THE VALUE OF CORPORATE ACCOUNTING REPORTS:

ARGUMENTS FOR A POLITICAL ECONOMY OF ACCOUNTING*

DAVID J. COOPER
University of East Anglia

and

MICHAEL J. SHERER
University of Manchester
Abstract
Existing research on the choice of accounting methods for corporate reports emphasizes private
interests. In particular, shareholders' interest predominate in studies of the effects of accounting
information on individual users Attempts at assessing the social value of accounting reports, using the
approach of marginal economics to information or the analysts of economic consequences also exhibit,
in their execution, a pronounced shareholder orientation This paper suggests that an alternative
approach, the Political Economy of Accounting, may be fruitful This approach seeks to understand and
evaluate the functions of accounting within the context of the economic, social and political
environment in which it operates. Research within this framework is identified as having normative,
descriptive and critical qualities, and the paper concludes with some illustrations of potential research
areas.

The major objective of this paper is to outline an alternative framework for relevant accounting
research. The intention is to reinforce recent calls (Burchell et al., 1980; Tinker, 1980) to understand
how accounting systems operate in their social, political and economic context in order that "better"
accounting systems might eventually be designed. In order to set our arguments for a political economy
of accounting in context, the first two sections of this paper review many of the current approaches to
assessing the value of corporate reports, whilst this review might be thought to duplicate others (eg.
Gonedes & Dopuch, 1974; Foster 1980a, Leftwich, 1980, Beaver, 1981), the synthesis offered in this
paper is quite different It is designed to highlight the emphasis by most accounting research on
individuals (especially shareholders) and a concern with market equilibrium and the associated passive
acceptance of the existing social and political context of corporate reporting. Implicit in our review is a
notion of social welfare that focuses on society as an aggregate (rather than an aggregation of
individuals), an emphasis on d is t r i b u t i v e as well as exchange ( a l l o c a t i v e ) dimensions of
wealth and power and a concern with socially necessary rather than market determined production.
This view of social welfare leads to the conclusion that the study of the i n s t i t u t i o n a l context of
accounting is a legitimate and necessary area of study for accounting research The dominant concern
with shareholder interest has limited the development of research about how accounting systems
operate and for designing corporate accounting reports which may lead to a fundamental improvement
in social welfare.
 It is important at the outset to highlight a crucial tension in this paper which arises out of our
concern to create accountings that are valuable in society Our position, that the objectives of and for
accounting are fundamentally contested, arises out of the recognition that any accounting contains a
representation of a specific social and political context. Not only is accounting policy essentially political
in that it derives from the political struggle m society as a whole but also the outcomes of accounting
policy are essentially political in that they operate for the benefit of some groups in society and to the
detriment of others However, it does not follow I and this is how the tension in the paper manifests
itself that an improvement in accounting policy can necessarily be achieved within the accounting
domain Rather, there as the implication that the politically determined nature of the value of accounting
prevents any such resolution within accounting itself.
Social welfare as likely to be improved if accounting practices are recognised as being
consistently partial, that the strategic outcomes of accounting practices consistently (if not invariably)
favour specific interests in society and disadvantage others Therefore, we are arguing that there already
exists an established, if implicit, conceptual framework for accounting practice. This paper offers an
analysis of the value of corporate accounting reports which recognises both the tactical discontinuities
and variations in accounting policies, including the possibility that actual policy outcomes may be an
imperfect match with the underlying intentions and motivations and the strategic consensus and
patterns of outcomes that more or less consistently support financial and shareholder. Interests in
society In order to achieve an improvement in accounting practices (to make them more accountable to
society m a democratic way) at is important to cast aside the ideological mask which hides the reality of
accounting research, to identify how accounting research justifies current political arrangements and
patterns of advantage and disadvantage, and how accounting research similarly implies that such
arrangements and patterns are immutable, efficient and even effective That as the purpose of this
paper.
The first section of this paper reviews many of the studies which mm to assess the usefulness of
corporate accounting reports for users. These studies, because they have implications merely for the
private value of information, provide little guidance for the design of, and choice between, alternative
accounting reports that are intended to contribute towards social welfare.
