Policy Competition in a Spatial Economy 

Abstract

We incorporate policy competition among governments into a quantitative spatial model. We develop a new approach to quantifying the interdependence of policymaking among a large number of competing jurisdictions that allows us to fully characterize an endogenous network of potentially millions of bilateral strategic policy linkages. These linkages are summarized by the “policy impact” of each jurisdiction – the extent to which a jurisdiction changing its policy affects policy choices of others.
Policy impact is heterogeneous across jurisdictions and the primary driver of the welfare effects of policy competition. Applying our model to a network of 3109 U.S. counties competing in local sales taxes, we consider several centralized interventions to limit tax competition and document that (1) tax competition is harmful: federal interventions increase welfare, albeit heterogeneously across places; (2) ignoring endogenous policies from a quantitative spatial model underestimates welfare effects up to 30%. 

To explore the underlying mechanisms, we rank counties by the magnitude of their policy impact. We show that interventions targeting high-policy impact jurisdictions raise welfare by six times more than targeting low-tax jurisdictions, as is commonly done through minimum tax rates. Although policy impact is endogenous, we document that jurisdiction size acts as a strong proxy. Beyond the tax application we consider, the framework for endogenizing policy and implications are applicable to tariffs and trade wars, environmental policy competition, regulatory policies, and spending competition.