You are here:
Social Psychology of Stated Preference
The proposed research will explore the social psychology of expressed willingness to pay (WTP) for environmental improvements in order to gain understanding of question-wording effects, examine whether structured value elicitation techniques can mitigate them, and explore the comparative advantage of the contingent valuation method (CVM) and more interactive and discursive methods of assessing social value. Sensitivity of CVM to question wording (e.g., embedding effects) is troublesome if WTP is presumed to directly reflect underlying preferences, but from the standpoint of a constructive theory of preferences and from knowledge gained from surveying other subjective phenomena, such phenomena are normal. They can be interpreted as reflecting framing or focus effects, in which wording influences WTP by altering the cognitive shortcuts individuals use to construct their responses. Two experiments are proposed: 1) to examine framing effects and analyze respondents cognitive processes to determine whether cognitive focusing mediates the effects, and 2) an exploration of ways to minimize focus effects in eliciting social value.
9525571Principal Investigators: Dietz, Thomas
Stern, PaulTechnical Liaison:Research Organization:
George Mason UniversityFunding Agency/Program: NSF/ValuationGrant Year: 1995Project Period: October 1, 1995 - December 31, 1999Cost to Funding Agency:
- Project Reports
- Final Report
Description/Objective of Research
The problem of valuation is central to environmental policy. But valuation is difficult because many environmental policies have substantial impacts on aspects of the world whose social value may not be adequately captured by market prices. Contingent valuation methods (CVM) have been used as an alternative or supplement to market prices as a means of assessing social value. If they work, these methods have the advantage of capturing values that are not captured in market prices. CVM proceeds by asking a representative sample of people what they would be willing to pay to preserve or create some environmental benefit or avoid some environmental cost.
This project has two goals. First, we have linked willingness to pay (WTP), the concept that underpins contingent valuation, to the emerging social psychological literature on environmental concern. This is theoretically important as it bridges two of the more robust literatures in the environmental social sciences. It is also of practical importance because it advances understanding of the beliefs and values that underpin WTP and that are related to conflicts about environmental policy.
Second, we have engaged the emerging discussion on the use of deliberative processes in environmental policy and conducted an experiment to compare a standard CVM survey with a method of eliciting WTP that involves group deliberation. This provides a test of deliberative theory and may suggest methods for CVM studies that are less prone to some of the methodological weaknesses noted in the literature.
Since the project has multiple goals, we have used multiple methods in collecting data. Our investigation of the relationship between WTP and the social psychology of environmental concern is based on a national telephone survey (n=418). A question wording experiment with a convenience sample of college students was used to determine the effects of information provided about the impacts of climate change and the method by which payment would be made (payment vehicle) (n=228 with 12-17 respondents in each of 16 experimental conditions). The comparison of conventional WTP survey methods with deliberative methods used both a mail survey and a series of nominal group sessions. In both cases a random sample of college students age 30 or over was selected from a large suburban university. (The age restriction was to ensure that respondents had experience managing household budgets and thus would take their income constraint seriously in providing their WTP.) The mail survey generated 135 completed responses with a response rate of 66%. A total of 16 groups were held using the nominal group process to structure discussion. Groups varied in size from 3 to 7, with a total of 68 participants.
In each part of the study we used open-ended WTP items. While the literature favors close-ended, ballot-like items for estimating consumer surplus and social value, our goal is to understand the covariates of WTP. Thus the simplicity of the open-ended item seemed preferable to the more complex design and analysis required for closed-ended WTP. Because the distribution of open-ended WTP have about half the responses at 0 and have a very long-tailed distribution for the non-zero responses, all analyses are conducted using smoothed logs, where half the smallest non-zero response is added to each response before taking the log. Models were estimated by ordinary least squares with robust standard errors. Other methods (e.g. logit, ordinal probit, robust regression) yield essentially the same results.
Accomplishments and Research Results
Social Psychology of WTP. Before examining the social psychological determinants of willingness to pay, we conducted an analysis comparing the prevalent theories of environmental concern: the Stern/ Dietz values/ beliefs theory, the norm activation theory that was initiated by Schwartz, Dunlap et al.’s New Ecological Paradigm (NEP) approach, the “cultural” theory implemented by Dake and Inglehart’s post-materialist theory. Our analysis suggests that a theory that takes account of values, beliefs and norms (VBN) provides the best predictions of pro-environmental behavior. It also suggests that environmental concern might best be conceptualized as support for the environmental movement and its goals. Our work indicates that there are at least three dimensions to this support: consumer behavior, political activism and willingness to sacrifice (Stern, Dietz, Abel, Guagnano and Kalof 1999). Because gender has been one of the most consistent predictors of environmental concern and a substantial literature discusses gender and values, we have also looked in some detail at gender differences in values (Kalof, Dietz, Stern and Guagnano 1998).
We hypothesize that WTP as expressed in a CVM survey on environmental policy is a subset of the more general concept of willingness to sacrifice. Willingness to sacrifice (WTS) includes WTP money in taxes, contributions or consumer prices but also includes the “hassle” factor involved in changes in lifestyle and a willingness to reduce material standard of living for environmental protection, time spent in activities to minimize environmental impact and other factors. The survey contained two open-ended WTP items, one asking about gasoline taxes, the other about preservation of tropical forests and three general “willingness to sacrifice” items. We find that the two WTP items load on the same factor as the three WTS items. The NEP scale that measures beliefs predicts both WTP measures and WTS. Income is positively related to WTP but not WTS. Traditional values are negatively related to both WTP measures but not WTS while altruism is positively related to WTS but not WTP. WTS is a strong predictor of WTP, net of other variables but values, beliefs and income retain some effect. This suggests that while WTP is closely related to the more generic concept of WTS, respondents are taking account of other factors in answering WTP questions. Some of these factors (i.e. income and beliefs) are consistent across the content of the WTP question, others are specific to the environmental problem being assessed. (We anticipate submitting a paper based on this work in early January.)
