T H S N T N T C B : R C, C C P G P

Economic Studies 86 T����� H���� S���������� N����� T������ ��� ��� N����� �� T�������’ C���������� B�������: R��������� C��������, C���������� C����...
Author: Bryan Holmes
7 downloads 0 Views 1MB Size
Economic Studies 86

T����� H���� S���������� N����� T������ ��� ��� N����� �� T�������’ C���������� B�������: R��������� C��������, C���������� C���������� ��� ��� P����� G��� P������

T����� H���� Sustainable Nature Tourism and the Nature of Tourists’ Cooperative Behavior: Recreation Conflicts, Conditional Cooperation and the Public Good Problem

Department of Economics, Uppsala University Visiting address: Postal address: Telephone: Telefax: Internet:

Kyrkogårdsgatan 10, Uppsala, Sweden Box 513, SE-751 20 Uppsala, Sweden +46 18 471 11 06 +46 18 471 14 78 http://www.nek.uu.se/


ECONOMICS AT UPPSALA UNIVERSITY The Department of Economics at Uppsala University has a long history. The first chair in Economics in the Nordic countries was instituted at Uppsala University in 1741. The main focus of research at the department has varied over the years but has typically been oriented towards policy-relevant applied economics, including both theoretical and empirical studies. The currently most active areas of research can be grouped into six categories: • • • • • •

Labour economics Public economics Macroeconomics Microeconometrics Environmental economics Housing and urban economics

______________________________________________________________________ Additional information about research in progress and published reports is given in our project catalogue. The catalogue can be ordered directly from the Department of Economics.

© Department of Economics, Uppsala University ISBN 91-87268-93-0 ISSN 0283-7668

Sustainable Nature Tourism and the Nature of Tourists’ Cooperative Behavior: Recreation Conflicts, Conditional Cooperation and the Public Good Problem: HELDT, Tobias Doctoral dissertation to be publicly examined in Hörsal 2, Ekonomikum, Uppsala University, June 3, 2005, at 10.15 a.m. for the degree of Doctor of Philosophy. HELDT, Tobias 2005, Sustainable Nature Tourism and the Nature of Tourists’ Cooperative Behavior: Recreation Conflicts, Conditional Cooperation and the Public Good Problem; Department of Economics, Uppsala University, Economic Studies 86, 148 pp, ISBN 91-87268-93-0. This thesis consists of five essays. Essay I (with David Vail) explains why neither Maine, USA's comparatively laissez faire economic and land use institutions nor Dalarna, Sweden's more heavily regulated economy seems well designed to make tourism a powerful economic development engine. The paper focuses on three clusters of institutions that have a major influence on tourism's scale, economic structure, and long term sustainability. Labor laws and labor market institutions, Land ownership and property rights and Commodity taxes. The paper employs institutional contrasts between Dalarna and Maine to frame hypotheses that will guide future studies of sustainable tourism in forest regions. Essay II (with David Vail) Snowmobiling growth in North America and Sweden creates challenges in “governing the commons.” Snowmobiling contributes to the economy of distressed rural regions and enhances residents’ quality of life; but quasi-open access to winter landscapes also breeds conflicts: among snowmobilers, with landowners, with other recreationists, and with environmentalists and ecosystem health. Case studies in Sweden and Maine are used to illustrate how innovative governance institutions, complemented by infrastructure investments, can mitigate conflicts, re-align incentives, and internalize costs. Essay III (with Kreg Lindberg and Peter Fredman) Many natural areas are visited by diverse groups of recreationists, and in some cases the presence or behavior of one group may negatively impact the experience of another group. This recreation conflict may lead to access restrictions for the "offending" group. However, the magnitude of the gains and losses from such management interventions remain unknown. The present study utilizes choice experiment analysis to provide estimates of the economic value to cross-country skiers of conflict reduction from various levels of snowmobile presence. Essay IV uses a natural experiment to study the impact of an informal sanctioning mechanism on individuals’ voluntary contribution to a public good. Cross-country skiers’ actual cash contributions in two ski resorts, one with and one without an informal sanctioning system, are used. I find the contributing share to be higher in the informal sanctioning system (79 percent) than in the non-sanctioning system (36 percent). Furthermore, a CC-function, i.e. the relationship between expected average contributions of other group members and the individual’s own contribution, is elicited and compared between the two systems. Essay V tests for conditional cooperation and social comparisons in a natural field experiment, using decisions from a sample of cross-country skiers in Sweden on the issue of voluntary cash contributions to the preparation of ski tracks. Testing by experimentally varying the beliefs about others’ behavior, I find the share of subjects giving a contribution to be significantly greater in the group receiving information about others’ behavior than in the group that does not. Regression analysis cannot reject that subjects are affected by social comparisons and express a behavior classified as conditional cooperation. Tobias Heldt Department of Economics, Uppsala University and Department of Economics and Society, Dalarna University, SE-781 88 Borlänge, Sweden

Table of Contents

Acknowledgements ……………………………………………………………….…….. iii 1. Introduction 1.1 Sustainable Nature Tourism and the Nature of Tourists’ Cooperative Behavior ……………………………………………………….…………………… 1 1.2 The Public Good Problem in a Nature Tourism Context ………………………. 3 1.3 Field Experiments and Experimental Economics ………..…………………….. 4 1.4 Summary of Articles ……………………………………………..…………….. 6 Essay I. Institutional Factors Influencing Size and Structure of Tourism: Comparing Dalarna (Sweden) and Maine (USA).1 Current Issues in Tourism, 3(4): 283 – 324. Purpose and Method……………………………………………….….…............. 283 Labour Market Institutions: Costs, Rigidities and Incentives ……………………286 Sales and Excise Tax Wedges: Effects on Costs, Competitiveness and Incentive ……………….....................................................................…................ 294 Land Ownership and Property Rights: Recreational Access to Private and Public Land ………………………………………………………………....…… 299 Research and Policy Implications ……………….……………………..………... 311 Essay II. Governing Snowmobilers in Multiple Use Landscapes: Swedish and Maine (USA) Cases.2 Ecological Economics, 48(4): 469-483. 1. Introduction: Snowmobiles’ Contradictory Roles in Rural Life and Landscape …………………………………………………………………..…… 469 2. Conceptual Lenses ……………………………………………………………. 473 3. Field Research Methods ………………….………………..………...……….. 475 4. Mitigating Snowmobile Conflicts in Sweden’s Southwestern Mountains .…... 476 5. Taming Maine’s Snowmobilers: Clubs, Contracts, and Renewed Conflicts …. 478 6. Findings and Discussion: Evolutionary Adaptation of Governance Regimes ... 481 Essay III. The Value of Reducing Recreation Conflict: A Choice Experiment Evaluation of Snowmobiler-Skier Conflict 1. Introduction ……………………………………………….….….......................... 1 2. Managing Recreation Conflict by Tools of Economics …………………………. 3 2.1 Recreation Conflict …………………….………………...…..................... 3 2.2 Dimensions, Trade-Offs and the Relevance of Economics …..………….. 4 2.3 The Choice Experiment Approach …………………………..……………7 3. Methods ……………………….………………………….….….......................... 8 3.1 Study Area ……………………………………………………………….. 8 3.2 Data Collection …….……..…………………………………...…........... 10 1 2

I am grateful to Channel View Publications for permission to include the paper in this thesis. I am grateful to Elsevier for permission to include the paper in this thesis.


3.3 Analytical Model …..…………………………………………………… 12 3.4 Data Analysis………………………………………………..…………... 14 4. Results ……………………………...…………………….….…......................... 15 4.1 Choice Experiment Results ……………………………………………... 16 4.2 Welfare Measures ………..……………………………….…...…........... 21 5. Discussion ………………………………………………….…………………... 23 Essay IV. Informal Sanctions and Conditional Cooperation: A Natural Experiment on Voluntary Contributions to a Public Good 1. Introduction ………………...……………………………………………….…… 1 2. Field Context and Theory …………………………………………….…………. 3 2.1 Tourist Skiers as One-Shot Public Good Players ………..………………. 4 2.2 A Traditional Public Good Game ……………………………………..…. 4 2.3 Informal Sanctioning of Free Riders to Sustain Cooperation ……….…… 5 2.4 Conditional Cooperation ………………………………………….……… 7 3. Experimental Design …………………………………………………..……….. 10 3.1 Data Collection …………………...…………………………………………... 10 4. Analysis and Results ………………...…………………………………….…… 14 4.1 Result 1: Voluntary Contributions Differ Between Systems…….……… 14 4.2 Result 2: The Shape of the CC-Function Differs Between System …….. 17 5. Summary and Conclusions …………………………………………………….. 20 Essay V. Conditional Cooperation in the Field: Cross-country Skiers’ Behavior in Sweden 1. Introduction ………………...……………………………………………….…… 1 2. Field Experiment Context and Collection of Data ……………….……………… 3 2.1 Experimental Design, Field Conditions and Data ………..……………….5 3. Related Literature and Behavioral Hypotheses ………………………...……..… 9 4. Analysis and Results ………………………………………………………….... 12 4.1 The Total Sample ……….………………………………………………. 13 4.2 The Control Sample …………………………..………………………... 13 4.3 The Tourist Sample …………………………….………………...…...… 15 4.4 Further Testing of the Conditional Cooperation Hypothesis ………..….. 16 5. Conclusions …………………………………………………………….…..…... 19


Acknowledgements The accomplishment of this thesis project has been made possible thanks to both formal and informal encouragement and support from a large number of people. Most important, this thesis would not have existed without my supervisor Lars Hultkrantz. Thanks for believing that I had the ability to pull this through and for supporting me in times when others may have not. Ever since my undergraduate studies, throughout the following years of Ph. D. studies, I have felt encouraged to let my own decisions and research interests lead the way. Thanks Lasse! Thanks also for all your sharp economic intuitions and suggestions for improving my essays and for hooking me up with David Vail. I would like to thank my co-supervisor and co-author of two of the essays, David Vail, for giving me the opportunity to learn about life in Maine, USA, and of course for all discussions, suggestions and your mentorship in how to write academic English. A special thanks for encouraging me to continue working on what would later become the first article in this thesis. I also thank my experimental economics supervisor Jan-Eric Nilsson, who has always been a keystone in the Borlänge economic research environment (known under various disguises as T&S, CTEK, VTI/TEK, et. al.), for generous guidance and detailed comments on the manuscript in the final stages of this race. I also thank my co-supervisor Chuan Zhong Li for always accepting to help me with my statistical enquiries and for encouraging future research in environmental economics. Håkan Jerker Holm has contributed with useful comments and suggestions as the discussant at my final seminar. Kreg Lindberg and Peter Fredman, my co-authors and research fellows, have contributed by sharing my interest in outdoor activities and for initiating and leading our common essay project. I also thank Maria Vredin-Johansson for all discussions and exiting work in LISREL within our “Latent variable travel behavior” research project, and for insightful comments on my cross-country skiing field studies. Lena Nerhagen has encouraged me and helped me make the Grövelsjön and Sälen case studies come through. Participating researchers and staff at the research environment in the Borlänge and Uppsala departments have also been important. Among colleagues and graduate students, I especially thank Kenneth for sharing a passion for betting on horses and golf, Lars for being as lousy as I am at predicting football results and for sharing my MFF passion,


Maria for also talking kids and family, Martin for being a cheerful roommate, and all floor ball players for the fun and exercise. Eva Holst and Agneta Grummas provided help with administrative matters. During my graduate studies, I have had the privilege of visiting the Department of Resource Economics and Policy, University of Maine. I want to thank Kevin Boyle for hospitality and inspiring discussions. The trip to Maine for me and my family would have been much harder without the help and hospitality of Per Gårder with friends. Financial support from Dalarna University, the Foundation for Promotion of Expertise Relating to Tourism, the Swedish Research Council for Environment, Agricultural Science and Spatial Planning (FORMAS) and Christer Berch’s and E E:son Borgström’s Foundation is gratefully acknowledged. For help in authorising my field research, I am grateful to Christer Rosén, Säfsen Resort, Karl-Erik Westergård, Hovfjället, Lars-Axel Magnusson, Dalarna County Board, Charlie Ekberg, Grövelsjöns Fjällstation and Inge Wadman, Torsby Skidgymnasium. I am also grateful for voluntary labor in the field by Bernt Gunnarsson and participating members of SK Bore, Solweig and Lars Todell, Lena Heldt-Strandberg, Torbjörn Strandberg, Stefan Pettersson and Anders Skött. I also thank all my family and friends for encouragement and support. A special thanks to Magnus Thor for useful comments on one of the articles. Important occasions that have helped keeping my mental spirit up are among others: Tipsbolaget, Hattrick®, Skinz (no1) - Pondus football matches and dart games. If Lasse’s granting of an individually based study plan after the first six months of courses was the best that could have happened to my thesis project, the births of my two children are the best things that could have happened to me on a personal level. Thanks Alice and Vincent for giving balance to my daily life. Without the love and support from Susanna, I would not have made it. The “Heldt-Cassel Ph.D.-study relay” hereby ends.

Tobias Heldt Uppsala, April 28, 2005.


1. Introduction 1.1 Sustainable nature tourism and the nature of tourists’ behavior This thesis consists of five essays with one common denominator. They are all using a winter tourism activity as an application to reach their conclusions. The first three essays have a specific focus on strategies and management for sustainable nature tourism development, using methods of institutional economics and non-market valuation. The last two essays provide input into the discussion of the nature of human cooperation by studying individuals’ behavior in a public-good situation. Field experiment is used as the method and the behavior of Swedish cross-country skiers in tourist areas provides the application.


The starting point for the essays in this thesis is a research project on strategies for economically sustainable nature tourism in forest regions. Specifically, the problem of tourism use of recreational land when the possibilities of excluding visitors are limited, has been central. This introductory chapter shows that this specific focus and context have been useful for providing new insights into economically relevant problems in two major areas; applied tourism economics and the use of field experiments within economics. The research interest in economically sustainable1 nature tourism originates in the view of tourism as an economic growth opportunity by policy makers in Sweden and other industrial nations. Nature tourism is typically seen as a clean, green and labor intensive way of bringing rural development to regions suffering from high unemployment and out migration. Even if policy makers’ predictions of increased demand for nature tourism are true, institutional factors affecting tourism production, like labor law, taxes and property rights, may be ill-designed to facilitate growth. Furthermore, nature in general, and 1

As used here, sustainability refers to two aspects; that it functions within ecologically carrying capacity limits and that it contributes to social vitality within host regions. For a strategy to be economically sustainable, it should also meet the regular requirements for a profitable business.


recreational land in particular, is a multiple use resource. In combination with open access, the risk of short-term overuse and the potential of a long-term sustainable yield are two forces to balance. Nature may not be able to accommodate a growth of visitors without effects on the quality of the tourist experience. The tourism use of recreational land may not only lead to conflicts between different tourism uses, but also with other land-use interests, like forestry and environmental preservation, especially when certain types of uses are viewed as mutually exclusive.

The results in this thesis contribute to the understanding of how a selection of institutional factors influences the potential for growth in sustainable nature tourism. Moreover, it is shown how economic theory and methods can be used to provide inputs for the analysis of tourism management actions in a situation involving conflicts in land use.

The thesis’ contribution to the use of field experiments in economics originates in the existence of Allemansrätten2, and the fact that tourists’ use of recreational land can be seen as a one-shot action. Under Allemansrätten, visitors cannot be excluded from access, which implies that recreation facilities like camp grounds and trails can be seen as public goods. This context offers a suitable environment for applications of non-market valuation techniques, notwithstanding if the purpose is theory and method development or to provide user values to be used in cost benefit analysis of management actions. Moreover, in tourist areas, facilities on recreational land offer a real setting where individuals’ voluntary contribution behavior to fund a public good can be tested in a natural field environment.

In this thesis, two conjectures from economics laboratory are tested in natural field experiments. In general, testing whether results obtained in the labs are transferable to field environments is crucial for the value of experimental economics as an important method in economics. Essay four in the thesis adds to the literature of testing social


The Right of Common Access, further described in Vail and Heldt (2000).


comparisons in a natural field experiment. Essay five includes the first test of the impact of informal sanctions on voluntary contributions in a natural experiment.

1.2 The public good problem in a nature tourism context One latent, overall research question for the essays in this thesis is: “how can recreational public goods3 be financed when open access prevails?”. From this question follows the general problem analyzed in this thesis: the tourism use of natural environments with limited possibilities of excluding visitors. Basically, this problem consists of components stemming from the open access feature; overuse of the resource and underprovision of high-quality tourism services. Resource theory, based on an assumption that individuals only care about their own material self-interest, predicts overuse of the resource, since open access misaligns the individual’s incentive to expand resource use and the collective interest in restricting exploitation. Underprovision is a predicted result, since open access causes a free-rider problem (also known as a public good problem). Rational self-interest individuals have no incentive to voluntarily contribute to the funding of a public good.

The tourism use of a recreational public good adds another dimension to the problem. Tourism is a one-shot activity by nature. Field studies have indicated that open access common property regimes can survive in the long term, if there is a small number of members in the group and there exist strong mutual obligations and ties among members (Ostrom 2000). Tourists visiting a destination are typically not likely to meet any of the two criteria. In contrast, local inhabitants frequently interacting in the destination can be viewed as playing a repeated game.

When exclusion of visitors cannot be practiced, the rationale for direct pricing does not exist. Recreational public goods may then be financed through taxes, indirectly as a mark up on other goods and services or by voluntary contributions by users, i.e. privately provided public goods. 3

Recreational land is best considered as an impure public good with rivalry in use, in contrast to a pure public good characterized by non-rivalry in consumption and non-excludability of benefits (Vail and Hultkrantz 2000). Non-rivalry means that the consumption of one individual does not detract from the consumption possibilities of others, while non-excludability means that once a public good is produced, it is not possible to exclude anyone from consuming its services.


However, theories based on the self-interest hypothesis do not predict the existence of privately provided public goods. Theory predicts that since rational and selfish individuals have the incentive to free ride on others’ efforts and this is common knowledge, nobody will contribute to a public good. Yet, in real life, phenomena like charitable giving and successful collective actions are frequently observed. Furthermore, testing the self-interest hypothesis in laboratory games with contributions to public good has given results clearly refuting the zero contribution prediction (for a survey of PG games, see Ledyard 1995).

Today, theoretical as well as lab and field experiments on public goods have explained the observed cooperative behavior by social interaction, social preferences and conditional cooperation. Economists have for a long time pointed out that people do often care about the well-being of others, and that social interactions have economic consequences (e.g. Smith 1759, Becker 1974, Sen 1994). Recently, it has, for example, been shown that social interactions are important (Manski 2000), that social preferences are heterogeneous (Fehr and Fischbacher, 2002) and that a fraction of individuals are prepared to cooperate, conditional on others’ cooperation (Ostrom 2000, Fischbacher et al. 2001). Furthermore, theories of fairness and reciprocity have been proposed to explain these observed results (Fehr and Gächter 2000, Sugden 1984). These findings and propositions have served as a starting point for the fourth and fifth essays in this thesis.

1.3 Field experiments and experimental economics The use of field experiments in economics is a new and growing field of research. One example of this trend is the February 5th (2005) launching of a web site, “Field experiment, Bibliography” (www.arec.umd.edu/fieldexperiments, thanks to John A. List, University of Maryland) directed at facilitating public access to this field of research. The site contains about a hundred published papers and half as many working papers and forthcoming papers divided into three categories4, following the definition of field


Artefactual field experiments, framed field experiments and natural field experiments.


experiments in a recent publication in Journal of Economic Literature (Harrison and List 2004).

Harrison and List (2004) do not propose a bright line to define some experiments as field experiment and others as something else. Instead they use six factors to determine the field context of an experiment; the subject pool used, the information that subjects bring to the task, the commodity, the nature of the task and trading rules, the stakes and the environment in which the subjects operate. However, only varying one of the six factors does not make it a field experiment. For example, running a lab experiment using a physical commodity (e.g. chocolate bars or text books) instead of an induced valuation of a virtual good, does not constitute a field experiment (Harrison and List 2004). As a reference point, a traditional lab experiment takes place in a lab or classroom, uses a subject pool of students, an abstract framing of the task and a pre-specified set of rules.

This thesis includes one natural field experiment and one natural experiment. A natural field experiment can be considered an ideal field experiment in which a subject can be observed in a controlled setting, but the subject does not perceive any of the controls as unnatural.5 (Harrison and List 2004). Further qualifications of a natural field experiment is that it uses a non-student subject pool and has field context in either the commodity, task or the information provided. To contrast, a natural experiment arises by chance and the researcher simply observes naturally occurring situations. The treatment itself is considered the experiment and the naturally occurring comparison group acts as the control group. Ideally, the treatment effect is measured as the difference in outcome before and after for a treatment group with before and after outcomes for a non-treated group.

The value of experimental economics critically depends on whether results obtained in labs are transferable to field environments. Any of the six factors proposed to define a field context, could be a factor influencing behavior in a lab experiment and throw doubt on the generality of results. As suggested by Smith (1982) the scientific question that


The standard ethic of experimental economics to not practice any deception also applies.


appears after replicable results have been found in the laboratory, summarized as parallelism, is the applicability of results to field environments. Systematically testing the parallelism of lab results to field environments greatly benefits from the definitions proposed in Harrison and List (2004). If parallelism cannot be established, the research process continues. In such cases, there is much to learn from field experiments when returning to the lab. Behaviors that occur when control is somewhat lost in the field may be an indicator of a feature previously neglected in the lab Therefore, field experiments should be seen as a complement to traditional laboratory experiments and as a help to design better lab experiments (Harrison and List 2004). The following section shortly introduces the thesis essays and summarizes their main findings.

1.4 Summary of articles Essay I in this thesis is co-authored with David Vail, Bowdoin College, Maine USA. It was published in Current Issues in Tourism 2000. This essay discusses general problems for sustainable nature tourism in forest regions by comparing the conditions for tourism development in two regions, whose local economy in the past was heavily reliant on selling its natural resources. Downturn and transition in natural resource industries made the regions suffer from unemployment and out migration. Today, politicians in both regions once more hope to sell their natural resources, but this time bundled as nature and cultural based tourism experiences. This paper explains why neither Maine, USA's comparatively laissez faire economic and land use institutions nor Dalarna, Sweden's more heavily regulated economy seems well designed to make tourism a powerful economic development engine. The paper focuses on three clusters of institutions with a major influence on tourism's scale, economic structure, and long-term sustainability. Labor laws and labor market institutions are important determinants of tourism employment, job quality, product mix, production methods, and regional competitiveness. Land ownership and property rights both influence the incentives facing land owners, tourists, and tourism businesses and stresses on the ecosystem carrying capacity. Commodity taxes affect the absolute and relative 6

prices of various tourist services and, via feedback effects on demand, influence tourism's aggregate scale, activity mix, and transportation/location patterns. The paper employs institutional contrasts between Dalarna and Maine to frame hypotheses that will guide a larger comparative study of sustainable tourism in forest regions. Perhaps most controversially, we hypothesize that Sweden's venerable right of common access (Allemansrätten), as currently implemented, impedes sustainable tourism development. An appendix sketches the current state of tourism in the two regions.

Essay II is a snowmobiling case study, where the problem is taken from one of the hypotheses on property rights in Vail and Heldt 2000. It was published in Ecological Economics 2004.

Snowmobiling growth in North America and Sweden creates challenges in “governing the commons.” Snowmobiling contributes to the economy of distressed rural regions and enhances residents’ quality of life; but quasi-open access to winter landscapes also breeds conflicts: among snowmobilers, with landowners, with other recreationists, and with environmentalists and ecosystem health. Common pool resource and impure public goods theories are used to interpret these conflicts and strategic interactions. Case studies in Sweden and Maine yield insights about conditions for sustainable management of multiple-use landscapes when property rights are complex and stakeholders diverse. The case studies utilize key informant interviews, tourist surveys, and a contingent valuation exercise to illustrate how innovative governance institutions, complemented by infrastructure investments, can mitigate conflicts, re-align incentives, and internalize costs. Local self-governance has evolved with state facilitation, but without changes in fundamental property law. Three keys are trail agreements between snowmobile clubs and landowners, clubs’ norm formation and rule enforcement, and public-private investment in quality trails. These arrangements are strained by hot spot congestion, free riding, and unresolved conflicts with other recreationists.

Essay III is coauthored with Kreg Lindberg, Oregon State University, USA and Peter Fredman, European Tourism Research Institute, Sweden. It is written to appeal to the


audience of the journal Forest Science and although the case is from Sweden, discussions and results are put in the context of a recent policy problem in the U.S. The paper is invited for resubmission to Forest Science.

Many natural areas are visited by diverse groups of recreationists and, in some cases, the presence or behavior of one group may have a negative impact on the experience of another. This recreation conflict may lead to access restrictions for the "offending" group. However, the magnitude of the gains and losses from such management interventions remains unknown. The present study utilizes choice experiment analysis to provide estimates of the economic value to cross-country skiers of conflict reduction from various levels of snowmobile presence. The values are substantial, though they vary widely depending on the level of conflict reduction and the cost measure utilized in the calculations. In addition, the results indicate protest behavior. The presence of protest behavior and the sensitivity to cost measures have implications for other choice experiment analyses.

Essay IV is the first of two applying experimental methods in the field to explore conditions for cooperative behavior.

In a natural experiment, this paper studies the impact of an informal sanctioning mechanism on individuals’ voluntary contribution to a public good. Cross-country skiers’ actual cash contributions in two ski resorts, one with and one without an informal sanctioning system, are used. I find the contributing share to be higher in the informal sanctioning system (79 percent) than in the non-sanctioning system (36 percent). Previous studies in one-shot public good situations have found an increasing conditional contribution (CC) function, i.e. the relationship between expected average contributions of other group members and the individual’s own contribution. In contrast, the present results suggest that the CC-function in the non-sanctioning system is non-increasing at high perceived levels of others’ contribution. This relationship deserves further testing in lab.


