Practical tools for value transfer in Denmark – guidelines and an example
5 Applying the guidelines – An illustrative example
To show the practical use of the guidelines in chapter 4, we will now apply them to the case of restoring the Skjern River wetlands and river landscape. This case illustrates the transfer of both use values (both consumptive and non-consumptive use) and non-use values. The valuation of specific components might deserve more in-depth information, consideration and discussion of the conditions at the policy site, but also at the study sites. Thus, this example should not be regarded as an authoritative valuation of restoring the area.
Dubgaard et al. (2003) is a very good example of how to conduct a cost-benefit analysis (CBA) of a nature restoration project. Dubgaard et al. (2003) used a unit value transfer methodology to asess the following social benefits:
• Value as a factor of production (farm land, reed production etc.)
• Ecosystem services (retention of nutrients, flood risk reduction etc.)
• Consumptive outdoor recreation values (hunting, angling)
• Non-consumptive outdoor recreation values (hiking, boating, wildlife observation, etc.)
• Non-use value which individuals place on the mere existence of biological diversity.
The first two benefit components were valued using market prices and the replacement cost method. As the focus of this value transfer guide is the Stated Preference methods (CV and CE) and the Revealed Preference methods (TC and HP), this example will consider the last three benefit components.
Dubgaard et al. (2003) make extensive use of unit value transfer to estimate the social benefits of restoring the Skjern River. In this chapter we will illustrate how the value transfer guidelines could be applied to the affected non-market, use and non-use value components; and how this could change the magnitude of the estimated benefits in Dubgaard et al. (2003).
The example will be described in terms of each of the following seven steps of the practical guidelines:
1) Identify the change in the environmental good to be valued at the policy site
(i)Type of environmental good
(ii)Describe baseline, magnitude and direction of change in environmental quality
2) Identify the affected population at the policy site
3) Conduct a literature review to identify relevant primary studies
4) Assess the relevance and quality of study site values for transfer
(i) Scientific soundness; the transfer estimates are only as good as the methodology and assumptions employed in the original studies
(ii) Relevance; primary studies should be similar and applicable to the “new” context
(iii) Richness in detail; primary studies should provide a detailed dataset and accompanying information
5) Select and summarize the data available from the study site(s)
6) Transfer value estimate from study site(s) to policy site
(i) Determine the transfer unit
(ii) Determine the transfer method for spatial transfer
(iii) Determine the transfer method for temporal transfer
7) Calculating total benefits or costs
Step 1: Identify the change in environmental good to be valued at the policy site
(i) Type of environmental good
This is both a type L and E environmental good (see chapter 4, step 1); i.e. a wetland providing landscape aesthetic and recreational opportunities (consumptive and non-consumptive recreation), and providing biodiversity. Thus, the project will impact both on use and non-use values.
(ii) Describe baseline, magnitude and direction of change in environmental quality
Dubgaard et al. (2003) describe the Skjern River restoration project below. Note that: (i) there are impacts both on recreation and biodiversity, (ii) this is a restoration project, and (iii) the short-term change in quality/quantity of the environmental good as a result of the described project is well known since this is an ex-post study (the River Skjern restoration project was completed in 2003).
.
With an average discharge of 35 m³/s the Skjern River is the largest Danish river by volume. It has a catchment area of 2,500 km² and a length of 95 km. The river discharges into the Ringkøbing Fjord - a shallow 300 km² costal lagoon, which is connected with the North Sea by a floodgate. The Skjern River delta and Ringkøbing Fjord have been designated as an international bird protection area for wading birds and as a EU habitat area. The river system is home to a number of red-listed species in Denmark. The location of the area is presented in Figure 2.
Figure 2.
