Introduction

The advent of corporate governance has rejuvenated interest in the effectiveness of systems and procedures by which enterprises and their managements are held accountable. The design of management control systems therefore becomes of increasing interest to an enlarged audience of stakeholders. As a consequence, the managerial behavior induced by control systems has resulted in public debate. The troubles at Baring Brothers, the Shell Transport forward contract fiasco, and others can be interpreted at a micro-behavioral level (Macintosh, 1994) as rational, legitimate responses to the systems employed in these companies. A more traditional interpretation indicates dysfunctional behavior.

Beginning with the premise that all human beings wish to show themselves in the most favorable light, it is inevitable that providers of information will attempt to manipulate reports to suit their own purposes (Prakesh and Rappoport, 1977). “Earnings management” is one critical and controversial area where manipulation can occur. Reported income may not reflect true economic achievement, but instead an apparent and misleading signal which is nevertheless consistent with the message the superior wants to hear (Merchant and Rockness, 1994). The means of managing earnings is informed in this research by six broad categories of dysfunctional behavior (Birnberg, Turpolec, and Young, 1983), namely smoothing, biasing, focusing, gaming, filtering, and “illegal acts.” A scenario for each category was presented to managers who responded on a scale of 1–5 as to the acceptability of a subordinate manager's action (Bruns and Merchant, 1990). The main research enquiry is to discover whether a consensus emerges as to the acceptability or otherwise of subordinate actions. The corollary is an examination of the tolerance levels different managers display to dysfunctional behavior.