Why do some local government bureaucracies perform well, delivering goods and services to all citizens equally, while in other cities accessing government services is a difficult task, with the result that those with political contacts are better able to take advantage of government services than others? This project took an important step towards understanding why governments perform as they do by developing a theory of government performance and collecting data on a direct measure of that performance in the implementation of a food distribution program in Argentine cities. The existing literature on government performance has not directly addressed the question of why there is variation in the quality of government performance. Whereas some authors assume politicians always seek to create high-performing bureaucracies, others take the opposite view and assume that politicians prioritize the development of clientelist networks or party machines through the strategic provision of government services.
In this project, the principal investigator developed a theory of government performance that explicitly treats the quality of performance as a political choice. It met the need, highlighted by other researchers, to identify testable hypotheses concerning the conditions under which we should expect to see better or worse government performance. It further innovates by analyzing the behavior not only of individuals potentially enmeshed in patron-client networks, but also of those outside such networks, and therefore highlights the trade-offs in support from different sectors of the population that poor government performance entails. Furthermore, whereas most existing literature focuses on general election competition only, the theory developed here also examines how intra-party competition, in the form of primary elections, may exercise a separate effect on government performance.
Empirically, there are very few direct measures of government performance in the existing literature. The theory developed in the dissertation was tested on an original dataset of the quality of Argentine city government performance in social welfare implementation. This dataset of local government practices was developed from an original questionnaire administered to the head of social welfare policy in over 100 cities across three different provinces broadly representative of Argentine social and political conditions. The research employed statistical analysis of questionnaire responses, supplemented by selected case studies, to test for a relationship between the nature of political competition and the quality of government performance. The questionnaire can serve as a template for the collection of similar data in different country and regional settings.
After nearly two decades of neo-liberal reform and decentralization in Latin America, local governments play an increasingly important role in the lives of citizens, especially for the region's poor who rely on government support to meet many of their basic needs. The findings of this dissertation will improve our understanding of why some local governments perform better than others and thereby allow for a better assessment of the likely consequences of future waves of decentralization in Latin America and elsewhere in the developing world.
