Combining revealed and stated data to examine housing decisions using discrete choice analysis

This paper uses stated preference and revealed preference data, separately and jointly, to examine individuals, housing choices. In particular, it combines an established revealed preference approach, discrete choice analysis of actual housing purchases, and a relatively new stated preference approach, choice-based conjoint analysis of hypothetical housing purchases, to provide us with a better understanding of the factors driving housing decisions. These methods are appropriately combined since they reflect the same decision process: households select the dwelling that provides the best combination of attributes from the feasible set of alternatives. In total, this paper analyzes three data sets: only actual housing purchases, only hypothetical housing purchases, and both types of purchases combined. Analysis finds that actual and hypothetical housing purchases are similar decision processes with respect to some attributes, yet are dissimilar with respect to other attributes. Nevertheless. combining the two types of data improves identification of both categories of attributes, Journal of Economic Literature Classification Number R21. (C) 2001 Elsevier Science.