04pouta (ds)

04pouta (ds)

5/4/00 4:12 pm

Page 119

Contingent valuation of the Natura 2000 nature conservation programme in Finland E. POUTA1, M. REKOLA1, J. KUULUVAINEN1, O. TAHVONEN2 AND C.-Z. LI3 1

University of Helsinki, Department of Forest Economics, PO Box 24, FIN-00014 Helsinki, Finland Finnish Forest Research Institute, Helsinki, Finland 3 Department of Economics, University of Dalarma, Sweden 2

Summary This paper shows how contingent valuation studies can produce relevant information for public nature conservation decisions. The study analyses the preferences of Finnish households for a nature conservation programme, Natura 2000 Network, by applying a dichotomous choice referendum model of the contingent valuation survey. In order to study the influence of attitudes and beliefs on a choice between the status quo and the new conservation project, an attitude–behaviour framework is applied. Beliefs concerning the outcomes of the nature conservation policy and evaluations of their importance describe how attitudes towards the programme are formed. The choices in the referendum are explained using a logit regression model and are found to be a function of attitude and socio-economic variables. The probability of a person supporting the proposed conservation level depends significantly on the income, age and background (urban–rural) of the respondent. The estimated model of choice behaviour is used to calculate the average willingness to pay for the Natura 2000 Network, which is also compared to the costs of the conservation programme.

Introduction Reallocating forests from timber production to nature conservation has been a prominent trend in northern Europe and North America during recent decades. On many occasions, this has also caused heated public discussion (Hellström, 1999). A recent example of this kind of nature conservation debate was connected with the implementation of the European Union’s nature © Institute of Chartered Foresters, 2000

conservation programme, Natura 2000 Network, in Finland. The programme met strong negative public opinion in Finland for various reasons. It was claimed that the programme would severely restrict forestry, endanger the profitability of forest industries and weaken private landowners’ rights concerning their property. In addition, the discussion suggested that the Finnish public is satisfied with its present environment and protected areas should perhaps be reduced rather Forestry, Vol. 73, No. 2, 2000

04pouta (ds)

5/4/00 4:12 pm

120

Page 120

F O R E S T RY

than increased. However, the discussion had been dominated by various vested-interest groups, such as agricultural producers and environmentalists, and the preferences of the average Finnish household and the welfare effects of the programme had not been studied. This paper shows how contingent valuation (CV) can produce relevant and useful information for public decision-making and conflict-solving concerning nature protection programmes. Cost–benefit analysis (CBA) is the traditional method used in welfare economics for studying the effects of public policy. In applying CBA to environmental decisions, reliable tools are needed to measure non-market benefits in monetary terms. CV is a method where the total monetary value of the environmental benefits of a carefully defined project is measured using survey research methods. The method has gained wide international recognition and a large number of studies concerning a wide range of natural resources have been made in several countries. For example, forest protection has been studied by Kriström (1990), wildlife and fish conservation by Brookshire et al. (1983), and biodiversity by Spash and Hanley (1995). In the US, CBA and CV have been put into practice by various public agencies despite some controversy concerning the CV method (Mitchell and Carson, 1989; Arrow et al., 1993; Hausman, 1993; Portney, 1994; Bishop et al., 1995). Also, in many European countries the application of the method has stimulated public awareness and influenced decisions (Kuik et al., 1992). The CV literature has also applied theories from social sciences other than economics for the purpose of explaining respondents’ choice behaviour (Cummings et al., 1986; Peterson et al., 1988; Kerr and Cullen, 1995). In this study, the attitude–behaviour framework is used to explain contingent valuations. Ajzen and Fishbein (1975) have developed a widely used model of the attitude–behaviour relationship, which has been used to explain choices, e.g. voting or buying a commodity. The framework has been applied in some CV studies, mainly for theoretical and methodological purposes (Ajzen and Driver, 1992; Barro et al., 1996). Here the model is applied to produce empirical results relevant to the conservation decision-making point of view. The contribution of this study is, firstly, the

