This paper studies politically feasible policy solutions to inequities in local public goods provision. I focus in particular on the entwined issues of high property taxes, geographic income disparities, and inequalities in public education prevalent in the United States. It has long been recognized that with a mobile population, local administration and funding of schools leads to competition between districts. By accounting for heterogeneity in incomes and home qualities, I am able to shed new light on this phenomenon, and make novel policy recommendations. I characterize the equilibrium in a dynamic general equilibrium model of location choice and education investment with a competitive housing market, heterogeneous wealth levels and home qualities, and strategic district governments. When all homes are owner-occupied, I show that competition between strategic districts leads to over-taxation in an attempt to attract wealthier residents. A simple class of policies that cap and/or tax the expenditure of richer districts are Pareto improving, and thus politically feasible. These policies reduce inequality in access to education while increasing expenditure for under-funded schools. Gains are driven by mitigation of the negative externalities generated by excessive spending among wealthier districts. I also discuss the policy implications of the degree of homeownership. The model sheds new light on observed patterns of homeownership, location choice, and income. Finally, I test the assumptions and implications empirically using a regression discontinuity design and data on property tax referenda in Massachusetts.