal Auditors and Internal Consulting" (Internal Auditor, June 1996, p. 10).
Despite the abundance of this qualitative evidence regarding the economies ob-
tained by the joint production of auditing and other services, a quantitative
identification of such economies has eluded those who have undertaken empirical
studies with this aim. Most authors believe, however, they can identify the presence
of these economies of scope as the cause of their observed indications. Moreover,
the reasons for this invisibility are clear. Firstly, the analysis is in itself difficult due
to the complexity of the possible existing interaction between auditing and other
services. I I To this should further be added that the lack of data on the costs of the
two lines of activity have made it necessary to use indirect methods which suffer
from substantial shortcomings and provide some contradictory results. In particu-
lar, those carrying out the studies know the maximum prices or costs of the audit
services but not those of the non-audit services. 12 For this reason, the empirical
results based on econometric studies are arguable since they are partial, and, conse-
quently, the qualitative evidence might be, in this case, more revealing. Despite
these problems, the content of the most relevant studies will be outlined at this
point, classifying them into two groups depending on whether or not they take ac-
count of differences in the audit workload.
a) The price of the audit is positively correlated to the provision of non-audit
services. The pioneering study in this field is that of Simunic (1984), who tried to
verify whether the provision of consultancy services has effects external to auditing.
To this end, he compared the prices charged for audits depending on whether they
were carried out on clients to whom other services were provided. On examining a
sample of companies audited by large firms he found that the audit fees were higher
when the client also required non-audit services from its auditor. Assuming that
demand for auditing is non-elastic and that audit firms do not cross-subsidize be-
tween activities, this apparent surcharge would be the means whereby the auditor
shares in the economies of scope. This study was repeated with different samples in
both the United States and other countries, with similar results. 13 Both their meth-
odology and interpretation of results might be misleading, however, in the light of
subsequent studies:
• On the one hand, it is possible that spurious correlations are being obtained in
these works by using data on fees and not having access to internal cost data. To
\I See, along these lines, the work of Gaver and Gaver (1995).
12 The study by O'Keefe, Simunic and Stein (1994) also suffers from this problem, despite using inter-
nal data from a firm. Specifically, the service data used is just the group to which each client belongs in
terms of the percentage of non-audit fees in relation to audit fees (O'Keefe el aI., 1994, pp. 250-2). For
this reason, the authors can only try to examine the existence of external economies which flow from
non-audit services towards auditing and not in the other direction.
13 See Turpen (1990), for the United States; Barkess and Sinmett (1994), for Australia; Ezzamel, Gwil-
liam, and Holland (1996, 1998), for Great Britain; and Firth (1997a) for Norway. They all found a
positive relationship between audit fees and fees for non-audit services.
avoid this problem, Abdel-Khalik (1990) used more refined econometric tech-
niques, specifically the Heckman-Lee method, to verify self-selection by clients
in the direction of acquiring non-audit services or not. By applying this proce-
dure he was incapable of detecting whether audit fees differed amongst clients
based on whether they acquired non-audit services or not.
• On the other hand, the acquisition of non-audit services could arise when the
client company undergoes changes or difficulties which lead it to demand such
services and which simultaneously increase the cost of auditing its financial
statements. Palrnrose (1986b) found, just like Simunic (1984), that the audit
price increases with the volume of non-audit services acquired by the client.
Nevertheless, she noted in a sub-sample that this higher price was found both
when the supplier of these non-audit services was the auditor himself and when
the supplier was another firm. This datum denies the validity of the interpreta-
tion that the price difference is an indicator of external effects between the two
types of service (although it says nothing regarding their existence, obviously).
b) The increased audit price is caused by greater difficulty and/or workload.

Unlike the case in the previous works, Davis, Ricchiute and Trompeter (1993) used
internal data from an audit firm, based on which they were able, firstly, to replicate
the results obtained in the previous studies and, secondly, throw doubt on the inter-
pretations which had been made of these results until then. Specifically, they found
that the price of auditing is greater for clients also acquiring other services, but this
price increase can be explained by greater auditor workload, measured in hours.
The result is thus consistent with the observation mentioned earlier and the conse-
quent conjecture at the time by Palrnrose (1986b).
In conclusion, little can be inferred from the studies into audit prices regarding
the existence of economies of scope. Nevertheless, this result is not surprising if it
is borne in mind that, if they were manifest in prices, it is as or more likely that they
would be so in the prices of the services, which are unknown and which, even if
known, would be very difficult to compare. Furthermore, some of these studies do
not consider the possible distorting effect which the practice of introductory pricing
could have on initial audits, an aspect which will be dealt with in more detail later.
4.2.4 The existence of diseconomies of scope
Together with these positive effects, manifested by lower production costs, the
provision of non-audit services by staff of the same audit firm or connected firms
could raise different types of problem which, in economic terms and from their
similarity, could be classified as diseconomies of joint production, or scope. In this
respect, two possible negative effects should be distinguished, revealed as poor

quality of the audit and of the non-audit services:
a) Negative effect on the audit. It has been argued that the provision of non-
audit services could affect audit quality, in particular by adversely affecting the
auditor's independence. Since this is the central point, the whole of Chapter 5 will
be devoted to discussing this aspect, applying the methodology developed in previ-
ous chapters. It suffices to say at this point that although this problem is potentially
serious, the market can and in fact has provided sufficient corrective incentives in
countries where auditing is more developed. Both fIrms and professional associa-
tions have to this end developed effective safeguards and counterweights, which
enable them to benefIt from the positive economies which result from jointly pro-
viding a wide variety of services without compromising auditor independence.
From the point of view of public regulation, the analysis leads in Chapter 7 to rec-
ommending a policy aimed at facilitating the development of those safeguards
which are provided by market forces. This policy might include intervening, at
most, to oblige audit fIrms to disclose data on the diversifIcation of their fees.
b) Effects on non-audit services. It has sometimes been argued that the provision
of some services by professionals in an audit fIrm, or a fIrm related to it, gives rise
to unavoidable conflicts of interest which compromise the quality of such services
or their provision in accordance with the standards of the corresponding profession.
In Europe, this problem has generated considerable debate regarding legal services,
for which reason the argument will focus on this area, although the analysis and
conclusions can be applied to any other sector. Section 6.4.2 will question the im-
portance of the problem and present the incentives that the market already provides
to contain it. It will also develop an alternative argument, according to which criti-
cism against auditors providing non-audit services simply reflects the resistance of
established fIrms and professionals to the new form of competition represented by a
more integrated provision of services.
5 EFFECTS OF NON-AUDIT
SERVICES ON AUDITOR
INDEPENDENCE
This Chapter investigates the effects of providing non-audit services on auditor

independence in fact and in appearance. Section 5.1 examines how the provision of
non-audit services affects each of the economic causes or factors, whose effect on
independence in fact was examined generally in Chapter 3, and explains the empiri-
cal evidence relevant to forming a judgement on the matter. The conclusion of this
dual examination is that independence in fact is not prejudiced by the provision of
non-audit services-rather it may be strengthened, particularly in firms with a di-
versified clientele. Section 5.2 then discusses the accumulated empirical evidence
regarding the apparently negative effects of providing non-audit services on the
appearance of independence. Finally, Section 5.3 explains the principal safeguards
adopted by professional bodies and firms to ensure quality, avoid conflicts of inter-
est and improve the perception of independence in the field of non-audit services.
5.1 Effects of non-audit services on the economic

incentives for independence in fact
The effects of providing non-audit services on independence in fact will be studied

with an adaptation of the conceptual model used previously in Chapter 3 to analyze
economic incentives for independence in a general manner. The treatment will be
more straightforward here, however, since this analytical framework has already
been explained.
The provision of non-audit services affects all elements of the condition of inde-
pendence defined by expression [12] and shown in Figure 3.2. Their fundamental
effect, however, consists of an increase in specific assets, associated both with cur-
rent clients (with a double effect in terms of the condition of independence, by
81
increasing both the quasi-rents, v, associated with the client whose audit is subject
to scrutiny, and those associated with all other clients, Q) as well as the assets spe-
cific to the firm (P). Since they are more important, these assets will now be
discussed in some detail before commenting on other, relatively secondary, conse-
quences. The latter include modifications to bargaining power in relation to sharing
the quasi-rents, incentives to change auditor, the cost of collusion between auditor
and client and possible liability costs. Finally, the available empirical indications
regarding independence in fact will be examined.
5. 1. 1 Higher quasi-rents linked to each client make diversified

auditors more independent
There is little doubt that the provision of non-audit services increases the volume of
specific assets and the amount of quasi-rents associated with the auditor-client rela-
tionship. This point is formally justified below before demonstrating that they have
a positive effect on incentives for independence amongst those auditors with a di-
versified clientele.
5.1.1.1 The increase in client-specific assets when the auditor provides

non-audit services
A simple extension of the model developed in Section 3.3.4 is useful in analyzing

the effects of the provision of non-audit services on auditor independence in relation
to each of his clients and, in particular, to estimate how prices and quasi-rents asso-
ciated with each client vary.
The academic literature has distinguished various types of effect between the dif-
ferent services provided by auditors. 1 There is no need for an analysis of all
possible variants, however, to show that the provision of services encourages inde-
pendence of the auditor who has a diversified client portfolio. A simple model
suffices in which it is assumed that the provision of non-audit services reduces the
recurrent auditing costs by a coefficient (J om). Consequently, the actual cost of
recurrent audits is me. This assumption implies that the audit precedes the provision
of non-audit services in time.
Distinguishing with a superscript s all variables associated with the provision of
non-audit services, the total quasi-rents when non-audit services are provided is
1 For example, distinctions can be made simply between economies of scope of an informative nature,
or knowledge spillovers, between those which can flow from auditing and similar services, in the oppo-
site direction or in both directions, and between those which can affect fixed, variable or both types of
cost. In addition, some effects are client-specific, whilst others are based on the accumulation of gen-
eral skills. Different treatments can be seen in the works of Beck et al. (1988a), Parkash and Venable
(1993) and DeBerg et af. (1991).
Effects of non-audit services on auditor independence 83
(v'-v:; +P"o -p'). Total auditing quasi-rents will be taken as those existing when the
auditor does not provide non-audit services (C-c) which, as obtained in Section
3.3.4, are equal to (v-v. +Po-p). The total quasi-rents of non-audit services are then
given by the difference between the total quasi-rents with non-audit services and the
total quasi-rents from auditing: (v'-V:; +P"o -p') - (v-v.+Po-p). Assuming, further-
more, that the bargaining conditions are such that the auditing quasi-rents are
attributed to the auditor in proportion a and those from non-audit services in a pro-
portionft, then:
v' - v:; = a(v- v. + Po- p) + ft {(v' - v:; + P"o - p')-(v- v. + Po- p)} [21]
By formulating and resolving the corresponding equations, the content of which

is similar to that of expressions [13] to [15],
(1-7]) v' + 11 v:;

v'=p'-mc+ (1+r) [22]
( 1-nlv' + ny': y.:

11" =s{(P"o-C) + '''1+r'{' }+(1-s) 1:r [23]
v'- 11" = a (v-v.+Po-p) + ft {(v'- v:; + P"o -p')-(v-v. +Po-p)} [24]
v:; = 0 [25]
the following prices result:
= CJ1-11) {fJc (1-m) + a(C- c))

[26]
P"o 1+r
(11 + r) {fJ c (1 - m) + a (C - c)]

p'=cm+ 1+r [27]
and the value of the quasi-rents received by the auditor who provides non-audit
services is:
v' = a (C - c) + ft (1 - m) c [28]
whilst the quasi-rents of the client are given by:
~ = (1 - a) (C - c) + (1 - ft) (1 - m) c [29]
in such a way that the total quasi-rents (those received by the auditor plus those
obtained by the client) are now:
v'+l{ =C-cm [30]
It can then be seen that in relation to the situation without non-audit services, the
value of both quasi-rents increases since both v' > v and l{ > Vc- In fact, both val-
ues are increased by an amount equal to the appropriation which each party makes
of the cost saving:
v' - v =fJ (J - m) c [31]
l{ - Vc = (J - fJ) (J - m) c [32]
5.1.1.2 The effect of higher quasi-rents on independence
It would be erroneous to interpret the above as showing that the provision of non-
audit services makes the auditor more dependent on his client,z since it also makes
him more dependent on all his other clients, which generates a compensatory incen-
tive which could be more powerful. In terms of the independence decision
examined in Chapter 3, the provision of services increases the quasi-rents associated
with the client in question, v, and with them the cost of independence, but it also
increases the value of the quasi-rents associated with all other clients, Q, and there-
fore the cost of dependence. Since in the independence condition defined by
expression [11], the impact of these consequences is affected by different coeffi-
cients ('7, on one side; 1l and k'l' on the other, as well as by the number of clients, n,
which defines the size of Q), the net impact will depend on these coefficients and on
the number of clients. Whether the predominant effect is one way or the other de-
pends on client diversification: as from a certain degree of diversification, the
dominating effect is positive, and thus independence is reinforced by the provision
of services.
Analytical demonstration of this affirmation is simple. Subtracting the two terms
from expression [11], the increased cost of being dependent (i.e. the incentive for
independent conduct) is
[33]
and assuming uniform clients, which enables us to substitute Q by [n-lJv), it is

transformed into:
[34]
2 Grout et al. take care not to do so, introducing a crucial qualification C ... at this level of analysis ... "
[1994, p. 330]) which essentially seems to mean " ... assuming that the auditor has a single client".
It is now clear in this expression that the effect of an increase in v depends on the
sign of the content of the fIrst brackets. Also, by deriving this expression in relation
to v, it turns out that the effect of an increase in the quasi-rents, v, on the increased
cost of being dependent, will be positive if
1l"+1l"kln-l)-17>0 [35]
a condition which, when expressed as a function of the number of clients, n, be-
comes:
n>J+ 17 -1l" [36]

1l"kq
and reveals that (purely through the assets specifIc to current clients) the provision
of non-audit services favors the economic incentives towards the independence of
those auditors with a high degree of client diversifIcation and prejudices the inde-
pendence of those others with little client diversifIcation (always provided,
predictably, that the probability that the client who receives a qualifIed report will
change auditor, 17, is greater than the probability of insolvency or detection of an
unduly favorable report, n; otherwise, this effect on independence of providing
services will be positive for any number of clients).
5.1.2 The increase in firm-specific assets favors independence
The ambiguity of the effect on independence of providing non-audit services, based

on the results just obtained, relates solely to the changes in quasi-rents which are
associated with current clients {in short, to changes in the value of the assets which
are specifIc to these clients).3
The provision of non-audit services has a positive effect on independence, fur-
thermore, attributable to the increase in volume of assets specifIc to the firm. This
increase derives from the fact that many of the assets required to provide non-audit
services will lose value as a result of possible audit irregularities, with the conse-
quences already analyzed in Section 2.2.1.3. It was then also shown that, more
generally, the provision of non-audit services extends the possibilities for using
3 In addition, the analysis is biased against independence due to its static nature. In particular, the
model does not consider the corrective incentives which the market generates so that if the provision of
non-audit services is in fact advantageous in relation to the cost saving, but harmful in terms of inde-
pendence, firms invent and adopt patterns which reduce these harmful consequences. An example, in
the terms of Figure 3.2: if the provision of services increases the value of assets specific to each client, v
and Q, but reduces the effect kq, firms could increase their diversification whilst maintaining a similar
level of incentives. Section 5.3 deals with other similar possibilities.
firm-specific assets, of both the directly productive type (basically knowledge) and
those of a contractual nature. In particular, it enables reputational assets to be used
to safeguard transactions connected with these services. The effect of providing
such services (compared with a hypothetical situation in which such provision is
restricted) on the volume of frrm-specific assets is thus to increase their value, and it
can therefore be expected that the cost incurred by the auditor will also rise if he
acts in a dependent manner. This analysis is confrrmed empirically in the results
obtained by Davis and Simon (1992), already commented on in Section 3.4.2: it can
be expected that an auditor affected by an adverse reputational incident will experi-
ence difficulties in maintaining his prices and/or gaining new non-audit service
contracts. The provision of such services thus provides an additional incentive for
maintaining a level of quality consistent with the reputation of the frrm.
5.1.3 Other effects
Although the main effect of providing non-audit services is felt through changes in
the quasi-rents, as just shown, other possible consequences are also worthy of men-
tion. Let us briefly look at the most important.
a) Bargaining power in the allocation of quasi-rents associated with non-audit
services. When assessing how the bargaining power of auditor and client can
change as a consequence of the provision by the auditor of non-audit services, the
general analysis of which of the two parties is better informed regarding the vari-
ables which can affect the bargaining can be reproduced as a first approximation.
One would thus be tempted to think that, in relation to recurrent work, it is unlikely
that there will be differences compared with auditing as such, as analyzed in Section
3.3.1.2. With respect to ad hoc work, the auditor will generally know better than
the client the costs of the consultancy work, both that incurred by the auditor and
the cost normally incurred by the client. However, clients also develop knowledge,
albeit generic, of the cost of this type of work. Moreover, they are better aware than
the auditor of alternative prices of competitors. Consequently, it seems unlikely
that there are major imbalances in the bargaining power of the two parties.
This effect can be evaluated with more precision by a realistic simulation of dif-
ferent bargaining powers in each segment, analyzing how the volume of the
auditor's quasi-rents varies in relation to those of the client with and without non-
audit services. Assuming that m = 0.7 and that the other parameters take the
following values: c = 100, C = 150, '7 = 0.04, r = 0.05, the ratio between the audi-
tor's and client's quasi-rents increases or diminishes as a function of the parameters
a and} as shown by Figure 5.1. The provision of non-audit services increases the
relative amount of the auditor's quasi-rents only if the bargaining circumstances are
such that the latter appropriates a proportion of the quasi-rents from non-audit
services exceeding the proportion appropriated from auditing, as shown by the fact
that the points where the curves intersect is equal to the ratio between the amounts
of the quasi-rents (wherej3 = a).
2.00
1.75
1.50
1.25
1.00
0.75
0.50
0.25
~----~---- __----~--~-----TP
0.2 0.4 0.6 0.8 1.0
Figure 5.1. Effect of providing non-audit services on the ratio between the quasi-
rents appropriated by the auditor and client
Explanation: The ratio between the values of the quasi-rents which the two parties receive changes from
vlvc (the horizontal straight lines, which do not vary withP) to v'1V;:.
If in fact the provision of non-audit services provides the auditor with greater
bargaining power, the quasi-rents and initial price discount would increase even
more, as demonstrated in Section 5.1.1.1. The effect on independence would be
positive or negative depending on the conditions examined in Section 5.1.1.2 and,
in particular, on diversification of client portfolio. It was demonstrated in that Sec-
tion that, with a sufficiently diversified portfolio, the provision of services
intensifies the economic incentives for the auditor to act with independence in rela-
tion to each of his clients.
b) Probability of auditor switching. There is controversy in the literature as to
the effect of providing non-audit services on the client's attitude. Goldman and
Barlev (1974) were the first to argue that consultancy services make the client more
dependent on the auditor and, therefore, in the event of disagreement, the reduce the
credibility of a threat by the client to replace the auditor. These authors consider
that the provision of non-audit services increases the power and independence of the
auditor in relation his clients because the auditor moves from resolving what the
authors consider a routine problem for the benefit of third parties to resolving non-
routine problems to the benefit of the client himself (1974, p. 715). Other authors,
however, disregard this effect because, as already mentioned in Section 3.3.1.2,
they only deal with the auditor's quasi-rents. For example, to Beck, Frecka and
Solomon (1988a, p. 54), the economic tie depends on the current value of the audi-
tor's future profits associated with the relationship, but they do not take into account
that these profits depend on whether or not the client breaks with the auditor, and
this decision will vary with the client's quasi-rents.
c) Costs of collusion between auditor and client. Theoretically, it seems reason-
able to assume that the provision of other services, the prices of which are difficult
to compare, could enable the client to remunerate the auditor for a "soft" attitude on
the part of the latter. Technically, if there are collusion costs between auditor and
client, these could be reduced by the provision of services. Nevertheless, this effect
is marginal-whatever the regulatory framework, collusion between client and
auditor will probably not give rise to significant transaction costs and it is even less
likely that they will be sufficiently high to reach an order of magnitude comparable
with the other costs that are present in this case. 4
d) Specific assets connected to excess capacity. It is reasonable to assume that
the difficulties in relocating resources will be greater, the narrower the range of
services offered by firms. For this reason, those firms specializing in the provision
of a single type of service will be more affected by a break with a client. Conse-
quently, the effect of rules restricting the presence of auditors in providing related
services in this way is to prejudice the relocation of resources and, therefore, in-
crease the cost of demonstrating auditor independence.
e) Professional liability. Finally, with respect to the possible influence of pro-
viding non-audit services or otherwise, on professional liability, the empirical
observation should be mentioned that liability after an audit failure is increased by
the fact that non-audit services have been provided. 5
5. 1.4 Empirical evidence regarding the effects of the supply of

non-audit services on independence in fact
Traditionally, both audit firms themselves, impliedly by adopting preventive rules,

and professional bodies and sector reports, have acknowledged that providing serv-
ices could compromise auditor independence." Nonetheless, the many empirical
4 Pioneering works on the transaction costs of collusion are those of Tirole (1986 and 1992). It has
been argued in a ditTerent work why these costs are scarcely relevant in the case of auditing (Arruiiada
and Paz-Ares, 1997, pp. 52-3).
5 See Beck et al., (1988a, p. 53, n. 5, in relation to the Yale Express case).
6 See, for example, the Cohen report which indicates how "there is little question that the provision of
some other services to audit clients poses an obvious potential threat to the auditor's independence"
studies carried out have not provided evidence that independence in fact nor objec-
tivity is prejudiced by such provision. On the contrary, the numerous professional
analyses of the problem coincide in that in examining cases of audit failure it has
not been possible to detect an appreciable negative influence from provision by the
auditor of non-audit services. 7
In the academic field, various studies have, using different methods, also tried to
provide indirect evidence of the effect of non-audit services on auditor independ-
ence in fact. Before commenting on the most important of these studies, it should
be pointed out that the results obtained do not enable a negative influence on inde-
pendence to be inferred from the provision of non-audit services.
a) Shareholders do not penalize the supply of non-audit services. Glezen and
Millar (1985) observe that approval of the auditor's work by shareholders in the
United States was unaffected by disclosing the volume of other services which the
audit firm had been providing to clients as well, obviously, as auditing. The scale
of such non-audit services also failed to affect the rate of auditor approval. The
results thus suggest that the provision of services does not compromise independ-
ence in the view of shareholders.
b) Positive correlation between supply of non-audit services and auditor
switching. With the aim of examining whether the provision of other services preju-
diced independence, Beck, Frecka and Solomon (1988b) examined how auditor
rotation varied with the volume of such services, finding that the higher the volume
of services the lower the rotation. This finding, however, can be interpreted in
several ways, and is thus uninformative as to the effect which providing services
can have on independence. Firstly, the direction of the causality is arguable: on the
one hand, it can be argued that the low rotation, and therefore greater continuity,
increased the auditor's cost advantage in providing other services. Conversely, it is
possible that the high volume of services links client and auditor more closely in
such a way that both have an interest in greater continuity in their relationship. In
addition, this second interpretation can be assessed in either direction, positively or
negatively, depending on whether it is considered that the low rotation is due to
greater client dependence, (which enables the auditor to be more rigorous without
being dismissed), or to lesser auditor independence, (which enables the client to
achieve softer audits which prevent him having to dismiss the auditor). Finally, the
study was replicated, and different results were obtained. Thus, using another sam-
(1978, p. 102, emphasis in the original). Moreover, the use by firms of patterns aimed at avoiding
conflicts of interest reaffirm the possibility of their existence.
7 See AICPA (1969, p. 52), where the generic allegations formulated in the I 960s were examined, and,
in particular, Cohen (1978, pp. 96-8), for a study of cases litigated and a detailed examination of the
only cases in which these allegations were specified; also POB (1979, pp. 33-4).
pIe of companies, DeBerg, Kaplan and Pany (1991) found that the decision to
switch auditor did not depend on the volume of non-audit services which the auditor
provided to the client, both when these services were recurrent, and when not.
Similarly, in a study of the relationship between non-audit services, prices and in-
dependent indicators, using data from the 500 largest Australian companies, which
are obliged to divulge the amount of fees paid to their auditors for all types of serv-
ice, Barkess and Sirnnett (1994) concluded that no relationship at all could be seen
in this market between the volume of such services and replacement as the client's
auditor.
c) Evidence ofprice maintenance. Davis, Ricchiute and Trompeter (1993) argue
that the results of their empirical study of audit prices, already mentioned in Section
4.2.3.3, are conclusive that joint production does not compromise auditor objectiv-
ity. They base this assessment on the fact that the price of auditing per hour does
not vary with the provision of other services. s This conclusion is solely based on the
fact that they can explain the differences in audit service prices to clients, with and
without additional services, by differences in the amount of auditor work and, there-
fore, it can be expected that both types of client provide the auditor with the same
level of quasi-rents. It seems logical, however, that in this sector the additional
quasi-rents associated with the other services, if they exist, are received through the
price for such services since they are less subject to professional regulation, as the
authors themselves appear to acknowledge (p. 149).
d) Litigiousness is positively correlated with the supply of non-audit services. It
was observed in the United States between 1967 and 1994 that the proportion of an
auditor's income from the client has a highly positive correlation with the litigious-
ness of the audit, which has been interpreted as a lack of independence. 9 This result
is consistent with the negative effects of providing non-audit services on the ap-
pearance of independence: a higher proportion of services makes it easier to
convince judges and juries that there are problems of independence, making litiga-
tion more profitable as a result.
e) Attempts at experimental analysis. The limitations of studies based on sur-
veys and indirect indications have led to the use of simulation experiments in which
games are played between contracting parties with the help of computers. In these,
8 See their Table 4 (Davis e/ al., 1993, pp. 145-6).

9 See, in particular, Lys and Watts (1994) and Stice (1991). Lys and Watts (1994) find a relationship in
the same direction for other variables: poor stock exchange performance and client size, the non-struc-
tured character of the audit, and the existence of a qualified report. They argue that the relationship
with poor stock exchange performance is logical, given that a fall in stock prices is a necessary condi-
tion for the existence of damage. The relationship with client size could reflect better possibilities of
compensation for loss, or the effect of some hidden inter-relationship. Finally, the relationship between
litigiousness and qualified audits is explained by the almost certain existence of causes common to both
variables.
attempts are made to reproduce market circumstances before analyzing how the
results change when different structural parameters are varied. The complex ex-
periment designed by Dopuch and King (1991) constitutes an initial attempt to
analyze the impact of providing non-audit services on auditor independence with
this technique. Their central conclusion is that prohibiting auditors from providing
these services to their audit clients does not affect the level of informational effi-
ciency in the market, although it favors auditor-consultants specializing in either
auditing or consultancy, and could damage competition between them. Nonethe-
less, it should be borne in mind that the experiment is subject to considerable
limitations, described by Berg (1991). Corless and Parker (1987), using a method
closer to survey than experimentation as such, also tried to verify the contention that
the auditor's judgment suffered when knowing that his own fIrm was involved in
design of the client's internal control system. Their results refute the argument by
showing auditors to be equally tough in examining the accounts of clients to whom
their fIrms had provided non-audit services as those of other clients.
5. 1.5 The reaction of audit clients
As mentioned in Chapter 1, Agency Theory argues that clients employ the services
of an independent auditor to reduce conflicts with those providing fInancial re-
sources (both shareholders and creditors).'o They will therefore be sensitive to the
possibility of those using the information perceiving that the auditor is excessively
dependent on the client. Otherwise, they would suffer a penalty in more expensive
access to the capital market.
Clients therefore have an interest in adopting policies to offset any negative ef-
fects which the provision of audit and other services by the same auditor might
involve. Measures range from not contracting any type of service to contracting
only those with less risk of compromising independence, both in fact and in appear-
ance. For example, it has been noted that clients with characteristics which make it
likely that they could be subject to higher agency costs tend to contract recurrent
services, which may be thought to be more compromising, to a lesser extent from
their auditors (Parkash and Venable, 1993). The observation of DeBerg et al.
(1991), already mentioned, that clients reduce the acquisition of recurrent consul-
tancy services from their auditors after a change of auditor, could also be similarly
interpreted; also that of Glezen and Millar (1985), to the effect that since sharehold-
ers continue to approve the continuity of auditors after disclosing the volume of
10 See Jensen and Meckling (1976), Fama and Jensen (1983a and 1983b) and Watts and Zimmerman
(1986).
non-audit services which they provide to the client, it does not seem that this provi-
sion prejudices independence.
5.2 Effects on the appearance of independence
5.2.1 Studies based on surveys
Several surveys have been carried out in many countries into the perception of
auditor independence amongst users and professionals, a large number of which
include questions regarding the effect of providing non-audit services. Their results
tend to indicate that the provision of such services prejudices the appearance of
auditor independence, albeit with substantial differences.
The conclusions from these studies are difficult to evaluate, however. Firstly,
they provide contradictory results, as mentioned. Moreover, the methodology,
based on opinion surveys, is subject to serious limitations, which will be mentioned
in Section 5.2.2. Finally, it is arguable which opinions should be taken into account
and, if so, with what weighting. This latter observation is more pertinent when
noting that one of the common results of these studies is that the provision of serv-
ices prejudices the perception of independence all the more the less the knowledge
amongst respondents regarding the role of the auditor.
a) The appearance of independence suffers with non-audit services. The pio-
neering work in this field was a study published by Schulte (1965), which even
gave rise to formation of a committee within the AICPA, which reported favorably
on the provision of services to audit clients, as it did not see any harm to independ-
ence in fact, at the same time acknowledging that in this field the profession should
be careful with the appearance of independence (AICPA, 1969). That study was
quickly followed by an initial wave of works, such as those of Briloff (1966), Titard
(1971) and Hartley and Ross (1972), which provided contradictory results, even
within each one. Thus, there are indications in favor of the argument that providing
non-audit services prejudices the appearance of independence in Schulte (1965, p.
590), Briloff (1966, p. 492), Titard (1971, p. 51) and Hartley and Ross (1972, p.
44), whilst Schulte (1965, p. 590) and Hartley and Ross (1972, p. 44) contain con-
trary indications.
A second type of study, somewhat more rigorous, was that begun with Shockley
(1981), who investigated how four groups of experts assessed different factors
which could affect auditor independence. These groups of experts were formed
from partners in large firms, partners in small firms, bank executives responsible for
granting credit, and financial analysts. The factors which the author assumed would
contain a risk of dependency and which should be assessed were: a high degree of
competition in the audit market, the provision of accountancy services to the client,
the small size of the audit firm, and the duration of the relationship between auditor
and client. Only the latter factor turned out to be insignificant. The significance of
the other three, however, was very different: competition appeared as the principal
factor, followed by size and, to a much lesser extent, the provision of non-audit
services in the accountancy field (Table 5.1). Other similar studies include those of
Firth (1980, 1981) and Pany and Reckers (1984). Although these works provide
contradictory indications, the available evidence in general indicates that non-audit
services prejudice independence in appearance. I I
b) Differences of opinion amongst experts. As a consequence of their methodo-
logical weaknesses, the most revealing results of this type of study are perhaps the
different perceptions between groups. As Shockley's results (summarized in Table
5.1) reveal, opinions differ to a large extent between groups of experts regarding
those factors with greatest competitive potential, such as competition itself, the
provision of non-audit services, and firm size. It may, however, be suspected that
either their knowledge of the problem or their interests in relation to it may be di-
vergent.
The results of subsequent studies support the fact that there is a widely varying
knowledge of the problem. McKinley, Pany and Reckers (1985) conducted a sur-
vey amongst bank officials responsible for granting credit to examine their opinion
regarding different features of audits of the accounts submitted by companies ap-
plying for credit. They observed that the provision of non-audit services does not
affect perception amongst these decision-makers regarding auditor independence.
Similarly, Knapp (1985) examined how different factors in the auditor-client rela-
tionship environment affect perception amongst US professionals regarding the
capacity of auditors to resist client pressure when there is a conflict between the
two. In this study, the public used as reference was also made up of bank execu-
tives responsible for deciding to grant credits to undertakings. The results do not
confirm the hypothesis that non-audit services prejudice perception of independence
amongst fmancial users: 12 in the views of the experts, this factor has a minimal
explanatory capacity, whether alone or when interacting with other variables. 13
Lastly, the opinion as to auditor independence and the reliability of fmancial state-
II Other works have analyzed more specific and practical individual aspects. For example, in a study
into perceptions of independence, Lowe and Pany (\994, 1996) found that contracting on a small scale
with clients to provide common services to third parties (so-called co-contracting), did not harm the
perception of independence and that this perception was reinforced when the consultancy and auditing
services were provided by different divisions and were ad hoc, not giving rise to a stable relationship.
12 On the contrary, it enables them to conclude that the perception of independence increases with the
objectivity of the technical standard to be applied and with the financial weakness of the client.
13 See Knapp (1985, Table I, p. 206).
ments amongst bank loan officers and fmancial analysts making credit and
investment decisions is not affected even with extreme levels of non-audit services,
such as 90 percent of average audit fees over the past three years (Pany and Reck-
ers, 1988).
Table 5.1. Variance explained by significant effects by group of experts

