The purpose of this paper is to find out whether corporate diversification provides a favourable environment for earnings management (agency conflicts hypothesis) or whether it mitigates this phenomenon (earnings volatility hypothesis).
Based on a sample of US firms and making an explicit distinction between industrial and geographic diversification, univariate and multivariate analyses are used to test whether firm diversification has an impact on earnings management.
Results show that the average diversified firm in the sample has somewhat more earnings management problems than a similarly constructed portfolio of stand‐alone firms chosen to approximate the segments of the conglomerate. Consistent with the agency conflicts hypothesis, the authors find that geographic diversification increases earnings management whereas industrial diversification decreases it, consistent with earnings volatility hypothesis. Moreover, industrial and geographic diversification combined reinforce this phenomenon. These findings are consistent with the view that the costs of geographic diversification outweigh the benefits.
The paper makes an important contribution to the accounting literature by providing new and significantly different evidence on the relative roles of corporate diversification in the earnings management. By linking two streams of research, earnings management and corporate diversification, one is taken into the unexplored area of the sources of the difference in earnings management between diversified and focussed firms. More specifically, this study provides evidence that earnings management is more intensively practiced in geographically diversified firms and even more so in firms that are both industrially and geographically diversified.