illustration of how CV and CBA can produce rich and applicable information for nature conservation decisions. Here, CV is applied to a real world decision-making situation, which makes it an interesting case compared with many other hypothetical CV studies. In order to understand the preference formation, important in decision making, we use an attitude–behaviour framework to produce information on attitudes and beliefs about the Natura 2000 Network. The study implements two tests. The first test focuses on the effect of the scope of conservation on the probability of supporting the conservation programme. The second test deals with the effect of the institutional context of valuation, i.e. the conservation planning method, on CV results. Thirdly, the study compares benefits with costs regarding the forestry part of the programme. The paper is organized as follows: in the next section, we form an economic model for conservation benefit valuation. We then describe the decision-making setting concerning Natura 2000 in Finland and clarify the way the contingent valuation method is implemented in this study. The results include attitudes to and beliefs about Natura 2000, the proportions of supporters and opponents of the programme, a dichotomous choice model and a comparison of benefits and costs. Finally, we conclude with a discussion of the results.

The economic model The economic model of benefit measurement considers the utility of a respondent before and after the conservation project. It includes conservation level and the conservation planning method as arguments of the utility function. The utility level can be described by an indirect random utility function (Hanemann, 1984): Vijk = V(wi, xj, zk, ci) + ijk

(1)

where Vijk is the utility level of individual i; wi is the income of individual i; xj is the amount of nature conservation (e.g. hectares) related to a project j; zk is the policy planning method k (Natura 2000 versus nature conservation planning); ci is a vector of variables describing the respondent (tastes and socio-economic variables); and ijk is a stochastic component.

04pouta (ds)

5/4/00 4:12 pm

Page 121

VA L U AT I O N O F N AT U R A 2 0 0 0 P R O G R A M M E I N F I N L A N D

The welfare measure CS (compensating surplus) can be defined as an amount of money which can be subtracted from (added to) an individual’s income after environmental change without changing his/her utility level from the level which existed before environmental change: V(wi, x0, z0, ci) + i00 = V(wi – CS, xj, zk, ci) + ijk

(2)

where x0 = status quo of nature conservation; z0 = no policy planning method concerning nature conservation is taking place (status quo); CS = compensating surplus (willingness-to-pay). In the referendum question, an individual faces an offer to pay a given sum of money (BID) to gain better quality of environment. The probability of accepting the proposed project instead of the status quo can be written as follows: Pr(project) = Pr[V(wi – Bid, xj, zk, ci) + ijk > V(wi, x0, z0, ci) + i00]

(3)

where V(.) is the observable component of utility. If the cumulative distribution of error terms is logistic, the logit model can be used for the estimation. The probability of choosing the project can be written as follows: 1 F(∆v) = Pr(No) = –––––– 1 + ∆v

(4)

where ∆v = v1 – v0 is the change in welfare and F(∆v) is the cumulative distribution function of standard logistic variate = 0 – 1.

Valuation context and survey implementation Nature conservation programme In 1992 The European Union (EU) decided to coordinate its nature protection policy within the member countries by introducing the Natura 2000 Network of nature conservation areas in Europe. The aim of the programme is to protect natural habitats of wild fauna and flora in order to guarantee their ‘favourable protection level’. According to the programme, specific sites are chosen using purely biological criteria. When Finland joined the EU in 1995, the EU directives

121

concerning the Natura 2000 Network became legally binding in Finland. The deadline for national proposals to be sent to the EU Commission was June 1995, and Finland did not ask for this schedule to be postponed. This explains the urgency of Natura planning during 1996 and 1997 by the Finnish environmental administration bodies. The first proposal for the conservation sites was finally published in April 1997. In the proposal, 95 per cent of sites consisted of existing nature conservation areas. However, the proposal also included 160 000 ha of newly protected land and 524 000 ha of newly protected water areas. Most of the newly protected area was under private ownership. In some parts of these areas all economic development, like forestry, was totally prohibited and in other parts these activities were restricted so that they did not endanger the specified nature values. The response towards the proposal was extremely negative; The Ministry of Environment received almost 15 000 submissions. In April 1998, the Ministry published a revised proposal that was 263 000 ha smaller than the original one. The decision-maker, the Finnish Government, realized the difficulties of the planning process. To increase understanding of the problem and to assist the decision-making, an environmental impact assessment of the Natura 2000 proposal was conducted in autumn 1997. The environmental impact assessment concentrated on the biological impacts of conservation, socio-political impacts and economic costs of the programme (Hildén et al., 1998). The contingent valuation study was executed as a separate part of the environmental impact assessment. It was financed by the Ministry of Environment, and conducted by the Finnish Forest Research Institute and the Department of Forest Economics, University of Helsinki. CV scenario A CV questionnaire includes a scenario describing the project which changes the level of a particular public benefit and the context of paying for it. The scenario should include the following items: (1) a detailed description of the benefit, or the commodity; (2) the baseline of the provision; (3) the circumstances under which the benefit is

