Comparative research on the development of modern welfare states treats the U.S. welfare state as an aggregate, and often this work must discard the U.S. case in any testing of hypotheses because it is exceptional. This holds true for those theories which argue that welfare states have resulted primarily from the exercise of power by social democratic parties. Since the New Deal, the role of state governments in welfare activity has grown. Yet, there has been no explicit geographical study of the formation of the U.S. welfare state, especially in terms of variation at the state level. This research project will use a number of theories to examine the geographical dimensions of the U.S. welfare state from the 1930s to the 1960s. Programs for unemployment insurance, aid to families with dependent children, and general relief will be analyzed across states over time in order to determine whether they conform to one of three models of welfare state development. Welfare program variables will be examined using pooled cross-sectional time series analysis in a time dependent analysis. The results will be used to test hypotheses about the regional structure of U.S. politics over the time period in question, and to assess the effectiveness of competing explanatory models in different regional contexts. This research will make an important contribution to an understanding of the causes of welfare state formation in the United States. It will be the first extensive longitudinal analysis of the applicability of competing theories of welfare state formation across different regional divisions.