Assessing different forms of accounting theory

Category: Accounting

The systematic recording, reporting, and analysis of financial transactions of a business. The person in charge of accounting is known as an accountant, and this individual is typically required to follow a set of rules and regulations, such as the Generally Accepted Accounting Principles. Accounting allows a company to analyze the financial performance of the business, and look at statistics such as net profit.

The definition suffices to show, broadly, what accountants do (i.e., keep records of financial transactions) and why they do it (i.e., to enable people to analyse an organisation's financial performance. It also indicates that there are generally accepted accounting principles. However, the use of the phrase generally accepted suggests that some authorities do not accept them. In other words, opinions differ. Hence there are different theories of accounting.

All theories of accounting draw upon economics-accounting is about money. However, some theories also draw on political science, psychology, and sociology. There can also be an ethical element to theories of accounting.

This essay describes some of these theories, and explains how and why they differ. First, however, it explains what is meant by the term theory.

What is theory?

The Encarta Dictionary (English UK) provides five definitions of theory, three of which are irrelevant to this essay (2: speculation; 3: idea formed by speculation; 4; hypothetical circumstances), namely:

1: Rules and techniques. The body of rules, ideas, principles, and techniques that applies to a subject, especially when seen as distinct from actual practice.

5: Scientific principle to explain phenomena. A set of facts, propositions, or principles analysed in their relation to one another and used, especially in science, to explain phenomena.

The first Encarta definition, as regards accounting theory, suggests that an accounting theory should describe what accountants should do. Such a theory would be both prescriptive and proscriptive, and as such should draw largely from economics, and maybe ethics. The fifth Encarta definition suggests an accounting theory should explain what accountants actually do. Such a theory would be ethically neutral, and would draw on economics and whatever other discipline the theorist believed relevant-psychology, sociology, political science, for example.

The fifth Encarta definition of theory is close to Popper's (1963) view of theories. Popper argued that theories evolve to solve problems. They are never shown to be "right". They are only shown to be wrong. The best theories therefore are the ones that explain the most yet are never shown to be wrong, despite many people trying to show they are wrong. Good theories, said Popper, are capable of being falsified.

Types of theory of accounting

As indicated, there are two definitions of theory relevant to accounting-those that are prescriptive and those that seek to explain-and these different definitions give rise to different types of theory. In accounting the former are called normative theories, the latter positive (Gaffikin, 2005). There is a major problem with normative theories: it is impossible, in principle, to determine the best normative theory.

This is not a theory. It is a theorem. It comes from Demski (1973). Stated informally, it states:

No set of standards exists that will always rank alternatives in accordance with preferences and beliefs-no matter what these preferences and beliefs are, as long as they are consistent in admitting to the expected utility characterization. (p. 721)

The theorem therefore demands that, even if one had a single, perfectly rational, supremely intelligent accountant, the accountant could not devise any principle of accounting in which it could be shown that some chosen method of bookkeeping (e.g., FIFO) was always preferable, by the accountant's own criteria of what was preferable, to some alternative method. Thus the theorem suggests, not only that any body (e.g. the IASB) that tries to impose universal standards will fail to satisfy some users of the standards on at least some occasions, but also that the body will, on occasion, fail to satisfy itself.

Accordingly, this essay looks only at positive theories. It describes two: positive accounting theory (PAT) and Legitimacy theory.

Positive accounting theory (PAT)

PAT was developed by Watts and Zimmerman (1978). Gaffikin (2005) describes its main features:

In PAT the firm . . . is described as a collection of contracts-a nexus of contracts. Contracts are necessary in order to get self seeking individuals to agree to cooperate. . . . However, there will be contracting costs associated with the contracts . . . . PAT holds that firms will seek to minimize the contracting costs and this will affect the policies adopted, including the accounting policies. (p. 9; emphasis original)

The theory, in that it aims to explain how accountants behave, also makes predictions. Thus the theory, in that it can be tested (i.e., falsified), is "scientific" in Popper's (1963) sense of the term. The predictions of PAT are based on three hypotheses: the bonus plan, the debt covenant, and the political cost hypotheses (see, e.g., Gaffikin, 2005).

The bonus plan hypothesis

The bonus plan hypothesis derives from agency theory (Jensen and Meckling, 1976). Agency theory states that agents have different interests from those that employ them. Thus, for example, company directors (agents) have different interests from shareholders (their employers). Shareholders are interested in high dividends and high prices of the shares they own. Company managers are interested in other things-high salaries, comfortable offices, "status", and so forth. The theory is not new. In essence, as Jensen and Meckling admit, it was appreciated by Adam Smith, writing in the 18th century. Adam Smith observed that people in positions of authority tend, to greater or lesser extent, to "cheat" the system. The novel part of agency theory is that cheating carries costs, and policing cheating carries costs, too.