04pouta (ds)

5/4/00 4:12 pm

Page 122

122

F O R E S T RY

provided; and (4) the method of payment (Fischhoff and Furby, 1988; Mitchell and Carson, 1989). Respondents have to be aware of all these items in order to be able to make a realistic evaluation of the project. After the description of the scenario a willingness-to-pay (WTP) question is asked. In this study, the first item, the benefit, was the actual extent of the conservation area described along with the percentage of conserved land area of Finland and the types of habitats in the Natura 2000 programme. The extent of the new conservation areas was one of the main issues in public discussion and was therefore also emphasised in the CV. In addition to studying the proposed magnitude of conservation areas, two other alternatives were included: one with a size between the existing amount and the proposed amount and the other with an even larger size than that originally proposed by the environmental administration. Thus the second item, the change to the baseline of the provision, was described as percentage change in land area, a 3, 6 or 9 per cent increase in the conservation level (Table 1). The third item, the circumstances under which the benefit is provided, describes the institutional structure that takes place with the provision. As stated above, the planning of Natura 2000 has been highly controversial in public discussion, which would be a potential problem with the implementation of the CV. Respondents’ opinions about Natura 2000 do not reflect their opinions about nature conservation in general. In order to study this problem, we

formulated contingent valuation questionnaires in two different contexts: in the first, we associated conservation directly with the Natura 2000 programme, and in the second, the project was introduced within the context of a general question about nature conservation planning. The policy planning method in the latter context was a revision of a national nature conservation plan without reference to the Natura 2000 programme. This may seem somewhat hypothetical; however, it is based on an existing plan of the Ministry of Environment according to which all nature conservation areas in Finland will be investigated and their contribution to biodiversity evaluated. The fourth item is the mechanism through which the project is financed. In the case of nature conservation in Finland, payment is made through taxes. The increase in taxes was wellfounded because of the requirement of compensating private landowners in Finnish nature conservation legislation. Willingness-to-pay measurement The WTP question in the questionnaire was in the dichotomous-choice format and it was illustrated by means of a table with two alternatives (Table 1). The first alternative is the status quo of conservation. The second one is either a 3, 6 or 9 per cent increase in the conservation level. The table also presents hypothetical expenses (BID) of a household which are connected to the increased conservation alternative. Expenses were varied within each conservation level and between

Table 1: Referendum setting in the case of a 3 per cent increase in nature conservation area Option 1

Option 2

Nature conservation area Change to current area

Same as current area

3% larger than current area

% of the land area of Finland

11.3%

11.6%

Conserved nature types

Currently conserved: swamps, shores, bird habitats, eskers, wilderness areas, old growth forests, groves

In addition to option 1: rich fens, springs, lakes, rivers, river deltas

Change in the income tax of your household

No change

280 FIM increase in income tax

04pouta (ds)

5/4/00 4:12 pm

Page 123


them1. Each respondent was asked to make a choice between the status quo and an alternative. Attitudes and beliefs In its simplest form, the attitude–behaviour framework (Ajzen and Fishbein, 1975) can be seen as a series of hypotheses linking beliefs to attitudes, attitudes to intention, and intention to behaviour. In the context of CV, the dichotomous-choice behaviour can be seen as an intention explained by attitudes and further beliefs. According to the attitude–behaviour framework, attitudes are a function of salient beliefs about the attitude-object. Each salient belief links the object to an attribute or outcome of the behaviour in question. The attitude (Att) is determined by the strength of these beliefs (bi) and evaluations (ei) associated with the attributes concerning behaviour; in mathematical form it is Att = ∑(bi,ei). Here beliefs concern outcomes or attributes of nature conservation policy and evaluations are formed according to their importance. Attitudes toward the proposed conservation project were first measured. Depending on the policy planning method, respondents’ attitudes were indicated by the nature of their responses to statements such as: ‘The Natura 2000 nature conservation programme is generally speaking . . .’ or ‘Increasing the amount of nature conservation areas is generally speaking . . .’. Three scales (necessary–unnecessary, supported–objected, positive–negative) were used to build a sum variable. Respondents were classified into four attitude groups by dividing the range of the sum variable (POLICYAT) into four equal categories (Table 2). Belief measurement considered those aspects of the conservation programme that were most under discussion in the media: landowner rights, conservation of species and biotopes, national economic costs, requirements of the EU and informing the public. To determine the belief strength (bi) of the six statements concerning the outcome of conservation programmes, respondents were asked 1

