Book excerpt: Decentralization has been advocated by donors and development agencies as an important factor broadening citizen participation and improving local governance, thereby, promoting poverty reduction from the bottom up. On the basis of a comprehensive review of 19 country case studies documented in the literature, this paper questions this assumption. The authors find that an unambiguous link between decentralization and poverty reduction cannot be established. In some of the poorest countries characterized by weak institutions and political conflicts, decentralization could actually make matters worse. Interestingly, the poverty impact of decentralization would appear to depend less on the physical country setting, for example a country's size or quality of infrastructure, than on the capacity and willingness of policy makers to ensure a pro-poor devolution process. Two important policy lessons emerge from this study. First, in an environment where the central state is not fulfilling its basic functions, decentralization could be counterproductive and, therefore, should not be a donor priority. Secondly, in countries that are fulfilling their functions, decentralization could be a powerful tool for poverty reduction, improving representation of the poor and better targeting of service delivery. To fully reap the potential benefits of decentralization, donors' intervention in these countries should focus on providing technical support and improving the co-ordination of their aid policies at both the local and national level.