Implications of a Split-Rate Real Property Tax

Since the time of Henry George, there has been interest in his “single tax” on land values only, and in the related notion of taxing improvements less heavily than land, if improvements are part of the property tax base. In the United States, though, the real property tax rate almost everywhere applies equally to the values of both land and improvements; several Pennsylvania cities have been the primary exceptions. Since July 2003, however, two Virginia cities—Fairfax and Roanoke—have had legal authority to tax improvements at a lower rate than land, and some other localities in the state have expressed an interest in such a “split-rate tax.”

This study uses data on individual property parcels from the tax rolls of three Virginia localities to explore the initial redistribution of real property tax liabilities under a split-rate tax, compared to the current uniform tax on land and improvements. Emphasis is on the limiting case of a pure land tax, but the pattern of redistribution would be the same with less rate differentiation, although the changes would not be as large.

The three localities are quite different from one another: Roanoke, a central city of nearly 100,000 residents, has been experiencing slow loss of population but modest growth in employment; Chesterfield County, a first-tier bedroom county in the Richmond metropolitan area with over 250,000 residents, has been experiencing rapid growth of population and nearly as rapid growth in employment; and Highland County, a small rural county on the West Virginia border with fewer than 2,500 residents, continues its long-term population loss and has been experiencing a more rapid loss of jobs.

In all three localities, the move to an equal-yield split-rate tax would reduce the residential share of the real property tax while increasing the business tax share—although the general character of business properties is quite different in Highland County than in the other two jurisdictions. Indeed, there are differences in inter-class tax changes among the three areas due, in part, to different land use patterns. In all three localities, though, there is substantial variation in the magnitude of tax change within classes, with the exception of the vacant land class.

In Roanoke, it was possible to identify the property parcels’ census tract locations, and this enabled us to explore relationships between tax changes and various population and property attributes. One of the strongest relationships is between residential tax change and residents’ income. In general, larger residential tax increases (or smaller decreases) occur where incomes are higher and where poverty rates are lower. Other relationships are weaker, but some still are substantial. For example, larger increases (smaller decreases) also are associated with higher owner-occupancy rates, an older population, a larger white percentage of the population, and larger home size, while larger residential tax reductions are associated with crowded housing.