Background:

Optimal tax rates have been a recurrent theme in tax policy discussions. In particular, the link between tax structure and economic growth has revived the Laffer hypothesis that implies the existence of a parabolic relationship between level of taxation and achieved revenues. Our own hypothesis is that urban areas have an equally significant impact on willingness to pay taxes and that there can be as much or even more heterogeneity among urban sizes as there is among different territories.

Methods:

We contrast the Laffer hypothesis for the Spanish case under different spatial scenarios using microdata for 2009 provided by the country’s Institute for Fiscal Studies (IEF). We employ a cross-sectional sample of tax filers, estimated by means of least squares.

Results:

The presence of a parabolic relationship between the tax burden and tax collection enables us to obtain a revenue-maximizing tax rate of around 30%, twice the average rate currently applied. In our analysis by regions, we find significant variations that justify the existence of a high level of fiscal decentralization. However, this variability is also found to be significant in our analysis by size of city. The maximum tax revenue of the two main cities -Madrid and Barcelona- is obtained at a tax rate of 34%, with all other cities reaching their maximum revenue levels at similar rates.

Conclusions:

The smaller the city, the lower the rate that maximizes tax revenue, with the gap between the real rate and the optimal rate being greater the smaller the city.