The authors ask whether shareholders benefit or react differently when managers are honest, direct, and take responsibility regarding poor company performance. Or does the potential for shareholder indifference, litigation, or reputational damage provide an incentive for managers to be vague and to blame others? Systematic problems, such as a downturn or a recession, affect companies broadly; thus, companies experiencing only idiosyncratic problems are considered. The companies that blame themselves tend to provide more details about the source of the problem and thus are able to correct the problem. Blaming others by using vague language may prevent companies from making timely and necessary changes.

Although the market may initially react negatively to companies communicating poor results to investors, the authors provide evidence that given time, the market highly values honesty. Poor equity market performance is observed for both the companies that blame themselves and the companies that blame others before the public disclosure. But the equities of companies that blame themselves demonstrate improvement in various performance metrics (such as profit margins and earnings growth rates) in contrast to the equities of companies that blame others.

Companies in the sample blame others in about two-thirds of the observations. The authors note that the sample time period includes the September 11 attacks and the financial crisis of 2008 as well as the onset of the Sarbanes–Oxley Act in 2002. Prior research is cited to explain why firms blame outside forces versus themselves, one of the reasons being the fear of litigation. The greatest number of companies that blamed themselves was found in the business equipment industry, followed by the money industry. The greatest number of firms blaming others also was found in the money industry because the money industry is the largest industry in the sample.

The authors are unable to find conclusive evidence that corporate governance can explain whether a company is more likely to assign blame for poor performance in an announcement. But the P/E of those that blame themselves tends to be higher than that of the companies that blame others (the only statistically significant governance variable).

The sample is constructed of companies that make public statements identifying the source of their problems. Only one observation per company is considered, which makes this research a study of situational honesty. First, the authors collect press releases that assign blame (using keywords) from the LexisNexis database for 1993–2009, producing 8,658 results. The sample is then scrubbed to remove releases that were duplicates, government issued, or third-party issued, and the authors retain only initial releases, resulting in 1,723 press releases. Next, the sample is filtered for companies that are traded on the NYSE, NASDAQ, and AMEX or the London, Toronto, Paris, or Tokyo exchanges. Data for each issuer must also exist in the CRSP and WorldScope databases. The resulting sample is 446 press releases. Finally, the authors assign either “blames self” or “blames others” to each press release and require that the designation be unambiguous, leading to a sample size of 150 press releases. The money industry represents 21% of the sample, business equipment represents 18%, and shops and other represent 15.33%.

The authors also consider whether well-governed companies are pressured by their boards to make announcements and whether such companies are more likely to blame themselves. Companies are matched based on market capitalization, book-to-market ratio, and risk-adjusted market performance for the prior month. A logit regression that uses various governance measures (obtained from the RiskMetrics Directors database) as independent variables is conducted on the sample and matched companies.

The results of this research are encouraging because one would hope that companies are incentivized to be not only self-aware but also honest when identifying and disclosing internal problems. It would be interesting to see whether these results hold with a larger dataset, beyond 2009, over multiple market cycles, and amid evolving regulatory environments.