Big Eight Other CPA Commercial Financial
Partners Partners Loan Officers Analysts
Effect
relative relative relative relative
w2 w2 w2 wJ
w1 w1 w2 w2
Competitive environment 0.062 0.218 0.138 0.726 0.067 0.298 0.117 0.480
Provision of accounting
0.010 0.035 0.052 0.274 0.009 0.040 0.092 0.377
related services
Small size of audit firm 0.205 0.722 0.148 0.658 0.035 0.143
Audit firm's tenure 0.001 0.004
Competition-size
0.005 0.018
interaction
Services-size interaction 0.002 0.007
Total 0.284 1.000 0.190 1.000 0.225 1.000 0.244 1.000
Note: The relative W2 is an estimate of the proportion of variance explained by a specific effect relative
to the total variance explained by all significant effects.
Source: Shockley (1981, p. 793).
c) Differences of opinion among different types of auditor. A second explana-

tion of the differences between groups is that those responding to the survey had
different interests which slanted their opinions and/or replies. An indication of this
is provided by the results of Shockley's work shown in Table 5.1. Comparing the
first columns gives the impression that small audit firms consider those factors
which prejudice their competitiveness as negative to independence: market compe-
tition itself and a provision of accountancy services, for which they are perhaps not
in a good position. It is revealing that, on the contrary, they do not consider the
factor which the other experts (not just the partners in large firms) consider funda-
mental, namely firm size, as significant.
d) Differences of opinion between countries. Perceptions of auditor independ-
ence vary between countries to a large extent, an understandable response to the
diversity in accounting traditions. In this respect the differences which have been
observed between German first level auditors (those who can carry out any type of
audit, irrespective of client size, known as wirtschaftsprilfer) and the Americans, are
interesting. The survey by Dykxhoom and Sinning (1981) shows that German
auditors considered that their independence would not be affected in various situa-
tions in which the American regulations (and subsequently also EU regulations) did
consider they would be compromised, if not in fact at least in terms of appearance.
This was the specific case with situations in which the auditor carried out all or part
of the client's accountancy work.
The case of Spain might be illustrative of the differences and problems which
arise, as well as of the lack of information amongst the public regarding this matter
in countries where there is little tradition for external auditing. Comparison with
the results obtained from previous studies in a country with a longer tradition, Great
Britain, is also instructive. The study by Garcia Benau and Humphrey (1992) into
perceptions of audit work included two relevant questions in terms of the topic
under discussion here. The responses to these questions, summarized in Table 5.2,
are particularly interesting because of the ambiguities and contradictions they re-
veal:
• Firstly, to the question whether auditors should identify means for improving
management efficiency, all the groups surveyed (auditors, fmancial executives
and users of fmancial statements) replied positively without significant differ-
ences between them, and with similar answers to those of their British
counterparts. 14 By not asking how the information required to improve man-
agement should be channeled, the reply is thus not very informative. It seems
clear, however, that the public was aware of the fact that the auditor can accu-
mulate information relevant to improving management. If it is considered that
the provision of non-audit services is one means, surely the most effective, to
stimulate and channel the production of this information, it could be taken that
the opinion expressed impliedly favors such provision.
• Secondly, to the question whether they believe that audit firms should not pro-
vide consultancy services to their clients, the average opinions were slightly
unfavorable amongst auditors (3.4) and slightly favorable amongst fmancial ex-
ecutives (3.7) and users (4.2), although all were close to the middle of the
opinion scale, from 1 to 7. Nevertheless, these average answers probably suffer
from substantial upwards deviation (i.e. against the provision of services) be-
cause the drafting of the question could be slanted. The problem is highlighted
when observing that respondents were asked whether they agreed or not that "An
audit firm should: ... not provide management advisory services to its audit cli-
ents" (Garcia Benau and Humphrey, 1992, p. 290, emphasis added). It may be
suspected that the average reply to this answer was significantly different from
14 The answers of the British and Spanish experts are of questionable comparability, however, due to the
different average response rates to the questionnaire (38.2% in Great Britain; 15.3% in Spain).
the complementary result which would have been obtained by asking alterna-
tively whether they were in agreement or not with audit ftrms in fact providing
consultancy services. The reason is that the drafting used in the study could give
respondents a negative predisposition, transmitting prejudice regarding these
services, unlike the possible alternative.
• In addition, a third aspect of the results is striking and worrying: the fact that as
many auditors as fmancial executives and users unanimously were of the opin-
ion, (in average inter-group terms), that auditors should not basically act to
obtain proftts, is at least of concern. Interpreted literally, these responses appear
to reveal an anachronistic and idealized vision of auditing activities, and of the
economic system itself, amongst the respondents, a view which, if conftrmed,
could give rise to difftculties when formulating sensible professional rules.
Table 5.2. The opinion of experts on different questions relating to the auditing
process and possible regulations on audit firms in Great Britain and Spain
Kruskal-
Ad. Financial Users Wallis
Question Country u ltors Directors
Test
Auditors should be identifying ways to Britain 5.5 4.5 5.0 S'
improve management efficiency Spain 4.8 4.6 5.0
An audit finn should not provide manage- Britain 1.6 3.1 3.5 S'
ment advisory services to its audit clients Spain 3.4 3.7 4.2 S'
An audit firm should not act primarily Britain 1.9 3.3 3.2 S'
to make a profit Spain 4.6 4.7 4.7
An audit firm should not be able to earn Britain 6.0 5.4 5.0 S'
more than 15% from anyone client Spain 4.7 4.8 4.7
Notes: The averages were based on a 7 point answer scale, in which I represented that the respondent
"strongly disagree" and a 7 "strongly agree" with the corresponding statement. S' = across-group
difference significant at I % level of statistical significance.
Source: Garcia Benau and Humphrey (1992, pp. 286 and 290).
5.2.2 General limitations of the stUdies based on surveys
A common problem with the type of study referred to in the previous section is that
they use artiftcial data from surveys in which the opinion of the public or different
experts is sought regarding certain variables or, at most, simulated situations. The
researcher is essentially confronting respondents with hypothetical situations and
asking for their opinion on them; speciftcally, how they believe that auditor inde-
pendence would be affected by a situation of conflict of interest arising because one
or more of the factors is present, whose effect they are trying to study (e.g. financial
weakness of the client, the fact that the auditor provides non-audit services, compe-
tition in the audit market, etc.). This methodology places respondents in a situation
where the elements in play are presented in a very explicit manner. As a result,
there is a greater risk that respondents will be involuntarily inclined towards hyper-
sensitivity to the variables presented to them as a result of the design of the survey
itself and therefore provide biased responses. It suffices to consider the huge pre-
cautions taken in market research to avoid this type of problem. Unfortunately, the
large budgets for market research are not usually available for studies such as those
referred to here. Consequently, the experimental design precautions within the
scope of the authors of these works are extremely modestY
The research of McKinley, Pany and Reckers (1985) and Pany and Reckers
(1987, 1988) is relatively unscathed by this criticism. Instead of carrying out a
survey and asking directly about the effect of providing non-audit services on inde-
pendence, they presented a dossier of financial information regarding a business to
a series of bank loan officers and/or fmancial analysts, for them to evaluate whether
or not a loan should be granted to them or an investment made in their stock. The
subjects had to take a decision in this respect and also assess the quality of the
financial statements and independence of the auditor. All the dossiers were the same
except, as far as we are concerned here, in one respect: they included different lev-
els of non-audit services provided by the auditor. As the individuals considered a
single dossier, they were unaware of the real focus of the study and their replies
were thus not conditioned by the "demand effect", which usually leads respondents
to answer in line with what they believe are the preferences of the questioner. In the
first of these studies, McKinley et al. (1985) observed that the analysts did not con-
sider that there were differences in the quality of the financial statements or in the
independence of the auditor when the non-audit services ranged from zero to 30
percent of the audit fee. These results were confirmed in two later, more detailed
studies. Firstly, Pany and Reckers (1987) showed that a judgment as to auditor
independence is conditioned by the method used to elicit the response. Specifically,
the observed increases in the provision of non-audit services only harmed percep-
tion of independence when the respondents are explicitly questioned on the level of
such services but not when each individual observes a single degree of expendi-
ture. 16 Pany and Reekers (1988) found that in this experimental context the
provision of non-audit services has no effect, even with a high level of services. 17
15 See, for example, Knapp (1985, pp. 209-10) and Shockley (1981, pp. 791 and 796-7).
16 See, however, Schepanski, Tubbs and Grimlund (1992, especia\1y pp. 130-1) for a critical analysis of
these results.
17 It is of course possible, as tends to occur when insignificant experimental results are obtained, that
this absence of effect is due to the fact that the variability introduced into the experiments is incapable
of generating an appreciable effect.
5.2.3 The correlation between supply of services and qualified

auditor opinion
One very direct although, as will be shown, deceptive manner of verifying whether
independence, or the quality in general of auditing, are affected or otherwise by the
provision of non-audit services consists in examining the correlation between the
provision of such services and various indicators which could at the same time be
related to audit quality.
The work of Wines (1994) falls into this context. Taking a sample of 76 Austra-
lian companies obliged to disclose information of this type, he found a correlation
between issuing a qualified audit report and the volume of non-audit services pro-
vided by the auditor to the client. This result, however, does not imply a judgement
as to the quality of the auditing in fact, but as to the appearance of quality and, more
specifically, the appearance of independence, as acknowledged by the author him-
self who does not express an opinion on independence in fact except to indicate that
his results tells us nothing about it, neither does he allude to technical competence.
This negative correlation between qualified audit report and volume of non-audit
services provided by auditor to client cannot provide information regarding inde-
pendence in fact because of the possible existence of a self-selection bias: there is
no certainty-on the contrary the opposite can be suspected-that the fmancial
conditions of companies who acquire non-audit services are similar to those who do
not acquire them. It seems reasonable to assume that, since the acquisition of non-
audit services is an unnecessary expense, companies in a worse fmancial situation
will abstain from acquiring them. The weakness of its balance sheet will thus be the
cause, firstly, of not acquiring such services and, secondly, of qualifications or res-
ervations in its accounts, to the extent that the directors would wish to conceal the
company's weakness. The fact that, as Wines himself indicates and verifies, his
results are provided by the average companies in his sample, supports the existence
of some type of bias. In this respect, the doubt regarding Wines' work increases
when observing that, with data from a more representative Australian sample (2,094
auditors), no relationship appears between providing services and issuing qualified
reports (Barkess and Sirnnett, 1994, p. 105), a result also confirmed by a later study
(Craswell, 1998).
5.3 Private safeguards of independence with respect to the

supply of non-audit services
In the first Chapter we examined how independent auditing of an undertaking'S

financial statements mainly serves to reduce the conflicts which can arise in con-
tracting with those providing fmancial resources, whether shareholders or creditors.
For this reason, clients are interested in users of the audited accounting information
perceiving the auditor as independent of the client.
Such client concern leads to a concern amongst auditors themselves and it is un-
derstandable that they implement policies of all types aimed at reinforcing their
independence. This section examines some of the formulas adopted for this pur-
pose by firms and professional self-regulatory bodies, focusing on the provision of
non-audit services. The objective is twofold. Firstly, to complete the empirical
indications which have just been presented. To the extent that one purpose of these
formulas is to contain potential problems of independence (particularly those relat-
ing to the appearance of independence), they provide evidence that such problems
exist. Secondly, it is necessary to be aware of their content before tackling the
regulatory problem in Chapter 7, since this will be less pressing to the extent to
which any shortcomings are resolved, or at least alleviated, by private decisions,
whether taken by firms or professional associations.
5.3. 1 Self-regulatory safeguards
Both audit firms, and their professional organizations, must maintain a good reputa-
tion to survive. For this reason, they have taken care over the years to adopt various
standards and organizational patterns which reduce the risks which could harm both
auditor independence in fact and, in particular, the appearance of independence.
The general guidelines of the Federation of European Accounting Experts (FEE) are
representative of the spirit of these standards, when stating the following:
In considering whether to undertake other assignments for an entity of which he is

auditor, the auditor's duty is to satisfy himself that the assignment will not affect his
ability to act in an objective, independent and impartial capacity. He must especially
make sure that he will not be in a position to take any decision on behalf of the audit
client. He must evaluate the matters discussed in section I above and take any neces-
sary steps to safeguard his objectivity and to satisfy those with a legitimate interest
that he has done so (FEE, 1995, Section 4.2.1).
Four main patterns can be distinguished in the specific implementation of this

general philosophy:
a) Different rules for non-audit and audit services. Non-audit services present
several characteristics which distinguish them from auditing. Firstly, the consultant
must gather data and formulate conclusions, whilst in auditing he is only attesting to
the reliability of information issued by others. In addition, the nature and content of
the work is determined by agreement with the client and the latter is the sole benefi-
ciary. These characteristics mean that quality problems have various different
properties in the two types of activity. Coherently with this, professional nonns
usually dictate distinctive procedures and standards of conduct, adapted to the par-
ticular features of each type of service. For example, in addition to common
standards of professional competence, due professional care, adequate planning and
supervision and use of sufficient relevant information, the AICPA, in the case of
non-audit services, mandates those of serving the client's interests, establishing an
understanding with him in which the responsibilities of the parties and the nature,
scope and limitations of the services to be provided are set out, and notifying him of
possible conflicts of interest, significant reservations regarding the scope or benefits
of the work and the significant engagement fmdings or events (AI CPA, 1996, Sec-
tions 6 and 7). In addition, a contract for providing non-audit services can restrict
the information which can be used, unlike in the case of auditing (AICPA, 1996,8).
Furthermore, audit services connected with such services must be carefully sepa-
rated, informing the client of the different standards which will be applied (AI CPA,
1996, 77) and with clearly separate reports issued (AICPA, 1996, 78).18 In this
field, the FEE is less specific, simply indicating that:
Advisory work and representation of a client in financial and tax matters is normalIy
compatible with an audit by the same auditor, though the auditor must adopt safe-
guards to ensure that his objectivity is not impaired by the review of that work in the
course of the audit or by adoption of a posture of advocacy. Advisory work which
constitutes a professional or technical clarification of a particular matter or an expert
description of alternatives is unlikely to give rise to any issue impairing the auditor's
objectivity (FEE, 1995, Section 4.2.2).
b) Generic restrictions on the content of all services: abstaining from partici-

pating in decision-making. Opinion is unanimous that the auditor must not overstep
the dividing line between providing advice and taking decisions. For example, the
AICPA already referred in 1975 to the crucial importance of "strict adherence to an
advisory role" (p. 20). The FEE is also very clear in this respect:
Any management role undertaken by the auditor would give rise to everyone of the
types of threat [to objectivity] listed in § 3.7.[1') The auditor should not therefore take
18 See the detailed analysis of these services in AICPA (1996), as well as the ethical code of the Inter-
national Federation of Accountants (IFAC, 1996).

l ' To the FEE, specific risks and threats relating to objectivity include: (a) the "Self-Interest Threat or
Risk", which can arise from factors such as a direct or indirect financial interest in the client, a wrongful
desire to increase fees or obtain additional work or the fear oflosses due to an excessive dependence on
a single client; (b) the "Self-Review Threat or Risk", in any service related to a previous audit or non-
audit assignment; (c) the "Advocacy Threat or Risk" when faced with adversarial proceedings or
situations; (d) the "Familiarity or Trust Threat or Risk", if the auditor becomes over-influenced by his
knowledge of the client; and (e) the "Intimidation Threat or Risk", if the auditor suffers coercion from
the client or an associate of the client (FEE, 1995, pp. 9-10).
any decision on behalf of the company or its management in the course of the audit
assignment or of the provision of other services to the audit client. He should neither
play any managerial role in the audited company, nor hold any managerial position in
it. In situations where the auditor advises or assists the audit client on any matter, he
should make clear that full responsibility for the decisions taken and for the financial
statements remains with the management of the audited company (FEE, 1995, Sec-
tion 4.2.3).
c) No provision of particular services. One of the most common measures in

this respect has been a prohibition on providing those services which could com-
promise independence, either in fact or appearance. This includes in the US for
instance, the profession's rules under the SEC Practice Section of the AICPA which
prohibit an audit firm from providing executive recruitment services for SEC audit
clients or actuarial services for insurance company clients. In addition, the profes-
sion has recently developed extensive guidance regarding the performance of
internal audit services for audit clients.
d) External review. In countries in which self-regulation has been able to de-
velop fully, professional organizations have also developed specific external review
procedures to monitor auditing quality. In the case of the United States, for exam-
ple, these include review of all litigation in which audit failure is alleged, in the case
of firms in the AICPA SEC Practice Section, and investigation by the AICPA Ethics
Division of firms generally when complaints are made against partners of audit
failure, resulting to possible disciplinary measures which can include termination of
AICP A membership.
5.3.2 Organizational safeguards
Firms with a better reputation likewise place maximum emphasis on avoiding pos-
sible conflicts of interest which could compromise the independence of their
auditors in their audit work. To this end, they usually take additional precautions in
order to reduce such conflicts to the economically viable minimum. To illustrate
this it suffices to mention some typical solutions, distinguishing between those
linked to the audit process and those applied to the provision of non-audit services.
5.3.2.1 General internal safeguards of objectivity
Throughout their existence, audit firms have developed numerous guidelines aimed
at controlling the quality of audit services and, in particular, to protect them against
possible threats to objectivity. The most important include:
a) Internal rules on independence. Most fIrms require that their partners and
staff follow predetermined rules on independence. In many cases, this is imple-
mented through lists of relationships, activities and services which are prohibited, as
well as through the dissemination of a restricted list of clients and entities that are
restricted for investment. In some fIrms, partners and managers have to report their
personal investments. Firms also take signifIcant disciplinary action against indi-
viduals who fail to comply with policies on independence.
b) Partner evaluation and compensation methods. Two features of the incentive
structure are particularly important in encouraging partners to take a broad perspec-
tive with less influence accorded to specifIc clients, offIces or regions. Firstly, the
practice of remunerating partners according to several performance variables which
include quality, risk management, client service and subordinates' development.
This is not based primarily on revenue generation or personal clientele as some-
times the case with law fIrms. Secondly, partners' compensation is frequently based
on the global results of the fIrm, not on the results of anyone offIce, region or
country.
c) Quality control procedures. The quality of audit work is assured by the im-
plementation of quality control measures, which fIrstly protect the audit process
and, secondly, review the results of such process. Thus, in many fIrms audit work is
reviewed no only by the person who prepared the work but also by a designated
reviewer. Moreover, the standard auditing process includes procedures to resolve
disagreements in the audit team. Similarly, the audit engagement partner has the
obligation to consult on those decisions requiring signifIcant judgement with fIrm
experts not associated with the audit engagement. Professional standards also re-
quire that a partner independent of the audit engagement team review and consent
to all signifIcant audit decisions. Also, audit engagement partners serve public
companies for no longer than a certain number of years (seven in the United States
for the SEC Practice Section of the AICPA) to promote a fresh perspective and
independence from company management. At a more general level, fIrms periodi-
cally assess the quality of their audit work, including conformance with its
independence policies, through annual internal inspection programs. Typically,
personnel from one offIce inspect the work of another offIce under the direction and
oversight of a global risk management unit. Lastly, the quality control systems of
fIrms in the AICPA SEC Practice Section are subject, at least once every three
years, to peer review by another fIrm, which reports its results to the public.
d) Dynamic nature of objectivity protection guidelines. Firms analyze regulatory
action and litigation where audit failure is alleged to identify process failures, in-
cluding those failures relating to independence. Procedures are then adopted to
mitigate the risk of future failures.
5.3.2.2 Specific internal safeguards with regard to non-audit services
The development and importance of non-audit services has led finns to design
organizational patterns and structures with the function of protecting audit inde-
pendence, as in the following cases:
a) Examination of decisions regarding the provision of new services. Before
commencing a new line of business, finns in general carefully examine the implica-
tions of providing each new service in terms of independence and, in particular,
whether the service should be provided only to non-audit clients.
b) Staff specialization. Most activities are, for reasons of functional specializa-
tion, carried out by staff belonging to separate departments or divisions. In general,
staff providing non-audit services are usually different from audit staff and there-
fore do not nonnally report to the same partners. In one large Spanish firm, for
example, only 13.4 percent of non-audit work was carried out in 1996-1997 by staff
from the audit divisions. Furthennore, in this finn, the audit staff are always differ-
ent from those engaged in legal and tax advisory work. This type of specialization
indicates that the largest part of the economies of scope are linked to the processes
of training and developing organizational structure and contractual safeguards, with
knowledge spillovers playing a minor role, if any.
c) Divisionalization. When the same finn offers audit and non-audit services
under the same name, a different division is usually responsible for each series of
services. These divisions are frequently organized as profit centers within the
auditing firm. These units usually have their own management and exert little if
any influence over the audit partner evaluation or compensation processes. Coordi-
nating divisions in such cases can be problematical, as illustrated by the attention
focused on this matter by the business economics and management literatures. The
basic reason is the risk that the coordination may negate the incentives sought by
decentralization. It must be expected that divisions will act in a less coordinated
manner than a functionally organized business would. Moreover, some audit firms
even instruct their divisions not to take account of the effect on other divisions in
their pricing decisions in order to reinforce auditing independence. 2o These aspects
are confinned by the empirical findings in the United States that when consultancy
and audit services are provided by different divisions, less harm is done to the per-
ception of independence. 21 In addition, when each business unit is a separate legal
entity with its own partners and human resources, and it develops its own methods
20 See Leibman and Kelly (1992, pp. 436-7).

21 See on this Pany and Reekers (1984) and Lowe and Pany (1995, \996).
and policies. This additional separation provides more insulation for audit partners
and less sharing of income than a common partnership structure.
d) Rules regarding abstention from involvement in executive tasks. In order to
ensure that auditors do not take decisions appropriate to client executives, fIrms
have for many years developed and applied a whole series of specifIc procedures
and criteria of the following types: balance between client and auditor participation;
appointing an employee of the client as project manager; minimum qualifIcations
and experience of the staff allocated; clear defmition in the contract of the responsi-
bility of executives and the advisory role of the auditor; rules by which work
programs must detail the roles of the client and auditor; specifying that the client
will decide how the project proceeds; and documentation of decisions, both required
and actually taken.
e) Multi-firm structure. Firms have been developing highly complex organiza-
tions which manage to take advantage of economies of scale in certain activities
whilst the various connected fIrms at the same time display considerable autonomy.
As explained in more detail in the next section, this independence is usually moti-
vated by the need to adapt effectively to differing geographical and functional
markets. Furthermore, this autonomy channels away potential conflicts of interest
which may arise and also makes it possible for fIrms to cut their links if these con-
flicts outweigh the advantages of continuing together, as one of the principal
organizations, the Andersen WorldWide network, has been doing with respect to its
Andersen Consulting unit.
5.3.3 The federation affirms
The large audit fIrms are fundamentally federations of professional cooperatives of

national scope, organized under a franchise formula. 22 As such, they share intangi-
ble assets connected particularly with the production and transmission of knowledge
and reputation. Most professional service providers are tending to organize them-
selves on a global scale under hybrid forms in which, by way of federation, fIrms in
different countries and with different specialties link themselves together through
22The terms "cooperative" and "franchise" are not being used in a legal but economic sense. The
cooperative character arises because it is the most qualified workers or professionals who are the resid-
ual owners of the assets and rents of the undertaking. [n this respect, see Fama and Jensen (1983a, pp.
315-7; and 1983b, pp. 334-7). It is also appropriate to talk of "franchise" using this term as a hybrid
business form in which different productive units share the use of certain assets and, in particular,
commercial reputation. For the economic rationale of franchise undertakings in this sense, see, princi-
pally, Brickley and Dark (1987), Ones (1996), Gallini and Lutz (\992), Kaufmann and Lafontaine
(1994), Klein (1995), Klein and Murphy (\988), Klein and Saft (\985), Krueger (199\), Lafontaine
(1992 and \993), Norton (\988 and 1995) and Rubin (1978). This economic concept of franchise
includes but goes far beyond the legal concept used in Continental Europe.
agreements creating reciprocal obligations to implement various common func-

tions. 23 In a very schematic manner, the purpose of this type of organizational
architecture is twofold, firstly seeking to achieve economies of scale at the center
(a) and, secondly, creating conditions favoring efflciency at the periphery (b). The
independence of the different firms making up each network is clear when consid-
ering their capacity to split off and go their own way when they disagree with the
strategy adopted by the network of which they form part (c).
a) Central economies of scale. The organizational duality of these federations
enables them to achieve economies of scale in those tasks in which such economies
are decisive. This is particularly the case with two series of intangible assets which
are of growing importance: knowledge and reputation. The fInns structure them-
selves to produce, accumulate and transmit information. To this end they
implement multiple research, data base and training functions centrally. With re-
spect to reputation, the activity of these federations ranges, depending on the
circumstances, from the development of a common trade mark to, in particular, the
establishment and supervision of minimum quality standards covering both services
and processes. The following are usually included in the typical services provided
by the centers of these federations of firms: research and development tasks con-
nected with the services of all the firms; managing common problems affecting new
and incipient markets such as those of Eastern European countries until recently;
quality control of input, processes and products, ensuring end clients of uniform
quality over a wide geographical area; staff training, a fundamental activity in this
sector, as shown by the investment made by audit fIrms/ 4 and covering certain
risks, subsidizing entry into incipient markets and activities and smoothing fluctua-
tions in local activities.
2J The activities of the international networks of audit firms are controlled to a large extent by their
partners at national level. This particular feature has been attributed to regulatory restrictions at na-
tionallevel on the ownership, management and control of audit firms, with the effect of reserving such
functions to native nationals (See, for the situation within the European Union, Buijink et al. [1996, pp.
113-34]). This is consistent with the claim that most consulting firms, who do not suffer such con-
straints, seemingly organize themselves in a much more integrated and conventional fashion, with
national firms acting more like fully owned subsidiaries of an international corporation. (According to
Buijink et al. [1996, p. 129) on the basis of a survey of the consulting industry carried out by the United
Nations [1993]). Substantial differences also exist between auditing firms, however. For instance, the
failure of the merger between KPMG and Ernst & Young was partly attributed to the fact that Ernst &
Young was much more integrated than KPMG, which conforms more to the model of a federation of
firms which franchise a global brand (Kelly, 1998).
24 For example, in 1997 the leading firm in Spain devoted 10 percent of its fee income to training, when
the average amongst Spanish firms does not reach half this figure. In fact, audit firms have been largely
functioning as an executive supply source. For data and examples, see Molero (1995, p. 64).
b) Incentives for efficiency through decentralization. This federated structure,

however, enables the connected fInns to maintain independence,25 based on differ-
ent typical patterns: there are no connected shareholders or cross-holdings between
them, there are no common partners, they have separate management, their market
image is different, and none have direct or indirect decision-making power over the
others. It is thus not a case of subsidiaries but of independent undertakings which
voluntarily enter into contracts with each other. 26 A two-fold series of positive
consequences derive from this hybrid structure:
Firstly, each finn preserves a more specialized identity and can thus apply and
develop the organizational patterns it requires for success in its market. These indi-
vidual patterns affect both the professional and environmental dimensions.
Professional specialization is essential when connected fInns provide different pro-
fessional services. For example, it enables auditors, consultants and lawyers to be
integrated into different fInns with each one thereby developing its own codes of
conduct and human resources policy. Applying these codes of conduct overcomes,
on the one hand, the problems which can arise because different professions are
subject to different obligations to their clients in different countries. Moreover,
adapting staffmg policies to the specific features of the activity is of prime impor-
tance in this sector in which human capital is very much the most basic and valuable
asset. 27 As a result, the conflicts of interest which the provision of multiple services
25 For example, the decision taken by Arthur Andersen in 1989 to divide itself into two business units,
separating off consultancy activities into Andersen Consulting, gave this new firm virtually full finan-
cial, strategic and operational autonomy (Nanda and Yoshino, 1996, p. 2). The distinction between the
two units was based on the fact that from the beginning they emerged as networks of separate firms (in
general, there were separate firms in each country, wholly owned locally), with their own partners,
management, employees, premises, business, capital and client portfolios; different corporate images
were established through their advertising, and different services offered; none of the firms has a sig-
nificant influence on the operational, financial or accounting policy of the others; and the economic
impact of any business between Andersen Consulting and a client of the Arthur Andersen audit firm on
the latter is minimal (Rankin, 1990).
26 This type of structure is not only used by large firms but also small firms. A good example is the
group formed in Spain by Audihispana and Bufete Cuatrecasas, organized by means of two independent
companies: Audihispana, which carries out audits, and Cuatrecasas, which provides legal services.
Both firms have totally separate shareholders, although they operate under a coordinated system. The
audit firm is a company with five partners, whilst Bufete Cuatrecasas, devoted to legal and tax work, is
a limited company with 25 partners. They are bound by a collaboration agreement under which they act
as a group and, when necessary, in close harmony. (According to the Chairman of Audihispana [Ex-
pansion, 23 September 1994]).
27 This aspect of the problem is illustrated by the conflicts which arose by the presence of consultants
and auditors within a single firm before and after the Andersen Consulting hive-off. The case is docu-
mented in the Business Management literature (see, for example, Robey [1991, pp. 152-3] as well as
"Civil War at Arthur Andersen" [The Economist, 17 August 1991, pp. 67-8]). Reconciling these two
lines of activity continued to be a source of conflict. It seems that Arthur Andersen's consulting activi-
ties increasingly competed with those of Andersen ConSUlting. The decision taken by the consulting
unit in December 1997 to split completely indicates that the benefits do not always exceed the cost of
can give rise to is eliminated or reduced. In other cases, specialization has more to
do with environmental adaptation to the type of market being served; e.g. the audit
arm of Arthur Andersen has an organizational structure with a geographical base in
which each national fInn, and even each branch, has considerable autonomy. On
the other hand, Andersen Consulting, the business unit created from it in 1989 and
which is now trying to split itself off fully, was organized by products and clients,
concentrating to a lesser extent on specifIc local factors, which is logical when
bearing in mind that it concentrates on serving large global clients. 28
In addition, fIrm autonomy introduces strong incentives for their partners by
linking liability with decision-making capacity. In particular, it enables those part-
ners who have greater knowledge as a result of their geographical and professional
specialization also to take decisions and bear most of the consequences of such
decisions. 29 The only correction required in this mechanism is the possibility that a
connected fIrm might harm the reputation of the whole and, therefore, the others.
To avoid this, exhaustive controls over the quality of inputs (particularly training
and promotion) along with the processes and services provided by each connected
fIrm, are introduced throughout the whole of the network.
c) The independence of the firms comprising each network. Within each net-
work, the various fIrms are independent in relation to both the other fIrms in each
country and their international organization. This latter aspect is clearly highlighted
when observing the diversity of reactions when two international networks merge.
In such cases, after representatives of the international networks reach agreement,
each national fIrm takes a vote amongst its partners on the merger. It is then com-
mon for a certain number of national affiliated fIrms to decide not to merge, in
which case they either maintain their independence, albeit as network associates, or
they join a different international network. The second option serves to demon-
strate the autonomy of national fIrms since, as a result of attempts at merging their
network, national fIrms generally receive offers from other international networks
the tie between the different branches making up this kind of network. There has been abundant com-
ment in the financial press of the conflict between these two firms leading eventually to their full break
up. See, for example, White and MacDonald (1997) and Kelly (I 997a-g). In a related problem, doubts
have been expressed as to the capacity of service firms connected with auditors to attract top quality
graduates (see, for example, "The Globalisation of Corporate Law: Red Tape Around the World," The
Economist, 23 November 1996, pp. 81-2).
28 See White and MacDonald (1997) and Kelly (I 997t). It is significant that Andersen Consulting
reorganized itself, giving even more priority to product markets, immediately after its decision to break
with Andersen WorldWide (Cardador and Cano, 1997).
29 As Fama and Jensen (I983a) observed, one of the two patterns which is systematically seen in the
division of responsibilities in very different types of organization consists of the fact that if the same
individuals manage and control decisions, they also bear the residual risk.
to join them. This was the case, for example, with the fIrms affIliated to Deloitte in
the United Kingdom, the Netherlands and Belgium, which decided to join Coopers
& Lybrand when their international network was merged to create Deloitte Ross
Tohmatsu, in 1989. 30 Opposition to merger is frequently based on disagreements
regarding the distribution of power in the new fIrm. For example, the additional
influence which the partners in Price Waterhouse would acquire in the fIrm as a
result of its merger with Coopers & Lybrand meant that the merger was much less
global than fIrst thought, when national fIrms in different countries, including in
Northern Europe and several South American countries, decided to remain outside
the merger. 3! Furthermore, a change in the network does not only occur as a result
of mergers but is relatively common as a result of all types of strategic disagree-
ment. Partly as a consequence, it is not uncommon, even amongst large fIrms, for
the same network to have two or more national member fIrms in the same country.32
30 See Ridyard and de Bolle (1992, p. 18).

