Do high-profile investigations of corruption deter corruption in non-targeted firms: indirect evidence from reductions in real earnings management

We examine whether high-profile corruption investigations effectively deter corruption beyond those companies being investigated by examining Brazil’s investigations of its largest companies, known as Lava Jato (Car Wash). Corruption is unobservable unless detected, however, the way firms manipulate their financial reporting in order to cover up their corruption activities can be detected. As such, we begin by focusing on those companies targeted by the investigations, where corruption was found to exist by the investigations, to identify how companies cover up their corrupt activities. We examine both accruals earnings management (AEM) and real earnings management (REM) as candidates for the cover up, but find evidence supporting only REM. With the investigations’ launch, these companies halted their corrupt activities to avoid further prosecution, which, by necessity, required a reduction in the REM employed to cover up their corruption. This suggests that a reduction in REM in companies suspected of corruption, when the threat of exposure increases, would be indicative that corruption likely existed. Next, we test and find that other publicly-scrutinized companies not under investigation reduced REM with the investigations’ launch. Thus, providing indirect evidence of the effectiveness of a high-profile investigation to deter corruption when the likelihood of corruption is high.