Abstract
This paper studies an overlooked phenomenon in the provision of public goods, namely local production of a national public good, such as the manufacture of fighter planes (which contribute to national defence) in many different jurisdictions across a country. Because local production of the national good raises local incomes, each jurisdiction seeks to raise its share of the good’s production. A subset of jurisdictions then forms a minimum winning coalition, which offers equal production shares to its members and smaller (possibly zero) shares to non-members, while choosing the provision level of the national good. The outcome is inefficient, with production inefficiently concentrated and the public good also overprovided (because income benefits reduce the good’s perceived marginal cost). Empirical results confirm the prediction that the location of production is important in determining Congressional support for federal programme spending.