31 See, for example, Cardador (I 998a and 1995b).
32 Ridyard and de Bolle refer to different past cases of both medium and large firms (1992, p. IS).
6 EFFECTS OF NON-AUDIT
SERVICES ON MARKET
COMPETITION
The previous Chapter showed that the provision of non-audit services by auditors
generates substantial cost savings. This Chapter will examine the consequences of
providing these non-audit services on competition in audit and non-audit service
markets. Sections 6.1 to 6.3 firstly look at how these cost savings are transferred to
clients as price reductions and, specifically, how this transfer is manifested in each
market (auditing and non-audit services) at each point in time (when contracting the
initial engagement and thereafter), depending on the competitive conditions in each
of these markets at each contractual stage. In this respect it will be shown that the
criticism leveled against the use of introductory pricing (or "lowballing") in audit-
ing, and against the use of auditing as a "loss leader" for selling other services, have
no foundation from the point of view of the public interest. When they occur, both
practices are benign and are a simple spontaneous consequence of the inter-tempo-
ral competition which is unleashed when, as in this sector, there are substantial
learning and switching costs. Similarly, Section 6.4 shows, firstly, that the provi-
sion of non-audit services by auditors enhances competition and quality in the
markets for such services. Contrary arguments, usually put forward by firms spe-
cializing in providing these services, who propose restrictions on the freedom of
auditors to offer services, appear rather to be the result of the private interest of such
producers in restricting competition.
6.1 The argument regarding auditing as a "loss-leader"
The provision by auditors of non-audit services has been quite frequently criticized
with the argument that they use auditing as an inducement to gain the client and
then sell him consultancy services, which are generally considered more profitable.
109
This argument is often expounded in the fmancial and professional press and some-
times appears in sector reports. 1
The argument usually contains two basic criticisms relating to the alleged conse-
quences of such a commercial policy for, fIrstly, the independence of the auditor
concerned and, secondly, for competition in the audit market. It is alleged, on the
one hand, that the auditor who provides non-audit services will pay more attention
to them than to auditing and will be more dependent on the client. It is also said
that this policy distorts competition, endangering the survival of professionals ex-
clusively involved in auditing.2
On analysis, both criticisms prove groundless. Chapter 5 has already dealt in
depth with the consequences for independence of providing non-audit services.
This Chapter will not therefore be concerned with this part of the argument/ except
to recall that the provision of non-audit services, far from being detrimental to inde-
pendence, can enhance it, particularly if the diversifIcation of fIrms is ensured.
Attention will be paid, on the other hand, to examining how the provision of non-
audit services not only does not harm but has a benefIcial effect on competition in
both audit and non-audit service markets. The rest of the Chapter is therefore di-
vided into the following three sections:
a) Firstly, Section 6.2 examines how the argument is based on factual and theo-
retical considerations which are at least dubious. The reason is that it is far from
I See various commentaries in relation to different countries and points in time in Benoit (1987 p. 19),
Berton (1989) and "Accountancy: All Change" (The Economist, 17 October 1992), in the media. As far
as sector reports are concerned, it is found, in its eclectic zeal, in the EC Green Paper (1996), Section
4.11.
2 It is stated, for example, that "auditors move into companies because the latter must have their ac-
counts monitored, but they then take advantage of their position to offer related services in such a way
that in many cases auditing is not the field generating most of the auditor's income and neither, of
course, the area generating most profit. For this reason [... ] professionals and firms are capable of
substantially reducing prices to gain a client, knowing they wiIl then be able to recover possible losses
many times over from the fees achieved from other services [.. ,], Consequently [.. ,], those auditors who
do not invoice related services have difficulty in surviving in a world where the price is used as an
inducement and not as adequate remuneration for the services to be provided" (Gonzalo Angulo, 1995,
p. 618, own translation).
3 One aspect of the debate which may have been side-lined in the general analysis in Chapter 5 should
be expressly mentioned, It has been said regarding quality that "the professional or firm will generally
concentrate more effort on those other activities which earn more profit to the detriment of the time and
attention properly due to auditing, which has simply been the gateway into the business" (Gonzalo
Angulo, 1995, p, 618, own translation), This type of comment overlooks the fact that it is, by defini-
tion, impossible in a joint production situation to identify the cost of each product and, therefore, their
profitability, Moreover, the price discount in one element making up a joint sale of services does not
compromise the quality of that element more than the others, Finally, even allowing the argument that
the audit is used as a gateway, the auditor who lowers the quality of his audits would have less opportu-
nity to use it as an inducement, and would therefore suffer a penalty which would restrict his
hypothetical opportunism,
Effects of non-audit services on market competition 111
clear, based on both empirical data and theory, that introductory pricing, or low-
balling, is in fact widely practiced.
b) Despite these doubts as to the extent of introductory pricing, Section 6.3 goes
on to accept its existence as a working hypothesis, but demonstrating why it is
likely that, despite its bad press, this pricing policy would still be efficient. The key
lies in the fact that, through ex ante competition before contracting in the fIrst year,
the discount on initial prices eliminates the monopolistic rents which can appear
when the auditor receives quasi-rents in subsequent years, the latter deriving from
the existence of learning and switching costs (it should be repeated here that the
amount of these two costs, which are generally present in the case of auditing,
would increase if, as seems likely, there are economies of scope between auditing
and non-audit services). This conclusion is reached after ruling out, in Section
6.3.2, that the use of introductory pricing could, in this sector, be the instrument of a
long term strategic objective by which fIrms would use it to eliminate competition
in order to raise prices after they have achieved a monopoly position.
c) Finally, Section 6.4 complements this analysis in terms of efficiency by pre-
senting various arguments regarding the distributive aspect of the problem. This
positive analysis aims at investigating which undertakings would be likely to benefIt
and which would be harmed if regulatory proposals of a prohibitionist nature were
applied.
6.2 Doubts as to the prevalence of introductory pricing
Despite dissemination of these arguments regarding the use of introductory pricing,

or lowballing, the extent of its use by fIrms is doubtful from both the empirical and
theoretical points of view. Empirically, there appear to be systematic differences
between countries, which could result from the diversity that exists in both market
conditions and regulatory environments. Secondly, on the theoretical side, it is
possible that the conditions prevailing in this market (mainly the intertemporal cost
structure, and the fact that fIrms have a reputation) favor implicit contracting at
prices which are valid for more than a single year, the existence of which would
neutralize arguments, like the one referred to, that they are based on a sequence of
entirely separate yearly prices. This possibility would, however, be compatible with
the occasional presence of introductory prices in some contracts which, despite
being of a rather anecdotal nature, may be used in the debate to try to prove the
widespread use of such prices.
Firstly, the available empirical evidence on introductory pricing is contradictory,
as mentioned in Section 4.2.3.3. On the one hand, such prices have been apparently
seen in different empirical studies in the United States but, however, indications
have also been found in Australia of initial prices which actually exceed cost. 4
Finally, in Spain it seems that a policy of constant prices is usually applied, perhaps
influenced by the regulatory imposition of a multi-year duration for initial contracts
between auditor and client and fIxing fees, or the principles for calculating them, in
writing for the whole of the period of appointment. 5
Secondly, in theoretical terms, the parties to the audit contract can, and perhaps
in fact do, even though in an implicit manner, contract prices for several years.6
Thus, the parties would expect a certain price for the services provided over the
period of their relationship, which would be adjusted annually to correct changes in
the workload involved and the audit market. This multi-year "contract" must be
implicit, being thus subject to solely reputational safeguards. The reason is that the
formal contract itself needs to be of an annual duration to facilitate the automatic
economic sanction, by means of cancellation, by the market against those auditors
who compromise their reputation for independence, a matter already referred to in
Section 1.3. As a result, it makes less sense to think in terms of annual prices and, if
it is done, it is not possible to establish from the outset whether an American-style
introductory pricing policy will be used, one of decreasing prices as seen in Austra-
lia, or constant prices, such as those which seem to be applied in Spain. Moreover,
it is foreseeable that different inter-temporal pricing patterns will be found depend-
ing on the contractual and competitive conditions in each country.
Finally, the fact that the allegations that auditors transfer the costs of one activity
to another have been formulated in both directions also contradicts the widespread
use of a particular introductory pricing pattern. Firstly, the fInancial press and vari-
ous commentators fear that auditing is used as a "loss-leader" to move into
consultancy work. The results of various empirical research carried out in the
United States are, on the contrary, compatible with the possibility that it is the con-
sultancy work which plays this role. Simunic (1984), in particular, and various later
studies, have shown that audit fees increase when the client also acquires other
4 For the United States, see Simon and Francis (1988) and TU11'en (1990) and, for Australia, Francis
(1984). After analyzing the US empirical evidence, the authors of a review of the literature concluded
that the available evidence did not support the notion that non-audit services enable large firms to
charge an additional premium on their audit fees, contradicting the results obtained in the first empirical
study into the problem by Simunic (1984). See Yardley, Kauffman, Cairney and Albrecht (1992, p.
172).
S See Section 2.2 of the Technical Audit Standards (ICAC, 1991). It can also be seen that standard
form contracts of the main professional organizations usually establish a constant real price without
including any changes to the nominal price other than as a result of inflation or substantial changes in
the workload, using terms for the pU11'ose such as those contained in the model contracts prepared by
the Instituto de Censores Jurados de Cuentas de Espana (ICJCE, undated) and the Registro de Econo-
mistas Auditores (REA, 1993, p. 639).
6 For a more complete explanation of this argument, see Arrunada and Paz-Ares (1997, pp. 50-1, n.
58).
services from his auditor, although such studies merit the reservations already men-
tioned in Section 4.2.3.3.
6.3 The efficiency of introductory pricing
Assuming that introductory pricing, or lowballing, is used to some extent in audit-

ing, two main explanations are relevant to understand such use: either such prices
result from competition between ftrms to win business which will earn them profits
in the future, or that those who do so are trying to become a monopoly in the hope
of offsetting their current losses by raising prices when they have achieved one.
Both possibilities will be examined in detail in the next two sections. Neverthe-
less, it should first be mentioned that observation of substantial differences in initial
audit prices could be the result of at least two other causes apart from the discount
initially induced by the possibility of receiving quasi-rents from audits and services
in successive years and predatory practices aimed at eliminating competitors in the
audit market. These other possible causes are the greater efftciency (a) or lower
quality (b) of the auditor who offers a lower price. For the purpose of this study, it
sufftces briefly to clarify the potentially positive nature of both.
a) With respect to the differences in productive efftciency, this is self-explana-
tory. It should only be pointed out that when efftcient competitors are castigated or
obstructed as a result of confusion (often promoted by those who suffer the conse-
quences of competition), the driving force behind economic development ends up
being destroyed. This could well be the case with the provision of non-audit serv-
ices since, as there are economies of scope, the initial price discount reflects the cost
saving, and thus greater efftciency, achieved by providing both sets of services.
b) Quality differences are also positive provided that they vary at above the level
which would generate undesirable external effects for other auditors or for the pub-
lic interest, since there is differential demand for auditing in terms of quality, in line
with the contractual costs of clients, as explained in Section l.l. For this reason,
price regulation should not be used to combat the provision of quality below the
level which is considered an acceptable minimum, for whatever reason. Such
regulation runs the serious risk of turning into a restraint on competition, without,
moreover, being effective in achieving its objective. Quality control should thus be
based on action of a different type, whether by laying down entry and practice con-
ditions or, better, sanctioning poor professional practice.
6.3.1 Introductory pricing as a consequence of the competition

for quasi-rents
6.3.1.1 Quasi-rents and introductory prices in auditing
Initial competition for profits received ex post in future engagements is the most
credible explanation for the use of introductory pricing in the context in which audit
firms operate. These future profits are technically known as "quasi-rents" and are
generated as the result of the start-up or learning costs of initial audits and the
switching costs which all clients must incur when they change auditor. In addition,
if there are economies of scope between audit and non-audit services, these econo-
mies constitute an additional source of quasi-rents, as shown in Section 5.1.1. All
these factors mean that continuity in their relationship is advantageous for both
client and auditor. The overall benefit can be shared out in several ways, as ex-
plained in Sections 3.3.1.2 and 5.1.3 .a, but it is fairly likely that a large part is
received by the auditor. If the provision of auditing is competitive, it can be ex-
pected that auditors compete downwards in their prices for initial engagements,
knowing that they will be able to make up the likely initial loss from the profits
from future work. If initial engagements are generally more often for auditing and
there are economies of scope, a larger discount on the initial audit prices will tend to
be seen when the auditor is equipped to provide such services.
The fully competitive and optimal nature (from the public point of view) of this
inter-temporal competition will be studied next. 7 A better understanding of this
phenomenon will avoid the still common error of considering the possible practice
of introductory prices at below the cost of the initial audit (lowballing) as anti-com-
petitive or harmful to independence, an error which sometimes leads to introducing
or proposing inappropriate policies with the intention of preventing it.
6.3.1.2 The effect of non-audit services on quasi-rents and introductory

prices
Discounting the introductory price, or lowballing, is the natural consequence of

competition in a context of learning costs or, more generally, when the supplier
enjoys quasi-rents ex post, the current value of which is offset in present value
terms by aggregating them with the entry discount. This discount is simply a
symptom that initial competition is very strong and is therefore an aspect which
7 Inter-temporal competition was originally studied in the literature on franchise bidding. See, in this
respect, the works of Demsetz (1968), Stigler (1968) and Posner (1972), as well as the critique of Wil-
liamson (1976), related to situations in which a repeated auction is necessary and there are substantial
physical assets.
should be positively assessed as an indication of the competition in the sector.

Moreover, this assessment is positive, independently of the judgement merited by
the quasi-rents, since the latter are the cause rather than the result of introductory
pricing. (The reader who has recently acquired a mobile telephone has very likely
benefited from this type of competition: companies virtually give away the tele-
phones knowing that the client is bound to them, generating substantial profits in
the future. It is this tie, together with a certain degree of competition, which is the
cause of the initial discount, but the tie would exist even if the initial hook-up price
had not been discounted, the latter being a mere consequence).
Under these conditions, the fact that an auditor provides non-audit services to his
audit clients, assuming as seems credible that there are economies of scope between
auditing and such services and that firms are in competition, affects the pricing
policy in the way examined in Section 5.1.1.1. Briefly, it was then shown how the
existing economies, or synergy, between auditing and other services has the effect
of raising the cost of changing auditor and, therefore, the value of the individual
quasi-rents. Initial competition then produces a greater degree of discounting, or
lowballing, in the initial period.
The effect of providing non-audit services on the price of the initial audit can be
estimated using the model developed in that Section. Under the conditions defined,
it turns out by comparing expression [16] with [26], and [17] with [27], that both
the price of the initial audit and that of recurrent audits is less when non-audit serv-
ices are provided:
(l - ,,) [j3 c (l-m)]
Po = Po - 1+ r [37]
p' = P - c(l-m) {i -fi(" + r) } [38]

l+r
where, it should be recalled, Po reflects the prices of the initial audits, p the prices of
recurrent audits and the superscript s identifies the respective prices of initial and
recurrent audits when non-audit services are provided, the only consequence of
which is to reduce the cost of recurrent audits by a proportion (1 - m). The conclu-
sion is that under these conditions both prices will be less when non-audit services
are provided. 8
By way of illustration, this price reduction can be simulated taking realistic con-
ditions for the relationships between the different costs, the likelihood of a break in
8 It is clear that the difference between the initial audit prices with and without services is negative in
expression [37] and the same applies to the difference between the prices of recurrent audits in expres-
sion[38], sincefJ(IJ-r) < J.
the relationship and the discount rate (assuming, specifically, C = 150, c = 100, 1] =
0.04 and r = 0.05), and introducing an additional assumption of equality and con-
stancy in the division of quasi-rents (a = fJ = 0.50). As shown in Figure 6.1, the
degree of discount increases with the economies of scope, and is nil when the latter
do not exist (in terms of the model, when the cost of the audit is not affected, and
therefore m = 1).
140 c
120 Po
100 F=========~~~~==========~~
80
60
40
20
m
0.2 0.4 0.6 0.8 1.0
Figure 6.1. Prices of initial and recurrent audits depending on the extent of the
economies of scope
Note: Simulation taking C = 150, c = 100, 1] = 0.04, r = 0.05 and a = jJ = 0.50.
Considering the parameters in this simulation, the price discount in relation to

the cost of the initial audit would in principle equal 15.24 percent (= [C - pJ / C =
[150-127.14]/150). On the other hand, if non-audit services are provided and it is
assumed that the economies of scope provide a value equivalent to a reduction of
20% (m = 0.80) in the recurrent audit costs, the initial price would be 118 and the
discount in relation to cost 21.33% (= [C -p~] /C= [150- 118] /150). The additional
increase from lowballing is therefore small, moving from 15.2% to 21.33%, and for
the most part (71 %) is the result of learning and rotation costs of the auditing itself,
not the provision of non-audit services.
In reality, it is also inappropriate to attribute so much importance to initial pric-
ing. It can be expected that any competitive capacity of firms based on whether
they provide non-audit services or not will depend rather on the current value (at the
time of contracting) of the total costs which clients expect to incur. Unless they are
irrational, they will choose between the two types of auditor on the basis of this
updated cost, defmed not only by the price of the initial audit but, in particular, by
the expected prices of future recurrent audits. Under the simulation conditions, the
price of the recurrent audit would fall drastically, from 102.l4 to 83 monetary units,
thanks to the cost saving from the joint provision of audit and non-audit services. 9
It would thus be this direct result of the greater efficiency inherent in economies of
scope which would determine the relative viability of the two types of firm. The
difference in the initial price is in reality relatively small, although its prominence in
the debate is understandable as it is most visible to competitors who are less effi-
cient in providing both types of service.
6.3.1.3 The fallacy of differential profitability of joint services
It is quite often stated that non-audit services are supposedly more profitable than
audit services. lo The preceding analysis illustrates the inappropriateness of such
statements. Firstly, the existence of economies of scope between audit and non-
audit services makes it impossible to identify the cost of each type of service. II
Secondly, the fact that the audit is generally the fust service provided l2 means that
the discount is seen in auditing and that those providing solely audit services are
prejudiced. To say that the non-audit services are more profitable than auditing
would be an even more serious mistake than stating that in the case of lowballing
the recurrent audits are more profitable than the initial ones, overlooking that in a
multi-period investment project annual profitability is of little fmancial significance.
In the case of providing non-audit services, the statement is even less meaningful,
since to this inter-temporal dimension is added the fact that the economies of scope
make it non-viable to estimate the cost, and therefore individual profitability, of
each service.
9 In the terms of the model, these effects appear in the audit price since this represents payment for all
services. In reality, such effects should not necessarily be seen in the audit prices but also in those of
the non-audit services, although always translated into a drop in price.
10 See, for example, Berton (1989) and Gonzalo Angulo (1995, p. 618).
11 Gigler and Penno (1995, p. 329) demonstrate how an independent price fixing policy can generate
inefficiencies.
12 In a sample of US companies it was estimated that non-audit fees are much less in the first year of
audit than in subsequent years. In the first year, fees from non-audit services represented a weighted
average of 12.35% of auditing turnover; in subsequent years this percentage increased to 64.13% (Tur-
pen, 1990, p. 66).
6.3.1.4 Undesirable effects of a possible prohibition on introductory

prices
Analysis of the problem also illustrates that the consequence of prohibiting intro-
ductory pricing (in the very unlikely event that such prohibition could be effectively
enforced!3) works against the objectives normally intended by such a prohibition in
terms of both independence and competition:
a) Firstly, such a prohibition would raise the total volume of quasi-rents associ-
ated with a client. The reason is that the prohibition would increase the cost of
replacing the auditor (the aspiring auditor would no longer be able to offer a dis-
count on the initial audit) and, therefore, the quasi-rents received by the current
auditor.!4 Curiously, such a rule would thus have the consequence of reducing the
auditor's independence in relation to the client when such independence depends on
the volume of these individual quasi-rents (which could happen in cases of little
client diversification, as examined in Section 3.3.1).
b) Furthermore, a prohibition of this type would be harmful to competitive con-
ditions: as it will be impossible for the potential auditor (or consultant) to discount
the initial price, the rotation cost would increase and the current auditor could
charge a higher price, thus earning an economic rent. After prohibiting the discount
on the initial audit, the price would equal the cost of this initial audit, whilst the
prices in successive years would be fixed at the level which prevents entry by a
competitor, a level which would be increased as a result of the higher rotation costs.
In short, not only would it increase the initial price but also successive prices.
(Imagine a rule or oligopolistic agreement which prevented mobile telephone op-
erators from offering connection discounts: as a change of operator would be more
costly, it is likely that each of them could charge higher prices to their customers,
who would thus pay higher prices for both the hook-up and subsequently!5).
These consequences may contradict the intuition of those who naively believe
that prohibiting the initial discount will improve independence.!6 The reason is that
13 It is doubtful whether prohibitions could be enforced because, even if it could not be avoided by
billing innovations, firms would resort to competition in areas other than price, stepping up service
quality (Stigler, 1968).
14 See, on this point, DeAngelo (1981 a, pp. 124-5) and Grout el al. (1994, pp. 329 and 343).
15 In fact, this outcome could be seen during the course of 1997 in the mobile telephone market in
Spain: the year began with nil hook-up prices for low-range telephones, which rose to 10,000 pesetas in
February and 15,000 in July. The increases were simultaneous, although the two companies concerned
denied any type of agreement between them. See "Airtel y Telef6nica entierran la guerra de los telefo-
nos m6viles al subir los precios un 50%" (Expansion, I July 1997, p. 3) and "Defensa de la
Competencia investiga un posible pacto entre Airtel y Telef6nica" (Expansion, 2 July 1997, p. 4).
16 This, for example, was the apparent intention of a Texan Act of 1991 which prohibited lowballing
with the argument that " ... who performs or offers to perform a service involving auditing skills for
such an assumption is based on the unfounded belief that without an initial discount
there will be no quasi-rents in future years. Belief, in other words, that forcing up
the price in the first year to at least the cost for that year will also force a reduction
in the prices of future years to the level of cost. What would tend to occur, on the
contrary, is that the current auditor would have less competition since his competi-
tors cannot make an introductory discounted offer but, at most, their best price
would be equal to the cost of the first year. Consequently, the current auditor can
establish a higher price in recurrent audits. As his costs do not change as a result of
the rule, the individual quasi-rents, i.e. the difference between price and cost, will
be greater. It is for these reasons that a large part of the literature considers that
professional concern regarding lowballing could have more to do with preserving
and maintaining monopolistic rents than with the alleged objective of preserving
independence.!7
Considering that this analysis in terms of inter-temporal competition was already
applied to auditing many years ago (DeAngelo, 1981a), it is discouraging to fmd
that the European Commission prefers to stimulate debate on the problem in Europe
on erroneous bases without responding to scientific knowledge in the field, and it
thus can be read in its Green Paper that:
The growing intensity of competition for audit 'business', and especially for the audit
of large 'prestige' companies, is also a cause of concern. There is no doubt that com-
petition sometimes results in low-cost and perhaps even below-cost tenders. The
procedure of calls for tenders which ensures transparency and competition, should
not have as a consequence that auditors quote an audit fee which does not allow them
to carry out their work in accordance with professional standards. Some observers in-
fer that the successful tenderers expect to recoup the balance of the full cost of the
audit from non-audit consultancy services (Green Paper, section 4.11).
Rules in force in some European countries (Austria, Belgium, Greece, Italy and
Portugal) merit criticism for the same reason, as they are aimed at preventing intro-
ductory pricing or discounting. The rules range from prohibiting discounts on
tariffs varying with client size to the supervision of fees and audit hours by profes-
sional bodies and administrative price fixing (Table 6.1). Outside Europe, there are
also restrictions on the level of fees in Australia where, although there are no regu-
latory restrictions, the ethical standards of professional auditors' associations
prohibit lowballing to obtain work. Also, in Japan fees are determined by a tariff
compensation that is less than the direct labor cost reasonably expected to be incurred in performing the
service creates a presumption of loss of independence" (cited by Grout e/ uf., p. 325).
17 This was already the opinion of DeAngelo (1981 a, p. 125). Dewatripont (1994, p. 344) has recently
expressed a similar opinion, from outside the accounting field.

laid down by the Japanese Institute of Certified Public Accountants (nCPA) and
approved by the Japan Federation of Economic Organizations (Keidamen).
Table 6.1. Rules in force in different countries aimed at preventing introductory

pricing or discounting on initial audits ("Iowballing")
Rules for Price must Restric-
Country calculat- be fixed tions on Explanation
ingfees previously lowballing
Australia No No Yes Lowballing is prohibited by the ethical standards of the

profession, but this prohibition is not enforced in practice
Prohibition on systematically undercutting the audit fees
Austria Yes No Yes indicated in the guidelines issued by the regulatory body,
which are based on the size of the client and hours of work.
Belgium No No Yes Time and fees must be notified to the professional body
which reviews their adequacy.
Prohibition on negotiating fees. The price per hour is
Greece Yes Yes Yes fixed by a Supervisory Board which also lays down
directives regarding the number of hours (based on size)
and monitors compliance.
A mandatory tariff must be applied, except by "regulated
Italy Yes No Yes audit firms" or Societa di Certijicazione, for whom only
the number of hours but not price per hour is regulated.
Japan Yes Yes Yes Fees are determined by a tariff laid down by the JIePA
and approved by the Keidanren.
Portugal Yes No Yes Directives on tariffs include minimum prices based on
client size with an absolute minimum of 150,000 escudos.
Sources: For European countries, adaptation of Buijink, Maijoor, Meuwissen and Witteloostuijn (1996,
pp.76-7); for Australia and Japan, specific survey (see details in Sources to Table 3.1).
6.3.2 The implausibility of predatory behavior through cross-

subsidization between different kinds of service
The argument in the previous section can explain the existence of introductory
pricing in auditing by the convergence of two circumstances: firstly, the presence of
learning and rotation costs and, secondly, competition between supplier firms. The
analysis also shows that the provision of non-audit services intensifies discounting
of the initial audit when there are economies of scope between audit and non-audit
services. The existence of these economies is clear, particularly in the light of the
qualitative indications mentioned in Section 4.2.3.3. Nevertheless, since the pres-
ence of these economies of scope has not been corroborated in quantitative studies
on a completely reliable basis, attention should be paid to another possibility, even
if only to demonstrate its unlikely nature: that the low prices have only a predatory
purpose. According to this argument, the intention of auditors practicing this policy
would be to subsidize their audit services for a period of time in order to raise the
prices after the competition has disappeared.
In the debate on the problem, arguments of this type are in fact sometimes de-
fended. ls Auditors who provide non-audit services are then accused of distorting
their pricing policy in the audit market. The concern expressed in this respect is
generally that the profits obtained from these non-audit services are used to subsi-
dize auditing activities, which could harm competition in the audit market.
Following this argument, it seems that these firms are basically being accused of
implementing predatory pricing in auditing and subsidizing it out of future income
from other services. The argument is difficult to sustain, however, when analyzing
the specific circumstances of the case, from both the point of view of traditional and
more modem industrial economics: 19
a) The idea in principle that a firm established in a market can discourage the
entry or, its equivalent, encourage the exit of competitors, by lowering the sales
price was put forward many years ago by Bain (1949). The viability of these
predatory practices, however, which never enjoyed much support as, since Telser
(1966), it was doubted whether predatory pricing constituted an effective strategy
for achieving a monopolizing objective, has in recent decades lost the little credit
which it had. 20 In these industries, monopolizing intention would also make little
sense. Firstly, all firms are capable of successively implementing this strategy.
Moreover, both markets, auditing and non-audit services, are highly competitive in
the view of those participating in them. Finally, in all market segments there are
potential entrants who do not suffer from appreciable entry barriers.
b) At the present time, accusations of predatory pricing against a business with
market power, or even a monopoly, tend to be discarded, unless they are based on
18 This usually happens in a somewhat confused manner since, on occasions, they are mixed with other
accusations of anti-competitive conduct which merit even less attention. This is the case with accusa-
tions of unfair competition, according to which the provision of non-audit services would give rise to
expropriatory activities. This type of criticism, however, merits little credit if, as is usually the case, it
is not substantiated by making a claim or beginning legal or administrative proceedings. Moreover, the
argument based on the greater financial muscle of the large firms also has little substance due to the
nature of the capital, basically human, and this added to the scant general consistency of this type of
argument.
19 Initial works on non-audit services already examined the lack of credibility of the predatory pricing
argument in this context (Simunic, 1984, p. 698, n. 13).

20 It is said, for example, that "although Bain's point of view prevailed for 40 years, many economists
felt uncomfortable when applying it to the analysis of competition policy. To condemn a business for
making consumers pay too little seems paradoxical. Even more important, it was not clear how a low
price could discourage entry" (Tirole, 1988, p. 368). See also Demsetz (1997, pp. 208-10).
infonnational asymmetries regarding cost,21 a situation which does not seem to

correspond with this activity, characterized by the presence of multiple suppliers
with relatively accurate knowledge of the production costs. For this reason, the
argument would not be sustained even if a hypothetically highly unfavorable situa-
tion is considered, in which a finn is in a dominant position in the market.
6.4 Elements for a positive analysis of the provision of

non-audit services and the prohibitionist proposals
Up to now the problem has been analyzed from the point of view of the collective
interests in play. The objective has been to identify the pros and cons of whether
auditors mayor may not provide services which are similar or complementary to
their main activity in an abstract manner. To this end, the productive consequences
of the different regulatory possibilities have been examined. To complement this, it
is now appropriate, although in an inevitably summary manner, to indicate some of
the distributive consequences which will enable us to glimpse some of the private
interests which may be affected.
The analysis will be tackled by asking who could be hanned by audit firms pro-
viding non-audit services. In this respect it is clear that the interests of two types of
finn would be hanned: firstly, those audit finns which are inefficient, both gener-
ally and specifically in providing such services and, secondly, many non-audit
service providers. Both types of firm will suffer as a result of having to face com-
petition which has achieved a comparative advantage-the cost savings generated
by joint production of both types of service and the organizational innovations nec-
essary to make this viable.
6.4. 1 Consequences for firms specialized in auditing
When assessing the possible impact of providing non-audit services on different

types of audit finn, it is not easy to advance beyond excessively general apprecia-
tions, such as that just mentioned. To do this, indirect indications must be relied on,
21The problem was reformulated by considering informational asymmetry at the beginning of the
1980s. The basic works are separate articles by Milgrom and Roberts in which they deal with practices
aimed at preventing entry (I 982a) and inducing exit (I 982b) amongst competitors. According to these
models, predatory pricing policy can make sense for the purpose of signaling to possible entrants or
existing competitors that, if they persist in their policy, they will suffer losses and that therefore they
should not enter or leave the market, respectively. For this reason, modem competition policy usually
takes into account that a price below marginal cost in the short term does not constitute an indication
that the price is predatory, since prices below cost can be perfectly competitive, but tends to require
indications that the monopolist has the intention of eliminating the current or potential competition, as
indicated by Tirole (1988, p. 373).
which only enable some types of fIrm to be identifIed which may be more or less
affected, or others which could be benefIted as a result of their greater capacity for
evading a possible prohibitive rule.
a) Effects of the supply of non-audit services on different kinds of firm. A dis-
tinction should initially be made when examining the effects on providers of
exclusively audit services. If auditors are harmed, it would not be so much those
who carry out auditing exclusively, but those who provide auditing services but not
services of the type most suitable for achieving economies of scope. For example,
those who teach accountancy and combine professional auditing practice with
teaching do not devote themselves exclusively to auditing. Their capacity to com-
pete in audit work, however, could be harmed by a widespread practice involving
other auditor-consultants using auditing as a "loss-leader" and therefore discounting
the prices of their audits. From this point of view, any prohibition on providing
non-audit services would perhaps not be so much in the interests of accountants
(who to a large extent presumably provide such services to their audit clients), as
those accountants who, even though they do not generally devote themselves exclu-
sively to auditing, provide other services to clients other than auditing clients.
b) Differential ability to evade the rule. It has repeatedly been pointed out that a
prohibition on providing non-audit services would lead to evasion of the law by
some professionals splitting their fIrms, formally hiving off auditing but maintain-
ing a unifIed control in fact. 22 Even those who maintain that this form of evading
the law would be easy to prevent by controlling professional fIrms (e.g. by prohib-
iting auditors or auditing fIrm partners from holding interests in businesses which
advise their audit clients), acknowledge that not even in this way would all prob-
lems be resolved, but only the means provided to prevent the breach from being
widespread. 23
This possibility of evading the law means that the effect of a prohibition is dis-
tributed in a particularly undesirable manner, by harming those fIrms of better
reputation. To gain an idea of this phenomenon it suffIces to consider which types
of professionals would be in the best position to evade the prohibition. It seems
logical to assume that formulas involving legal evasion will be less costly to those
professionals who, as a result of their smaller reputational capital, would have less
to lose if the regulatory authorities or those who may be affected try and prove that
they are getting round the law. From this point of view, application of a prohibitive
22 The authors of the NERA report stated their conviction in this respect in regard to the rules in force in
France and Portugal and did not rule it out in the case of Italy (Ridyard and de Bolle, 1992, p. 67). The
EC Green Paper is of the same view (Section 4.13).
23 See, for example, Gonzalo Angulo (1995, p. 6\8).
rule would probably in practice penalize fInns which provide greater quality more
and, therefore, indirectly benefIt those who provide lower quality.
6.4.2 Consequences for providers of other services
Examination of the positioning of different types of fInn suggests that those mainly
harmed by the entry of auditors into the provision of non-audit services would be
those providing such services. From this point of view two explanatory hypotheses
can be formulated: According to the fIrst hypothesis, which will here be called
"service quality erosion", those specializing in the provision of such services basi-
cally complain that their joint provision, or provision in some way linked to
auditing, lessens the quality of such non-audit services and, in particular, infringes
professional requirements. The second hypothesis, which will be classifIed as "de-
fense of monopoly in service provision" raises the possibility that these proposals in
reality constitute a masked defense of the dominant position achieved by such pro-
viders in the marketplace. In the following discussion it will be argued, and some
evidence provided, that the fIrst hypothesis is groundless. The most plausible ex-
planation of the opposing reactions is therefore that they result from a defense of
monopolistic interests.
6.4.2.1 The hypothesis of service quality erosion
The central conflict regarding collaboration between auditors and other profession-
als has focused in Europe on collaboration with lawyers. It is argued that legislation
imposes, or that clients demand (or it is said that they demand), different duties
from the two professions, maintaining that whilst auditors have a duty to be inde-
pendent of their clients, lawyers, on the other hand, have an opposite duty. The legal
dimension of the problem will not be dealt with here, however. It should only be
mentioned that there is considerable disagreement amongst lawyers themselves in
terms of whether this argument has a legal foundation and how to regulate the sec-
tor. For the purposes of this book, the legal answer is of little interest since, as well
as depending on the legislation in each country, professional associations aim to
change the law in order effectively to restrict real possibilities of collaboration be-
tween auditors and lawyers. 24
The question must thus be raised in the fIeld of legislative policy. The basic
question is knowing whether the law should restrict such collaboration, or not. A
24 This is particularly the case with the report issued by the European Council of Bar Associations on
the European Commission Green Paper, in which it is proposed that the auditor's role is made equiva-
lent to that of the professions which enjoy authority conferred by public authority, such as judges or
civil servants, demanding that auditor independence be determined by a strict system of incompatibili-
ties (HEI etemo dilema del limite de actividad," Expansion, 7 May 1997 p. 48).
negative answer will be defended here, basically because it is a matter which the
parties, particularly clients, seem perfectly capable of deciding for themselves. Any
restriction would thus be at least superfluous and, more seriously, run the risk of
being counter-productive. In short, the matter should be left to the judgement of the
client himself. Clients can assess the advantages to them of joint service provision:
a lower number of providers, lower costs, better quality and, above all, a better
guarantee of quality. Clients are also able to appreciate the risks which may be
involved in this combining, if in fact they really exist. 25 Even those who tend to
mistrust consumer rationality will appreciate that the client for this type of service is
a business client, not a poorly informed consumer. This strengthens the presump-
tion that he is capable of forming and, as appropriate, modifying a sensible
judgement on the matter. He will become, in short, a well-informed purchaser
aware of the conflicts which may be involved in the links between carrying out the
two activities. 26
A greater degree of collaboration between professions is most probably dictated
by demand. It basically responds to two main forces. Firstly, changes in business
circumstances, where problems are of growing complexity and scale, necessitating a
global and coordinated provision of different specialized services which require a
similarly increasing degree of expertise. Secondly, the increasing demand for serv-
ices of guaranteed quality on a global scale, the ever-costly safeguard of which
means that it is in everyone's interest to optimize the use of reputational capital.
This is achieved by providing different services under the same quality assurance
structures which, in their most visible manifestation, take the form of a name with a
good reputation.
As a result of this second factor-the demand for quality safeguard-those pro-
fessional activities which have previously been carried out or are still being carried
out under a system of restrictions which has prevented the development of effective
quality safeguard formulas, are perhaps most likely to be linked with auditing. This
is usually the consequence when the traditional systems of professional corporatism
deteriorate. In such cases, the association restrictions cease to protect quality. They
25 In this matter, experience in the advertising field is illustrative since, being a more dynamic and Iittle-
regulated sector, the main firms adopted one-stop shopping strategies much earlier. The response of
clients varied, with some clients seen to value the advantages of geographical and service consolidation
more, but others, on the other hand, continuing to give priority to the benefits provided by diverse
providers. See, for example, "Advertising: a Passion for Variety" (The Economist, 30 November 1996,
pp. 82-3).
26 For this reason, audit firms themselves and their professional associations have the greatest interest in
structuring the linkage between services in such a way that both conflicts of interest and the appearance
of conflict are minimized, as described in Section 5.3. It should be mentioned here that such structures
for preventing conflicts include forming separate firms without common partners, and prohibiting or
abstaining from providing services which could be or appear more conflictive.
generally still have sufficient force, however, to prevent the appearance of the types
of safeguard inherent in the provision of such services under a system of free mar-
ket competition, for example the creation and development of firms organized in
such a way that they can guarantee quality as a result of their size and structure.
Specifically, this could be the case with legal services, subject in many countries to
a series of professional association restrictions which, because of their partial nature
(particularly in those countries when in practice there are no effective entry barri-
ers), do little to safeguard quality. If this consideration is well-founded, linking
legal and audit services would provide an "external effect" in this field (a social
benefit which its protagonists to not appropriate): the stimulus for the legal profes-
sion to change and satisfy a demand for legal services of guaranteed quality, a
demand which is nowadays to a large extent unsatisfied.
6.4.2.2 The defense of monopoly hypothesis
The first works concerned with this matter in the United States were already warn-
ing of the role of competitors in demanding regulation to restrict the provision of
services. Some textual references to this discussion should be made as a result of
their similarity to the statements which have proliferated in Europe more recently.27
For example, the President of a large consultancy firm and of the Association of
Consulting Management Engineers testified before the US Congress that the provi-
sion by auditors of management consultancy services represented "a potential
compromise of the audit, which requires independence". Along the same lines, the
Chairman of a staff recruitment consultancy saw particular danger (to the independ-
ence of the auditor) in the latter providing recruitment services and the fact that
there was a transfer of executives.28 It is not surprising, therefore, that with this
debate now starting up in Europe, it is the lawyers who are also becoming the
champions of auditor independence.
It is obviously very difficult to verify the hypothesis that provision of non-audit
services is opposed in order to protect monopoly rents, settling therefore instead for
presenting some clarifying signs. The most simple evidence emerges when ob-
serving the origin of the opposition to freedom of action in this field. This
opposition does not come, as might be expected in a situation of alleged abuse, from
new entrants into the market whose attempts could be prejudiced but, on the con-
trary, from comfortably installed professionals and firms. From this point of view,
it may be that if the entry of audit firms in fact constitutes a threat, it is not so much
27The following references are taken from Benston (1979-80, pp. 26-7).
28The transfer of executives has a certain importance for the efficient management of human resources
amongst large audit firms, who make use of a professional career structure of the "up or out" type (for
an introduction to the functioning of such up or out systems, see Milgrom and Roberts [1992, pp. 379-
82]).
to competition but to continuity in enjoying a certain degree of market power on the

part of such firms. This argument is supported when verifying that protests against
the presence of audit firms are closely correlated with the existence of competitive
barriers, since they are in fact concentrated in the legal field and do not exist in the
tax advisory or management consultancy fields. Neither can it be argued that the
size of the large audit firms could give rise to problems in terms of a potentially
dangerous concentration of the respective service markets. The reason is that the
degree of concentration in these service markets is very low, even when taking a
very narrow definition of the relevant market. 29
6.4.3 Towards a new equilibrium