Essay V further dwells into the nature of cooperative behavior of skiers departing in a hypothesis of conditional cooperation. In a laboratory one-shot public good game, Fischbacher, Gächter and Fehr (2001) classify 50 percent of the subjects as conditional cooperators. Outside the lab, using a student sample, Frey and Meier (2005) find that people behave pro-socially, conditional on others’ behavior. This paper tests for conditional cooperation and social comparisons in a natural field experiment, using decisions from a sample of cross-country skiers in Sweden on the issue of voluntary cash contributions to the preparation of ski tracks. Two test procedures are used. First, testing for correlation between beliefs about the contribution of others and own behavior and second, experimentally varying the beliefs about others’ behavior. Using the latter approach, I find the share of subjects giving a contribution to be significantly greater in the group receiving information about others’ behavior than in the group that does not. Regression analysis cannot reject that subjects are affected by social comparisons and express a behavior classified as conditional cooperation.


References Becker G. S. (1974). A Theory of Social Interactions. Journal of Political Economy, 82(6): 1063-1093. Fehr E. and S. Gächter (2000). Fairness and Retaliation - The Economics of Reciprocity. Journal of Economic Perspectives, 14: 159-181. Fehr E. and U. Fischbacher (2002). Why Social Preferences Matter – The Impact of Nonselfish Motives on Competition, Cooperation and Incentives. The Economic Journal, 112: C1-C33. Fischbacher U., S. Gächter and E. Fehr (2001). Are People Conditionally Cooperative? Evidence from a Public Goods Experiment, Economic Letters, 71: 397- 404. Frey B. S. and S. Meier (2005). Social Comparisons and Pro-social Behavior: Testing Conditional Cooperation in a Field Experiment. Forthcoming in American Economic Review. Harrison G. W. and J. A. List (2004). Field Experiments. Literature, 42(4): 1009-1055.

Journal of Economic

Ledyard J. O. (1995). Public Goods: A Survey of Experimental Research, in A.E.Roth and J. Kagel (eds.), The Handbook of Experimental Economics. Princeton: Princeton University Press. Manski C. F. (2000). Economic Analysis of Social Interactions. The Journal of Economic Perspectives, 14(3): 115-136. Ostrom E. (2000). Collective Action and the Evolution of Social Norms. Journal of Economic Perspectives, 14: 137-158. Sen A. (1994). The Formulation of Rational Choice. American Economic Review, 84(2): 385-391. Smith A. (1759). Inquiry into the Nature and Causes of the Wealth of Nations. Buffalo: Promethus Books, Reprint (1991). Smith V. L. (1982). Microeconomic Systems as an Experimental Science. American Economic Review, 75: 265-272. Sugden R. (1984). Reciprocity: The Supply of Public Goods through Voluntary Contributions. The Economic Journal 94: 772-787. Vail, D. and L. Hultkrantz (2000). Property Rights and Sustainable Nature Tourism: Institutional Adaptation and Mal-adaptation in Dalarna and Maine. Ecological Economics, 35 (2): 223 - 242.


Discussion Web Site: http://www.commerce.otago.ac.nz/tourism/current-issues/homepage.htm

Institutional Factors Influencing the Size and Structure of Tourism: Comparing Dalarna (Sweden) and Maine (USA)* David Vail Bowdoin College, Brunswick, Maine, USA

Tobias Heldt Center for Transportation Economics, Dalarna University, Borlänge, Sweden This paper explains why neither Maine, USA’s comparatively laissez faire economic and land use institutions, nor Dalarna, Sweden’s more heavily regulated economy, seems well designed to make tourism a powerful economic development engine. The paper focuses on three clusters of institutions that have a major influence on tourism’s scale, economic structure, and long-term sustainability. Labour laws and labour market institutions are important determinants of tourism employment, job quality, product mix, production methods, and regional competitiveness. Land ownership and property rights influence both the incentives facing landowners, tourists, and tourism businesses and stresses on ecosystem carrying capacity. Commodity taxes affect the absolute and relative prices of various tourist services and, via feedback effects on demand, influence tourism’s aggregate scale, activity mix and transportation/location patterns. The paper employs institutional contrasts between Dalarna and Maine to frame hypotheses that will guide a larger comparative study of sustainable tourism in forest regions. Perhaps most controversially, we hypothesise that Sweden’s venerable right of common access (allemansrätten), as currently implemented, impedes sustainable tourism development. An appendix sketches the current state of tourism in the two regions.

Purpose and Method This paper is the starting point for a study of strategies for economically sustainable nature tourism in forest regions. It focuses on three clusters of institutions that have a major influence on tourism’s scale, economic structure, and long-term sustainability in Dalarna County, Sweden, and the US state of Maine. Tourism scale refers to magnitudes such as annual visitor numbers, guest nights at commercial establishments, total tourist spending, economic multiplier effects, and employment levels.1 Tourism’s core structural characteristics are the mix of commercial and non-commercial recreation, seasonal distribution of visits, land ownership and business organisation patterns, and methods of producing tourist services. Sustainability, as used here, refers to tourist activities that meet two conditions: they function within ecological carrying capacity limits and they contribute to durable economic prosperity and social vitality in host regions. The three key institutional clusters are described, compared, and contrasted:

… Labour laws and labour market institutions, which are important determinants of tourism employment, job quality,2 product mix, production methods, and regional competitiveness. 1368-3500/00/04 0283-42 $16.00/0 Current Issues in Tourism

© 2000 D. Vail & T. Heldt Vol. 3, No. 4, 2000


Current Issues in Tourism


… Land ownership and property rights, which affect tourism’s economic potential by influencing the incentives facing landowners, tourists and tourist businesses, as well as the level of stress on ecosystem carrying capacity. … Commodity taxes influence the absolute and relative prices of various tourist services and, through feedback effects on demand, affect tourism’s aggregate scale, activity mix, and transportation/location patterns. This paper uses institutional contrasts between Dalarna and Maine to frame the hypotheses that will guide our future investigations. Later sections present several such hypotheses and corollaries and flesh out the assertion that most Maine institutions lie near the laissez faire end of the political-economic spectrum, while Swedish tourism reflects more extensive state intervention in the economy and property rights. Labour, land, and tax institutions are by no means the only ones shaping tourism, but they are particularly important. Future research will include comparative investigation of additional public institutions, such as transportation networks, education and training, and tourism research. A co-evolutionary institutional perspective We employ a very basic notion of institutions, derived from the work of economist Douglas North and summarised by Folke et al.: By institutions we mean the humanly devised constraints that shape human interaction and the way societies evolve through time. Institutions are made up of formal constraints (rules, laws, constitutions), informal constraints (norms of behavior, conventions and self-imposed codes of conduct), and their enforcement characteristics. (Folke et al., 1997: 3) Starting from this broad understanding of institutions, it should be clear that tourism in industrial societies is enmeshed in a web of institutions, with strands running from markets to property rights and from environmental ethics to public policies. This is the matrix within which tourism’s key players – tourism service producers, trade associations, land owners, travel agents, seasonal employees, labour unions, local officials, national governments and, of course, leisure travellers themselves – make decisions and take action. Our premise is that, if we hope to explain the behaviour of individuals and organised interests and identify their cumulative effects, we must first understand the institutional rules and norms that shape behaviour. This essay attempts to unravel webs made up of both deliberately designed and informally evolved institutions and to show the major impact they have had on the scale and structure of tourism in two places, Maine (USA) and Dalarna (Sweden). It demonstrates how particular institutional differences operating within broadly similar societies can lead to quite different economic, equity and environmental outcomes. The emphasis on institutionalised relationships and behaviour also reflects an interpretation of institutions as an important part of every society’s accumulated ‘capital’. Institutional assets, such as shared values, knowledge and ethical standards can contribute to or detract from a collective purpose, in this case sustainable tourism development. Institutional capital can be accumulated through deliberate investments (e.g. in legislation, education, formation of private organisations) or it can depreciate through neglect. From an evolutionary

Institutions Determining Tourism’s Scale and Structure


perspective, a lack of institutional resilience in the face of changing conditions and collective purposes can render ineffective or even counterproductive institutions that were once highly effective. For example, several features of both Swedish and American labour market policy have proven to be maladapted to the policy objective of expanding the number of high quality tourism jobs in distressed rural regions. In contrast, institutions may adapt and prove resilient in changing circumstances. For example, private landowner associations have been created in both Maine and Dalarna to reduce transaction costs and capture infrastructure economies of scale in back country recreation. Government measures to promote such private institutional innovations are particularly intriguing (Becker & Ostrom, 1995; Folke et al., 1997). Dalarna and Maine: A ‘quasi-experiment’ A few words are in order about why Dalarna County and interior Maine were chosen as a basis for assessing the potential for sustainable tourism development. Well before tourism became a mass phenomenon toward the end of the last century, Dalecarlia’s visitors, including story-teller H. C. Andersen and heritage park builder Arthur Hazelius, and travellers in Maine’s vast northern wilderness, such as naturalist Henry David Thoreau and novelist Nathaniel Hawthorne, were shaping a cultural mystique that still surrounds these places. The mystique was reinforced in the popular imagination by turn-of-the-century artists such as Carl Larsson, Anders Zorn and Winslow Homer. A century later, Dalarna and Maine still retain the special allure of places that are accessible and yet ‘off the beaten path’, with distinctive folkways and stunning lake and mountain scenery.3 Dalarna and Maine are both predominantly rural, with extensive mountain ranges and thousands of lakes and ponds. With the return of marginal farmland to forest, over 90% of the land area is forested. Three-fourths of Dalarna’s land area is privately owned, while the Maine figure is close to 95%. The southern parts of both regions are within an easy day’s drive for prospective visitors from major metropolitan centres: Stockholm, Göteborg, and Oslo; Boston, New York, and Montreal. Economic conditions in both places historically rose and fell with the fortunes of natural resource-based industries, which, for complex reasons, are now mature or declining. Policy makers seeking new rural development opportunities have identified tourism as a clean, green, and labour intensive way to grow: exploiting underutilised natural and human capital, diversifying the economic base, creating jobs, and revitalising rural communities. This ambition is reflected in Maine’s 1995 tourism Marketing and Development Strategy: ‘Within the context of an emerging global economy, the growth of the tourism industry . . . brings jobs and stability to the state’ (MOT, 1995: 2). The strategy sets a tourism growth target of 3% per year, compared to 2+% for the overall economy. Dalarna’s County Board echoes that, ‘There are big hopes in the county for an increase in tourism and employment within tourism. These hopes take as their starting point that Dalarna is [already] a major tourist county and that tourism internationally is one of the fastest growing industries’(LD, 1998: 3). Beyond its contributions to income, employment and tax revenues, Dalarna tourism is expected to help

Current Issues in Tourism


maintain quality of life in host communities via better physical facilities, social services and cultural activities. There is, in particular, a widespread conviction that ‘Nature’ has a unique and increasingly important place in the leisure life of very large nearby populations. Maine tourism promotions make much of the fact that the state has the largest contiguous forest ‘wildland’ east of the Mississippi River, including the only designated wilderness waterway, the northern terminus of the 3000 km Appalachian Trail, and 40,000 moose to shoot with camera or rifle. Sahlberg paints a similar picture of interior Sweden: The environment in its entirety with clean air and clean waters, the lack of noise and crowding has got qualities that seldom are found at other places in Europe . . . These conditions together with the Swedish right of common access make these areas unique in Europe. (Sahlberg, 1998: 229) In sum, Dalarna and Maine offer a fruitful basis for comparative study of tourism development for two reasons. First, there are important historic, geographic and economic parallels. Second, policy makers are convinced their jurisdictions are uniquely situated to make tourism a major engine of economic growth. This paper explains why, in fact, neither Maine’s comparatively laissez faire institutions nor Dalarna’s more heavily regulated tourism seems well designed for that purpose. Although Dalarna and Maine do not offer a controlled laboratory to investigate institutional variations, the overlay of specific differences on basic institutional similarities gives us a promising ‘quasi-experiment’ for generating and testing hypotheses.

Labour Market Institutions: Costs, Rigidities and Incentives Dalarna’s high cost tourism labour There are no regional or sectoral differences in Sweden’s labour laws and institutions; thus what goes for Sweden also goes for Dalarna. Historically, Swedish wage formation and the regulations governing many labour market institutions were based largely on conditions in manufacturing industries, where full-time, career-long jobs were the norm. Today service sectors employ most workers, and part-time and limited-term employment have become much more common. Public policies and collective bargaining arrangements designed for other sectors and earlier times have a major impact upon the rural tourism labour market. Unionisation, collective agreements and average tourism wages Sweden has no statutory minimum wage. Instead, collective agreements between employee organisations and employer organisations or individual employers regulate pay arrangements within each line of business. For tourism industry employees, the 1997 starting hourly wage for adult workers was SEK 65 ($8.70), but the average gross wage reported by the Hotel and Restaurant union, which organises most tourist industry employees, was SEK 96/hour ($12.80). Gross wages include vacation pay and supplements for night and weekend work4 (HoR, 1998). Table 1 shows the wide range of annual earnings in various tourism-related sub-sectors. (The ‘All Tourism’ figure is a weighted average

Institutions Determining Tourism’s Scale and Structure


Table 1 Average annual gross earnings in tourism sub-sectors (1994) a Sector All Services All Tourism Hotel and restaurant Ground transport Air transport Travel bureau Recreation, culture, sport

Average Earnings SEK169,100 ($22,500) 158,500 ($21,100) 114,300 ($13,300) 166,400 ($22,200) 281,000 ($37,500) 192,600 ($25,700) 159,400 ($21,300)

% of service sector average 100 94 68 98 167 114 94


Excludes mandatory employer-financed benefits Source: TD, 1998a

across occupational groups, e.g. hotel managers and chambermaids; airline pilots and ground crews.) Average earnings in tourism are held down by the preponderance of relatively poorly educated, low paid, and frequently seasonal hotel and restaurant workers. As a group, tourism employees have significantly less education than all service sector employees: 31% have just a nine year elementary school education, compared to 19% of all service employees; only 16% have post-secondary education, versus 32% of all service employees (TD, 1998a). Roughly 70% of employees in tourism’s various sub-sectors are trade union members, which is about average for Sweden’s private services. Union membership varies within tourism, with high-skill, year-round workers such as cooks and buffet managers 90% unionised, and less skilled and more seasonal groups, like chambermaids, at about 50% (HoR, 1998). Unionisation raises wages above the competitive market level, and the much higher rate of union membership in Dalarna compared to Maine probably explains part of the big wage differences revealed in the sub-section on Maine, below. Unions also work to secure better non-wage conditions for their members. Swedish labour law and the collective agreement The generic term for statutes governing work conditions, workers’ rights and collective bargaining is labour law. Labour legislation has given workers several rights, including the right to participation in workplace decision-making, paid vacations, limited working hours, improved working environment, and increased employment security. An employee’s most important protection is The collective agreement, a written, signed contract between the labour organisation and employer organisation or individual employer which specifies obligations for both parties. It functions for the employee in the same way as legislation, such as the Working Hours Act or the Annual Leave Act. Apart from provisions concerning pay and work conditions, the collective agreement gives employees supplementary insurance protection and pension contributions, beyond the mandatory employer contributions discussed below. In the USA and several other countries minimum wage laws provide one protection that the collective agreement gives in Sweden, where as mentioned, there is no such legislation. The collective agreement is widely viewed as giving workers more effective protection in Sweden than in most other countries (LO, 1999).


Current Issues in Tourism

Table 2 Mandatory employer contributions (% of gross wage) Old age pension Survivors’ pension Sickness insurance Parental leave insurance Occupational injury insurance Labour market contribution General payroll Total

6.40 1.70 7.50 2.20 1.38 5.84 8.04 33.06

Source: LO, 1999

Mandatory employer contributions and total labour costs Tourist sector employers make mandatory contributions, currently 33.06% of the gross wage (see Table 2). These benefits extend to tourism’s part-time and seasonal employees. In addition, full-time employees receive 25 days of annual paid vacation, while seasonal workers receive a 13% hourly wage supplement in lieu of vacation. Self-employed tourist entrepreneurs must pay fees similar to mandatory employer contributions (at 31.05% of salary), as well as income tax. Combining mandatory employer contributions with gross wages, the average labour cost for hourly tourism workers is approximately SEK 128/hour ($17.07), or about twice the Maine figure reported below. In addition to the mandatory contributions, most employers make supplementary insurance and pension contributions, currently 5.15% of the gross wage. Unemployment insurance, layoff legislation and seasonal employment patterns Another institution affecting tourism’s labour costs, incentives, and labour market flexibility is the unemployment insurance system. To be eligible for unemployment benefits, a person must have worked at least six months in the past year; the majority of seasonal tourism employees receive benefits during the ‘off season’. Eligible workers receive up to 80% of their immediate past monthly wage, to a maximum of SEK 580 ($77) per day (Eriksson, 1999; HoA, 1998). This generous eligibility rule acts like a subsidy to seasonally dependent industries like tourism and their part-year employees. The artificial increase in year-round compensation probably inflates tourism’s labour supply and reduces wages, since employees know they will be compensated when laid off. Employers can more readily vary labour input and production costs in response to seasonal demand fluctuations. Research confirms that Sweden’s comparatively generous unemployment compensation increases labour market flexibility and reinforces underlying seasonal employment fluctuations (Hultkrantz et al., 1987). Employment security legislation, known as LAS, governs layoffs, prior notification, etc. In general, LAS works against seasonal and temporary employment, but in the tourist industry specifically, seasonal layoffs are allowed. To benefit from LAS protections, a worker must have had at least six months of employment within the last two years.

Institutions Determining Tourism’s Scale and Structure


Regulations regarding normal work hours and layoffs also affect labour market flexibility. Since tourist demand fluctuates widely by season, many employers have an interest in extending work hours at peak seasons and cutting them in slack periods. By keeping more permanent, year-round employees, they can reduce recruitment, training and management costs. The 1982 Working Hours Act (SFS, 1982: 673) set the normal work week at 40 hours, averaged over a four week period. However, a 1995 law allows averaging of hours over a full year with the agreement of the collective bargaining agents (HoR, 1998). This mechanism is increasingly used by the larger tourism employers, which number 20 to 30 in Dalarna (Persson, 1999). Such flexibility reduces seasonal employment fluctuations and further expansion of work-time averaging might also reduce financial pressure on the unemployment compensation system. Maine’s ‘cheap labour’ tourism Maine tourism operates in a quasi-free market environment, encouraging employers to adopt a ‘cheap labour’ strategy by hiring relatively unskilled workers and producing services with labour-intensive methods. The key public institutions, elaborated below, are a low statutory minimum wage, low payroll taxes, limited employer-financed benefits, and minimal restrictions on layoffs. Two key features of the private institutional environment are free market ideology and negligible unionisation of tourism workers. We will refer below to the notion of a ‘livable wage’. This concept has been defined by a prestigious public–private body, the Maine Economic Growth Council, as the minimum full-time earnings required to pay for a two-person household’s basic necessities. Operationally, the livable wage is set at 185% of the official poverty threshold. In 1998, the last year for which comprehensive occupational wage data are available, the livable wage threshold was $9.46/hour. In 1968, 68% of Maine workers received a livable wage. As a point of reference, the Economic Growth Council set a goal of 85% livable wage jobs by the year 2005. The minimum wage and tourism wage formation Since late 1997, the US minimum wage has been $5.15/hour (SEK 38.6), or 64% of the Swedish level.5 Several states, including four of Maine’s five New England neighbors but not Maine itself, have enacted higher minimum wages. The US Congress adjusts the legal minimum upward periodically, but its long-term purchasing power has declined. In fact, ‘hospitality industry’ lobbying has been a major force against higher federal and Maine minimum wages.6 An individual with no dependants would have to be employed full time at $5.25/hour just to reach the official US ‘poverty line’. As noted above, Maine’s livable wage threshold was $9.46 in 1998. The state’s average wage was $12.07 in that year (MDOL, 2000). Wage formation in Maine’s tourist services is not significantly distorted by market power on either side of the labour market. At least three-fourths of tourism employment is in three sub-sectors, eating places, lodging establishments and retail outlets. On the supply side, almost no workers in these sub-sectors belong to labour unions or are covered by collective bargaining agreements.7 On the demand side, most businesses serving tourists are very small relative to their local labour markets. (There are exceptions, such as ski resorts in sparsely populated rural areas.) There is no complete or unbiased survey of tourism wages and

Current Issues in Tourism


Table 3 Average hourly wages in Maine’s key tourism sub-sectors Sector Retail sales Food & alcohol Lodging Recreation Total

% of total tourist Average hourly wage spending 1994 1998 44 $7.59 $8.35 24 6.46 7.10 17 6.76 7.44 10 7.08 7.79 95 1998 weighted average $7.42



62.6 53.3 55.8 58.4 55.2

Sources: MDOL, 2000; MEGC, 1998: 5; Vail et al., 1998: 43

salaries, but based on survey data from lodgings and restaurants plus state wage statistics for key tourism services, we are confident Table 3 presents accurate orders of magnitude for the majority of tourism’s hourly employees. Normative evaluation of these low wages is complicated by the fact that many workers take tourism jobs by choice, seeking seasonal or part-time work to complement activities such as education, domestic work, farming and forestry. It is nonetheless pertinent that the median wage for motel room maids, retail sales clerks, and non-tipped food sector employees is barely above the official poverty threshold for an individual. In 1998 only about one-fourth of hourly paid tourism workers received the $9.46 wage needed to meet a small family’s basic needs budget (Vail & Kavanaugh, 2000: Appendix 1). Prominent Maine policy makers urge an increase in tourism promotion, arguing that it will create jobs; but such a strategy is questionable when most tourism jobs pay so poorly. Policy makers seem also not to grasp that a cheap labour strategy in effect subsidises tourists and owners of labour-intensive tourist businesses. The cheap labour strategy also has hidden social and fiscal costs, including higher state and local taxes to finance health and welfare benefits for the working poor and the near-poor (BLE, 1991). Of course thousands of self-employed and wage earning tourism workers earn much more than the average hourly wage. Even among the hourly workers the wage distribution is skewed, with the median 15–20% below the average in restaurants and lodgings (MDOL, 1997: 153). In the special conditions of southern and coastal Maine, where tight labour markets have pushed summer unemployment rates down to 1.7–2.5% since 1997, even low-skill, entry-level workers receive wages approaching the averages in Table 3. The average salaries of managerial employees in lodgings and restaurants (c. 20% of employees) are equivalent to hourly wages of $9.31 (SEK 69.8), and $9.51 (SEK 71.3), respectively, and operators of expensive southern Maine restaurants report that their tipped waiters earn more than $15 (SEK 113)/ hour in the peak season. At the ‘top of the heap’ are several thousand business owners, craftspeople, entertainers, sporting guides and other highly skilled tourism service providers. By our estimate, 38% of all people engaged in tourism occupations earned a livable wage in 1998, dramatically below the 68% figure for the entire Maine workforce (Vail et al., 1998: 42; Vail & Kavanaugh, 2000: Appendix 1). Mandatory and voluntary employer contributions and total labour costs All employers contribute to the federal Social Security (SS) and state Unemployment Insurance (UI) funds and make Workers’ Compensation (WC)

Institutions Determining Tourism’s Scale and Structure


Table 4 Cost of mandatory employer contributions Social Security Unemployment Insurance Workers’ Compensation Total

% of payroll 6.25 3.2b 2.5c 11.95


Average hourly cost $0.46 SEK 0.24 0.19 $0.89 SEK

3.45 1.80 1.43 6.68


Based on an average tourism wage of $7.42 in 1998. Currently, small employers pay 2.8% on the first $7000 of an employee’s wages plus a 0.4% surcharge to maintain the UI system’s solvency. c Estimated average for all tourism employment. Sources: Kimball, 1999; Murphy, 1999 b

insurance payments to cover job-related accidents, injury and illness. WC insurance premiums vary according to sectors’ past health and safety records. In three low risk, low stress activities that contribute much tourism employment – hotels, restaurants and retail clothing stores – WC as a share of payroll costs is 2.7%, 2.2% and 1.3% respectively. Riskier and more stressful tourism activities, such as amusement parks (3.7%) and alpine ski operations (16.2%), have considerably higher WC premiums. Table 4 presents a ‘best guess’ of industry-wide mandatory benefits in tourism and recreation. The key implication of these approximations is that mandatory contributions add only about one-third as much to total labour costs in Maine as in Dalarna. Conceivably, voluntary employer contributions might make up much of the difference. Under federal and state law, employers are not required to underwrite health insurance, pensions, vacations, or medical and family leave. Although such benefits are fairly common among firms with over 500 employees, they are far less common among the small businesses typical of tourism. Indeed, if we focus on the most important employee benefit, health insurance, the proportion of small employers offering coverage is declining, as a cost-cutting reaction to fast escalating premiums (St John et al., 2000). The following data from a survey of Maine lodging and restaurant operators are the best available, although they are probably biased upward. 8 (See Table 5). It is difficult to assess the economic significance of voluntary benefits because the survey responses did not include monetary amounts; further, comparable information is lacking for the many other tourist services. America’s big-ticket employee benefit is employer-subsidised health insurance, which adds as much as 5–7% to payroll costs. Considering the low frequency of employer subsidised health insurance and other benefits reported above, a reasonable upper bound estimate for all voluntary employer contributions is 5% of payroll, or $0.37/hour. Combining information about wages, mandatory payments and voluntary contributions, we estimate that the average hourly labour cost in Maine tourism is in the neighbourhood of $8.64/hour (SEK 64.8), with $7.42 in wages and $1.22 in benefits. Benefits thus represent 16.5% of gross wages. In sum:

… Maine’s average tourism wage is about the same as Dalarna’s lowest wage9

Current Issues in Tourism


Table 5 Voluntary contributions by lodging and restaurant employers (% of respondents providing the benefit to hourly workers in any amount) Contribution to health insurance Paid holidays Paid vacations Paid sick leave Free or subsidised meals Profit sharing or bonuses

Lodgings (%) 34 25 19a 20 46 23

Restaurants (%) 27 32 53a 20 98 35


Typically applies only to full-time, year-round workers with at least one year of experience, and typically one week of paid vacation, increasing to two or three weeks after several years of service. Source: Austin Associates, 1997 (unnumbered pages)

… Dalarna’s average hourly compensation, including seasonal workers’ supplemental vacation pay and unemployment compensation, is about 70% above Maine’s; and … Dalarna’s average hourly labour cost is twice Maine’s. Institutions affecting seasonal employment flexibility Under Maine law, employers with more than 100 workers must give 90 days notice prior to plant closings or mass layoffs. Large employers must also make severance payments to workers with three or more years seniority. These laws have no effect on the great majority of tourism sector firms and do not apply at all to seasonal employees. Employers have broad discretionary power to fire individual workers for cause. Although terminated employees can demand a written explanation for their firing, it is difficult and costly to win compensation or reinstatement through the legal appeals process (Murphy & Hanson, 1995). In brief, there is practically no legal obstacle to terminating seasonal employees nor any penalty for laying off workers when tourist demand is slack. Of course, employers may have economic incentives to retain highly skilled workers through slack seasons and to take account of future recruitment and training costs when making layoff decisions. Particularly in Maine’s south coastal region, with its low unemployment and year-round demand base, such considerations probably deter some layoffs. On the labour supply side, few seasonal workers are eligible for unemployment compensation in Maine. Benefit levels are, in any case, low. In contrast to Sweden, where tourism workers have an incentive to apply for seasonal jobs because they qualify for unemployment compensation at 80% of past wages, America’s tough UI eligibility criteria do not strengthen the incentives to seek seasonal employment, for example as ski lift operators, fishing guides, ‘burger flippers’, chambermaids, parking lot attendants, or whale watch crews. Hypotheses We hypothesise that tourism in Dalarna and Maine is affected in six specific ways by the labour laws and other institutions described above.