In the next section some recent developments which purport to address the question of the
social, as against the private value of corporate accounting reports are discussed. These attempts, too,
provide an incomplete analysis of the social value of accounting information. Their deficiency lies in their
focus on issues of efficiency (rather than effectiveness) and their emphasis on a small set of users and
the producers of accounting reports.
An alternative framework for analysing the role of accounting information, designated as a
political economy of accounting, is presented thereafter. A political economy of accounting emphasizes
the infrastructure, the fundamental relations between class an society It recognizes the institutional
environment which supports the existing system of corporate reporting and subjects to critical scrutiny
those issues (such as the assumed importance of shareholders and securities markets) that are
frequently taken for granted in current accounting research. Finally this alternative paradigm of political
economy of accounting is applied to three examples of potential research.
 PRIVATE VALUE APPROACHES
There is a long tradition of accounting theory being concerned with the interests of the users of
accounts (Sterling, 1972). This theoretical concern with users has influenced practice in the form of
conceptual frameworks offered by professional bodies (AICPA, 1973; FASB, 1978b; Stamp, 1980; Macve,
1981). In this section we review the user orientation in order to highlight two features, the emphasis on
shareholders and the partial equilibrium approach adopted. Emphasis on users in a partial equilibrium
context may indicate necessary conditions for the private value of information. It does not provide a
sufficient basis for prescriptions about socially desirable accounting policy and reports.
Corporate reports and individual shareholders
Accounting theory has long been concerned with the interests of individual private
shareholders. Whilst many theories have concentrated on aiding shareholders in decisions concerning
their income, wealth and even utility 2 (Edwards & Bell, 1961; Chambers, 1966; Sterling, 1970; Beaver &
Demski, 1974), much of the empirical research has been limited to studies of shareholder usage and
understanding of accounting reports. Shareholders usage and understanding has been assessed in two
ways; firstly, by the application of techniques to measure readability, and hence understanding, of
accounting reports (e.g. Smith & Smith, 1970; Still, 1972; Haried, 1972, 1973; Adelberg, 1979); secondly,
by shareholders', or heir representatives', responses to questionnaires about their use, and hence
understanding, of financial reports (e.g. Epstein, 1975; Lee & Tweedie, 1977, 1981; Chang & Most, 1979;
Advisory Committee on Corporate Disclosure, 1977).
In addition to certain technical difficulties and inconsistencies inherent in the types of empirical
tests used, 3 both approaches suffer from problems of interpretation. There is an absence of references
to any theories of how investors do or should use accounting information. Possible theories might
include the bounded rationality model (Clarkson, 1962), or the portfolio model (Ball & Brown, 1969)
Without such a theory against which to evaluate the empirical findings, it is impossible to determine
whether "poor" usage or understanding is a material factor affecting individuals' actions. Other
problems of interpretation include the focussing on parts rather than the full contents of the accounting
reports (Gonedes, 1978) and the level of self-insight required of respondents (Nisbett & Wilson, 1977).
A second limitation of this type of empirical research concerns the benefits which are expected
to accrue from a concentration on the interests of the individual shareholder. The prescriptions derived
from this research include calls for accounting reports to be simplified, accounting policy makers to
concentrate on the needs of naive investors, and the need for education of individual shareholders in
accounting and financial matters (e.g Tweedie, 1981). A potential consequence of these prescriptions
would be to redistribute wealth from one group of "knowledgeable" shareholders to another group of
"naive" shareholders (Findlay, 1977). Indeed, it is an implicit value judgement of this type of research
that such a re-distribution is a beneficial consequence in itself. In effect, shareholders are depicted as
individuals operating within an environmental vacuum and this allows the design of corporate
accounting reports to be considered as if it were only of private interest But the omission of any
consideration for the immediate environment, the capital market, in which the shareholder class
operates, ignores wider effects which may ensue from such prescriptions Research into shareholder
usage and understanding cannot by itself assess whether the above re-distribution would lead to a more
appropriate allocation of resources within the capital market, let alone to a higher level of welfare for all
members of the economy. As Butterworth et al. argue (1981, pp. 58--62), understanding individual
responses may be of interest in contributing to a general understanding of accounting (elaborating users
and their settings); but It is unlikely that individual behaviour translates to aggregate market responses
(Scheling, 1978).
