The research presented in this book is a comparative study of collective community decision-making regarding federally funded redistributive policy outputs. The study covers federal allocations by the Office of Economic Opportunity (OEO) to thirty middle-sized American cities from 1968 to 1970. At a theoretical level, three but interdependent community subsystems are distinguished: citizens, organizations, and political entrepreneurs. The propositions that are formulated and tested relate to the role of subsystem preferences in determining redistributive policy outputs and to the processes through which these preferences are translated into collective community decisions. The key analytic concepts are purposive actors, assets and liabilities, influence resources, and collective action. The major methodological innovation of this study is it's use of Coleman's Mathematical Model of Collective Action (1973) in comparative urban research. In a sample, twenty to thirty-five political entrepreneurs were questioned in each city considering their interests and influence in two kinds of anti-poverty policies: economic development providing jobs to the poor, and redistributive welfare spending. The use of the model is empirically successful. Collective elite action explains indeed a substantial proportion of the variance in federal OEO allocations to cities.