How to Select Targets for Investigations? Evidence from Financial Reporting Enforcement

• Applicants: Professor Dr. Marius Gros; Professor Dr. Martin Nienhaus
• Subject Area: Management and Marketing
• Term: since 2022
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 509830657

Most financial reporting enforcement systems rely on investigations to foster compliance with rules. Different options exist how regulators can select targets for investigations: The most common approach is to initiate investigations based on a discretionary risk-based assessment or due to a specific complaint. Alternatively, regulators may use a randomized selection model or a combination of risk-based and random selection. However, to the best of our knowledge, no research so far has shed light on the consequences of different selection mechanisms in financial reporting enforcement; and even the broader public oversight literature is sparse on this topic. Hence, in this research project, we study the value of risk-based versus random selection mechanisms in financial reporting enforcement. Analyzing the selection mechanism is important because a critical element of enforcement systems is the ex-ante deterrence effect. Specifically, due to the fear of being investigated, caught, and penalized, firms have incentives to be compliant. We know from other fields such as public crime enforcement that the severity of the penalty is not the main driver of deterrence, but rather the perceived detection probability. A key determinant of this detection probability is the way firms are selected for investigation, and thereby firms' assessment of their ability to evade investigations. Hence, we believe it is critical to shed light on the different selection mechanisms. In theory, random selection mitigates that regulators may be biased in their selection due to corruption, laziness, or incompetence. This concern is aggravated in the context of enforcement as prior research suggests that enforcement institutions are captured politically. Moreover, with random investigations, firms must constantly fear being selected for investigations, while a risk-based approach may allow firms to develop evasion strategies once they understand the enforcers' selection criteria. However, an obvious advantage of risk-based selections is that they are more effective in targeting the ‘right' firms, and are thus more cost-efficient. Given the conflicting arguments, we believe that the value of risk-based versus random selection in enforcement is unclear, and we ultimately consider it as an empirical question. Analyzing this question is difficult because most enforcement institutions shroud their investigation process in secrecy and data is scarce. For those settings where data is available, we either do not have information on the selection mechanism or the regulator does not use random selection. We circumvent this issue by using a proprietary data set. Specifically, we obtain publicly unavailable data on enforcement investigations from the BaFin. The BaFin provides us with detailed data on all enforcement investigations of German public firms. Another important advantage of the German setting is that the selection model uses both risk-based and randomized investigations.

• DFG Programme: Research Grants