Understanding Heterogeneous Preferences in Random Utility Models: The Use of Latent Class Analysis

  • Author(s) / Creator(s)
  • Consumer preferences for goods and services are characterized by heterogeneity. Accounting for this heterogeneity in economic analysis will be useful in estimating unbiased models as well as for forecasting demand by including individual characteristics and proving a broader picture of the distribution of resource use decisions or policy impacts. However, many empirical economic analyses assume homogeneous preferences among consumers. Alternatively, previous analyses considered preference heterogeneity a priori by: 1) including demographic parameters in demand functions directly or through the utility function (e.g. Pollack and Wales 1992); or 2) by stratifying consumers into various segments and estimating demands separately on each stratum. For these analyses, economists traditionally focus on demographic variables. There is empirical evidence that these methods identify sources of heterogeneity. For example, Famulari (1995) showed that stratifying households by demographic categories significantly improved tests of consistency with the axioms of revealed preference. Boxall et al. (1996) studied recreation demand in a traditional travel cost model framework and found that stratification reduced the percentage and mean error of violations of the choice axioms. These studies examined traditional demand analysis, rather than considering the individual choice behaviour typified by the random utility framework. Also there are few economic studies which examine individual-specific variables other than sociodemographic factors.

  • Date created
  • Subjects / Keywords
  • Type of Item
  • DOI
  • License
    Attribution-NonCommercial-NoDerivatives 3.0 International