Corporate reports and aggregate shareholder behavior
Research indicating capital market "efficiency" (or otherwise) with respect to published financial
information might be thought to hold promise for understanding and designing accounting systems. By
explicitly taking account of the effects of the aggregate behaviour of investors in a market environment,
studies of the impact of accounting information on stock market prices might provide insights into
shareholder use of accounting reports and the choice between alternative reporting methods.
Foster (1980) suggests that alternative accounting reports may affect the cash flows of
individual firms, the covariance of cash flows of individual firms with the market, the risk-return
characteristic of the entire capital market and the information set used by market traders. Only the first
two effects seem to have been systematically investigated Studies in the U S (Ball & Brown, 1969) and in
the U K (Firth, 1977) indicate that there ~s some information content in accounting earnings reports but
this information is not "timely" since the market price changes precede publication of the accounting
reports Reviews of the empirical literature (Foster, 1978) also suggests that the stock market can "see
through" and adjust for changes in accounting policy which do not affect the economic position of the
reporting firm. Thus the wealth of evidence from many of the empirical studies of the relationship
between published accounting information and stock market prices seems to indicate that the private
shareholder cannot make consistent abnormal gains from using such information.
However there are a number of reasons why accounting policy makers should be wary of using
market responses to assess the private and particularly the social value of the reports chosen Firstly
there are considerable problems in defining and hence assessing the efficiency of reformation markets
(Fama, 1976; Beaver, 1981). A second difficulty with the efficient market literature occurs because most
existing tests are, in fact, joint tests of the capital asset pricing model and the efficiency of the
information market Interpretation of the results of tests is accordingly ambiguous (Ball, 1978, Foster,
1979). Many of the above difficulties might be removed by careful theoretical work about the meaning
of efficiency (Beaver, 1981), explicit modelling of shareholder reaction to potential gains (Ohlson, 1979,
Patell, 1979) and more sophisticated econometric work (e.g. dealing with beta stationarity, the
normality of security returns and measures of the market portfolio).
Perhaps of more importance in relation to the value of accounting reports is that this literature
is concerned with the efficiency of the market for information rather than the efficiency of the market
for the securities themselves. It may well be that the empirical results indicate the private value (or
otherwise) of information (Ohlson, 1979, Patell, 1979) But only in the most unlikely of circumstances is it
possible that capital market reactions also indicate the social value of information or have implications
about the desirability of alternative accounting measures or disclosures. Work linking capital market
reactions and the value of information (e.g. Beaver & Dukes, 1972a, Ohlson, 1980) may eventually
provide guidance to accounting policy makers. Under present institutional arrangements (in particular,
those that give rise to the free rider effect, reformation asymmetry and market incompleteness) it is
however logically invalid to suggest that capital market efficiency tests can be used to assess the
desirability of alternative accounting measures or disclosures (Beaver and Demski, 1974; Gonedes and
Dopuch, 1974).
The desirability of alternative accounting measurement systems can only be assessed if the
objectives of the accounting function within society are made explicit (Ronen, 1979). And, if such
objectives are concerned with the efficiency of resource allocation and the distributions of wealth within
the economy as a whole rather than one particular market, then capital market research is of limited
value. Focusing on informational efficiency in the capital market may contribute towards an efficient
allocation of capital resources from the perspective of the shareholder class, but the resulting
equilibrium may not be efficient for other members of the economy.
Consideration also needs to be given to the important but frequently neglected question of
efficiency intra and inter all the mechanisms for allocating capital in the economy. Capital is allocated by
several markets, including the property and labour markets (for human capital) as well as by a number
of publicly owned institutions and planning authorities such as nationalized and regulated companies,
and national and local governments.