Location of the Skjern River Project Area. Source: Dubgaard et al. (2003, fig. 1)
(ii) Describe (expected) change in environmental quality
According to Dubgaard et al (2003),: The primary purpose of the Skjern River project is to re-establish a large coherent nature conservation area with good conditions of life for the fauna and flora connected with wetlands. Before the 1960s the Skjern River floodplain was managed as extensively grazed meadows and hayfields. During the 1960s the lower 20 km of the river were straightened and diked. Pumping stations were established and 4,000 ha of meadows were drained and converted to arable land. In 1987 the Danish Parliament decided to initiate studies of restoration possibilities. Detailed surveying and designing started in 1995 and re-meandering work began in 1999. The river restoration works were completed by mid-2003.
Of the 4,000 ha reclaimed in the 1960s 2,200 ha were included in the project. The entire project comprises the following initiatives (Danish Forest and Nature Agency, 1998):
• The lower 19 km of channelled river have been turned into a 26 km meandering course.
• The River has been laid out with several outflows to the Fjord, which, in time, will create
a delta of app. 220 ha.
• Creation of a lake of approximately 160 ha.
• Re-establishment of the contact between the River and riparian areas by permitting
periodical floods of land within the project area.
• Transfer of 1,550 ha of arable land to extensive grazing.
The project will improve the water quality of the Skjern River system, living conditions for the wild flora and fauna, together with the recreational value of the area. The flora of riparian areas and the River will become more diversified and is expected to include rare species like Elisma natans in flowing water and calamus, water soldier and cowbane in still water. The area will become increasingly attractive to breeding birds, especially species specifically found in wetlands, reed, and meadows. A significant factor is the establishment of a large coherent area with improved possibilities for nesting and feeding. Key areas free from hunting and disturbance will be established to ensure resting, foraging, and breeding possibilities for birds and mammals. Bird species like kingfisher, bittern, water rail, crake, reed bunting, reed warbler, bearded tit, ducks and geese are expected to breed in the area. Large amounts of ducks and geese have already been registered and the populations of migrating and resting birds are expected to increase. In addition to a varied bird life, an increase in the population of endangered amphibian and reptile species is expected, and populations of otter in central Jutland will be able to migrate as a result of the removal of man-made barriers in the landscape. Improved water quality, environmentally friendly maintenance practices, and the re-establishment of spawning grounds will have a positive effect on the salmon and trout populations in the River system.
The River discharges into the Ringkøbing Fjord, which is a shallow costal lagoon considerably affected by excessive loads of nutrients. A major programme is underway aiming at reducing nutrient emissions to the Fjord. The Skjern River project will contribute to this programme due to the retention of nutrients and other particles in the wetlands of the river valley. The reduction is partly obtained by the transfer of arable land to more extensive land uses, but more significantly by the re-creation of the natural ecology and hydrology of wetlands, which will filter and absorb nutrients and other particles in the river water during flooding.
Finally, the nature restoration project will increase the possibilities for recreation in the area. The size of the project area facilitates activities such as hiking and biking, boating, camping, studies of flora and fauna, angling and hunting. Accessibility has been improved by the establishment of new trails, access to grazing areas and the construction of outdoor recreation facilities.
Step 2: Identify the affected population at the policy site
Due to its sheer size, the Skjern River has few, if any, perfect substitutes in Denmark. There are areas in Western Jutland that have biological characteristics, which are quite similar to the Skjern River area. The Skjern River has an original Atlantic salmon population and is an important stopover for migrating birds along the North Sea. Thus, the Skjern River can be considered a unique environmental good, and is therefore of national importance. Thus, all Danish households should be used as the affected population when calculating the non-use values. For consumptive and non-consumptive use values we will of course only consider the respective user groups.
For the remaining steps we will report separately on each of the three types of values considered: i) consumptive use, ii) non-consumptive use, and iii) non-use values.
Step 3: Conduct a literature review to identify relevant primary studies
CONSUMPTIVE USE VALUES
Both hunting and angling in the area will be affected by the restoration projects, and increased benefits are expected from both these types of consumptive recreational activities. As Dubgaard et al. (2003) point out there does not exist any Danish primary valuation study on hunting, and only one primary study on recreational fishing (which is part of a Nordic study; see Toivonen et al. 2000). There are also other primary valuation studies of recreational fishing for trout and salmon in both Norway and Sweden (based on a search in EVRI (www.evri.ca) and VALUEBASESWE in appendix F). However, since the guidelines recommend using national valuation studies, and this primary Danish study values recreational fishing for the same species (salmon and trout), we report only this study here.
