3 Corresponding author regarding the Danish part of the study: AKF (Institute for Local Government). Arguments for the existence of “public” preferences as part of sets of multiple orders that individuals possess have received the attention of some of the most influential individuals of the 20th century. In that article, J-S also distinguishes a paternalistic version of true altruism, in which it is not the utility level but the environmental situation that is at stake.
He notes that this latter result may seem counterintuitive because consumers have already reduced their consumption of the externality-causing good to some extent due to paternalistic altruism. The explanation is that the social optimal has also changed, partly as a result of altruism. Complicating issues are the common thread in the psychological literature, also with venerable roots (e.g., Lapiere, 1934), which asserts the lack of well-defined preferences over unknown choices or their application in unknown, hypothetical situations (Fischhoff, 1991).
And how the questions are asked will be a very sensitive matter, because we would not want to tip off the respondent about the identity of the socially "correct" answer. What is labeled here as "pure public good" corresponds to the provision of the good after the students are likely to graduate. The rest of the paper is organized as follows: Section 2 contains a more complete description of the study method, including the framework and the management policies as well as information about the samples (sizes, characteristics and response rates).
Description of Research Method
7 In Denmark, most socioeconomic background variables were drawn from official registers and were also available to non-respondents. Three different frameworks constituted the introduction to these questions about desired intensity levels; (4) WTP questions;6 (5) requires the respondent to indicate whether he considered himself (including immediate family), other potential users, flora and fauna in the area, or future generations when answering questions about desired intensity levels; and (6) standard questions about socio-economic background.7 A question about what or who the respondent had in mind when answering the intensity level questions was included to obtain some evidence of what the respondents thought (or at least wanted to claim) that they were doing. a version of this question asked which of the four motivations was most important in influencing the responses given. Choose one of the levels listed as the best and one as the worst based on what you think should be available to visitors.
The Radnor area is unusual in that it preserves a small portion of the hilly Tennessee forest in Metropolitan Nashville. Some of these were computer manipulated to more accurately illustrate the text description of the respective ones. Together, the framework and the description of the indices (in text and photographs) constituted the description of the "good" being asked about.
Note that the frame text made up only a small portion of the total space used to describe the good. We hypothesized that if these were expressed, they would be seen as differences in mean levels and ranking of most and least preferred intensity levels from the reference point. For the benchmark case we expect that the typical pattern will be that the most preferred intensity level is lower than the least preferred level (eg the most preferred intensity level might be 3, while the least preferred might be to be 5).10 If public frames successfully induce sympathetic preferences, we expect these to be revealed as an inversion of the "form" of the relative.
If the third frame were successful in inducing committed preferences, we would expect these to appear as what might be called an accentuation of the referent's private pattern. The focus of the survey is on the analysis of respondents' preferences for the intensity of forest and lake use, as expressed in the answers to the most and least preferred questions on management intensity, because, as suggested by the survey from others, the WTP questions could themselves be a trigger for individualistic or private preferences (Stern and Dietz, 1994). In the US, a higher percentage of questionnaires with the dedicated frame were completed and returned than in the other two versions.
However, this difference appears to be more a function of the geographic distribution of observations. Unfortunately, respondents residing in zip codes near Radnor Lake received a higher proportion of the framed version than respondents in zip codes further away).
Empirical results
In the US, 78% of respondents who received the committed frame, but only 58% of respondents who received the public, chose the highest intensity level as their least preferred. In Denmark, 85% of respondents who received the mandatory framework chose the highest level as the least preferred, while 80% of those who received the public framework made the same choice. As a first approximation, the means of the most and least preferred levels can be used to summarize the differences in the distribution of the responses to these questions.
These tools are presented in table 2, where we also show the expectations regarding the relative size of. However, simple t tests for differences in means do not account for differences. In general, we would expect the signs of the parameters in the probit models for the relative order of most preferred and least preferred levels to be similar to the parameters found in the ranked probit models for the most preferred levels, because of the way in which the binary variable is defined.
The coefficient for the public frame is positive (as expected) in six of the eight regressions, but only significant in four of these regressions. 18 Distance is not included in the Danish estimates because multicollinearity between VISITAREA and DISTAN CE affects the reliability of the coefficient estimates. Users are familiar with the parks as they are currently managed and likely largely approve of the existing rules.
A comparison of the distribution of these responses per frame (not shown here) suggests that the frames have, at best, a limited impact on the self-reported problem. Here, the effects of the frames on self-reported concern are more formally examined in probit models, where the concern dummy variables are the endogenous variables. Focusing on the effect of the frames on self-reported concerns, it appears that in some cases the frames had a significant impact on this concern.
In all five of these cases, the sign of the significant parameter corresponds to the a priori expectations. The effect of self-reported concerns on the most preferred levels is summarized in Table 6 (the same socio-economic and activity variables that were included in the “original” regressions reported in Appendix Two are also included, but the parameters for these variables are not reported).21. This is consistent with the results in Table 5, in which the public frame had a small effect on the respondent's reported consideration.
This is what the analysis of the respondents' concerns and the desired level of intensity suggests.
Summary and conclusion
This difficulty may explain why the public frame did not have a clear effect on expressed preferences. On the other hand, this may just mean that we had better luck choosing the text we used to try to elicit committed preferences compared to that for the public frame (ie, the text was “stronger” for the committed frame). Furthermore, it may also be the case that our a priori hypothesis about the form of expression of public altruism was too simple (or even wrong).
Our expectation was that those respondents who were asked to think about others would choose higher intensity levels because these higher intensity levels would, at least implicitly, satisfy a wider variety of tastes and thus allow more people to use the natural area for a wider variety purposes. However, it can be argued that the opposite is true, i.e. the publicly or altruistically motivated person who thinks in terms of external effects may be paternalistic and opt for lower intensity levels to protect other users from the noise and general disturbance associated with high accompanying usage levels. This same person may even want to use a vehicle in the woods himself (high intensity level), but as a public (paternalistic) altruist he may vote for rules that would make it impossible to protect others.
In this case, it would be difficult or impossible to reveal any differences in private and public preferences, at least in the way we have pursued here. Such disregard for the implications of serving many people would be consistent with the results reported in Table 6, where preference intensity levels were regressed on self-reported concerns (and other variables). In any case, there remains much room for additional empirical exploration of preference mutability, but we believe that the technique used here, which focuses on preference patterns for described policies in the absence of monetary considerations, is a promising technique for undertaking this.
34; The Importance of the Ethical Voter: An Estimate of Altruism", in Stefano Zamagni, (ed), The Economics of Altruism. Philosophy, Economics and the Contingent Valuation Method" in Hausman, Jerry A. ed) Contingent Valuation: A Critical Examination. Cambridge Studies in Philosophy and Public Policy. Cambridge; New York and Melbourne, Cambridge University Press, 1988, pages x, 271.
Intensity Index
Models for the most and least preferred intensity index Table A2.1 Regression of framing effects in U.S
