Post

This @iojournal.bsky.social article from @lorenzocrippa.bsky.social et al. is one of the most important recent pieces on corruption, American politics, and democratic backsliding. It is an empirical research note, but I think its theoretical implications are much bigger. 1/n🧵

Since Persson, Rothstein, and Teorell’s important 2013 @govjournal.bsky.social article, corruption scholars have hotly debated whether corruption is better understood as a principal-agent problem or a collective action problem.

The distinction matters. Principal-agent approaches emphasize delegation, information asymmetries, monitoring, and sanctions. Collective action approaches emphasize assurance-game dynamics: actors may prefer a clean system, but they do not want to be the only ones playing clean: norms matter!

An implication of collective action approaches is that monitoring alone may fail when corruption is systemic. If actors expect monitors, firms, or officials to be corrupt, incremental reforms may not reassure them enough to shift behavior toward a cleaner equilibrium. Better = big-bang approaches!

At the same time, a lot of work shows that monitoring-based and incremental anti-corruption reforms can work. In particular, I'd highlight the now growing literature on audits as well as more reform-oriented work, such as the excellent work of @kbersch.bsky.social

After Trump suspended FCPA enforcement, investors rewarded firms with prior FCPA histories: markets believed lower enforcement made corruption-risk firms more profitable. In assurance-game terms, monitoring sustained the clean equilibrium; removing it shifted expectations toward the corrupt one.