Factors affecting residential property development patterns

Journal of Real Estate Research, The, Jan-Mar 2003 by Smersh, Greg T, Smith, Marc T, Schwartz, Arthur L Jr

Abstract This article is the winner of the Real Estate Development manuscript prize (sponsored by the Urban Land Institute) presented at the 2002 American Real Estate Society Annual Meeting.

This study uses a disaggregated data set, county property appraiser data, to track the number of new single-family housing units built in each section (square mile) of Alachua County, Florida by the year built over a twenty-year period. It explores the role of transportation, large-scale development, employment nodes, existing patterns of development and regulation on the spatial pattern of development. The results of the model suggest that the variables tested are important determinants of the growth pattern in Alachua County, but that much of the growth pattern is not explained by the explanatory variables employed in the model.

Factors Affecting Property Development Patterns

The pattern of residential development within the context of metropolitan growth and development has been the subject of an extensive literature. Among the streams of literature have been monocentric and policentric models, rent gradients and population density, and spatial mismatch and jobs/housing balance. Less examined have been the factors that determine the specific location of residential development from among the number of potentially suitable sites available. Textbook treatments (e.g., Miles et al., 2000; and Kone, 1994) of site selection suggest that factors that are important in locating a residential development include:

* Physical suitability for development: slopes, soils, hydrology, land availability

* Legal restrictions, government regulations (zoning and other land use controls)

* Existing land use patterns and location of other residential development

* Access, including proximity to interstate highways

* Distance to employment sources

* Distance to shopping

* Availability of amenities (water, restaurants and shopping, golf, parks)

* Neighborhood factors: age of surrounding housing stock, schools, crime

However, multiple sites may be suitable when evaluated across the range of criteria, yet one is developed. Further, development may move in a single direction or sector of a city although suitable sites are available in other areas. This suggests that certain factors may be more important than others in determining the location of new projects. McDonald and McMillan (2000: 136) note, "little is known about the spatial patterns of development because data is usually highly aggregated spatially." This article uses a disaggregated data set, county property appraiser data, to track the number of new housing units built in each section (square mile) by the year built for Alachua County, Florida. Alachua County forms the metropolitan area for Gainesville. As the home of the University of Florida, it is an area with a defined employment node. It is also a significantly smaller city than has been the target of the limited research to date, allowing a focus on fewer locational variables, as it does not have a major airport, public transportation or a number of employment nodes. Alachua County also has a large-scale development, Haile Plantation, which may have had a major influence on the development pattern.

This article explores the role of transportation, large-scale development, employment nodes, existing residential development and regulation on the spatial pattern of development. As discussions turn to smart growth, compact development and the alleviation of sprawl, it is important to understand the forces that contribute to observed development patterns. Specifically, the growth pattern of single-family housing over the past twenty years is assessed.

Influences on Spatial Patterns of Development

McDonald and McMillan (2000) completed research that most directly impacts on this work. They studied the choice of development location for different forms of land use (industrial, commercial and residential) in the Chicago metropolitan area. For residential development, they found that proximity to commuter rail stations, highway interchanges and suburban employment nodes had negative effects on residential development. These variables were all specified as the inverse of distance to allow the marginal effect of each variable to decline rapidly. Distance to downtown and to O'Hare Airport also had negative and statistically significant coefficients. The authors conclude that residential development is more likely at greater distances from an interchange, and that commuter rail and employment subcenter have no effect on residential locations. As a result, the pattern of residential development is scattered. The results may reflect the already built-up nature of some sites, the high value of such sites and the negative externalities associated with such sites as suitable residential location choices. Because of these negative externalities, households desire to locate where they have access to these locations but are not so close as to be negatively impacted by the externalities.