Managers’ unethical fraudulent financial reporting: The effect of control strength and control framing
- Kelvin X. Liu
- Arnold M. Wright
- Yi-Jing Wu
Journal of Business Ethics
In response to numerous recent cases involving materially misstated financial information arising from fraudulent financial reporting, companies, auditors, and academics have increased their focus on strengthening internal controls as a means of deterring such unethical behaviors. However, prior research suggests that stronger controls may actually exacerbate the very opportunistic behavior the controls are intended to curb. The current study investigates whether the efficacy of an implemented control is conditioned on not only the strength of the control (weaker or stronger), but also on how the firm frames the purpose for implementing the control (e.g., monitoring or coordinating). A monitoring purpose frames controls as reducing managers’ opportunities to engage in self-interested behavior, while a coordinating purpose frames controls as facilitating coordination between the firm and its managers. We posit that the efficacy of stronger controls to reduce unethical fraudulent reporting depends on the control frame. Using an experiment, this study investigates the interactive effect of control strength and control frame on managers’ fraudulent reporting decisions. As predicted, our results show that when controls are framed for monitoring purposes, stronger controls result in less fraudulent reporting than weaker controls. Conversely, when controls are framed for coordinating purposes, stronger controls result in more fraudulent reporting than weaker controls. Our results suggest that an inconsistency between the firm’s choice of the control strength and the control frame reduces the efficacy of the implemented control to curb unethical reporting behaviors. Furthermore, supplemental analysis shows that managers’ rationalization helps explain the interactive effect of control strength and communicated control purpose on fraudulent reporting.