What determines the level of local business property taxes?

Conventional economic wisdom holds that the optimal level of business taxes recovers the exact cost of providing public services to businesses. I present an alternative model and derive conditions under which a self-interested decisionmaker picks tax rates to maximize an objective function that depends on business tax revenues, workers' labor earnings, and business profits. I show that, in this model, optimal business tax revenues may either be more or less than the cost of providing public services to a city's businesses. Data from approximately 70 large U.S. cities over a 16-year period reveals that the mean commercial effective property tax rate is about 1.6 times the mean home effective tax rate. Conventional wisdom suggests that commercial effective tax rates (ETRs) should rise with the share of spending on business services, while the alternative theory suggests that commercial ETRs should rise with a city's desirability as a business location. Empirically, I explore three proxies for desirability as a business location: net-of worker compensation value added; adjusted land values; and population, since standard urban theory suggests that larger cities are more desirable business locations. I report results from a large variety of regressions using various empirical proxies for a city's share of spending on business services, desirability as a business location, and an assortment of control variables on commercial ETRs. These regressions provide no support for the conventional wisdom - commercial ETRs are inversely related to the share of spending on business public services. There is mixed evidence to support the alternative hypothesis that commercial ETRs rise with a city's desirability as a business location. Commercial ETRs also rise with median home values but are not consistently associated with a city's economic or fiscal conditions.