Institutions Determining Tourism’s Scale and Structure


Hypothesis 1 Maine’s cheap labour and Sweden’s expensive labour induce tourism service producers to adopt relatively labour-using production methods in Maine and relatively labour-saving methods in Dalarna. One crude indicator of this effect is that tourism (spending) accounts for roughly 4.4% of Dalarna’s employment, versus 3.9% of gross county product;10 while Maine tourism accounts for 8.1% of employment versus 4.9% of gross product. Another way to describe the difference is that Maine tourism generates a full-time equivalent (FTE) job for each $64,100 of tourist spending, while it takes $106,700of spending to create a FTE job in Dalarna. (Calculations based on official tourism employment and expenditure data: Longwood, 1997; LD, 1998; NUTEK, 1999.) Corollary 1: Swedish tourism generates relatively high quality jobs, especially taking into account seasonal workers’ health care coverage, supplemental vacation pay, and unemployment compensation eligibility. Corollary 2: Maine tourism generates a large number of jobs, but the combination of labour laws and market forces results in low median wages, meagre benefits, and little unemployment compensation for less-skilled workers. Hypothesis 2 Labour costs are a major determinant of many tourism service prices. For instance, labour accounts for 30% of production cost in Swedish hotels and restaurants, compared to 15% in wood products and 12% agriculture and forestry and other major rural sectors (TD, 1996: 15). Maine tourists have a comparatively strong incentive to consume labour-intensive commercial services, while Dalarna tourists have a stronger incentive to substitute their own labour for purchased services. (This calculus is presented formally in the section below headed Sales and Excises Tax Wedges.) A crude example is the comparative prevalence of restaurant meals (in Maine): 24% of Maine tourist spending is at eating and drinking establishments, compared to 18% in Dalarna. The importance of this difference is suggested by the fact that, per krona of expenditure, restaurant meals create twice as much employment as food purchased at grocery stores and prepared by the tourist (Vail et al., 1998: 43; LD, 1998: 25). The senior author’s ‘participant observation’ in both places suggests that table service is more common in Maine restaurants, while cafeteria and buffet service is more common in Dalarna. Hypothesis 3 Labour laws and employment security measures tend to reinforce seasonal employment fluctuations in both places, but for different reasons. In Dalarna, seasonal workers’ eligibility for unemployment compensation increases the supply of workers willing to accept non-permanent jobs and enhances employers’ ability to hire and fire workers according to demand fluctuations. Working against that effect, several collective bargaining agreements exempt tourist businesses from the standard 40 hour workweek, and this tends to dampen seasonal fluctuation in jobs. Maine’s highly seasonal tourism labour demand is reinforced by the lack of layoff restrictions or collective bargaining agreements protecting job security. Hypothesis 4 Sweden’s high cost per hour of labour drives up tourist service prices – particularly for labour intensive services – putting Dalarna at a competitive


Current Issues in Tourism

disadvantage vis-à-vis lower-wage European destinations offering similar winter and summer attractions. In contrast, Maine’s low service sector wages and labour costs give it a competitive advantage vis-à-vis nearby north-eastern regions with higher labour costs and similar tourist attractions (e.g. Cape Cod, upstate New York). Corollary: The magnitudes of Maine’s cheap labour advantage and Dalarna’s disadvantage are not clear and should not be exaggerated. Impacts depend on numerous factors, including labour’s share of total tourist trip costs, the accuracy of tourists’ information about prices at competing destinations, and the ‘cross elasticity of demand’ for substitute destinations. Hypothesis 5 Maine’s cheap labour tourism tends to be self-perpetuating and contributes to socio-economic problems in some tourist-dependent communities. Jobs with low pay, few benefits, and no career prospects tend to attract less skilled and less motivated workers, reinforcing the low productivity/low wage syndrome. The social, cultural, and economic vitality of host communities may be sapped by a concentration of low paying, seasonal jobs that put local people in a quasi-servant relationship with high income visitors (Vail et al., 1998; Vail & Hillard, 1997). Hypothesis 6 Both Dalarna and Maine could achieve their tourism development goals more effectively if their labour market institutions converged toward a ‘middle way,’ reducing the cost of semi-skilled labour in Dalarna and raising wages and benefits in Maine. Corollary 1: Sweden’s 1995 law allowing year-round averaging of the 40 hour workweek is a step in the right direction, reducing employers’administrative costs and increasing the number of year-round tourism jobs. Corollary 2: A state strategy to make ‘livable wages’ more widespread in Maine tourism would have three core components: more extensive training for front line tourism employees; measures to encourage the spread of ‘high performance’workplace practices; and strengthening of labour market policies, including the minimum wage, the earned income tax credit, and unemployment insurance coverage for qualifiying seasonal workers (Vail & Kavanaugh, 2000).

Sales and Excise Tax Wedges: Effects on Costs, Competitiveness and Incentive Sales and excise taxes affect resource allocation and consumption, either for better or for worse, by widening the wedge between production costs and retail prices. The economic impact of such taxes depends on several factors, including the absolute tax rate, relative taxes on different goods and services, the producers’ market power, and consumers’ price responsiveness. Consumers’ responsiveness to a commodity tax is reflected in demand elasticity, which is influenced by the strength of preferences for the taxed good, prices of substitute goods (substitution effect), the proportion of budgets spent on the good (income effect),

Institutions Determining Tourism’s Scale and Structure


and the length of time they have to adjust consumption in response to tax-induced price changes. Tax wedges and the tourist activity mix Many tourist ‘products’ can be characterised as composite goods that combine purchased goods and services with own-produced services. For example, driving to a nature reserve combines a car and fuel with the driver’s labour; and preparing a campsite meal combines store-bought food, utensils, stove, etc. with the cook’s labour. In these and many other activities, tourists face choices: commercial transportation versus self-chauffeuring, hotel versus tent, restaurant versus home cooked meal. Henrekson (1998) raises the question: what causes a certain activity to be performed in the taxed market sector, in the untaxed ‘black market’, or as self-production? Broadly, the choice can be viewed as tourists optimising subject to a budget constraint that includes both their time and money. Preferences and opportunity costs regarding leisure, wage earning work, and unpaid work all influence the outcome (Becker, 1968). Even though tourists may not carry out an explicit calculus of monetary and opportunity costs, taxes can influence decisions whether to purchase or selfproduce services in two ways: directly via the value added tax (VAT) and indirectly via employers’ ‘pass-through’ of mandatory contributions. In particular, even if commercial production of services is more efficient than tourists’ self-production, the price inflating effect of these levies may induce self-production and depress tourism expenditures and employment. Påhlson (1997) has modelled the necessary conditions for a service to be bought on the market, when labour accounts for the major part of total costs. The consumer optimises by purchasing a service when: Tourist’s pre-tax wage Producer’s productivity 1 [ Producer’s pre-tax wage ] - [Tourist’s productivity in self-production ] > [ 1 – percentage tax wage ]

Intuitively, the incentive to purchase a tourist service rather than produce it for oneself is stronger: (1) The higher the tourist’s wage rate relative to the producer’s wage; (2) The lower the consumer’s productivity relative to the commercial producer; and (3) The smaller the overall tax wedge. In Sweden the ratio 1/(1 minus %-tax wedge) lies in the range of 2.7 to 4.1, while in USA the ratio is between 1.4 and 1.9. In other words, there is a much stronger incentive for self-production in Swedish tourism. In this rational actor framework, if productivity in commercial and own production were the same, a tourist’s pre-tax wage rate would have to be at least 2.7 times greater than wages paid by the commercial producer to justify purchasing the service. Given Sweden’s compressed wage structure, that is not the case for many tourists and services. (Because of lower US taxes, the American wage gap need only be 40% to warrant purchase of commercial tourist services; the gap between the average US wage and the typical tourism wages is much greater than in Sweden.) Conversely, if the tourist and producer had the same wages, commercial labour productivity would have to be at least 2.7 times greater than in self-production to justify purchasing a service (Henrekson, 1998).

Current Issues in Tourism


Sweden’s 25% value added tax and tourism In July 1990, Sweden’s value added tax (VAT) was raised from 22% to 25% and Parliament eliminated special lower rates on parts of the travel and tourism industry. Thus the tax on restaurants, previously half the general rate, jumped to 25% and VAT was extended to previously exempted services including commercial transportation, commercial camp sites, ski lifts, and amusement parks. Apparently little consideration was given to likely competitive effects.11 Table 6 summarises Sweden’s current VAT rates for core tourist services; they are substantially higher than the EU average. Compared to the other Nordic countries, with which Sweden competes most directly for foreign tourists, only Denmark’s VATs approach Sweden’s high levels on most tourism services. A state-commissionedstudy showed that in the first two years of the new VAT rates, travel volume, tourism sector profitability, and employment all declined, while bankruptcies increased (NUTEK, 1993: 24). Hultkrantz (1995a) estimates that the VAT increases, combined with other tax revisions, reduced Swedes’ internal recreational travel spending by at least 15%.12 The tax increase also had significant substitution effects, with more Swedes vacationing at their own cottages, visiting friends and relatives, and travelling abroad. The higher VAT on commercial lodgings, restaurants, and other marketed tourism services raised their retail prices by 15% to 25%, while the cost of visiting privately owned leisure homes was unchanged. Although it is difficult to separate the effects of a higher VAT from those of a contemporaneous national economic recession and exchange rate fluctuations, the evidence strongly suggests that domestic and foreign tourists are both price-responsive. The VAT increase led to fewer foreign visitors and more Swedes travelling abroad (TD, 1996: 14; Sahlberg, 1996). There have been no studies of the tax reform’s specific effects on Dalarna tourism, but the overall impact must have been the same as for Sweden as a whole. In sum, the higher VAT worked against commercial tourism by changing tourists’ incentives in two key ways. First, the higher total cost of a Sweden vacation strengthened residents’ and foreigners’ incentive to travel elsewhere. Second, the tax increase strengthened incentives to substitute untaxed self-produced services for taxed commercial lodging, meals and entertainment. Data are not available to estimate demand elasticities for distinct tourism packTable 6 Value added tax rates: Sweden and the EU Activity General VAT Restaurants Car rentals Ski lift tickets Amusement parks Hotels, hostels, campsites Commercial transport Theatres Museums Source: TD, 1996:18, 19

Sweden (%) 25 25 25 25 25 12 12 6 0

EU average (%) 19.3 14.6 18.7 n.a. 11.9 9.9 8.7 9.1 4.5

Institutions Determining Tourism’s Scale and Structure


ages (e.g. alpine ski holidays, heritage tourism, and fishing) or to estimate the cross-elasticity of demand for commercial vs. self-produced services. National time series data show the expected inverse price–quantity relationship in foreigners’ recreational travel to Sweden; however, microeconomic analysis reveals a complex pattern of price responses. For example, Danes and Germans visiting Sweden to enjoy nature are not very price sensitive, in contrast to Norwegians and Finns, who arguably can opt for close substitutes at home (Nordström, 1999). In the most intensely competitive parts of Dalarna’s tourism industry, for instance lodgings in the Siljan region, both tourist numbers and the after-tax price received by producers are depressed by high taxes. Lower profit margins and employment in the short run and lower investment over the long term are plausible consequences of depressed operating revenues. While Sweden’s high VAT may depress after-tax income and employment in Dalarna, the added tax revenue from foreign tourists is a benefit from the national perspective. Particularly if foreign demand for the internal components of Swedish holidays were less than unitary elastic, the higher prices stemming from Sweden’s high VAT would increase national tourist earnings. This may well be the case for Dalarna, where unique natural and cultural features make it a special tourist destination. Similar issues are under discussion in Australia, which is considering alternative tourism development strategies (Clarke, 1997). Vehicle fuel taxes Sweden has among the world’s highest excise taxes and retail prices for personal vehicle fuel. The current tax of SEK 5.54/litre ($0.74) represents 59% of the February 2000 pump price for low octane gasoline.13 High fuel taxes and a low 12% VAT on commercial transportation probably induces some tourists to opt for commercial transportation and reduces total travel to some remote northwestern Dalarna destinations. These effects are not likely to be large, however, since fuel cost is just a small fraction of the total holiday trip price. The fact that 90% of Dalarna’s visitors arrive by car, despite excellent train and bus service, reflects a strong preference for the flexibility and convenience of automobile travel (Nerhagen, 1997: 7). Maine’s low taxes on tourism services The United States has no general sales or value added tax, but roughly 30% of Maine’s $3 billion/year in state and local revenue comes from a 5.5% retail sales tax. Like many other states, Maine imposes a higher tax (currently 7%) on meals at expensive restaurants and commercial lodgings, which cater largely to tourists. The 10% car rental tax affects primarily tourists and business travellers. In contrast with these tax surcharges, amusement and recreation services are completely exempt from sales tax, as are basic groceries. Maine’s sales tax is in line with all but one of the neighbouring New England states and lower than nearby Canadian provinces.14 In sum, compared to Sweden, Maine’s sales tax adds relatively little to the total cost of a tourist visit and presumably has less influence on the incentive to self-produce services. It is estimated that tourist spending directly generated $206 million in state tax revenues in 1996, with $118 million coming from sales and excise taxes and $27


Current Issues in Tourism

million from the gasoline tax (Longwoods Associates, 1997: 25). Systematic records are not kept on the other side of the fiscal ledger: the state and local government spending required to support the tourist economy (e.g. road maintenance, water supply systems, waste disposal, police, fire and hospital services). Maine municipalities cannot levy their own sales taxes to capture part of the tourist revenue stream or cover the cost of public services that underwrite commercial tourism. A previous study indicated that Maine could (and should) raise an additional $26 million in sales taxes targeted to tourist spending, on the principle that visitors should pay for the public services they require and for the environmental wear and tear they cause. That goal could be accomplished by raising the restaurant and lodging taxes another 1%, adding 1% to the sales tax on recreational equipment, and extending the sales tax to amusements and recreational services. Tourist demand is predicted to be highly inelastic over such small retail price increases, thus tourist spending would be only minimally affected and the tax incidence would be borne primarily by visitors (Vail et al., 1998: 74–5). At present, low octane (‘regular’) gasoline is priced at about $1.50/gallon in Maine (SEK 2.1/litre), or about one-fourth of the Swedish price. Perhaps this enormous differential reflects in part the northeast US’ more intense retail competition and lower refining and fuel transport costs, compared to Sweden. However, the big difference is that state and federal excise taxes – $0.37/gallon (SEK 0.64/litre) – are only about one-tenth of Sweden’s. Recent proposals for significant federal and state gasoline tax increases – whether to generate revenue, induce energy conservation, improve ambient air quality, or cut greenhouse gas emissions – have met with intense resistance from motorists as well as the powerful petroleum-automobile-highway construction lobby. The regressivity of the gasoline tax and Americans’ almost total dependence on cars, trucks, vans, and sport utility vehicles for commuting, shopping and leisure activities make cheap gasoline a ‘knee jerk’ populist issue. Despite several legislative attempts, Maine’s gasoline tax has not been increased since 1991. Late in 1998, following his landslide re-election, Maine’s governor proposed a modest $0.05/gallon (SEK 0.10/litre) tax increase, with the revenues dedicated to a backlog of urgent road and bridge repair needs. Despite the minimal increase and the widely recognised need for better road maintenance, the legislature enacted only a $0.03 increase. Hypotheses regarding the impact of sales and excise taxes on tourism Hypothesis 1 Sweden’s high tax on most tourism services has a significant depressing effect on the number of foreign and Swedish visitors to Dalarna as well as their contribution to the region’s economy. In general, consumer demand seems to be elastic over price increase caused by the tax; however, demand effects vary for different activities and tourist groups. The decline in tourist numbers is mitigated somewhat by more tourists who substitute cheaper self-produced services for more expensive taxed services. Maine’s much lower retail sales taxes add comparatively little to total trip costs and are in line with competing states; thus they have little effect on tourist numbers, total expenditures, or the composition of spending.

Institutions Determining Tourism’s Scale and Structure


Corollary: Maine could incrementally increase sales taxes on lodging, amusements and recreational equipment purchases without significant adverse effects on tourist numbers or total spending. In other words, consumer demand is inelastic over small increases in total trip cost. The added tax revenues would enable the state to cover public expenditures for tourist infrastructures and services and help pay for protecting prime natural attractions. Hypothesis 2 Dalarna’s high VAT on most tourist services, combined with high labour costs, encourages tourists to self-produce services instead of purchasing labour-intensive commercial services. The latter would contribute substantially more to regional production and employment. Corollary: Sweden’s high VAT strengthens the incentive of tourists and businesses to engage in untaxed black market transactions for services like cottage and campsite rentals and sporting guides. Hypothesis 3 Sweden’s low VAT on theatres (6%) and museums (0%) and Maine’s sales tax exemption on ‘amusements and recreation’ have the beneficial effect of encouraging tourist spending on these value-added and culturally enriching activities. No more than 10% to 14% of total tourist spending is currently channelled to such activities, so these low tax rates help boost demand in an under-developed – and relatively labour-intensive – part of the tourism economy (See Appendix 1). Hypothesis 4 Maine’s comparatively low taxed and cheap gasoline encourages more day and weekend visits, focusing on activities like ‘sun and surf’, outlet mall shopping, skiing, and fall ‘leaf peeping’. At a time when leisure travel is increasingly compressed into short holidays, this may give a destination like Maine a significant competitive advantage in its market. Corollary: The extreme car-dependence of tourism in both places raises sustainable development questions: it intensifies peak period traffic congestion and increases ambient air pollution and greenhouse gas emissions per visitor.

Land Ownership and Property Rights: Recreational Access to Private and Public Land The forests, mountains, lakes and rivers of Dalarna and Maine are prime tourist attractions. According to one set of estimates, nature-based tourism in and around Maine’s 4 million hectare North Woods accounts for nearly $1 billion/year in gross tourist spending (including dining, lodging, shopping and transportation) (MAS, 1996). Even for tourists attracted primarily by cultural events, shopping, or just relaxing, beautiful natural settings comprise the backdrop for Dalarna and Maine vacations. In economic terms, recreational landscapes are best understood as impure public goods with limited exclusion possibilities and a degree of rivalness in use. Dalarna and Maine are similar in that most land is privately owned (Dalarna 75%; Maine 94%) and is working forest, not only a recreational resource. Outdoor recreation in a multi-function ecosystem coexists with and


Current Issues in Tourism

affects other recreation, non-recreational land uses, and basic ecosystem functions. For example, snowmobiling, a popular winter-spring activity in Dalarna’s mountains, affects cross-country skiing, commercial forest management, Sami reindeer herding, and critical habitat for fish, fowl and mammals. Within these similarities, institutional differences influence nature tourism’s economic potential and environmental sustainability. In Dalarna the right of common access to private land (allemansrätten) and small fragmented land holdings are key contextual features. In Maine’s North Woods there is a a prevalence of large industrial land ownerships combined with a tradition of ‘permissive access’. This section explores how these and other property institutions affect landowners’ and tourists’ incentives and create pressure on ecosystem carrying capacity. Open access, fragmented ownership, and development of nature tourism in Dalarna Most of Dalarna’s undeveloped land is held in 24,000 private ownership’s, although a national park and almost 100 nature reserves cover 8% of forest land. Some of the latter parcels, especially in the mountains, are important for tourism. For instance, one nature reserve embraces Sweden’s highest waterfall, Njupeskär. There are also three large industrial forest owners which together hold almost 50% of Dalarna’s land area. Municipalities, the Church of Sweden and the Dalarna County Board are other minor landowners (SCB, 1999). Fragmented ownership characterises the county’s more densely settled regions. The Lake Siljan region has the most fragmented land holdings in Europe. It was one of few areas exempted from Sweden’s major land partitioning in late 18th and early 19th centuries. Narrow strip holdings, a metre or two wide but several kilometres long, are common (Westholm, 1992). In the past, this was a problem primarily for forestry operations, but recently it has also emerged as an obstacle to commercial tourism development. A major recreation park was delayed several years because so many landowners were involved; and the combination of fragmentation and open access has impeded commercial beach and marina development. In the mountains, property rights and land use controversies surround the expansion of alpine ski areas, use of motorised recreational vehicles, Sami people’s usufruct rights, and environmental damage in fragile alpine zones subject to the combined impact of recreation, industrial forestry and reindeer herding. The problematic side of allemansrätten – Sweden’s right of common access As a matter of principle, access to outdoor recreational opportunities is open and free in Sweden. The right of common access rests on the implicit but incorrect presumption that recreational landscapes are pure public goods. Allemansrätten is not explicitly defined in law;15 rather, it is a set of customary rules and judicial decisions governing activities not mentioned in laws articulating the state’s obligation to defend landowner claims. These use rights have been preserved from archaic law, so in a sense current rights still reflect relative resource scarcities and prices from the 13th century. Over the centuries, some valuable species have become scarcer (e.g. chanterelle mushrooms and cloudberries), yet they have

Institutions Determining Tourism’s Scale and Structure


remained free to everyone, as do distinctly modern attractions like white water streams and snowmobile trails. Allemansrätten has roots in a custom allowing free travel in roadless country, including the right to stay overnight and collect natural products to eat and drink. Today, this custom allows recreational trespass and primitive camping on private land. However, allemansrätten also entails an ethical responsibility to avoid harm to nature or economic damage to the owner, for instance to crops and timber stands (Colby, 1988). Allemansrätten has nearly universal public support and has been used to rationalise the extension of rights well beyond its traditional domain, particularly in sport fishing and snowmobiling. This includes situations where open access conflicts with efficient resource use, rights of aboriginal Sami people, and interests of other recreationists. A national recreation policy based on open access is not fully compatible with a rural development strategy based on promoting commercial tourism. In sum, compared to Maine, Dalarna’s nature tourism is growing in conditions less friendly to commercial exploitation of nature. Alpine skiing is the key exception for complex reasons, but especially because a natural force – gravity – makes it easy to exclude free riders by charging for lift tickets (Vail & Hultkrantz, 2000). Swedish citizens, foreign tourists, and even commercial tourist businesses have the right to use private property for recreation without paying user fees. Nonetheless, this right falls short of true ‘open access’, since certain activities are prohibited (e.g. hunting, fishing, felling trees, disturbing wildlife habitat) as is trespassing on certain lands (e.g. cropped fields, homesteads). Privately owned landscapes thus resemble, but are not identical to, open access common pool resources. The following examples are stylised (deliberately one-sided) to highlight problems that arise in this institutional setting. The lack of mechanisms to exclude recreational users or secure payment for nature’s – and landowners’– services has caused two familiar common pool resource problems: short-term congestion and long-term resource depletion. Snowmobiling in Dalarna’s mountain region is a prime example. At the Easter weekend peak, snowmobiles may congest popular trails to the point of causing traffic slowdowns, increased accidents, and degraded air quality. These ‘reciprocal externalities’ reduce amenities to all snowmobilers. Snowmobile traffic is also experienced as a ‘unidirectional externality’ by other recreationists, particularly cross-country skiers. Unregulated snowmobiling may impede non-recreational activities, such as Sami reindeer herding and logging operations. In fragile environments, unregulated snowmobiling may also have detrimental long-term effects, for example, disrupting bird and mammal breeding, depleting fish stocks (via ice fishing), and destroying young tree seedlings (Anttila, 1999; Hultkrantz, 1995b). The development of instruments for pricing recreational land use must be compatible with allemansrätten but take advantage of its limitations. First, free access applies only to individual recreationists; second, as mentioned, fishing and hunting rights are not included. In principle, allemansrätten does not apply to organised commercial recreation, which should open the possibility of charging for some land uses. To date, contractual landowner compensation has only occurred in a few circumstances, such as military manoeuvres and large-scale sporting events. The National Farmers’ Association (LRF), which represents


