ACCOUNTING CHANGE BEHAVIOR: THE RELATIONSHIP BETWEEN EARNINGS ADJUSTMENT AND FIRM-SPECIFIC EXPLANATORY FACTORS. AN EMPIRICAL ANALYSI

The general purpose of this study was to provide evidence relevant to one general accounting question: what factors influence the choice of firms' accounting systems? Evidence was sought by observing changes in accounting procedures by firms. Two major empirical questions were addressed: (1) What factors explain how accounting changes are used to alter the level of reported earnings? (2) What factors explain how accounting changes are used to reduce fluctuations in or "normalize" reported earnings?;Two alternative but not mutually exclusive, views of management behavior were used to develop potential explanatory factors. An "economic" view, consistent with recently developed "positive theory" of accounting was originally adopted. This view is consistent with management belief in sophisticated capital markets and assumes accounting change behavior is influenced by the cashflow consequences of alternative accounting rules. Using this view, factors related to the exposure of the firm to political costs, the nature of the firm's lending agreements, the existence of bonus compensation contracts, and the percentage of stock ownership held by management were hypothesized as relevant in explaining accounting change behavior.;A "naive" view, consistent with management's belief in naive capital markets was also examined. This view assumes that change behavior is influenced by the earnings-measurement implications of alternative accounting rules. Using this view, factors related to the firm's past and current earnings realizations were hypothesized as relevant in explaining accounting change behavior.;The empirical investigation suggested that accounting change behavior may be influenced by political cost and lending agreement concerns, but the results with respect to these factors were not consistent across all tests and the findings were not unambiguous. The empirical investigation also indicated that accounting change behavior is influenced by bonus compensation arrangements and the firm's past and current earnings realizations. The results with respect to these factors were consistent and strong. Firms with bonus arrangements were found to adopt accounting changes that increased earnings and normalized earnings. Income-increases were also found to be associated with relatively poorer pre-change levels of earnings and poorer prior periods earnings growth. Normalization behavior was found to be positively related to large pre-change divergence of earnings from expectations.;Findings suggest that both cashflow and earnings-measurement implications of accounting alternatives may be relevant to explaining accounting procedure choice, but that variables related to the earnings-measurement implications of accounting changes are more strongly related to accounting change behavior.