Association for Public Policy Analysis & Management

In this study we examine whether the salience of local government finance method (taxation vs. debt finance) affects household mobility. To do so, we use the Fiscally Standardized Cities (FiSC) database for the 150 largest cities in the U.S. to investigate the salience of local government finance policy and household mobility. We estimate empirical models to determine whether households react to local government finance choices by moving in or out of cities.

According to the Ricardian Equivalence Theorem, choices between tax financing and debt financing by local governments should not matter in a closed economy with full salience of financing policy and other necessary assumptions. However, Wellisch and Richter (1995) challenge the neutrality thesis of local public finance by developing a model that predicts the neutrality result does not hold in general if we allow for household mobility and debt serviced by local residence-based taxes. The main reason for the non-neutrality result is that those taxes create locational distortions. Consequently, the model predicts that people react to changes in local government finance policy by moving into or out of jurisdictions.

In our paper we develop a discrete time dynamic optimization model in which we analyze two cases: a Ricardian case in which a representative agent is immobile, and a non-Ricardian case in which a representative agent can move to some other city in order to maximize life-time utility. In this model, we find that households will tend to move out when the local government chooses to use debt financing over tax financing in order to avoid future tax increases necessary to service the debt. Our crucial assumption here is that financing options are fully salient to households. Then, we use a panel data set of 150 FiSCs in the U.S. over the period from 2007 to 2012 to estimate empirical models of mobility and government finance choices. This database provides a consistent picture of local government financing across cities that is essential to our modeling approach. The migration data are collected from the U.S. Census Bureau migration flow estimates based on the American Community Survey (ACS). We test whether households react to the local government finance choices in terms of voting with their feet. Our preliminary results for out-migration reveal that households react to changes in local government tax policy, but not to changes in debt policy. We interpret the different reactions as revealing that tax policy changes may be more salient than debt policy changes. We also plan to analyze in-migration as well, providing a comprehensive analysis of this topic.