Recreational benefits of urban forests: Explaining visitors’ willingness to pay in the context of the theory of planned behavior
Introduction
The importance of forests for recreation and leisure activities is widely recognized. The systematic assessment of recreational and other non-market benefits is essential in order to incorporate these positive externalities of forests into policy and planning decisions. An established method to assess monetary values of non-market commodities is the contingent valuation method (Bateman et al., 2002; Mitchell and Carson, 1989). Contingent valuation (CV) surveys investigate individuals’ willingness to pay for the good being valued. Therefore, a hypothetical market for this good has to be constructed. The fact that economic values are derived from individuals’ preferences stated within a hypothetical scenario is one of the reasons why there are reservations regarding this method (e.g. Kahneman and Knetsch, 1992). An extensive literature dealing with this and other difficulties with CV studies shows that the method is controversial (e.g. Carson et al., 2001; Hausman, 1993). One argument raised by critics is that CV models only capture a minor share of the variance of willingness to pay and therefore the reliability and validity of the estimates have to be questioned.
Approaches to improve the model's quality include extending conventional economic models to include attitudes and other psychological predictors (Harris et al., 1989; Mitchell and Carson, 1989). A theoretical framework for this approach is provided by the theory of planned behavior (Ajzen, 1991). According to this theory, the immediate antecedent of a behavior is the intention to perform the behavior. Moreover, behavior depends on the perceived behavioral control, i.e. individuals’ perceived capability to perform the behavior. The three stipulated predictors of the behavioral intention are: attitude towards behavior, perceived behavioral control and subjective norm, which represents the social pressure to perform the behavior. This model of attitude-behavior relationships has been applied in CV studies assessing willingness to pay for different types of goods, e.g. leisure activities (Ajzen and Driver, 1992), abatement of forest regeneration cuttings (Pouta and Rekola, 2001) or drugs indicated for the treatment of Alzheimer's disease (Werner et al., 2002). In the context of contingent valuation, willingness to pay (WTP) is interpreted as behavioral intention. Impacts of attitudes, perceived behavioral control and subjective norms on WTP are analyzed in order to improve the understanding and explanation of WTP responses.
The overall impact of the psychological predictor variables proposed by the theory of planned behavior on the goodness of fit of models explaining individuals’ WTP has been assessed in the following two studies. Luzar and Cossé (1998) analyzed the impact of subjective norms and attitudes towards the good to be valued. The addition of these two variables improved the explanatory power of the models, as indicated by an increase in the adjusted R2. Werner et al. (2002) tested the influence of all three predictor variables according to the TPB, and also ascertained an increased adjusted R2 in the extended model. Both studies concluded that CV models explaining respondents’ WTP can be improved by applying an attitude-behavior based framework.
But WTP responses comprise further aspects than the bid levels analyzed in these studies: respondents to a CV survey are, in fact, faced with three decisions. They have to decide whether to accept the hypothetical scenario or not, whether to state a positive or zero bid and how much to bid (Alvarez-Farizo et al., 1999). The presence of individuals who are not willing to pay anything for the good under analysis is frequently observed in CV studies with open-ended question formats (Boyle and Bergstrom, 1999). Genuine zero bids can be separated from protest bids with debriefing questions to find out the reasons for zero WTP (e.g. Freeman III, 2003). In general, those respondents who do not accept the hypothetical scenario are said to be protest voters.
There are different strategies for coping with protest votes in the analysis of WTP responses. Ad-hoc solutions include removing protest votes from the sample. This procedure is problematic as it induces a sample selection bias if the group of protesters differs from the remainder of the sample (e.g. Jorgensen et al., 1999). Therefore, an analysis of protest votes has to be included in a CV study.
Previous studies integrating the theory of planned behavior (TPB) into CV analysis have focused on the explanation of bid levels. The present study additionally implements the TPB into the analysis of protest votes. It has been shown that attitudes towards paying for the good to be valued may also help to explain protest votes (Jorgensen and Syme, 2000). In our study, we extended the analysis of protest votes to take into account other psychological predictors proposed by the TPB. Respondents’ decisions whether to reply with a positive bid or a genuine zero bid were not analyzed in this study. In fact, we observed only a few genuine zero bids because we assessed use-values of a resource hold by actual users. Therefore, we decided to integrate them into the analysis of the bid levels. Based on these considerations, we split the analysis of WTP responses into two steps. First, we identified protest votes and second, we examined the amount non-protesters were willing to pay.
