Accounting Research and Theory: The Age of Neo-Empiricism
Gaffikin, Michael, Australasian Accounting Business & Finance Journal
The theorising in accounting prior to 1970 was rejected as not providing sufficiently general theories. Informed by theories in economics and finance (and other disciplines such as psychology) and with the aid of computers, attempts to theorise accounting took a new direction. Large data collection and analysis emphasized a purportedly more systematic empirical approach to developing theory.
Key words: accounting; neo-empiricism; capital markets research; behavioural finance; efficient markets hypothesis; positive accounting theory
Around 1970 there was a dramatic change in the approach to accounting research. Several reasons have been suggested for this change in methodological direction by those reviewing the development of accounting thought. To many, a major distinction is a change in direction away from attempts to prescribe a theory of accounting to developing theory from a description of extant practices. To advocates of the latter, previous attempts to develop a theory of accounting were futile as there could never be agreement over many of the inputs into a theory such as the postulates, principles but most specifically the assumptions. Although a very inaccurate description the two approaches are labelled normative (the prescriptive theories that dominated prior to 1970) and positive (the descriptive research that has dominated mainstream accounting research since 1970).
With its emphasis on description, the most defining characteristic of mainstream research since 1970 is its commitment to empiricism. In their book on accounting theory, Henderson, Peirson and Brown (1992) refer to this research as neo-empirical research: a most apt nomenclature. As mentioned, the dominating characteristic was empiricism. It is "neo" (new) because, although earlier research had relied on empiricism in that it sought to establish "theory" from best practice, the emphasis after 1970 was on a more systematic use of empirical evidence. This was largely made possible with the availability of large financial databases to which sophisticated statistical techniques were applied to test hypotheses. This, in turn, was greatly facilitated by the increasing availability and use of computers.
All neo-empirical accounting research has the underlying assumption of the efficiency of markets - the efficient markets hypothesis (EMH). It is referred to as an "hypothesis" because despite more than forty years of research designed to test the hypothesis all attempts to date have failed to confirm it. Therefore, consistent with the generally accepted process of theory construction which states (simply) that a theory is a confirmed hypothesis, it remains an hypothesis. The EMH emerged in the 1960s from the work of researchers at the University of Chicago trained in economics and finance and working in the areas known as portfolio theory and employing the capital assets pricing model (CAPM). It was then taken up by accountants also working and studying at the University of Chicago and, as stated, it has been the cornerstone of a considerable amount of research over the last forty years. The EMH is an assumption about the relationship of information to security prices. The research area is usually referred to as capital markets research but the EMH has had implications for other research areas as well. Two Australians working at the University of Chicago, Ray Ball and Phillip Brown, are regarded as the first to engage in capital markets research in accounting and their work (Ball and Brown, 1968) has had a vast number of citations. Another seminal work was that by Bill Beaver (1968).
There are various forms of market efficiency which, however, only discriminate on the degree of efficiency similar to how economists describe how perfect a market is: perfect or imperfect etc. Nevertheless, in the literature reference is made to three types of EMH, the strong form (markets very efficient), semi-strong and weak (markets not very efficient). …