For hunting, there is no study in the Nordic countries of duck hunting. However, there is one primary valuation study for small game hunting in the mountains in Sweden (see Dalin 2000 in appendix F), and several in the US (for an overview see e.g. Walsh et al. 1990, 1992; Smith and Kaoru 1990; and Appendix 3, chap. 3).
NON-CONSUMPTIVE USE VALUE
Dubgaard et al. (2003) base their non-consumptive value transfer on a value per visit (activity day) from a Danish valuation study of another unique area (Mols Bjerge; Dubgaard 1996) However, also more recent Danish SP studies of forest areas, applying state-of-the-art methodology, should be considered for transfer to the Skjern River of less uniqueness than Mols Bjerge. Recent Danish and Nordic valuation studies of forest recreation areas have been reviewed, and can provide values for less unique areas than the Mols Bjerge. The most relevant study is Termansen et al. (2004) (see Appendix A) that value a day trip (i.e. activity day) to forests in Denmark in general. They find a mean WTP of 20-28 2005-DKK per activity day (which is lower than the more unique Mols Bjerge, valued at 38-56 2005-DKK).
NON-USE VALUE
Since Dubgaard et al. (2003) did their transfer study, one new very relevant Danish primary study has been performed that could be used in a unit value transfer exercise if the change in quality of the two resources can be assumed to be of the same magnitude. This is the Store Åmose-study by Lundhede et al. (2005) – see Appendix A.
Step 4: Assessing the relevance and quality of study site values for transfer
(i) Scientific soundness
CONSUMPTIVE USE VALUES
Recreational fishing
Toivonen et al. (2000) is a CV study (according to appendices J and K), with a survey instrument that was well constructed and carefully tested, but this was a mail survey with a response rate of 45 % in Denmark (somewhat higher in the other Nordic countries). In these mail surveys of the general public we very seldom get higher than 50 % response rate, and then of course the representativeness of the sample can be questioned. One large advantage of this CV study is that it was constructed for value transfer (and to test the validity of value transfer between the Nordic countries). However, this study, as most multi-country CV studies, is based on CV scenarios that are trade-offs between what is optimal design in each country and the need to use the same CV scenario and questionnaire in all countries to avoid that differences in valuation are due to methodological differences (but only reflect differences in individual preferences of people in different countries). Thus, some of the scenarios might be hypothetical in some countries even if they are very relevant in others. To conclude, the scientific soundness of the study is sufficient to serve as a basis for transfers.
Hunting
The Dalin (2000) is a high quality TC mail survey of hunters with a very high response rate (85%). Consumer surplus (CS) per hunting day was found to be 82 2000-SEK.
NON-CONSUMPTIVE USE VALUES
Termansen et al. (2004) was also based on a high quality TC study of recreationists, with an acceptable response rate (> 50 %).
NON-USE VALUE
Lundhede et al. (2005) performed a CE study in 2005 of a sample of both the entire Danish population and the local population, asking them to value a nature restoration project in the wetlands area of Store Åmose, which has no close substitutes. This study also values a nature restoration project, just like the Skjern River project, but Store Åmose is also very important as an archaeological site. The Skjern River project focuses more on preservation of biodiversity and is a more comprehensive project. The CE study of Store Åmose is based on a large number of observations (n=1,636). It had a response rate of above 50 % (51%), which shows that this internet based survey could easily achieve as high response rates as mail surveys.
(ii) Relevance/similarity in change in environmental goods valued
CONSUMPTIVE USE VALUES
Recreational fishing
Toivonen et al. (2000) value recreational fishing in a reasonably similar context to the Skjern River restoration project, as this study considers the same direction of change, i.e. an improvement in recreational fishing opportunities, but by opening up a new river to fishing rather than improving an existing one (which is the case in the Skjern River).
