On the Choice of Accounting Method for Partnerships

Abstract

The problem of an optimal choice of an accounting method for a partnership, as a model of group activity inferred from the theory of syndicates of Wilson [1968], was originally introduced to the accounting literature by Demski [1973]. A partnership is a collection of individuals who jointly engage in an income-generating activity and then cooperate to share the income that results. The distinguishing characteristic of a partnership's sharing of income is that it is assumed to yield a Pareto optimal allocation among members. Within the context of this model, Demski demonstrated that the selection of an accounting method may be viewed as a well-defined optimization problem, which, because of its formal analytical nature, eschews vague criteria espoused by accounting theory such as fairness and neutrality. 1 Through the convention of allowing an information system to be interpreted as an accounting method, Demski observes that there may exist an information system which is unanimously preferred by each member of the partnership despite a degree of heterogeneity among members which would invalidate the application of neutrality. In brief, Demski established that there may exist some system which is ex ante Pareto optimal; that is, its selection maximizes each member's