There is a vast body of literature on performance-related pay (PRP), with strongly held views from opponents and proponents. This study reviews this literature, disaggregating the available evidence by the different public sector contexts, particularly the different types of public sector jobs, the quality of the empirical study, and the economic context (developing country or OECD settings), with the aim of distilling useful lessons for policy-makers in developing countries. The overall findings of the review are generally positive across these contextual categories. In particular, the findings from high quality studies, based on a simple scoring method for internal and external validity, of PRP in public sector-equivalent jobs show that explicit performance standards linked to some form of bonus pay can improve the desired service outcomes, at times dramatically. This evidence primarily concerns "craft” jobs, such as teaching, health care, and revenue administration, apparently negating (at least in the short term) the behavioral economics concern about the crowding out of intrinsic incentives. The available evidence suggests that if policy-makers are sensitive to design and vigilant about the risks of gaming, then PRP may result in performance improvements in these jobs in developing countries. However, it is difficult to draw firm conclusions from the review about the effect of PRP in core civil service jobs for three reasons. First, there are very few studies of PRP in these organizational contexts. The work of senior administrators in the civil service is very different from that of many private sector jobs and is characterized by task complexity and the difficulty of measuring outcomes. Second, although some studies have shown that PRP can work in even the most dysfunctional bureaucracies in developing countries, there are few cases illustrating its effectiveness or otherwise outside OECD settings. Finally, few studies follow PRP effects over time, providing little information on long-term effects and adjustments in staff behavior. We conclude that more empirical research is needed to examine the effects of PRP in the core civil service in developing countries
