“…Theoretically, a rational government official will choose to be corrupt, if the benefits of corruption exceed the costs, where the costs is estimated as the probability of being caught times the penalty for being caught (Klitgaard, 1988). Hence, local political corruption level is affected by factors related to the benefits of corruption (e.g., local government revenue, number of regulations that can be subverted), local people's ability and incentive to monitor (e.g., the strength of local Freedom of Information Act, local media coverage of politics, local education level, local religiosity, local wealth level, local voter preference heterogeneity) and government official accountability (e.g., the appointive power of state governor, voter turnout in elections, government decentralization level) (Meier and Holbrook, 1992;Fisman and Gatti, 2002;Glaeser and Saks, 2006;Campante and Do, 2014;Cordis and Warren, 2014;Niu et al, 2022). Given that most of these factors are long-lasting, we think that corruption is likely to be long-lasting and therefore, firms may want to take actions to create a long-term sustainable wedge between the reported numbers and the actual numbers.…”