This paper presents a model linking ethnic fragmentation in cities to the amount and type of public goods provided. Using data from US cities, metropolitan areas, and urban counties, the authors find that productive public goods such as education, roads, libraries, sewers, and trash pickup are inversely related to ethnic fragmentation. Ethnic fragmentation is also negatively related to the share of local spending on welfare. These results are mainly driven by observations where white majority cities react to varying minority group sizes. The paper concludes that ethnic conflict is an important determinant of local public finances. The authors argue that when individuals have different preferences, they are less willing to contribute to public projects. In cities with polarized ethnic groups, politicians with ethnic constituencies provide less efficient public goods. Interest groups with an ethnic base value only the benefits of public goods that accrue to their groups and discount benefits for other groups. This leads political actors to divert more public resources to private patronage. The paper also discusses how income inequality, migration, productive public goods, and interest group politics may affect the predictions of the model. It relates the polarization described in the model to ethnic divisions, noting that conflicts over public policy are increasingly determined by racial cleavages rather than class cleavages. The empirical evidence shows that public goods provision is lower in more ethnically fragmented localities. The share of spending on welfare is also negatively related to ethnic fragmentation. The share of expenditure on police increases with ethnic fragmentation, and spending on health and hospitals strongly increases with ethnic fragmentation in the metro and county samples. The results suggest that more ethnically diverse jurisdictions have higher spending and higher deficits/debt per capita, but devote lower shares of spending and less per capita spending to core public goods like education and roads. The higher spending in more ethnically diverse jurisdictions is financed in part by higher federal transfers rather than by local taxes. These results are consistent with political economy theories in which heterogeneous and polarized societies value public goods less and are collectively careless about fiscal discipline. The paper concludes that ethnic divisions are an important determinant of local public finances and that the problem of urban public goods in America appears intractable due to ethnic divisions. The results suggest that policymakers should focus on increasing incentives for mixing and desegregation rather than increasing segregation and decentralization to handle ethnic polarization.This paper presents a model linking ethnic fragmentation in cities to the amount and type of public goods provided. Using data from US cities, metropolitan areas, and urban counties, the authors find that productive public goods such as education, roads, libraries, sewers, and trash pickup are inversely related to ethnic fragmentation. Ethnic fragmentation is also negatively related to the share of local spending on welfare. These results are mainly driven by observations where white majority cities react to varying minority group sizes. The paper concludes that ethnic conflict is an important determinant of local public finances. The authors argue that when individuals have different preferences, they are less willing to contribute to public projects. In cities with polarized ethnic groups, politicians with ethnic constituencies provide less efficient public goods. Interest groups with an ethnic base value only the benefits of public goods that accrue to their groups and discount benefits for other groups. This leads political actors to divert more public resources to private patronage. The paper also discusses how income inequality, migration, productive public goods, and interest group politics may affect the predictions of the model. It relates the polarization described in the model to ethnic divisions, noting that conflicts over public policy are increasingly determined by racial cleavages rather than class cleavages. The empirical evidence shows that public goods provision is lower in more ethnically fragmented localities. The share of spending on welfare is also negatively related to ethnic fragmentation. The share of expenditure on police increases with ethnic fragmentation, and spending on health and hospitals strongly increases with ethnic fragmentation in the metro and county samples. The results suggest that more ethnically diverse jurisdictions have higher spending and higher deficits/debt per capita, but devote lower shares of spending and less per capita spending to core public goods like education and roads. The higher spending in more ethnically diverse jurisdictions is financed in part by higher federal transfers rather than by local taxes. These results are consistent with political economy theories in which heterogeneous and polarized societies value public goods less and are collectively careless about fiscal discipline. The paper concludes that ethnic divisions are an important determinant of local public finances and that the problem of urban public goods in America appears intractable due to ethnic divisions. The results suggest that policymakers should focus on increasing incentives for mixing and desegregation rather than increasing segregation and decentralization to handle ethnic polarization.