Jiyoon Kim, Ph.D., Political Science, University of Montreal, CP 6128, succ. centre-ville, Montreal, QC H3C 3J7, Canada
Recently, a growing body of studies finds a negative impact by ethnic heterogeneity of a locality on local fiscal policy. Among these studies, research by Alesina, Baqir and Easterly argues that ethnic diversity of a locality decreases investment-type public expenditure. With county and city level data, the authors find that a lower share of public spending goes to “nonexcludable” or “core” public goods in ethnically heterogeneous localities than in homogenous localities. They emphasize that the finding indicates more patronage goods provided in ethnically heterogeneous localities and express concerns about inefficient redistribution and unhealthy public finances as a result (Alesina, Baqir & Easterly, 1999).
Nevertheless, a relevant question to their findings still remains without receiving sufficient research attention. Do patronage-type public goods indeed consist of a larger share of local budget in ethnically diverse localities than homogeneous localities? Namely, has the spending on patronage-type public goods indeed risen at the expense of investment-type public goods? In the same vein, provided that local government spends less on investment-type public goods (i.e. education, road construction, sewage system and garbage pick-up), where does exactly the rest of budget go? That is, at the expense of investment-type public goods, who are the groups that receive targeted spending and patronage-type goods?
This paper aims to answer to these questions by exploring the patterns of local fiscal policy of American localities. I examine the functional budget items in a more specific way to find out which budget item systematically increases in ethnically diverse localities as the budget outlay on “core” public goods decreases.
For this study, I collected the data from the Census of Government of 2002 for detailed budget information of county governments of the United States. For demographic data, I used County and City Data Book released in 2000. Minding of a possible endogeneity problem due to Tiebout sorting, Two-Stage-Least-Square is used for the estimation.
The result shows no support for the argument of Alesina et al. Although it is a robust that the shares of spending on education, highway and welfare decrease as a community becomes more heterogeneous, there is no sign of an increase in patronage-type public spending. In particular, the share of spending on the wage of public employment, which is suggested as one form of patronage by the same authors in another paper, is not significantly associated with the degree of ethnic diversity of a locality. Instead, the larger share of public expenditure in ethnically heterogeneous locality goes to public services related to health and hospitals, public safety, natural resources and parks and recreation.