ARTICLE IN PRESS

Journal of Forest Economics 12 (2006) 75–81

1104-6899/$ -

doi:10.1016/j

�CorrespoE-mail ad

www.elsevier.de/jfe

A note on benefits and costs of adjusting forestry to

meet recreational demands

Goran Bostedta,�, Leif Mattssonb

aDepartment of Forest Economics, Swedish University of Agricultural Sciences,

S-901 83 Ume (a, SwedenbSouthern Swedish Forest Research Centre, Swedish University of Agricultural Sciences,

P.O. Box 49, S-230 53 Alnarp, Sweden

Received 18 May 2005; accepted 16 December 2005

Abstract

Economic studies concerning environmental functions of forests are often partial in the

sense that they focus on either benefits or costs. In other words, benefit/cost analyses

indicating whether it is economically motivated to change forestry to make it compatible with

environmental demands are relatively rare. This benefit/cost analysis deals with the forest as

recreation environment, where benefit estimates from a study conducted in the county of

Vasterbotten, Sweden, are compared with cost estimates from another study in the same

county. It is shown that adjustments of forest management to meet recreational demands do

largely affect both benefits and costs, and that the results are sensitive to how soon the effects

on the recreation environment occur after the adjustments of forest management.

r 2006 Elsevier GmbH. All rights reserved.

JEL classification: Q230

Keywords: Benefit/cost analysis; Contingent valuation; Forest management; Recreation

values; Silvicultural systems

see front matter r 2006 Elsevier GmbH. All rights reserved.

.jfe.2005.12.002

nding author.

dress: goran.bostedt@sekon.slu.se (G. Bostedt).

ARTICLE IN PRESS

G. Bostedt, L. Mattsson / Journal of Forest Economics 12 (2006) 75–8176

Introduction

One of the scientific pioneers in Sweden concerning the economics of timberproduction was Holmertz (1873), while Lindgren (1976) was one of the very first inthe country to deal scientifically with the economics of forest environmental concern,primarily by analyzing costs to forestry of setting aside specific recreation areas. Thetraditional forest economic research in Sweden is thus about 100 years older than theeconomic research on environmental functions of the forest. This is an importantreason why there is still a lot to be done on the scientific knowledge about the latter.Not surprisingly when considering the youth of forest environmental economics, ithas to a large extent been characterized by partial analyses of either benefits or costs,each of which with no ambitions to really make a complete benefit/costs-analysis.

One interesting exception was the benefit/cost analysis carried out by Bojo (1985).The background to his analysis was the question of whether a forest area close to themountains in the county of Jamtland in northern Sweden should be set aside as strictnature reserve or whether industrial forestry should be allowed in the area. Using theTravel Cost Method (Clawson and Knetsch, 1966) to estimate benefits and forestdata from the area to estimate costs, the results showed that the first mentionedalternative was superior to the latter. This was primarily because of the area’srelatively low value to forestry regarding volume of standing timber as well aspotential timber production.

Like the study by Bojo, most (of the few) scientific and empirically based benefit/cost analyses conducted deal with the problem of choosing between forestry with noenvironmental concern, or no forestry at all, i.e. strict nature reserve or nationalpark. Consequently, there is a lack of knowledge about the benefit/cost relationshipfor environmental concern in forestry on the main part of the Swedish forest area,where – according to the current Forestry Act – production of timber is to becombined with protection of environmental values on the same land. In other words,more knowledge is needed about the benefit/cost relationship when forestry ischanged (but not prohibited) to meet different environmental demands.

The aim of this research note is to shed some light on the economic effects ofmaking forestry more compatible with demands from forest recreation, when thetwo forms of land use occur on the same land. In doing this, we merge results fromempirical studies on benefits and costs, respectively, concerning the county ofVasterbotten in northern Sweden. The following section describes briefly theunderlying studies. In the subsequent section we present a benefit/cost analysis andresults. The research note ends with conclusions and final remarks.

Underlying studies

In Sweden, the Right of Common Access makes forests for recreation a non-market priced environmental good. To estimate the recreation value in economicterms of the forests in the Vasterbotten county, Mattsson and Li (1993, 1994a)

ARTICLE IN PRESS

G. Bostedt, L. Mattsson / Journal of Forest Economics 12 (2006) 75–81 77

conducted a survey study of 2000 citizens in the county using the ContingentValuation Method (CVM), where both a continuous and a discrete question formatwere applied (Mitchell and Carson, 1989; Garrod and Willis, 1999). The survey wasdesigned so that it focused on local forest recreation for the citizens (i.e. not foresttourism), and so that the recreation value assessed referred to on-site use value(hiking, camping, berry-picking, etc. ‘‘inside’’ the forest) as well as off-site experience(enjoyment of the forest landscape scenery from ‘‘outside’’ the forest). According tothis study, the recreation value amounted to about half of the timber productionvalue. As an integrated part of this research Mattsson and Li (1994b) made anotherCVM survey, using the same two question formats, focused on preferences andvaluations held by the public in the county regarding forest landscapes produced bydifferent silvicultural systems. This survey covered a random sample of 800 citizensin the county. It was shown that the recreation value in economic terms of the forests(including the value components mentioned above) varies a lot depending on whatkind of forest landscape there is, i.e. what silvicultural system that is practiced.Holgen et al. (2000) made a follow up analysis based on the same data used byMattsson and Li (1994b), also dealing with the effects of silvicultural systems on therecreation value of the forests.