Hunting
Although the Swedish study (Dalin 2000) considers small game hunting for other species, and in the mountains rather than on wetlands, we assume that the activity quite closely resembles the one in the Skjern River area. Although this type of hunting in Sweden and Denmark is considered to be closely related culturally and institutionally, one should in practice compare the market prices of hunting in the two countries, expected catch rates, average distance (and, thus, costs and consumer surplus) to the site for the hunters, and quality of the hunting experience at the study and policy sites..
NON-CONSUMPTIVE USE VALUES
The Termansen et al. (2004) study is relevant as it a Danish study (and captures the preferences of the same pulation as the one that is considdred to be affected here, i.e. the national Danish population. It provides an estimate for forest recreation per activity day in general in Denmark, and not for a unique area as the Mols bjerge. However, one should keep in mind that general forest recreation could be somewhat less attractive than the experience of walking in the fields along the Skjern River, and thus, Termansen et al. (2004) study should be used as a lower estimate.
NON-USE VALUE
For non-use values Dubgaard et al. (2003) transferred a reported willingness-to-pay (WTP) per household per year from a CV study of Pevensey Levels in the UK (Willis et al. 1996), which seems to be very similar to the Skjern River. However, in the Pevensey Level study the costs per respondent is stated in terms of a few pence per year per households, and respondents then will anchor their WTP on this very low amount and provide an estimate of the expected costs rather than expressing their “true” maximum WTP based on the benefits they expect to receive. Then, instead of transferring this unit value estimate as WTP/household/year as the guidelines recommend, they generalized to the whole UK population and converted this to a per ha value per year. They did this to correct both for differences in size of the study and policy site, and the differences in population size between the UK and Denmark. However, by doing this they assume that households´ WTP increase proportionally with size, i.e. a strict assumption that is seldom fulfilled. To avoid this rather arbitrary assumption, the guidelines suggest that we use the WTP per household per year. Thus, the UK WTP per household per year should be converted to DKK in the year of study using Purchase Power Parity (PPP) corrected exchange rates, and then adjusted with the Danish Consumer Price Index (CPI) to 2005-DKK using the same procedure as outlined for the hunting value (see Step 6 below). However, now we have the new valuation study of Lundhede et al. (2005). This study is also valuing a nature restoration project, just like the Skjern River project, but the Skjern River project is more comprehensive.
(iii) Richness in detail; primary studies should provide a detailed dataset and accompanying information
CONSUMPTIVE USE VALUES
With regard to recreational fishing, both Toivonen et al. (2000) and Hansen (2005) have the advantage of being primary studies that were designed with value transfer in mind, even though the scenarios run the risk of being more hypothetical (i.e. in both cases respondents were asked to value an environmental good that was to become available close to their house).[17]
The Dalin (2000) study has the disadvantage of reporting values for small game hunting in total only, and not distinguishing between different species.
NON-CONSUMPTIVE USE VALUES
The Termansen et al. study provides very detailed values for non-consumptive
use per activity day, based on different estimation models for the Random Utility Model (RUM) of the Travel Cost study.
NON-USE VALUES
Lundhede et al. (2005) performed a CE study in 2005 of a sample of both the entire Danish population and locally of a nature restoration project in the wetlands area of Store Åmose, and very detailed data are available. An assessment of similarities and differences between Store Åmose and the Skjern River was collected in order to increase the quality of the unit transfer exercise.
Step 5: Select and summarize the data available from the study site(s)
NON- CONSUMPTIVE USE
Termansen et al. (2004) value a day trip to forests in Denmark in general. They found a mean WTP of 20-28 2005-DKK per activity day, while Dubgaard et al. (2003) found an estimate of 38–56 2005-DKK per activity day for non-consumptive use of the more unique Mols Bjerge area. We will use the former estimate as a lower estimate.
