Association for Public Policy Analysis & Management

Background: From cities and counties to school districts and water districts, nearly 90,000 local taxing bodies exist in the United States. While the existence of numerous competing governments is often advanced as a norm for the fiscal operations of a federal system, the reality is that the proliferation of governments has resulted largely from the vertical layering of multiple function jurisdictions rather than from the horizontal partitioning of same-level jurisdictions. The governments in this case do not appear to be competing on service and quality; rather, they share and compete over commonly shared resources—the overlapping tax revenue. This paper is dedicated to providing empirical evidence of the existence, direction, and magnitude of this overlap effect in the context of property tax policy. Also, this paper fills the gap in the existing literature on property tax competition by shedding light on the evolving polycentric governance embedded (Ross and Hummel, 2012) across irregular boundaries and interconnected through multilayer relations.

Method: In this paper, I estimate the causal effect of a property tax increase at one level of government on fiscal decisions of overlapping governments. To address endogeneity concerns in jurisdictional overlap, I use different quasi-experimental approaches based on close bond referenda, comparing the regression discontinuity (RD) with difference-in-differences designs (DiD). I compare the fiscal responses of cities and water districts that overlap with school districts whose referenda barely passed to those where school districts’ referenda barely failed. Around the 50% vote-share threshold, this simulates an experiment where tax increases in a shared tax base are almost randomly assigned to cities and water districts. I also exploit this feature by comparing trends in tax and tax-supported debt before and after the referenda between the winners and losers of marginal referenda. Drawing on bond referenda data, fiscal data, and geospatial data, I use GIS to match each city and water district experiencing a school bond referendum to its tax and debt decisions.

Findings and Implications: I find that, overall, a tax increase in a school district induced an increase in tax and tax-supported debt among the overlapping cities and water districts. The results were remarkably consistent with regarding to model results and across samples. I also find some evidence that bond passage in a close referendum led to significant decreases in cities’ property tax rates in the short term. Further, the results imply that these impacts vary by taxing capacity and also intergovernmental constraints of overlapping jurisdictions; more populous cities, cities having smaller capital budget, and younger water districts were more affected by an increase in school property taxes. This finding suggests the presence of asymmetric power relations that resonate unequal power dynamics within uncoordinated overlapping institutions. The results of this study provide efficiency and equity implications for policy and practice in a fragmented fiscal landscape.