Current Issues in Tourism

thousands of small Dalarna landowners, has pressed the issue by initiating litigation to win compensation for organised tourist activities on members’ land. The courts have thus far gone against landowners, ruling that tourists should have free access as individuals even when they participate in commercial activities like guided backpacking and mushroom-picking tours. The open access tradition, backed by the court’s ruling, weakens owners’ incentives to invest in land stewardship and some types of value-added capital: they cannot recoup the expense of efforts like trail maintenance and river bank erosion control. Even where fee collection is legal, for instance at picnic sites and canoe put-ins, there are practical difficulties in recouping investments, such as monitoring and fee collection costs. (Note that these particular costs would not disappear with public land ownership.) In short, tourists’ ability to free ride on landowners’ investments mis-aligns incentives in ways that are bad for both the environment and the local economy. When owners’ and users’ incentives conflict with their sense of moral obligation, it is unwise to rely exclusively on ‘voluntarism,’ as allemansrätten does. Ski lift tickets and fishing licences are two means currently used to align landowners’ and tourists’ incentives. Guided tours with a fixed ‘package price’, for instance fishing and cross-country skiing tours, are another potential way to extract economic rent for use of nature’s services. Scale misalignment between small landholdings and large efficient management units Fragmentation makes it difficult for landowners to capture potential economies of scale in tourism investments. Three examples are: (1) economies in building and maintaining physical infrastructures, such as road and cross-country ski trail networks that cross property lines; (2) development of tourism destination ‘packages’ that combine parking areas, cabins, cafés, and boat launches on different properties; and (3) lower transaction costs for promotion, fee collection, and monitoring. These latent opportunities are an incentive for collective action by landowners, an approach with many precedents in rural Sweden. Fish conservation districts (fiskevårdsområden), are an example, whereby riparian owners sell fishing licences to control access and capture rent. On a much larger scale, investors in the Sälen and Idre ski resorts coordinate promotion, local transportation, fee setting, and social events. And the State Forest Management Company (Assi Domän), although a single large owner, follows an integrated back country recreation management strategy. In sum, collective action has the potential to be a ‘win-win-win’ game, where landowners, the local economy, and the environment all benefit. Joint resource management institutions face internal governance problems, however, and success is not guaranteed. Idre/Grövelsjön: a mini-case study To highlight the problems and potential associated with property rights, scale economies, and rent-capturing initiatives, we sketch the evolution of outdoor tourism in Idre/Grövelsjön, in Dalarna’s far northwestern corner (see map Figure 1, Appendix 2). More than a century ago, Arthur Hazelius set the stage for Dalarna tourism by founding Stockholm’s Skansen, the world’s first ‘outdoor-life’-museum, which celebrated Dalecarlian folk culture. The fledgling

Institutions Determining Tourism’s Scale and Structure


Swedish Tourism Association (STF) helped spread knowledge of Sweden’s varied regional landscapes and local cultures. In 1907 Dalarna’s first year-round hotel with its own ‘touristlift’ was advertised. By 1937 the ‘Högfjällshotel’ in Sälen and STF’s mountain station in Grövelsjön were in operation. Grövelsjö tourism was and remains dominated by small family owned hotels with XC skiers and hikers as the main guests. But nearby Idre municipality has developed a large alpine skiing resort presided over by the Idre Fjäll (IF) corporation as an outgrowth of the late-1970s alpine skiing boom. It owns or controls a large area around the main mountain, Grängesvålen, which hosts more than 30 ski runs, 60 km of XC tracks and hundreds of rental cottages (Idre Fjäll, 1999). By controlling core economic facilities over a sizable area, IF can employ both direct and indirect means to appropriate land rents. Compared to most other nature-based recreation, alpine skiing has a more straightforward way to charge users via lift tickets. Land rents can also be appropriated indirectly through oligopoly pricing of the complementary services and accommodations that make up a ski holiday package. Apart from alpine skiing areas, however, there is not much scope for exercising such market power in Dalarna. For instance, the lodging sector in most of the county is highly competitive and suffers from excess capacity. XC-skiing, like most nature tourism, lacks a direct mechanism to charge for use of nature or for costs borne by landowners. Under allemansrätten, fees cannot normally be charged for entrance to undeveloped areas, whether private or public. Thus a landowner’s costs for laying out and maintaining a ski track or thinning timber stands to improve aesthetic quality can only be captured via the price of lodging or a complementary service like ski equipment rental and ski instruction, which may not be supplied by the landowner. Introducing user fees might seem an easy way to overcome the problem, but there are several limits to fee collection. Permission to fence a recreational area and charge fees requires that the owner ‘develop’ the area by constructing facilities that seriously disturb natural features, as in the case of a golf course or recreation park. Moreover, commercial facilities such as campsites, fishing and bathing piers, toilets and parking lots, cannot be provided exclusively to one group of users, such as a tour operator’s paying clients (Vail & Hultkrantz, 2000). In Grövelsjön, with its family hotels, rental cottages, and STF mountain station, family traditions are strong and tourist operations are small scale compared to Idre Fjäll. Winter season is dominated by XC-skiers seeking silence along mountain tracks in the semi-wilderness surrounding Lake Grövlan. Snowmobilers also drive to mountain lakes for ice fishing, and a few alpine skiers use the two ‘single-lift’ establishments. This small business profile has made achievement of scale economies in service provision difficult, although there is cooperation in marketing and promotion. Land holdings are not as small as around Lake Siljan, but fragmented ownership problems are present. To overcome open access problems and achieve scale economies in building and maintaining the XC ski network, tourist business operators and surrounding private cottage owners have formed a ‘track making’ association. The businesses make and groom the trails while cottage owners pay a yearly fee to the association and benefit by being able to charge vacationers higher rents. Without the trail network, there would be few winter tourists. The rights issue of a commercial


Current Issues in Tourism

track crossing numerous private holdings is solved by simple contracts with landowners. The association has just a small budget to compensate them, and what makes the trail network economically viable is its indirect value in attracting tourists and their spending to the area. Voices have been raised for introduction of a direct ‘ski fee’ that would include a contribution to local nature conservation efforts (Westergård, 1999). In contrast to Grövelsjön, Idre Fjäll controls the land in its XC-ski network and faces no multi-owner conflicts. It charges SEK 50 ($6.60) per day for use of its trails; for visitors staying in IF lodgings, the fee is included in the room or cottage fee. IF is concerned about the free-rider problem of non-paying visitors, mostly local residents, using the illuminated and well groomed tracks; but for cultural and aesthetic reasons it has chosen not to fence the trails or gate the entrance (Idre Fjäll, 1999). Residents’ rights, land ethics, and farsighted stewardship Dalarna’s ambition to expand nature tourism raises an equity issue: the restriction of local residents’ customary land use entitlements. More visiting hunters, fishers, campers and snowmobilers adversely affect their well-being. Traditionally, the rural landscape was both their playground for active and contemplative recreation and a source of subsistence via hunting and gathering. The right of common access does not protect residents’ quality of recreational experience or compensate for amenity losses when they are ‘crowded out’ by tourists. The Swedish Tourist Authority foresees a related danger that visitors, especially foreigners, will lack environmental understanding and a sense of responsibility for nature protection. In short, tourism growth may erode the land ethic that made allemansrätten ‘work’ in the past (TD, 1998b: 44). Finally, nature tourism promotion is not backed by a comprehensive natural resource inventory, local carrying capacity estimates, or a long-term resource management plan. This institutional deficiency has already had conspicuous effects such as fish stock depletion in high altitude lakes, degradation of thin mountain valley soils, and disruption of wildlife breeding. (Unregulated snowmobiling is implicated in all of these.) An overarching analysis of ecosystem carrying capacities and a land use management plan would help to resolve the two property rights conflicts that have generated most heat in Dalarna: proposed expansion of the Sälen-Idre ski area into the Städjan National Park, and the Sami people’s rights to herd reindeer, hunt, fish and harvest timber. A more thorough analysis and forecast of the demand for Dalarna’s various natural attractions would also facilitate more effective resource management and tourism promotion. Commercialising outdoor recreation in Maine’s North Woods Maine’s 94% private land ownership ranks it near the top among US states. Over 90% of surface area is forested and most is industrial forest. Much outdoor recreation makes use of private land, as Table 7 reveals. In addition to the listed activities, almost all hunting and alpine skiing take place on private land. Public lands, though not extensive, are also important for tourism and recreation. Federal lands, Acadia National Park and a section of the White Mountains National Forest, host five million yearly visitors. The 150 km long Allagash Wilderness Waterway is the only federally designated wilderness east of the

Institutions Determining Tourism’s Scale and Structure


Table 7 Proportion of Maine’s recreational activity on private land (1992) White water rafting Trail biking Camp sites Snowmobiling Hiking XC skiing

100% (put in and take out sites) 98% 98% (many maintained by Maine Forest Service) 67% 60% 59%

Source: Irland, 1993: 10

Mississippi River. There are also 61 parcels of state lands, and thousands of lakes and ponds are held in public trust by the state. The jewel in the crown is the 80,000 hectare Baxter State Park in the heart of the North Woods, with 1600 metre Mt Katahdin marking the northern end of the 3000 km Appalachian Trail (see map Figure 2, Appendix 2). Many of Maine’s 464 towns and cities also have public parks and commons, beaches and boat landings. Finally, a small but growing area of ecologically and recreationally valuable land is being acquired by land trusts and conservation organisations. Almost all of the 4 million hectare North Woods are held by about 15 large owners. By custom, though not in law, there has been ‘free and easy’ recreational access to most of this land going back to H. D. Thoreau’s first visit in the 1840s. In this century, paper corporations with familiar names like International Paper and Boise Cascade have been the dominant owners and a powerful presence in Maine’s economy and politics. However, in recent years the number and composition of forest land owners have been in flux, with paper corporations consolidating, selling, or developing timberland as part of worldwide asset management strategies. Recreational land use is legally governed by a system of ‘permissive access’, whereby owners can prohibit access or charge fees. Prospective users have a nominal obligation to seek permission. In practice, however, hunters, fishers, hikers and other recreationists have faced few restrictions or fees. As late as 1982, only 8% of Maine’s undeveloped land was posted against trespassing (Irland, 1996: 61). Recreational users have extensive rights to cross forested or unfenced land to reach lakes and large ponds. Rivers and ‘floatable’ streams are also open to the public, though there is no general right of trespass to reach them. Trees and other vegetation belong exclusively to the owner, while fish and game are a ‘public trust’, requiring only a state licence (Hasbrouke, 1988). These property rights obviously differ in important ways from Sweden’s allemansrätten. Over time, recreationists and sporting organisations came to view access as a de facto right. This is reflected in the fact that the Inland Fisheries and Wildlife Department (IFW) has a ‘landowner relations coordinator’ who mediates owner-user relations and educates sportsmen about their responsibilities: ‘Everyone MUST respect all landowners and their rights . . . All landowner wishes have to be followed by all outdoor recreational participants to help ensure access to private property in the years to come’ (IFW, 1998b). State agencies promote a ‘public–private partnership’ in many ways. For example, IFW facilitates contractual negotiations between landowners and snowmobile clubs over trail layout and maintenance; the Maine Forest Service maintains nearly 100 campsites on


Current Issues in Tourism

private land; the Department of Conservation coordinates dam releases for whitewater rafting; and a 1978 law strictly limits the liability of landowners who allow free access (ALT, 1996; Hasbrouke, 1988; IFW, 1998a; Irland, 1993).

Contested terrain For most of this century, permissive access enabled recreationists to use vast stretches of undeveloped private land for hunting, fishing, canoeing, etc. In general, recreational use levels were compatible with the forest owners’ timber management goals. In fact, free access helped them maintain the goodwill of politically influential groups at little opportunity cost. Over the past quarter century, however, a complex set of changes has induced many owners to restrict access, charge user fees, develop prime shore front real estate, and liquidate forest holdings. These trends accelerated in the latter 1990s, leaving the future of nature-based tourism in doubt. The major factors changing landowners’ costs and incentives can be catalogued as follows:

… expanded timber harvesting, mechanised logging, and herbicide spraying brought industrial forest management into greater conflict with recreational uses; … construction of nearly 40,000 km of logging roads brought once remote areas within reach of motorised vehicles; … noisier and more intrusive recreational vehicles (sport utility vehicles, all-terrain vehicles, snowmobiles, jet skis, and even light planes) became common; … tourism promotions informed a broader public about the North Woods’ recreational ‘gems’ and brought visitors who lacked the traditional users’ land ethic; … a nationwide shift toward concentrated day, weekend and holiday trips intensified peak congestion pressures; … wear and tear on infrastructures, waste disposal costs, fire hazards, and vandalism increased; … the market value of undeveloped land for seasonal homes escalated, especially on lakes and ponds and near alpine ski areas, reflecting consumers’ growing wealth and easier access. The paper corporations’ strategic perspective on forest holdings has also evolved. Forest land formerly viewed as a raw material ‘mine’, an illiquid asset, and a cost centre is increasingly considered a multiple use resource, a liquid asset, and a profit centre. This shift is part of the industry’s globalisation and restructuring, driven by shareholder demands for higher, faster returns on investment (Irland, 1996: 41). Together, these changes have brought about four landowner and government reactions that are sketched here: gating the North Woods and collective management by large landowners; cumulative, irreversible real estate development; reactive land use regulation; and public acquisition of land. (For a more thorough discussion, see Vail & Hultkrantz, 2000.)

Gating and cooperative management Commercialised recreation has a long history in Maine’s North Woods, with alpine ski areas in existence for decades and commercial sporting camps for

Institutions Determining Tourism’s Scale and Structure


more than a century. Since the 1970s, forest owners have also contracted with white water rafting outfitters for base camps, take-out places, and water releases from company-owned dams. In 1998, 81,000 tourists took commercial rafting trips (AP, 1998). Nonetheless, the past two decades mark the first period when large owners have gated private roads into their holdings, excluded some users (e.g. all terrain vehicles), and set up check points to collect fees from users. A positive result of gating is that it increases landowners’ ability to control congestion pressures at peak sites and times and reduce conflicts between recreation and timber management. New revenue earning opportunities also strengthen owners’ incentives to practise land stewardship and invest in value-added facilities, such as campsites, cabins and boat launches. On the other hand, the equity of gating and fee collection can be questioned, since traditional user groups have lost their free access entitlement. In 1971, 24 owners of 1.2 million hectares, mostly remote wildlands in northwest Maine, formed a non-profit corporation, North Maine Woods, Inc. (NMW), to manage recreational activity. NMW collects day fees and campsite fees from roughly 200,000 yearly users and manages nearly 1000 campsites as well as parking lots, boat launches, and waste disposal systems. Landowner cooperation includes liability insurance, road maintenance, fish stocking, legislative lobbying, and contracting with sporting organisations. The latter activity has features of a repeated cooperative game. For instance, NMW contracts with the Maine Snowmobilers Association over the layout and maintenance of a 100 km section of a major US–Canada trail network. Snowmobile club members have assured access to trails, campfire sites, and ice fishing ponds for a modest annual fee of $25. In return, the clubs maintain trails and enforce basic ‘rules of the road’, especially keeping to marked trails, avoiding logging operations, and preventing drunk driving (Cowperthwaite, 1999; Fisk, 1999; Irland, 1993: 8–9). The IFW facilitates contract negotiations and allocates a share of snowmobile license fees to trail maintenance. Through joint land use management, NMW’s members achieve economies of scale with infrastructures and cut the transaction costs of promotion, gating, monitoring, and fee collection. Contracts with sporting groups appear to be a ‘win-win’ arrangement, with users gaining stable, low cost access and the owners benefiting from the users’ self-policing. Government agencies, trusted by both sides, play an important brokering role. Cumulative and irreversible real estate development Maine’s North Woods are in the midst of a real estate boom that combines two processes: accelerated turnover of paper corporation land and expanded seasonal home construction. The Land Use Regulation Commission (LURC), which regulates land use in Maine’s ‘unorganised territories’, believes that, ‘Cumulatively, second home development may have the greatest impact on natural and recreational resources in the region’ (LURC, 1997: 67–8). The extent of home building is not known precisely, but a government official estimated in the early 1990s that 1000 cabins had been built ‘in recent years’. Hundreds of existing structures are being enlarged and winterised (Irland, 1993). This is most intense in the west-central mountain and lake region, near ski resorts and on lake fronts that are within an easy day’s drive of metropolitan Boston. Lease fees for


Current Issues in Tourism

many corporate-owned cabin sites have increased by 1000% in recent years, reflecting soaring consumer wealth. Several industrial landowners are ‘cashing in’ by selling off formerly leased cabins (Fisk, 1999). Each hectare of wildlands converted to leisure home lots has a ‘shadow conversion’ effect on several more hectares, as access roads and drainage ditches are built, electricity lines are installed, ‘NO TRESPASSING’ signs appear, roads are gated, and game corridors are disrupted. The solitude of an entire lake can be destroyed by one jet ski speeding across its surface and one brightly lighted home on its shore (Irland, 1993). For decades, paper corporations have bought and sold forest land, but the process has accelerated and a new type of industrial landowner has appeared. A noteworthy sequence began in 1994, when Scott Paper sold its Maine paper mills and 400,000 hectares of forest, including parcels of tremendous scenic, recreational and ecological importance, to SAPPI (South African Pulp and Paper). In 1998, SAPPI re-sold the lands to Plum Creek Timber, a western US land development company with a reputation for ‘liquidation’ clearcutting and speculative land sub-division. Although Plum Creek has contracted to supply pulpwood to SAPPI’s mills, it has also created a ‘real estate development trust’ to oversee commercial land development. The same year, Bowater Inc. sold a million hectares – Maine’s largest single holding – containing prime recreational lands it had purchased just a few years earlier. About half of the parcel was sold to another out-of-state land management company. Finally, a few months later, Georgia Pacific sold 175,000 hectares to an anonymous institutional investor. All told, more than 30% of Maine’s North Woods and 15% of the state’s entire land area changed hands in 1998 and early 1999; still more land is on the market. The new owners are a mix of mill-owning corporations and real estate developers plus family and institutional investors with reputations for ‘patient capital’. However, the growing presence of land speculators and developers and the irreversible impacts of real estate development all bode ill for the future of back country nature tourism (Austin, 1997b). Inchoate political responses: Increased regulation and public ownership North Woods land use is nominally guided by the Land Use Regulation Commission’s 1997 Comprehensive Land Use Plan. However, LURC’s ability to prevent real estate developments that undermine sustainable tourism is limited by numerous factors, including landowners’ rights, Maine’s prevailing ideology, limited information, the economic interests of politically appointed commissioners, and pressures from various interest groups. It is thus not surprising that LURC’s responses to development pressure have not followed a ‘comprehensive plan’, but rather have been reactive and piecemeal. Nonetheless, the LURC staff’s critical appraisal of the situation has led to new growth management initiatives based on two principles: first, channel intensive seasonal home building and high volume tourism away from inappropriate areas and into ‘areas most appropriate for development’; second, apply more stringent zoning criteria to sub-divisions and major developments (LURC, 1997). LURC is in the process of identifying ‘Recreation Protection Subdistricts’ that ‘support or have opportunities for significant primitive recreational activities’ and adding zoning restrictions to prevent incompatible development. In

Institutions Determining Tourism’s Scale and Structure


principle, ‘All new development must meet the requirement that “provision be made for fitting the project harmoniously into the existing natural environment in order to assure that there will be no undue adverse effect on existing uses, scenic character and natural and historical resources”’ (LURC, 1997: 148). In practice, the commissioners face twin dilemmas: ‘Seasonal housing development is most likely to occur in areas with high recreational values’, and landowners’ rights often supersede zoning criteria. Recently, LURC resisted pressure from a powerful sportsmen’s organisation by rejecting a boat ramp on an undeveloped lake. It has also refused permission for a commercial sporting camp complex on several remote ponds (Austin, 1997a; Fisk, 1999). However, the Commission admits that its current standards contain few firm and specific guidelines, which strengthens land owners’ ability to avoid or appeal decisions they do not like (LURC, 1997: 148). In sum, LURC lacks the authority, facts, and personnel to manage North Woods development according to a true ‘comprehensive plan’. In the past, even when the pace of land conversion was slow, LURC’s actions tended to lag behind trends on the ground. In the future, the Commission’s capacity may be overwhelmed by accelerated land use conversion. The past decade has been marked by political struggles over property rights in the North Woods, though for most of the period forest management practices, rather than recreation and nature conservation,were the central issues. Ten years ago, Maine voters passed a referendum to invest $35 million in ‘Land for Maine’s Future’, but priority was given to protecting land under development pressure in more populous regions, especially on and near the coast. Only one sizeable North Woods tract (11,000 ha) was acquired. The momentum for additional public purchases was slowed for several years by a split between two conservationist factions, one advocating selective purchase of parcels with special scenic, recreational, biological and watershed protection values, and the other proposing a contiguous Maine Woods National Park of c. 1 million hectares. Both approaches are ambitious by any historical standard and would require federal funding to complement state bond revenues.16 Both are also opposed by right-wing property rights activists (the so-called ‘wise use’ movement) and viewed sceptically by paper corporations that fear a loss of working forest. Public anxiety was aroused by the massive 1998–99 land sales and the prospect of losing still more wildlands through gating, liquidation timber harvesting, and seasonal home construction. In November 1999, Maine voters approved a second Land for Maine’s Future bond, this time for $50 million, with good prospects for at least $25 million in local and federal matching funds. The state has already purchased a 200 metre wide ‘beauty strip’ from Plum Creek Timber, protecting over 100 km of prime lake and river frontage and widening the Appalachian Trail corridor. (The state has power of eminent domain to acquire private land at ‘fair value’, but in practice, all public purchases have been on a ‘willing seller-willing buyer’ basis.) Important acquisitions by two conservation groups actually preceded the state’s action. In the fall of 1998, The Nature Conservancy paid International Paper $35 million for 80,000 hectares of remote forest land on the US–Quebec border, including important wild rivers and several 1200 metre peaks. Recreational uses will be managed by North Woods, Inc. In March 1999, the New England Forestry Foundation contracted for a $30 million conservation


Current Issues in Tourism

easement that will prevent real estate development and sustain a mix of forestry and recreation on 300,000 hectares of land owned by the heirs of a 19th century ‘timber baron’. In the meantime, Maine’s Bureau of Parks and Lands has moved toward a more rational, multi-tier fee structure for the state parks: at peak seasons, fees are higher at several congested sites, and lower for Maine citizens, recognising their implicit property right. Part of the added revenues will be earmarked for natural resource protection and restoration (BPL, 1998). These actions are not anything like a coherent land use strategy for the North Woods. Maine is just in the early stages of designing a plan that would fulfil politicians’ rhetorical commitment to sustainable multiple use resource management. Hypotheses regarding property rights institutions Hypothesis 1 Dalarna’s combination of predominantly private land ownership with (quasi)open access maintains a cherished public right, but at the expense of weakening and mis-aligning the incentives facing landowners, tourists, and commercial recreation businesses. This situation detracts from both the preservation of tourism’s natural capital base and from developing tourism’s economic potential. In Maine, private land ownership combined with the spread of restricted public access creates stronger economic development incentives, but it also builds momentum for irreversible real estate development, concentrated in areas with high scenic and recreational value. Corollary 1: Sweden needs to rethink the rights and responsibilities associated with allemansrätten. Maine needs to rethink the high degree of landowner sovereignty and the determining role of market forces. Corollary 2: Both jurisdictions need more effective planning and management of competing land uses at the landscape and ecosystem scale. A key challenge is to develop and use estimates of ecosystem carrying capacity as a way to reduce short-term peaking pressures as well as long-term resource depletion, especially irreversible land use conversion. Public ownership of unique and fragile lands is part of a sustainable nature tourism strategy, but it is not ‘the’ solution. Hypothesis 2 Expanded commercial tourism may be a route to increased rural employment and economic growth, but it constricts recreational opportunities for local residents, representing an opportunity cost and an equity dilemma. Tourism growth may also generate socio-cultural conflicts between local residents and tourists (see Vail et al., 1998). Hypothesis 3 With an open access regime such as Dalarna’s, a shared ‘land ethic’ helps to counteract self-interested and shortsighted behaviour on the part of landowners and recreational users. However, under mounting peak demand pressures, voluntary action based on such an ethic will likely prove inadequate for environmentally sustainable nature tourism.

Institutions Determining Tourism’s Scale and Structure


Corollary 1: Land ethics are a form of cultural capital that may depreciate as tourism becomes more dependent on consumers lacking traditional users’ knowledge and sense of responsibility for ‘nature’. Corollary 2: Land stewardship, and especially the priority given to preserving non-use values, erodes when corporate managers are pressured to maximise rents or liquidate land assets by shareholders. Weak land use regulatory institutions reinforce this tendency. Hypothesis 4 Cooperative recreation management is an important tool for overcoming the scale mis-match between small landholdings and large effective management units. Landowner collaboration can achieve economies of scale, reduce transaction costs, and raise the return on investment in value-added tourism facilities. Corollary: Collective bargaining between landowners and recreational users, such as sporting organisations and commercial tour operators, can produce ‘win-win-win’ situations, enhancing land owners’ incentives and incomes, tourists’ quality of experience, and resource conservation. Governmental agencies can play an important role in facilitating such arrangements.

