Measuring values for environmental public goods: incorporating gender and ethnic social effects into stated-preference value-elicitation methods
Date
2011
Authors
Jones Ritten, Chian, author
Kling, Robert W., advisor
Bernasek, Alexandra, committee member
Tavani, Daniele, committee member
Loomis, John B., committee member
Journal Title
Journal ISSN
Volume Title
Abstract
This dissertation explores the theory and seemingly paradoxical results between the economic literatures of contingent valuation method and risk aversion and the interdisciplinary literature of perceptions of risk, specifically with regard to race and gender effects. While a review of the contingent valuation literature shows no systematic gender or race differences in willingness to pay to reduce risks associated with nonmarket goods and services, the risk aversion literature finds systematic gender and race differences in levels of aversion to risks. Women are found to be more risk averse than men and Blacks and Hispanics less risk averse than whites. It is hypothesized that an individual measure of willingness to pay to reduce risks associated with nonmarket goods should be directly related to individual levels of risk aversion and consistent with individual perception of risk. The results from the risk perception literature also find systematic gender and race differences. These results are consistent with the risk aversion literature for gender effects, but inconsistent for Blacks and Hispanics who are found to perceive more risks than whites. To explore this inconsistency, a theoretical model is constructed that connects the contingent valuation theory to that of risk aversion and perceptions of risk. Insights from the risk perception literature are used to create a model of absolute risk aversion in order to make a tractable connection to risk aversion and stated valuation in CVM. Data from a previously collected dataset by Loomis et al (2009) is fit to the model. The results reinforce the inconsistency found between the risk perception and contingent valuation literatures and indicate a possible shortcoming of traditional methodology used by contingent valuation studies and the need for use of proper payment vehicles. The existence of social preferences has been well established in the experimental literature and is formally modeled in this dissertation by incorporating influences of self-interest, altruism, reciprocity, fairness, and commitment in the context of stated willingness to pay in contingent valuation methodology. The models suggest that the existence of social preferences may explain some of the inconsistency between the relevant literatures. A dichotomous choice stated valuation study of various vaccination programs was conducted among college students at Colorado State University. The finding indicate gender differences in willingness to pay for vaccination programs and that the payment vehicle may have substantial effects on valuation. The inclusion of social preferences is a substantial improvement to modeling of valuation and when not included, may lead to underspecified models that ignore existing gender effects.