The authors empirically test whether earnings quality improves when International Financial Reporting Standards (IFRS) are adopted and when measures of investor protection (i.e., board independence, enforcement of security laws, protection of minority shareholder interests, enforcement of accounting and auditing standards, integrity of the judiciary, and tolerance of individual freedoms) are strong. They find that earnings quality improves when IFRS are adopted by a regime that has stronger investor protection. But only adopting IFRS or only having strong investor protection, rather than combining the two, does not appear to be sufficient to improve earnings quality.

Earnings quality requires both a means of transparently reporting accounting information and some level of investor protection in a given country. The authors posit that IFRS are the mechanism for transparency and that if there is sufficient investor protection (measured using various metrics), earnings quality will improve because managers cannot manage earnings to management’s advantage.

Understanding the relationship between these factors and earnings quality allows investors to make better investment choices by favoring investments when earnings quality is high or by taking risk-mitigating actions when earnings quality is poor.

Earnings quality is measured with discretionary accruals (current accruals less nondiscretionary accruals, which are found using current- and previous-year information). The authors then regress (fixed effects) discretionary accruals on a dummy variable for IFRS being adopted (1 if adopted and 0 if not), a variable for investor protection, and a set of control variables. The variable for investor protection is based on one of six possible measures: board independence, enforcement of security laws, protection of minority shareholder interests, enforcement of accounting and auditing standards, integrity of the judiciary, and tolerance of individual freedoms. The first five measures are evaluated on a scale of 1 to 7 by the World Economic Forum.

The firm-level sample data are from 46 different countries (not all countries are equally represented in the number of firms) between 2000 and 2007. The authors perform six different regressions, using a different measure for investor protection in each regression because the six measures have a high degree of correlation with each other.

In the regressions, the IFRS and investor protection variables are not statistically significant, which leads to the hypothesis that IFRS adoption and a strong level of investor protection are both necessary for better earnings quality. Performing the regressions again with an additional interaction variable—IFRS multiplied by the given investor protection variable—provides support for this hypothesis. In the second set of regressions, the authors find the interaction variable is generally significant at the 1% level and always with the correct sign. Finally, they perform a number of robustness checks to confirm the results.

The authors empirically demonstrate that earnings quality requires both a transparent means of disseminating information and an environment that protects investor interests. I am not entirely convinced that IFRS are necessarily the best means of information dissemination, but I am reasonably convinced that both pieces—effective information dissemination and investor protection—are necessary for improved earnings quality.