Optimal constitution design for state vs. local governments

In Bassetto with Sargent (2004) and Bassetto and Lepetyuk (2007), I developed a model that can be used to quantitatively assess the consequences of departures from Ricardian equivalence on the mix of public goods provided by the government. Those papers assume that mobility is exogenous and that no factors are in fixed net supply within a jurisdiction. While both assumptions are reasonable approximations for a U.S. state or a national government, they are clearly unsatisfactory for studying local public finance. I consider here a model where households make endogenous location decisions, subject to differing degrees of frictions at the state vs. local community level. In particular, opportunities to move across states occur only occasionally, while relocation within a state is always possible, at a cost. I use the model to compare quantitatively the performance of the golden rule of public finance with the incentives stemming from capitalization of local public goods into land prices. The golden rule (borrowing only to pay for capital improvements) is an imperfect instrument insofar as it gives a very crude distinction of public goods into two categories. By contrast, capitalization is only effective if endogenous mobility and fixed factors are sufficiently important.