Research and Policy Implications Dalarna and Maine suffer from regional economic dualism. The economies of southeastern Dalarna and southern Maine have diversified and are thriving, but other regions – Dalarna’s Bergslagen and northwestern mountains and Maine’s northern interior and ‘downeast’ coast – face chronic economic distress, with high unemployment, out-migration of young people, and difficulty maintaining a base of public and commercial services. We agree with policy makers who believe tourism has both untapped capacity and demand-side potential to contribute to regional economic revitalisation and diversification – but with qualifications. As this paper’s introduction (and the following Appendices) stress, sustainable tourism growth is not a sure thing. Some prime attractions face carrying capacity limits, and there is intense national and international competition to attract tourists. Just as important, there are serious institutional impediments to sustainable tourism. In particular, we have hypothesised that core public policies, reinforced by long traditions and powerful interest groups, are ‘part of the problem’. In the Dalarna case, three national policies are most clearly implicated. First, the high cost of inexperienced and low-skill labour blocks employment creation in small-scale tourism services. High labour costs drive up the price of those services, which means fewer tourists. High commercial service prices, especially meals and lodging, induce tourists to self-produce them. And high labour costs induce businesses to opt for labour-saving production methods. These are not new claims: the contradiction between Sweden’s employment goals, on the one hand, and the high cost of inexperienced and low-skilled labour, on the other, has been a much studied and hotly debated issue throughout the 1990s. Second, Sweden’s high value added tax on most tourism-related services reinforces the first two effects described above. In particular, it strengthens affluent Swedes’ incentive to travel abroad and the less affluent to visit friends and relatives or


Current Issues in Tourism

vacation at home. It also leads some price-sensitive foreigners to vacation elsewhere. Third, and perhaps most controversially, we hypothesise that the venerable allemansrätten, as currently implemented, also impedes sustainable tourism development. Free public access to undeveloped land, combined with sharp demand peaks, contributes to short-term congestion and long term depletion of prime natural attractions. And free access to private land weakens owners’ investment incentives. Our hypotheses regarding Sweden’s labour, tax and property institutions raise politically contentious issues; therefore they need to be further refined and tested before a public critique of existing policies or a proposal for policy reforms is offered. In any case, it is crucial to recognise that the hegemony of national policy in these areas severely limits the options of Dalarna policy makers. The inconsistencies between these three institutional design features and what is needed for a truly sustainable tourism development are not limited to Sweden. Every central and north European nation now promotes tourism growth as an antidote to rural socio-economic distress and depopulation. For every one of them, however, some combination of high labour costs, high commodity taxes, and free recreational land access slows and distorts such growth. Low-skilled labour, for example, is comparably expensive in Germany, Switzerland, Norway, Austria and Denmark (EPI, 1999: Chapter 8). The value added tax stands at 20% or higher in Norway, Finland, Denmark and Austria (Brors, 2000). Further, Norway, Finland, Denmark, Germany, Austria and Switzerland each has its historically rooted version of the right of common access (Jäggi, 1998). The European Union members among these nations currently receive EU investment subsidies for tourism development in their distressed rural regions. The economic and social payoffs are vitiated, however, by the institutional obstacles described here. It is possible to imagine these nations reaching a clearer understanding of the contradictions embedded within these institutions and moving to harmonise them, either ‘upward’ toward a common European standard or ‘downward’ toward the more free market US pattern. In the United States’ federal system, Maine has greater constitutional authority than Dalarna to shape economic and land use institutions, although in basic labour regulations, federal statutes, such as the 1938 Fair Labor Standards Act, prevail. Maine’s freedom of manoeuvre is also limited by powerful free market and private property rights traditions, not to mention competition with other states for tourist dollars (Vail et al., 1998). Although Maine’s ‘cheap labour’ institutions undoubtedly boost sheer tourism employment, we hypothesise that the low quality of most jobs contradicts the state’s livable wage goal, an equitable sharing of economic benefits, and the revitalisation of distressed rural regions. We also suspect that core property institutions are maladapted for sustainable tourism development. Ownership of Maine’s vast North Woods is turning over rapidly and wildlands real estate development is accelerating, bringing irreversible changes to a vast multiple-use landscape. Maine’s heretofore weak land use regulations and very limited public ownership of prime scenic and recreational lands jeopardise the future of nature tourism. Nonetheless, there are positive signs on several fronts, including cooperative recreational land management, purchases of land and conservation easements by conservation groups, a more systematic approach to land use zoning in the ‘unorganised territories’, and

Institutions Determining Tourism’s Scale and Structure


increasing public ownership. However, the public policy response has in essence been reactive and it may be ‘too little, too late’ to cope with the magnitude of change in Maine’s North Woods. Viewed from a larger national and international perspective, policy makers in several US states, as well as Maine, face a similar choice between fundamentally different tourism strategies. On the one hand, they can perpetuate the present institutional mix of competitive labour markets, low taxes, and quasi-laissez faire land use, which results in income patterns, community impacts, and environmental stress that bear a resemblance to East European or third world tourism (think of Maine as a ‘chilly banana republic’). Alternatively, they may espouse the goals of livable tourism wages, host community vitality, and ecosystem health. Reaching these goals would require institutional innovations centring on greater human capital investments, progressive labour market policies, higher tax rates on tourism services, and more farsighted management of natural endowments. Our hypotheses help to bring these larger stories of institutional adaptation and maladaptation into focus. Since the implications of our critique extend beyond regional tourism economies to core national policies, they are certain to be controversial. Thus, this paper’s conclusions should be considered provisional: a launching point for continuing research. Correspondence Any correspondence should be directed to Professor David Vail, Department of Economics, Bowdoin College, 9700 College Station, Brunswick, ME 04011-8497, USA ([email protected]). Notes *



3. 4. 5.

This paper contains values in US dollars and Swedish kronor for various years between 1994 and 1999. Since exchange rates have fluctuated over the relevant period, the rule of thumb exchange rate used here is SEK 7.5/US$1 (In 1999–2000, the rate has fluctuated between SEK 8.4 and SEK 9.3/US$1). This should not impart any significant bias to the calculations or conclusions in the paper. Estimates of tourism’s economic scope and importance are ambiguous and often exaggerated. ‘Satellite tourism accounts’ have recently been derived from national income accounts, to reflect the fact that tourist services span many distinct industries and make up only part of total spending in most of them. Critics contend that satellite accounting remains crude. In any case, precise accounts for sub-national jurisdictions like Dalarna and Maine do not exist (Hansen & Jensen, 1996; Nordström, 1996; Smith & Wilton, 1997). Vail et al. (1998) suggest that quality tourism jobs have a cluster of attributes including a ‘livable wage’, benefits (such as health insurance, paid vacation and pension contributions), job security, opportunities for skill acquisition and advancement, and workplace congeniality. Maine’s 5000 km of ‘rock-bound coast’ are also a tourist magnet, not discussed in this paper. The overtime supplement applies to work between 8 pm and 6 am, Monday to Friday, and 4 pm to 6 am on Saturday or Sunday. The current supplement for hotel and restaurant workers is SEK 13.50/hour ($1.80), rising to SEK 23 ($3.07) in 2000 (HoR, 1998). Federal law requires that employees be paid 1.5 times normal wages for work beyond 40 hours/week; however, hotels, motels, restaurants and other eating establishments are exempt.


Current Issues in Tourism

6. Workers who earn tips, for example restaurant waiters and bar tenders, may be paid less than the minimum wage as long as combined wages and tips are at least $5.15/hour. Employers who offer meals as a regular condition of employment may deduct the cost of meals from wages. 7. American workers are among the least unionised in any industrial society. Roughly 5.4% belong to unions in ‘services’ and 5.6% in ‘wholesale and retail trade’, two core tourism sectors. Unionisation is still lower among seasonal and part-time tourism workers (BLE, 1998; Murphy & Hanson, 1995: ch. 4). 8. The survey summarised here included a cover letter indicating that the purpose was to demonstrate the high quality of jobs in lodgings and restaurants. This information probably imparted a strategic bias in two ways: first, respondents had an incentive to inflate reported wages and benefits; second, survey recipients with low wages and benefits had an incentive not to respond. The survey response rates were: lodgings 32%; restaurants 36%. 9. For recent years, nominal wages track real US and Swedish wages well: Sweden’s consumer price index increased 25% from 1990 to 1997, compared to 24% in the USA (SCB, 1999: 495). 10. This figure is based on a ‘guestimate’ that half of gross tourism expenditures represent value added within Dalarna, while the rest ‘leaks’ into purchases of goods and services from elsewhere – e.g. food, fuel and vehicles. Sweden’s satellite tourism accounts do not estimate value added at the county level. 11. See Hultkrantz (1995a: 5–8), for a detailed analysis of how VAT changes affected tourism between 1990 and 1993. 12. In general, Sweden’s pre-1990 business taxes discouraged development of small businesses, which predominate in Dalarna tourism. The old tax code favoured capital intensive and highly leveraged corporations relative to family owned, labour-intensive, and equity financed firms. Credit rationing in Sweden’s financial markets also favoured large, established, capital-intensive firms (Henrekson & Johansson, 1999). The 1990 tax reform slightly reduced these biases and probably encouraged tourism, but not enough to offset the higher VAT discussed in the text (Hultkrantz, 1995a). 13. Carbon emissions and energy conservation taxes add another SEK 3.6/litre. 14. New England state sales taxes are: Connecticut 6% plus surcharges of 3% on car rentals and 4% on entertainment; Massachusetts 5% plus 3% on lodgings; Rhode Island 7% plus 5% on lodging; and Vermont 5%. New Hampshire, the outlier, has no general sales tax, but an 8% tax on lodgings (Rafool, 1997). 15. Until recently, this right was not mentioned in statutes; it is now included in the Environment Preservation Act and the Swedish Constitution. Unlike neighbouring Norway, however, Sweden still has no explicit legislation defining the content and limits of allemansrätten. 16. Pending federal legislation, based on the Land and Water Conservation Act, would yield Maine $20–30 million/year for land acquisition. 17. Dalarna’s surface area is 28,194 square km and the 1998 population was 282,198. Northwestern Dalarna has a continental climate with cold, snowy winters and average temperatures of 12–14 C in summer and –10 to –12 C in winter. Southeastern Dalarna has an average summer temperature of 16 C and –6 to –8 C in winter. Annual precipitation averages 60–70 cm, with up to 100 cm in the mountains (SCB, 1999). 18. Most skiers drive to Dalarna. There is no train service to the mountains and the scattered layout of ski lifts and lodgings makes a car handy. 19. Maine is the most northeasterly US state, bordering Canada’s Quebec and New Brunswick provinces. Two-thirds of its 1.2 million residents are concentrated in the southern third of the state. The surface area is 86,500 square km and there are 5650 km of Atlantic Ocean coastline. The average summer temperature is 21 C and winter is –5.5 C. Annual snowfall ranges from 152 to 228 cm (MOT, 1999). 20. Approximately 36% of all overnight visitors in a survey reported shopping at L.L. Bean, while 35% shopped at factory outlet stores (Longwoods, 1997: 66). 21. These employment estimates should be taken with a grain of salt, since tourism employment cannot be estimated directly. Maine’s State Planning Office’s (SPO) uses

Institutions Determining Tourism’s Scale and Structure


an input-output model to estimate total tourism employment, including inter-industry linkages and multiplier effects (Rose, 1997). Comparing 49,900 full-time tourism jobs with average monthly employment of 615,000 in 1996 also gives rise to a problem of commensurability problem, since the total employment figure does not measure year round employment. 22. The lodging sector is viewed as a bellwether for the tourism sector, since it depends heavily on out-of-state visitors. 23. General touring spans activities ranging from museum visits to shopping and from lobster dinners to autumn leaf peeping.

Appendix 1: Sketches of Tourism in Dalarna and Maine Dalarna tourism

1 7

Dalarna, located in Sweden’s central interior, has a long tradition as a tourist destination, with internationally known attractions such as the homes of celebrated painters Carl Larsson and Anders Zorn, the Vasaloppet cross-country ski race and mid-summer celebrations around Lake Siljan. Dalarna’s culture, nature, and major events present an image to foreign tourists of Sweden’s ‘essence’. Small lakes embedded in vast spruce and pine forests characterise much of the landscape, but there are also rolling agricultural landscapes in south and east. To the north and west, Dalarna is mountainous, with two major alpine ski resorts, Idre/Grövelsjön and Sälen (Sweden’s largest) offering downhill and XC opportunities to the 12 million Swedes and Norwegians who live within a day’s drive. Central Dalarna, around Lake Siljan, has strong traditions in music and dance, native costume, vernacular architecture, and crafts (including the famous carved Dala horse). The landscape is dotted with clusters of small red cottages on small plots that escaped Sweden’s great partitioning. Major Siljan attractions include Leksand’s Summerland, Mora’s Santa Land park, Rättvik’s mammoth Dalhalla open air concert hall, and Dalarna’s most visited site, the Clas Ohlson department store, with over half a million yearly visitors. Notwithstanding the county’s reputation and attractions, recent tourism trends give reason for concern. Its national ranking as a tourist destination, measured by guest nights, has fallen from third to fourth in recent years. This is closely linked to the stagnation of skiing, the big growth sector of the early 1990s. According to the Swedish Tourist Authority winter sports holidays are declining throughout Europe. The largest single group of tourists, Swedes visiting relatives and friends (accounting for 19% of total guest nights), is also on the decline. Overall, Dalarna’s tourism economy is nearly stagnant, although tourism is growing nationally and Dalarna policy makers are banking on tourism to boost the economy. Overall, Dalarna has a fairly mature tourist economy and a great deal has already been done to develop capacity and promote demand for the summer and winter activities with greatest market potential. The roster of special events ranges from world skiing championships to summer jazz and folk festivals. Increasing Swedish future tourist spending will depend on building a clientele for the spring and fall ‘shoulder seasons’, attracting more long weekend travellers, and expanding high value ‘niche market’ tourism. The big unknown is whether Dalarna, and for that matter Sweden, can attract significantly more foreign tourists. They currently account for only 5% to 10% of commercial guest nights (LD, 1998).


Current Issues in Tourism

There are two distinct seasonal tourist peaks: mid-June to August and Christmas to Easter. Although 80% of foreigner’s visit in the summer, just 40% of total guest nights are recorded from May through September: 60% are recorded in the winter months, with alpine skiing the biggest draw. The Sälen area is Sweden’s skiing leader with a 30% market share. Sälen and Idre benefit from a competitive locational advantage over ski resorts farther north.18 During the January–February ‘sport holiday’ ski resorts operate close to full capacity, attracting parents with children of all ages. This national institution helps explain the large share of winter tourism in Dalarna, compared to Maine. Guest night statistics, expenditures, and employment Apart from the highly commercialised winter sport tourism in the mountains, which accounts for 30% of total guest nights, most Dalarna tourists are either inhabitants of the region or other Swedes visiting relatives and friends or their own second homes. There are about 2.9 million overnight visitors per year and total guest nights of 14.5 million. (Both figures are almost exactly half of Maine’s corresponding levels.) There are also 1.5 million day trips to Dalarna, less than one-tenth of Maine’s level. Two-thirds of Dalarna’s visitors but only one-third of Maine’s make overnight stays. As mentioned, foreign tourists account for just 5% to 10% of total guest nights. The largest cohorts are Danes, with roughly a 40% share, and Germans at 35%. Norwegians are Dalarna’s closest neighbours, but with many similar attractionsin their own country, they constitute only 10% (LD, 1998). Table A1 indicates the relative importance of major visitor groups, their spending, and the associated employment. Total 1997 tourism spending in Dalarna has been estimated at SEK 4 billion ($533 million), with direct employment in the tourist industry of 5000 full time equivalents (FTE) (LD, 1998). These figures are about one-sixth and one-eighth of Maine’s corresponding levels. Not all benefits from tourism employment remain in Dalarna, since many seasonal employees, especially at ski resorts, pay income tax in other counties. There is no accurate estimate of the proportion of tourist spending that ‘leaks’ out of Dalarna in the form of purchases of food, fuel, etc. produced elsewhere. Using a 50–60% ‘guesstimate’ for leakage (similar to Maine’s), tourism spending probably accounts for about 3.1 to 3.9% of Dalarna’s 1996 gross county product of SEK. 53.1 billion (NUTEK, 1999). This figure suggests that the economy’s dependence on tourism is considerably less than Maine’s and roughly on a par with Sweden as a whole. Not surprisingly, Table A1 reveals that visitors staying overnight in commercial lodgings spend far more and generate more employment per visitor-day than visitors to relatives, friends and seasonal homes. Interestingly, business and conference tourism creates more spending and employment per guest night than any other tourism category. Discussions of Swedish tourism commonly divide spending into four broad categories: lodging, transport, eating, and ‘doing.’ It is of note that ‘Activities’– the ‘doing’ part of tourism – account for just 14% of Dalarna’s tourist expenditures. At first glance this figure may seem low, considering that ‘activities’ ranging from skiing to folk festivals are what attract many tourists to Dalarna. In fact Dalarna’s ‘doing’ share is high compared to other Swedish regions and to Maine, presumably due to the centrality of alpine skiing, an expensive activity. None-

Institutions Determining Tourism’s Scale and Structure


Table A1 Dalarna tourism statistics (percentages) Type of tourist Foreigners Swedes Of whom: Visits to friends, relatives or own leisure homes Commercial lodging Winter Other Work-related Day visits

Visitor days (%) 5 95

Share of spending (%) 5 95

Share of employment (%) 5 95




30 10 5 10

40 10 15 10

40 15 20 5

Source: LD, 1998: 24

Table A2 Dalarna tourist spending and employment by activity Activity Lodging Groceries Restaurants Transportb Shopping ‘Activities’

% of Spending 26 18 18 14 10 14

% of Employment 48 8 16 4 5 19

Labour intensitya 1.8 0.4 0.9 0.3 0.5 1.4


% of employment divided by % of expenditure including petrol Source: LD, 1998: 25 b

theless, ‘Doing’ is probably tourism’s most underdeveloped component, in terms of untapped potential for value-added and employment (Stiernstrand, 1998). Table A2 reveals pronounced differences in the labour intensity of various tourist services, with lodging, tourism activities, and restaurant meals generating the most employment per krona spent. Grocery and other retail stores and transportation generate far less employment per krona. The data hint at the economic importance of encouraging growth in types of tourism that include commercial lodging and focus on commercial recreation activities, including value-added nature tourism. In reality, however, commercial guest nights increased by only 3.7% in Dalarna between 1988 and 1997, less than one-half percent per year and well below the 6.4% national growth (TD, 1998b). By this measure, Dalarna’s tourist industry has been nearly stagnant for a decade. The bright side of the picture economically is that stays at inexpensive camp grounds have declined sharply as visitors move ‘up market’ to more expensive and employment-creating accommodations. A statistical sketch of Maine tourism


Maine hosted 27 million trips in 1996,as shown in Table A3. Day trips made up over two-thirds of the total, reflecting the tremendous expansion in short dura-

Current Issues in Tourism


Table A3 Trips to/within Maine by North Americans: 1996 a

Maine residents Other Americans Canadians Total

Day trips 5.7m (31%) 6.8m (38%) 5.7m (31%) 18.1m

Number of trips Overnight trips 1.6m (18%) 6.3m (71%) 1.0m (11%) 8.9m

Source: Longwoods (1996: 12–15) a This study does not document foreign visitors other than Canadians

tion leisure travel in recent decades. Maine residents comprised about one-fourth of tourists and Canadians were by far the most important foreign group. The concentrated geographic origin of visitors is reflected in the fact that 30% of ‘marketable’ overnight visits were by Massachusetts residents and fully 71% were by residents of New England, New York, and Canada – the 30 million people who live within one day’s drive of Maine. Of 20.1 million visits by Americans, 63.7% were ‘marketable’ pleasure trips, 28.9% were visits to friends and relatives, and 7.5% were business-related (Longwoods, 1997: 38). In comparison to Dalarna, Maine has a much larger proportion of ‘day trippers’ and users of commercial lodgings. The 5.7 million users of commercial lodgings averaged 5.3 nights per stay in 1996, giving Maine a total of 30.2 million commercial guest nights, or roughly twice Dalarna’s level (Longwoods, 1997: 14). Survey research indicates that tourists spent $3.2 billion in Maine in 1996: $2.2 billion by visitors ‘from away’ and $1.0 billion by residents. However, approximately 60% of this spending ‘leaks’ out of the state’s economy in the form of federal taxes and direct and indirect ‘imports’ of food, fuel, materials, services, etc. A rough estimate of tourism’s 1996 in-state value-added is $1.3 billion, or 4.8% of gross state product (Longwoods, 1997: 23; MEGC, 1998: 5; Vail et al., 1998: 15). The spending breakdown in Table A4 reveals that three activities – shopping, eating, and sleeping – account for almost all tourist spending. (‘food & alcohol’ refers to spending in eating and drinking establishments: tourist spending on groceries is included in the retail category.) The large retail component reflects the fact that millions of visitors shop at factory stores and outlet malls. Freeport’s L.L. Bean retail store is Maine’s most popular tourist site, with over 4 million annual visitors.20 It is estimated that tourism spending directly supported 49,900 full-time equivalent jobs in 1996, making it the state’s largest employment generator, with Table A4 Components of tourist spending (1991) Sector Retail trade Food & alcohol Lodging Recreation & tourist activities Transportation Source: Vail et al., 1998: 43

% of Tourist spending 44 24 17 10 5

Institutions Determining Tourism’s Scale and Structure


Table A5 Seasonal distribution of marketable overnight trips January–March April–June July–September October–December

9% 23% 50% 18%

Source: Longwoods, 1997: 53

8.1% of all jobs. Input-output simulations suggests that when indirect employment is added, tourism accounted for 71,800 FTE jobs (this includes jobs in businesses supplying tourist service providers as well as multiplier effects)21 (Longwoods, 1997: 27; Rose, 1997: 6; Vail et al., 1998: 14fn, 16fn). In sum, tourism comprises a very labour-intensive set of activities in Maine, generating 8.1% of Maine’s employment from less than 5% of gross state product. Tourism is seasonally skewed, as revealed in Table A5. In contrast to Dalarna, two-thirds of all visits are concentrated in the summer months of June-September. Despite the state’s numerous alpine skiing facilities and outstanding snowmobile trail network, winter is a low tourist season, a sharp contrast with Dalarna, where winter is the peak season. Whereas Dalarna benefits from relative closeness to major population centres, Maine’s ski areas face a locational disadvantage vis à vis New Hampshire and Vermont resorts closer to Boston and New York. Skiing quality in the eastern US is also inferior to the Rocky Mountain states. On the other hand, Fall ‘leaf peeping’ is a growing market segment, attracting affluent but busy working couples and retired people from the US Northeast (Carrier, 1998). Roughly 70% of tourist spending is concentrated on the coast, where 9 of Maine’s 10 most frequently visited host communities are located (Longwoods, 1997: 70; Vail et al., 1998: 18). Nearly all non-Canadian visitors enter via interstate highway at Maine’s southern boundary, passing through the south coastal region and often stopping to eat and shop there even when their ultimate destination is the North Woods or downeast coast. This resembles the position of Dalarna’s Siljan region with tourists driving from greater Stockholm. Growing numbers, declining market share Policy makers claim that tourism is a prime growth sector, yet the available data show tourism value-added growing more slowly than overall state production during the macroeconomic recovery from 1991 to 1996: 0.5%/year compared to 2.5%/year. From 1994 to 1996, the growth tempo picked up, with marketable guest nights increasing by 2.8%/yr and lodging employment rising 2.9%/yr;22 the tempo also appears to be accelerating during the current full employment boom. Preliminary 1997–98 data show lodging expenditure rising 7%–8%/year (MEGC, 1998: 22; Longwoods, 1997: 17; SPO, 1998: 12; Vail et al., 1998: 15). Somewhat anaemic growth throughout 1996 dropped Maine’s national rank as a tourist destination from 37th to 38th, measured by total spending. Even within the generally slow growing northeast tourism market, Maine’s share of marketable trips fell from 4.5% to 3.8%. As explained above, Dalarna has experienced a similar market share deterioration. There are several possible explanations for Maine’s declining share. The ‘hospitality industry’ and the Maine Office of Tourism claim that the state spends too little on marketing and promotion compared to other states. Comparative advertising statistics bear this out, yet


Current Issues in Tourism

underlying demand trends also seem to work against Maine’s major tourist attractions. Maine’s national ranking in the category ‘outdoor trips’ has actually improved from 24th to 20th, but outdoor tourism is declining in popularity as a fraction of all trips and in absolute expenditure. Yet another of Maine’s strengths, ‘general touring’, has been growing at only 1%/yr.23 Nationally, tourist spending growth is fastest for services where Maine’s competitive position is weak: casinos, theme parks, cruises, ski trips, and major events. Among these, only skiing has much current presence and clear future promise in Maine, although the state does increasingly promote cruises and special events (Longwoods, 1997: 6, 10, 30). These broad demand trends suggest that if Maine policy makers want tourism to be an engine of income and employment growth for the next century, they must encourage investment in services that match demand trends or find a more effective strategy to expand existing market niches, especially nature-based tourism.

Appendix 2: Maps

Figure 1 Dalarna County and its location in Sweden and Europe Source: Dalarna (2000)

Institutions Determining Tourism’s Scale and Structure


Figure 2 The Maine Woods

References Androscoggin Land Trust (ALT) (1996) Maine Landowner Liability Explained. Lewiston, Maine. Anttila, S. (1999) The snowmobile issue as a matter of common-pool resources. ETOUR Report. Mid-Sweden University. Associated Press (AP) (1998) Rafting industry hits new record. Times Record (24 December). Brunswick, Maine. Austin Associates (1997) Maine Lodging and Restaurant Wage Survey. For the Maine Office of Tourism (Department of Economic and Community Development). Augusta, Maine. Austin, P. (1997a) Lake access controversy may just be getting started. Maine Times (5 June). Austin, P. (1997b) Webber Timberland heir sells: Liquidation could follow. Maine Times (6 November). Becker, C.D. and Ostrom E. (1995) Human ecology and resource sustainability: The importance of institutional diversity. Annual Review of Ecology and Systematics 26, 113–33. Becker, G.S. (1965) A theory of the allocation of time. Economic Journal 75, 493–517. Brors, H. (2000) Lägre skatter tvingas fram. Dagens Nyheter (10 August). Bureau of Labour Education (BLE) (1991) Reducing the Health Care Cost Burden on Maine Consumers and Taxpayers. Orono, Maine: University of Maine. Bureau of Parks and Lands (BPL) (1998) Maine Department of Conservation Fee Schedule. Augusta, Maine.


