Project R-5488

Does respondent inefficiency drive hypothetical bias in discrete choice experiments for non-market goods? (Research)

This research is situated in the micro-economic field of economics that aims at unveiling preferences and by extension people's willingness to pay (WTP) for those goods that do not have a market value or are difficult to proxy. To allow such an act, stated preference methods have been developed, among which discrete choice experiments have been gaining attention. In such an exercise, people are generally asked to make repeated choices between several alternatives described by a number of attributes having a number of levels. Consequently, through nonlinear regression using models based on random utility theory (RUT) the indirect utility each attribute contributes to the good or service under revision can be assessed. However, RUT assumes the respondent to be a perfect homo economicus. This has been found to be a demanding assumption. Moreover, it has been shown that the WTP estimates are often inflated when compared to real expenditures or their experimental counterparts that do have pecuniary implications. These aspects are in need of further investigation. Therefore, it is my aim to: (1) verify the influence of context variables on how people make their choices, (2) verify whether adding pecuniary implications to the experiment causes shifts in how people make their choices, and (3) develop an econometric model that allows representing a decision making strategy in which people not only ignore information about attributes, but additionally disregard entire alternatives.

01 October 2014 - 30 September 2017