The demand for professional services nowadays requires that they have certain
characteristics, such as globality and, in particular, assured quality. There is room
to suspect that many of the traditional professional structures which still govern the
supply of some of these services have not been capable of satisfying this demand.
As a result, audit firms and, in the case of legal services, Anglo-Saxon law firms-
which have enjoyed more freedom of internal organization-have been able to
compete in the new environment at an advantage. Nevertheless, in the medium
term it is possible that specialized professional practice (by both service and coun-
try) will develop new organization formulas which will enable lost ground to be
recovered. This is encouraged by the fact that in almost all service sectors, firms
co-exist of very different sizes and degrees of specialization. This generally occurs
to a greater extent the less the restrictions on such services. For this reason, the
social objective in this field must to a certain degree be epistemological or cogni-
tive: it is desirable that the market should function freely to discover the optimum
firm typology which is best adapted to the new context of an increasingly global
and changing economy.
29An illustration of the minimal consequences in terms of concentration of service provision by audi-
tors is provided by the much discussed merger of the law firm of J. & A. Garrigues and Andersen ALT,
which took place in Spain at the beginning of 1997. The merger had virtually no consequences on
concentration in the Spanish market since the Herfindahl index moved from 668 to 917, a level at which
the market is considered deconcentrated, according, for example, to the criterion of the US Department
of Justice's Merger Guidelines (1992, p. 84).
6.4.3.2 The competition between multidisciplinary and specialized

practice
In this respect, what is today perceived by the professions most affected solely as a
danger will surely constitute the stimulus they need to renew their structures. The
competitive advantage of audit and some large Anglo-Saxon legal firms could in
fact be that they are capable of providing a type of service of a much more uniform
quality in a much broader geographical ambit than that covered by professional
organizations and their members. 30 The restrictions affecting the latter are to a large
extent the result of the restrictions which have for a long time governed their activi-
ties, preventing them from developing, at least until recently/i the characteristics
required today, thereby compromising quality of service and their opportunities for
survival. 32
It should not therefore be completely ruled out that, after a period of adaptation,
the specialized professions recover some of the positions which they are currently
losing, although adopting organizational formulas adapted to the new requirements
of demand. Moreover, this seems to have already been happening for some years.33
In some countries a certain recovery of position has in fact already been seen by
specialized providers. This is the case with tax advice in Great Britain, which was
traditionally carried out in Great Britain by accountants or auditors. In the 1990s,
however, there was a considerable increase of involvement by other professionals in
this activity, particularly lawyers, but specialized firms have also appeared on the
initiative of other professionals (Jack, 1993). In Great Britain there has also been a
notable reduction in the overlap between audit and non-audit service clients since
many companies now put both services out to tender. 34 The apparent about-tum of
the firm which was most prominent in the past in pursuing a "one-stop shopping"
strategy can also be interpreted along the same lines. A little while ago, KPMG
tried to construct a firm capable of providing the whole range of professional serv-
30 The increasing competition from law firms forming part of multinational networks could be already
as significant as that from law firms linked with audit firms.

31 See, in general, OECD (\ 985). For example, up to a few years ago, Spanish law firms could not have
more than 20 lawyers or open more than one office in each Bar Association area or be formed as a
company (TDC, 1992, p. 32).
32 Analysis by competition authorities is along the same lines, generally being favorable to the intro-
duction of new business structures in the professional service field. See, particularly, the OECD report
on the professions (OECD, 1985, pp. 70-3). With respect to the specific case under study here, the
favorable assessment of multidisciplinary practice, including law, of the Australian Trade Practices
Commission in its review of the competitive position of auditing, is of interest (TPC, 1991, pp. 12-3; or,
for a summary, Kestigian, 1992, p. 16).
33 The financial press constantly reports how, for example, European law firms link up with foreign
firms, often justifying this by the need to provide global services. Such links usually begin with a client
exchange agreement and proceed to the use of a common name or include common elements and
eventually result in a merger.
34 See "Accountancy: All Change" (The Economist, 17 November 1992).
ices required by large international companies, including auditing, consultancy and

legal services. After only a few years it already sees itself more as a specialist than
as a provider of all types of service. 35 It is otherwise clear that not all audit fIrms
enjoy comparative advantages in efficiently providing non-audit services. In this
respect the mushrooming of consultancy fIrms specialized in selling their services to
audit fIrms themselves is revealing (Teng, 1994). Even within the group of the Big
Five fIrms there are substantial differences, with one of them (Deloitte and Touche)
rejecting until now the idea of building an international network for the provision of
legal services.
These cases show that the market not only provides incentives for trying out new
formulas but also for controlling possible problems which they can give rise to,
whilst fIrms recognize where their true comparative advantages lie in an environ-
ment which is constantly changing. Moreover, the fact that non-audit services are
not always profItable can also be interpreted as an indirect indication that improper
means are not being used in this strategy, which would have made the gamble more
certain. This all suggests legislators should be tolerant with such attempts, and give
the participants sufficient time to discover where the new equilibrium lies in terms
of optimum specialization of different types of fIrm.
6.4.3.3 Competitive situation of the markets for non-audit services
The possibility of competing with audit fIrms, in what to them are complementary
services, is even clearer when looking at the competitive conditions in which these
services are provided. These conditions are highlighted in various empirical studies
which demonstrate how the auditor is generally only one of several actual and many
other potential providers of such services. These studies also reveal that auditors
must compete not only with specialized service providers but also with other audit
fIrms, and that their clients do not seem to have the impression that the auditor's
contribution is essential in most services:
a) Competition between large firms in non-audit services. In a sample of 298
fIrms used in the Palrnrose study (1986), all of which had one of the large fIrms as
auditor, 259 of them had during the previous year acquired other services, as well as
auditing, from one or more of these large audit fIrms. Of these, 35 fIrms (13.5 per-
cent) had acquired some or all of such services from a fIrm other than their own
auditor (1986, p. 407).
b) The subjective importance of the non-audit services. Beattie, Brandt and
Fearnley (1996) carried out a study specifIcally aimed at ascertaining how fInance
35 See, for example, "A Glimmer of Hope" (The Economist, 5 April 1995).
directors of listed companies and partners in the 20 main British audit firms per-
ceived the importance of non-audit services provided by the auditor. The results,
summarized in Table 6.2, show that services are acquired from both auditor and
third parties and that some of them are purchased mainly from third parties (finance,
information technology). Moreover, in terms of the services which auditors pro-
vide, these are concentrated in the areas closest to the accounting cycle (tax,
assistance in preparing accounts and regulatory advice), from which the authors of
the study concluded that "the bulk of non-audit services that auditors provide is not
management consultancy" (p. 96).
Table 6.2. Views on the importance of non-audit services and purchasing profiles in
Great Britain
Audit partners sample:
Mean score Finance directors (in respect of the client "X" that
sample: generated the largest total fee)
Non-audit (J = not important. percent of companies Percent of client X
service
5 = very important) purchasing service companies purchasing Percent of
service non-audit
Finance Audit From From From From fee to
current current
Directors partners elsewhere elsewhere incumbent
auditor auditor
Corporate taxation
(inc. VAT, PAYE) 3.69 4.55 82% 26% 88% 21% 37.3%
Accounting advice 3.41 4.34 80% 6% 78% 3% 7.9%
Due diligence
(e.g. acquisitions) 3.23 4.31 73% 27% 77% 14% 21.7%
Stock Exchange
circulars 4.29 66% 12% 11.0%
Regulatory advice 3.79 41% 15% 4.2%
Corporate finance 2.13 3.60 20% 59% 38% 26% 5.5%
Accounts prepara-
tion assistance 1.99 2.41 44% 11% 32% 4% 2.7%
IT 1.71 3.32 16% 50% 32% 34% 5.0%
Source: Beattie et al. (1996, pp. 95-6).

7 NORMATIVE POSSIBILITIES
FOR NON-AUDIT SERVICES
In this Chapter, the analysis in previous chapters is applied to shed light on the
normative question: What if any is the most appropriate type of regulation of non-
audit services provided by auditors? Firstly, it briefly explains the background to
the debate, highlighting the most relevant points raised in the many reports which
have dealt with this topic. A normative proposal is then articulated. This basically
consists of leaving market forces free to fmd an efficient equilibrium in terms of
both the mix of services and the right combination of quality safeguards, with
regulation intervening, at most, to oblige audit fIrms to disclose the percentage of
their total fee income which their biggest client accounts for.
7.1 The public debate
The debate on auditors providing non-audit services, which arose in the United
States in the 1960s, crossed the Atlantic in recent years and took root in Europe,
where the most recent pronouncements on the matter have abounded, with several
expert committees looking into the matter. Firstly, as might be expected, it has
concerned committees of sector scope, occupied with the problems of auditing it-
self. It has also drawn the attention of those responsible for examining general
problems in relation to corporate governance.
7.1.1 The initial debate in the United States
The debate reached a high point in the 1970s in the United States. In a political
environment rarefIed by the Watergate crisis and the unearthing of a series of finan-
131
cial scandals with a political background, 1 a Senate sub-committee examined

auditing and accounting information practices. In 1976, the sub-committee, chaired
by Senator L. Metcalf, reported that one of the main causes of the alleged lack of
independence oflarge audit firms was the scope of the services which they provided
to their clients. The report was specifically concerned with the effect of providing
various management advisory services (executive recruitment, marketing analysis,
plant layout, product analysis, and actuarial and fmancial management services) as
well as tax services on the more traditional accounting and auditing. Based on the
fact that the provision of management advisory services "necessarily involves the
[then] 'Big Eight' firms in the business operations of their clients", it concluded that
this provision "conflicts with the 'Big Eight' firms' obligation to be independent in
fact of their clients [... because they] cannot act effectively as independent auditors
when they have financial and professional interests in the business operations of
their clients" (Metcalf, 1976, p. 50). The report was particularly critical of the pro-
vision of executive recruitment services. On this regard, it claimed that if the
executives recruited were unsuccessful, the value of placement departments would
be reduced, and therefore firms would have an interest in covering up the failure of
those they had placed (p. 51). Consistently, in Recommendation 8, the report pro-
posed prohibiting auditors from performing non-accounting management advisory
services.
During the same period, an independent report from a working party chaired by
M. Cohen (1978) did not, however, detect any appreciable empirical relationship
between sub-standard audits and the provision of management consultancy services
(1978, pp. 95-lO3). Indications were found in just one case in which the auditor's
independence was compromised, relating to advice on a merger. None were found,
on the contrary, in either audits subject to litigation (pp. 94-5) or in tax services (p.
97). On the other hand, the report verified the advantages provided by joint provi-
sion of audit and management consultancy due to the existence of positive external
effects between the two. It also acknowledged that auditors often point out defi-
ciencies, weaknesses and opportunities for improvement to their clients. It thus
considered logical that the client would ask the auditor for help in rectifying them.
This possibility even increases the incentives to detect shortcomings.
The matter subsequently lost its controversial tone. Firstly, it became depoliti-
cized, and was restricted to the more technical arenas and aspects. Secondly, the
response of professional organizations and firms seems to have been at least effec-
tive in avoiding the appearance of harm to independence. The debate has
occasionally been reopened, as recently by the President of the Securities and Ex-
change Commission, Arthur Levitt, when criticizing some of the new activities of
some audit firms. It is interesting to note, however, that even in this case the criti-
I See Briloff (1976), Benston (1979-80, p. 30) and Metcalf (1976, p. III). For earlier references, see
Cannichael and Swieringa (1968).

Normative possibilities for non-audit services 133
cism focused on those actIvItIes which auditors themselves have conventionally

considered dangerous from the point of view of the appearance of independence,
such as "participating in management activities of audit clients [or] selling services
that leave them auditing their own work".2 On doing so, a similar line was main-
tained to that set out by the Staff report on auditor independence from the Office of
the Chief Accountant, in March 1994 (SEC, 1994).
7.1.2 The eclecticism of Cadbury and other reports on corporate

governance
In the field of corporate governance, the pioneering reference in a European context

is the Cadbury Report on the fmancial aspects of running British companies. In
relation to the question here, the Report ruled out the notion of prohibiting auditors
from providing additional services. The Report firstly considered that this would
limit the freedom of companies to choose their sources of advice and could increase
costs needlessly. Moreover, it considered that these drawbacks would not be offset
by any increase in auditor independence, which it considered solely potential. Con-
sequently, it recommended that each audited body should publish the fees paid to
auditors for non-audit work. The two relevant paragraphs of the report, entitled
"'Quarantining' audit from other services", stated literally:
Among the propositions made to the Committee to strengthen the objective relation-
ship between auditors and management, one was that audit finns should not provide
other types of service to their audit clients. The argument runs that such a prohibi-
tion would remove any pressure on the auditors to give way to management on audit
matters in order not to jeopardise their other business services; and that it would re-
move any incentive for auditors to take on audits at rates which could risk corner-
cutting in the hope of obtaining more remunerative non-audit work.
Such a prohibition would limit the freedom of companies to choose their sources
of advice and could increase their costs. The Committee was not persuaded that any
potential gains in objectivity would outweigh these disadvantages. It does, however,
strongly support full disclosure of fees paid to audit finns for non-audit work. The
essential principle is that disclosure must enable the relative significance of the com-
pany's audit and non-audit fees to the audit finn to be assessed, both in a UK context
and, where appropriate, a world-wide context. We recommend that the 1991 Regu-
lations under the Companies Act on the disclosure of remuneration for non-audit
work should be reviewed and amended as necessary in order to apply this principle.
We also regard it as good practice for audit committees to keep under review the non-
audit fees paid to the auditor both in relation to their significance to the auditor and in
2 See Rankin (1997).

relation to the company's total expenditure on consultancy (Cadbury Committee,

1992, pp. 38-9).
In Spain, the matter was dealt with in a similar manner in Section 11.2 of the
Olivencia Report on auditor independence. The Report concluded by stating that:
In general terms, the greatest guarantee of independence is based on the reputation of

audit firms, since they all have a strong commercial reason to preserve and increase
it. It cannot be ruled out, however, that they are sometimes subject to pressures
which lead to them relaxing their standards of conduct and, therefore, it seems desir-
able to urge boards of directors and audit committees to be attentive to such
circumstances of audit firms which could indicate a risk situation. Specifically, it is
recommended that the degree of diversification of the audit firm be verified and that
those firms should not be contracted in which the fees received from the company
represent a significant percentage, for example exceeding 10 percent, of their total in-
come.
It is also common for the audit firm to provide the company with other types of
professional services (consultancy, legal advice, etc.) This Committee, whilst con-
sidering this practice perfectly legitimate, recommends that various precautions be
taken, such as including the income from these areas in calculating the degree of di-
versification and publicly disclosing the total fees which the company pays to the
audit firm for services other than aUditing.3
Consequently, the report, in Recommendation 21, advises "that the board of di-
rectors and audit committee monitor situations which could involve a risk to the
independence of external auditors and, specifically, check the percentage which the
fees paid under all headings represent of the total income of the audit firm and that
professional services other than auditing be publicly disclosed".
7. 1.3 The opinion of sector reports
It suffices to include a representative sample of the more recent of the numerous

reports and positions adopted, dealing in greater depth with those which have had
and continue to have more influence on the debate and regulatory initiatives in this
field.
7.1.3.1 The Auditing Practices Board
One of the first pronouncements on this matter in Europe was by the Auditing Prac-
tices Board, a body created to promote the development of auditing in Great Britain
3 Own translation from CNMV (1998. Section 11.2).

and Ireland and made up of a balanced number of auditors and non-auditors (com-
pany directors, shareholders and users of accounting information).
In its 1992 Discussion Paper, it inclined towards a tolerant attitude, although in-
troducing, as a protective measure, shareholder participation in the decision to
acquire services from the audit fIrm (APB, 1992, p. 11). This idea took shape in the
fmal recommendations issued in 1994, including that the audit committee be re-
sponsible for approving the contracting of the auditor to provide other services.
The same report also recommended that the audit fIrm partner responsible should
not undertake commercial tasks in relation to the other services provided to the
client by the fIrm (APB, 1994, pp. 35-9).
7.1.3.2 Professional Bodies
The question of auditor incompatibilities has also been dealt with by professional
bodies of all types. For example, the International Federation of Accountants
(lFAC) , in its Code of Ethics for Professional Accountants (1996), includes the
following points: "A professional accountant in public practice should not concur-
rently engage in any business, occupation or activity which impairs or might impair
integrity, objectivity or independence, or the good reputation of the profession and
therefore would be incompatible with the rendering of professional services" (Sec-
tion 11.1). However, "[t]he rendering of two or more types of professional services
concurrently does not by itself impair integrity, objectivity or independence" (Sec-
tion 11.2). Moreover, "[t]he simultaneous engagement in another business,
occupation or activity umelated to professional services which has the effect of not
allowing the professional accountant in public practice properly to conduct a pro-
fessional practice in accordance with the fundamental ethical principles of the
accountancy profession should be regarded as inconsistent with the practice of pub-
lic accountancy" (Section 11.3). In short, the IFAC proposed that auditors should
be very rigorous in their work, but considered that integrity, objectivity and inde-
pendence were not undermined by providing other types of service to the client.
The Federation of European Accounting Experts (FEE) adopted a similar view
and put forward various measures to avoid these negative consequences, already
referred to in Section 5.3.1. In particular, it considered that:
An auditor can provide to audit clients other services beyond performance of the
audit, provided that the provision of such services does not impair the auditor's ob-
jectivity and that the provision of such services is not forbidden in the national law or
under the relevant professional rules.
The provision of other services to an audit client entails benefits as it will increase
the auditor's understanding of the client's business and may result in a better audit.
However, it may entail threats to the auditor's objectivity and the appearance of in-
dependence (FEE, 1995, Section 4.2).
The so-called European Contact Group, created at the end of 1993 by the eight
largest auditing organizations to examine possible responses to the gap opened up
between public expectations and audit possibilities, reached similar conclusions. In
July 1996 its opinions gave rise to a report containing a series ofrecommendations. 4
In the specific area of providing other services to audit clients, the report considered
that such services encouraged an understanding of the client's business and opera-
tions thus facilitating a higher quality audit. Nevertheless, it also echoed the
concern raised amongst some outside commentators and therefore recommended
restricting the services to "those whose nature and quality are consistent with the
auditor's professional image and which are unlikely to impair his objectivity".5 It
also recommended that the auditor should never be involved in the client's decision-
making, nor in preparing fmancial statements and other activities, all in identical
terms to Section 4.14 of the European Commission Green Paper, as mentioned in
the Section 7.1.4.
7.1.3.3 The MARC Report
The most detailed examination to date of the rules governing auditing activities in
European Union countries generally was commissioned from the Maastricht Ac-
counting and Auditing Research Centre (MARC) by the European Commission,
resulting in the so-called "MARC Report" (Buijink et al., 1996). The authors of the
Report were roundly in favor of the provision of non-audit services when stating,
firstly, that laws and regulations on auditor independence contain
many measures regarding the supply side ofthe audit market including restrictions on
advertising, and on the scope of services. There is no evidence that these restrictions
make a direct positive contribution towards audit quality but there is convincing evi-
dence on the negative effects of these type of restrictions on (intra-Union)
competition. The fact that (intra-Union) competition is itself an important mechanism
for stimulating audit quality suggests that these restrictions may even have an indirect
negative effect on audit quality. Hence, it is recommended that current national re-
strictions regarding the scope of services, fee setting, unsolicited offering of services,
advertising, and the length of tenure should be removed (Buijink et al., 1996, p. XI).
The content of the above in relation to non-audit services is clarified by the re-
port's authors under the heading of "Scope of Services". Their assessment is that
4 See European Contact Group (1996).

5 Idem, p. 9.
there is no evidence that the existing restrictions contribute to a direct improvement

in quality, but that there is, on the other hand, evidence that they harm competition
and, therefore, also indirectly undermine quality, insofar as quality generally tends
to enhance it. They therefore recommend abolishing such restrictions. In addition,
the authors of the report are more specific in their conclusions when they detail a
recommendation more favorable towards the provision by auditors of non-audit
services. After setting out the potential risks to audit quality, they thus state that:
"empirical research has established that MAS [i.e., non-audit services] contributes
to the detection ability of auditor; MAS gives the auditor more knowledge about the
client. In addition, there is no convincing empirical evidence that MAS creates
independence problems"(Buijink et aI., 1996, p. 141).
7.1.4 The European Union Green Paper
At the end of 1996, the European Commission produced its Green Paper on the role,
the position and the liability of the statutory auditor with the European Union, in
order to promote a detailed debate on the scope of EU intervention in this field and
the need to adopt new measures. The ambiguity of this Green Paper, however,
typical of many consensual texts, has meant that its content can be interpreted in
very different ways. The reason is that the meaning of the Green Paper can radi-
cally vary if the reader overlooks the odd subjunctive or treats as recommendations
what are potentially contradictory statements of the "believe that" or "could be
done" variety.
According to the European Commission, auditors must satisfy growing expecta-
tions, not only in the more conventional areas in relation to the accuracy of financial
statements and the continuity and solvency of companies, but also in relation to the
existence of fraud, compliance with legal obligations and even "the responsible
behavior of the company in environmental and social fields".6 The Commission
rules out the need to engage other professionals in these matters and specifically
states the following in Section 3.35:
Although it is unreasonable to expect the statutory auditor t~ make judgements on

matters outside his competence and expertise, it can be argued that auditors should
accept that their responsibilities will tend to increase in line with public expectations.
The auditor can, given time and a sufficiently clear consensus on what is expected,
avail himself of the necessary expertise in areas which go beyond strict financial
audit.
6 See Sections 3.7 to 3.35 of the Green Paper.

The scope of external effects which could possibly form the basis of a mandatory
audit rule is impliedly extended. On this matter there is a risk of overlooking the
cost involved in any extension of functions. What is important for the issues under
discussion here, however, is that this extension in the scope of auditing contradicts a
reduction in the services which fIrms provide: they need multi-disciplinary re-
sources to undertake these tasks and it would be diffIcult to use such resources
economically in a framework of providing exclusively audit services. The Com-
mission summarizes different positions on the matter as follows:
The question whether the auditor should provide other services to his audit client has
been the subject of much debate. It has been advocated that the provision of other
services to an audit client is likely to increase the auditor's understanding of the cli-
ent's business and operations and therefore results in a better audit. Is has also been
argued that there is no evidence that the volume of non-audit fees in relation to the
level of audit fees is a threat to objectivity. Nevertheless, there is an equally strong
belief that the provision of a broad range of non-audit services to an existing audit
client can impair audit objectivity. Safeguards are necessary to respond to these con-
cerns (Section 4.12).
After this summary, the Green Paper ventures to formulate what appears to be a
list of normative possibilities, including, fIrstly, those already implemented by the
large firms and, secondly, those of disclosing fees. It then comments that applying
an absolute prohibition on providing non-audit services would resolve the problem
of the appearance of independence. Section 4.13 ends, however, with a warning
that an absolute prohibition would not be viable:
Such safeguards could consist in confining the provision of other services to those
whose nature and quality are consistent with the auditor's professional image and
which are unlikely to impair his objectivity, or in requiring the fee for both audit and
non-audit services to be disclosed. A solution which would consist in preventing the
auditor from providing any other service to his audit client would clearly solve the
appearance problem. However, practice has shown that this solution is not a viable
alternative, as it is too easy to circumvent by having the service provided by an affili-
ated or associated firm.
Finally, Section 4.14 proposes a series of restrictions detailed in terms similar to

professional ethical codes pursuant to which the auditor must avoid any situation
which could compromise his independence: 7
7 This final Section of the Green Paper is virtually identical to the recommendations of the European
Contact Group, created in 1993 by the eight main networks of firms, whose conclusions were referred
to in Section 7.1.3.2.
In all cases the auditor should ensure that he and his finn are not involved in the
management or decision making of his client. The auditor should not be engaged in
the preparation of the financial statements of his client, be involved in the valuation
of assets or liabilities for purposes of recording them in the financial statements, act
for the client in the resolution of litigation which may have a material impact on the
financial statements, or perfonn services having a direct impact upon senior man-
agement such as their recruitment.
The European Commission is not considering new legislative measures or

amendments to existing legislation, however. The only possible exception at the
moment is a sectoral directive on freedom of establishment and freedom to provide
services for auditors. It also intends to set up a committee with special responsibil-
ity for auditing matters to review the international standards on auditing developed
by the International Federation of Accountants (lFAC) in order to determine
whether the application of these standards meets EU requirements (European
Commission, 1998).
7.2 Analysis of the main regulatory possibilities
Throughout this work it has been examined how the provision of non-audit services
has a beneficial effect because it reduces costs and improves the technical compe-
tence of the auditor. Furthermore, freely acting market forces are capable of
generating incentives for audit firms to provide a consistent level of quality and not
disappoint the expectations of economic agents. In this way they correct potential
harmful effects which such services could have on independence in fact or, more
likely, on the appearance of independence. For both reasons, rules prohibiting the
provision of non-audit services are in general inadequate, particularly those of a
general nature such as those in some European countries referred to in Section
7.2.1.1, apart from the question of their scant effectiveness in practice.
The most important guiding principle of regulation should instead be to provide
enough scope for audit firms, self-regulatory bodies and audit clients to discover
through competitive market interaction both the efficient mix of services and the
corresponding quality safeguards, adjusting for the costs and benefits of each possi-
bility. In this process, allowance should also be made for mistakes that might
eventually be made, as firms and clients adapt to new technological circumstances.
The most basic reason to entrust this epistemological task to the market is that the
incentives and the ability of market participants seem perfectly capable in this case
of guiding such a discovery process. Regulators, on the other hand, seem to lack
both the required knowledge and the right incentives to define the efficient frame-
work. The lack of knowledge is inherent to their position as neither producers nor
clients. The defective incentives are rooted in at least two potential biases. Simply
put, they may tend, firstly, to exaggerate eventual external effects and consequently
require higher quality and quantity than optimum, as this additional quality involves
no cost for them. Second, they are bound to be captured by private interests alien to
the audit market, as shown by the variety of prohibitive or restraining regulations
and the analysis in Section 6.4.2.2.
From this point of view, prohibitive rules are inadvisable. The only coercive
regulation which could merit a certain consideration in the area of concern here is
that which, on a complementary basis, is aimed at facilitating or accelerating this
process of discovery in the free market. Specifically, this facilitating function could
be implemented by introducing rules governing the minimum diversification of fees
and/or mandatory disclosure by audit firms. Different variants of these three sets of
normative possibilities (prohibition, diversification and disclosure) will be exam-
ined next. After that, the advantages, within disclosure rules, of a minimalist rule
obliging auditors to disclose the proportion of their total income which they receive
from their largest client will be explained.
7.2. 1 Rules forbidding or restricting the supply of non-audit

services
7.2.1.1 Restrictions on provision of non-audit services
There is considerable variation in the types of service which different domestic

legislative systems allow auditors to provide to their audit clients. Table 7.1 sum-
marizes the current regulations in the main OECD countries. 8 Regulatory models
can be classified into three main types:
a) Countries which allow all types of non-audit services to be provided. These
include Australia, Canada, Ireland, Luxembourg, The Netherlands, Sweden and the
United Kingdom. In all these countries, and in those which allow certain types of
non-audit services to be provided, auditing is nevertheless subject to the generic
rules protecting auditor independence.
8 Within the European Union, regulation of these matters is in the hands of the authorities in each
Member State, with substantial differences found, since there is no EU regulation in terms of providing
non-audit services. The Eighth Council Directive of 10 April 1984, based on Article 54.3.g) of the EC
Treaty, relating to the authorization of persons responsible for the legal control of accounting docu-
ments (84/253/EC), only provides in Article 24, with respect to independence, that "Member States
shall prescribe that such persons shall not carry out statutory audits which they have required if such
persons are not independent in accordance with the law of the Member State which requires the audit".
Table 7.1. Services other than auditing which audit firms are allowed to provide
without restriction to their statutory audit clients and within the same legal entity
Bookkeeping Financial
Legal Corporate
Country and Tax Consultancy advisory
services recovery
accountancy services
Australia' Yes Yes Yes Yes Yes Yes
Austria' No Yes Yes Yes Yes Yes
Belgiumb Not No No No No No
Canada' Yes Yes Yes Yes Yes Yes
Denmark' Yes Yes No Yes Yes Yes
Finland' No Yes Yes Yes Yes Yes
Franceb No No No No No No
Germany" No Yes Yes Yes Yes Yes
Greece' No Yes No Yes Yes Yes
Ireland' Yesd Yes Yes Yes Yes Yes·
Italy No No No Nof No No
Japang Yes No No Yes Yes No
Luxembourg' Yes Yes Yesh Yes Yes Yes
Netherlands' Yes Yes Yes Yes Yes Yes
Norway' No Yes Yes Yes Yes Yes
Portugal' Yes Yes No Yes Yes No
Spain' No Yes Yes Yes Yes Yes
Sweden' Yes Yes Yes Yes Yes Yes
UK' Yesd Yes Yesh Yes Yes Yes·
USA'·i Noi Yes k No' Yesk Yes k Yes
Notes: 'The provision of other services must not endanger the independence of the statutory auditor,
however. bThe ban on other services only holds for the provision of audit services within the same
legal entity. Auditing networks provide them in co-operation with separate legal entities. tThe provi-
sion of bookkeeping and accountancy services is allowed on an ad hoc and non-recurring basis. d The
provision of bookkeeping/accountancy services is forbidden in the case of public interest companies
unless this is of a routine clerical nature. 'However, the statutory auditor is not permitted to act as
liquidator, examiner or receiver to that company except in a solvent liquidation (member's voluntary
liquidation). fHowever, Regulated Auditing Firms are allowed to provide consulting services in rela-
tion to the accounting organization. gOnly qualified tax accountants (a separate profession) are able to
provide tax services; consultancy services are restricted in scope; and corporate recovery is not per-
formed by auditors. These restricted services can be provided to clients by networks of firms that have
different legal ownership but share common interests and goals, however. h In practice, legal services
cannot be provided by auditing firms within the same legal entity, however, in view of the rules of the
legal profession. i Refers to U.S. SEC independence rules, which are required of SEC registrants. Inde-
pendence restrictions apply to other entities and partners/owners of a multi-disciplinary firm. j With
some exceptions such as non-material recordkeeping and small foreign affiliates. k In performing those
services the audit firm cannot make management decisions or act in the capacity of management.
I Except legal services that are not material to the registrant and do not involve representing the client in
a public forum.
Sources: For EU countries, adapted from Buijink et al. (1996, p. 74); for other countries, specific survey
(see details in Sources to Table 3. I).
142 The economics oj audit quality
b) Countries which prohibit the provision of any type of non-audit service, as in

