The role of politics in national social policy developments has been a frequent subject of comparative policy research. Empirical attempts to isolate the relative impact of political forces on social policy outcomes, however, have produced contradictory results. In some analyses, a nation's level of economic development and the age of its social security system are reported to be more consistent and better predictors of social policy efforts than political variables.' In other studies, political forces, ideologies, and institutions are said to be key elements of the social policy process.2 The findings of these research efforts have been inconsistent because different sets of countries and time frames have been analyzed and different methodologies and statistical procedures have been employed. In addition, political forces have consistently been defined in very limited terms. The usual practice has been to define politics only as conventional political activity and to measure it in terms of type of party system, degree of interparty competition, or level of voter participation. Such descriptions and operationalizations assume that citizens exert an influence on policymaking only through democratic constitutional channels of participation. Yet, historical accounts suggest that more unconventional forms of political activity, such as displays of citizen unrest and organized protest movements, have been significant social policy stimulants within particular countries.3 The impact of such unconventional types of political activity has been investigated in some empirical studies; again, the findings have been ambiguous. For example, Frances Fox Piven and Richard Cloward present data in their research to support the argument that, in the United States, citizen protests led to increased enrollments in welfare programs.4 In a recent exchange, Robert Albritton challenges these conclusionss and is, in turn, disputed by the first two authors.6 In another analysis of welfare activities in the American states, Edward Jennings examines the impact of more generalized citizen unrest and finds some relationship between protest and program enrollments; however, he adds that other factors, such as institutionalization, must also be considered.' In still other analyses, social unrest is identified as an important determinant of social welfare policies and is operationalized in a more structured fashion. Peter Flora, for example, cites the significance of labor movements and organized left-wing parties in his examination of the introduction of social insurance systems in western Europe. Again, the