Focus Effects. Previous theoretical work had suggested that focus effects may be important in WTP responses. We conducted a survey experiment with 2 crossed experimental conditions. The first condition used four versions of the kinds of impact information given to the respondent: no impact information, information on the area that might be flooded from sea level rise, information on estimated heat related deaths or information about both. The second condition varied the payment vehicle suggested: taxes, prices, contributions or no vehicle specified. In previous work, we have found that payment vehicle influences both the stated WTP directly and indirectly through an interaction effect with values and beliefs. We hypothesized that information provided about impacts might have a similar effect. However, none of the treatment effects were significant in our experiment. In response to these results, we altered the design of the experiment to compare conventional CVM with a more discursive method allowing for more power in that analysis. These results suggest that CVM surveys may be more robust with regard to question wording effects than is sometimes suggested.
Conventional versus Discursive WTP. One of the basic concepts underpinning our project was the idea that people may make more rational decisions after a structured group discussion than acting in isolation (Dietz and Stern 1998). Our experiment to examine this hypothesis had two conditions. In one, respondents were mailed a standard CVM survey. The survey described the impacts of climate related sea-level rise on the U.S. and the possible health effects and described a policy of tree planting to reduce greenhouse gas emissions. Respondents were asked to think about what factors they would take into consideration before answering a WTP question on tree planting. After answering the WTP question they were asked to list the things they took into account and identify the three most important factors in their decision. In the discursive protocol, small groups met with a group facilitator. The WTP question was posed to the group. Following nominal group protocol, group members first were asked to silently write things to take into consideration in answering the question, then the individual lists were pooled into a group list. After discussion, the participants were asked to privately complete the survey instrument.
Because of the amount of data involved, we have not completed our analysis of this experiment. We are comparing the central tendency and variability of stated willingness to pay between the two modes of elicitation. We are also comparing the survey and group process with regard to the reasons participants say they considered in making their decision and the influence of those reasons on stated WTP. Our current plans are to submit papers based on this analysis to refereed journals during Spring 2000.
Our results to date lend some support to the use of CVM in environmental policy. While WTP is related to the key variables of the values/ beliefs/ norm theory of environmental concern, and especially to the willingness to sacrifice dimension of that theory, WTP is not identical to willingness to sacrifice. This is consistent with the notion that respondents are offering an expression of their concerns that takes into account other factors, such as their income. The negative results from our question wording experiment also lends support to the use of CVM in that changes in payment vehicle and in the kinds of impact information provided in a survey had no effect on stated WTP. Our results also suggest that variation in individual willingness to pay is driven substantially by differing beliefs about the sensitivity of the biosphere to human intervention and to some extent by differing values regarding environmental change.
Dietz, Thomas and Paul C. Stern. 1998. “Science, values and biodiversity.” BioScience 48(6): 441-444.
Kalof, Linda, Thomas Dietz, Paul C Stern, , and Gregory A Guagnano. 1998.
“Value Priorities and Gender in the United States.” Fairfax, Virginia: Human Ecology Research Group, George Mason University. (Currently under review)
Dietz, Thomas, Paul C Stern, Linda Kalof, and Gregory A Guagnano. 2000. “A Social Psychological Model of Stated Willingness to Pay” Fairfax, Virginia: Human Ecology Research Group, George Mason University. (Currently in preparation)
Stern, Paul C, Thomas Dietz, Troy Abel, Gregory A Guagnano, and Linda Kalof. 1999. “A Social Psychological Theory of Support for Social Movements: The Case of Environmentalism.” Human Ecology Review 6: 81-97.
- Project Status Reports
The kind of logic used in making decisions depends on the context in which the decision is made. Decisions about expensive choices, such as the purchase of a car, house, or college education, usually involve research, conversation, reflection, and comparison of alternatives. Such decisions probably are well described by rational choice theory. Routine decisions, such as the purchase of non-durable consumer goods, are repeated frequently and allow for learning over time. Thus, rational choice theory is applicable here as well. But decisions that assign value to environmental goods and services, including non-consumptive uses, may not be well described by traditional rational choice theory. Many environmental issues are novel, and the public will have limited familiarity with them. Indeed, being asked about such issues in a contingent valuation survey may be the first time many individuals have heard about such problems.
We suggest that when presented with novel phenomena, and when required to make a quick decision (as in responding to a survey), people use cues contained in the context of the question to decide how the question links to their core values. Different cues will highlight different values and will lead to different decisions – in the case of valuation surveys, to different stated preferences. The strength of the focus effect should depend on how familiar an individual is with the objects being described in the question.
In previous studies, we have shown that payment vehicles (taxes vs. contributions to a fund) lead not only to different stated willingness to pay but also to differences in the determinants of willingness to pay. We review those results and also present preliminary analyses of a national survey with an embedded experiment in which we manipulate question wording to focus respondents on different values for some relatively familiar and some relatively unfamiliar problems. We conclude by discussing the plans for the second phase of our study in which we are experimenting with deliberative approaches to environmental valuation.