Belgium, France and Italy. Nevertheless, in Belgium and France this prohibition
applies only to the provision of audit services by the same legal entity. Hence,
auditing networks provide them in co-operation with other legal entities. This pos-
sibility has been restricted in France to material services provided to listed
companies.
c) In other countries there is only a prohibition on providing services to audit
clients in respect of particular activities: book-keeping and accountancy is prohib-
ited in all countries except Australia, Canada, Denmark, Ireland, Japan,
Luxembourg, The Netherlands, Portugal, Sweden and the United Kingdom; tax and
management advisory services are allowed in all countries, except Japan; legal
services are forbidden in Denmark, Greece, Japan, Portugal and the United States;
and, finally, provision of corporate recovery services is forbidden only in Japan and
Portugal.
7.2.1.2 Evaluation of prohibitive rules
Two types of rules restricting the provision of non-audit services can be distin-
guished: general restrictions (a) and those which subject them to different types of
safeguard or limitation (b):
a) General prohibitions. Prohibitive rules of a general nature are clearly inad-
visable from the point of view of both the audit market (in which they damage
efficiency and quality) and in relation to competition in service markets. Such a
prohibition would not even be advisable for those auditors with little client diversi-
fication, since the effect of a prohibitive rule on independence is doubtful. It should
be recalled that not only would the quasi-rents associated with clients be reduced
but also the firm-specific assets and, moreover, the rule can easily be evaded by
auditors of lesser reputation. On the other hand, the cost in terms of loss of econo-
mies of joint production are substantial and certain. In addition, prohibitive rules
suffer from enforcement problems since the evidence from those countries where
they have been adopted is that many firms get round the prohibition by using sec-
ond trading names and similar formulas. One of the most worrying aspects of this
breach is its asymmetry, since it is most appropriate for firms of lesser reputation.
Competition is thus distorted in a direction which certainly runs contrary to both the
relative efficiency of the different firms and the intentions of the regulator.
b) Specific prohibitions. Restrictive rules of a more specific nature, both those
prohibiting the provision of certain types of service and those subjecting them to
specific safeguards, have a lower impact but also suffer from substantial problems
of the same type. In particular, they necessitate a high degree of knowledge
amongst those formulating them. Even if they are really necessary, it is likely that
the information necessary to evaluate them and, moreover, to implement them, will
be more accessible to audit firms or their professional bodies than the legislative or
regulatory bodies. Prohibitive rules might thus only be justified in respect of those
services or practices where the evidence of private decisions by firms ensures that
the rule is appropriate, thus confirming good practice in the sector and discouraging
the appearance of marginal suppliers. Even in these cases, however, the legislator
should exercise caution. Firstly, he must avoid rules having an anti-competitive
effect and, secondly, they must be sufficiently flexible to take into account the pos-
sibility that technical changes in the future will make the rule less than optimal.
7.2.2 Rules facilitating the work of the market and self-regulation
As a consequence of all these problems, the legislator would do well to abstain from
introducing prohibitive regulations. He will then be able to concentrate on redefm-
ing audit market conditions in such a way that there are stronger incentives for firms
to safeguard quality and for their associations to act as self-regulators. In order to
strengthen market incentives, the regulator can promote and facilitate competition
by means which include policies aimed at increasing transparency of information
and facilitating the creation of private quality safeguards. Specific measures would
involve both introducing new standards (for example, disclosure of client concen-
tration) and eliminating some of the restrictions currently in force (such as those on
advertising and unsolicited offering of services, the outcome of which has been
shown to enhance quality in the United States).
In this context, it would also certainly be appropriate to review the conditions
governing demand for auditing, matching the legal requirement for statutory audit-
ing to those cases in which the audit which can be produced actually reduces
contractual costs (taken in their widest sense to include all types of external effects
or costs and social benefits). By "audits which can be produced", reference is being
made here to the fact that, even if it is desirable that a certain type of undertaking be
obliged to audit its accounts, it makes little sense to introduce a mandatory rule in
this respect if there can be no supply of audit services of sufficient quality. Demand
and supply should develop in coordination, without leaping into a vacuum similar to
the one which was certainly created in many European countries following the in-
troduction of mandatory auditing.
7.2.3 Minimum diversification rules
The logic for encouraging client diversification is the direct result of the previous
analysis of the role of client-specific assets: if there is client concentration, these
assets can threaten independence; on the other hand, diversification favors inde-
pendence, as will be formally examined in the following section. The function of
diversification rules is to ensure that this latter situation prevails. It is for this rea-
son that diversification criteria are usually applied as a safeguard in many activities
in which there is a potential serious conflict of interest. 9 The rule should be more
effective in preventing at least those anomalous cases in which this concentration is
unduly high.
As a result of this independence-promoting effect, diversification provides bene-
ficial effects generally. Moreover, it is also particularly valuable in the field of non-
audit services since, as explained in Section 5.1.1.2, a large part of the effect of
these services on independence, (that associated with client-specific assets), depends
on the degree of client diversification.
7.2.3.1 Formal analysis
The model in Section 3.3.4 enables the logic of a diversification rule to be formally
illustrated. Let us assume, for the sake of simplification, but without any apprecia-
ble loss of generality, that audit firms do not have firm-specific assets (P = 0), that
the regulatory system includes no professional liability (L = 0), and that it is adopted
as an objective that auditors report when the probability of detection Jr reaches a
certain level. In this case, the condition which defines independent conduct in rela-
tion to a generic client, i, given by inequality [11], becomes:
[39]
By expressing Q as the product of the number of other clients, n-J, and the aver-
age value of the assets specific to one client, v , in tum a function of average costs
C and c ; assuming that the division of quasi-rents (a) is equal for all clients; and
substituting V, and v by the value of the specific assets obtained in [18], it turns out
that:
'7 a (C, - cJ < Jr {a (C, - cJ + kq (n-l) a (C - c)} [40]
and, resolving:
C, - C; Jr
-----''--...!..---:::---<- [41]
C;-c;+k,,{n-l) (C-c) '7
9 A recent example can be seen in a stock exchange context by an incident connected with the privati-
zation of the Spanish public company Auxini. The Privatization Consultative Board, an independent
public agency, believed that the advisor would not be subject to any incompatibility because it was also
responsible for temporarily maintaining the list price, or exchange value, of the acquiring company,
since for the latter service it merely received annual income of less than one million pesetas, insignifi-
cant compared with total invoicing of 6,600 million pesetas (see "La independencia del mi1l6n,"
Expansion, 24 May, 1997, p. 4).
Ideally, it would be desirable to defme the diversification rule or objective in

terms of quasi-rents, as in [41]. IO However this is usually impossible because the
amount of quasi-rents is unknown and difficult to ascertain objectively. On the
other hand, turnover figures are relatively easy to measure and monitor. For this
reason, diversification rules normally establish a maximum limit on the income
which one client can represent in terms of total income. In order that compliance
with this limit becomes equivalent to fulfilling condition [41] above, it has to be
assumed that the volume of individual quasi-rents is proportional to the amount of
fees. (In terms of providing non-audit services, this simplification could be impor-
tant if these services generate quasi-rents in a very different proportion than audit
services. It seems unlikely, however, that the difference could be of sufficient mag-
nitude substantially to undermine the effectiveness of the rule). Under this
assumption, if F j represents the fees received from a generic client, and F the aver-
age fee earned from other clients, it follows that:
Fj Jr
----'----,:: <- [42]
F j + klJ (n-l) F "
which expression, when resolved, enables the condition to be defined in terms that
the proportion of fee income received from each client is less than a certain limit:
[43]
It can thus be seen that the minimum diversification limit increases with the
likelihood of detection used as an objective, Jr. For normal parameter values, it also
does so with the impact which the discovery that an auditor has failed in his duty of
independence has on the value of the quasi-rents associated with the other clients,
k,l' Conversely, it reduces with the probability that the client changes auditor after
receiving an unfavorable report from him, the parameter ".
7.2.3.2 Existing rules regarding minimum diversification
Diversification of auditors' fee income is regarded as necessary by both profes-

sional bodies and legal systems. For example, the Federation of European
Accounting Experts (FEE) favors auditors diversifying their income:
10 In this respect, DeAngelo (198 I b. p. 192).

Because of the real and apparent risks oflosing his objectivity through self-interest in
over-reliance on a single client, the auditor should not accept an audit appointment
from an entity which provides his practice with an unduly large proportion of gross
practice income. The size of the threat and the possible safeguards will vary but the
auditor should recognise that the public perception of his objectivity is likely to be in
jeopardy if revenues from a single client or group of connected clients (all entities
that would fall within the consolidation scope) exceed a small percentage of his aver-
age yearly total revenues of the last 5 years. That small percentage lies generally
between 5% and 15% according to countries (FEE, 1995, section 4.3.1).
Table 7.2. Restrictions on over-dependence on a single client, specifying, where

appropriate, the maximum percentage of audit fees of a single client in relation to
total fee income
Existence of Maximum
Countries Explanation
restrictions percentage
Great Britain Yes 10 or 15% 10% for listed companies
Ireland 15% for other clients
United States Yes 15% Otherwise SEC may challenge
independence
Denmark Yes 20% Over three consecutive years
Austria Yes 30% Over one year
Germany Yes 50% Over five consecutive years
Australia Netherlands Yes The rules consider that receipt

Belgium Norway of a significant portion of fees
Canada Spain from one client or group dam-
France Sweden ages independence but does
Luxembourg not specify quantitative limits.
Finland Italy No
Greece Portugal
Japan
Sources: For European countries, adapted from Buijink e/ af. (1996, pp. 76-7); for other countries,
specific survey (see details in Sources to Table 3.1).
Moreover, rules in this field already exist in various countries. Specifically,

three types of rules regulating the diversification of audit firms' income are found:
The first group of countries, made up of the UK, Ireland, United States, Denmark,
Austria and Germany,1I restrict the maximum income from a single client in relation
11 At the time of writing, an exposure draft is also under discussion in Australia which would address
overdependence of the statutory auditor on a particular client.
to the total income of the audit firm, laying down a maximum percentage (Table
7.2). This percentage ranges from 10 percent for listed companies in the UK and
Ireland to 50 percent in Germany (although the latter is computed over a consecu-
tive five-year period). Some regulations also explicitly limit the existence of unpaid
professional fees due from prior years' services. In most countries making up the
second group of countries, there is a restriction defmed in very general and unquan-
tifIed terms, the practical importance of which thus varies. Finally, no restrictions
exist in this respect in Finland, Greece, Italy, Japan or Portugal.
7.2.3.3 Means available for achieving a diversification objective
The legislator has two main types of instrument available to promote client diversi-
fication amongst auditors. The first is by directly introducing a rule preventing
concentration beyond a certain limit, such as those in countries included in the first
category in the previous section. This regulatory method could perhaps be useful in
establishing a maximum concentration limit to avoid possible external effects,
mainly among audit firms. Nevertheless, such a rule is subject to substantiallimita-
tions which will now be referred to. As a result of these, it should playa secondary
role in terms of the objective of increasing diversification.
Strict rigid diversification rules are inadequate because the diversifying objective
can and must be compatible with freedom for auditors to work with different de-
grees of diversification and other quality safeguards, provided that this is known to
the public. The main reason for this is that the optimum level of auditor diversifi-
cation depends on the characteristics of the auditor himself as well as those of his
clients. Consider an auditor of good reputation who provides substantial guarantees
in terms of assets. It is likely that he will need a less diversified clientele for the
incentives which induce him to be independent to be equally intense as those of
another auditor who lacks these additional reputational and asset safeguards. Simi-
larly, if two auditors are equal in all respects, including in their degree of client
diversification, but the clients have very different characteristics in terms of their
solvency, the auditor who reviews the accounts of the less successful companies,
where there is therefore a higher probability of detection, 1[, will require, ceteris
paribus, lower diversification to enjoy the same incentives as his colleague.
It is impossible for a single diversification rule to satisfy these requirements. As
with all rigid rules, it would force a standard level which would be too high in some
cases and too low in others. Such a rule is not therefore advisable. Moreover, if it
is used, it should only be used as a minimum rule aimed at preventing clearly aber-
rant situations of client concentration. Otherwise, if an ambitious minimum
diversification rule is adopted, participants would be prevented from seeking an
efficient substitute from amongst the different types of independence safeguards.
As these safeguards are costly, it is desirable that substitution decisions (between

solvency guarantees and diversification, for example) be taken by those who have
most incentives and the information necessary to adapt to them in an optimal man-
ner. In this respect, the possibility should also be ruled out of highly detailed
regulation of these safeguard substitutes. There would be two types of problem
with such regulations: firstly, the existing difficulties in establishing rates of substi-
tution and, secondly, those which would arise when attempting to monitor
compliance. At most, some type of very general rule could be established, or one
for different types of firm or client.
7.2.4 Disclosure rules
7.2.4.1 Analysis of the rules' rationale: the variance of non-audit services

and assessment of auditor independence by the market
The least costly regulatory method of stimulating the diversification objective seems
to be by mandating that audit firms disclose indications regarding the degree to
which their fee income is diversified. In principle, the costs of this mandate in
terms of rigidity should be very low because it allows audit firms to adjust flexibly
to different clients and circumstances and to substitute across safeguarding mecha-
nisms. In the following two sections, it will be shown that some disclosure rules
may impose substantial costs without providing many benefits, however. The rea-
son is that they tend to focus on the amount or ratio of non-audit to audit fees. They
do so, however, on an individual client basis that tells nothing about diversification.
Nevertheless, these costs of disclosure might be reduced and at the same time the
informational content is increased if the correct rule is chosen. Let us clarify firstly,
however, that such disclosure could provide useful information to the market in the
case of non-audit services (a) and, secondly, point out some specific circumstances
in which collective action problems might prevent firms from voluntary disclosure
(b).
a) One reason why diversification disclosure is more valuable when the auditor
provides a multiplicity of services to his clients is that the market's capacity for
evaluating client concentration is probably reduced by the fact that he provides non-
audit services. When he solely provides audit services, it is relatively easier to
assess client concentration. In this case, although it is true that there are several
unknowns (mainly the fees received, but also the costs), even external observers can
gain some idea of the scale of the services provided. Just the contrary occurs when
non-audit services are supplied since, in principle, it is not even known whether the
auditor is providing them. To these difficulties in external assessment is added the
fact that both the range and variance are probably much higher for non-audit service
fees than for audit fees (see Table 7.3 for the variance and, for range, the data from
Grout, Jewitt and Whittington [1994, p. 326]12 for example). Consequently, in-
come, and presumably the quasi-rents from providing non-audit services, are
distributed less uniformly than those from auditing, and therefore the market could
have greater difficulty in assessing the degree of auditor independence if unaware of
the volume of services provided. Therefore the provision of non-audit services
makes disclosure more necessary because it substantially increases the variability of
income. When only audit services are provided, it is known who the auditors are
and approximately how much they earn from each client. On the other hand, this
assessment is much more difficult when they provide non-audit services.
Table 7.3. Variability of audit and non-audit fees in a sample of US firms

Standard Standard
Mean
deviation deviation /
(US$)
(US$) Mean
Audit fees:
In initial audit engagements 85,800 78,400 91.38%
In continuing audit engagements 193,200 294,400 152.38%
In total combined sample 151,300 240,300 158.82%
Fees for non-audit services provided
by the principal auditor:
In the year of the initial audit engagement 10,600 28,500 268.87%
[n continuing audit engagements 123,900 600,000 484.26%
[n total combined sample 79,700 471,000 590.97%
Source: Own preparation using sample data from Turpen (1990, p. 66).
b) The desirability of introducing a disclosure rule on a mandatory basis can be

disputed inasmuch that if the information were truly valuable and its disclosure
entailed no cost, firms would already be disclosing it on their own initiative, to
avoid being classified as bad quality providers. t3 The presence of collective action
problems or third-party effects may, however, prevent individual voluntary disclo-
sure from reaching optimal levels with regard to both the amount and content of the
12 Specifically in the case of Great Britain, Grout et al. (1994, p. 326) indicated that amongst the 100
companies making up the FTSE 100 index, fees from non-audit services accounted for between 23.3
and 0.02 times audit fees, although the maximum was atypical and the second highest figure was only
three times the audit fees. SEC data also reveals a very wide range, as well as a decreasing weight of
fees for non-audit services compared with audit services between 1979 and 1992 (SEC, 1994, pp. 29
and 32).
13 See on this Grossman and Hart (1980), Grossman (1981) and Milgrom (1981).
disclosed information. 14 For instance, the disclosure of some diversification indi-

cators could damage the confidentiality of the relationship with the affected client,
possibly revealing information of some strategic value to his competitors. Even if
the private cost is less than the private benefit, it may be the case that the net surplus
is positive only when all firms disclose. However, individual firms may be better
off not disclosing if some other firms do disclose, with the fmal outcome that no
firm discloses. IS Likewise, disclosed information might be more valuable when it is
used to gain knowledge of the industry or, more pertinently to the case under dis-
cussion here, it serves to compare across firms in the industry. In both uses, the
value of the information disclosed by anyone firm is not fully appropriated by it,
generating positive externalities. In addition, disclosure needs ex post verifiability
to function, which may require a centralized agent who accumulates and to some
extent monitors the disclosed data. In performing this task, regulatory bodies may
have an advantage over self-regulatory bodies when the latter represent firms with
diverse degrees of diversification, as it happens in this case. Moreover, in a context
in which the market itself is unaware of the concentration levels of other firms, the
initial discloser perhaps runs a certain risk of confusion amongst economic agents
who are unaccustomed to assessing such information. Furthermore, regulation may
enjoy some advantages over the market in standardizing the contents and the lan-
guage of disclosure.1 6 Lastly, third party effects may appear within the firms
themselves, particularly as a consequence of their hybrid structure. The existence
of inter-temporal or geographical variability within firms could incline them not to
disclose from concern that their image may appear distorted in some periods or,
mainly, some geographical areas. For instance, interests may diverge within a fed-
eration of firms. This could be an important consideration for those still developing
a global standard of quality.
Because of these possible contingencies, regulators should focus their task in this
field on eliminating those barriers that might hinder disclosure on a voluntary basis.
Candidates for this are derived directly from the previous discussion. In particular,
mandatory disclosure should be seen and structured as a solution to free-riding
problems among potential disclosers. The regulator could also defme diversifica-
tion scales in order to reduce the risk of disclosing information which might be
sensitive to the affected client. Moreover, the regulator himself, in promoting vol-
14 For an account of the main issues involved in corporate disclosure which holds a similar line to the
one argued here, see Easterbrook and Fischel (1991, pp. 276-314).
15 Jovanovic (1982), Verrecchia (1983) and Dye (1986) model situations in which uninformed parties
do not always infer bad quality from nondisclosure, due to the presence of disclosure costs. As a con-
sequence, voluntary disclosure is less than optimal. In particular, disclosure is not profitable for those
providers of bad but not the worst quality at some point in the quality scale.
16 This can be particularly important when limiting discretion controls the informational content of the
disclosure because discretion affects how the disclosed information is seen by recipients, as modeled in
Fishman and Hagerty (\990).
untary disclosure, would eliminate the risk of confusion borne by the initial
discloser. He can, fmally, act as ex post verifying agent and, by emphasizing the
substitutability of safeguards, he might even aspire to reduce the risk of misinter-
pretation. Let us examine now how to materialize these generic principles into
concrete rules.
7.2.4.2 Mandatory disclosure offees and fee ratios
There are various possibilities for drawing up a disclosure rule. The choice should
take into account its relative effectiveness and the costs it may generate. In par-
ticular, effectiveness will depend on whether it provides the market with
information useful to correctly assess economic incentives favoring independence.
With respect to the cost, some rules can affect competition: for instance, the disclo-
sure of fees makes competitive strategies more difficult and less profitable for audit
firms, reveals sensitive information about their clients and, in some cases, facilitates
the monitoring of collusive agreements in the audit market. In any case, given that
mandatory disclosure surely involves costs and benefits, its adequacy is an empiri-
cal question that carmot be solved. Only arguments and principles can be put
forward which, in conjunction with an examination of the previous regulatory expe-
rience, might perhaps be useful in designing efficient disclosure policies.
a) Disclosure rules. A few countries require auditors or their clients to disclose
fees (Table 7.4). Specifically, in Ireland, Norway, the United Kingdom, Denmark
and Belgium, the auditor must disclose fees received for both mandatory auditing
and non-audit services, except in Belgium where the obligation relates to service but
not audit fees. Likewise, in Italy clients must specify audit fees in their armual fi-
nancial statements, although this rule does not apply if the audit has been carried out
by one of the regulated firms, which are the larger ones. 17 Finally, in Australia, the
financial statements of publicly listed companies make full disclosure of remunera-
tion received by the auditor for both audit and other services.
b) The SEC experience. In the United States, the Securities and Exchange
Commission also, from 1978 to 1982, required a proxy disclosure,18 which obliged
SEC registrants to disclose the ratios of total and material individual non-audit
services fees to audit fees,19 as well as affirmative disclosure of whether the board
17 According to Buijink et af. (1996, pp. 80-1).

18 See the SEC's Accounting Series Release (ASR) no. 250 (SEC, 1978). One year later, the SEC issued
an interpretative release (SEC, 1979) describing several factors that auditors and registrants should
consider when contracting non-audit services.
19 Disclosure of fees was first contemplated but finally discarded. The precise requirement was to
disclose "the percentage relationships to the audit fees of the fees for the aggregate of all non-audit
or the audit committee had approved each such service. The requirement was re-
scinded in 1982, however, because the SEC itself came to believe that it was "not
generally of sufficient utility to investors" (SEC, 1982).20 It is not surprising after
this experience that the SEC report on auditor independence in March 1994 ruled
out the possibility ofre-introducing a fee disclosure rule. 21
Table 7.4. Regulations regarding public disclosure of individual fees
Country Required disclosure of audit fees Required disclosure of non-audit fees
Australia Yes Full disclosure in financial state- Yes Full disclosure in financial state-
ments of public listed companies ments of public listed companies
Ireland Yes Auditing firms need to disclose Yes Auditing firms need to disclose
Norway audit fees for statutory audits. If an fees for non-audit services
Britain auditing firm is (required to be) a
shareholder, the shareholding
needs to be disclosed
Denmark Yes Auditing firms need to disclose
audit fees with respect to listed
and large companies
Belgium No
Italy Yes Annual accounts need to disclose No

fees of the statutory auditors but not
the fees of the Regulated Auditing
Firms
Other No No
countries'
Note: 'including Austria, Canada, Finland, France, Germany, Greece, Japan, Luxembourg, Netherlands,
Portugal, Spain, Sweden and United States.
Sources: For European countries, adapted from Buijink et al. (1996, pp. 80-81); for other countries,
specific survey (see details in Sources to Table 3.1).
The US experience with the SEC disclosure requirement and its accompanying
interpretative release calls for caution when assessing the impact of disclosure on
services and for each non-audit service that results in a fee 0 three percent or more of the audit fee"
(SEC, 1978).
20 See SEC (1981; 1982; 1994, pp. 27-31). At the time of rescission, the membership provisions of the
accounting profession self-regulatory body, the SEC Practice Section of the AICPA (SECPS) were
revised in response to an SEC belief that information was still needed to continue monitoring non-audit
activity, expressed by the SEC in ASR 296 (SEC, 1981). As a consequence, since 1982 SECPS mem-
bers are required to disclose in their annual public report to the SECPS the total amount of fees earned
for audit, tax and management advisory services, as well as certain percentage relationships of their
non-audit fees to their audit fees on an aggregate basis for all of its SEC audit clients.
21 Auditing standards for publicly quoted corporations in the United States do require, however, that
each audit firm discloses annually to the audit committee of their clients the nature and amount of the
fees received for non-audit services provided to such clients.
parties' behavior. Accounting firms, in particular, claimed that it had produced a

curtailment of non-audit services. It was alleged that parties perceived that the SEC
was deprecating the benefits of management advisory services performed by the
auditor and that the SEC might question or criticize the independence of the auditor
considering only the disclosed ratios of non-audit to audit fees. 22 Seemingly, at that
time some clients were even setting arbitrary percentage fee limits on the non-audit
services provided by their auditors and deciding not to engage their auditors to
perform services discussed prior to the issuance of the interpretative release ASR
264, without considering the nature of the service and the potential impact on inde-
pendence. Even if there are discrepancies regarding the real impact of the SEC
rules,23 the controversy shows that some mandatory disclosures or, in particular,
regulatory activism on the matter risk over-emphasizing the negative effects of non-
audit services. (Unfortunately, it is impossible here to disentangle the two effects:
the impact of disclosure on the market, which might be considered less risky, from
that derived from possible criticism or action by a powerful regulator.) In both
cases, the diffIculties that less-informed market participants have exhibited in cor-
rectly evaluating auditor independence (remember their apparent bias against non-
audit services, documented in Section 5.2.1), may lead clients to acquire less non-
audit services from the auditor than optimum.
The case is worth a short digression as it calls for regulatory caution by showing
that, in general, regulatory activism of one sign of another may paradoxically
worsen the information of less informed parties, especially when the latter lack
incentives to expend resources in ascertaining the issues, relying instead on indica-
tions provided at no cost to them. When issues are not clear cut, regulators should
thus aspire to be neutral in the information they provide to the market. The risk is
of no lesser importance for those regulatory bodies with scant reputation, because
they usually perform their duties in markets with less informed parties.
Returning to the specific issue of disclosure, both of these rules, mandating dis-
closure of the fees received by client and type of service, and the revoked SEC rule
relating to the ratio of fees from each client for non-audit services compared with
audit fees, probably are unnecessary and costly. They are unnecessary because
22See SEC (1981, p. 3810; 1994, pp. 28-9).

23At least one study claimed that the rules did not significantly reduce the quantity of services provided
(Scheiner, 1984). This result is questionable, however, on at least two accounts. First, the estimation
did not consider the expected growth of non-audit services. In other words, the comparison was made
between quantities before and after the rules and not with and without the rules. Second, the impact of
the rules on clients with the largest ratios of non-audit to audit fees (the ones claimed to be most af-
fected) might have been underestimated. The impact on clients buying substantial non-audit services
was indeed tested, but this test was made considering as "substantial" those clients with total non-audit
services greater than the audit firm median. Both problems likely have the effect of substantially un-
derestimating the alleged reduction in non-audit services.
independence does not depend on the volume of individual fees or quasi-rents, but
on diversification, as demonstrated in Chapter 3. They are also costly because they
involve disseminating information the confidentiality of which is important to both
audit firms and their clients in developing competitive strategies and might also
facilitate possible oligopolistic pricing arrangements. On the other hand, it is un-
likely that fee disclosure will enhance competition if due consideration is given to
the process of intertemporal competition. Fee disclosure could at most modify the
relative bargaining power of clients and incumbent auditors. In that case, incum-
bent auditors would appropriate a smaller stream of quasi-rents. It could thus
discourage investment in assets specific to each client and, then, even reduce inde-
pendence ex post (in line with the previous argument regarding the safeguard role
played by client-specific assets when auditors are well diversified).
7.2.4.3 Mandatory disclosure of maximum individual concentration
Taking these factors into account, when choosing informative diversification indi-
cators, due weight should be given, firstly, to the risk of introducing a bias against
non-audit services and, secondly, the cost of losing price confidentiality. For these
reasons, if a disclosure rule is to be adopted, this rule should mandate a minimum
disclosure: a good candidate is the ratio between total fees from a single client (the
biggest one) and total fee income. This information seems sufficient to promote
market supervision and to accelerate the development of firms' reputations, mini-
mizing the risks of adverse effects. Furthermore, this type of solution has
occasionally been used by firms themselves on a voluntary basis. 24 Conversely, it is
irrelevant from the perspective of the arguments maintained here that the market
should know the relationship between fees from non-audit and audit services for
each and every client. This figure is irrelevant in terms of independence for all
those clients where the sum of the two items does not amount to a significant pro-
portion of the auditor's totalfees.
This judgment is grounded in the theory developed in Chapter 3. It was then
shown that the effect on independence of quasi-rents associated with each client
changes sign when the degree of diversification varies. In practice, it would suffice
for the diversification rule to oblige each auditor to disclose that the percentage
which the fees from his principal client bear to his total fee income is below a cer-
tain level established by the regulator. 25 To reduce the costs of identification, the
24 DeAngelo (1981 b, p. 193) mentions one case in this respect: the 1977 annual report of Peat, Mar-
wick, Mitchell, which stated that the single largest audit fee comprised only 0.5 percent of total
revenues.
25 When computing the level of diversification in all of these rules, it seems reasonable to apply a
decision unit criterion to both client and auditor. In practice, this means that the auditor's income from
all companies included in the group of companies of each client would be taken into account. For the
regulation could, as an alternative, defme a diversification scale comprising several

categories. Audit firms would then disclose where they fit within that scale in terms
of client diversification. Consideration could also be given to indicating the audi-
tor's position in the scale in each audit report alongside signature.
A disclosure rule limited to maximum concentration scarcely suffers from the
problems of general price disclosure. It firstly enables the acquisition by clients of
non-audit services to remain anonymous, and they thus provide no information to
their competitors, which would place auditors at a disadvantage compared with
other service providers. Furthermore, the interests of third parties including, princi-
pally, other clients (as discussed in Section 1.3) are met without the need to disclose
the identity of the principal client. Finally, the bias which could be introduced in
public perception of independence is less than with a rule of the type adopted by the
SEC and mentioned previously, thanks to its general nature and that diversification
represents a broader series of safeguards. This risk would be the greater, however,
the more explicit the disclosure obligation and the more enthusiastic the regulatory
body in publicizing or enforcing the rule. For this reason both the type andirnme-
diate recipients of disclosure must be determined with extreme care and the
regulator should take into account his role as opinion maker in this respect. Moreo-
ver, careful attention should be paid to how the market reacts, in order to detect
unforeseen costs and benefits which might make it advisable to modify or even
repeal the rule.
same reason, those audit firms with subsidiaries would have to compute the income they receive from
each client by all companies in their group.
Part 3
SUMMARY AND
CONCLUSIONS
8 LET KNOWLEDGE EVOLVE
IN THE MARKET
This book firstly examined the private mechanisms used to safeguard quality in
auditing, in order to defme rules capable of facilitating the performance of market
forces. Thus, a general theory of private quality assurance in auditing was devel-
oped, based on the use of "quasi-rents" to self-enforce quality dimensions. This
theory was then applied in the second part of the book to search for and define an
efficient regulation of the provision of non-audit services by auditors and auditing
firms to their audit clients. From examining the particular problems posed by these
services, several conclusions have been reached regarding their effects: they are
claimed to reduce total costs, increase technical competence and motivate more
intense competition. Furthermore, they do not necessarily damage auditor inde-
pendence nor the quality of non-audit services. This assessment leads, in the
normative sphere, to recommending a legislative policy aimed at facilitating the
development and use of safeguards provided by the free action of market forces.
Regulation should thus aim to enable the parties-audit firms, self-regulatory bod-
ies and audit clients-to discover through competitive market interaction both the
most efficient mix of services and the corresponding quality safeguards, adjusting
for the costs and benefits of each possibility. Particular emphasis has been placed
on the role played by fee income diversification and the enhancement through dis-
closure rules of market incentives to diversify.
159
8.1 Private quality safeguards and the role of regulation
8. 1. 1 Dimensions of auditing quality which are crucial for

regulation
The demand for auditing services arises from a need to facilitate dealings between
the parties involved in business relationships-shareholders, creditors, public
authorities, employees and customers, etc. Exchanges between such parties are
usually costly since informational asymmetries give rise to uncertainty as to
whether they will perform their contractual obligations. Accounts, along with other
instruments and safeguards, are used in this context of contractual opportunism to
enhance the likelihood of performance and demonstrate a willingness to perform to
contracting parties. The preparation of accounts is controlled by one of the parties
only (either the executives of a large company or an individual businessman in
smaller undertakings). This occurs because, in this way, they can be produced at
lower cost and the figures are also used-perhaps principally-for internal man-
agement and control purposes. This dual capacity of executives and businessmen as
agents to be monitored and those actually preparing the accounts reduces their value
to third parties. A review of these accounts by an independent expert, the external
auditor, is useful to enhance their reliability in the eyes of other contracting parties
who are not involved in their preparation. There is thus a demand amongst compa-
nies for independent auditing to reduce their contractual or transaction costs. This
has been verified empirically by observing that the increased contractual conflict
resulting from a greater specialization of ownership and control or a higher level of
indebtedness leads to more demand for auditing services. It has also been observed
that contractual costs fall largely as auditing quality is increased.
Auditing quality depends on the ability of the auditor to carry out a thorough ex-
amination of the accounts and detect possible errors or anomalies (technical
competence) and his willingness to provide an objective opinion in relation to them
(his independence). Careful consideration of these two dimensions and the quality
safeguards that are provided by the market are crucial for the regulation to be effi-
cient. In particular, technical competence can be hindered and independence
trivialized if auditors are not allowed to exercise their professional judgement (Sec-
tion 8.1.1.1). In addition, independence is not only a problem confined to the
parties directly involved but it also concerns third parties, mainly other clients (Sec-
tion 8.1.1.2).
8.1.1.1 Professional judgement
Auditing increases the informational value of the accounts and it is therefore desir-
able that the auditor's opinion should reflect as much information as possible.
Summary and conclusions 161
Hence, the exercise of professional judgement by the auditor becomes an important

