7. Conclusion
Our approach in this review is to define earnings quality broadly to be decision usefulness—in any decision by any decision maker. Thus the number of relevant articles is over 300, and by necessity our discussion is broad.
We emphasize two significant conclusions based on our survey of the earnings quality literature as a whole. First, because all of the proxies for earnings quality that involve earnings (i.e., properties such as persistence, timely loss recognition, smoothness, and small profits, as well as the ERCs) have at their core the reported accrual-based earnings number, these proxies are affected both by the firm’s fundamental performance and by the measurement of performance. Future research could more clearly recognize the distinction between performance and performance measurement when making predictions and evaluating results. This would help us as a profession to determine the contribution of the accounting measurement system to the quality of reported earnings.
Second, although all of the proxies based on reported earnings are affected by both fundamental performance and its measurement, the proxies are not equally affected by these two factors. Therefore, the proxies do not measure the same underlying construct. In addition, because the proxies focus on different elements of decision usefulness, we should not expect the proxies to work equally well in all circumstances investigated by researchers. We hope the breadth of the discussion of each proxy has shed light on the context-specific dimensions of quality captured by each proxy and on the sometimes subtle distinctions between them.
As part of our review process, we noted five areas of research that have received relatively little attention; we believe that further research on these topics would substantially enhance our understanding of earnings quality.