![Sample our Engineering & Technology journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5933/TOC-Subject-Sample-2021-Engineering-Tech.jpg"})
Abstract
Ecologists recommend preserving more of the old-growth forests in Norway, as half of the species have forests as their main habitat and many are in decline. We investigate benefits and costs over a 50-year period of increasing forest conservation from 1.4% of the productive forest area (the situation in 2007) to 2.8% (doubling), 4.5% (‘ecologists’ minimum’) and 10% (one goal suggested in public debate). The benefits are estimated based on a national contingent valuation (CV) survey of Norwegian households. Two independent measures of total costs are used: (1) the actual compensation amounts paid to forest owners and (2) results from a survey of forest owners’ minimum willingness to accept compensation to preserve. Results show that social benefits outweigh costs of the three conservation plans by a large margin. The middle option of 4.5% has the highest net present value. This result is robust to a range of assumptions, including considerations of potential hypothetical bias in willingness to pay estimates. The results of this cost–benefit analysis reflect the preferences of the general population, the authorities and the forest owners with respect to biodiversity and ecosystem services conservation, and supplement the expert opinion of ecologists.
Acknowledgements
We would like to acknowledge valuable inputs from the participants at the Biennial workshop in forest economics of the Laboratory of Forest Economics and the European Forest Institute 30 May to 1 June 2012 in Nancy, France. We would also like to thank Asbjørn Tingstad, Pierre-Alexandre Mahieu, Petter Gudding, Erik Framstad, Aina Johansen, Anne Stenger, and two anonymous referees for very useful comments.
Notes
1 A preliminary version of the paper was presented at the workshop ‘Forest Sector Modelling and Mulitfunctionality in Forests’ at INRA, Nancy, France, 31 May–1 June 2012.
1. Productive forest is forest with a production potential of at least 1 m3 per hectare and year, i.e. areas that are suitable for commercial forestry. Measuring conservation as a percentage of the productive forest area is a common way in many countries.
2. The so-called Aichi targets of the UN Convention on Biological Diversity specify 17% protection of non-marine nature types, which may be a goal for the even longer term.
3. The information available to consumers is rarely perfect, and hence it is important to consider the amount and type of information available to individuals when valuing public goods, especially complex goods such as biodiversity (e.g. information given in a questionnaire as discussed subsequently).
4. Note that Bergstrom and Taylor (Citation2006) point out that in the case of demand and/or supply uncertainty, terms capturing these effects will have to be included and the resulting WTP then measures option price.
5. Regarding the QUAL variable, it has been questioned in Norway to what extent voluntary conservation areas fill stated conservation gaps (Framstad and Blindheim Citation2010). However, this is a matter of setting strict enough biological requirements for new enrolments in the program, and is not our primary concern here.
6. There may be reasons to include also non-Norwegians, as some areas may be of at least Nordic or European significance. However, to be conservative, we limit the population to the domestic population.
7. Some forest areas, so-called zero-areas, are in principle protected by the current timber prices, as they are not worth harvesting. The compensation for enrolling these areas would be very low if not zero.
8. Note that transaction costs for conservation may be partly offset by some reduction in transaction costs related to existing support schemes for regular forestry activities. For simplicity, we do not include these here.
9. There are pros and cons of each approach and we do not reflect the full debate here. See Bateman et al. (Citation2002) for a practical discussion of both stated preference approaches.
10. Although it may be considered speculative to say something about the overall improvement in groups of species on the basis of the changes in forest reserves on a large geographical scale as considered here, decision-makers will in practice still need some expert opinion about such improvements to make an informed decision about which conservation plan to choose.
11. Experts from the Norwegian University of Life Sciences.
12. 1 NOK = 0.11 UK £ = 0.13 euro = 0.18 US $ (nominal exchange rates April 2013).
13. The age of 15 was set instead of the perhaps more common age of 18 as we think the preferences of the younger teenagers also should count, since many at that age already have part time or holiday jobs and must pay income tax. Note also that every respondent were asked for the WTP of their household (not individual WTP).
14. Note that we for simplicity treat the two WTP responses from each respondent as independent (i.e. no panel data considerations).
15. An alternative to this is to use an interval estimation approach (see e.g. Mahieu, Riera, and Giergiczny Citation2012) where WTP >0 is assumed to have a log normal distribution. We also tested this and found similar (but slightly higher) values compared to those reported in . See also Notaro and De Salvo (Citation2010) for a recent discussion on payment card estimation.
16. All other reasons than clear statements of no benefit from or no budget for conservation were interpreted as ‘protest zeros’.
17. In some instances being too specific on the ‘delivery date’ for the good may compromise the realism of the valuation scenario. We therefore instead made clear that the government currently is seriously considering different plans and that their decision and speedy implementation would depend on the results of this survey among other things.