attribute of auditing quality. By means of this professional judgement, the auditor
incorporates a picture of the business into his report based on facts known to him,
but which are difficult for third parties to verify. The informational value of audit-
ing to third parties is thus enhanced. As a consequence of the need to convey such
unverifiable information, it is not generally optimum for auditors to be merely inde-
pendent. What is required of auditors is that they exercise their professional
judgement independently. It would be prejudicial however if, in order to preserve
his independence, auditors are obliged to refrain from making a professional
judgement since this provides valuable information to those using the company's
accounts.
To examine this issue, let us assume that auditors can use two types of informa-
tion in their professional judgement: "hard" information, which is observable and
verifiable, and "soft" information which is observable but not verifiable. In the case
of auditing, hard information can probably be suppressed from the accounts but not
equivocally transmitted whilst soft information can be the subject of equivocation
and subjetiveness in transmission. In order to maximize the value of auditing, the
regulatory framework must make it possible for auditors to use and transmit this
soft information regarding the client to the market and third parties. To this end,
they must exercise their professional judgement based on both categories of infor-
mation. If the regulatory framework penalizes them excessively they will only
make use of the hard information. Consequently, audits will contribute less infor-
mation and may sometimes even be misleading. The latter may occur when
auditors are obliged to play safe and base their opinions on hard information. They
would then, for instance, disclose their reservations even when they consider that
the client's situation would not warrant such reservations after taking into account
the soft information available to them regarding the client, information that they are
not in fact allowed to transmit.
This analysis has important consequences for legislative policy. It might in prin-
ciple be thought that an appropriate stiffening of the system of professional liability
and penalties would suffice to bring about greater independence. This may well be
the case but unfortunately it is not the only consequence of making penalties more
severe. If the criteria or system used to distinguish between proper and improper
action is imperfect in that only part of the relevant information is used then audi-
tors' opinions will tend to be based on that information alone. Other information is
discarded, however relevant it may be in assessing the situation of the business
being audited. In other words, the informational value of auditing can be substan-
tially reduced if it is simply based on evidence verifiable by third parties (whether
judges or regulators), with all information which is available to the auditor but not
to third parties being wasted.
It is also likely that, when assessing the quality of auditing work, market-based
penalty mechanisms (particularly reputational damage) are capable of using a much
broader range of information than can be used by judicial and regulatory mecha-
nisms. A principal reason for this advantage is that the market is not restricted to
the use of specific types of evidence. Furthermore, it accumulates many individual
decisions, with biases in each one of them tending to balance out and end up being
of little importance. Moreover, such decisions are taken largely by professionals
who can be assumed to be well informed. Finally, negotiating costs in this case are
nil, as compared with their equivalent in the legal process or an out of court agree-
ment. As a result, much information which is observable by auditors but not
verifiable by the legal system is verifiable by the market and auditors will have no
hesitation in using it if the potential sanction comes from the market rather than the
legal system. Overall this means that it might well be desirable for regulation and
the judicial system to play more of a facilitating role than to provide a substitute for
market-based penalties.
8.1.1.2 The value of independence for third parties and other clients
Regulation should also take into account that the asymmetry and conflicts posed in
relation to auditing quality are more complex than in most markets since it is deal-
ing with contractual facilitation with at least three parties having an interest (clients,
auditors and users). In this respect, auditor independence is the most conflictive
attribute. Firstly, the role played by auditing in third party contracting means that
the client will require independence or otherwise depending on his situation. In
problematical cases, although the client in principle usually wants high quality
auditing, after his fmancial situation changes he may prefer a low quality audit in
terms of independence. More precisely, he would prefer a dependent auditor per-
ceived by third parties as being independent. This deceptive audit might enable him
either to contract with third parties on better terms than those available in his situa-
tion if the same is known to such third parties, or to postpone review of those
contracts which involve corrective action based on financial situation.
Third parties who contract with a client of an auditor in reliance on an audit
which they believe is independent are not the only parties prejudiced by a deceptive
audit, however. Other clients of the audit firm will also suffer when their name is
associated by the public with an auditor reputed to lack independence. Moreover,
unlike third parties these clients will have no recourse to the judicial system to ob-
tain compensation for the harm caused to them by this lack of independence. Much
of the conflict relating to independence thus arises not so much between the auditor
and those using the accounts but between the auditor and those other clients who
demand and have contracted an independent stance from him. As a result, a large
part of the conflict regarding independence does not manifest itself so much be-
tween auditors and those using the accounts of clients in a difficult situation as
between auditors and their other clients who do not wish the independence of their
auditor to be diminished. Regulation would therefore do better by modifying its
current emphasis on the effect that auditing failure has on third parties, because in
so doing it would probably also strengthen market incentives.
8.1.2 Market-based quality safeguards and the role of regulation
There need to be safeguards in most economic exchanges to ensure that each party's
pursuit of self-interest does not prejudice the outcome of the exchange, thus endan-
gering its very existence. As well as external mechanisms, most importantly the
legal system, market-based instruments develop to ensure contractual performance.
These essentially involve automatic penalties for breach by means of which the
party failing to perform is castigated when contractual parties or the market in gen-
eral find out that he has breached his obligations to a market member. These
mechanisms usually operate in a simple manner by the ability to charge a higher
price than the highest which could be charged if a breach is observed (the premium
being usually known technically as a "quasi-rent"). The most important example of
this type of remuneration is that associated with a good commercial reputation,
which enables a higher price to be charged compared with those who contract with-
out such reputation. The ability to continue charging this higher price will only be
ensured, however, if a level of quality is supplied which accords with expectations
and reputation. Breach generates a more or less automatic penalty and, therefore,
reputation acts as a quality safeguard.
There are various more or less costly ways in which an economic agent can be
placed in a position where he has an incentive to perform his obligations in the face
of a threat that he will be penalized by the market-i.e., his potential contracting
parties:
a) Low-cost safeguards. Low-cost strategies for generating performance incen-
tives are based on prior performance of obligations, a cross-market reputation
spread and the use of real assets with a value linked to a continued presence in an
activity or market. These are all applicable to the auditing field and because of their
low cost are preferable from the social point of view. Regulation would thus do
well to facilitate them and enhance their effectiveness.
• The most elemental strategy of this type consists in building a reputation by
providing higher than expected quality and refraining from opportunistic behav-
ior. This is of fundamental importance and forms the cornerstone of the system
of safeguards. It operates slowly, however, and can also result in allocative inef-
ficiencies when introductory pricing is used, even though these inefficiencies are
probably of a secondary nature. Moreover, the value of a reputational guarantee
is highly sensitive, a feature which is precisely the result of its high degree of ef-
fectiveness. This sensitivity, however, becomes a disadvantage when subjected
to opportunistic or arbitrary action. For this reason, regulation must ensure that
reputational penalties are in line with the breach of professional duties. In par-
ticular it must prevent those not in breach from being penalized, avoiding undue
loss of reputation being inflicted by competitors or supposedly independent third
parties, such as judges and regulators.
• The second low-cost strategy is the use of real assets with a value which would
deteriorate when clients are lost or the provider eventually has to abandon the
activity completely. These assets are technically known as "specific". In audit-
ing, the main such asset is the accumulated information on clients, their activities
and markets in which they operate. These specific assets can act as safeguard or
hamper quality. Which role they play depends on audit firm size and the degree
of client diversification. In a firm with a single client the threat of losing such
assets, which are to a large extent "client specific" (since they can only be used
to audit that client) may make the auditor more disposed to compromise his in-
dependence. In a firm with a diverse client base on the other hand, specific
assets act as a safeguard. The reason is that if its independence vis-a-vis one cli-
ent is compromised the auditor retains the assets specific to the particular client
but endangers the assets specific to other clients. If a lack of independence is
discovered other clients would tend to change auditor or penalize him in some
other way. This analysis of specific assets is of considerable importance to
regulation because, by showing the importance of client diversification for audit
quality, it provides the basis for choosing which variables to regulate and how to
do this in a market facilitating way.
• Finally, reputation or other safeguards used in one market can facilitate con-
tracting in other markets. Audit firms have used this strategy to sell non-audit
services which tend to be problematical to contract as a result of informational
asymmetry in relation to their quality. For this reason, auditors are natural pro-
viders of such services since they can ensure their quality by means of the
reputational assets which they have worked to build up in providing reliable
auditing services. The fact that such contractual resources can be used in a more
intensive and efficient manner means that it is inadvisable to adopt rules making
it more difficult to provide services. These contractual economies of scope flow
both ways. Thus, not only new non-audit services benefit from the reputation
acquired in providing audit services but the effect has been and is mutual: repu-
tation and other safeguards are utilized more intensively by broadening the set of
activities carried out by the firm. As a consequence, the quality of audit services
is also being ensured at a lower cost, or a better quality is provided when incur-
ring the same level of safeguarding costs.
b) High-cost safeguards. Other more costly strategies for assuring performance

are also possible. Firstly, advertising and other types of marketing expenses gener-
ate a "brand-name capital" which serves as a guarantee, even when it is not directly
informative: they constitute intangible assets whose rapid volatility as a conse-
quence of defective performance gives them a substantial advantage in effectively
acting as a quality safeguard. Secondly, regulatory authorities can also replicate
reputational incentives by raising the barriers to engaging in the activity. After that,
producers will have more incentive to avoid expulsion as a result of bad perform-
ance. Both possibilities are costly from the social point of view and some are
inapplicable to auditing. Advertising consumes actual resources and is in fact pro-
hibited under the regulations of many countries. There may be no reason to justify
this prohibition, however. On the contrary, some empirical evidence indicates that
the prohibition damages both competition and quality. Generating incentives by
entry barriers, on the other hand, poses all the problems inherent in a restrictive
regulation of professional services. Essentially it is of doubtful effectiveness and
involves a serious risk of monopoly and regulatory capture. It thus generates direct
inefficiencies by raising prices and consequently restricting supply as well as indi-
rect inefficiencies by unleashing competition or "rent-seeking" processes when
potential entrants compete to secure monopoly rents.
8.2 The provision of non-audit services: Consequences

and regulation
The second part of the book applies this view of how auditing should be regulated
in accordance with the functioning of private quality safeguards to examine a par-
ticular problem: that of determining the most suitable rules covering the provision
by auditors of non-audit services, often to their own audit clients. To this end, the
impact that the joint provision of audit and non-audit services has on costs, auditor
independence and market competition needs to be considered before reaching nor-
mative conclusions. It is argued that the provision of such services reduces overall
costs, raises the technical quality of auditing, enhances competition and need not
prejudice auditor independence or the quality of these non-audit services. The basic
recommendation derived from this analysis is to leave market forces free to fmd an
efficient equilibrium in terms of both the mix of services and the right combination
of quality safeguards. The role of regulation then becomes limited to that of
strengthening the safeguards provided by the market. In accordance with this role,
a rule of mandatory disclosure of client diversification could be contemplated in
order to facilitate the task of the market with regard to achieving the optimal degree
of auditor independence.
8.2. 1 Consequences of the supply of non-audit services
8.2.1.1. The reduction in total cost: economies of scope
Auditing has rarely if ever been the only product of auditing fIrms. On the contrary,
auditors have provided their clients with many types of service since the times when
external auditing began in the nineteenth century up to the present day. These
services are founded on economies of joint production or scope which outweigh
possible diseconomies or negative effects and are of two types, productive and
contractual:
a) Productive or, more precisely, transformation economies arise from the fact
that many of the major processes in auditing and these services are the same. For
example, the information required to evaluate and improve an internal control sys-
tem is to a large extent identical (this information flow is frequently referred to as
knowledge spillovers). Auditors are therefore in the best possible position to advise
on renewing such systems. These possibilities increase with the scope of the audit
and the complexity of the organizations audited since more specialized resources
are then required which often mean that a wide range of services need to be pro-
vided to make efflcient use of them.
b) Contractual economies (frequently referred to as "one-stop shopping") in tum
arise because it is costly to contract the services and it therefore becomes worth-
while making use of the safeguards (brand-name, reputation, conduct rules, control
systems amongst professionals, client confIdence) already developed when con-
tracting and ensuring quality in auditing, thereby reducing the total cost of
providing such services. Audit fIrms with a good reputation have an advantage in
expanding their activities into such services because they are in a better position to
provide clients with quality safeguards. However, these contractual economies of
scope may and will in general flow both ways, from audit to non-audit services and
vice versa. Thus, the contractual safeguarding of auditing will also be easier for
fIrms providing non-audit services than otherwise.
Both types of economies of scope contribute to enhancing the "technical com-

petence" of auditors, meaning by technical competence their ability to detect
shortcomings in accounts. In particular, the provision of non-audit services helps
the auditor to arrive at an informed professional judgement, using his knowledge to
evaluate the situation of the business beyond that indicated by narrowly defIned or
excessively mechanical book-keeping indicators. Secondly, non-audit services
make it possible for auditing fIrms to contract directly and make efflcient use of the
specialized experts and resources required for quality auditing in specialist sectors
at a lower cost, thus improving and extending their professional judgement. For
example, in order to audit a highly regulated undertaking properly at least one ex-
pert in that sector will be required. If consultancy services are provided to such
clients it will be more practicable to contract and make efficient use of such experts.
Economies of scope of the first type, associated with the joint use of information
to provide different services to the same single client, were very important in the
past and probably remain so amongst the smaller firms. In the large multidisciplin-
ary organizations, the advantage of joint provision of a wide range of services
seems increasingly to reside, however, in all types of contractual advantages, and
perhaps also in scale and network economies in the production of knowledge, which
is not specific to a single client, as in the case of the knowledge spillovers, but
which can be utilized on a general basis. This latter is particularly the case with the
investments which are necessary to arrange and control a network of offices pro-
viding services of uniform quality on a global basis: training centers, programs, data
bases, quality control systems and management and organizational systems gener-
ally. The relative importance of the different kinds of economies can be grasped
from the fact that nowadays different teams or even divisions and companies are
responsible within each multidisciplinary organization for providing each type of
service. This suggests that most client-specific economies of scope are contractual
in nature, because knowledge spillovers are not viable when information hardly
flows across such teams.
Most observers accept these economies of scope. Furthermore, it is likely that
they are becoming increasingly extensive, specially those of a contractual nature, as
the scope of auditing increases and businesses become more complex and their
activities more global. The debate centers rather on whether, in addition to these
positive effects, the joint provision of audit and non-audit services has negative
effects. The principal arguments relate to whether or not they are prejudicial to
auditor independence as well as to competition in audit and service markets. Let us
discuss these more controversial aspects of the problem.
8.2.1.2 Absence of harm to auditor independence
The available empirical evidence does not support the contention that auditor inde-
pendence is harmed by providing non-audit services, even to audit clients. No
causal relationship can be detected between providing services and a lack of inde-
pendence from the studies carried out in the United States since the 1970s into cases
of improper auditing or those based on indirect indicators. Providing such services
does, however, appear to result in a public perception that independence could be
harmed, especially in the view of market participants who are either poorly in-
formed or have some interest in the issue.
This empirical evidence is consistent with the theoretical analysis of the effect on
independence of providing such services. In essence, they result in an increase in
client- and firm-specific assets. The latter always have a positive effect on inde-
pendence and, in general, on quality, whilst the effect of client-specific assets
depends on the degree of client diversification, When this is sufficiently high, they
also encourage independence. It should not be inferred from this, however, that the
overall effect of providing such services is negative in the case of auditors with little
client diversification for the reason that although they increase the quasi-rents asso-
ciated with the clients, so also do firm-specific assets.
The above conclusion is reached from a static analysis which does not take ac-
count of the action and policies implemented by firms and the rules adopted by their
professional associations to increase both independence in fact and independence in
appearance. A consideration of such action and self-regulation reinforces the ar-
gument that providing services encourages independence. Firstly, firms' rules and
organizational patterns range between a radical abstention from carrying out man-
agement or decision-making functions through implementation of service provision
by firms which are connected to them contractually but have separate management
and assets. This is a suitable formula for achieving the economies of scale and
scope available in certain types of joint activities (basically investment in training,
reputation and quality control) and in achieving certain product attributes (global-
ization) whilst at the same time preserving the advantages of different firms
specializing in different types of service. These measures in the specific area of
non-audit services are implemented in addition to the many incentive and control
devices which are applied on a general basis in order to motivate individuals and
divisional units towards compliance with firms' quality standards, including inde-
pendence. Examples of these are basing the remuneration of partners on
performance variables which encourage them to take a broad perspective, including
the global results of the firm; or having personnel from one area inspecting the work
of another geographic area. Secondly, the rules adopted by professional associa-
tions are aimed at controlling potentially conflictive situations and in particular
preserving a public perception of independence. Specific rules adapted to their
characteristics are applied for this purpose in the field of professional services to
avoid possible confusion as to the professional standards applicable to them. For
example, the AICPA mandates serving the client's interests, establishing an under-
standing with him in which the responsibilities of the parties and the nature, scope
and limitations of the services to be provided are set out, and notifying him of pos-
sible conflicts of interest and significant reservations.
8.2.1.3. Lack of harm to competition in the audit market
It is also unlikely that non-audit services harm competition in the audit market.
They may certainly engender price structures that are prone to be confused, how-
ever. Evaluation of these issues must thus take account of the economic rationale of
introductory pricing and inter-temporal competition. In this respect, standard In-
dustrial Organization analysis shows that cost savings obtained from the joint
provision of the two sets of services will be passed on to clients as a price reduction
in both market (auditing and services) and at each stage (in the fIrst or subsequent
contracts) depending on the competitive circumstances governing each market and
each contractual stage. Using introductory pricing in auditing or using auditing as a
loss leader then appears as the simple spontaneous outcome of intertemporal com-
petition when there are learning and rotation costs.
In particular, competition for future profIts ("quasi-rents") explains the use of intro-
ductory pricing by audit fIrms. These future profIts are generated as the result of
the start-up or learning costs of initial audits and the switching costs which all cli-
ents must incur when changing auditor. The economies resulting from combining
auditing and other services are an additional source of such profits or quasi-rents.
All these factors mean that continuity in their relationship is advantageous to both
clients and auditors. If the audit market is competitive, auditors compete to lower
the prices of their initial work knowing that they can make up the probable initial
loss from profIts from future work. If the initial work tends most frequently to be
auditing and there are economies of scope a larger discount will also be seen in the
initial prices of auditing when the auditor is equipped to provide other services.
Moreover, this assessment is positive, independently of the judgement merited by
the quasi-rents, since the latter are the cause rather than the result of introductory
pricing. Readers who have recently acquired a mobile phone will probably have
benefited from this type of competition-telephone companies generally give the
telephones away knowing that customers will be tied to them, generating substantial
profIts in the future. It is this tie together with a certain degree of competition
which results in the initial discount. The tie would still exist however even if the
connection price had not been discounted, the latter merely being a consequence.
This intertemporal competitive process is optimal from the public point of view.
A better understanding of the phenomenon will help to avoid the still common
mistake of considering the practice of introductory pricing at less than the cost of
the initial audit, or "lowballing", as uncompetitive or prejudicial to independence.
This error sometimes leads to proposing or adopting rules aimed at preventing in-
troductory pricing. Such rules are self-defeating however, in terms of both
independence and competition. They fIrstly raise the total volume of quasi-rents
associated with the client since they increase the cost of replacing the incumbent
auditor (the aspiring auditor can no longer offer a discount on the initial audit) and
therefore also the quasi-rents earned by the current auditor. Furthermore, prohibi-
tion of lowballing is prejudicial to competitive conditions-it would be equivalent
to preventing mobile telephone companies from giving away telephones to new
subscribers. By impeding potential auditors from discounting the price of the initial
audit the rotation cost is raised and the incumbent auditor can charge a higher price.
Once discounting is prohibited, the price of the initial audit rises to equal its cost
whilst the prices in subsequent years continue at the level which prevents competi-
tors from coming in, a level which rises moreover because of the higher rotation
costs. These reasons support the idea that professional concern for lowballing
probably has more to do with preserving and increasing market power than with the
alleged objective of protecting independence.
8.2.1.4 Increased competition in the market for non-audit services
Some service providers (particularly bar associations in Europe) have opposed al-
lowing audit firms or firms contractually connected with them from entering into
the field of their own services. They contend that such entry harms competition and
threatens the quality of such services. An examination of the problem reveals that
both contentions are probably groundless, however. Firstly there is certainly a
positive effect on competition. Secondly, if the effect on the quality of services was
negative, clients seem quite competent to realize this, and therefore the action of the
free market itself would make a normative prohibition unnecessary. This lack of
merit in the criticism, together with amount of resources devoted to expounding it,
could lead to thinking that it has more to do with the interests of established profes-
sionals in protecting their market power than in defending professional quality
standards.
a) Firstly, entry of new suppliers can only increase competition, especially when
these new providers render services with greater added value at less cost (at least of
a more certain quality and over a greater geographical area). In this respect the fact
that criticism relates to the strategies and operations of firms whose service quality
is not questioned gives rise to the suspicion that the critics' concern is not really
based on fear of falling quality but increasing competition.
b) Secondly, if quality is questionable clients will be the first to reject the multi-
ple services offered by a single firm or by connected firms. Since the situation is
readily transparent it is also easy for clients to evaluate the possible consequences of
such connections. As a result, the market would provide effective corrective incen-
tives if service quality was really endangered, whether because of a conflict of
interest or any other reason. A decision as to the degree of specialization of firms
should therefore be taken in the area of the parties' freedom of contract.
c) The previous points lead then to think that proposals restricting freedom to
provide services in this field might well be the result of the private interests of some
service providers in reducing competition. The fact that prohibitive rules in some
European countries are generally ineffective also explains how interest in this type
of restriction can extend to those fIrms which, as a result of their lesser reputation,
would be in a position to circumvent the rule at low cost. Furthermore, in those
countries where the organization of the legal profession is more outdated, the entry
of established law fIrms, whether or not connected with multidisciplinary organiza-
tions, is doubly useful since it can provide a powerful spur towards much-needed
modernization in the profession. Encouraging signs can already be seen in this
fIeld, particularly in the development of larger professional fIrms with international
links which for both reasons are in a better position to meet the demand from in-
creasingly international businesses.
8.2.2. The regulation of non-audit services
Considering the previous analysis of the consequences of non-audit services,their

prohibition and regulatory constraint do not make much sense. Prohibitive rules of
a general nature-such as those in force in Belgium, France and Italy-are inadvis-
able from both the point of view of the audit market, where they are harmful to
effIciency and quality, and in relation to competition in service markets. Further-
more, they are said to suffer from serious enforcement problems, which might be
very harmful if high quality fIrms with good reputations cannot afford to get around
the rule and are therefore placed at a disadvantage. In addition, restrictive rules of a
particular nature, whether prohibiting certain types of service or making them sub-
ject to specifIc restrictions, are less negative but also suffer from substantial
problems. Above all such rules necessitate ample and detailed knowledge on the
part of those drawing them up. They are perhaps only justifIed therefore when they
confIrm good practice in the fIeld and discourage the appearance of marginal pro-
viders.
8.2.2.1 The market as a producer of knowledge
The legislator would consequently do well to refrain from introducing detailed

regulations. In particular, he should be more concerned for re-defIning audit market
conditions so that the market provides an incentive to fIrms and professional asso-
ciations to act as self-regulators, as has occurred with relative effectiveness in the
most advanced Anglo-Saxon countries in this respect. A guiding principle of regu-
lation should then be to provide enough leeway for audit fIrms, self-regulatory
bodies and audit clients to ascertain through competitive market interaction both the
effIcient mix of services and the corresponding quality safeguards, adjusting for the
costs and benefIts of each possibility. The main reason to entrust this discovery
function to the free market is that the incentives and the ability of market partici-
pants seem more capable of achieving it than regulators, as the latter lack both the
knowledge and the incentives to define the efficient framework. Their lack of
knowledge is inherent to their position as neither producers nor clients. The defec-
tive incentives are rooted in their proclivity, firstly, to exaggerate external effects
and consequently require higher quality and quantity than optimum, both of which
have no cost for them. They are also bound to be captured by private interests alien
to the audit market, as shown by the variety of prohibitive or restraining regulations
and the previous analysis on the competition in the market for non-audit services.
Letting the market drive the evolution of the industry does not avoid a need for
regulation. Rules are still needed for facilitating the smooth functioning and speedy
adjustment of the market. Regulators should then concentrate on promoting and
facilitating competition in order to enhance market incentives, including by means
of policies aimed at increasing informational transparency and facilitating the crea-
tion of private quality safeguards. Specific measures would involve both
introducing new standards (for example as to disclosing client concentration) and
eliminating some current restrictions (such as those relating to advertising and un-
solicited offering of services which have been demonstrated to be beneficial to
quality in the United States). In this context, it would also be appropriate to review
the conditions governing demand for auditing, matching the legal requirement for
statutory auditing to those cases in which the audit actually reduces contractual
costs (these taken in their widest sense to include all types of external effects).
8.2.2.2 Client diversification as a regulatory objective
Among the instruments potentially useful for a regulatory intervention of a facili-

tating character, fee income diversification shows good properties. Both common
sense and scientific analysis of the problem coincide in that client diversification is
an essential element in private safeguards of audit quality and, specifically, in
auditor independence. When an auditor's fee income is not concentrated in the
client whose accounts are being examined the auditor will be less dependent on that
client. The reason is that the anticipated cost of dependent conduct (motivated by
possible loss of clientele) will be very high compared with the cost of independent
conduct (associated with the loss of the client in question). In more technical terms,
in the auditor-client relationship both parties are investing in "specific assets" (par-
ticularly knowledge) which are only of value if the relationship continues. Because
of this, they have an interest in continuity and a potential risk to independence
emerges. This potential risk only materializes, however, when their revenue is
concentrated and disappears when it is diversified. To auditors with many clients
independent behavior can endanger the assets specific to one of them but it main-
tains or even increases the value of the assets which are specific to the remainder (as
well as those specific to the firm, principally its reputation).
As a result of this independence-encouraging effect, client diversification is in
general a desirable objective of audit regulation. As such it offers two further ad-
vantages. Firstly it bolsters use of a type of low-cost safeguard thanks to the

productive nature of the specific assets which add their directly productive func-
tionality to their capacity to generate contractual safeguards as a free by-product. It
is also an objective which is relatively easy to supervise and instrument, either by
the regulator or by the market. Furthermore it is also advisable in the particular
field of non-audit services since to a certain extent the effect of these services on
independence depends on the degree of diversification.
The possibility of substituting across quality assurance safeguards and developing
innovative safeguards in general makes it nonetheless inadvisable to adopt rigid
diversification rules which could restrict firms' activities unnecessarily. The reason
is that the optimum level of auditor diversification depends on the characteristics of
the auditor himself as well as those of his clients. A single diversification rule
would force a standard level, however, which would be too high in some cases and
too low in others. The diversifying objective should, on the contrary, be compatible
with freedom for auditors to work with different degrees of diversification and other
quality safeguards. This way, market participants are not prevented from seeking
an efficient substitute from amongst the different types of quality assurance safe-
guards. It would suffice to establish a standard of disclosure referring to one or
more indicators providing information as to diversification. In other words, diversi-
fication should be an objective of the rule but the rule should leave it to the market
to decide on a suitable degree of diversification and ancillary safeguards along with
the possibility of replacing it by other guarantee mechanisms.
8.2.2.3 Costs and benefits of mandatory disclosure
The limitations of direct regulation of diversification could make it then advisable

to use an indirect strategy to stimulate it and at the same time give sufficient flexi-
bility to firms and their clients to locate themselves at optimum levels in each
particular situation. This indirect strategy would involve improving transparency in
the audit market which will facilitate the operation of private mechanisms to safe-
guard quality and strengthen incentives to meet professional obligations, in
particular the need for independence. In this context, the provision of non-audit
services might make it all the more necessary for each audit firm to make specific
disclosure of its degree of client diversification since it is difficult for the market to
estimate the extent of non-audit work or even its existence. If auditors only provide
audit services it would be relatively easy to gain an idea of their client diversifica-
tion. A substantial and positive correlation can be expected between assets specific
to each client and audit fees. The latter in turn will depend on the size of each cli-
ent. Since the size of clients and firms is relatively easy to ascertain, the market
would then be in a position to estimate the diversification of each audit firm's cli-
entele. This is not so easy, on the contrary, when non-audit services are provided,
because projects vary greatly in size and their size is not as well correlated to the
size of the client as audit work.
Furthermore, mandatory disclosure may be justified when voluntary disclosure is
hindered by collective action problems or third-party effects. This is the case when,
despite the private cost of disclosure being smaller than the private benefit, individ-
ual firms may be better off not disclosing if some other firms do disclose, with the
fmal outcome that no firm discloses. Likewise, individual disclosure generates
positive externalities when the information provides knowledge on the industry or,
more pertinent to the audit case, is used in comparisons across firms. In addition,
disclosure needs ex post verifiability and a common language to function economi-
cally, both requiring some centralized action which can be provided by regulation in
a fast, even if frequently inadequate, manner. Confusion risks and conflicts within
networks of providers add to the constraints on a speedy development of voluntary
disclosure.
From this perspective, regulation can play a useful role by eliminating some of
the barriers hindering voluntary disclosure. In particular, mandatory disclosure
should be seen and structured as a solution to free-riding problems among potential
disclosers. The regulation could also define diversification scales in order to reduce
the risk of disclosing information which might be sensitive to the affected client.
Moreover, the regulator himself, in promoting voluntary disclosure, would elimi-
nate the risk of confusion borne by the initial discloser. The regulatory agency can
also act as an ex post verifying agent and, by emphasizing the substitutability of
safeguards, it might even reduce the risk of misinterpretation by market participants.
The choice of a disclosure rule among the multiple possibilities available, some
of which have already been adopted in some countries, must take into account its
relative effectiveness and the costs it may generate. Effectiveness depends on
whether it provides the market with information useful to correctly assess economic
incentives favoring independence. The main cost comes from a potential damage to
competition:
a) Benefits of disclosure. The benefits of disclosure arise when the market can
gain a more accurate idea of an auditor's incentive to be independent. For this rea-
son public disclosure is desirable in a standard form (to reduce processing costs).
The content of the disclosure should ideally relate to the economic variables on
which incentives depend (the value of specific assets). As this figure is not known,
it is preferable to use other more objective indicators such as fee income. With
respect to the computational scope of the rule (for both firms and their clients) it
would be appropriate to use the decision unit definitions which are included stan-
dard in most Corporation Laws. In addition, the relevant figure is the concentration
of fee income and not its composition. For this reason it would be worthless to
disclose data on the relative weight of auditing and service fees as this relationship
does not affect independence.
b) Costs of disclosure. The main cost of mandatory disclosure has an indirect
character as it is related to the problems that some rules could bring to competition.
In particular, full disclosure of individual fees, in force in some countries, probably
makes competitive strategies more difficult and less profitable for audit firms, re-
veals sensitive information about their clients and, in some cases, facilitates the
monitoring of collusive agreements in the audit market. In addition, the direct cost
of gathering and publishing data will increase with the amount of information re-
quired. Both direct and . indirect costs thus make it advisable to minimize the
amount of information required. Moreover, the US experience with the SEC disclo-
sure requirement reveals the possibility that some disclosure policies may engender
hidden costs of a more subtle character. From 1978 to 1982 SEC registrants had to
disclose the ratios of total and material individual non-audit services fees to audit
fees. This requirement was rescinded in 1982, however, because the SEC itself
came to believe that it was not sufficiently useful for investors. Audit firms claimed
that it had curtailed non-audit services because parties perceived that the SEC was
deprecating their benefits and that the SEC might question the independence of the
auditor considering only the disclosed ratios fees. The controversy shows that some
disclosures or, more generally, naIve regulatory activism on the matter risk over-
emphasizing the negative effects of non-audit services and may lead clients to
acquire less non-audit services from the auditor than optimum.
Taking these costs and benefits into account a rule making it obligatory to dis-
close maximum concentration (the contribution of the auditor's biggest client to his
total fee income) is preferable to a general disclosure of his fees. Awareness of the
maximum concentration figure will provide valuable information for all those di-
rectly or indirectly involved in the market. Clients of the audit firm will thus have
additional information to gauge the quality of the auditor, more specifically as to
some of the incentives which affect his decisions. In the same way, those using the
clients' accounts will in tum have additional information to assess their quality.
The information of those using the accounts of all of the auditor's clients will also
indirectly be improved.
In any case, given that mandatory disclosure surely involves hidden and unfore-
seen costs and benefits which differ across different rules, its introduction and their
adequacy are empirical questions which can never be answered fully. Arguments
are surely useful when designing disclosure policies to avoid mistakes. However,
careful attention to how the market reacts to the chosen rules and flexibility in
eventually removing or modifying them is also essential.
REFERENCES
AAKER, D. A. (1991), Managing Brand Equity, Capitalizing on the Value of a Brand Name
(New York: The Free Press).
ABDEL-KHALIK, A. R. (1990), "The Jointness of Audit Fees and Demand for MAS: A
Self-Selection Analysis," Contemporary Accounting Research, 6 (2), 295-322.
AICPA, American Institute of Certified Public Accountants (1969), "Final Report of the Ad
Hoc Committee on Independence," The Journal of Accountancy (December), 51-56.
AICPA, American Institute of Certified Public Accountants (1996), Statement on Consulting
Services (New York: AICPA).
ANDERSON, D., 1. R. FRANCIS and D. 1. STOKES (1993), "Auditing, Directorships and
the Demand for Monitoring," Journal ofAccounting and Public Policy, 12 (4), 353-375.
ANTLE, R. (1984), "Auditor Independence," Journal ofAccounting Research, 22 (1), 1-20.
ANTLE, R., and 1. DEMSKI (1991), "Contracting Frictions, Regulation and the Structure of
CPA Firms," Journal ofAccounting Research, 29 (supplement), 1-24.
ANTLE, R., B. NALEBUFF and S. BAIMAN (1991), "Conservatism and Auditor-Client
Negotiations Discussion," Journal ofAccounting Research, 29 (supplement), 31-59.
APB, The Auditing Practices Board (1992), The Future Development of Auditing (London:
CCAB Ltd.).
APB, The Auditing Practices Board (1994), The Audit Agenda (London: CCAB Ltd.).
ARRUNADA, B. (1995), Amilisis economico del notariado (Madrid: CON).
ARRuNADA, B. (1996), "The Economics of Notaries," European Journal of Law and
Economics, 3 (1),5-37.
ARRUNADA, B. (1998), "Oestaltung und Regulierung von Freiberuflichen Dienstleistun-
gen: Ein Allgemeiner Handlungsleifaden," in H. Herrmann and 1. Backhaus (eds.),
Staatlich Gebundene Freiberufe im Wandel: Rechtliche und Okonomische Aspekte aus
Wissenschaft und Praxis (Baden-Baden: Nomos Verlagsgesellschaft), 115-135.
ARRUNADA, B. (1999), "The Organisation of Professional Services: A Policy-Oriented
Analysis," in 1. Backhaus (ed.), Companion to Law and Economics (Cheltenham, UK,
and Northampton, USA: Elgar, forthcoming).
177
178 References
ARRuNADA, 8., and C. PAZ-ARES (1997), "Mandatory Rotation of Company Auditors: A

