Our findings have a number of policy implications. First, while broad efforts to support democracy and erode authoritarianism have had limited effectiveness, more targeted support to journalists and media in closed and closing settings to report on corruption could help reduce corruption. Our research demonstrates that even in the face of severe restrictions, significant reporting on government corruption persists. Such an approach is ethically fraught—raising questions about the mechanisms for participant selection, for instance—but it might offer a more focused (and evaluable) approach to holding backsliding governments accountable.

Second, our findings also underscore the importance of exploring non-journalistic efforts (e.g., civil society organizations working on transparency and anti-corruption efforts) to detect corruption in repressive settings. Recent advancements in computational techniques, such as analyzing anomalies in trade data and government contracting, have shown promise in identifying corruption. While these methods are more advanced in developed countries, the increasing availability of high-frequency trade statistics and open budgeting make it entirely possible to detect egregious corruption without directly endangering journalists. The policy challenge in such cases is to both deploy the computational tools and disseminate the results in ways that promote good governance and accountability, and strengthen civil society, even in backsliding and authoritarian settings.

Third, our findings suggest the importance of focusing on restrictive media regulatory frameworks, which can have a more immediate and pronounced impact on corruption reporting than formal legal changes targeting the press. While formal legal changes often attract significant attention from development practitioners and media watchdogs, the less visible nature of regulatory frameworks may lead to their underestimation.

See Policy Brief for more detailed information