NON-USE VALUE
Lundhede et al. (2005) found a mean WTP of 500 2005-DKK per year per person (in the national sample) for changing Store Åmose’s currently “low biological diversity” to a higher level described as “high biological diversity”. We can use this estimate directly if we assume that this qualitative change in the biodiversity attribute valued in the CE of Lundhede et al. (2005), is similar to a change of similar magnitude in the Skjern River project. However, the biodiversity is not exactly the same at Store Åmose as in Skjern River, and the stated WTP per individual should be converted to a ”per household” basis to better reflect the household as the smallest decision unit.. The meta-analysis of Lindhjem (2007) also showed that WTP was significantly higher when stated on an individual basis compared to a household basis (possibly because people find the household budget constraint more binding than personal budgets). This result supports the use of stated WTP per person as an expression of WTP per household, arguing that this is the correct unit for non-use values and that multiplying personal WTP by the household size to get WTP per household would result in upward biased results.
Step 6: Transfer value estimate from study site(s) to policy site
(i) Determine the transfer unit
According to our guidelines, the transfer unit for consumptive and non-consumptive recreational use should be consumer surplus (CS) per activity day (defined as one person performing the activity one day).[18]
For non-use values the transfer unit is WTP per household per year from a study site of similar national importance, or using the results from a study site of regional importance as a lower estimate of the non-use values. In order to avoid the possible added uncertainty of international value transfer we will use the latter strategy and transfer estimates from the national sample in the CE study of the Store Åmose wetlands (Lundhede et al. 2005).
(ii) Determine the transfer method for spatial transfer
For both use and non-use values we will use the unit value transfer, as more input data are needed for a proper function transfer, and function transfer in most cases do not perform better than unit value transfer (with adjustments).
iii) Determine the transfer method for temporal transfer
For transfer in time we either use DKK directly from a Danish study or convert the foreign currency into DKK in the year of the study using Purchase Power Parity (PPP) corrected exchange rates; see http://www.oecd.org/dataoecd/61/56/1876133.xls. Then we use the Danish Consumer Price Index (CPI) to adjust to 2005-DKK. Since we transfer from many recent Danish transfer studies, we avoid the added uncertainty of converting from one currency to another. We assume that the value of environmental goods increases at a rate proportional to the CPI.
CONSUMPTIVE USE VALUES
Recreational fishing
Since the existing primary Danish study for recreational fishing in a salmon and trout river (Toivonen et al. 2000) calculated a CS per fisherman per year (instead of activity day), on the assumption that this is a new river, we (realistically) assume that the number of fishing days per angler in this new river will be the same as the increased number of fishing days per angler for the existing angler, as a result of the Skjern River Project. Then we can just adjust the CS per fisherman from Toivonen et al. (2000) to 2005-DKK, which is 901 DKK for trout and salmon fishing (according to Appendix A - and using the same procedure as described below for hunting). Using this estimate multiplied by the number of existing fishermen (fishing for salmon and trout) will likely provide a lower estimate of the increased use value of recreational fishing, since the number of fishermen might increase and/or the quality of fishing for other species might also improve. (However, the largest improvement is expected for trout and salmon).
Hunting
Dalin (2000) reports a CS per hunting day of 82 2000-SEK. According to the guidelines we transfer unit values between countries, taking account of differences in currency, income and cost of living between countries by using the Purchase Power Parity (PPP) corrected exchange rates; see http://www.oecd.org/dataoecd/61/56/1876133.xls . In 2000 the PPP-value for 1 US $ is 8.41 DKK and 9.20 SEK. Dividing 8.41 by 9.20 we get that the 2000-SEK-value has to be multiplied by 0.914 (i.e. subtracting almost 9 %) to convert it into PPP-adjusted 2000- DKK. Then we use Appendix H to adjust the value to 2005-DKK. Looking at the column for Denmark in Appendix H, we see that the Consumer Price Index (CPI) for 2000 and 2004 is 123.6 and 133.8, respectively Dividing the CPI in 2004 with the CPI in 2000 gives a factor of 1.083 (i.e. an inflation rate for this period, up until January 2005, of about 8 %.). Thus the 2005-DKK value becomes 82 x 0.914 x 1.083 = 81 2005-DKK per activity day.