Current Issues in Tourism

Carrier, P. (1998) State tries to rake in leaf peepers. Maine Sunday Telegram. Clarke, H. (1997)Australian tourism industry policy: A new view. Tourism Economics 3 (4), 361–77. Colby, K.T. (1988) Public access to private land – allemansrätt in Sweden. Landscape and Urban Planning 15, 253–64. Cowperthwaite, A. (1999) Managing director, North Maine Woods Inc. Telephone interview. Dalarna (2000) Map of Dalarna, www.dalarna.se/karta, January. Economic Policy Institute (EPI) (1999) The State of Working America. Ithaca, NY: Cornell University Press. Eriksson, H. (1999) Director, HoA. Telephone interview. Fisk, A. (1999) Maine Land Use Regulation Commission. Personal communication. Folke, C., Pritchard, L., Berkes, F., Colding, J. and Svedin, U. (1997) The problem of fit between ecosystems and institutions. Beijer Institute Discussion Papers 108. Stockholm. Hansen, C. and Jensen, S. (1996) The impact of tourism on employment in Denmark: Different definitions, different results. Tourism Economics 2 (4), 283–302. Hasbrouk, S. (1998) The public access issue. Maine Fish and Wildlife (winter). Henrekson, M. (1998) Högre sysslesättning genom en utvidgad marknadssektor. Ekonomisk Debatt 7, 515–26. Henrekson, M. and Johansson D. (1999) Institutional effects on the evolution of the size distribution of firms. Small Business Economics12, 11–23. Hotell och Restauranganställdas Arbetslöshetskassa (HoA) (1998) Bra att Veta om Arbetslösheten. Stockholm. Hotell och Restauranganställdas Förbund (HoR) (1998)Kollektivavtalmellan SverigesHotell och Restaurangföretagare och Hotell och Restauranganställdas Förbund, 1998–2001. Malmö, Sweden. Hultkrantz, L. (1995a) On determinants of Swedish recreational domestic and outbound travel 1989–1993. Tax Reform Evaluation Report 7. Uppsala University. Hultkrantz, L. (1995b) Hushållning med knappa resurser. ESO Rapport 15. Stockholm: Allmänna Förlaget. Hultkrantz, L., Larsson S-O., Lundström M. and Paulsson B. (1987) Säsongsarbete och regionalpolitik. Cerum Report 2. Umeå, Sweden. Idre Fjäll (1999) Marketing folder and field studies by author at Idre Fjäll, Sweden 10–13 June. IFW (Maine Inland Fisheries and Wildlife Department) (1998a) Landowners and Land Users: How to Get Along: General Information. Augusta, Maine: Department of Conservation. IFW (Maine Inland Fisheries and Wildlife Department) (1998b)Summary of Maine Hunting and Trapping Laws and Rules. Augusta, Maine: Department of Conservation. Irland, L. (1993) Outdoor recreation supply in the Maine woods: Issues for the future. Renewable Resources Journal (Autumn), 6–15. Irland, L. (1996) Land, Timber and Recreation in Maine’s North Woods. Miscellaneous publication 730. Orono, Maine : Maine Agricultural Experiment Station. Jäggi, S. (1998)Allemansretten I Sveits, Tyskland og Østerrike. In B. Dahle (ed.) Friluftsliv I et Nytt Århundre. Oslo: Norges Idrettshøgskole. Kimball, F. (1999) Maine Bureau of Insurance. Personal communication. Landsorganisationen (LO) (Swedish Trade Union Confederation) (1999) The Swedish Labour Market – Facts and Figures. Stockholm. Länsstyrelsen Dalarna (LD) (1998) Turismen i Dalarna in i det 21:a århundradet. Falun, Sweden. Longwoods Associates (1997) Maine’s 1996 Travel Year and 1997 Advertising Evaluation Summary Report. Montreal: Maine Office of Tourism. LURC (Maine Land Use Regulation Commission) (1997) Comprehensive Land Use Plan (3rd rev). Augusta, Maine: Department of Conservation. Maine Audubon Society (MAS) (1996) Valuing the Nature of Maine. Falmouth, Maine. Maine Department of Labour (MDOL) (1995) Occupational Wage Rates, Maine Statewide. Augusta, Maine.

Institutions Determining Tourism’s Scale and Structure


Maine Department of Labour (MDOL) (1997) Occupational Wage Rates by Industry. Augusta, Maine. Maine Department of Labour (MDOL) (2000) Occupational Wage Data: 1998. Augusta, Maine. Maine Department of Transportation (MDOT) (1997) Strategic Transportation Plan. Augusta, Maine. Maine Economic Growth Council (MEGC) (1998) Measures of Growth 1998. Augusta, Maine: Maine Development Foundation. Maine Office of Tourism (MOT) (1995) Five Year Marketing and Development Strategy for Maine Tourism. Augusta, Maine: Maine Department of Economic and Community Development. M ai n e O ffi ce o f T o u ri s m (M O T ) ( 19 99 ) St a t e fac t s a nd i nf o rm a t io n. www.visitmaine.com/facts, February. Murphy, E. (1999) ‘And the legislature’s biggest chore is . . . ’. Maine Sunday Telegram (17 January). Portland, Maine. Murphy, W. and Hanson J. (1995) A Maine Guide to Employment Law. Orono, Maine: University of Maine, Bureau of Labour Education. Nerhagen, L. (1997) Faktorer som påverkar färdmedelsvalet vid längre skidresor. CTS Working Paper 11. Dalarna University College, Sweden. Nordström, J. (1996) Tourism satellite accounts for Sweden 1992–93. Tourism Economics 2 (3), 13–42. Nordström, J. (1999) Unobserved components in international tourist demand. Paper II in doctoral dissertation: Tourism and travel: Accounts, demand and forecasts, No 509. Umeå Economic Studies, Sweden. NUTEK (1993) Skattereformen och turismen. R: 1993: 24. Stockholm. NUTEK (1999) Regional Statistics Over Gross Regional Product for Dalarna. Dalarna: Länsstyrelsen. Persson, A. (1999) HoR Dalarna section. Telephone interview. Påhlson, A-M. (1997) Taxation and the market for domestic services. In I. Persson and C. Jonung (eds) Economics of the Family and Family Policies. London: Routledge. Rafool, M. (1997) State tourism taxes. Travel and Tourism 2. Denver, CO: National Conference of State Legislatures. Rose, G. (1997) Maine Regional Output Simulation Model. Augusta, Maine: State Planning Office. Sahlberg, B. (1996) Resandets Grundläggande Struktur och Dynamik. Turistdelegationen. Stockholm. Sahlberg, B. (1998) Ku ngar, Katastrofer, och Kryddor (om en världsspännande industri-turismen). Östersund, Sweden: ETOUR. St John, C., Ditre, J. and Pohlmann, L. (2000) At Risk: Small Business Health Coverage. Augusta, Maine: Maine Center for Economic Policy. Smith, S. and Wilton, D. (1997) TSAs and the WTTC/WEFA methodology: Different satellites or different planets? Tourism Economics 4 (1), 71–8. SPO (Maine State Planning Office) (1998) Maine Retail Sales Quarterly Report. Third Quarter 1998. Augusta, Maine. Statistiska Central Byrån (SCB) (1999) Statistisk Årsbok 1998 (Official Statistics of Sweden). Stockholm. Stiernstrand, O. (1998) Riktade studier för utveckling av svensk turism – studie gjord på uppdrag av Turistdelegationen U 1998: 1. Östersund, Sweden: ETOUR. Svensk författningssamling (SFS) (1982) Arbetstidslagen 673. Stockholm. Svenska Turistföreningen (STF) (1996) Svenska Turistföreningens Årsbok 1997. Dalarna. STF publication no 3183. Turistdelegationen (TD) (1996) Sunt Förnuft – Grunden För en God Byråkrati. Stockholm. TD (1998a) Turismen Behöver Välutbildade Medarbetare. Stockholm. TD (1998b) Hållbar Utveckling i Svensk Turistnäring. Stockholm. Vail, D. and Hillard, M. (1997) Taking the High Road: Human Resources and Sustainable Rural Development. Augusta, Maine: Maine Center for Economic Policy (Ford Foundation funding).


Current Issues in Tourism

Vail, D. and Hultkrantz, L. (2000) Property rights and sustainable nature tourism: Institutional adaptation and mal-adaptation in Dalarna and Maine. Ecological Economics. Vail, D. and Kavanaugh, W. (2000) Livable Wages in Maine Tourism? Augusta, Maine: Maine Center for Economic Policy. Vail, D., Lapping, W.R. and Cote, M. (1998) Tourism and Maine’s Future. Augusta, Maine: Maine Center for Economic Policy (Maine Community Foundation funding). Westergård T. (1999) Managing director of Lövåsgården. Personal communication. Westholm E. (1992) Mark, människor och moderna skiftesreformer i Dalarna. Geografiska regionstudier 25 (Phd thesis). Uppsala.

Ecological Economics 48 (2004) 469 – 483 www.elsevier.com/locate/ecolecon


Governing snowmobilers in multiple-use landscapes: Swedish and Maine (USA) cases David Vail a,*, Tobias Heldt b,c a

Department of Economics, Bowdoin College Brunswick, ME 04011, USA b Uppsala University, Sweden c Dalarna University College, Sweden

Received 4 November 2002; received in revised form 20 October 2003; accepted 20 October 2003

Abstract Snowmobiling growth in North America and Sweden creates challenges in ‘‘governing the commons.’’ Snowmobiling contributes to the economy of distressed rural regions and enhances residents’ quality of life; but quasi-open access to winter landscapes also breeds conflicts: among snowmobilers, with landowners, with other recreationists, and with environmentalists and ecosystem health. Common pool resource and impure public goods theories are used to interpret these conflicts and strategic interactions. Case studies in Sweden and Maine yield insights about conditions for sustainable management of multiple-use landscapes when property rights are complex and stakeholders diverse. The case studies utilize key informant interviews, tourist surveys, and a contingent valuation exercise to illustrate how innovative governance institutions, complemented by infrastructure investments, can mitigate conflicts, re-align incentives, and internalize costs. Local self-governance has evolved with state facilitation, but without changes in fundamental property law. Three keys are trail agreements between snowmobile clubs and landowners, clubs’ norm formation and rule enforcement, and public – private investment in quality trails. These arrangements are strained by hot spot congestion, free riding, and unresolved conflicts with other recreationists. D 2004 Elsevier B.V. All rights reserved. Keywords: Multiple-use landscape; Snowmobile; Common pool resource; Impure public good; Free riding; Governance regime

1. Introduction: snowmobiles’ contradictory roles in rural life and landscapes The central problem explored here is governing recreational access to multiple-use landscapes in an era of growing motorized off-road recreation. The ecosystems under study combine private and state

* Corresponding author. Tel.: +1-207-725-3596; fax: +1-207725-3691. E-mail address: [email protected] (D. Vail). 0921-8009/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.ecolecon.2003.10.014

land ownership, but with important common property features regarding public recreational access. Limited compatibility among land uses and limited possibilities for excluding motorized users have bred conflicts among stakeholders and with ecosystem health. Although this narrative centers on snowmobiling, analogous tales could be told about the impacts of all-terrain vehicles (ATVs), personal water craft (jet skis), and even floatplanes (aircraft), whose sounds, sights, and emissions increasingly pervade industrial nations’ backcountry landscapes (Havlick, 2002; Vail, 2001). We investigate how stakeholders


D. Vail, T. Heldt / Ecological Economics 48 (2004) 469–483

in Sweden’s western mountains and Maine’s northern forest have responded to emergent snowmobiling conflicts, reshaping governance institutions and building infrastructure, but not changing core property laws. Snowmobiling has complex and contradictory effects on rural lifestyles, economies, and ecosystems. It has unquestionably improved the quality of winter life for many rural residents, replacing isolated ‘‘cabin fever’’ with group-centered outdoor adventure. ‘‘Snow scooter driving is part of our identity up here in northern Sweden. It sits deeply in our soul that this is freedom’’ (Lindgren, 2003). Snowmobiling has buoyed economically distressed regions, suffering from chronic winter unemployment and ‘‘mature’’ agriculture, forestry, and mining sectors (Vail and Heldt, 2000). Snowmobiling also manifests cultural contradictions. It perpetuates machine-centered, nature-dominating rural traditions, but has also evolved into a family-, club-, and community-oriented activity, featuring ice fishing contests, barbeques, charities, and outdoor religious services (Anttila, 2001; Genthner, 2002). A few statistics convey snowmobiling’s scope, growth, and economic importance in the study regions. Swedes registered 224,000 snowmobiles in 2001, up 47% from 1990. This is roughly one snowmobile per 25 adults, with the ratio nearly oneper-two adults in northern counties. Maine’s 1.2 million residents registered 80,000 snowmobiles: a one-per-fifteen residents ratio said to be the highest of any US state (Rubin et al., 2001; SCB, 2000). In 2002, snowmobilers’ spent over US$300 million in Maine, a 500% increase from 1988. ‘‘Sledding’’ now equals alpine skiing as Maine’s top source of winter tourism revenue. It generates 2300 full-time equivalent jobs (BPR, 1988; Reiling, 1998; Mazzone, 2001). Not surprisingly, Maine’s Office of Tourism and local chambers of commerce hope to attract still more snowmobilers. Swedish policy, with regional exceptions, has been less encouraging. Snowmobilers’ economic contribution is less appreciated and, as we shall explain, they are implicated in more intense land use conflicts. In contrast to snowmobiling’s contributions to rural economics and welfare, it is implicated in several distinct conflicts, entailing what Manning (1999) terms ‘‘goal interference attributed to others’’

(p. 203). Knopp and Tyger (1973) began to document such snowmobiling conflicts three decades ago. Lindberg et al. (2001) and Havlick (2002) explain that these conflicts are typically asymmetrical, with motorized recreationists who impose costs, such as property damage and habitat disruption, often unaware of their interference with others’ goals. Fig. 1 introduces four principal conflicts uncovered by this study (one-way arrows indicate predominantly directional impacts; two-way arrows indicate reciprocal impacts.) Snowmobiling can also confer benefits on other actors. For instance, most Swedish backcountry skiers view snowmobilers as an emergency lifeline, and Maine’s forest owners rely on snowmobile clubs to help police ATV tresspassing (Gamble; Lindberg et al., 2001). Given this mix of detrimental and beneficial externalities, Marcouiller and Green’s (2000) identification of three degrees of compatibility—mutual exclusivity, competitive coexistence, and complementarity—is useful. The paper is organized as follows. First, the study areas are briefly described. Section 2 then draws on two theoretical traditions to explain snowmobiling conflicts and conflict resolution. One approach weaves together common pool and impure public goods theories, emphasizing complementarity among three types of capital: natural (winter landscapes), human-made (trail infrastructure), and social (governance institutions). The second approach combines institutional economics and governance theories to show how innovative governance arrangements reshape snowmobilers’ incentives, behavior, and norms. Section 3 describes field research methods. Sections 4 and 5 present the Sweden and Maine case studies. Finally, Section 6 summarizes findings, stressing the evolutionary adaptation of governance regimes to changing constraints.

Fig. 1. Patterns of conflict.

D. Vail, T. Heldt / Ecological Economics 48 (2004) 469–483

1.1. Sweden’s western mountains and Maine’s northern forest The study areas have core similarities, but also differences that make for fruitful comparison. Both are predominantly rural, with extensive mountain ranges, major watersheds, and hundreds of lakes and ponds. The land is over 90% forested. In Sweden’s Dalarna County, three-fourths is privately owned; the Maine figure is 94%. Both have long


outdoor recreation traditions. Storyteller Hans Christian Andersen lauded Dalarna’s rustic charms following an 1849 visit. H.D. Thoreau shaped Maine’s wilderness mystique with his contemporaneous Maine Woods essays. In recent decades, regional economies have become increasingly tourism-dependent. Tourism accounts for 4% of Dalarna’s employment, versus 2% for all Sweden; and 10% of Maine jobs, compared to 5% for the US economy (Vail and Heldt, 2000).


D. Vail, T. Heldt / Ecological Economics 48 (2004) 469–483

Sweden’s centuries-old right of common access, allemansra¨tten, ensures public access to undeveloped landscapes. Not codified in law, it is rather a set of customary rules and judicial interpretations regarding activities, such as camping, berry picking, and crosscountry (XC) skiing, where the state accepts no obligation to defend landowner claims. Under allemansra¨tten, XC skiing became a widespread pastime, with the Society for Promotion of Ski Touring founded in 1892. Today, one-fourth of Swedes, compared to just 3% of Americans, participate (McIntyre, 1999). In 1975, following heated debate about allemansra¨tten’s applicability to recreational vehicles,

Parliament passed the Terrain Driving Law, granting snowmobiles access to nearly all snow-covered land. (Twelve fragile alpine regions were excepted.) Today’s snowmobile conflicts stem largely from this statute, which, de facto, curtailed landowners’ and self-propelled recreationists’ rights and opened the door to environmental degradation. Maine’s 4-million-hectare northern forest is nearly all commercial timberland, long held by a handful of owners bearing names like International Paper. Legally, landowners could prohibit public access, yet most allowed recreational use of lands not involved in logging operations. Over many decades, ‘‘Mainers’’

D. Vail, T. Heldt / Ecological Economics 48 (2004) 469–483

came to view recreational access as an entitlement.1 In the 1970s and 1980s, with construction of a 40,000km logging road network, access was—sometimes grudgingly—extended to snowmobilers. As a practical matter, the transaction costs of excluding them or collecting fees would have been prohibitive, since sledders could drive nearly anywhere in the vast winter landscape.2 In the mid-1970s, Maine’s ‘‘snowmobile problem,’’ unlike Sweden’s, centered on lawless riders and high policing costs.

2. Conceptual lenses 2.1. Outdoor recreation and common pool resource (CPR) theory Textbook discussions of CPRs emphasize homogeneous natural capital stocks such as trees, fish, and aquifers (Tietenberg, 2003, p. 70). With open access, individuals’ incentives to intensify exploitation are misaligned with the collective interest in restraining exploitation, to enhance economic rents and conserve stocks. Open access causes reciprocal externalities, whereby each user’s harvest subtracts from the remaining pool and may also raise others’ extraction costs. Rational individual exploitation dissipates short-term income (scarcity rent) and may exceed maximum sustainable yield. As the following case studies convey, recreational landscapes do not meet a literal definition of open access, where there is no vested ownership, no limitation on use, and no user responsibility (Hanna et al., 1996). Most snowmobile trails cross private land; Swedes are indoctrinated that allemansra¨tten rests on a conservation ethic and respect for landowners’ interests; and Maine recreation brochures stress that the public’s backcountry adventures depend on landowners’ goodwill. CPR theory has recently been applied to recreational land use conflicts (Bosselman et al., 1999; 1

User fees are increasingly collected for backcountry recreation, like hunting and camping, but not snowmobiling (Vail and Hultkrantz, 2000). 2 Permissive access was also motivated by forest owners’ search for political allies against conservationist efforts to regulate clearcutting, pesticide spraying and other industrial forestry practices.


Marcouiller, 1998; Anttila, 1999). With snowmobiling, it applies most clearly to type 1 conflicts among snowmobilers. The key reciprocal externality involves hot spot congestion at peak times on prime trails, for instance, Easter Weekend on Sweden’s Gro¨nklitt massif. ‘‘Congestion’’ is shorthand for a nexus of amenity-reducing and cost-raising effects, including increased physical crowding, exhaust fumes, trail deterioration, accidents, and hostile encounters. It also includes what Manning (1999) terms ‘‘displacement costs,’’ when sledders react to congestion by traveling farther afield or riding less attractive trails. Hot spots are intensified by the broad social trend toward concentration of leisure into weekend bursts, redoubling pressure on trails within a few hours’ drive of population centers. Although snow-covered landscapes are quite resilient, unrestricted access nonetheless contributes to three depletion effects: soil and plant degradation when snow cover recedes, wear-and-tear on infrastructure (bridges, culverts, etc.), and trail mileage cutbacks when landowners gate their property in response to heavy traffic. Open access also entails a free rider problem. Quality trails require in-season grooming as well as investments in land clearing, culverts, bridges, and signage. With open access, individuals have minimal incentive to contribute to these investments, resulting in sub-optimal trail quality. This, in turn, weakens the incentive to ride on designated trails. Three basic approaches to CPR governance are privatization and markets, state ownership and regulation, and community norms and reciprocity. Whichever regime is crafted to govern snowmobiling, some free riding incentive persists due to high exclusion and fee collection costs. 2.2. Ecosystem services, multiple use landscapes, and impure public goods The forest, lake, and alpine ecosystems of western Sweden and northern Maine generate service flows— timber, natural beauty, stream flows, songbirds, wild game, etc.—that support both commodity production (forestry, farming, reindeer herding) and many forms of recreation. The mix and levels of these flows determine sustainable carrying capacity for any particular activity and shape conflicts and complementar-


D. Vail, T. Heldt / Ecological Economics 48 (2004) 469–483

ities among activities. Recreational carrying capacity is a controversial notion, used here heuristically, not literally. Capacity is rarely a fixed bio-physical limit. Typically, it is influenced by the resource management regime and by various actors’ ‘‘sensitivity to conflict’’ and ‘‘coping behaviors’’ (Manning, 1999, pp. 202 – 203). Although higher use levels commonly raise incremental costs, there is seldom a ‘‘bright line’’ beyond which growth causes irreversible damages or a system collapse. This is particularly true when users adjust their expectations or shift to less-congested sites and times (Manning, 1999; Prato, 2001; Seidl and Tisdell, 1999). Landscape scale is important both for carrying capacity and conflict management. Locally, XC skiing and snowmobiling may be mutually exclusive, while at a large landscape scale, they are compatible or even complementary. Thus, snowmobile tracks in remote areas facilitate skiing in deep snow and provide a rescue lifeline. In the study areas, the overarching land use complementarity is between snowmobiling and commercial forestry, where unused logging roads are integral to trail networks. The landscapes under study are best considered impure public goods, in contrast to pure public goods where exclusion is not feasible and no-one’s use subtracts from others’ welfare (Vail and Hultkrantz, 2000). Snowmobiling can certainly have detrimental third-party effects and disrupt ecosystem services. Symmetrically, other land uses may adversely affect snowmobilers. A distinctive feature of snowmobiling is the strong complementarity between two impure public goods: natural landscapes and trail infrastructure. Well-groomed, well-sited trails mitigate third party effects by inducing most sledders to steer clear of logging operations, XC ski tracks, and wildlife breeding areas. 2.3. Managing multiple use landscapes with vested ownership but common property features Williamson (2000) defines resource governance as, ‘‘an effort to craft order, thereby to mitigate conflict and realize mutual gains’’ (p. 599). Designing snowmobile governance to reduce, if not eliminate, the four types of conflict faces impediments. First, it is not politically feasible to rewrite fundamental property laws. Second, there are cultural obstacles to modify-

ing snowmobilers’ behavior and norms. Third, incentives to shirk make enforcement of laws, contracts, or informal rules costly. Finally, governance is complicated by trail networks crossing multiple ownerships and political jurisdictions. Focusing on strategic interactions between snowmobilers and landowners in western Sweden, Anttila (2001) models the evolution of local contracting arrangements. A ‘‘lose – lose’’ prisoners’ dilemma game has been replaced by a repeated cooperative game. In the pre-contract setting, landowners posted land against trespassers and threatened to prosecute violators: the worst case for snowmobilers. Enforcement, however, imposed heavy monitoring costs on the owners. Contracts for well-defined trails, with self-policing by snowmobile clubs, improved both parties’ payoffs. Snowmobilers avoided prosecution risk and collective trail investment brought higher quality riding. Landowners could limit sledding to specified areas and cut enforcement costs. Two conditions strengthened the players’ incentives to make and abide by such ‘‘win – win’’ (Pareto-improving) contracts. First, quality trails induced most snowmobilers to ride in designated corridors.3 Second, asymmetric information—snowmobilers’ knowledge of peer behavior—made self-policing more cost-effective than third-party enforcement. A key to evolutionary success was the clubs’ discouragement of ‘‘rogue sledding’’ through a mix of moral suasion and informing state rangers of violators. The vast literature on commons governance reveals no universal truths: ‘‘the devil is in the details.’’ In our case, organizing snowmobilers and landowners, mediating contract negotiations, sanctioning shirkers, and financing trail infrastructure all pose challenges. Social scientists have surveyed dozens of case studies and experiments in commons governance (Grafton, 2000; Ostrom, 2000; Agrawal, 2002). Synthesizing insights from diverse contexts, Agrawal (2002, pp. 62 –63) identifies over 30 features of resource systems, stakeholder groups, institutional arrangements, and ‘‘external environments’’ that influence the effec3

Maine’s forest managers claim most sledders keep to designated trails; the main exception and concern is sledding on plowed logging roads (Gamble; Medina). The Funa¨sdalen case described in Section 4 also supports this conclusion (Alexandersson, 2001).

D. Vail, T. Heldt / Ecological Economics 48 (2004) 469–483

tiveness and resilience of governance regimes. Given such complexity and context specificity, a deterministic theory of commons governance is not likely to emerge. Nonetheless, Grafton (2000) identifies facilitating conditions—not definitive causes—relevant to the snowmobile context:    


‘‘Rules of access [to the resource]. . .accepted by the community’’ Manageably small numbers of participants Strong ‘‘mutual obligations and ties among members’’ ‘‘Participation of most resource users [or representatives] in changes to collective rules’’ and in regime management and monitoring Easy detection of violators and ‘‘technology to exclude non-members’’ ‘‘Enforcement of rules with graduated sanctions against transgressors’’ (pp. 506, 512).