Critical Examination," International Review of Law and Economies, 17 (1), 31-61.
BAIN, 1. (1949), "A Note on Pricing in Monopoly and Oligopoly," American Economic
Review, 39, 448-464.
BALACHANDRAN, B. V., and N. J. NAGARAJAN (1991), "Imperfect Information, Insur-
ance and Auditors' Legal Liability," Contemporary Accounting Research, (supplement),
31-54.
BALACHANDRAN, B. V., and R. RAMAKRISHNAN (1987), "A Theory of Audit Partner-
ships: Audit Firm Size and Fees," Journal ofAccounting Research, 25 (I), 111-126.
BALVERS, R. 1., B. McDONALD and R. E. MILLER (1988), "Underpricing of New Issues
and the Choice of Auditor as a Signal of Investment Banker Reputation," The Accounting
Review, 63 (4), 605-622.
BARKESS, L. and R. SIMNETT (1994), "The Provision of Other Services by Auditors:
Independence and Pricing Issues," Accounting and Business Research, 24 (94), 99-108.
BEATIlE, V., R. BRANDT and S. FEARN LEY (1996), "Consulting? More like Compli-
ance," Accountancy (IntI. ed.), 118 (1.239), 95-96.
BEATTY, R. (1989), "Auditor Reputation and the Pricing of Initial Public Offerings," The
Accounting Review, 64 (October), 693-705.
BECK, P. 1., T. 1. FRECKA and I. SOLOMON (I 988a), "A Model of the Market for MAS
and Audit Services: Knowledge Spillovers and Auditor-Auditee Bonding," Journal of
Accounting Literature, 7, 50-64.
BECK, P. 1., T. 1. FRECKA and I. SOLOMON (l988b), "An Empirical Analysis of the
Relationship Between MAS Involvement and Auditor Tenure: Implications for Auditor
Independence," Journal ofAccounting Literature, 7, 65-84.
BECKER, G. S. (1964), Human Capital: A Theoretical and Empirical Analysis with Special
Reference to Education (New York: NBER-Columbia University Press).
BECKER, G. S., and G. 1. STIGLER (1974), "Law Enforcement, Malfeasance, and Com-
pensation of Enforcers," Journal of Legal Studies, 3 (January), 1-18.
BENOIT, E. (1987), "A Gentleman's Game," Financial World, 156 (19), 18-19.
BENSTON, G. 1. (1973), "Required Disclosure and the Stock Market: An Evaluation of the
Securities Exchange Act of 1934," American Economic Review, 63, 132-155.
BENSTON, G. 1. (1979-1980), "The Market for Public Accounting Services: Demand, Sup-
ply and Regulation," Accounting Journal, 2 (1),2-46.
BERG, 1. E. (1991), "The Impact of MAS on Auditors' Independence: An Experimental
Markets Study: Discussion," Journal ofAccounting Research, 29 (supplement), 99-106.
BERTON, L. (1989), "SEC Is Studying Independence of Firms Doing Consulting and Ac-
counting Work," The Wall Street Journal (November II).
BERTON, L. (1996), "Accountants Expand Scope of Audit Work," The Wall Street Journal
(June 17).
BOOTH, 1., and R. SMITH (1986), "Capital Raising Underwriting and the Certification
Hypothesis," Journal of Financial Economics, 15,261-281.
BRADBURY, M. E. (1990), "The Incentives for Voluntary Audit Committee Formation,"
Journal ofAccounting and Public Policy, 9 (I), 19-36.
BRICKLEY, 1. A., and F. H. DARK (1987), "The Choice of Organizational Form: The Case
of Franchising," Journal of Financial Economics, 18,401-420.
References 179
BRILOFF, A. 1. (1966), "Old Myths and New Realities in Accounting," The Accounting
Review, 41 (July), 284-295.
BRILOFF, A. 1. (1976), More Debits than Credits (New York: Harper and Row).
BRILOFF, A. J. (1994), "Our Profession's 'Jurassic Park' ," CPA Journal, 64 (8), 26-31.
BUIJINK, W., S. 1. MAIJOOR, R. MEUWISSEN and A. van WITIELOOSTUIJN (1996),
The Role, Position and Liability of the Statutory Auditor within the European Union, Fi-
nal report of a study commissioned by DG XV of the European Commission
(Luxembourg: Office for Official Publications of the European Communities).
CADBURY COMMITIEE, Committee on the Financial Aspects of Corporate Governance
(1992), The Financial Aspects of Corporate Governance (London: Burger Science
Press/Gee), December.
CAMPBELL, D. R., and L. M. PARKER (1992), "SEC Communications to the Independent
Auditors: An Analysis of Enforcement Actions," Journal of Accounting and Public Pol-
icy, 11 (4),297-330.
CARCELLO, 1. V., and Z. V. PALMROSE (1994), "Auditor Litigation and Modified Re-
porting on Bankrupt Clients," Journal ofAccounting Research, 32 (supplement), 1-30.
CARDADOR, G. (I 998a), "Arthur, Ernst, Deloitte y Peat se disputan las firmas latinoameri-
canas de Coopers," Expansion (May 27), 6.
CARDADOR, G. (1998b), "Los socios de Coopers paralizan su fusion con Price," Expansion
(May 11),3.
CARDADOR, G. and 1. M. del CANO (1997), "Pedro Navarro sera el responsable de Finan-
zas y Seguros de la nueva Andersen Consulting Worldwide," Expansion (December 22),
3.
CARLTON, D. W., and 1. W. PERLOFF (1994), Modern Industrial Organization 2nd ed.
(New York: Harper Collins).
CARMICHAEL, D. R., and R. 1. SWIERINGA (1968), "The Compatibility of Auditing
Independence and Management Services-An Identification of the Issues," The Ac-
counting Review, 43 (4), 697-705.
CARPENTER, C. G., and R. H. STRAWSER (1971), "Displacement of Auditors When
Clients Go Public," Journal ofAccountancy, 131 (June), 55-58.
CEPS, Centre for European Policy Studies (1995), "Corporate Governance in Europe," Re-
port of a CEPS Working Party (Brussels: CEPS), June.
CHOW, C. W. (1982), "The Demand for External Auditing: Size, Debt and Ownership In-
fluences," The Accounting Review, 57 (2), 272-291.
CHOW, C. W. (1983), "The Impacts of Accounting Regulation on Bondholder and Share-
holder Wealth: The Case of the Securities Acts," The Accounting Review, 58 (3), 485-
520.
CNMV (Comision Nacional del Mercado de Val ores) , Comision Especial para el estudio de
un Codigo Etico de los Consejos de administracion de las sociedades (1998), EI gobierno
de las sociedades cotizadas, Madrid, February 26.
COHEN COMMISSION (1978), The Commission on Auditors' Responsibilities: Report,
Conclusions and Recommendations (An Independent Commission Established by the
AICPA. Chairman M. F. Cohen), New York.
CORLESS, 1. C., and L. M. PARKER (1987), "The Impact of MAS on Auditor Independ-
ence: An Experiment," Accounting Horizons, 1 (September), 25-29.
180 References
CORPORATE GOVERNANCE COMMITTEE (1998), "Corporate Governance Principles-

A Japanese View," Final Report of the Corporate Governance Committee, Corporate
Governance Forum of Japan, May 26.
CRASWELL, A. T. (1998), "Threats to Auditor Independence: the Case of Non-Audit Serv-
ices," paper presented to the 21 st Congress of the European Accounting Association,
Antwerp (April 6-9).
CRASWELL, A. T., 1. R. FRANCIS and S. L. TAYLOR (1995), "Auditor Brand Name
Reputations and Industry Specialization," Journal ofAccounting and Economics, 20, 297-
322.
CRASWELL, A. T., 1. R. FRANCIS, and B. SNEDDON (1997), "Audit Quality Uncertainty
and Pricing of Initial Engagements," University of Sidney, mimeo (March).
DARBY, M. R., and E. KARNI (1973), "Free Competition and the Optimal Amount of
Fraud," Journal of Law and Economics, 16 (April), 67-88.
DA VIS, L. R., and D. T. SIMON (1992), "The Impact of SEC Disciplinary Actions on Audit
Fees," Auditing: A Journal of Practice and Theory, 11 (1),58-68.
DAVIS, L., D. RICCHIUTE and G. M. TROMPETER (1993), "Audit Effort, Audit Fees,
and the Provision of Nonaudit Services to Audit Clients," The Accounting Review, 68 (1),
135-150.
DeANGELO, L. E. (198Ia), "Auditor Independence, 'Low Balling', and Disclosure Regula-
tion," Journal ofAccounting and Economics, 3, 113-127.
DeANGELO, L. E. (1981 b), "Auditor Size and Audit Quality," Journal of Accounting and
Economics, 3, 183-199.
DeBERG, C. L., S. E. KAPLAN and K. PANY (1991), "An Examination of Some Relation-
ships Between Non-Audit Services and Auditor Change," Accounting Horizons,S (I), 17-
28.
DEFOND, M. (1992), "The Association Between Changes in Client Firm Agency Costs and
Auditor Switching," Auditing: A Journal of Practice and Theory, 11, 16-31.
DEJONG, D. V. (1985), "Class-Action Privileges and Contingent Legal Fees: Investor and
Lawyer Incentives to Litigate and the Effect on Audit Quality," Journal of Accounting
and Public Policy, 175-200.
DEMSETZ, H. (1968), "Why Regulate Utilities?," The Journal of Law Economics, 11 (55),
55-65.
DEW ATRIPONT, M. (1994), "Discussion of' Auditor Professional Judgement: Implications
for Regulation and the Law'," Economic Policy, 19 (October), 344-346.
DNES, A. W. (1996), 'The Economic Analysis of Franchise Contracts," Journal of Institu-
tional and Theoretical Economics, 152 (2), 297-324.
DONABEDlAN, B. (1993), "Accounting Self-Regulation and the Enforcement of Profes-
sional Codes," Journal ofAccounting and Public Policy, 12 (2), 87-112.
DOPUCH, N., and R. R. KING (1991), "The Impact of MAS on Auditors' Independence: An
Experimental Markets Study," Journal ofAccounting Research, 29 (supplement), 60-106.
DOWNES, D., and R. HEINKEL (1982), "Signaling and the Valuation of Unseasoned New
Issues," Journal of Finance (March), 1-10.
DYE, R. A. (1986), "Proprietary and Nonproprietary Disclosures," Journal of Business, 59
(2),331-366.
DYE, R. A. (1991), "Informationally Motivated Auditor Replacement," Journal of Account-
ing and Economics, 14, 347-374.
References 181
DYE, R. A. (1993), "Auditing Standards, Legal Liability, and Auditor Wealth," Journal of
Political Economy, 101 (5), 887-914.
DYKXHOORN, H. 1., and K. E. SINNING (1981), "Wirtschaftspriifer Perceptions of Audi-
tor Independence," The Accounting Review, 56 (1), 97-107.
EASTERBROOK, F. H., and D. R. FISCHEL (1991), The Economic Structure of Corporate
Law (Cambridge, MA: Harvard University Press).
ETTREDGE, M., and R. GREENBERG (1990), "Determinants of Fee Cutting on Initial
Audit Engagements," Journal ofAccounting Research, 28 (1), 198-210.
EUROPEAN COMMISSION (1998), "The Statutory Audit in the European Union: the Way
Forward," communication adopted on April 29, Official Journal of the European Com-
munities, C 143 (May 8).
EUROPEAN CONTACT GROUP (1996), "Responding to Market Expectations: An Action
Plan to Reduce the Expectations Gap," July.
EZZAMEL, M., D. R. GWILLIAM and K. M. HOLLAND (1996), "Some Empirical Evi-
dence From Publicly Quoted UK Companies on the Relationship Between the Pricing of
Audit and non-Audit Services," Accounting and Business Research, 27 (1), 3-16.
EZZAMEL, M., D. R. GWILLIAM, and K. HOLLAND (1998), "The Relationship Between
Categories of Non-Audit Services and Audit Fees: Evidence From UK Companies," paper
presented to the 21 st Congress of the European Accounting Association, Antwerp (April
6-9).
FAMA, E. F., and M. C. JENSEN (1983a), "Separation of Ownership and Control," Journal
of Law and Economics, 26 (June), 301-325.
FAMA, E. F., and M. C. JENSEN (1983b), "Agency Problems and Residual Claims," Jour-
nal ofLaw and Economics, 26 (June), 327-349.
FEE, Federation des Experts Comptables Europeens (1995), Audit Independence and Objec-
tivity, Position Paper approved by the Council of FEE on 27 June 1995, Brussels.
FEE, Federation des Experts Comptables Europeens (1996), The Role, Position and Liability
of the Statutory Auditor in the European Union, January, Brussels.
FIRTH, M. (1980), "Perceptions of Auditor Independence and Official Ethical Guidelines,"
The Accounting Review, 55 (3),451-466.
FIRTH, M. (1981), "Auditor-Client Relationships and Their Impact on Bankers' Perceived
Lending Decisions," Accounting and Business Research (summer), 179-188.
FIRTH, M. (1997a), "The Provision of Non-Audit Services and the Pricing of Audit Fees,"
Journal of Business Finance and Accounting, 24 (3-4), 511-525.
FIRTH, M., and A. SMITH (1992), "Selection of Auditor Firms by Companies in the New
Issue Market," Applied Economics, 24 (2),247-255.
FISHMAN, M. 1., and K. M. HAGERTY (1990), "The Optimal Amount of Discretion to
Allow in Disclosure," Quarterly Journal of Economics, lOS (2), 427-444.
FRANCIS, 1. (1984), "The Effect of Audit Firm Size on Audit Prices: A Study of the Aus-
tralian Market," Journal ofAccounting and Economics, 6 (August), 133-151.
FRANCIS, 1. R., and E. R. WILSON (1988), "Auditor Changes: A Joint Test of Theories
Relating to Agency Costs and Auditor Differentiation," The Accounting Review, 63 (4),
663-682.
GALLINI, N., and N. LUTZ (1992), "Dual Distribution and Royalty Fees in Franchising,"
Journal of Law Economics, and Organization, 8 (3), 471-501.
182 References
GARCiA-BENAU, M. A., and C. HUMPHREY (1992), "Beyond the Audit Expectations

Gap: Learning from the Experiences of Britain and Spain," European Accounting Review,
1,303-331.
GA VER, 1., J., and K. M. GAVER (1995), "Simultaneous Estimation of the Demand and
Supply of Differentiated Audits," Review of Quantitative Finance and Accounting, 5 (I),
55-70.
GIGLER, F., and M. PENNO (1995), "Imperfect Competition in Audit Markets and its Ef-
fect on the Demand for Audit-Related Services," The Accounting Review, 70 (2), 317-
336.
GILSON, R., and R. H. KRAAKMAN (1984), "The Mechanisms of Market Efficiency,"
Virginia Law Review, 70, 549-644.
GLEZEN, G. W., and A. MILLAR (1985), "An Empirical Investigation of Stockholder
Reaction to Disclosures Required by ASR No. 250," Journal of Accounting Research, 23
(2), 859-870.
GOLDMAN, A., and B. BARLEV (1974), "The Auditor-Firm Conflict of Interests: Its Im-
plications for Independence," The Accounting Review (October), 707-718.
GONZALO ANGULO, J. A. (1995), "Presentaci6n: La auditoria, una profesi6n en la encfU-
cijada de los noventa," Revista Espanola de Financiacion y Contabilidad, 24 (84), 595-
629.
GREEN PAPER on the Role, the Position and the Liability of the Statutory Auditor within
the European Union (1996), OffiCial Journal of the European Communities, C 321 (Octo-
ber 28).
GROSSMAN, S. 1. (1981), "The Informational Role of Warranties and Private Disclosure
about Product Quality," Journal of Law and Economics, 24 (December), 461-483.
GROSSMAN, S. 1., and O. H. HART (1980), "Disclosure Laws and Takeover Bids," Jour-
/lal of Finance, 35 (2), 323-334.
GROSSMAN, S. 1., and O. HART (1986), "The Costs and Benefits of Ownership: A Theory
of Lateral and Vertical Integration," Journal of Political Economy, 94 (4), 691-719.
GROUT, P., I. JEWITT, C. PONG and G. WHITTINGTON (1994), "Auditor Professional
Judgement': Implications for Regulation and the Law," Economic Policy, 19 (October),
307-344 and 349-351.
HANSON, R. K., and 1. D. BAKER (1996), "Taming the Beast of Liability," National Pub-
lic Accountant, 41 (8), 17-19.
HARTLEY, R. V., and T. L. ROSS (1972), "MAS and Audit Independence: An Image
Problem," Journal of Accountancy, (November), 42-51.
HILL, J. W., M. B. METZGER and 1. W. SCHATZBERG (1993), "Auditing's Emerging
Legal Peril Under the National Surety Doctrine: A Program for Research," Accounting
Horizons, 7 (1), 12-28.
HOGAN, C. E. (1997), "Costs and Benefits of Audit Quality in the IPO Market: A Self-
Selection Analysis," The Accounting Review, 72 (1),67-86.
IF AC, International Federation of Accountants (1996), Code of Ethics for Professional Ac-
colllitants, July.
JACK, A. (1993), "Survey of World Taxation (3): Lawyers Muscle in-Tax Advisers Are
Thriving," The Financial Times (London ed.), February 18.
JARRELL, G. A. (1981), "The Economic Effects of Federal Regulations of the Market for
New Securities Issues," Journal of Law and Economics, 24 (December), 613-675.
References 183
JENSEN, M. c., and W. H. MECKLING (1976), "Theory of Firm: Managerial Behavior,

Agency Costs and Ownership Structure," Journal of Financial Economics, 3 (4), 305-
360.
JETER, D. c., and P. ERICKSON SHAW (1995), "Solicitation and Auditor Reporting Deci-
sions," The Accounting Review, 70 (2), 293-315.
JOHNSON, H. (1992), "Fair Franchising," Managerial Law, 34, 1-24.
JONES, E. (1981), Accountancy and the British Economy 1840-1980: The Evolution of Ernst
& Whinney (London: Batsford).
JOV ANOVIC, B. (1982), "Truthful Disclosure of Infonnation," Bell Journal of Economics,
13,36-44.
KANODIA, C, and A. MUKHERJI (1994), "Audit Pricing, Lowballing and Auditor Turn-
over: A Dynamic Analysis," The Accounting Review, 69 (4), 593-615.
KAUFMANN, P., and F. LAFONTAINE (1994), "Costs of Control: The Source of Eco-
nomic Rents for McDonald's Franchisees," Journal of Law and Economics, 37 (October),
417-453.
KELLY, 1. (I 997a), "Andersen: Partners Likely to Decide Against Split," Financial Times
(April 28).
KELLY, 1. (I 997b), "Flotation Plan: Andersen Partners Vote Against Split," Financial Times
(May 1).
KELLY, 1. (1997c), "Andersen: Board Looks to UK for New Chief," Financial Times (May
6).
KELLY, 1. (I 997d), "Andersen Worldwide: Candidate Rejected by Partners," Financial
Times (June 2).
KELLY, 1. (1997e), "Andersen Worldwide: Partners Fail to Agree on New Chief," Financial
Times (June 26).
KELLY, 1. (I 997t), "Andersen Consulting: A Reinvention on a Global Scale," Financial
Times (July 25).
KELLY, 1. (1997g), "Andersen: Accounting Company Splits to Go to Arbitration," Finan-
cial Times (December 18).
KELLY, 1. (1998), "Virtual Firms Stays in Realm of the Unreal," Financial Times (February
14).
KESSLER, D., and M. McCELLAN (1996), "Do Doctors Practice Defensive Medicine?,"
Working Paper, no. 5.466 (Cambridge, MA: National Bureau of Economic Research),
May.
KESTIGIAN, M. (1992), "The TPC's Review of the Accountancy Profession," Australian
Accountant, 62 (July), 20-26.
KLEIN, B. (1995), "The Economics of Franchise Contracts," Journal of Corporate Finance:
Contracting, Governance and Organization, 2, 356-362.
KLEIN, B., and K. LEFFLER (1981), "The Role of Market Forces in Assuring Contractual
Performance," Journal of Political Economy, 89 (August), 615-641.
KLEIN, B., and K. MURPHY (1988), "Vertical Restraints as Contract Enforcement Mecha-
nisms," Journal of Law and Economics, 31 (October), 265-297.
KLEIN, B., and L. F. SAFT (1985), "The Law and Economics of Franchise Tying Con-
tracts," Journal of Law and Economics, 28 (May), 345-361.
184 References
KLEIN, B., R. G. CRAWFORD and A. A. ALCHIAN (1978), "Vertical Integration, Appro-

priable Rents, and the Competitive Contracting Process," Journal of Law and Economics,
21 (2),297-326.
KNAPP, M. C. (1985), "Audit Conflict: An Empirical Study of the Perceived Ability of
Auditors to Resist Management Pressure," The Accounting Review, 60 (2), 202-211.
KRAAKMAN, R. H. (1986), "Gatekeepers: The Anatomy of a Third-Party Enforcement
Strategy," Journal of Law, Economics, and Organization, 2, 53-105.
KREPS, D. M, and R. WILSON (1982), "Reputation and Imperfect Information," Journal of
Economic Theory, 27, 253-279.
KRISHNAN, 1. (1994), "Auditor Switching and Conservatism," The Accounting Review, 69
(1),200-215.
KRISHNAN, 1., and R. G. STEPHENS (1995), "Evidence on Opinion Shopping From Audit
Opinion Conservatism," Journal ofAccounting and Public Policy, 14 (3),179-201.
KRUEGER, A. B. (1991), "Ownership, Agency and Wages: An Examination of Franchising
in the Fast Food Industry," Quarterly Journal of Economics, 106,75-101.
LAFONTAINE, F. (1992), "Agency Theory and Franchising: Some Empirical Results,"
Rand Journal ofEconomics, 23, 263-283.
LAFONTAINE, F. (1993), "Contractual Arrangements as Signaling Devices: Evidence from
Franchising," Journal of Law, Economics and Organization, 9, 256-289.
LAZEAR, E. P. (1979), "Why Is There Mandatory Retirement?," Journal of Political Econ-
omy, 87 (December), 1261-1284.
LEIBMAN, 1., and A. KELLY (1992), "Accountant's Liability to Third Parties for Negligent
Misrepresentation: The Search for a New Limiting Principle," American Business Law
Journal (November), 347-439.
LOWE, D. 1., and K. PANY (1994), "Auditor Independence: The Performance of Consulting
Engagements with Audit Clients," Journal ofApplied Business Research, 10 (1), 6-13.
LOWE, D. 1., and K. PANY (1995), "CPA Performance of Consulting Engagements with
Audit Clients: Effects on Financial Statement Users' Perceptions and Decisions," Audit-
ing: A Journal of Practice and Theory, 14 (2), 35-53.
LOWE, D. 1., and K. PANY (1996), "An Examination of the Effects of Type of Engagement,
Materiality and Structure on CPA Consulting Engagements," Accounting Horizons, 10
(4),32-51.
L YS, T., and R. L. WATIS (1994), "Lawsuits Against Auditors," Journal of Accounting
Research, 32 (supplement), 65-93.
MAGEE, R. P., and M.-C. TSENG (1990), "Audit Pricing and Independence," The Ac-
counting Review, 65 (2), 315-336.
MASTEN, S. E., 1. W. MEEHAN Jr., and E. A. SNYDER (1991), "The Costs ofOrganiza-
tion," Journal of Law, Economics, and Organization, 7 (1),1-25.
MAUTZ, R. F., and H. A. SHARAF (1961), The Philosophy of Auditing (Sarasota, FL:
American Accounting Association).
MA YERS, D., and C. W. SMITH (1982), "Toward a Positive Theory of Insurance," Mono-
graph Series in Economics and Finance (New York: Solomon Brothers Center for the
Study of Financial Institutions, New Yark University).
McKINLEY, S., K., PANY and P. M. 1. RECKERS (1985), "An Examination of the Influ-
ence of CPA Firm Type, Size, and MAS Provision on Loan Officer Decisions and
Perceptions," Journal ofAccounting Research, 23 (2), 887-896.
References 185
MEDNICK, R. (1990), "Independence: Let's Get Back to Basics," Journal of Accountancy,

169 (1),86-93.
MEDNICK, R., and 1. PECK (1994), "Proportionality: A Much-Needed Solution to the
Accountants' Legal Liability Crisis," Valparaiso University Law Review, 28 (3),867-918.
MELUMAD, N., and L. THOMAN (1990), "On Auditors and the Courts in an Adverse
Selection Setting," Journal ofAccounting Research, 28 (1),77-120.
MENON, K., and D. D. WILLIAMS (1994), "The Insurance Hypothesis and Market Prices,"
The Accounting Review, 62 (2), 327-342.
METCALF COMMITTEE (1976), The Accounting Establishment: A Staff Study; United
States Senate, Subcommittee on Reports, Accounting and Management of the Committee
on Government Operations, 94th Congress, 2d Session; Senator L. Metcalf, president
(Washington DC: U.S. Government Printing Office), December. (Summary reproduced
in The Journal ofAccountancy [March J, 104-120).
MILGROM, P. R. (1981), "Good News and Bad News: Representation Theorems and Appli-
cations," Bell Journal of Economics, 12 (Autumn), 380-391.
MILGROM, P., and 1. ROBERTS (1982a), "Limit Pricing and Entry Under Incomplete
Information: An Equilibrium Analysis," Econometrica, 50, 443-460.
MILGROM, P., and 1. ROBERTS (l982b), "Predation, Reputation and Entry Deterrence,"
Journal ofEconomic Theory, 27, 280-312.
MILGROM, P., and 1. ROBERTS (1992), Economics, Organization and Management
(Englewood Cliffs, NJ: Prentice-Hall).
MOLERO, A. G. (1995), "Arturos: una raza aparte," Actualidad Econ6mica, no. 1947 (Oc-
tober 16) 64-66.
NANDA, A., and M. YOSHINO (1996), "Andersen Consulting-EMEAI: Reorganization for
Revitalization," Harvard Business School Case, 9-396-007 (revised version: June 17).
NARAYANAN, V. G. (1994), "An Analysis of Auditor Liability Rules," Journal of Ac-
counting Research, 32 (supplement), 39-59.
NELSON, P. (1970), "Information and Consumer Behavior," Journal of Political Economy,
78,311-329.
NICHOLS, D. R., and K. H. PRICE (1976), "The Audit-Firm Conflict: An Analysis Using
Concepts of Exchange Theory," The Accounting Review, 51 (April), 335-346.
NORTON, S. W. (1988), "An Empirical Look at Franchising as an Organizational Form,"
Journal of Business, 61 (2), 179-218.
NORTON, S. W. (1995), "Is Franchising a Capital Structure Issue," Journal of Corporate
Finance: Contracting, Governance and Organization, 2, 75-101.
O'KEEFE, T. B., D. A. SIMUNIC and M. T. STEIN (1994), "The Production of Audit
Services: Evidence From a Major Public Accounting Firm," Journal of Accounting Re-
search, 32 (2), 241-261.
OECD (1985), Competition Policy and the Professions (Paris: OECD).
PALMROSE, Z. V. (1986b), "The Effect of Nonaudit Services on the Pricing of Audit
Services: Further Evidence," Journal ofAccounting Research, 24 (2), 405-411.
PALMROSE, Z. V. (1988a), "An Analysis of Auditor Litigation and Audit Service Quality,"
The Accounting Review, 63 (1),55-73.
186 References
PALMROSE, Z. V. (I 988b), "Public Accounting Firms and the Acquisition of Nonaudit

Services by Public and Closely-Held Companies," Auditing: A Journal of Practice and
Theory, 8 (I), 63-71.
PALMROSE, Z. V. (1991), "Trials of Legal Disputes Involving Independent Auditors: Some
Empirical Evidence," Journal ofAccounting Research, 29 (supplement), 149-185.
PANY, K., and P. M. 1. RECKERS (1984), "Non-Audit Services and Auditor Independence-
A Continuing Problem," Auditing: A Journal of Practice and Theory, 3 (2), 89-97.
PANY, K., and P. M. 1. RECKERS (1987), "Within- Vs. Between-Subject Experimental
Designs: A Study of Demand Effects," Auditing: A Journal of Practice and Theory, 7 (1),
39-53.
PANY, K., and P. M. 1. RECKERS (1988), "Auditor Performance of MAS: A Study of its
Effects on Decisions and Perceptions," Accounting Horizons, 2 (June), 31-38.
PARK, S. H. (1990), "Competition, Independence and Audit Quality: The Korean Experi-
ence," International Journal ofAccounting, 25, 71-86.
PARKASH, M., and C. F. VENABLE (1993), "Auditee Incentives for Auditor Independ-
ence: The Case of Nonaudit Services," The Accounting Review, 68 (I), 113-133.
PAZ-ARES, C. (1996), La Ley, el mercado y la independencia del auditor (Madrid: Civitas).
POB, Public Oversight Board (1979), Scope of Services by CPA Firms (Public Oversight
Board of the SEC Practice Section of the Division for CPA Firms in the American Insti-
tute of Certified Public Accountants, AICPA, March).
POSNER, R. A. (1972), "The Appropriate Scope of Regulation in the Cable Television
Industry," Bell Journal of Economics, 3, 98-129.
PRATT, 1., and 1. D. STICE (1994), "The Effects of Client Characteristics on Auditor Liti-
gation Risk Judgments, Required Audit Evidence, and Recommended Audit Fees," The
Accounting Review, 69 (4), 639-656.
PREVITS, G. 1. (1985), The Scope of Services: A Study of the Development of the Concept of
Independence and the Profession's Role in Society (New York: Wiley).
RANKIN, K. (1990), "SEC OK's Consulting-Auditing Jobs at AA," Accounting Today, 3
(July 23), 3.
RANKIN, K. (1997), "Get Back to Work! SEC Chief Chides CPAs for Shirking Watchdog
Role," Accounting Today (January 6).
REA, Registro de Economistas Auditores (1993), "Modelo de carta de encargo," in Manual
de Auditoria: Inventario de objetivos y procedimientos de Contabilidad y Auditoria 2nd
ed. (Madrid: REA), 637-641.
RIDY ARD, D., and J. de BOLLE (1992), Competition in European Accounting (study
commissioned by the EEC Commission to National Economic Research Associates,
NERA, acting as project advisers D. Glynn, D. Gribbin and Ch. Nobes), (Dublin: Laf-
ferty Publications Ltd.).
RITTENBERG, L. (1977), Auditor Independence and Systems Design (New York: Institute
of Internal Auditors).
ROBERTS, R. W, G. W GLEZEN and T. W. JONES (1990), "Determinants of Auditor
Change in the Public Sector," Journal ofAccounting Research, 28, 220-228.
ROBEY, D. (1991), Designing Organizations (Burr Ridge, IL: Irwin).
ROTH, A. E. (1995), "Bargaining Experiments," in J. H. Kagel and A. E. Roth, (eds.), The
Handbook of Experimental Economics (Princeton, NJ: Princeton University Press), 253-
348.
References 187
RUBIN, P. H. (1978), "The Theory of the Firm and the Structure of the Franchise Contract,"
Journal of Law and Economics, 21 (April), 223-233.
SCHAEFER, 1., and R. 8. WELKER (1994), "Distinguishing Characteristics of Certified
Public Accountants Disciplined for Unprofessional Behavior," Journal ofAccounting and
Public Policy, 13 (2), 97-119.
SCHATZBERG,1. W. (1990), "A Laboratory Market Investigation of Low Balling in Audit
Pricing," The Accounting Review, 6S (2), 337-362.
SCHEINER, 1. H. (1984), "An Empirical Assessment of the Impact of SEC Nonaudit Service
Disclosure Requirements on Independent Auditors and their Clients," Journal of Ac-
counting Research, 22 (2), 789-797.
SCHEPANSKI, A., R. M. TUBBS, and R. A. GRIM LUND (1992), "Issues of Concern Re-
garding Within- and Between-Subjects Designs in Behavioral Accounting Research",
Journal ofAccounting Literature, 11 (autumn), 121-150.
SCHULTE, A. A. Jr. (1965), "Compatibility of Management Consulting and Auditing," The
Accounting Review, 40 (July), 587-593.
SCOTT, R. W. (1997), "Microsoft Tries to Enlist Firms in Consulting Wars," Accounting
Today (January 6).
SEC, Securities and Exchange Commission (1978), "Disclosure of Relationships with Inde-
pendent Public Accountants," Accounting Series Release, no. 250, June 29 (Washington
D.C.: Government Printing Office).
SEC, Securities and Exchange Commission (1979), "Scope of Services by Independent
Accountants," Accounting Series Release, no. 264, June 14 (Washington D.C.: Govern-
ment Printing Office).
SEC, Securities and Exchange Commission (1981), "Relationships Between Registrants and
Independent Accountants," Accounting Series Release, no. 296, August 20 (Washington
SEC, Securities and Exchange Commission (1982), "Relationships Between Registrants and
Independent Accountants," Accounting Series Release, no. 304, January 28 (Washington
SEC, Securities and Exchange Commission (1994), Office of the Chief Accountant (OCA),
Staff Report on Auditor Independence to the US House of Representatives (March).
SHAPIRO, C. (1983), "Premiums for High Quality Products as Returns to Reputations,"
Quarterly Journal of Economics, 98 (November), 659-679.
SHAPIRO, C. (1991), "Symposium on the Economics of Liability," Journal of Economic
Perspectives, S (3),3-10.
SHAPIRO, c., and 1. E. STIGLITZ (1984), "Equilibrium Unemployment as a Worker Disci-
pline Device," American Economic ReView, 74 (June), 433-444.
SHIBANO, T. (1994), "Over-Guarding the Guardians: Excessive Auditor Liability Can
Cause Suboptimal Investment," Working Paper (Berkeley, CA: University of California).
SHOCKLEY, R. A. (1981), "Perceptions of Auditors' Independence: An Empirical Analy-
sis," The Accounting Review, S6 (4), 785-800.
SIMON, C. 1. (1989), "The Effect of the 1933 Securities Act on Investor Information and the
Performance of New Issues," American Economic Review, 79 (3), 295-318.
SIMON, D., and 1. FRANCIS (1988), "The Effects of Auditor Change on Audit Fees: Tests
of Price Cutting and Price Recovery," The Accounting Review, 63 (2), 255-269.
188 References
SIMUNIC, D. A. (1984), "Auditing, Consulting, and Auditor Independence," Journal of