NON-CONSUMPTIVE USE VALUES
Again CS per activity day should be used, and this is carefully reported by Termansen et al. (2005). As the unit values are reported in the original study in terms of 2005-DK, there is no need to adjust them over time.
NON-USE VALUES
As stated above, we will use 500 DKK/person/year from Lundhede et al. (2005) as our estimate of the WTP/household/year.
7) Calculating total benefits or costs
(Only total benefits will be calculated here in the Skjern River case).
CONSUMPTIVE USE
For the Skjern River restoration project (and in many other cases) the most difficult part is to predict the increased or reduced number of hunting days and angling days (and/or increased number of hunters and anglers) as a result of the project, both in the project area and in adjacent areas (and/or the increase in CS per activity day; e.g. increased number of hunting days as a result of increased game density due to reducing the hunting area by 1,045 ha, and more restrictive hunting in the area where this is allowed). This should be based on expert estimates from Environmental Impact specialists in hunting and fishing, and from local, regional and national fishing and hunting associations. Sensitivity analyses with different assumptions for the change in number of hunting and fishing days should be performed.
Dubgaard et al. (2003) used expert estimates to find that 5000 recreational fishermen per year used the area. They assume that this number will not increase due to the project, but that their CS per year will increase (as a result of increased number of activity days and/or quality of the activity days). As we have no information about the increase in activity days, Dubgaard et al. just assume that the reported WTP per fisherman in Toivoinen et al. is representative for the annual WTP among the fishermen i Skjern River. This is a crude approximation, but in the absence of information about activity days for recreational fishing it is the best we can do since the national CV survey of recreational fishing reports WTP in this way. Thus, the annual benefit is 901 DKK/fisherman x 5,000 fishermen = 4,505,000 DKK. This benefit will accrue in all years to come, and thus we can apply equation 7 (p. 35) and calculate PV (with a 3 % social discount rate) as 4,505,000 / 0.03 = 150,166,167 DKK. With the recommended general transfer error bound of +/- 25-40%, this amounts to about 90 - 210 million DKK.
For hunting, Dubgaard et al. (2003), in absence of a valuation study to transfer from and expert assessment of the increase in number of hunting days, use the expert assessment of the expected increased price of a duck hunting licence per ha per year of 200 and 400 2003-DKK for public and private land, respectively (and 200 DKK/ha/year for areas surrounding the area). This should be seen as an estimate of increased producer surplus, and thus a lower estimate of the welfare impact of improved hunting (which could be seen as the minimum WTP among hunters to go duck hunting). Since we have no information of the number of hunting days, we unfortunately cannot utilise a transferred value per hunting day. Therefore, we will use the same approach as Dubgaard et al. (2003; p. 18-19): Private land: 240 ha x 400 DKK/ha/year; public land: 840 ha x 200 DKK/ha/year; and surrounding areas 1.045 ha x 200 DK/ha/year; which comes to a total of 473,000 2003-DKK/year. According to appendix H, the Danish CPI in 2003 (to give January 2004–values) was 132.3, while CPI for 2004 (January 2005) was 133.8; i.e. a percentage relative increase of CPI equal to 1 %. Thus, this amounts to 478,000 2005-DKK/year. Assuming again that this benefit will accrue in infinity, we can use the same PV equation as above, i.e. PV = 478,000 / 0.03 = 15,933,333 DKK. With the recommended general transfer error bound of +/- 25-40 %, this amounts to about 9,6 – 22,3 million DKK.
NON-CONSUMPTIVE USE
Again we need an estimate of the additional number of non-consumptive activity days to multiply with the unit value of WTP/activity day to arrive at total non-consumptive use values. Dubgaard et al. (2003) make an expert assessment (based on the difference in the visitation numbers for the Tipper Penninsula near the Skjern River area and the visitation to the Mols Bjerge) and recommend using an increase 90-100,000 visits (which they state are on the lower side). Since each visit is expected to last one day, this is also the number of increased activity days. Thus, annual total non-consumptive benefits are 100,000 activity days x 20-28 DKK/activity day = 2-2.8 million DKK. With infinite time horizon and a 3 % social discount rate this gives a PV of about 67-93 million DKK. With the recommended general transfer error bound of +/- 25-40%, this amounts to about 40-130 million DKK.