The literature suggests that state intervention is often necessary. In Grafton’s (2000) words, ‘‘The institutional costs to overcome the free rider problem can often be very high, requiring some form of coordination across resource owners, users, and beneficiaries’’ (p. 513). A core insight is that often states can most effectively mitigate conflicts and promote mutual gains with a facilitative strategy, rather than public resource ownership or regulation. Given deeply embedded cultural norms of private property and public recreational access in Sweden and Maine, a facilitative strategy is intuitively plausible. Ostrom (2000) observes that, ‘‘In all known selforganized governance regimes that have survived for multiple generations, participants invest resources in monitoring and sanctioning the actions of each other to reduce the probability of free riding’’ (p. 138) Appealing to evolutionary game theory, Ostrom interprets players’ interactions in terms of distinct personality types and strategies. Rational egoists, the amoral utility maximizers familiar from neoclassical economics, seek to optimize short-term payoffs. They readily flaunt rules and free ride on others’ investments. Should they predominate, the game may degenerate to a prisoner’s dilemma. Narrow, myopic self-interest may be countered, however, by the strategies of two other types of player: conditional cooperators and willing sanctioners. Not


altruists, they attempt to secure their multi-period interests by improving collective payoffs. Cooperators follow cooperative strategies so long as enough others do so, while sanctioners accept short term costs in order to enforce cooperative game rules. Such a mix of personalities and strategies seems to be present among snowmobilers, helping to explain how governance has evolved. Social norms evolve along with governance structures and can be their ‘‘glue,’’ reinforcing external incentives and sanctions. Following Ostrom (2000), norms are ‘‘shared understandings about actions that are obligatory, permitted, or forbidden’’ (p. 143). Williamson (2000) stresses that while incentives and regulations can change behavior relatively quickly, norms evolve on a longer time scale. Indeed, America’s experience in managing motorized recreation indicates that rational egoism is tenacious (Havlick, 2002). Nonetheless, our case studies suggest that, with three decades of behavior modification, a new snowmobiling ethic is spreading, reflected in a greater sense of responsibility toward landowners, other recreationists, and nature conservation.

3. Field research methods The Swedish cases build on a foundation of tourist survey research in mountain regions, including skier – snowmobiler and landowner – snowmobiler relationships (Fredman and Heberlein, 2003; Hultkrantz and Mortazavi, 1999; Lindberg et al., 2001). Attitudinal surveys and contingent valuation experiments in Dalarna County are used to assess the extent and severity of skier – sledder conflicts and the value placed on trail separation (Heldt, 2000; Heldt and Nerhagen, 2001). This analysis is supplemented by a semi-structured key informant interview with a regional recreation official (Magnusson) and written communications from the chair of a working group on snowmobile traffic (Martinsson). Assessment of Funa¨sdalen’s trail management innovations relies on a project consultant interview (Alexandersson) and reports of the project’s initial year (Alexandersson, 2001; Bjo¨rke´n, 2001). The Maine section centers on a case study of the Rangeley Lakes region. It utilizes eight semi-structured key informant interviews plus meeting minutes


D. Vail, T. Heldt / Ecological Economics 48 (2004) 469–483

and consultant reports from a 2-year outdoor recreation planning process for the Rangeley region. The process, called ‘‘Conservation Works!’’, was initiated by the Maine Audubon Society (EPO, 2000; Irland Group, 2001; Jones, 2002). State-level informants include the Maine Snowmobile Program’s director of landowner relations (Peppard) and the Maine Snowmobile Association (MSA)’s executive director (Myers). Rangeley interviewees include presidents of the snowmobile and cross-country ski clubs (Ellis; Foltz), forest managers for large landowners (Gamble; Medina), and ‘‘Conservation Works!’’ principal consultants (Giffen; Irland).4 Field research findings are presented in a narrative and primarily qualitative form.

4. Mitigating snowmobile conflicts in Sweden’s southwestern mountains A century ago, the Swedish Tourism Association was built around a network of mountain hiking and skiing huts. Skiers, primarily well-educated urbanites, viewed the mountains as rightfully their place for exercise and contemplation. The 1975 Terrain Driving Law dramatically curtailed rights to solitude in nature, previously guaranteed by allemansra¨tten. Snowmobiling’s rapid growth and skiers’ high ‘‘sensitivity to conflict’’ have made the law controversial ever since. Mountain tourism areas also experience conflicts among snowmobilers and with landowners and environmentalists. High latitude and severe climate make Sweden’s alpine ecosystems fragile. Above tree line landscapes are open, making snowmobiles visible and audible over long distances and leading many skiers to view sledding and skiing as incompatible. Fifty-six percent of skiers surveyed in Sa¨len municipality claim that snowmobiles interfere with skiing. This understates the extent of conflict sensitivity, since ‘‘displaced’’ skiers who sought solitude elsewhere were not interviewed (Heldt and Nerhagen, 2001). This section uses two studies in mountain tourism areas and one in a working landscape to compare the effectiveness of alternative local governance mecha4 Interviews were conducted in person or by telephone, with the interview protocol pre-mailed to respondents.

nisms in mitigating conflicts. In Sa¨len, governance efforts initially stressed norms of courtesy and reciprocity to reduce snowmobiler – skier conflict. Ultimately, the state used zoning to separate trails. In Funa¨sdalen, stakeholders responded to types 1, 2, and 3 conflicts with multi-party contracting and a stakeholder-owned corporation. In both areas, motorized and self-propelled recreation are generally considered incompatible at a micro-scale, but potentially compatible at landscape scale. The third example briefly revisits Antilla’s analysis of snowmobile club –landowner contracting. The governance regimes sketched here focus on conflicts among people, but snowmobiling also degrades alpine ecosystems. Indeed, it was signs of environmental damage, publicized by the Swedish Environmental Protection Agency (EPA) and Society for Nature Conservation, that prompted official investigations of air and water pollution, fish stock depletion (via ice fishing), destruction of fragile tundra, and disruption of wildlife reproduction (SOU, 1993, 1995). Although the governance arrangements described here have reduced those impacts, continuing EPA criticism is one reason Swedish policy makers do not aggressively promote snowmobiling as a rural development engine (Martinsson). 4.1. Conflict resolution in Sa¨lenfja¨llen: from failed voluntarism to state zoning Sa¨lenfja¨llen, Sweden’s top winter tourist destination, illustrates the limitations of voluntary compliance with social norms to govern snowmobiling, particularly when sledders are heterogeneous and monitoring costs are high. Most land around Sa¨len is owned by alpine ski resorts. However, most XC skiing and snowmobile trails are public, managed by the municipality and county. Three distinct types of snowmobile actors compete with skiers for access: local club members, snowmobile rental and ‘‘safari’’ businesses, and tourist sledders. The first two can be thought of as multi-period players. Their repeated interactions and reciprocal ties strengthen incentives to make long-term agreements. Tourist sledders, in contrast, are ‘‘one-shot’’ players, in theory less easily influenced by other players and needing tangible incentives or sanctions to comply with game rules (Ostrom, 2000).

D. Vail, T. Heldt / Ecological Economics 48 (2004) 469–483

For most of the 1990s, community organizations (including public agencies, the mountain rescue service, snowmobile club, police department, large landowners, and tourism businesses) jointly implemented the norm-based regime. It emphasized mutual respect between snowmobilers and skiers and appeals to use only designated trails. Specific measures included print media, trailhead signs, and trail separation, with snowmobiles nominally banned from ski tracks (Magnusson). Skiing conditions improved somewhat, but skiers still suffered amenity losses, due both to normviolating sledders and to snowmobiling’s sheer growth in southwestern Sweden, which is within a day’s drive of Scandinavia’s four largest cities.5 Lobbying by angry skiers and conservationists led Parliament to add Sa¨lenfja¨llen to Sweden’s system of Regulation Zones in 1997, prohibiting snowmobiling except for one trail monitored by county rangers. This proved to be a ‘‘win – lose’’ approach. Two-thirds of skiers perceived conditions to be improved with zoning, but two-thirds of snowmobilers claimed that their situation was worse. Indeed, the combination of zoning restrictions and Sa¨len’s hot spot location has intensified type 1 congestion conflicts (Heldt and Nerhagen, 2001). Grafton’s criteria for effective self-governance, cited above, cast light on Sa¨len’s inability to resolve the sledder – skier conflict. The increase in one-time tourist sledders from outside the community makes it difficult to achieve voluntary compliance based on ‘‘mutual obligations and ties among members.’’ In retrospect, it would have been more effective to reinforce ‘‘moral suasion’’ with greater investment in high-quality trails, strengthening snowmobilers’ incentive to ride them. However, fund raising is difficult when tourist sledders can free ride and lack any long-term stake. As Ostrom (2000) observes, ‘‘without some form of coordination to enable all individuals to agree upon, monitor and sanction contributions to the provision of a public good, the good is underprovided’’ (p. 138). In 2001, Sa¨len’s six snowmobile rental and safari businesses overcame the problem through joint investment in trail expansion. The municipality agreed 5

Between 1985 and 2000, the fraction of adult Swedes who had snowmobiled in the mountains increased from 9.4% to 16.1% (Fredman and Heberlein, 2003).


to share costs and manage grooming. Acting as joint monopolists, they eliminated tourist sledders’ free riding by raising rental fees and putting the added revenue into trails. Under this private – public partnership, Sa¨len’s trails are improving, to the benefit of sledders and skiers alike. Yet, since snowmobile owners can still free ride, trail quality will remain less than optimal. 4.2. ‘‘Northern Europe’s finest snowmobile area’’: a governance success in the making Funa¨sdalen illustrates innovative local governance to attack types 1, 2, and 3 conflicts simultaneously. It joins several interest groups and combines land use contracts, a stakeholder-owned corporation, trail user fees, and zoning. This is Europe’s first trail system where grooming is fully funded by user fees (Alexandersson, 2001; Bjo¨rke´n, 2001). The actors and conflict patterns are similar to Sa¨len’s, but Funa¨sdalen’s greater economic reliance on XC skiing has made resolution of snowmobiler – skier conflicts a top priority. The Funa¨sdalen project is an exception to Sweden’s tendency not to promote snowmobile tourism. It aims to strengthen the economy by expanding snowmobiling and skiing compatibly. Its perceived importance is reflected in US$750,000 of seed capital for 450 km of trails, jointly financed by Sweden’s EPA and the European Union’s fund for distressed rural regions. Notwithstanding this state backing, planning, and management are grounded in the community. Local stakeholders have laid out ski and sled trails and formulated access rules, using long-term contracts with landowners. To formalize commitments, the various constituencies chartered a for-profit corporation to manage trail maintenance, fee collection, and monitoring. It is 51% owned by the landowners’ association, with area tourism businesses and the snowmobile club as minority owners. Alexandersson (2001) believes a side benefit has been to strengthen local democracy by increasing participants’ mutual responsibility and legitimating new forms of collective action. To ensure high-quality skiing and snowmobiling within one landscape, trails are carefully separated and the state has created snowmobile prohibition zones. As in Sa¨len, zoning is particularly important


D. Vail, T. Heldt / Ecological Economics 48 (2004) 469–483

to control tourist snowmobilers, who lack solidarity with other players. Snowmobilers caught violating the rules are fined US$60 (four times the daily fee), however, monitoring costs are high, so some free riding and nuisance effects are inevitable. To encourage cooperation, trail fees are kept low: US$15/day, US$45/week, and US$105/season (at 8 SEK/1 US$). By comparison, Sweden’s cheapest alpine skiing pass costs US$30/day. The fee is too low to deter peak congestion. Funa¨sdalen’s blend of infrastructure investment, user fees, landowner contracts, and zoning has significantly reduced free riding and conflicts with skiers and private landowners. Reduced riding off-trail and in ecological protection zones also mitigates environmental damages. The state facilitates these arrangements in four ways: underwriting trail investment, establishing prohibition zones, policing with county rangers, and rendering a judicial ruling that open access under the 1975 law does not hold for developed trails.

mentary support from government and tourism businesses. The clubs also promote responsible riding norms through member education and trail signs. In 2001, the state strengthened clubs’ ability to recruit members and spread norms by giving them responsibility for preparing drivers for Sweden’s mandatory snowmobile license exam. Although present arrangements improve on the pre-contract situation, they perpetuate under-investment in trails and therefore incomplete compliance with responsible sledding norms, particularly by nonclub members. This is confirmed by Anttila’s (2000) survey of Swedish Farmers Federation officials, 85% of whom acknowledge that contracts reduced farmers’ frustrations and property damage. But over 60% report continuing snowmobile problems. Given the 1975 law’s open access guarantee, the central challenge is to strengthen incentives to ride on designated trails. That requires more investment than current arrangements can mobilize.

4.3. Resolving a commons dilemma via contracting

5. Taming Maine’s snowmobilers: clubs, contracts, and renewed conflicts

Outside tourist areas, the central snowmobile conflict pits snowmobilers against farm and forest operators. Open access under the 1975 law results in externalities such as logging road obstruction, tree seedling destruction, and soil compaction. These depress landowners’ economic returns, although estimates of income losses are small: roughly US$1.5 million/year nationally for farmers and US$0.15 –.25 million for woodlot owners. Nonetheless, costs and emotional stress can be significant for individual farmers and woodlot owners (SOU, 1994). Skier conflicts and environmental damage are not severe in these working landscapes because skiing is less common, noise is buffered by forest, and ecosystems are more resilient. For a decade, snowmobile clubs have contracted with landowner associations. The potential for Paretoimproving outcomes hinges on the credibility of clubs’ commitment to confine snowmobiling to designated trails. In turn, sledders’ incentive to stay on trails depends on trail quality. Trail infrastructure is largely maintained through club dues and volunteer labor, unlike Sa¨len and Funa¨sdalen where trails are considered a quasi-public good and receive supple-

Maine’s government has had a longer and more extensive role than Sweden’s in managing and also promoting snowmobiling. Department of Conservation (DOC) intervention in the 1970s was prompted by a simmering type 2 conflict: landowners frustrated by ‘‘rogue sledders’’ threatened to gate their land. The state’s prime objective was to preserve traditional public access to private wildlands, without burdening the state’s budget or entangling government in micromanagement. State promotion of snowmobiling for its rural economic contribution came later. The DOC approached its task strategically, creating incentives for snowmobilers to organize and landowners to negotiate trail licenses. For snowmobilers the prime incentive was an integrated, high-quality trail network, funded largely from snowmobile fees fuel, and taxes. Landowners’ received no monetary rents, but gained control over trail location, exemption from legal liability, and low enforcement costs. Under scores of local licensing arrangements, the trail network has reached 20,000 km, including 4000 km of the Interconnected Trail System (ITS), snowmobilers’ ‘‘superhighway,’’ connecting Maine

D. Vail, T. Heldt / Ecological Economics 48 (2004) 469–483

with neighboring New Hampshire and Canada. Laying out the ITS necessitated state brokering of landowner –club negotiations, but on-going management is largely decentralized. The MSA has evolved into Maine’s largest sporting organization and a political force, with 280 local clubs and 15,000 dues payers (32,000 family members). The state’s landowner relations director observes, ‘‘Snowmobilers. . .learned at the very beginning that to have a trail system, they needed the support of the landowner. The clubs organize the landowner appreciation program and have been strong lobbyists for landowners in the Legislature’’ (Peppard). MSA’s public image has also been enhanced by the civic and charitable works, previously noted. These arrangements meet several of Grafton’s criteria for effective commons governance, notably ‘‘mutual obligations and ties among members’’ and participation by resource owners’ and users’ in rule making and monitoring. However, other supportive conditions, particularly ‘‘a manageably small number of participants’’ and easy apprehension of rule violators, are not fulfilled and contribute to unresolved conflicts. The incentive ‘‘glue’’ binding the arrangement is the state’s Trail Grant program, which rebates most sled registration revenues and 0.62% of state gasoline taxes to clubs. The clubs contribute volunteer labor and member dues. Between 1986 and 2000, Trail Grants increased sixfold, from US$286,000/year to US$1,787,000/year and the state’s share of trail infrastructure outlays grew from 54% to 67% (BPL, 2000). Recent Trail Grant initiatives seek to boost the rural economy by attracting more sledders, particularly non-residents who pay higher registration fees, take longer average trips, and spend more per day (Irland Group, 2001). Initiatives include supplementary trail grants to town governments and grants for bridges to upgrade the ITS and link trails to tourist towns. 5.1. Growing congestion and persistent free riding Maine’s post-1990 snowmobiling boom challenges the sustainability of current institutional and financial arrangements. Rapid growth has resulted from promotional campaigns, infrastructure development (trails, parking, sled rental operations), and changes


outside the northern forest (sprawl and low-snow winters to the south have attracted more non-resident snowmobilers). Non-resident registrations tripled to 15,000 in the 1990s. From 1971 to 1991, registrations grew by just 15%, while trail mileage doubled. By contrast, registrations surged 50% in the 1990s, with very little trail expansion. More sledders means good economic news for towns like Rangeley, located at a junction of two ITS trails. Whereas many tourist businesses formerly closed and unemployment soared in the winter, snowmobilers now inject US$5 –7 million of spending and support 75 jobs (Irland Group, 2001). Rangeley’s development goal is to re-double ‘‘sledding days.’’ Local snowmobilers, however, are ambivalent. Many feel crowded off their own trails on weekends and harassed by reckless drivers. Trail grooming costs are outstripping the budget from Trail Grants, club dues, and business contributions. Club members, whose thousands of hours of volunteer labor maintain the trails, are burning out and growing resentful of free riders (Ellis; Giffen). In Ostrom’s terms, they are ‘‘conditional cooperators’’ rethinking their strategy. Registration fees and gasoline taxes prevent pure free riding, but just one-third of snowmobile owners pay club dues and even fewer share in volunteer labor. Resident sledders arguably have a latent property right and should pay less than outsiders, yet, club members actually pay more through their dues and ‘‘sweat equity.’’6 5.2. New conflicts in landowner relations Maine’s trail network, 95% on private land, is a notable achievement in multiple-use landscape management. As in Swedish non-tourist areas, it functions like a repeated cooperative game. Licensing arrangements are sustained by the players’ reputations (trust) and low enforcement costs. However, new problems have arisen with changes in land ownership and owners’ objectives, combined with growing thirdparty effects. Trail expansion, sought by clubs and tourist towns, is threatened by emergent conflicts, both with large landowners who license most of the

6 The club and local businesses urge non-resident sledders to buy US$10 booster stickers, but most do not (Ellis).


D. Vail, T. Heldt / Ecological Economics 48 (2004) 469–483

ITS corridor and with residential owners along trails linking the ITS to tourism communities.7 Rangeley’s 250 km of trails traverse three large forest ownerships. The owners’ longstanding relationship with the Rangeley Snowmobile Club is now threatened by a third party: all-terrain vehicle drivers, who trespass on snowmobile trails and logging roads in all seasons. ATVs undercut both recreational and industrial land uses, tearing up trails and causing timber stand damage, soil erosion, and vandalism.8 One forest products corporation has passed ATV exclusion costs onto the snowmobile club, obliging it to install gates that are locked except in winter. Another owner is considering closing some trails to all recreation, a practice already common in more populous Maine regions. Neither owner is inclined to license a new expedition trail that the town and club want to relieve congestion and boost the economy (Ellis; Gamble; Medina; Myers). The state has responded with increased mediation efforts and enforcement of ATV speeding and drunk driving laws. Its strategy is to tame ATVers, as it did snowmobilers in the 1970s, by offering trail grants, safety education, and other services as incentives for club formation and self-policing. Recruitment has been limited, however: trail grants are a weak incentive for ‘‘rational egoist’’ ATVers who can trespass onor off-trail with a low probability of apprehension (Fleming, 2003; Myers; Peppard). A second emerging conflict centers on trails connecting the ITS with tourist towns. Trails near overnight destinations like Rangeley cross numerous leisure home lots, requiring complex right-of-way negotiations. Transaction costs for contracting and yearly trail relocation are rising, a situation exacerbated by ownership turnover and real estate development, as industrial landowners ‘‘cash out’’ recreationally valuable lands. Rangeley’s problem crystallized with condominium construction along a link trail. Many owners are ‘‘urban refugees,’’ intolerant of noisy, smelly snowmobiles and their sometimes rowdy owners. This parcelization and sprawling development near gateway 7 A related conflict involves the hostility of some town residents toward crowds of sledders. In one town, this precipitated a court order to re-route trails (Bell, 2002). 8 In the 1990s, ATV registrations doubled to 41,000/year and sales now outstrip snowmobiles. Symptomatic of the governance challenge, fewer than half of ATVs are state-registered. (Peppard).

towns threatens to make long-compatible land uses mutually exclusive (Ellis; Vail and Hultkrantz, 2000). 5.3. Snowmobilers, skiers, and environmentalists In contrast to Sweden’s intense sledder– skier conflicts, Maine’s principal type 3 conflict has snowmobilers on the receiving end of detrimental externalities caused by ATVers. Indeed, the media often remark on the lack of friction between snowmobilers and XC skiers, snowshoers, hikers, or dog sledders (Wonsavage, 2002). In Rangeley, key informant interviews and ‘‘Conservation Works!’’ stakeholder meeting minutes failed to uncover serious conflicts (Ellis; Foltz; Giffen; Peppard). Skiers’ low ‘‘sensitivity to conflict’’ is probably linked to their sport’s comparatively recent arrival and limited spread. In much of Maine, snowmobilers ‘‘were there first,’’ and skiers do not claim prior rights. Further, few skiers participate in multi-day expeditions (Maine currently lacks a hut-to-hut trail). Most take day outings to public lands or commercial ski tour centers, where groomed tracks are separated from snowmobile trails and engine noise is muffled by forests. Rangeley’s experience is instructive. Its ski club began as a branch of the snowmobile club, which lent it grooming equipment. Together with landowners, the clubs jointly planned 75 km of ski tracks to minimize conflicts. Safety gates are placed at trail intersections. However, both clubs have expansion plans, so future conflicts are anticipated. (Ellis; Foltz) Regional recreation planners are attempting to fashion a new multistakeholder compact building on the two clubs’ norms of trust and conflict resolution. The goal is, ‘‘to develop a more permanent system of trails for all types of activities, including hiking, skiing, horseback riding, snowmobiling, cycling and ATVs’’ (Jones, 2002). This process will test whether simultaneous expansion of motorized and self-propelled recreation is feasible at a large landscape scale. Snowmobiling’s environmental impacts, widely investigated and publicized in Sweden, have received little study in Maine. Major conservation organizations have not raised environmental objections to snowmobiling, although they lobby vigorously for expansion of motorless tracts on public land. Indeed, the Maine Audubon Society has been a prime mover in Rangeley’s multiple-use landscape planning. Climatic and

D. Vail, T. Heldt / Ecological Economics 48 (2004) 469–483

ecological differences from Sweden partially explain this complacency: There is no fragile tundra and little spring season sledding to threaten wildlife reproduction. The most noted impact is local air and noise pollution, although there is a suspicion that snowmobile access to remote ponds contributes to illegal ice fishing and game fish depletion (Giffen; Peppard). 5.4. Policy responses to emergent governance challenges For a quarter century, state-supported local governance arrangements kept types 1 and 2 conflicts in check. Conflicts with other recreationists and environmentalists were minimal. However, this institutional model appears increasingly maladapted to challenges inherent in snowmobiling’s growing popularity: hot spot congestion, free riding, and inadequate trail infrastructure. The model’s resilience is also being tested by two exogenous forces: the ATV boom and changing land ownership. The Maine Legislature has offered symptomatic relief by increasing state allocations to trail grants, policing, and safety programs (but notably not to environmental impact analysis).9 Some key informants doubt that money alone can sustain present governance arrangements. They argue for more direct state financing, management, maintenance, and policing of the trail network (Ellis; Irland).