On the cost-side, Holgen and Lind (1995) made a study focused on the question ofwhat happens to timber production and the economy of forestry if it is made morecompatible with the recreational demands. They combined results on forestrecreational demands, as expressed by the preferences held by the public in theVasterbotten county (Mattsson and Li, 1994b), and forest data from the NationalForest Inventory (Ranneby et al., 1987). By using this in the Hugin-system(Bengtsson, 1981; Wilhelmsson, 1989), a comparative analysis was made of twoforestry programmes for the county, called the ‘‘reference programme’’ and the‘‘alternative programme’’. In the ‘‘reference programme’’ forestry remains the sameas during the 1990s, while in the ‘‘alternative programme’’ forestry is brought in linewith the recreational demands – as far as this can be done with regard to siteconditions, etc. required for different silvicultural systems. The results showed, notsurprisingly, that seen from forestry’s point of view the ‘‘alternative programme’’ isinferior to the ‘‘reference programme’’, but also that this inferiority of the‘‘alternative programme’’ is smaller when the two programmes are comparedregarding the profitability of the forestry than with regard to volume of timber thatcan be harvested.

Benefit/cost analysis and results

The knowledge on preferences and valuations in economic terms held by thepublic regarding forest landscapes produced by different silvicultural systems(Mattsson and Li, 1994b; Holgen et al., 2000), and on the economic outcome forforestry given different silvicultural systems (Holgen and Lind, 1995), make itpossible to approach the answer to the central question: Is it economically

ARTICLE IN PRESS

G. Bostedt, L. Mattsson / Journal of Forest Economics 12 (2006) 75–8178

worthwhile, in the county of Vasterbotten, to adjust forest management with respectto the areal distribution of different silvicultural systems in order to meet therecreational demands from the public? Or in other words: Should the forestmanagement regarding to silvicultural systems continue to be practiced in line withthe ‘‘reference programme’’ or should it be transformed according to the ‘‘alternativeprogramme’’, seen from welfare economic point of view?

The areal distribution of the four main silvicultural systems in the county ofVasterbotten is approximately the following:

�

25% Natural regeneration using seed trees � 2% Single-tree selection � 70% Artificial regeneration after clear-cutting � 3% Shelterwood system

This distribution of silvicultural systems would continue to be about the samegiven the ‘‘reference programme’’.

The ‘‘alternative programme’’, i.e. the one where forest management is brought inline as much as possible with the recreational demands, involves a change to thefollowing areal distribution of silvicultural systems:

�

31% Natural regeneration using seed trees � 27% Single-tree selection � 8% Artificial regeneration after clear-cutting � 34% Shelterwood system

Such a transformation from the ‘‘reference programme’’ to the ‘‘alternativeprogramme’’ would result in an additional cost to forestry amounting to SEK 2280million, calculated as the loss in present value over a 100-year period using adiscount rate of 3%.

What is then the additional recreational value of such a transformation? Therecreation values of the forest landscapes that the four different silvicultural systemsproduce are:

�

Natural regeneration using seed trees: SEK 2801 per citizen and year � Single-tree selection: SEK 1251 per citizen and year � Artificial regeneration after clear-cutting: SEK 750 per citizen and year � Shelterwood system: SEK 3700 per citizen and year

Using these recreation values and the areal distribution of silvicultural systemsaccording to the ‘‘reference programme’’ as a weight (i.e. 0.25�2801+0.02�1251+?, etc.), gives a weighted recreation value of SEK 1361 per citizen and year.If instead the distribution according to the ‘‘alternative programme’’ is used asweight (i.e. 0.31�2801+0.27�1251+?, etc.), a weighted recreation value ofSEK 2524 per citizen and year is obtained. The difference, SEK 1163 per citizen and

ARTICLE IN PRESS

G. Bostedt, L. Mattsson / Journal of Forest Economics 12 (2006) 75–81 79

year, can be seen as a measure of the additional recreation value of conducting thetransformation from the ‘‘reference programme’’ to the ‘‘alternative programme’’.

