Regret minimization or utility maximization: it depends on the attribute
In this study we show how the coexistence of different decision rules can be accommodated in discrete choice models. Specifically, in this paper we present a generic hybrid model specification that allows for some attributes being processed using conventional linear-additive utility-maximization-based rules, while others are being processed using regret-minimization-based rules. We show that on two revealed and stated choice datasets particular specifications of hybrid models, containing both regret-based and utility-based attribute decision rules, outperform—in terms of model fit and out-of-sample predictive ability—choice models where all attributes are assumed to be processed by means of one and the same decision rule. However, in our data differences between models are very small. Implications, in terms of marginal willingness-to-pay measures (WtP), are derived for the different hybrid model specifications and applied in the context of the two datasets. It is found that in the context of our data hybrid WtP measures differ substantially from conventional utility-based WtP measures, and that the hybrid WtP specifications allow for a richer (choice-set-specific) interpretation of the trade-offs that people make. Keywords: random regret, random utility, hybrid choice models, willingness to pay
When requesting a correction, please mention this item's handle: RePEc:pio:envirb:v:40:y:2013:i:1:p:154-169. See general information about how to correct material in RePEc.
For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: (Neil Hammond)
If references are entirely missing, you can add them using this form.