Varieties of connections, varieties of corruption: Evidence from bureaucrats in five countries

Why do some bureaucrats engage in corruption for personal gain, yet others for political gain? We show that these forms of corruption frequently do not coincide and offer an explanation: bureaucrats hired based on political and personal connections have different identities and incentives which compel them to engage in corruption for political and personal gain respectively. List experiments with a unique sample of 6400 bureaucrats in five countries in Africa and Asia support our argument. As theoretically expected, effects are strongest for bureaucrats whose political patrons remain in power (for corruption for political gain) and who do not need corruption gains to sustain their households (for corruption for personal gain). We also find that personal connections matter more than political connections for bureaucratic recruitment across surveyed countries. Our findings underscore the importance of studying varieties of bureaucratic corruption and of supplementing the politicization literature with studies of personal connections in bureaucracy.