This paper clarifies the theoretical bases for expecting women to be more supportive of government programs than are men. We identify several factors not developed in past literature: (1) differences in emotional responses to social problems, (2) gendered differences in the awareness of those problems among one's own kin, (3) differences in the perceived fairness of existing social institutions, (4) differences in the perceived efficacy of government programs, and (5) variations in the preferred form that those programs should take. Testing hypotheses using data from a 1995 survey of public opinion covering five policy domains, we find that the gender gap in these domains is substantial, comparable to that associated with race and partisanship. These differences are attributable in part to factors identified previously in the literature--such as gendered differences in perceived opportunity--but are also affected by differences between men and women in the perceived efficacy of government programs (e.g., the extent of fraud and abuse) and differences in the preferred form those programs should take.