An Empirical Analysis of the Incidence of Location on Land and Building Values

Although it is universally acknowledged that property values are first and foremost a function of location, the extent to which location affects land versus building values has not been empirically examined. Traditional valuation models either make no attempt to separate land and building values or make implicit, untested assumptions about the extent to which various location features impact land and buildings.

This paper tests various assumptions concerning the incidence of location factors on land and building values and evaluates the composition of total value between the two parts. The research builds on research conducted in 2000 under a David C. Lincoln Institute Fellowship in Land Value Taxation that explored the use of modern computer-assisted mass appraisal (CAMA) tools to estimate land values in urban residential areas, often with comparatively few vacant land sales. That research concluded that CAMA models combining vacant and improved sales can be used to predict land values with acceptable reliability, even when some neighborhoods lack vacant land sales altogether. Thus, the phase-in of a site valuation tax scheme in which buildings were untaxed or taxed at a lesser percentage than land could continue to use the same sales-based mass appraisal tools commonly used for improved residential properties.

Utilizing the same three data bases as the prior research project, this paper evaluates the extent to which location affects land and building values and how total property value is broken out between the two components. The research results indicate that, while location impacts both land and buildings, on a percentage basis the impact on land is far greater. It also suggests that traditional attempts to separate values between land and buildings are likely unreliably and may well under-estimate the contribution of the land component. A reliable decomposition would seem to requires the incorporation of both vacant and improved sales, at least until more empirical experience is gained with respect to typical land-to-building ratios among various property types and market areas.