Accounting Research, 22 (2), 679-702.
SMITH, C. W., and J. B. WARNER (1979), "On Financial Contracting: An Analysis of
Bond Covenants," Journal ofFinancial Economics, 7,117-161.
STICE, J. D. (1991), "Using Financial and Market Information to Identify Pre-Engagement
Factors Associated with Lawsuits Against Auditors," The Accounting Review, 66 (July),
516-533.
STIGLER, G. 1. (1964), "Public Regulation of the Securities Market," Journal of Business,
37,117-142.
STIGLER, G. 1. (1968), "Price and Non Price Competition," Journal of Political Economy,
72 (I). Reproduced in G. 1. Stigler, The Organization of Industry (Homewood, IL: Irwin),
23-28.
TDC, Tribunal de Defensa de la Competencia (1992), Informe sobre el libre ejercicio de las
profesiones: Propuesta para adecuar la normativa sobre las profesiones colegiadas al
regimen de la libre competencia vigente en Espana (Madrid: Tribunal de Defensa de la
Competencia), June.
TELSER, L. (1966), "Cut-throat Competition and the Long Purse," Journal of Law and
Economics (October), 259-277.
TENG, A. Y. (1994), "CPAs Need Consultants Too: An Outsourcing Guide," Accounting
Today (May 2).
TEOH, S. H., and T. 1. WONG (1993), "Perceived Auditor Quality and the Earnings Re-
sponse Coefficient," The Accounting Review, 68 (2), 346-366.
TIROLE, 1. (1986), "Hierarchies and Bureaucracies: On the Role of Collusion in Organiza-
tions," Journal of Law, Economics and Organization, 2, 181-214.
TIROLE,1. (1988), The Theory of Industrial Organization (Cambridge, MA: MIT Press).
TIROLE, 1. (1992), "Collusion and the Theory of Organizations," in 1.-1. Laffont (ed.),
Advances in Economic Theory, Econometric Society Sixth World Congress, Barcelona,
1990 (Cambridge: Cambridge University Press), 151-206.
TITARD, P. L. (1971), "Independence and MAS: Opinions of Financial Statement Users,"
Journal ofAccountancy (July), 47-52.
TITMAN, S., and B. TRUEMAN (1986), "Information Quality and the Valuation of New
Issues," Journal ofAccounting and Economics, 8, 159-172.
TPC, Trade Practices Commission (1991), "Study of the Accountancy Profession," Issues for
Discussion (March 22).
TURPEN, R. A. (1990), "Differential Pricing on Auditors' Initial Engagements: Further
Evidence," Auditing: A Journal of Practice and Theory, 9 (2), 60-76.
U.S. DEPARTMENT OF JUSTICE and FEDERAL TRADE COMMISSION (1992), Hori-
zontal Merger Guidelines, (April 2. Revised: April 8, 1997).
VERRECCHIA, R. E. (1983), "Discretionary Disclosure," Journal of Accounting and Eco-
nomics, 5, 179-194.
WAKEMAN, L. (1981), "The Real Function of Bond Rating Agencies," Chase Financial
Quarterly, 1,18-26.
WAITS, R. L., and 1. L. ZIMMERMAN (1980), "The Markets for Independence and Inde-
pendent Auditors," Working Paper, August (Rochester, NY: University of Rochester
Graduate School of Management).
References 189
WATTS, R. L., and J. L. ZIMMERMAN (1983), "Agency Problems, Auditing, and the
Theory of the Firm: Some Evidence," Journal of Law and Economics, 26 (October), 613-
633.
WATTS, R. L., and J. L. ZIMMERMAN (1986), Positive Accounting Theory (Englewood
Cliffs, NJ: Prentice-Hall).
WHITE, J. B., and E. MACDONALD (1997), "Generation Gap: At Arthur Andersen, Ac-
countants Face Unlikely Adversary: The Consulting Operation They Created Becomes
Rival for Power, Prestige. One Company, Two Cultures," The Wall Street Journal Europe
(April 24), I and 4.
WILLIAMSON, O. E (1975), Markets and Hierarchies: Analysis and Antitrust Implications
(New York: The Free Press).
WILLIAMSON, O. E. (1976), "Franchise Bidding for Natural Monopoly-In General and
with Respect to CATV," The Bell Journal of Economics, 7 (spring), 73-107. Revised in
The Economic Institutions of Capitalism: Firms, Markets, Relational Contracting (New
York: The Free Press, 1985),326-364.
WILLIAMSON, O. E. (1979), "Transaction-Cost Economics: The Governance of Contrac-
tual Relations," Journal of Law and Economics, 22, 233-261.
WILLIAMSON, O. E. (1983), "Credible Commitments: Using Hostages to Support Ex-
change," American Economic Review, 73 (September), 519-540.
WILLIAMSON, O. E. (1985), The Economic Institutions of Capitalism: Firms, Markets,
Relational Contracting (New York: The Free Press).
WILSON, R. (1983), "Auditing: Perspectives from Multi-Person Decision Theory," The
Accounting Review, 58 (2), 305-318.
WINES, G. (1994), "Auditor Independence, Audit Qualifications and the Provision of Non-
Audit Services: A Note," Accounting and Finance, 34 (I), 75-86.
YARDLEY, 1. A., N. L. KAUFFMAN, T. D. CAIRNEY and W. D. ALBRECHT (1992),
"Supplier Behavior in the U.S. Audit Market," Journal of Accounting Literature, 11, 151-
185.
INDEX
Aaker, D. A., 29 Audit firms

Abdel-Khalik, A. R., 79 as a hybrid form of organization, 104
Accountancy, 17,43, 110, 128 as cooperatives, 104
as franchised units, 104
Accounting Series Release (ASR), 151-3
contractual arrangements, 106
Advertising, 165 federation of firms, 104-8
as a quality-assurance strategy, 32 independence within each network,
regulation, 32, 143 106,108
Agency costs, 4, 8, 16, 91 internal organization and rules, 168
Albrecht, W. D., 112 ownership structure, 27
Alchian, A. A., 25 Audit market
American Institute of Certified Public extent of competition, 55, 169
Accountants (AICPA), 69, 89, 92, product differentiation, 56
100-2, 152, 168 Audit quality
Andersen Consulting, 27, 104, 106-7 and multiple services, 164
Anderson, D., 5 and underpricing of IPOs, 4
concept and attributes, 4, 8-9, 160
Antle, R., 12, 43, 76-7
control procedures, 102, 107
Appointment of auditors
safeguards, 163-5
by a government body, 4
Auditing Practices Board (APB), 3,
by lottery, 5
134-5
Arruiiada, B., 33, 36, 41,49, 88, 112
Auditing profession, 6, 34
Arthur Andersen, 73, 106-7
Auditor independence
Audihispana, 106 and bargaining power between
Audit committee, 5, 133-5, 152 auditor and client, 86-7
Audit failures, 17,37-8,88, 101-2 and internal rules, 102
191
192 Index
and non-audit services, 81-108 Beattie, Y., 129-30

as an economic decision, 43-63 Beatty, R., 4, 29
auditor dilemma, 44 Beck, P. 1., 17,50,55,82, 88-9
concept, 43-4
Becker, G. S., 21, 25
conflict with clients and third parties,
15-7,162-3 Belgium, 63, 108, 119-20, 142, 146,
151-2,171
costs of collusion, 88
economic incentives, 82 Benchmarking, 73
importance of reputation, 86 Benoit, E., 110
limitations of the studies based on Benston, G. 1., 6, 126, 132
surveys, 92-6
Berg, 1. E., 91
organizational safeguards, 101-4
Berton, L., 73, 110, 117
probability of auditor switching,
54-5,87-8 Big firms, 26, 52, 64
professional liability, 88 Big Eight firms, 94, 132
quality control procedures, 102 Big Five firms, 8, 70, 129
quasi-rents, 84-5 Big Six firms, 62
rules of conduct, 101 Boletin Oficial del ICAC, 6, 34
self-regulatory safeguards, 99-101
Bolle, 1. de, 17, 108, 123
specific assets, 85-6
Bonding costs, 4
value for clients and third parties, 16,
162 Booth, 1., 5
Auditor objectivity, 44, 90, 100 Bradbury, M. E., 4
Auditor rotation, 17,45,54-5,87-8 Brand-stretching, 29
Australia, 26, 33, 52, 55, 63, 78, 112, Brandt, R., 129
119-20,140,142,146,151-52 Brickley, 1. A., 16, 104
Austria, 32, 55, 63,119-20,146,152 Briloff, A. 1., 92, 132
Auxini, 144 Bufete Cuatrecasas, 106
Baiman, S., 12 Buijink, W., 6, 16,33,40, 55,63, 105,
Bain, 1., 121 120,136-7,141,146,151-2
Baker, 1. D., 62 Cadbury Report, 133-4
Balachandran, B. Y., 58, 60-1 Caimey, T. D., 112
Balvers, R. J., 3, 29 Campbell, D. R., 37
Bargaining power of auditor and client, Canada, 33, 40, 63, 140, 142, 146, 152
50-I, 86-7 Cano, 1. M. del, 107
empirical evidence, 51-3 Carcello, J. Y., 39
experimental results, 52
Cardador, G., 107-8
Barkess, L., 78, 90, 98
Carlton, D. W., 26
Barlev, B., 87
Carmichael, D. R., 132
Barriers to entry
Carpenter, C. G., 9, 29
as a quality-assurance strategy, 33
Index 193
Centre for European Policy Studies probability of auditor switching,

(CEPS), II, 13 54-5,87-8
Certified public accountant (CPA), 34 specific assets, 49-57
volume of auditor quasi-rents, 49-53
Chartered accountant, 6, 34
total quasi-rents to share between
Chow, C. W., 6, 8
auditor and client, 49
Cinco Dias, 4
Craswell, A. T., 8, 25-6, 52, 98
Co-contracting, 73, 93
Crawford, R. G., 25
Cohen Commission, 70, 88-9,132
Credence goods, 9,15,17,26
Cohen, M. F., 70, 88-9, 132
Credible commitment, 21
Comisi6n Nacional del Mercado de
Creditors, 3-4, 8, 91, 98,160
Yalores (CNMY), 134
Cross-subsidization
Continuity of the audit-client
between kinds of service, 121
relationship, 44
Current contracts, 45,58
Contractual asymmetry, 5
Darby, M. R., 10
Contractual costs, 4, 22, 28, 113, 143,
160,172 Dark, F. H., 16, 104
Contractual frictions, 56, 76 Davis, L. R., 60, 79, 86,90
Contractual safeguards, 4 DeAngelo, L. E., 9, 17, 28, 50, 52, 55,
118-9, 145, 154
Corless, J. c., 91
DeBerg, C. L., 77, 82, 90-1
Corporate Governance
good practices, II Dedicated assets, 54
Cost function Defective audits, 20, 32, 61
intertemporal cost structure, 56, 111 Defensive auditing, 9-14
recurrent audits, 51, 82, 115 Defond, M., 8
rotation costs, 51 Dejong, D. Y., 61
start-up costs, 56, 114, 169
Demand for external auditing, 3-8, 160
Costs of collusion, 88 of differing quality, 7
Costs of dependence, 47, 57 regulatory demand, 6
clients' ability to sanction auditor voluntary demand, 6-8
failure, 59 Demsetz, H., 114, 121
loss of quasi-rents linked to current
Demski, J., 76-7
clients, 58-9
loss of quasi-rents linked to potential Denmark, 33, 40, 55, 63,142,146,151-2
clients, 59-60 Dewatripont, M., 119
losses due to liabi lity, 60-4 Disclosure rules, 174
risk aversion, 64 disclosure of fees, 151-2
sanction by disappointed clients, 58 disclosure of maximum individual
Costs of independence, 48 concentration, 154-5, 175
194 Index
economic rationale, 149-50, 174 on independence in appearance, 92-8

Diseconomies of scope in the provision on independence in fact, 81-92
of non-audit services, 79-80 on lowballing, 114-7
Diversification, 28, 36, 56, 82-5, 118, on market competition, 109-30
142, 144,147, 155, 164-5, 168, Efficient wages, 21
172-3 Empirical evidence
as a regulatory objective, 172-3 correlation between audit quality and
difficult to asses when non-audit audit firm size, 29
services are provided, 174 cost of developing a reputation, 25
of fee income, 143-5, 159, 172 demand of auditing of differing
rules, 145-7 quality, 7
Divisionalization deterioration of incentives to
as safeguard, \03 perform, 34
Donabedian, B., 34 economies of scope, 77-9
effect of audit quality in improving
Dopuch, N., 91
contractual terms, 3
Downes, D., 5 how SEC supervision facilitates
Dye, R. A., 50, 53, 61,150 litigation against auditors, 39
Dykxhoorn, H. J., 94 independence in appearance, 92-8
Easterbrook, F. H., 150 positive effect on audit quality of
unsolicited offering of services,
Economies of scale, 77, \05, 167
32
Economies of scope pricing, 78-9, 90, III
empirical evidence, 77-9 reaction of audit clients, 91
in non-audit services, 75 relative bargaining power of auditor
in real-estate transactions, 41 and client, 51-3
in the provision of non-audit specialization of audit firms, 8
services, 71-79 through experimental analysis, 90
of contractual origin, 76, 166
Empirical evidence on non-audit services
of technical origin, 71, 75-6,166
and auditor switching, 89
relative importance of technical and
and independence in appearance,
contractual economies of scope,
92-8
76-7,167
and independence in fact, 88-92
Effects of non-audit services and litigiousness, 90
cost reduction, 166 and qualified opinion, 98
different opinions on the appearance effect on independence, 88-98
of independence, 94-6
effect on prices, 112
on audit firms, 122-4
evolution of non-audit fees, 117
on auditor independence, 81-108,
the opinion of shareholders, 89
167-8
on competition in non-audit markets, Entry barriers, 165
122-30, 170-1 Ernst & Young, \05
on competition in the audit market, Ethical codes, 100
113-22, 168-9
Index 195
Ettredge, M., 52 Frecka, T. 1., 17,50, 88-9

EU Green Paper, 71, 74,110,119,123-4, Garcia Benau, M. A., 95-6
136-8 Gatekeepers, 41
ambiguity with respect to non-audit
Gaver, J., 78
services, 137-8
contradictions, 138 Gaver, K. M., 78
expectations on auditing, 74, 137 Generally accepted accounting principles
normative possibilities on non-audit (GAAP)
services, 138 differences in interpretation, 38, 50
recommendations, 139 Germany, 41, 55, 63, 70, 146, 152
European Contact Group, 136, 138 Gigler, F., 56, 117
European Union Gilson, R., 5
directives, 6, 140 Glezen, G. W., 52, 89,91
regulatory prospects, 139
Goldman, A., 87
Expansion, 64,106,118,124,144
Gonzalo Angulo, 1. A., 31, 110, 117, 123
Experience goods, 9
Great Britain, 17, 33,62-3, 74, 78, 95-6,
Explicit contracts, 20 128,130,134,146,149
External effects Greece, 33, 63,119-20,142,146-7,152
between audit and non-audit services, Green Paper on the Role, the Position
31 and the Liability of the Statutory
of client-specific assets, 28 Auditor within the European Union.
Ezzamel, M., 78 See EU Green Paper
Fama, E. F., 27, 91,104,107 Greenberg, R., 52
Fearnley, S., 129 Grimlund, R. A., 97
Federation des Experts Comptables Grossman, S. 1., II, 149
Europeens (FEE), 43, 99-10 I, 135-6, Grout, P., 11-4,50,55,84,118-9,149
145-6
Gwilliam, D. R., 78
Finland, 6, 33,40,63,146-7, 152
Hagerty, K. M., ISO
Firth, M., 8, 29, 78, 93
Hanson, R. K., 62
Fischel, D. R., 150
Hart, O. H., 11,149
Fishman, M. 1.,150
Hartley, R. Y., 92
France, 40,63,123,142,146,152,171
Heinkel, R., 5
Franchise bidding, 114
Hill, 1. W., 60
Franchising
Hogan, C. E., 4
audit firms as franchised units, 104
Holland, K., 78
free riding on quality and brand
reputation, 16 Human capital, 25, 27, 54, 59, 72-3, 106
Francis, 1. R., 8,26,29,52, 112 Humphrey, c., 95-6
196 Index
Implicit contracts, 20-1, 111 risk of confusion, 26, 150-1, 174

Independence in appearance, 43-4, 71, Ireland, 6, 63,135,140,142,146,151-2
92-8, 168 Italy, 36,40,55,63,119-20,123,141-2,
empirical evidence, 92-8 146-7, 151-2, 171
limitations of surveys, 96-7 Jack, A., 128
Independence in fact, 10,43, 71, 81-92, Japan, 6, 33, 63,119-20,142,146-7,152
98,139,168
Jarrell, G. A., 6
effect of quasi-rents, 53
Individual auditors
Jensen, M. c., 4-5, 27, 91,104,107
and barriers to entry, 34 Jeter, D. c., 32
Informational asymmetry, 15,21-2,29, Jewitt, 1.,11-4,50,55,84,118-9,149
76,122,164 Joint audits, 40
between auditor and client, 51-2 Jones, E., 52, 75
Informative content of auditing, 10-5, Jones, T. W., 52, 89, 91
161,175
Journal of Economic Perspectives, 60
and non-audit services, 14
Jovanovic, B., 150
hard information, 12-3, 161
optimum rules, 12 Kanodia, C, 50-1
soft information, 11-4, 161 Kaplan, S. E., 77, 90
Informative value of accounts, 3 Karni, E., 9
Initial public offers (IPOs), 3 Kauffman, N. L., 112
Insolvency, 20, 34,47, 85 Kelly, A., 103
Instituto de Censores Jurados de Cuentas Kelly, J., 103, 105, 107
de Espana (ICJCE), 112 Kessler, D., 10
Instituto de Contabilidad y Auditoria de Kestigian, M., 128
Cuentas (lCAC), 6, 112
King, R. R., 91
Intangible assets, 14,27,32,72, 104-5,
Klein, B., 21-2, 25, 28, 104
165
Internal auditing, 5, 77, 101
Knapp, M. c., 93, 97
Knowledge spillovers, 75
Internal Auditor, 77
Korea, 4
Internal rules on independence, 102
KPMG, 73,105,128
International Federation of Accountants
(IFAC), 100, 135, 139 Kraakman, R. H., 5, 41
Intertemporal competition, 114, 154, 169 Kreps, D. M, 37
Introductory pricing Krishnan, Jagan, II
as a consequence of competition, Lafontaine, F., 16, 104
114-20 Laventhol & Horwath
as part of a quality-assurance effects of bankruptcy on its clients,
strategy, 26-7 59
in consulting, 27
price of initial audits, 52
Index 197
Lawyers' opposition to non-audit Mccellan, M., 10

services, 124 McDonald, B., 3, 29
Lazear, E. P., 21 Mckinley, S., 93, 97
Leffler, K., 21-2, 28 Meckling, W. H., 4-5, 91
Leibman, 1., 103 Mednick, R., 43, 62
Loss-leader (auditing as), 109-12 Meehan Jr., 1. W., 54
Lowballing, 26, 50-2, 55, 57, 109-20, Melumad, N., 61
169
Menon, K., 4, 59
and implicit contracting, 112
and the EU Green Paper, 119 Merger Guidelines, 127
consequences, 169 Metcalf Report, 70-1, 132
difficult to enforce prohibition, 123 Metcalf, L., 70-1, 132
doubts on its real extent, 111-3 Metzger, M. B., 60
economic rationale, 113-4
Meuwissen, R., 6, 16,33,40,55,63,
effect of non-audit services, 114-7 105,120,136-7,141,146,151-2
effects of its prohibition, 118-9
Microsoft, 73
in mobile telephones, 115, 118,
169-70 Milgrom, P. R., 37,122,126,149
regulations, 119-20, 169 Millar, A., 89, 91
see also Introductory pricing Miller, R. E., 3, 29
Lowe, D. 1., 93, 103 Minimum level of quality, 33-5
Luxembourg, 63,140,142,146,152 Model of the auditor decision, 55-7
Lys, T., 60, 90 Molero, A. G., 105
MacDonald, E., 107 Monitoring costs, 4
Magee, R. P., 50 Moral hazard, 5, 21
Maijoor, S. 1., 6, 16,33,40,55,63, 105, Mukherji, A., 50-1
120,136-7,141,146,151-2 Nagarajan, N. 1., 61
Mandatory auditing, 6, 143 Nalebuff, B., 12
Mandatory auditor rotation, 36, 40, 51 Nanda, A., 27, 73, 106
Mandatory disclosure, 140, 150-1, 153, Narayanan, V. G., 62
165, 173-5
Nelson, P., 9
MARC Report, 6,16,33,40,55,63,105,
120,136-7,141,146,151-2 Netherlands, 33, 63,108,140,142,146,
152
Market versus liability in promoting
audit quality, 14 Network economies, 77, 167
Masten, S. E., 54 Nichols, D. R., 43
Mautz, R. F., 43 Non-audit services
accounting-related, 71
Mayers, D., 5
clients' rationality, 125
198 Index
consequences, 165 Pany, K. V., 77, 90, 93, 97, 103
°
consequences for providers of other Park, S. H., 4
services, 13 Parkash, M., 82, 91
defense of monopoly hypothesis,
126-7 Parker, L. M., 37, 91
divisionalization, 103 Paz-Ares, c., 36, 88, 112
economies of scope. See Economies Peck, J., 62
of scope Peer review, 76, 102
effects. See Effects of non-audit
Penno, M., 56, 117
services
efficient use of experts, 73-4, 137 Perl off, 1. W., 26
empirical evidence. See Empirical Pong, c., 11-4,50,55,84,118-9,149
evidence on non-audit services Portugal, 55, 63, 119-20, 123, 142,
executive recruitment, 70, 101, 126, 146-7,152
132 Positive Accounting Theory, 9
forces driving its growing demand,
Posner, R. A., 114
72, 125, 128
general versus particular problems, Potential contracts, 45
70 Pratt, 1., 61
history, 74 Predatory behavior, 120-2
importance, 129-30 Previts, G. 1., 74-5
management advisory services, 95-6,
Price Waterhouse, 108
132,142,152-3
market concentration, 127 Price, K. H., 43
organizational safeguards of Probabilities
independence, IOJ of auditor switching, 54-5, 87-8
quality erosion hypothesis, 124 that the client will change auditor, 45
reaction of audit clients, 91 that the client's financial situation
regulation. See Regulation of non- deteriorates, 45
audit services Productive economies of scope. See
relative importance of fees, 149 Economies of scope
relative profitability, 117 Professional ethics, 43, 100, 119-20, 135,
self-regulation. See Self-regulation 138
specialization of staff, 103 Professional judgement, 9-14, 61,160-2,
taxation services, 12, 70, 130 166
types, 69
Professional liability, 10, 16,27,39,45,
unfair competition, 121 47,60,88,144,161
Norway, 6, 33, 63, 78,146,151-2 compensatory function, 60
OEeD, 63,128,140 consequences, 10
Olivencia Report, 134 current importance, 61
One-stop shopping, 76, 125, 128, 166 joint and several, 16,62
in the advertising industry, 125 level of third party compensation, 16
premiums, 62
Palmrose, Z. V., 29, 39, 47, 61, 79,129
Index 199
proportionate liability, 62-3 cause of sub-optimal technology, 25

Proportionate liability. See Professional comparative advantage, 39,139
liability defective incentives, 140, 172
Public Oversight Board (POB), 70, 89 differential effects based on evasion
ability, 123
Qualified audit opinion, 44, 85, 90, 98
diversification rules. See
Quality assurance Diversification rules
attainable objectives for regulation, its reinforcement as a quality-
37,41 assurance strategy, 34-5
competitive equilibrium, 23-4 mandatory auditing. See Mandatory
high-cost safeguards, 32-4 auditing
low-cost safeguards, 25-32, 163 market-facilitating approach, 37-9,
purchasers' expectations, 22 140,171-2
regulation as a facilitator, 38 of advertising, 32
regulatory strategies, 40 of auditor rights on dismissal, 55
sources of quality assurance, 29 of auditor switching, 55
types of safeguards, 20-24 of minimum diversification, 143
up-market strategy, 26 of non-audit services. See Regulation
Quasi-rents, 32-3, 35-7, 71,111,113-4, of non-audit services
116,118-9,142,144-5,149,154, of professional liability, 63
159,163,168-9 of quality standards, 39
and auditor independence, 44-59 of unsolicited offering of audit
and non-audit services, 82-90 services, 33, 143
appropriation by parties, 49, 53, 83 perverse effect of regulatory
as quality safeguard, 20-5 activism, 153
associated with other clients, 53 possible strategies, 41
definition, 19,21 production of information, 39
effect on independence, 53, 84 separate for small and large firms, 34
individual quasi-rents, 25, 44, 47,50, Texan rule against lowballing, 118
53,58,115,118-9,145 Regulation of non-audit services, 139-55
Ramakrishnan, R., 58, 60 arguments against prohibition, 142-3
Rankin, K., 106, 133 existing rules, 140-2
market-facilitating approach, 139,
Real estate transactions
143
title assurance, 40
prohibition, 140
Reckers, P. M. J., 93, 97, 103
Rent seeking, 25, 33
Registro de Economistas Auditores
Reputation
(REA), 112
audit and non-audit services as rental
Regulation of reputation, 5
as a protector of reputation, 38 potential effects in consulting and
as a safeguard strategy, 34, 37, 40 auditing markets, 31
attainable objectives, 37,41
200 Index
Residua\\oss, 4 Simon, C. 1., 6

Ricchiute, D., 79, 90 Simon, D. T., 52, 60, 86, 112
Ridyard, D., 17, 108, 123 Simunic, D. A., 55, 78-9, 112, 121
Risk aversion, 64 Sinning, K. E., 94
Risk-neutrality assumption, 46 Size of the firms
Rittenberg, L., 43 and audit quality, 28
Roberts, 1., 37, 52, 122, 126 Small auditing firms, 36, 92, 106
Roberts, R. W., 37, 52,122,126 and barriers to entry, 34
Robey, D., 106 Smith, A., 8, 29
Ross, T. L., 92, 108 Smith, C. W., 5,15
Roth, A. E., 53 Smith, R., 5
Royal Mail case, 12 Sneddon, B., 26, 52
Rubin, P. H., 16, 104 Snyder, E. A., 54
Schaefer, 1., 29 Solomon, 1., 50, 88-9

Spain, 6-7, 34, 36, 55, 63, 70, 95-6,
Schatzberg, J. W., 52, 60
105-6,112,118,127,146,152
Scheiner, J. H., 153
Specific assets, 53, 88
Schepanski, A., 97 advantage as quality safeguard, 28
Schulte, Jr., A. A., 92 as a quality-assurance strategy, 27-9
Scott, R. W., 73 caused by excess capacity, 53
Search goods, 9 client-specific, 28, 45, 85, 154, 173-4
Securities and Exchange Commission connected to excess capacity, 88
(SEC), 6, 37-9, 55, 60, 63,101-2, dedicated assets, 54
132,141,146,149,151-3,155,175 firm-specific, 28-9, 45, 82, 85, 172
Self-regulation, 33, 40, 71, 99-108, 143, temporal specificity, 54
168 Staff specialization
of auditor independence, 99 as safeguard, 103
rule on abstention from decision- Standard form contracts, 112
making, 100, 104
Stein, M. T., 78
specific internal safeguards, 103
Stephens, R. G., 11
specific rules of conduct, 99
Stice, J. D., 61, 90
Shapiro, c., 21-2, 24, 60
Stigler, G. 1., 6, 21,114,118
Sharaf, H. A., 43
Stiglitz, 1. E., 21
Shareholders, 3-5, 8, 12,55,60,62,89,
91,98,106,135,152,160 Stock Exchange., 6
Shibano, T., 61 Stokes, D. 1., 5
Shockley, R. A., 92-4, 97 Strawser, R. H., 8,29
Signaling, 6, 122 Substitution
of auditing as a safeguard, 5
Simnett, R., 78, 90, 98
Index 201
Sweden, 6, 33,40,63, 140, 142, 146, Variability of audit and non-audit fees,
152 149
Swieringa, R. 1., 132 Venable, C. F., 82,91
Tangible assets, 27, 72 Verifiable information, 10-1
Taylor, S. L., 8, 25 Verrecchia, R. E., 150
Technical competence, 9-10,15,17,28, Voluntary demand for auditing, 3
33,61,71,98,139,159-60,166 Voluntary disclosure, 39, 148-51, 174
Technical economies of scope. See Wakeman, L., 5
Economies of scope
Watts, R. L., 6, 9, 34, 60, 90-1
Telser, L., 121
Welker, R. B., 29
Temporal specificity, 54
White, J. B., 107
Teng, A. Y., 129
Whittington, G., 11-4,50,55, 84, 118-9,
Teoh, S. H., 29 149
The Economist, 62, 106, 110, 125, 128-9 Williams, D. 0.,4,59
Thoman, L., 61 Williamson, O. E., 21,25,54, 114
Tirole, 1., 22, 37, 88, 121-2 Wilson, E. R., 8,29
Titard, P. L., 92 Wilson, R., 5, 37
Titman, S., 6 Wines, G., 98
Trade Practices Commission (TPC), 128 Witteloostuijn, A. Van, 6, 16,33,40,55,
Transaction costs, 3 63,105,120,136-7,141,146,151-2
Tribunal de Defensa de la Competencia Wong, T. J., 29
(TDC),128 Yardley, 1. A., 112
Trompeter, G. M., 79, 90 Yoshino, M., 27, 73,106
Trueman, B., 6 Zimmerman, 1. L., 6, 9, 34, 91
Tseng, M.-C., 50
Tubbs, R. M., 97
Turpen, R. A., 78,112,117, 149
United Kingdom, 17,33,62-3, 74,78,
95-6, 128, 130, 134, 146, 149
United States, 6, 8,10,16-7,33-4,37,
41,52,55,59,62-3, 70, 74, 78,
89-90,101-3,112,126,131,142-3,
146,151-2,167,172
Department of Justice, 127
Securities Act of 1933, 6
Up or out policies, 27,126

Benito Arruñada (Auth.) - The Economics of Audit Quality - Private Incentives and The Regulation of Audit and Non-Audit Services-Springer US (1999)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Benito Arruñada (Auth.) - The Economics of Audit Quality - Private Incentives and The Regulation of Audit and Non-Audit Services-Springer US (1999)

Uploaded by

Copyright:

Available Formats

The Economics of Audit Quality

The Economics of Audit Quality

Private Incentives and the Regulation

SPRINGER SCIENCE+BUSINESS MEDIA, LLC

ISBN 978-1-4419-5082-6 ISBN 978-1-4757-6728-5 (eBook)

Printed on acid-free paper

All Rights Reserved

1.1 The demand for financial auditing 3

1.2 The dimensions of quality 8

1.2.1 Technical competence and independence 8

2 The safeguard of quality 19

2.1 General analysis of safeguarding mechanisms 20

2.1.1 Explicit and implicit safeguards 20

2.2 Strategies adopted to create quality-safeguarding incentives 24

2.2.1 Low-cost strategies 25

2.2.2 High-cost strategies 32

2.4 The regulation of quality: Attainable objectives 37

3 Auditor independence as an economic decision 43

3.1 Concepts of auditor independence 43

3.2 Economic factors behind auditor independence 44

3.3 The cost of independence: the loss of client-specific assets 49

3.4 The costs of dependence 57

4 The supply of non-audit services by auditors 69

4.1 Definition and types of non-audit services: the importance of a

4.2 Economies of scope 71

4.2.1 Auditing technology and firms' complexity 71

4.2.4 The existence of diseconomies of scope 79

5 Effects of non-audit services on auditor independence 81

5.l Effects of non-audit services on the economic incentives for

5.2 Effects on the appearance of independence 92

5.3 Private safeguards of independence with respect to the supply of

6 Effects of non-audit services on market competition 109

6.1 The argument regarding auditing as a "loss-leader" 109

6.2 Doubts as to the prevalence of introductory pricing 111

6.3 The efficiency of introductory pricing 113

6.4 Elements for a positive analysis of the provision of non-audit

7 Normative possibilities for non-audit services 131

7.1 The public debate 131

8 Let knowledge evolve in the market 159

8.1 Private quality safeguards and the role of regulation 160

8.2 The provision of non-audit services: Consequences and regulation 165

guarding quality and normative proposals, making a substantial contribution to

As with any other goods or services it is appropriate to examine the quality of

1.1 The demand for financial auditing

1.1.1 Financial auditing as a contractual facilitator

Auditing is generally understood in terms of Agency Theory as a type of bonding

information asymmetries leading to performance failures.

9 See, for example, Titman and Trueman (1986).

1.1.2 Empirical evidence

The contractual costs of businesses vary substantially depending on their structure.

1.2 The dimensions of quality

1.2.1 Technical competence and independence

As an initial approximation to understanding the meaning of "quality" in auditing

1.2.2 The importance of professional judgement and the danger

fixed certain volume of information. The problem becomes more complicated

1.2.2.1 The informative content of auditing

1.2.2.2 General normative consequences: the argument for rules which

any, as modeled by Antle, Nalebuffand Baiman (1991).

1.2.2.3 Particular normative consequences: non-audit services as a source

If the informative value of auditing is taken as an objective and, therefore, there is

1.3 The conflict with third parties and among clients

29 See Smith and Warner (1979).

2.1 General analysis of safeguarding mechanisms