Abstract Consciously or unconsciously, all choices are made by testing alternatives against a predetermined set of criteria. There has been little discussion in the accounting literature about the criteria to be used in choosing an accounting model, except to say that the resulting information should be relevant and reliable. This paper goes beyond that, by formulating seven criteria an accounting model should satisfy if it is to be preferred over competing models. The criteria are virtually axiomatic, and it is shown that our present body of generally accepted accounting principles fails to meet any of them. The criteria point, in fact, to the choice of an accounting model based, not on historical cost, but on ‘value to the business,’ i.e. deprival value. The paper also discusses the debate, which has been going on for at least fifty years, between proponents of matching costs and revenues as the basis of income measurement and those who think that income should be measured by measuring changes in net worth, thus giving conceptual primacy to the balance sheet and changes in it. The superiority of the balance sheet approach is demonstrated, and the paper shows how the liability for and the expensing of abandonment costs, which will be incurred at the termination of a project such as the exploitation of a natural resource such as an oil field, can be accounted for more elegantly by means of a balance sheet approach than by a matching approach, contrary to what has been argued in the literature.
David Solomons (Wed,) studied this question.