To reach an estimate of the additional benefit that is comparable with the estimateof the additional cost, the figure SEK 1163 is multiplied with the 179,000 citizens(between 17 and 74 years of age) in the county of Vasterbotten, which gives a totalbenefit increase of SEK 208 million per year. The present value of this figure, over a100-year period using a discount rate of 3%, is SEK 6579 million. The conclusion isthat the benefit increase of conducting the transformation from the ‘‘referenceprogramme’’ to the ‘‘alternative programme’’ is almost 3 times larger than the costincrease of doing so.

The high present value of the benefit increase is crucially dependent on theassumption that the transformation from the ‘‘reference programme’’ to the‘‘alternative programme’’ has an immediate positive effect on the recreation value.In reality, a transformation to the ‘‘alternative programme’’ would have a smalleffect initially, and then gradually increase the recreation value. If we assume,somewhat simplistically, that the transformation does not generate any effect on therecreation value for the first 30 years, and that the transformation has full effectthereafter, the results of the benefit/cost calculation becomes very different. Thepresent value of the benefits increase is now SEK 2498 million, i.e. only 1.1 timeslarger than the cost increase. Suddenly, the argument for making the transformationfrom the ‘‘reference programme’’ to the ‘‘alternative programme’’ becomes muchweaker.

Conclusions and final remarks

In Sweden, forestry and forest recreation are activities that can share the sameland relatively well (with less conflicts than between forestry and biodiversityconcern), which is indicated by results from several studies (e.g. Lindhagen andHornsten, 2000). Nevertheless, forestry can be adjusted to increase the value offorests as recreation environment. The results of this benefit/cost analysis for thecounty of Vasterbotten indicate that if the forestry is transformed from what it wasduring the 1990s to largely consider recreational demands, and if the positive effectson the recreation environment of such a transformation do not take too long time tofully occur (no longer than approximately 30 years), then such a transformation isprofitable seen from a welfare economic point of view. The shorter time (than 30years) it takes, the more profitable the transformation would be – and vice versa. Inother words, the results are sensitive to the perspective of time.

Another time-dependent aspect is that there is no absolute scientific knowledge,especially in the long run, about the outcome of different silvicultural systems interms of timber volume and quality. It is fair to say, that the dominating focus of theSwedish research in the field has been on artificial regeneration after clear-cuttingand natural regeneration using seed trees, while research on single-tree selection andshelterwood system is more sparse (e.g. Ekelund and Hamilton, 2001). This means

ARTICLE IN PRESS

G. Bostedt, L. Mattsson / Journal of Forest Economics 12 (2006) 75–8180

that forecasts of timber volume and quality given alternative areal extents ofdifferent silvicultural systems – and thus costs in the benefit/cost analysis – areapproximations.

This is also true for the benefits, i.e. estimates of the recreation values. Therecreation values used in the benefit/cost analysis originate, as mentioned earlier,from CVM surveys where both continuous and discrete question formats wereapplied. In recent decades the CVM has been the dominating scientific method forassessing the economic value of non-market priced environmental goods (Carsonet al., 1996), and there are continuous efforts to develop the method and itsapplications (e.g. Boman et al., 2005). In this context, the main advantage of theCVM compared to other methods is that it allows estimation of the recreation valuegiven a broad spectrum of alternative forest characteristics. Nevertheless, assess-ments of the economic value of non-market priced environmental goods, like forestsas recreation environments, are always approximations.

Furthermore, preferences held by the public should not be regarded asunchangeable over time. Concerning market priced goods, new products or brandsof products are constantly introduced on the market, while other disappear. A basicreason to this is changes in consumers’ preferences, under the influence ofcomplicated and often worldwide market mechanisms. However, when it comes towhat kind of forest environments people in Sweden want for their recreation it seemslike the preferences are more stable over time. For example, Lindhagen andHornsten’s (2000) conclusion from a comparative analysis of public preferencesregarding different forest recreation environments in 1977 and 1997 was that‘‘changes are relatively small’’.

Important to consider is also that the forest recreation is not equally spread allover the 3.2 million hectares of forest land that is in the county of Vasterbotten(Mattsson and Li, 1993; Anonymous, 2004a). This, combined with the fact that thepeople in the sparsely populated Vasterbotten county (0.25 million inhabitants intotal, Anon., 2004b) are largely concentrated to a few population centers (70% of theinhabitants living in the two major towns), means that meeting the recreationaldemands would not require a transformation from ‘‘normal’’ to ‘‘recreationalfriendly’’ forestry all over the forest area in the county, which, in turn, would reducethe costs to forestry considerably. Consequently, in the benefit/cost analysis the costsare most likely exaggerated. More accurate cost estimates thus require that economicanalyses are combined with spatial modelling (cf. Horne et al., 2005).

