Microsimulation of Residential Land Development and Household Location Choices

Abstract

This paper investigates single-family residential development for housing market equilibria using microeconomic theory and disaggregate spatial data. Mixed logit models and notions of price competition are used to simulate household location choices for three different household segments, assuming job sites of household members are known. Consistent with bid-rent theory, housing market equilibrium was reached in an iterative fashion. The spatial allocation of new households in the region of Austin, Texas illustrates the potential shape of things to come, with endogenously determined home prices and demographic distributions, based on job access. As expected, positive spatial autocorrelation in home prices and household distributions are observed.

The role of location in residential location choice models: a review of literature

“…One explanation is that density of jobs has no consistent effect on residential location choice, but that households avoid locations with a high unemployment rate. Kim et al (2005) and Zhou and Kockelman (2008) find school quality to have a positive influence; as it is a non-interaction variable, this holds for all households. Chen et al (2008) include school quality as interaction variable for households with children and without children and find it to be slightly stronger for households with children, but positive and significant for both.…”

“…Zolfaghari et al (2012) see a positive price impact when constraining the choice set based on price, indicating that households choose the best available alternative within a certain range. Ratio of price to household income is also significant when it is the only price-related variable in the model specification (Belart 2011;Bürgle 2006;Lee and Waddell 2010a;Waddell 2006;Weisbrod et al 1980;Zhou and Kockelman 2008). Walker and Li (2007) observe that price sensitivity decreases with rising income; integration of the ratio and, ultimately, the logarithmic transformation seems reasonable.…”

“…Generally, it is obtained from a transport model and needs to be calculated for every choice set alternative. Commuting time is usually a negative influence on the residential utility of a location (Axhausen et al 2004;Guo and Bhat 2007;Habib and Miller 2009;Lee and Waddell 2010a;Zhou and Kockelman 2008;Zolfaghari et al 2012). If commuting time is not available, studies use commuting distance as network-based distance or Euclidean distance instead (Belart 2011;Bürgle 2006;Chen et al 2008;Srour et al 2002), which has also always been negative but has a lower significance.…”

“…When given, residential unit size can be included as an absolute value, or in interaction with household size, to capture space per person (Belart 2011;Bürgle 2006). Zhou and Kockelman (2008) model the location of single-family house owners with the lot as location and include lot size, as well as its ratio to household size, as an explanatory variable. Axhausen et al (2004) subtract the observed mean space per person in a zone from the individual space per person to capture regional differences.…”

Residential mobility and location choice: a nested logit model with sampling of alternatives

“…Kim et al (2005) and Pinjari et al (2008a) developed NL and joint binary-ordered logit models, respectively, by using categories of locations to significantly reduce the dimensionality. Others have simply used non-MNL models with sampling of alternatives and ignored sampling biases (Yagi and Mohammadian 2008;Zhou and Kockelman 2008); this strategy, as demonstrated by Nerella and Bhat (2004) in their explorations of this subject, is highly discouraged.…”