This study explores the relationship between four measures of earnings quality and
investor-protection, hypothesizing that favorable values of each earnings attribute
(considered individually) occur in countries whose institutional characteristics provide
relatively strong investor-protection. The results, based on K-means cluster analysis
between institutional characteristics and earnings attributes, are mixed. When earnings
quality is measured based on earnings smoothness, the results are consistent with the
study's hypothesis since earnings is less smooth in countries whose institutional
characteristics are strong. The results for accruals quality and earnings predictability are,
however, inconsistent with the hypothesis, since countries with strong investor-protection
have less favorable values of these measures than weak investor-protection countries.
Finally, no relationship is evident between investor-protection and earnings persistence.
Thus no clear conclusions can be drawn about how investor-protection affects earnings
quality. The regression results are similar to those of the cluster analysis.
There are several possible reasons for the mixed results obtained here. Although the
earnings attributes have been widely used in empirical studies of U.S. data, they may not be
well-specified when applied to international data. Second, variation in firm size and industry membership across countries may lead to substantially different incentives and
opportunities for earnings management in these countries. Finally, factors other than
investor-protection may influence earnings quality. One such factor is the extent to which
stakeholders rely on accounting reports in their decision making. Accounting numbers are
extensively used for contracting in the developed countries, and managers have more to
gain through earnings management. This contracting role of accounting numbers may
explain why earnings quality, measured by accruals quality and earnings predictability,
appears to be relatively low in the developed nations.
The findings of this study provide some, though limited, insights into cross-country
differences in earnings attributes and the linkage between these earnings attributes and the
protection of investors' rights. One important implication of our study is that conclusions
about earnings quality depend critically on how earnings quality is defined. Another
possible implication is that factors other than investor-protection may significantly
influence the characteristics of international accounting data. Thus investors and other
financial-information users should not assume that a country's investor-protection features
necessarily signal the reliability of accounting reports issued by its firms. Finally, our results
suggest that commonly-used earnings attributes such as accruals quality and earnings
persistence may not convey earnings quality in an international context. Future research
that develops globally appropriate earnings-quality measures is necessary.