123

whether they agreed or disagreed (fully, somewhat, or neither agree or disagree) using a 5-point scale (–2 to +2). A belief evaluation (ei) was obtained by asking respondents whether they agreed or disagreed with the importance of the outcome, which was coded along a scale of 1–5. Sample The sample used for the mailed questionnaire consisted of 2400 Finns aged from 18 to 70. The random sample was drawn from the census of Finland. After the first mailing, reminder postcards were sent to the respondents.2 This produced a response rate of 46 per cent, i.e. 1085 completed forms.

Results Attitudes to and beliefs about nature conservation From the standpoint of managing conflicts concerning the project, combining an attitude–behaviour framework with a valuation study furnishes material for understanding the arguments behind conservation preferences. Attitudes toward the proposed nature conservation policy (POLICYAT) were quite positive, with 66 per cent of respondents belonging to the positive attitude groups (Table 2). Table 3 illustrates how attitude can be equated with respondents’ beliefs about the outcomes of the project and evaluation of the importance of these outcomes. One variable which significantly affected general attitudes toward the programme with a high negative coefficient was respondents’ perception of the ability of the programme to take into account landowners’ rights. On the other hand, the respondents’ perceptions about the ability of the programme to conserve animal and plant species had a positive effect on their attitudes. The respondents’ opinions about the

BIDS (FIM) in different samples were (3 per cent): 60, 170, 280, 400, (6 per cent): 120, 340, 560, 800, (9 per cent): 180, 510, 840, 1200. BID design was based on a pilot study. BIDS were designed to increase linearly as the area to be conserved increases. This is due to another subsample in the same study that utilizes a multiple-choice question format where the respondent can see all projects and their BID levels at the same time. 2 No reminder survey was sent due to the time limitations.

04pouta (ds)

5/4/00 4:12 pm

Page 124

124

F O R E S T RY

Table 2: Nature conservation attitudes (POLICYAT). Share of respondents in four attitude groups. n = 990*

Attitude toward proposed nature conservation policy (POLICYAT)

Very positive attitude

Positive attitude

Negative attitude

Very negative attitude

18.4%

47.6%

23.0%

11.0%

* Due to item non-responses, the number of observations is decreased from 1085 to 990.

Table 3: Explaining attitude toward Natura 2000 with beliefs about the outcomes of the programme and their evaluation. Linear regression, regression coefficient (B) and significance level (T) B Beliefs about the outcomes of the programme The programme takes land owner rights well into account The programme conserves animal and plant species well The programme conserves biotopes well The programme does not cause considerable costs to the national economy By implementing the programme, Finland follows EU regulations well Informing the public is taken into account well in planning the programme Evaluation of the outcomes of the programme Importance of land owner rights Importance of conservation of animal and plant species Importance of conservation of biotopes Importance of cost to national economy Importance of following EU regulations Importance of informing the public Constant R2 n*

T

0.2565 0.1141 0.0681 0.0767 0.0100 0.0082

0.0000 0.0018 0.0663 0.0042 0.7404 0.7554

–0.1124 0.3326 0.2549 –0.0638 0.1464 –0.0292 –4.4989 0.529 863

0.0013 0.0000 0.0000 0.0663 0.0000 0.4608 0.0000

* Due to item non-responses the number of observations is decreased from 1085 to 863.

importance of species and biotope conservation also had high and significant positive coefficients in explaining their attitude formation. Choices between the conservation project and the status quo Figure 1 shows the dependency within a dichotomous choice between the probability of supporting the conservation programmes and their expense. When a 3 per cent increase in the status quo conservation level was proposed, the majority of respondents (53.3 per cent) supported the project. However, when the proposed increase was 6 or 9 per cent, the majority of the respondents supported the status quo conservation level.

