It is widely acknowledged that democracy affects economic growth and social welfare. The existing political economy literature, however, rarely disaggregates political institutions in ways that generate testable propositions about causal mechanisms. Our project seeks to advance this important strand of research by reconceptualizing both the independent and dependent variables. Specifically, we argue that stronger party systems, characterized by electoral competition between few disciplined parties, are more likely to implement effective government policies that generate inclusive economic growth than weaker systems with undisciplined or multiple coalition parties. We then show that government investments in early childhood education and care (ECEC) and family in-kind benefits are especially good at promoting long-term social welfare. To test our argument, we estimate the effect of party institutions on such future-oriented government spending by exploiting the timing of major changes to party discipline and fragmentation in an original dataset covering the last forty years in OECD countries. Overall, we show that political systems in which parties are large and disciplined are more likely to spend on effective public policies.