Some forms of cheating are subtle. Gaffikin (2005) notes, for instance, that shareholders might favour some high-risk activities, because, on average, they may be highly profitable. However, directors will tend to dislike high-risk activities because, if they fail (which some, by definition, will) the directors will get the blame, and thus lose their jobs or, at minimum, their bonuses.

The debt covenant hypothesis

The bonus plan hypothesis suggests not only that managers will prefer activities that maximise their bonuses, but also that they will prefer accounting practices that maximise their bonuses. One way of doing this is to move earning from the future to earnings from the present. This is the debt covenant hypothesis (Gaffikin, 2005)..

The political cost hypothesis

The political cost hypothesis states that organisations are sensitive to the political implications of their performance. For example, when oil companies make record profits, there may be calls for lower petrol prices, or higher taxes on oil companies. Thus, in the wake of high oil company profits in 2004 and 2005, oil companies in some countries switched their accounting practices to LIFO valuations, which tend to minimise profits (Gaffikin, 2005).

Criticisms of PAT

There are three major criticisms of PAT. First, it assumes all participants in organizational behaviour are rational. This is not necessarily the case. Second, although PAT purports to be objective, it cannot empirically justify its choice of assumptions. Therefore it is not truly objective (Gaffikin, 2005). Third, despite its claim to be an empirical theory, evidence in support of it is scant. Thus, for example, Sterling (1990) writes:

Although it would not be proper for me to infer anything from zero data, I must confess the suspicion that the absence of a report is due to failure to find even one instance of support for PAT. Until something beyond W&Z's ipse dixit is forthcoming, I will remain massively sceptical that such support exists. (p. 126)

In this connection, Mouck (1992) argues that PAT, despite its claims to be scientific, lacks substance.

Systems-oriented theories

Systems-oriented theories are like PAT in that they aim to describe and explain the behaviour of accountants. However, unlike PAT, they do not assume accountants, company managers, or shareholders are rational. Instead, they seek to explain corporate behaviour, including that of accounting, in socio-political terms. Thus some systems-oriented theories are political, others are more social (Godfrey et al., 2003). Legitimacy theory is more social

Legitimacy theory

Legitimacy theory takes as a starting point that corporations wish to be perceived as ethical-as "legitimate"-even if they are not. The theory makes no claim as to what, ultimately, is "legitimate", because ethical standards change. Thus, for example, many corporations boast about being environmentally "aware". Such a boast, however, appears recent.

Related to these criticisms, it arguably does not describe how organisations behave. Legitimacy theory suggests, in contrast to PAT, that organisations are predominantly concerned with being perceived as legitimate-that is, honest. In this, legitimacy theory suggests that corporations exist, not so much to make money, but to change the world for the better. Corporations have a "mission".

Legitimacy uses Max Weber's ideas of substantive rationality. Weber (1864-1914) was a pioneering sociologist and political philosopher who argued that societies are formed largely as a consequence of the institutions that comprise them. In this regards, ethically, institutions may choose from a range of values and actions and may seek to make them consistent. Of course, different societies, or the same ones over time, may choose different values.

Legitimacy theory impacts upon accounting because, in seeking to appear legitimate, corporations can report on the "good" things they do. Thus, Deegan and Unerman. (2006) report of a number of accountancy studies that have tried to test legitimacy theory by investigating company accounting practices that do publicise the "good" things that companies engage in. Thus, for instance, Cadbury's Get Active campaign publicized Cadbury's philanthropy to schools. Similarly, environmental impact reports are used as standard by FTSE 100 companies. (Godfrey, 2006)

Criticisms of legitimacy theory

The major problem with legitimacy theory is that it makes predictions that are indistinguishable from those of the political cost hypothesis in PAT. Another problem is that, in that it makes no assumptions of market efficiency, it is even less open to refutation than PAT. A third criticism is that it is one thing to show that companies seek to appear legitimate. It is another thing to show that they genuinely are legitimate.

A final criticism, one that can be made against any positive theory of accounting, is that such theory has had little impact on how accountants actually practice their trade. As Inanga and Schneider (2005) report, academic researchers have different interests from those of practising accountants. Thus much accountancy theory-including legitimacy theory-may amount to little more than abstract research, little related to the real world.


Accountancy theory seeks to do two things: provide guidelines to practitioners and to describe and explain what practitioners do. The former, in an absolute sense, cannot be done, for it is a theorem that no single system of accounting suits all circumstances. The latter appears problematic because there does not seem to be an agreed frame of reference for accounting theory. That is why there are many theories of accounting.