The effect of ethnic diversity on municipal spending
In an influential paper on ethnic diversity and government spending, Alesina Baqir and Easterly (1999, “Public goods and ethnic divisions,” Quarterly Journal of Economics) found that greater ethnic fragmentation decreased city spending on “productive” public goods but increased overall government spending and led to budget deficits. In a later paper (2000, “Redistributive public employment,” Journal of Urban Economics) they found that ethnic fragmentation is associated with higher levels of city employment. This goal of this study is to explore the effect of ethnic diversity in terms of the geographical family origins on four municipal-level fiscal measures: spending, employment, taxes, intergovernmental grants and spending on specific functions provided. The groups studied are “British Isles,” “Germanic,”“Latin European,” “Scandinavian,” and “Eastern European.” The groups French Canadian, Arab, Greek and American are measured separately. (The data on diversity comes from the Census; people voluntarily classify themselves by group.) We also investigate the effect of diversity on specific categories of municipal spending such as police, roads, sewers, housing and health. For comparison purposes, the measures reflecting the traditional definitions of diversity are also included: non-Hispanic white or white, black, American Indian, Asian or Pacific Islander and Hispanic.
Data was collected from a national cross-sectional sample of 417 municipalities in the United States with populations over 25,000. The dependent variables were collected from the 2002 Census of Governments and were based on fiscal measures for the whole municipality. These municipal-level measures were expenditures per capita, employment per capita, taxes per capita, and intergovernmental (state and Federal) grants per capita. We also analyze per capita spending on specific functions provided by the municipality, including police, fire, roads, sewers, parking, parks and recreation, health, housing and welfare.
The primary finding of the paper is that in the fragmentation regressions, greater geographical ancestry fragmentation decreases municipal spending, employment, taxes and grants. In contrast, greater fragmentation into the traditional categories increases spending and taxes, has no effect on employment, and decreases grants. The specific municipal function regressions give mixed results in terms of the geographical ancestry ethnic fragmentation variable. Whereas, it increases fire and road spending and decreases police, sewer and housing spending. The traditional ethnic fragmentation variable increases seven of the nine spending categories including roads and sewers.