Since the early 1990s, we have witnessed a dramatic rise in the number of anti-corruption agencies (ACA) – separate, specialised institutions providing core leadership in anti-corruption activities – that are clustered regionally in West Africa, South East Asia and the Caribbean. Following the successful cases of Hong Kong and Singapore, more than 80 countries have established anti-corruption agencies to date. However, the literature finds them largely to be ineffective in curbing corruption, due to weak mandates, the lack of independence from the executive, the lack of political will or insufficient resources.
This raises three important questions: first, why do states adopt a seemingly ineffective institution? Second, why do they make this specific institutional choice, when other less costly (especially with regards to the political constraints ACAs can impose) options, such as strengthening conventional law enforcement mechanisms, are available? Third, given that corruption is seen as culturally specific, what might explain the uniformity in the adoption of this institution?
The literature has so far provided little evidence for the dynamics driving the diffusion of anti-corruption agencies. This paper fills this gap. Building on existing theories of policy diffusion, we theorise that the diffusion of anti-corruption agencies may be directly driven by donor coercion and norm entrepreneurs or indirectly through neighbourhood competition and norm emulation. We also ask whether a strong civil society demanding the need for an ACA effects the establishment of an institution, as well as a change in political leadership. We use survival analysis on a new global data set on anti-corruption agencies to test these different theoretical accounts.