The choice between the status quo and the conservation programme are explained using the logit model. Table 4 presents the empirical results of this model. The first explanatory variable is change in income tax per household (BID). The sign of the (BID) variable is negative as expected, higher (BID)s reduce the probability of choosing the proposed conservation project. Three conservation levels are coded using two 0/1 dummy variables (PROJECT). However, the probability of supporting conservation was not dependent on the level of conservation. The policy variable (POLICY) describes the Natura 2000 nature conservation planning method (0) and revision of national nature conservation plan without connection to Natura 2000 (1). The policy frame had

04pouta (ds)

5/4/00 4:12 pm

Page 125


125

Table 5: WTP and its aggregation, 11.6 per cent conservation area of the land area WTP/household Median* 196 FIM Mean† 600 FIM Number of households 2 290 100 Aggregated WTP Using median* 449 million FIM Using mean† 1 374 million FIM Costs (Hildén et al., 1998) 241–796 million FIM

Figure 1. The probabilities of supporting three different alternatives (increasing the nature conservation area corresponding to a 3(), 6(), or 9() per cent increase in expenditure in the respondent’s household). Table 4: Logit model explaining the willingness-topay, regression coefficient (B) and significance levels (T) Variable BID PROJECT† PROJECT(1) PROJECT(2) POLICY(1) INCOME AGE LIVING(1) POLICYAT Constant

B

T

–0.0014

***

–0.2140 –0.0165 0.4363 4.02E-06 –0.0367 –0.4927 1.1838 3.8874

** *** *** *** *** ***

Significance of Wald test statistic: ***P < 0.005, **P < 0.01, *P < 0.05; 76.81 per cent predicted correct; –2Log Likelihood (constant only) 1262.303; –2Log Likelihood (model) 873.339; n = 914. † Coding of the PROJECT variable: 3 per cent increase in nature conservation level PROJECT(1) = 0 and PROJECT(2) = 0; 6 per cent increase in nature conservation level PROJECT(1) = 1 and PROJECT(2) = 0; 9 per cent increase in nature conservation level PROJECT(1) = 0 and PROJECT(2) = 1. 3

* Assumption: WTP ]– ∞, ∞[. † Assumption: WTP [0, ∞[.

an effect on the willingness-to-pay. Respondents were more willing to pay for the programme if it was offered as a general nature conservation programme instead of Natura 2000. Household income (INCOME) in Finnish marks (FIM) is included in the model. Other socio-economic variables are the respondent’s age (AGE), living environment (LIVING), coded urban (0) and rural (1),3 and attitude toward the proposed policy (POLICYAT). Of the socio-economic variables, a higher income, lower age and urban living environment positively affected the probability of choosing the conservation project.4 Attitudes towards the programme have a high correlation coefficient and significantly influence the model. Willingness-to-pay and cost estimates The estimation results from the logit model are used to calculate the mean and median willingness-to-pay (Table 5). If WTP is assumed to be positive, the following equation can be used to calculate the mean WTP (Hanemann, 1989): mean(WTP) = (–1/b1) · ln(1 + exp(b

0

+ b2

MEAN(PROJECT(1)) + . . . + b MEAN(POLICYAT)))

(5)

where b1 is the value of the coefficient of the (BID), and b0 is the constant term, and b2, . . ., b8 are the other variables in the model.

The living environment was originally measured in five scales: (1) City centre, (2) City suburban, (3) Municipality centre, (4) Municipality village, (5) Rural area belonging to city or municipality. Options (1) and (2) were coded as urban (0), and options (3), (4), and (5) as rural (1). 4 To study the self selection bias between policy treatments an ANOVA was used. Age was the only variable which had statistically different means between samples so that respondents in Natura treatment had a mean age of 43.6 years and in the other treatment mean age was 41.1.