The benefit estimates may also be exaggerated, because these estimates are basedon the underlying assumption that there is no decreasing marginal value in theincrease of the extent of ‘‘recreational friendly’’ forests. However, increasing, e.g. theshelterwood system from 3% to 34% of the forest area (one of the most importantcomponents of transforming the forestry to be ‘‘recreational friendly’’, according tothe analysis presented) is a large increase, and it is far from unlikely that such anincrease would involve a decreasing marginal value. Other and more detailedeconomic analyses have shown a decreasing marginal value when the ‘‘stock’’ of theforest recreation resource is increased, e.g. an analysis of the hunting value givendifferent moose population densities (Mattsson, 1990).

ARTICLE IN PRESS

G. Bostedt, L. Mattsson / Journal of Forest Economics 12 (2006) 75–81 81

There is obviously a lack of knowledge on how to balance forestry and forestrecreation in an economically efficient manner. In the further research that is needed,economics should be combined to a certain extent with other disciplines, e.g. forestproduction and silviculture, forest management and planning, sociology andgeography. In Sweden, forests close to population centers are often owned by themunicipalities, and results from the research should thus be of interest for this forestownership category in particular.

References

Anonymous, 2004a. Statistical Yearbook of Forestry. National Board of Forestry, Jonkoping.Anon, 2004b. Statistical Yearbook of Sweden. Statistics Sweden, Stockholm.Bengtsson, G., 1981. Oversiktlig beskrivning av Hugin-systemet. Mimeo, Department of Forest Survey,

Swedish University of Agricultural Sciences, Ume(a.Bojo, J., 1985. Kostnadsanyttoanalys av fjallnara skogar – Fallet V(al(adalen. EFI Research Report.

Stockholm School of Economics, Stockholm.Boman, M., Berg, C., Alroth, S., Bostedt, G., Mattsson, L., Gong, P., 2005. Environmental accounting

through adaptation of contingent valuation methodology: a proposal. Submitted for publication.Carson, R.T., Wright, J.L., Carson, N., Alberini, A., Flores, N., 1996. A Bibliography of Contingent

Valuation Studies and Papers. Natural Resource Damage Assessment, Inc., La Jolla.Clawson, M., Knetsch, J., 1966. The Economics of Outdoor Recreation. Johns Hopkins University Press,

Baltimore.Ekelund, H., Hamilton, G., 2001. Skogspolitisk historia. Report 8A. Swedish National Board of Forestry,

Jonkoping.Garrod, G., Willis, K., 1999. Economic Valuation of the Environment – Methods and Case Studies.

Edward Elgar, Cheltenham.Holgen, P., Lind, T., 1995. How do adjustments in the forest landscape resulting from environmental

demands affect the costs and revenues to forestry? J. Environ. Manage. 45, 177–187.Holgen, P., Mattsson, L., Li, C.Z., 2000. Recreation values of boreal forest stand types and landscapes

resulting from different silvicultural systems – an economic analysis. J. Environ. Manage. 60, 173–180.Holmertz, G., 1873. Studier I Skogstaxation. Norstedts, Stockholm.Horne, P., Boxall, P., Adamowicz, W.L., 2005. Multiple-use management of forest recreation sites –

a spatially explicit choice experiment. For. Ecol. Manage. 207, 189–199.Lindgren, J.E., 1976. Rekreationsanpassat skogsbruk – En ekonomisk bedomning. Rapport 1,

Troedssonska Forskningskuratoriet for Skogsbrukets Utveckling, Kungl. Skogshogskolan, Stock-holm.

Lindhagen, A., Hornsten, L., 2000. Forest recreation in 1977 and 1997 in Sweden: changes in publicpreferences and behaviour. Forestry 73, 143–153.

Mattsson, L., 1990. Moose management and the economic value of hunting – towards bioeconomicanalysis. Scand. J. For. Res. 5, 575–581.

Mattsson, L., Li, C.Z., 1993. The non-timber value of northern Swedish forests – an economic analysis.Scand. J. For. Res. 8, 426–434.

Mattsson, L., Li, C.Z., 1994a. Sample non-response in a contingent valuation survey – an empirical test ofthe effect on value inference. J. Leisure Res. 26, 182–189.

Mattsson, L., Li, C.Z., 1994b. How do different forest management practices affect the non-timber valueof forests? – An economic analysis. J. Environ. Manage. 41, 79–88.

Mitchell, R.C., Carson, R.T., 1989. Using Surveys to Value Public Goods: The Contingent ValuationMethod. Resources for the Future, Washington, DC.

Ranneby, B., Cruse, T., Hagglund, B., Jonasson, H., Svard, J., 1987. Designing a new national forestinventory of Sweden. Stud. For. Suec. 177.

Wilhelmsson, E., 1989. Modell och verklighet vid regionala avverkningsberakningar – N(agrarestriktioners inverkan p(a virkesproduktionen vid berakningar med Hugin-systemet. Report 48,Department of Forest Survey, Swedish University of Agricultural Sciences, Ume(a.