A Markov Chain Monte Carlo Analysis of the Effect of Two-Rate Property Taxes on Construction

The paper examines the effect of taxing structures at a lower rate than land on the level of construction in 15 Pennsylvania municipalities between 1972 and 1994. These municipalities are compared with 204 similar Pennsylvania municipalities that tax land and structures at the same rate. To reduce the standard errors of estimate below those of a quasi-maximum likelihood estimation, the model parameters are estimated with a Markov chain Monte Carlo method, the Gibbs sampler. In contrast to earlier tax analyses, the estimates indicate that cities with two-rate taxes enjoy significantly higher levels of construction than they would with one-rate taxes.