Hasler B., 1), Jacobsen, J.B., 2), Lundhede, T.H., 2), Martinsen, L., 1), Thorsen, B.J., 2). Værdisætning af natur- og kulturgoder. Et metodestudie af betydningen for ændringer i skala og betalingsformat. 2008. Institutioner: 1) Afdeling for Systemanalyse, Danmarks Miljøundersøgelser, Aarhus Universitet.  2) Skov og Landskab, Det Biovidenskabelige Fakultet, Københavns Universitet. Danmarks Miljøundersøgelser, Aarhus Universitet. 78 s. - Faglig rapport fra DMU Nr. 695 . 

 

Summary

This project report aims at adding to the development of methods used to value public goods, focusing of developments and tests of importance for the practical policy relevance of stated preference methods used for evaluating nature and environmental projects. When used for practical decision making in Denmark the credibility of the methods and results has been questioned, among other reasons because the aggregated benefits have shown to be very high compared to the costs. Several causes for overestimated aggregate benefits are examined in this report, and split samples are used for testing these potential causes for overestimation and biased aggregation. The tests are applied in a CE-study of the Danish population’s willingness-to-pay for a nature restoration project in the Store Aamose area in Western Zealand, which is supposed to improve biodiversity and recreational opportunities as well as continued preservation of significant, cultural heritage sites buried in the top-soil in this specific agricultural area.

 

Prior to the launch of the whole survey three focus groups were used to test the questionnaires. One of the main foci of these was participants’ perception of the valuation questions, and the questionnaire was revised following each focus group. The survey was internet based, and was conducted by the professional survey institute GALLUP. The survey was submitted in December 2006.

 

These particular issues are examined:

• Scale and scope sensitivity and interpretations by the respondents,

• Aggregation issues:

• whether the WTP should be aggregated over households or individuals, and

• whether the WTP should be interpreted as a one time payment or an annual payment each year over a indefinite time horizon.

 

Scale and scope sensitivity

The project in Store Aamose is only one of several projects currently considered to be of relevance for the coming years’ publicly funded nature restoration projects. At the same time a number of national nature parks are considered launched or are already launched. The obvious question raised by the policy makers is, if the value derived from an environmental valuation study of one of the projects depends on the overall scale of the Danish nature restoration activities? For instance, would people state the same willingness-to-pay for this particular project if they knew it was going to be one out of three or seven projects producing environmental goods and services of similar character, i.e. these projects could be considered as substitutes? The test conducted examines whether the WTP differs if one single site is improved, compared to a situation where the improved site is number three or eight in a line of improved sites. This approach is applied by presenting three different splits of respondents with the same initial and background information and questions, but where split 1 is informed that the choice experiment only deals with improvements of Store Aamose, in split 2 the respondents are informed that two national parks will also be implemented prior to Store Aamose, and in split 3 they are informed that seven national parks will be implemented prior to Store Aamose. Pictures and a map of these sites are provided, and in the choice experiment the same choice of different protection options for Store Aamose is presented for all respondents in all three splits. The only difference is that the respondents in the “scope splits” 2 and 3 are informed that the national parks will be implemented, and that these projects are paid for by Danish taxpayers. A follow up question was asked whether the respondents thought that they valued Store Aamose, or that they valued all the national parks presented and Store Aamose.

 

Aggregation

The question of aggregation from the respondents of the single questionnaires to the Danish population was tested by split samples as well. The payment unit was tested, i.e. whether respondents perceive the payments as individual or household payments. In the basic split we used an increase of the individual tax payments as the payment vehicle, while the alternative split presents that the payment vehicle is an increase of the household tax level. Follow up questions are asked addressing whether the payment was perceived as household or individual payments in each of these splits. Further, a test was applied to test whether the respondents perceived the payments as a single one-time payment or as an annual payment occurring every year over a 20 years period. The results of the basic split with an annual payment were compared to a split where respondents were told that the payment was a one-time payment. Follow up questions were asked to both of the splits on whether they perceived the payment as an annual or one-time payment.

 

Results

The results clearly demonstrate that respondents, through their choices in the experiment, can relate to the level of attributes, and apply internal consistency and scope sensitivity within the choice experiment. The results of the external scope test are not as clear. For biodiversity improvement in Store Aamose WTP is reduced somewhat when the national parks are introduced, indicating that the national parks work as substitute for the Store Aamose-restoration. The WTP does not change differently between the two splits of two and seven national parks respectively, and the data from the two split were pooled for further assessment. The results from this pooled assessment show that the WTP for biodiversity improvement in Store Aamose reduces between 10 and 20 % as a result of the introduction of national parks/substitutes. For the protection of artefacts this picture is not repeated as WTP is not declining for this attribute when the national parks are introduced. This can be explained by the fact that Store Aamose and the national parks are not substitutes regarding artefact protection; the artefacts in Store Aamose are unique. For the artefacts the preferences seem to be more lexicographic than the biodiversity protection.

 

The results of the differences between eliciting individual and household WTP measures reveal that a surprisingly large number of respondents are uncertain to whether they are answering as households or individuals, but that most respondents’ perceive the payments as individual. A direct result could therefore be to ask respondents for individual WTP in coming valuation studies. However, testing the hypothesis that the household WTP exceeds the level of the individual WTP, reveal an unclear picture. The results indicate altruism because the sum of individual WTP exceeds the sum of household WTP. A more nuanced conclusion is that the researcher should take care that i) elicitation units are stated clearly, ii) payment vehicles are in accordance with the elicitation unit applied and iii) debriefing questions are applied to test if respondents perceived this key aspect of the study correctly. To assess the differences between household and individual payments we used the answers to the debriefing questions to pool the data across the samples, and this improved the credibility of the models. We conclude that the most conservative advice is to ask for the households WTP to avoid aggregation bias.

 

The level of annual WTP is lower that the one-time payments as expected, but the difference is not large. Most respondents answer that they anticipate that the payment is annual.

 

Full report i pdf-format (899 kB)