Does firm-level political risk influence earnings management?

Abstract This study diverges from mixed findings in the literature on political uncertainty and earnings management by reporting a significant positive association between the firm-level political risk ( FLPR ) measure proposed by Hassan et al. (Q J Econ 134(4):2135–2202, 2019) and both accrual-based and real earnings management. This aligns with the predictions of agency theory and the political cost hypothesis, indicating that firms exposed to higher political risk are more prone to heightened earnings manipulation. Additionally, we find that in the face of increased political risk, firms tend to substitute accrual-based earnings management with real earnings management, which is relatively harder to detect. This study further identifies a non-linear ‘U’-shaped association between FLPR and both accrual-based and real earnings management, suggesting significant manipulation at both low and high political risk levels, with the least manipulation at a moderate level. This non-linear association is primarily observed in firms that are smaller in size, pay lower abnormal compensation to their CEOs and are less likely to be monitored by lenders. Thus, emphasising the role of external monitoring mechanisms in driving the non-linear association between FLPR and earnings management.