18. The actual coordinates of the reserves are currently not public information.
19. A more realistic scenario would be for example to assume a plan that implements, e.g. 10% of the conservation increase each year. However, this complicates calculations, both for costs and benefits, and is not in line with what respondents were presented in the survey. We therefore, stick with this simplifying assumption.
20. It was made clear that the answer would not be used to estimate compensation for this forest in particular, to remove any worries about our intentions and potentially reduce strategic answering.
21. Amounts were chosen in consultation with forestry experts to cover the typical value range of rents from timber in a Norwegian forest. The list also included ‘don't know’.
22. A common area unit for forestry in Norway, corresponding to 0.5 hectares (ha).
23. Note that compared to the CV survey of the environmental benefits above, where the good people value is the same across respondents (except of course for subsamples getting different-sized conservation plans), the WTA stated in the forest owner survey may depend both on the different characteristics of the forest property as well as the preferences of the forest owner. That is the reason for collecting the extra information about the specific plot forest owners would like to enrol.
24. The cut-off point of 25 ha was chosen because, from an ecological perspective, smaller properties would not contribute significantly to conservation purposes.
25. There was some slight overrepresentation of higher educated, males in the age group 30–44, which is often observed in such surveys. Since public information about the income distribution is not easily available, we instead compared the income data for respondents and non-respondents in the Internet panel. A t-test revealed no significant difference in income levels between respondents and non-respondents. We decided not to weigh the data.
26. An alternative procedure for excluding protesters of requiring both statements to be related to protest yield the following intermediate mean WTP estimates: NOK 857 (797.4, 916.6) for 2.8%, NOK 1009.9 (862.4, 1157.3) for 4.5% and NOK 1080.6 (996.4, 1164.8).
27. We also conducted this test by comparing mean WTP when only WTP>0 respondents are included (and outliers excluded). We found that WTP for the 4.5% plan was greater than for the 2.8% plan (one-sided test, p < 0.05) and that the WTP for the 10% plan was greater than for the 2.8% plan (one-sided test and two-sided test, p < 0.01).
28. Two-sample t-test with unequal variances, showed no significant difference between means. This held both with and without zero responses included.
29. As a validity check we ran some simple regressions (results left out for brevity) for our specific samples explaining variation of WTP. These show that WTP vary with variables expected from theory and previous empirical research, such as income and frequency of recreational use. Lindhjem and Navrud (Citation2009, Citation2011a) also carry out validity checks of some of the subsamples from the same dataset and find responses reliable.
30. Here we choose the conservative approach to scale up by households since respondents are asked to state their WTP on behalf of their household. However, this approach is not without problems, and an individual WTP approach could also have been chosen. See Lindhjem and Navrud (Citation2009) for a discussion.
31. Coefficient of 0.775 on log productive forest area, t-value of 19.43.
32. There were some signs of non-response bias, as analysed by Nybakk et al. (Citation2009), in that respondents had significantly (p < 0.05) higher education and were significantly (p < 0.05) younger than non-respondents. None of these variables were significant in the analysis of Lindhjem and Mitani (Citation2012), so we decided for simplicity not to weigh the data.
33. Taking the mean of the two estimates in of Lindhjem and Mitani (Citation2012). The corresponding standard error of this estimate is 220.3.
34. An infinite horizon would increase the benefits compared to costs.
35. NOU (Citation2012) recommends a risk free real discount rate of 2% and states that an average project should have an adjustment for so-called systematic risk of 2%. It is a relevant discussion whether forest conservation should be seen as having elements of ‘insurance’, i.e. that risk adjustment should be downwards. However, WTP will also be higher with higher national income, indicating some systematic risk. The net effect would be uncertain, and we stick to the standard approach here.
36. The international debate on the choice of discount rate is extensive and a ‘second wave’ of research was initiated in the wake of the Stern Review from 2006, which recommended lower discount rates for climate mitigation than economists traditionally had used in CBA. Recommendations on discount rates currently vary between countries. The UK Green Book from 2003 (HM Treasury Citation2003), for example, differ from the Norwegian NOU (Citation2012). NOU (Citation2012) recommends 3% until year 75 and then 2% beyond, while HM Treasury (Citation2003) recommend several time horizons and lower rates.
37. Note that for certain combinations of discount rates and elasticities with respect to income, the benefit side may not converge.
38. The extent of hypothetical bias has also been found to vary between countries and cultures (Ehmke, Lusk, and List Citation2008). Since Norwegians are generally not known to prefer politeness over honesty and since respondent honesty tends to increase and yield lower WTP in Internet survey settings where interviewers are not present (Lindhjem and Navrud Citation2011b), we believe that the hypothetical bias factor in our case at least is not in the high end of the distribution.
39. They made, however, such a recommendation for the value of time and the value of statistical life (both elasticity = 1).