Spatially-variable property tax and optimal tax composition in congested monocentric cities: George, Pigou, Ramsey and Strotz unified

We explore the optimal spatial structure of the property tax in a monocentric city with transportation congestion. To show how the availablity of revenue instruments impacts the properties of the optimal property tax, we discuss three regimes with different combinations of a head tax, congestion tolls, and property tax. In each regime, the tax composition is optimally determined to finance a predetermined expenditure. Under the first regime, with all the three taxes available, the first-best allocation is achievable only when the property tax is unused. Under the second regime, where only head and property taxes are available, a spatially variable property tax substitutes for the unavailable congestion tolls; the head tax (or subsidy) fills the gap between the property tax revenue and the predetermined expenditure. Moreover, property tax is negative (positive) close to the city center (boundary). Under the third regime, where only property tax is available, the role of the property tax is twofold: to substitute for the unavailable congestion tolls and to finance the predetermined expenditure. In general, for both roles, the optimal tax rates are spatially variable. Furthermore, if its role as a substitute for congestion tolls dominates its other role, the property tax may be negative close to the boundary.