Title:

The mathematical and economic foundations of accounting measurement

In many respects, the present state of the theory of accounting measurement resembles that of probability theory before the path breaking analysis of A.N. Komolgorov. In accounting, as in "preKomolgorov" probability theory, there have been numerous attempts at providing an act of axioms for accounting measurement, all of which have either been ignored or subjected to varying degrees of criticism. By building on these prior attempts, the present thesis proposes an alternative net of axioms and then investigates its implications for accounting measurement in general. The unifying conception has been alluded to already. The thesis endeavours to show that the theory of accounting measurement is, in fact, grounded upon three axioms, and it is the specification of the information assumed given by these axioms, which is the source of many (if not all) of accounting's problems. The remainder of the thesis deals with the more important of these problems. Thus, chapter three concerns itself with the statistical estimation and identifiability of accounting measurement rules; chapter four, with the commonly encountered models (or interpretations) of the axiom system alluded to above; chapter five, with some numerical methods for estimating the replacement cost of asset disposals (a necessary piece of datum if we are to provide the axiom system with a replacement cost interpretation), whilst chapter six, relying on the capital theory of Irving Fisher, deals with the economic foundations of accounting measurement. There are two major conclusions which emerge from the study. Firstly, by summarizing the antecedent conditions which must be satisfied before it is possible to generate accounting measurements, the "axiomatic method" provides a useful framework from which to determine (and organize) the relative importance of measurement problems in accounting. However, much remains to be done if the method is to achieve its "ideal" function as a watershed or "clearing house" for measurement problems in accounting. Secondly, Irving Fisher's "capital theory" possesses far greater potential for accounting theory than has hitherto been realized. Specifically, by deriving Fisher's "investment opportunity locus" from first principles, as distinct from assuming it to be exogeneously specified, it is possible to provide an economic rationale for each of the measurement systems alluded to in chapter four.