Within this wider set of allocation mechanisms, the choice of an accounting measurement
system becomes much more complex. These mechanisms are directly affected by the accounting
information produced by listed corporations Many allocators of capital resources other than the stock
market also have an interest in the choice of accounting measures and disclosures. Although in the
future it may be shown that what is "good" for the stock market is also "good" for these other allocators
of capital and hence the economy as a whole, it is a question that is rarely addressed in the literature.
CORPORATE REPORTS AND OTHER USERS
There have been a number of attempts to assess the use of published accounting reports by
external users other than shareholders, including employees (e.g. Cooper & Essex, 1977; Foley &
Maunders, 1977, Carlsson et al., 1978), lenders (e.g. Libby, 1979), tax authorities (e.g. Mace, 1977a;
Lawson, 1980), government (e.g. Enthoven, 1973; Gambling, 1974) and researchers (e.g. Gonedes &
Dopuch, 1979). Even if homogeneity within the user groups is assumed, conflicting objectives between
these user groups may result in different preferences for the content and form of accounting reports.
These conflicting objectives and hence preferences may be satisfied by publishing different accounting
reports directed towards each specified group (Revsine, 1973; ASSC, 1975), although the cost to the
economy in terms of resources used would be greater than if only one set of accounts were published.
The alternative approach is to design general purpose accounting reports (Ijiri , 1975; Mace 1977b)
which attempt to satisfy the preferences of all user groups. The difficulty with this approach is that any
inter-user conflict over preferences must be dealt with by a decision to choose one method of
accounting over another, i.e social choice. However, as has been demonstrated by Arrow (1951) for the
general case and by Demski (1973) for the case of financial reporting, it is logically impossible to make
social choices that are rational, reflect individual preferences and are not dictatorial. 6 In other words,
the recognition, per se, of multiple external users for accounting reports is insufficient for assessing the
overall economy-wide value of accounting reports. It fails because the approach lacks a theory about
how these users should or do interact with each other in the social and political environment. Such a
theory would need to articulate differing interests in society, their interactions and means of resolution.
Corporate reports in a contracting context
One of the more recent developments in theories of corporate reporting has been a shift from
an emphasis on the use of accounting in predicting variables "of interest" to a concern with the use of
accounting in contractual relationships between corporate stakeholders. As this development is well
summarised m Butterworth et.al (1981), only the main issues will be discussed in this paper.
Following Coase (1937) and Alchlan & Demsetz (1972), corporations have been viewed as a set
of rater-related contracts between participants. In this view a significant role for accounting reports and
information is as a monitoring device to record the behaviour of contracting parties involved. There are
several critical issues for the value of corporate reports Firstly what are the desired attributes of such a
monitoring system (eg Ijiri, 1975, 1981). Given the complex way the contractual system adapts to
"imperfections" in the monitoring system and the difficulty of identifying an optimal system, attention
has been placed on adaptive processes in the contractual relationship (Watts, 1977; Leftwlch, 1980,
Milne & Weber, 1981). It would appear, however, that this attention has been stronger in asserting the
existence of processes which result in perfect adaptation rather than in demonstrating how they
operate. Finally attention has been focused on the behaviour of the parties m the relationship and the
design of the reward systems to share the risks and returns derived from the interaction (Holmstrom,
1979, Harris and Raviv, 1978).
Applications of the contracting approach with its emphasis on agency relationships, have been
used to provide explanations for the development of GAAP (Butterworth et al., 1981) and the auditing
function as mechanisms for increasing the amount of disclosure above that which managers might
privately wish to produce (Watts, 1977, Ng, 1978) Similarly, the continued existence of institutions such
as the FASB in the U S and the ASC in the U K., and the regulation of accounting by government has been
explained in terms of public good characteristic of accounting reports and hence the social value which
the production of accounting reports can confer on the economy as a whole (Benston, 1976, Findlay,
1977, Foster 1980b) The identified deficiencies of these explanations (e.g. Leftwich, 1980) do not detract
from the insight derived from the approach.
