Conjoint Preference Estimation for Residential Land Use Policy Evaluation.
Residential environments were defined in terms of seven attributes pertaining to off-lot visual environmental quality, on-lot space-using characteristics, and house price. Each survey respondent traded off several pairs of attributes through a preference ranking of combinations of attribute levels. Metric estimates of the parameters of a separable utility function (one for each respondent) were calculated from these rankings using a conjoint measurement algorithm, a nonmetric multidimensional scaling procedure. Simulations of hypothetical markets used the estimated utility functions to project market shares. The two alternatives of each simulation traded off a stated improvement in the view from the front or backyard against a specified modest reduction in a component of private space, with all else equal. In seven of the eight hypothetical markets a majority of homebuyers were projected as choosing to sacrifice modest amounts of space for a visual quality improvement. (For Essays on the Multidimensional Analysis of Perceptions and Preferences.) 43 pp. Ref. (Author)