04pouta (ds)

5/4/00 4:12 pm

126

Page 126

F O R E S T RY

If the negative values of the WTP are allowed, the median is a proper description of average WTP. That is: median(WTP) = –(b0 + b2 MEAN (6) (PROJECT(1)) + . . . + b8 MEAN (POLICY))/b1. Means and medians of the WTP for 3 per cent increase level were FIM 600 and 196, for the 6 per cent level FIM 517 and 44, and for the 9 per cent increase level FIM 593 and 185, respectively. Differences between mean WTPs are not statistically significant.5 Because WTPs for different levels of the protection are in the same range it seems that respondents have not necessarily been sensitive to the scope of the conservation. This further refers to the possibility of the so called embedding problem (Kahneman and Knetsch, 1992; Hanemann, 1994). Problems with using means as welfare measures in aggregation have been widely discussed (Hanemann, 1984; Carson, 1991; Ready and Hu, 1995). Use of the median is often recommended, because it is not very sensitive to the unobserved tails of density function. Because costs have been estimated only for the revised Natura proposal, WTP aggregation is limited here to the 3 per cent increase level (Table 5). This level was also closest to the project under political discussion and decision making. Here both the mean and the median are used in the aggregation. The population in aggregation is 2.29 million Finnish households. The aggregated WTP using the median is 449 million FIM, and using the mean 1374 million FIM. It is not a purpose of this study to make a detailed cost–benefit analysis. However, we feel that it is interesting to briefly describe the cost estimates available and compare them to the benefits estimated in the CV study. The cost estimation was limited to the costs of forestry on private and state-owned land. It was, however, likely that they would cover most of the total costs (Hildén et al., 1998). Forestry costs are

calculated using two methods. First, the lower estimate, 241 million FIM, assumes that the forests conserved have the same timber volume per hectare as the average stands in the forest district. Second, in the upper estimate, 796 million FIM, the conserved forests are assumed to have timber volume equal to the mature stands. Some general assumptions behind the cost estimation can be mentioned. For example, the stand inventories have average timber volumes for the district, the rate of return is 4 per cent, and timber prices are fixed. When comparing costs and benefits, the conservative estimate of benefits, based on the median, exceeds the lower estimate of the costs of the project but was under the upper estimate. The benefit estimate based on the mean was almost double compared to the upper cost estimate.

Conclusions According to the results of this study, the probability of supporting the increase in nature conservation was dependent on its expense to households. Most of the respondents supported a 3 per cent increase in the conservation level. The probability of choosing the increased conservation was dependent on the age, income and living environment of the respondent, so that conservation was supported more by the young, high-income and urban population. However, the proposed increase in the size of conservation area had no statistically significant effect on the support. The primary belief statements against the programme were connected to landowner rights and costs to the national economy. On the other hand, the main statements to explain positive attitudes towards the project were connected with the importance of the flora and fauna, and biotope conservation. Aggregated benefits were calculated for the project at a 3 per cent increase in conservation

5 To analyse sample non-response bias the income and age of respondents were compared to the Finnish population. Non-response bias was corrected using population values for age and income in valuation function (Mitchell and Carson 1989, p. 273). The WTPs for a 3 per cent increase level were FIM 629 and 246, for the 6 per cent level FIM 543 and 93, and for a 9 per cent increase level FIM 622 and 234, respectively. According to this analysis it seems that sample non-response bias was not a serious problem in this study.

04pouta (ds)

5/4/00 4:12 pm

Page 127


level. The cost estimates available on this level were based on the revised plan by the Finnish environmental administration. The benefits exceeded the project’s cost to the forestry industry even when the most conservative estimates of willingness-to-pay were used. However, it should be kept in mind that both cost and benefit estimates rely on several critical assumptions. This study was able to reveal the preferences of the general public in the form of monetary benefit measures that can be used in the cost-benefit analysis. The survey also measured other issues that are needed for an informative CBA, like the beliefs and attitudes of the public. After the implementation of the study, the Finnish government passed the Natura 2000 Network policy in Finland in August 1998. The main results of the CV study were also included in the resolution of the government. It can be concluded that the survey, despite its limitations, gives decision makers a wider perspective of citizen preferences than do the viewpoints of the interest groups, which dominated the discussions in the media.

Acknowledgements We are grateful for comments by three referees and to the Ministry of Environment and Academy of Finland for financial support.