Nevertheless, there seem to be at least two problems with this approach. Firstly it tends to
elevate markets to the status of an immutable and ideal benchmark. That is, markets are treated as the
standard by which other institutional arrangements are to be judged (Demsetz, 1969). Market failures
such as information asymmetry and non-excludability may be recognized but by assuming the perfect
adaptability and omniscience of market participants, other institutional possibilities are dismissed
(Leftwich, 1980). There is no recognition of the social, contrived nature of markets (White, 1981) or of
their historical specificity (Routh, 1975). Consequently this approach almost invariably reinforces the
existing market system or recommends reduction in intervention in market operations so that the
market can operate according to its logic. In short, the emphasis on market efficiency which is inherent
in this approach relies on the belief (derived from marginal welfare economics), that market efficiency is
a necessary condition for social welfare improvements. The problem with this belief, which is considered
further in the next section, is that it is based on extremely dubious assumptions (Graft, 1957) and an
untenable instrumentalist philosophical position (Tinker et al., 1982).
 The second problem with the contracting approach is common to all the approaches discussed
in this section A concern with "users" of corporate accounts (for decision making involving prediction or
for stewardship) may be able to address issues of private value but does not seem able to deal with the
social value of these reports. By focusing on one subset of participants in society -active market agents -
it ignores issues of social welfare which incorporate the well being of all members of society. It has been
recognised that the welfare of producer of accounting reports (Cyert & Ijiri, 1974) and corporate
managers (Watts & Zimmerman, 1978) may be affected by corporate reports. It has also been
recognised that accounting reports may themselves influence firms' financing and operating decisions
(Prakash & Rappaport, 1977; Heald, 1980, Butterworth et al., 1981) Yet all these developments cannot
avoid the problem of partial equilibrium approaches to valuing accounting reports, namely their failure
to model the total interaction between these reports and all individuals and classes in society.
SOCIAL VALUE APPROACHES
It is perhaps not surprising, given the claim of accountants to be professional (Sterling, 1974),
the increasing concern about the value for money of accounting reports in society (Briston & Perks,
1977), and the public concern about the accounting profession (U.S. Senate, 1976, Davidson, 1979), that
increasing attention has been placed on the social value of corporate reports. In this section we review
the general equilibrium approach to the economics of information and the analysis of economic
consequences. Both approaches attempt to understand and explain the production and use of
accounting reports from an economy-wide perspective and hence directly address the broader issue of
the social value of accounting information.
General equilibrium analysis
General equilibrium economic analysis (GEEA) typically involves the analysis of information in a
market context. It aims to identify conditions which result in economic efficiency in the allocation of
resources through time between all market participants. In particular, GEEA of information seeks to
identify the role of information in these allocations. Ohlson & Buckman (1981) synthesise much of the
literature on the welfare implications of public information within the framework of marginal
economics. Whereas Hirschleifer (1971) suggested that public information was socially useless, Ohlson
& Buckman (1981) demonstrate how this information will affect the sharing of risks in an economy and
so has welfare implications. The nature of these implications is not well specified, however, as they
depend on market arrangements and the set of available resources to be traded.
The failure to produce specific welfare implications is also evident in the analysis of the
production of private information. Hirshleifer (1971) and Demski (1974) have argued that there are
incentives for one individual to privately produce information (thereby consuming recources) in order to
make gains at the expense of another who does not have this information. Demski & Feltham (1976)
also suggest that it may be worthwhile for an individual to pay to suppress information if other
individuals may act on this information in an unfavourable way.
In short, then, the welfare implications of GEEA are unclear but depend on (exogenously
determined) markets, alternatives and initial resource endowments For instance, Hakansson (1981)
shows how groups that differ in initial endowments (eg. wealth) will take different positions on
accounting disclosures Further, the use of GEEA to explain the function of accounting in society and to
provide criteria for evaluating alternative accounting systems, is limited by its high level of generality
and abstraction. Without detailed specification of, for example, utility functions, cost structures of
information and production technologies, it is not possible to identify which forms of accounting or
institution are efficient Further, Ohlson & Buckman (1980) indicate the sensitivity of the conclusions of
the analysis to changes in the attributes of the model of the economy. For instance, they suggest that
the conclusions derived from a simple (analytically tractable) model may not apply for more complex
models (which includes such "complexities" as three time periods, heterogeneous expectations and
endogeneous information production).
