Repository logo

Measuring citizens' preferences for protecting environmental resources: applications of choice experiment surveys, social network analysis and Deliberative Citizens' juries




Geleta, Solomon, author
Loomis, John, advisor
Janmaat, Johannus, advisor
Davies, Stephen, committee member
Kroll, Stephan, committee member
Buyere, Brett, committee member

Journal Title

Journal ISSN

Volume Title


Many reasons have been suggested as explanation for observed differences in citizens' environmental conservation projects policy choices and willingness-to-pay (WTP) values. Some people attribute this distinctive decision behavior to contrasts in the overall policy outcome expectations (preference heterogeneity) and/or differences in reactions to the changes in the environmental attributes (response heterogeneity). Others attribute this to differences in individual choice rationales, personalities, encounters, and past and present experiences. In other words, regardless of the possibility that outcomes are the same, people do not have the same emotions, convictions, disposition, or motivations. In three separate essays, I investigate the possible reasons for the observed differences in citizens' environmental conservation policy choices and examine how preference and response heterogeneity arise. In the first essay, I ask if a priori environmental damage perception is a source of heterogeneity affecting conservation option choice decisions. In the second, I investigate if social networks (interactions among decision-making agents) affect choice decisions. In the third, I investigate if preferences change when decision making agents are allowed to deliberate among peers. For the first essay, I conducted an on-line choice experiment (CE) survey. The survey asked questions that help to measure citizen preference for protecting environmental public goods, ascertain the value local residents are willing-to-pay (WTP), and determine how preference heterogeneity arises. CE attributes included groundwater use (measured by share of total water use from groundwater), aquatic habitat (measured by count of spawning kokanee salmon return), natural habitat health (measured by the sensitive ecosystem area reclaimed), and rural character (measured by a decrease in urban sprawl and/or a decrease in population density in rural areas). I used a special property levy as the vehicle of payment. Random parameter logit (RP) and latent class (LC) models were estimated to capture response and preference heterogeneity. The results suggest that (1) both preference and response heterogeneities were found for the choices and all environmental attributes respectively (2) respondents who have a higher value for one environmental good will have a higher value for other environmental goods, and (3) a priori damage perception could be one of the sources of response and preference heterogeneity. In the same survey, I included people's egocentric networks, interactions, environment related activities and perceptions to empirically evaluate whether social network effect (SNE) is a source of systematic differences in preference. I estimate consumer preferences for a hypothetical future environmental conservation management alternative described by its attributes within a Nested Logit Model: nesting broader and distinct conservation options within choices impacted by individual's network structure. The results show that some network centrality measures capture preference heterogeneity, and consequently the differences in WTP values in a systematic way. Third, I compare the value estimated based on the traditional choice experiment (CE) with the results obtained using the citizen jury (CJ) approach or a group-based approach or also called the "Market Stall" in some literature. I estimate the effect of deliberation on conservation choice outcomes by removing any significant differences between the people who participated in the CJ (people who volunteered to be contacted again after deliberation treatment) and those people who did the survey twice but did not volunteer for CJ (control group) in terms of their socioeconomic status and be able attribute the changes in preferences to deliberation treatment only. CJ approach involved two 90 minute deliberations held over two days to discuss and consider their preferences and WTP values with other household members. Results show that deliberation improves individuals' valuation process and there is observed difference in choice outcomes between the deliberation treatment and control groups. Both preference and response heterogeneity relatively vanish when people were allowed to deliberate among peers.


Rights Access


citizens' preference
preference heterogeneity
social network analysis
deliberative citizens' juries
choice experiment
response heterogeneity


Associated Publications