The main purpose of the research reported here was to explore the potential of the theory of planned behavior to improve the models explaining two different aspects of WTP responses. Taking our survey of the recreational values of an urban forest as an example, we extended conventional models to include the psychological predictors proposed by the TPB. This extension was applied in both steps of our data analysis in order to assess its ability to explain first protest votes and second how much non-protesters would be willing to pay. We tested the contribution of the psychological variables in both steps and compared conventional and extended models by evaluating different indicators of goodness of fit. A further purpose of the study was to determine how WTP responses were influenced by respondents’ characteristics, i.e. which variables affected their decisions whether to give a protest vote and how much to bid.
The remainder of this paper is organized as follows. In the next section we present the survey and questionnaire design for our empirical research. General aspects of the theory of planned behavior in the context of contingent valuation and its implementation in our example are outlined. We also describe the statistical methods and hypothesized outcomes of our data analysis. Results are presented in Section 3. Finally, Section 4 offers a discussion of our findings and some conclusions.
Section snippets
Survey and questionnaire design
Many studies of forest recreation behavior and the economic valuation of forest recreation benefits rely on on-site surveys carried out in the forest region under consideration. Whereas this procedure is very efficient in terms of surveying many forest visitors within a short time, it is subject to sample selection bias. Frequent forest visitors are overrepresented in the sample and the characteristics of non-visitors cannot be detected. Therefore, we decided to combine an on-site survey in the
Results
The results presented here are based on the off-site mail survey of randomly chosen residents of Zurich. Out of the 558 individuals answering the questionnaire, 65 (12%) reported that they had not visited the forests around Zurich during the previous 12 months. These people then did not have to answer the questions on forest visiting behavior and on willingness to pay for forest visits.5
Discussion and conclusions
Conventional models explaining willingness to pay typically have low explanation power. Mitchell and Carson (1989) proposed an R2 of at least 15% for reliable results of contingent valuation studies. But even if this benchmark is reached, a large proportion of variation remains unexplained. Therefore, extensions of the conventional approach have been proposed, e.g. developing CV models based on attitude theory (Harris et al., 1989). A theoretical framework was provided by the theory of planned
Acknowledgements
We thank several colleagues and the anonymous reviewers for valuable comments on preliminary versions of the survey and of this paper. Special thanks are due to Peter Elsasser for advice concerning the development of the questionnaire, and Jürgen Meyerhoff for valuable remarks on an earlier draft. Financial support from “Grün Stadt Zürich” and the WSL Research Programmes “Land Resources Management in Peri-Urban Environments” and “Forest Dynamics” is gratefully acknowledged.
References (35)
- et al.
Perceptions versus objective measures of environmental quality in combined revealed and stated preference models of environmental valuation
Journal of Environmental Economics and Management
(1997) The theory of planned behavior
Organizational Behavior and Human Decision Processes
(1991)- et al.
Contingent value measurement: on the nature and meaning of willingness to pay
Journal of Consumer Psychology
(1992) - et al.
Information bias in contingent valuation: effects of personal relevance, quality of information, and motivational orientation
Journal of Environmental Economics and Management
(1996) - et al.
Predictive habitat distribution models in ecology
Ecological Modelling
(2000) - et al.
Improving the contingent valuation method: a psychological perspective
Journal of Environmental Economics and Management
(1989) - et al.
Protest responses and willingness to pay: attitude toward paying for stormwater pollution abatement
Ecological Economics
(2000) - et al.
Valuing public goods: the purchase of moral satisfaction
Journal of Environmental Economics and Management
(1992) - et al.
Willingness to pay or intention to pay: the attitude-behavior relationship in contingent valuation
Journal of Socio-Economics
(1998) Attitude and belief questions as a source of context effect in a contingent valuation survey
Journal of Economic Psychology
(2004)