In this paper it is shown with the help of a small sample that accounting is ethically loaded; that there exists ethical frustration caused by situational factors related to accounting; and that most probably the situational pressures may also change the level of ethics in the Kohlbergian sense; and that in studying accounting-related ethical problems empirically, the paper-and pencil tests and interviewing may give biased results.

The accountant's model of the world is supplemented with many discretionary practices. The general structure of this model is determined by the duality of double-entry bookkeeping, the system of stock accounts and flow accounts, principles of valuation, allocation, and control. Financial accounting is strongly controlled by legislation concerning who is obliged to do accounting, what is to be accounted for, how allocations should be made, how valuations should be carried out, and what should be presented in public. But even in these procedures there are many discretionary practices, the use of which is guided by conventions and general ideals like, for example, by a 'true and fair view' or a 'good accounting practice'.

In the realm of managerial accounting, the scope of discretion is even greater. To some extent it is, of course, constrained by the demonstration effect of practices present in financial accounting [Johnson et al., 1987], but in principle the hands of the accountant are relatively free.

In this paper we present the results of our small-sample investigation into how the ethical dimension was handled in the daily work of some chief executives and heads of accounting departments. We will start by presenting a few theoretical concepts, which explain the way in which the study was conducted, and proceed by illustrating what our interviewees told us about their ethical problems relating to discretionary practices in accounting. Finally, an attempt will be made to draw some generalizations from our observations.