References Ajzen, I. and Driver, B.L. 1992 Contingent value measurement: On the nature and meaning of willingness-to-pay. J. Consum. Psychol. 1(4), 297–316. Ajzen, I. and Fishbein, M. 1975 Belief, Attitude, Intention, and Behavior: An Introduction to Theory and Research. Reading, MA, Addison-Wesley. Arrow, K., Solow, R., Leamer, E., Portney, P., Randner, R. and Schuman, H. 1993 Natural resource damage assessment under the Oil Pollution Act of 1990. National Oceanic and Atmospheric Administration panel. Federal Register, vol. 58, No 10. Barro, S.C., Manfredo., M.J., Brown, T.C. and Peterson, G.L. 1996 Examination of the predictive validity of CVM using attitude–behavior framework. Soc. Nat. Res. 9, 111–124. Bishop, R.C., Champ, P.A. and Mullarkey, D.J. 1995 Contingent valuation. In The Handbook of Environmental Economics. D. Bromley (ed.). Blackwell, Oxford, UK.

127

Brookshire, D.S., Eubanks, L.S. and Randall, A. 1983 Estimating option prices and existence values for wildlife resources. Land Econ. 59(1), 1–15. Carson, R.T. 1991 Constructed markets. In Measuring the Demand for Environmental Quality. J.B. Braden and C.D. Kolstad (eds). Elsevier Science Publishers BV, Amsterdam. Cummings, R.G., Brookshire, D.S. and Schulze, W.D. 1986 Valuing Environmental Goods: An Assessment of the Contingent Valuation Method. Rowman and Allenheld, Totawa, NJ. Fischhoff, B. and Furby, L. 1988 Measuring values: a conceptual framework for interpreting transactions with special reference to contingent valuation of visibility. J. Risk Uncertainty 1, 147–184. Hanemann, W.M. 1984 Welfare evaluations in contingent valuation experiments with discrete responses. Am. J. Agric. Econ. 66, 332–341. Hanemann, W.M. 1989 Welfare evaluations in contingent valuation experiments with discrete responses data: reply. Am. J. Agric. Econ. 71, 1057–1061. Hanemann, W.M. 1994 Valuing the environment through contingent valuation. J. Econ. Perspect. 8(4), 19–44. Hausman, J.A. 1993 (ed.) Contingent Valuation: A Critical Assessment. North Holland Press, Amsterdam. Hellström, E. 1999 Environmental forestry conflicts in Europe and the United States. In World Forests, Society and Environment. World Forests, Volume 1. M. Palo and J. Uusivuori (eds). Kluwer Academic Publishers, 276–284. Hildén, M., Tahvonen, O., Valsta, L., Niininen, I., Ostamo, E., Leppänen, J. and Herkiä, L. 1998 Natura 2000 verkoston arviointi. Ministry of Environment, Helsinki. Kahneman, D. and Knetsch, J. 1992 Valuing public goods: the purchase of moral satisfaction. J. Environ. Econ. Manage. 22, 57–70. Kerr, G.N. and Cullen, R. 1995 Public preferences and efficient allocation of a possum-control budget. J. Environ. Manage. 43, 1–15. Kriström, B. 1990 Valuing Environmental Benefits using the Contingent Valuation Method. Umeå Economic Studies, 219. Umeå Sweden, 169pp. Kuik, O., Navrud, S. and Pearce, D.W. 1992 Benefit estimation and environmental decision making. In Pricing the European Environment. S. Navrud (ed.). Scandinavian University Press. Mitchell, R.C. and Carson, R.T. 1989 Using Surveys to Value Public Goods: the Contingent Valuation Method. Resources for the Future. The Johns Hopkins University Press, Washington DC. Peterson, G.T., Driver, B.L. and Gregory, R. (eds) 1988 Amenity Resource Valuation: Integrating Economics with Other Disciplines. Venture Publishing Inc., State College, PA.

04pouta (ds)

5/4/00 4:12 pm

128

Page 128

F O R E S T RY

Portney, P.R. 1994 The contingent valuation debate: why economists should care. J. Econ. Perspect. 8(4), 3–17. Ready, C.R. and Hu, D. 1995 Statistical approaches to the fat tail problem for dichotomous choice contingent valuation. Land Econ. 71(4), 491–499.

Spash, C.L. and Hanley, N. 1995 Methodological and ideological options. Preferences, information and biodiversity preservation. Ecol. Econ. 12(1995), 191–208.