Measuring Firms’ Investment in Accounting Resources : Implications for Malfeasance Versus Optimal Incompetence

Despite acknowledgement in the literature that low quality earnings (i.e., those that are subsequently restated) arise not only from managerial malfeasance but also underlying differences in firm fundamentals, many studies focus on the former and relegate the latter to the generic rubric of “innate” factors. We seek to expand this black box by studying one specific mechanism, which is firms' investment in accounting resources. This represents investments in hardware, software, and accounting staff that focus on the accounting function. We provide a framework that illustrates how this cost-benefit tradeoff is influenced by firm characteristics, and we demonstrate that restatements and these innate factors are linked because earnings quality arises endogenously from firm-value maximizing investments in accounting resources. We test our framework empirically and propose two proxies for the firm's (unobservable) investment in accounting systems: filing timeliness of the 10-K and spelling errors in the 10-K. In addition to uncovering correlations consistent with our framework, we document that our composite measure of accounting resources is negatively associated with the likelihood of a restatement in the subsequent year. We apply these insights to prior work, in particular the role of the business cycle in accounting fraud. We find that intertemporal variation in the frequency of accounting restatements is correlated with similar patterns in firms' investments in accounting resources. This evidence raises the alternative explanation that the alleged role of the business cycle on accounting fraud can be driven by time-series variation in accounting resources. Our results are robust to controlling for the endogeneity of accounting resources using an instrumental variables design that exploits regulatory changes in firms' filing deadlines