Two further hmitations of GEEA relate to the use of marginalist (neo-classical) economics as the
theoretical basts of the analysts. The first general limitation relates to the exclusion of effectiveness
considerations. The unwillingness of marginalist economics to enquire into the factors that shape
preferences and motivate market demands (and the corollary of failing to consider demands which are
not supported by resources or are not expressed in a market) means that GEEA fails to distinguish
between efficiency, the relationship between resources used and outputs achieved, and effectiveness,
the relationship between the outputs achieved and the satisfaction of society's needs and expectations
(Pfeffer & Salancik, 1978)7 Whilst recent reviews demonstrate the theoretical disarray of much of the
social science literature which addresses issues of effectiveness (particularly that relating to
organizations -- cf. Kanter and Brmkerhoff, 1981, Goodman & Pennings, 1977) it is nonetheless evident
that political issues cannot be divorced from economic analysis in relation to social choices The
fundamental issues revolve around power and whose interests predominate in society In order to
address questions of effectiveness m relation to accounting reports, it is necessary to make explicit a
social welfare function which enables trade-offs to be made between individuals and classes of
individuals in society This is not to say that it is the responsibility of an accounting researcher (or
accountant) to produce a social welfare function, such functions should be articulated in the political
arena However, it is the case that an analysis by an accounting researcher (or accountant) that fails to
incorporate that arena has little credibility when discussing values and ' better " choices. Because GEEA
emphasizes the significance of individuals and avoids making inter-personal comparisons of utility, it is
not an appropriate method of analysts for facilitating the choice between alternative accounting
measurement systems which have differing distribution effects in terms of wealth and/or welfare
between classes in the economy.
The second limitation of marginal economic theory for guidance about designing accounting systems
relates to what might be termed its micro foundations Marginal economics implies that individual
preferences for accounting information should be treated as exogenous wants although it has been
suggested that such preferences may be influenced by learning through action (Sterling, 1972, March,
1978; Einhorn & Hogarth, 1981). Similarly, accounting researchers might also question the normative
and descriptive validity of the notion of rational choice assumed by marginal economic theory. That is to
say, individuals may not only be unable to behave consistently (Tversky and Kahneman, 1974) but they
may also wish not to do so (Slovic & Tversky, 1974). Thus, some of the seemingly innocuous assumptions
about rational choice that form the basis of much of the public choice literature (including Arrow's
Impossibility theorem) may indeed be contested.
These observations notwithstanding, it seems probable that developments in GEEA, as they
incorporate more "realistic" models of the economy, will not only produce more robust analytical
welfare implications relating to the efficiency of resource allocation in a capitalist economy but also will
provide some insight into the roles of both publicly and privately produced information in this process
However, such implications and insights will be dependent on the essentially value laden
conceptualization of the variables analysed Differing views of such variables as efficiency, effectiveness
and information would lead to a range of general economic equilibrium analyses, each offering differing
implications and interpretations. There are alternative approaches to the value of corporate accounting
reports which offer a better basis for informed debate and discourse than the currently dominant one of
a single analytical view derived from marginal economics and a particular conceptualisation of the
variables.
Economic consequences analysis
Economic consequences analysis (ECA) has recently emerged as an alternative approach for
understanding and valuing the role of accounting reports in a broader societal context In contrast to the
largely abstract approach of GEEA, the ECA literature tends to be empirical and seems to have the
potential for assessing a wider range of effects of changes in the accounting measurement system and
hence for understanding the social role of corporate accounting reports (FASB, 1978a) Any changes to
the status quo can, it is suggested, be assessed in terms of the "impact of accounting reports on the
decision making behaviour of business, government, unions, investors and creditors" (Zeff, 1978, p.56).
It seems unfortunate, however, that the "rise of economic consequences" (Zeff, 1978) seems to have
been motivated, at least in the United States, by a desire by large corporations to counter attempts to
change the existing reporting systems and levels of disclosure. To date, it would seem that accounting
researchers have generally reiterated the complaints of investors and businessmen about the
consequences of changes in required accounting practice (AAA, 1978).