NON- USE VALUES
Since the Skjern River is unique in a national context, we will transfer the unit value found for Store Åmose for the national sample, for an increase from “low to high” for the biodiversity characteristic of the Choice Experiment (as this is the same general increase in biodiversity as we expect as a result of the restoration project). All Danish households will be defined as “affected”, and thus their mean WTP per year multiplied by the number of Danish households will be a measure of the annual non-use value of the Skjern River project.
If we assume that Skjern River is unique, and therefore benefits accrue to all Danish households, annual non-use value is 500 DKK x 2,516,682 households (according to Statistics Denmark 2006) = 1,258,341,000 DKK. With infinite time horizon and 3 % social discount rate this gives a PV of about 41.944 million DKK. With the recommended general transfer error bound of + 25-40%, this amounts to about 25,166 – 58,721 million DKK. This assumes that the Skjern River is considered to be a unique resource for Denmark. If we assume that there are close substitutes at the national, but not on the regional level, the observed WTP should be multiplied only by the regional population of Jutland which (according to Statistics Denmark) was 2,504,036 in 2006. With an average household size of 2.21 this gives 1,133,048 households. This could serve as a sensitivity analysis for calculating the total non-use benefits, which account for 1,133,048 households x 500 DKK/household = 566,524,000 DKK annually. PV with infinite time horizon is then 18,884 million DKK. Thus, with the same general transfer error the total non-use value is about 11,330 – 26,437 million DKK.
As this example shows, the total benefits are dominated by the non-use values. It is therefore very important to select a study to transfer from that was based on the population at the same geographical level as we have defined for our “affected population” at the policy site. By using a WTP estimate from a national survey to value a good where only a regional population is affected (as we assume in the sensitivity analysis above), we might underestimate total WTP if WTP per regional household for a regional good is higher than WTP per national household for a national good (assuming a distance decay in WTP), and overestimate if the opposite is true. Also, one should select a study where the reported WTP is not biased downwards due to anchoring on the stated (low) costs per household per year.
Another general lesson is that in value transfer, one should try to use the value as reported in the identified, suitable primary study, without conversion to an area-based unit[19]. This conversion involves strict assumptions that can easily be critized and therefore reduces the reliability and acceptability of the estimated aggregate benefits, and thus also the outcome of the CBA. A recent meta-analysis of CV and CE studies of non-timber benefits of forests (mainly-non-use value) in Norway, Sweden and Finland also clearly shows that WTP does not depend on the size of the forest area (Lindhjem 2006). This is probably due to the fact that people care for and state their WTP for the biodiversity preserved and other complex aspects of forests, rather than the sheer size of the area.
Footnotes
[17] In Hansen (2005) the respondents were asked to value the availability of a a new (restored) wetland/river (in terms of higher biodiversity, improved quality of landscape and free accessibility) close to their home or summerhouse, while Toivonen et al. (2000) asked respondents to value two different rivers (one with trout and salmon, and one with whitefish species) near their home becoming available for recreational fishing.
[18] Increased producer surplus (net income) to the private owners of the hunting and fishing rights (as described by Dubgaard et al. 2003) should be added to the increased consumer surplus described in this report. When transferring consumer surplus estimates from a study site, one should be clear on whether an access fee had been paid when the fishermen stated their WTP or not. If there is a zero access fee there will be a higher CS, and thus no producer surplus (since the landowners have no income). If there is an access fee, the net producer surplus should be added to the CS to get net social surplus of the recreational activity.
[19] Note, however, that in Dubgaard et al. (2003) it is not this area-based conversion of WTP that results in low aggregated benefits, but rather the fact that WTP per household is very low due to transfer from a CV survey where respondents anchor their WTP on the reported very low costs of restoration per household when divided on the national population.
Version 1.0 December 2007, © Danish Environmental Protection Agency