Doctoral Dissertation Research in Political Science: Institutions or Culture or Both? A Comparative Study of Government Performance

University Of Pittsburgh, Pittsburgh PA

Investigators

Abstract

Why some governments perform better than others has been a core question in political science for decades, but it has lately taken on debate characteristics. One line of argument holds that variables exogenous to government -civic culture or social capital-explain differential performance. Another line contends that there is, at best, a minimal link between social capital and government performance and that research should concentrate on institutional analyses and focus on variables endogenous to governments. This project intends to achieve theoretical reconciliation between currently competing explanations to one of the most enduring puzzles of the discipline of political science. These competing explanations arise from different research traditions and different conceptualizations of government performance. Therefore, the project has two objectives. The first is to clarify the concept and improve measurement of government performance by defining it on two equally important dimensions: administrative effectiveness and policy responsiveness. The second objective is to explain variance along both dimensions, arguing that administrative effectiveness is mainly associated with endogenous or institutional variables, and policy responsiveness is more closely associated with exogenous or societal variables. The project uses a combination of quantitative and qualitative research methods with data from local governments of two federal systems: the U.S. and Germany. Such a case selection serves two purposes: 1) it provides a sample of cases large enough in order to perform statistical analyses and reveal general patterns of relationships, 2) it allows for cross-national verification of the results avoiding the fallacy of deriving general conclusions from the idiosyncrasies of a single national context. Thus, the level of analysis is within system, the results are verified across systems. The research methods employed include statistical analysis of the available data, elite interviewing for gathering original data, as well as analysis of available documentation. Factor analysis of data on the commonly used indicators of government performance is used to empirically determine the dimensionality of the government performance measure. Regression analysis helps to determine the general patterns of relationships with societal and institutional variables. The data for these two types of analyses is compiled from the local government data sets of the U.S. Census Bureau and German Association of Cities, surveys of city governments, the Government Performance Project and German Association of Cities), mass surveys (Putnam's Bowling Alone dataset and German Social Survey) and coding from the available documentation, such as the municipal codes of the U.S. cities and from German local government constitutions. The quantitative analyses are complemented by comparative case studies of four selected cities from each country in order to validate the results and illuminate the processes by which the various societal and institutional variables influence different aspects of government performance. The project's intellectual merit is both theoretical and practical. The project refines one of the most enduring puzzles within the discipline - the determinants of government performance. In addition it contributes to the growing intellectual debate on the effects of social capital. Finally, in order to improve governance, it is necessary to accurately measure government performance and this project will serve as a guidance tool for assessing and improving the performance of specific governments.