As the 1980 Supplement to the Journal of Accounting Research illustrates, studies using ECA
have almost invariably evaluated the consequences of accounting reports solely in terms of the
behaviour and interest of the shareholder and/or corporate manager class (Selto & Neumann, 1981).
Many of the empirical studies have attempted to assess the stock market reaction to changes in the
content of published accounting information. For example, Griffin (1978) suggested that the market
reaction to SFAS 8 (Translation of Foreign Currency Transactions) was a consequence of the additional
currency hedging which management was likely to undertake to minimize the fluctuations in corporate
earnings. Similarly, the possibility that management of "full-cost" oil and gas corporations would reduce
exploration activity as a consequence of a switch to "successful efforts" accounting under SFAS 19 may
have explained the market reaction to such corporations (Lev, 1979). But as Foster (1980a) has
observed, the inconsistency of the results m the numerous tests of the market reaction to SFAS 19 is
indicative of a general failure in such tests to specify a theory of expected effects (and thence identify
necessary control variables). Vigeland (1977) found that the stock market did not react to SFAS 2 This
result may be interpreted as indicating that investors' expectations about managements' R&D decisions
were unaffected by accounting standards. Two disclosure requirements enacted by the SEC have also
been evaluated in terms of stock market impact: line of business (segmented) disclosure (Horwitz &
Kolodny, 1977; Collins & Dent, 1979), and replacement cost disclosures (Beaver et al, 1980, Ro, 1980).
Similarly, Morris (1975) found that the U K. stock market found little new information in early inflation
adjusted corporate reports. All these studies have relied on the efficient markets paradigm in order to
measure any impact of accounting changes on market price. Even allowing for the criticisms, discussed
previously in the first section of the paper, which this paradigm has recently encountered and for the
possibility that such studies can be interpreted as negating the efficient market hypothesis itself (Lev,
1979, p 501 ), the concentration on stock market reaction hardly fulfills the expectations established for
the ECA approach by May & Sundem (1976) who had emphasised the importance of obtaining evidence
of the total economy-wide impact of changes in published accounting information.
 There have also been a few studies, within the ECA framework, which have directly investigated
the impact of changes in accounting reports on corporate management decisions. Dukes et al (1980) and
Horwitz & Holodny (1980) investigated the impact of SFAS 2 on corporate R&D expenditures. The
inconsistent results of these studies may be due to differing samples but even when a reduction in R&D
expenditure occured, these studies are unable to detect whether there was a change in the nature and
distribution of R&D throughout the economy In short, the focus is on firm-level effects and not
economy-wide effects. Similarly, Wilner's experiments (1982), albeit in a laboratory, supported earlier
surveys which suggested that SFAS 8 affected managers' exchange risk decisions And Biddle (1980)
found that the LIFO-FIFO choice affected the size of inventories held by corporations The theoretical
justification for the effects identified in these studies tends to be based on some version of agency
theory (Jensen & Meckling, 1976; Watts, 1977). This theory suggests that corporate managers will have
an incentive, based either on direct changes in expected corporate cash flows (e.g. cost of capital, tax,
information processing, political or production costs) or on changes to the managers' own compensation
systems, to adjust their behaviour as a result of changes in accounting methods.
The relevance of these studies to an assessment of the social value of accounting reports can be
questioned on a number of grounds. Firstly, there has been very little modelling of the process by which
managers respond to accounting changes. The outputs, managerial attitudes and actions, are observed
in association with the inputs, the change in accounting reports, but no evidence is presented
concerning the way managers actually arrive at their attitudes or actions Thus, the whole issue of why
managers act, or have the attitudes they do, is subsumed by the rationality assumptions in agency
theory. Alternative explanations such as the nature of the information itself (Kelly-Newton, 1980) or the
role of corporate personality (Sorter & Becker, 1964) are not explicitly considered Secondly, these
studies have been unable to distinguish between operating decisions brought about by new accounting
standards and operating decisions which would still have been taking because of changes in other
factors (Ball, 1980).