Description
Since the mid-1990s, anti-corruption initiatives have gained prominence on the international agenda, breaking the so-called "corruption taboo" in policy circles and emerging as a critical political issue in both developed and developing countries. The early 2000s witnessed a surge in anti-corruption parties, policies, and movements, while protests against corruption have become increasingly common since the late 2010s. However, the underlying drivers of these protests remain underexplored. This paper contributes to the growing body of literature on anti-corruption mobilizations by conducting the first systematic time-series cross-sectional study of anti-corruption protests. Leveraging multiple Big Data sources, we develop regression models to examine the determinants of corruption-related protests. Protest data is sourced from Armed Conflict Location and Event Data (ACLED), with corruption-related protests identified through keyword analysis and classification models. Our explanatory variables include metrics of change in the quality of political institutions and economic output, as well as shifts in corruption perception indices, Google Trends data on corruption-related keywords, and event data on corruption investigations and accusations from the Integrated Crisis Early Warning System (ICEWS). Our findings suggest that fluctuations in perceived corruption levels significantly predict an increase in corruption-related protests. Moreover, consistent with existing research, we find that grand corruption—linked to high-level government officials—serves as a much stronger protest trigger than petty corruption.
