Abstract We document and analyze the time-series and cross-sectional patterns in 6,920 voluntary and mandatory accounting changes made over the 1969-88 time period. We report on the types, frequency, and earnings effects of voluntary accounting changes (VACs), and the economic characteristics of firms making these changes. We also document industry/year clusterings in accounting changes and the relation between VACs and various macroeconomic factors. We similarly describe mandatory accounting changes (MACs), compare economic characteristics of early versus late adopters of MACs, and investigate whether a timing relation exists between the adoption of VACs and MACs (i.e., do firms make VACs to offset the impact of a MAC?). We structure our analysis of VACs around two perspectives of accounting choice: "managerial opportunism/earnings management" and "efficient (optimal) contracting." Motivation for VACs under managerial opportun- ism/earnings management includes post-contract opportunism to affect wealth redistributions between managers and stockholders or between bondholders and stockholders, and attempts to hide poor operating performance or to smooth reported earnings. Under efficient contracting, VACs are rational responses to changes in firms' investment opportunity sets. Thus, VACs are made to minimize contracting costs and to maximize firm value. Our study contributes to the accounting choice and change literature in several ways. First, we provide a synthesis of the issues surrounding accounting changes and consider their impact and pervasiveness. Second, by examining a broad set of accounting changes and by drawing out time-series and cross-sectional patterns of accounting changes, we enable researchers to identify more readily the types of accounting changes and is- sues offering the greatest opportunities (or further study. Third, we provide evidence on the efficient contracting and managerial opportunism/earnings management perspectives of accounting choice and change.
Pincus et al. (Wed,) studied this question.