Paper short abstract:

Theory and evidence on the impact of decentralisation on corruption remain mixed. This paper presents analytical approaches to incorporating political drivers of decentralisation and corruption into understandings of decentralisation and its relationship with corruption.

Paper long abstract:

The theoretical literature and the empirical evidence on the impact of decentralisation on corruption remains inconsistent. Some models argue that decentralisation should reduce corruption through more accountable and transparent local governance, while others suggest more autonomy and actors increase opportunities for corruption. At the same time, empirical studies find that the relationship between decentralisation and corruption is far from clear.

Interestingly, these findings echo inconsistent evidence on the impact of decentralisation on development outcomes such as services and conflict management. This ambiguity is in part due to inadequate specification and understanding of both decentralisation and corruption. It also has roots in a gap between the theoretical benefits of decentralisation and variations in the implementation of decentralisation policies as expressed in the autonomy enjoyed by local governments, the functioning of intergovernmental relationships, local accountability, and the coherence of reforms across political, administrative and fiscal dimensions.

Similarly, studies often do not disaggregate corruption. Corruption involves different organisational dynamics: for example, between 'top-down', in which higher level officials buy in lower ones, and 'bottom-up' in which rents collected locally are distributed upwards. Regional and global factors such as organised crime and economic interdependencies also shape corruption at both national and local levels.

Both these qualities of decentralisation and dynamics of corruption are shaped by the political drivers behind and within decentralisation processes, and the nature of political settlements that influence patterns of rent-seeking within a given context. This paper presents analytical approaches to better incorporating these factors into understandings of decentralisation and its relationship with corruption.