6. Findings and discussion: evolutionary adaptation of governance regimes The Sweden and Maine case studies underscore the usefulness of conceptualizing rural landscapes as impure public goods with certain common pool characteristics. In the snowmobiling case, natural capital— the undeveloped landscape—is complementary with human-made capital—trail infrastructure. Rivalness among landscape uses, combined with legal and practical limits to excluding snowmobiles, leads to four principal types of conflict. Type 1 conflicts among 9

Legislators are also considering doubling the allocation of gas taxes to trails. Trail grants currently receive just half of sledders’ US$2 million/year gas taxes. (Rubin et al., 2001) A backlog of road maintenance and a powerful highway lobby make this an uphill battle.


snowmobilers center on reciprocal congestion externalities and free riding on trail infrastructure. Type 2 conflicts between sledders and landowners involve property damages and nuisance effects. Type 3 conflicts pit snowmobilers against other recreationists, particularly cross-country skiers in Sweden and ATV riders in Maine. And type 4 conflicts involve tensions between snowmobilers and environmentalists over detrimental effects on air and water quality, fragile terrain, and wildlife habitat. Lesser conflicts involve snowmobilers and tourist community residents and snowmobiling demands on the limited resources of Sweden’s ranger patrol and Maine’s warden service. This conflict typology sheds light on more general controversies surrounding motorized recreation in ‘‘the commons.’’ Thus, types 3 and 4 conflicts are at the heart of Sweden’s national debate over snowmobiling in the region surrounding the nation’s highest peak, Kebnekaise, as well as the Bush administration’s decisions to reduce roadless areas on US federal lands and permit snowmobiling in Yellowstone and Grand Teton National Parks. With snowmobiling, managing the commons centers on governance regimes and infrastructure investments to improve compatibility with competing land uses and values, particularly commodity production (crops, timber, reindeer), self-propelled recreation, and ecosystem health. The case studies trace the evolution of local institutional arrangements, backed by crucial state assistance but a limited state management role. Following passage of Sweden’s 1975 Terrain Driving Law, mountain tourism areas like Sa¨len tried several methods to ‘‘tame’’ snowmobilers. Moral suasion, trail separation, and selective zoning mitigated sledder – skier conflicts and ecosystem damage. However, the limited extent and quality of trails, combined with the growing popularity of snowmobiling, intensified congestion and strengthened incentives to shirk rules and norms. In contrast, the governance model being developed in Funa¨sdalen appears well designed to reduce all four types of conflict while enhancing snowmobiling’s local economic contribution. The model, institutionalized in a multi-stakeholder corporation, involves land-use planning to separate ski and snowmobile trails, zoning to protect fragile ecosystems, investment in high-quality trails, and trail fees to prevent free riding and finance grooming. The Swedish state offers crucial support in


D. Vail, T. Heldt / Ecological Economics 48 (2004) 469–483

the form of trail financing, zoning restrictions, ranger patrols, and a court ruling permitting fees on developed trails. Funa¨sdalen’s model is catching on, with two neighboring counties now seeking EU funds for their own experiments (Lindgren, 2003). For a quarter century, Maine’s snowmobile governance arrangements were tolerably effective. They centered on snowmobile club – landowner contracting and local, volunteer-based trail maintenance, with the state offering financial incentives and organizational assistance. With the rapid growth of snowmobiling since 1990, however, type 1 hot spot congestion and free riding problems have intensified. Types 2 and 3 conflicts have also worsened, as a result of changing land ownership and the explosive growth of all-terrain vehicle use. The Rangeley Lakes study indicates that the ‘‘old regime’’ is increasingly maladapted to evolving conditions. More Trail Grant funding, more wardens patrolling the trails, and tighter ATV regulation offer symptomatic relief. However, the core arrangements—local club contracting and volunteer trail labor—have become increasingly problematic with more congested trails, increased maintenance requirements, and more complicated and contentious annual licensing negotiations. It is tempting to conclude that Maine should adopt the Funa¨sdalen model, but a sustainable Maine snowmobiling regime requires more extensive and directive state intervention, for two reasons. First, Maine’s policy makers, more than Sweden’s, promote snowmobiling for its rural economic development contribution. Second, Maine’s snowmobile economy depends on a more complex, larger-scale public good, the 20,000 km statewide trail network that crosses hundreds of private ownerships. In reality, Maine’s political culture is not well-adapted to extensive state management of recreation on private lands. The DOC bureaucracy, landowners, and the Snowmobilers’ Association would probably oppose such an intrusive role (Gamble; Myers; Peppard). Nonetheless, meeting the challenges of expanding the ITS, securing connector trails, and establishing trail user fees is not likely under the present institutional arrangements. Several institutional and technical innovations might mitigate remaining conflicts in both settings. For instance, sharply differentiated trail fees might encourage more mid-week riding and less weekend

and holiday traffic. Discounted state registration fees might induce more snowmobilers to join local clubs. Zonal trail licenses might cut free riding without excessive transaction costs (Hultkrantz and Mortazavi, 1999). Differentiated registration fees could accelerate the replacement of noisy, polluting two-stroke engines with four-stroke snowmobiles. Also, on-board transponders (e.g., embedded in license plates) combined with a network of field sensors could facilitate more complete, lower cost monitoring of driving violations. The studies reported here contribute to a growing literature on institutional arrangements for ‘‘governing the commons.’’ They suggest the significant potential of local institutions, with state support, to mitigate four types of conflict linked to motorized off-road recreation. Maine’s Rangeley case reveals a need for new governance and investment approaches in the face of rapid snowmobiling growth and changing land ownership, while Sweden’s Funa¨sdalen experiment shows the potential of multi-stakeholder cooperation and multiple-use landscape management to improve snowmobiling’s compatibility with landowners’ and skiers’ interests, as well as ecosystem health. References Agrawal, A., 2002. Common resources and institutional sustainability. In: Ostrom, E., et al. (Ed.), The Drama of the Commons. National Academy Press, Washington, DC, pp. 41 – 86. Alexandersson, U., 2001. Pa˚ skoter i Funa¨sdalsfja¨llen: Slutrapport fra˚n skoterprojektet Ha˚llbar utveckling i Funa¨sdalsfja¨llen. Ha¨rjedalen Municipality, Sveg. Anttila, S., 1999. The snowmobile issue as a commons dilemma: a problem of concept formation. Report 1999:2. Fja¨llforskningsinstitutet. Mid-Sweden University. Ha¨rno¨sand. Anttila, S., 2000. Sno¨skotera˚kning pa˚ privat mark. Rapport 2000:1. ¨ stersund. Fja¨llforskningsinstitutet. Mid-Sweden University. O Anttila, S., 2001. Voluntary Provision of Snowmobile Trails on Private Land in Sweden. Unpublished. Gotland University College, Visby. Bell, T., 2002. Whose town is it anyway? Portland Sunday Telegram 3, 1K (Mar.). Bjo¨rke´n, K.-E., 2001. Ha¨rliga va¨stra Ha¨rjedalen—har Europas ba¨sta skoterledsomra˚de. Snowmobile 2, pp. 4 – 5. Bosselman, F., Peterson, C., McCarthy, C., 1999. Managing Tourism Growth. Island Press, Covelo, CA. BPL(Bureau of Parks and Reaction), 2000. Off-Road Vehicle Division Report: Current Operations and Future Opportunities. Maine Department of Conservation, Augusta. BRP (Bureau of Parks and Reaction), 1988. A Brief Explanation of the Snowmobile Program. Maine Department of Conservation, Augusta.

D. Vail, T. Heldt / Ecological Economics 48 (2004) 469–483 EPO (Environmental Policy Options, LLC), 2000. Profile of Tourists Visiting the Rangeley Area. For Maine Audubon Society, Falmouth, ME. Fleming, D., 2003 Task force to address problems with ATV’s. www.pressherald.com, 19 March. Fredman, P., Heberlein, T., 2003. Changes in skiing and snowmobiling in Swedish mountains. Annals of Tourism Research 30 (2), 485 – 488. Genthner, C., 2002. Saturday’s Ricky Craven run to benefit five charities. Maine Sunday Telegram 20, 2k (January). Grafton, R., 2000. Governance of the commons: a role for the state? Land Economics 76 (4), 504 – 517. Hanna, S., Folke, C., Ma¨ler, C. (Eds.), 1996. Rights to Nature. Island Press, Covelo, pp. 1 – 10. Havlick, D., 2002. No Place Distant: Roads and Motorized Recreation on America’s Public Lands. Island Press, Covelo, CA. Heldt, T., 2000. Cross-country skiers valuation of externalities from snowmobile. Presented at the 9th Nordic Tourism Research Conference. Bornholm, Denmark. October 12 – 14. Heldt, T., and Nerhagen, L., 2001. Turskid-och skotera˚kning i Sa¨le¨ r ‘‘Pa˚ ra¨tt led i Sa¨lenfja¨llen’’ ur led? Working paper nfja¨llen: A 2001:3 in Transportation and Tourism. Dalarna University College. Borla¨nge. Hultkrantz, L., Mortazavi, R., 1999. Landowner participation in the regulation of outdoor recreation: problems in snowmobiling regulation. Tourism Economics 4 (1), 33 – 49. The Irland Group, 2001. Rangeley region’s economy current situation and outlook. For Conservation Works! Maine Audubon Society, Falmouth. Jones, J., 2002. Conservation works! Update. Habitat, vol. 18 (4) Maine Audubon Society, Falmouth, Maine, p. 11. Knopp, D., Tyger, J., 1973. A study of conflict in recreational land use: snowmobiling vs. tour-skiing. Journal of Leisure Research 5 (3), 6 – 17. Lindberg, K., Fredman, P., Heldt, T., 2001. Skiers and snowmobilers in So¨dra Ja¨mtlandsfja¨llen: are there recreational conflicts? ETOUR Working Papers 2001:12. European Tourism ¨ stersund, Sweden. Research Institute. O Lindgrenm, M., 2003. Betalleder info¨rs fo¨r skoter Swedish Radio. http://sr.se/cgi-bin/ekot., 21 April. Manning, R., 1999. Studies in Outdoor Recreation. Oregon State Univ. Press, Corvalis. Marcouiller, D., 1998. Environmental resources as latent factors of production in tourism. Tourism Economics 4 (2), 131 – 145. Marcouiller, D., Green, G., 2000. Outdoor recreation and rural development. In: Machlis, G., Field, D. (Eds.), National Parks and Rural Development. Island Press, Covelo, pp. 3 – 50. Mazzone, P., 2001. Snowmobile Town. Maine Times 27, 24 – 27 (December). McIntyre, N., 1999. Internationella Tendenser. In: Sandell, K., So¨rlin, S. (Eds.), Friluftshistoria. Carlssons Fo¨rlag, Stockholm, pp. 236 – 259. Ostrom, E., 2000. Collective action and the evolution of social norms. Journal of Economic Perspectives 14 (3), 137 – 158. Prato, T., 2001. Modeling carrying capacity for national parks. Ecological Economics 39, 321 – 331. Reiling, S., 1998. An Economic Evaluation of Snowmobiling in


Maine. Maine Agricultural and Forest Experiment Station, Orono. Publication 2281. Rubin, J., Hart, S., Morris, C., 2001. Gasoline Consumption Attributable to Snowmobile Use in Maine. Margaret Chase Smith Center, University of Maine, Orono. SCB (Central Statistical Bureau), 2000. Fordon enligt bilregistret, ¨ rebro. Statistiska meddelanden TK20, table 4. O Seidl, I., Tisdell, C., 1999. Carrying capacity reconsidered: from Malthus’ population theory to cultural carrying capacity. Ecological Economics 31, 395 – 408. SOU (Offical State Investigation), 1993. Naturupplevelser utan buller. Slutbeta¨nkande fra˚n utredningen om a˚tga¨rder mot buller i fja¨llomra˚den och ska¨rga˚rdar. Report 1993: 51 Stockholm. SOU, 1994. Skoterko¨rning pa˚ jordbruksmark och skogsmark. Report 1994: 16. Stockholm. SOU., 1995. Ha˚llbar utveckling i landets fja¨llomra˚den. Beta¨nkande av Miljo¨va˚rdsberedningen. Report 1995: 100 Tietenberg, T., 2003. Environmental and Natural Resource Economics, 6e Addison Wesley, Boston. Vail, D., 2001. Engines in the wilderness: governing motorized recreation in Maine’s North Woods. Presented at the US Society for Ecological Economics Conference, Duluth. Vail, D., Heldt, T., 2000. Institutional factors influencing the size and structure of tourism: Dalarna (Sweden) and Maine (USA). Current Issues in Tourism 3 (4), 283 – 324. Vail, D., Hultkrantz, L., 2000. Property rights and sustainable nature tourism: institutional adaptation and mal-adaptation Dalarna and Maine. Ecological Economics 35 (2), 223 – 242. Williamson, O., 2000. The new institutional economics: taking stock looking ahead. Journal of Economic Literature. XXXVIII (3), 595 – 613. Wonsavage, D., 2002. An outdoor Kinship links Snowmobilers, nordic skiers. Maine Sunday Telegram. 13 Jan.

Key Informant Interviews (winter, spring 2001) Alexandersson U. Consultant. Ha¨rjedalen Municipality. Sveg. Foltz, J. President, Rangeley Lakes Cross Country Ski Club. Rangeley. Ellis, K. President, Rangeley Lakes Snowmobile Club. Rangeley. Gamble, G. New England Lands Manager. Mead Paper Corporation. Rumford. Giffen, A. Project director, Rangely Lakes, Conservation Works! Land and Water Associates. Hallowell. Irland, L. Principal, The Irland Group. Consultants to Conservation Works! Winthrop. Magnusson L.-A. Regional forest manager. Transtrandsfja¨llen. Sa¨len. Martinsson, T. Working group chair, Snowmobile traffic in Sa¨len. Malung Municipality. Medina, S. Regional forest manager. Seven Islands Land Company. Bangor. Myers, B. Executive Director, Maine Snowmobile Association. Augusta. Peppard, Sgt. D. Director of Landowner Relations. Maine Department of Inland Fisheries and Wildlife. Augusta.

Essay III

The Value of Reducing Recreation Conflict: A Choice Experiment Evaluation of Snowmobiler-Skier Conflict1 by Kreg Lindberga, Tobias Heldtb and Peter Fredmanc

1. Introduction Public lands in the US and abroad are visited by a range of outdoor recreation user groups, including both "traditional" groups like hikers and cross-country skiers and "new" or nontraditional groups like mountain bikers and snowmobilers. Though conflicts can exist within a given group, or between traditional groups, they more commonly arise from the introduction or expansion of new, typically motorized or mechanized forms of recreation. The behavior or simple presence of these new user groups may interfere with the experience of traditional visitors. The resulting conflict, combined with negative impacts on the natural environment, frequently lead to calls for separating new users from traditional users or for banning them altogether (Manning 1999; Jackson et al. 2002). Understandably, the new users resist access limitations, so land management agencies (and politicians) are faced with a difficult decision. Snowmobiling in Yellowstone National Park is a recent and high-profile example of this phenomenon. As part of its decision-making process, the National Park Service (NPS) correctly notes that management interventions, such as a ban, generate both positive and negative impacts, and that these impacts should be evaluated and


Submitted to Forest Science, 23 March, 2004, invited resubmission Sept. 14. a) Oregon State University, USA, b) Dalarna University and Uppsala University, Sweden, c) European Tourism Research Institute, Sweden. The authors thank John Loomis, Benedict Dellaert, Chuan Zhong Li and Lars Hultkrantz for comments on the manuscript, Dan Williams for comments on the survey instrument, and Troy Hall for input on recreation conflict and related topics.


compared to the extent possible (MACTEC, BBL, and RTI 2002) 2. However, NPS' ability to do so depends on its ability to measure the impacts of interventions and to assess the value of these impacts. Unfortunately, relevant data is sorely lacking, and evaluators are in the unenviable position of making arbitrary assumptions about the effect of changing snowmobile regulations on cross-country skiers.3 The present study estimates the effect of snowmobile access restrictions on cross-country skier (hereafter "skier") visitation rates and consumer surplus, using a case study from Sweden. Although this study preceded the Yellowstone evaluation, that evaluation highlights its relevance. Through benefits transfer, study results can inform analyses and decisions, such as those undertaken for Yellowstone and other public lands. More generally, the study illustrates the use of choice experiments for such purposes and enhances understanding of how conflict affects the value recreationists place on their experience. In addition, the study assesses the presence of protest behavior in choice experiments. The latter issue is relevant to choice experiment applications in recreation and other contexts. The paper begins with a description of recreation conflict and a discussion of economics as a tool for understanding and managing conflict. The relevance of choice experiments as a means of measuring economic value is especially discussed. This introductory section is followed by a methods section, including a description of recreation conflict in context of snowmobiling in the 2

In November 2000 NPS decided to phase out snowmobile use in Yellowstone over three years and enforce a total ban beginning the winter of 2003. The International Snowmobile Manufacturers Association, the State of Wyoming, and others brought a lawsuit against the decision, and an agreement was reached to delay implementation pending further analysis and public consultation. During this process, 360,000 public comments were received, of which 80 percent favored banning snowmobiles (Seelye 2002). Despite these comments, the NPS ultimately chose not to ban snowmobiles but rather to limit the number of entries, require a trained guide, and require the best commercially available technology (e.g., cleaner and quieter four-stroke engines). The NPS outlines several concerns with snowmobile use in Yellowstone, including 1) nonmotorized users expressing dissatisfaction with the sound, odor, and quantity of snowmobiles, 2) public safety concerns due to the level of snowmobile accidents, unsafe users, and inherent winter risks, and 3) negative resource effects from snowmobile emissions, harassment and unintended impacts on wildlife from groomed trails and their use, degradation of air quality values, and impacts on the natural soundscape (USDI NPS 2002:v). 3

In the Yellowstone case, two assumptions are that 1) snowmobile access restrictions will not affect the frequency of non-snowmobile visitation and that 2) non-snowmobile visitors will experience a five percent reduction in consumer surplus due to the one-year delay in restricting snowmobile access (MACTEC, BBL, and RTI 2002:3-13, 3-54).


Swedish study area as well as descriptions of data and data collection. The paper ends with sections presenting results and discussion.

2. Managing recreation conflict by tools of economics 2.1 Recreation conflict Researchers have defined recreation conflict in various ways, but perhaps the most common and basic is "goal interference attributed to others" (Jacob and Schreyer 1980; Manning 1999). This concept is based on expectancy theory, recreation motivations, and discrepancy theory. In brief, people engage in recreation activities to achieve certain goals, such as solitude, exercise, or escape from the stresses and environment of everyday life. The presence or behavior of other recreationists may lead to a discrepancy between these desired or expected goals and those that are actually achieved, which may lead to dissatisfaction. For example, Jackson and Wong (1982) found that skiers rated "feeling a part of nature," "tranquility," and "solitude" as important motivations, and the experience of such goals may be jeopardized by the presence of snowmobiles. Indeed, when asked in that study to identify what they disliked about snowmobilers, 85 percent of respondents noted sentiments like "the noise and smell of the machines destroys the natural setting" (1982:55). As the authors note (1982:59), "skiers choose their activity precisely for the reasons which make it susceptible to impact, whereas snowmobilers choose theirs precisely for the reasons which may generate those impacts." Conflict may exist even when the dissatisfied user does not come into direct contact with "offending" users or the tangible results of their presence, such as noise generated by snowmobile engines (Blahna et al. 1995; Carothers et al. 2001; Vaske et al. 1995). Conflict may arise from "indirect contact" and reflect social values, including the appropriateness of selected recreation activities. For example, a skier may experience conflict simply because she feels snowmobiling is not an appropriate use of the area. In addition, based on past interaction or stereotypes, she may not like snowmobilers. In other words, the simple presence of users and/or uses in an area may generate conflict. 3

Considering both direct and indirect contact as sources of conflict, it is not surprising that conflict is more common between different types of recreationists (inter-group conflict) than between recreationists engaging in the same activity (intra-group). In addition, research indicates that conflict is often asymmetric, with many recreationists engaged in traditional activities perceiving conflict caused by those using newer technology, but the reverse being less common (Vaske et al. 2000). In the present context, skiers are more likely to dislike encountering snowmobilers than are snowmobilers to dislike encountering skiers. There is a relatively long history of research on visitor response to crowding, with a focus on three behaviors: displacement, product shift, and rationalization (Manning 1999). The former is a behavioral response involving substitution to alternative sites, while product shift and rationalization are cognitive responses. The management response depends in part on the source of conflict. When it results from direct contact, spatial (zoning) or temporal separation of conflicting uses can be an effective approach. However, if the source is indirect contact, separation may not resolve it. Educational programs can also be effective, either through establishing basic etiquette that might lessen conflict or, of particular value for indirect contact, by increasing tolerance for other types of users and activities (Manning 1999; Ramthun 1995; Vaske et al. 1995). A ban on access by the offending user group is a more dramatic alternative, one that may be necessary if education is not feasible or effective and/or if that user groups also generates unacceptable impacts on the natural environment. 2.2 Dimensions, trade-offs, and the relevance of economics In situations of recreation conflict or crowding, the focus typically is on how management intervention can lead to gains for included visitors (those not excluded due to access restrictions) and the natural environment (and ultimately to those in society who value reduction of negative environmental impacts). Less attention is paid to the losses resulting from these interventions, or the financial cost of implementing them. One would expect that interventions would occur only when the gains outweigh the losses (Lindberg et al. 1997: 463), so the inattention paid to the balance between the two is surprising. A likely cause is the difficulty of quantifying the impact of intervention, how impacts affect the various stakeholders, and whether, overall, the effects are 4

more positive than negative. The following briefly describes a simplified typology of impacts, how they affect stakeholder groups, and the relevance of welfare economics for measuring and comparing these effects.4 A ban on snowmobile access is used as illustration, and the primary non-snowmobile group is assumed to be skiers.5 First, included visitors, in this case skiers, may increase in number and may enjoy a higher-quality experience due to the elimination of conflict associated with snowmobiling. Although there have been several studies of recreation conflict, including some relating to snowmobiling, the Yellowstone analysis illustrates the lack of data on how conflict reduction affects visitation rates and the value of the skier recreation experience. Second, excluded visitors (snowmobilers) may lose because they need to forego the experience or find a substitute site, which may be of lower quality and/or more expensive to access. The extent of their loss depends on the value placed on the pre-ban experience and the availability and quality of substitutes. In addition, if displacement to substitute sites leads to crowding at those sites, both the new (displaced) and old snowmobilers may experience a loss. Third, local communities may be affected beyond effects accruing to residents as included or excluded visitors. Depending on the local availability of substitutes, the ban may lead to a decline in snowmobiler visitation. If this decline is not compensated for by an increase in skier or other user groups, there will be an overall decline in visitation and (potentially) the local economic benefits associated with it, such as producer surplus, tax revenue, and personal income. The welfare effects of these impacts will depend on factors such as the availability of alternative employment and reservation wages. Matthews (1999a, b) reports that implementation of the snowmobile ban in proposed wilderness and roadless management areas in the Lolo National Forest was estimated by the local chamber 4

Welfare economics refers to the economic theory concerned with the welfare of society and priorities to be observed in the allocation of resources. 5

As the Yellowstone example illustrates, other non-snowmobile visitors may include snowshoers and those travelling by other forms of motorized recreation, such as snow coach and car.


of commerce to cost Superior, Montana up to $500,000 in lost economic impact, of which producer surplus and local income would be a portion (other areas of the forest remain open to snowmobiling, and Forest Service officials estimate the lost economic impact at just $44,000). In their Yellowstone evaluation, the NPS estimated that a snowmobiling ban would lead to a reduction of $21.1 million in output and 499 jobs in the 5-county Greater Yellowstone Area. This represents less than a 1% decline in both variables (USDI NPS 2002), though the relative impact would be greater in the gateway community of West Yellowstone.6 Fourth, a ban would reduce negative impacts on human health and the natural environment, which in turn would increase the welfare of citizens who value such a reduction. In addition, those who oppose snowmobile use in some or all natural areas on symbolic grounds would experience a welfare gain. Conversely, those who oppose access restrictions in principle or who value maintenance of local economies would experience a welfare loss (Berrens et al. 1998; Xu et al. 2003).7 Impacts, and associated values, have a spatial dimension. If snowmobilers substitute to, and skiers substitute from, alternate sites, impacts may simply be re-distributed. These may lead to important impacts at the level of local communities and recreation sites without affecting the state or national level. Assuming that the local level is of importance and/or that close substitutes do not exist, the respective gains and losses may be substantial. There has been increasing recognition of tradeoffs within and across dimensions and stakeholder groups (Hall 2001; Lawson and Manning 2002; Manning 1999). Nonetheless, frequently it simply has been assumed (often only implicitly) that the gains outweigh the losses, as reflected by the sentiment that turning some visitors away “would be the price paid for higher-quality experiences by all those 6

McManus, Coupal, and Taylor (2001:13) estimate that a ban on snowmobiling in Yellowstone (and Grand Teton) could decrease snowmobile expenditures in Wyoming by up to $36.8 million. Regarding the general economic importance of snowmobiling, Bell (2002) reports that snowmobile tourists bring $300 million annually to the state of Maine, while overnight snowmobiler trip expenditure in Canada is estimated at C$121.9 million in 1999 (Pannell Kerr-Forster 2001). 7

At the institutional level, land management agencies also have made local economic issues a priority. For example, in his 2000 “State of the Forest” speech, former USDA Forest Service chief Mike Dombeck noted that one of the six key areas for his agency’s recreation strategy is to “develop business opportunities for under-served and low income communities” and that “[t]he objective of our recreation strategy is to expand upon recreation opportunities in a manner that benefits local communities” and improves landscape health (Dombeck 2000:8).


receiving permits” (Hendee and Lucas 1973:207). The challenge is to determine whether this assumption is accurate, whether the balance of gains and losses (financial and otherwise) is, overall, positive. The metric of economic value and the goal of maximizing total value across stakeholders is a means for determining whether access restrictions do lead to an increase in societal well-being. It is recognized that this economics approach may not capture all the values relevant to the issue, and that it rests on the assumptions inherent in economic valuation (Driver et al. 1991). Nonetheless, it can provide important insight into the magnitude of costs and benefits from management regulations such as numerical restrictions. The relevance of economics to recreation management decisions has long been recognized (Wagar 1974; Fisher and Krutilla 1972). Loomis (1993) illustrates its application to the context of congestion (crowding), a popular context for economists (Cicchetti and Smith 1973 is an early example, while Jakus and Shaw 1997 provide a recent overview).8 2.3 The choice experiment approach Although conflict is related to congestion, economic evaluations of recreation conflict have not been found in the literature. The present research provides estimated welfare changes resulting from progressive conflict reductions that may stem from management actions ranging from spatial separation to a snowmobile access ban. The benefits of these actions reflect different causes of conflict, notably direct and indirect contact. Choice experiments, included in surveys completed by skier visitors, were used to estimate the effect of conflict reduction on both market share (level of visitation) and willingness to pay. Choice experiments (CE) are a type of conjoint analysis and one of several stated preference (SP) approaches, of which contingent valuation (CV) is perhaps the best known (Louviere et al. 2000; Haab and McConnell 2002). In simplified terms, dichotomous choice CV asks respondents 8

Economics also provides a conceptual framework for understanding the cause of conflicts and potential responses. The disturbance caused by snowmobiling is a negative externality that does not enter into snowmobiler decisions regarding how often, where, and in what manner to go snowmobiling. In practice, it is difficult to internalize such


to decide whether they are willing to trade a specified amount of money (a variable level of one attribute) for a specified good or service (a fixed level of a second attribute, such as access to a recreational resource or a quality enhancement to that resource). In conjoint models, respondents are asked to rate, rank, or choose amongst multiple alternatives, with the latter being choice experiments. Each of the alternatives is characterized by multiple attributes with varying levels (e.g., levels of entry fees and levels of snowmobile presence), and each set of alternatives presented to respondents is called the choice set. In the context of outdoor recreation, respondent choices enable analysts to evaluate how visitors react to changes in attribute levels and thus to predict market share amongst locations given the levels of each attribute they contain. For example, Haider and Ewing (1990) evaluated the effect of price, distance to beach, and other variables in respondent choice amongst hypothetical Caribbean destinations. When a monetary variable, such as entry price, is included as one attribute, choice experiments can be used to estimate willingness to pay for changes in other attributes.

3. Methods 3.1 Study area The current study was conducted in the Swedish mountain region, which is attractive for both snowmobiling and cross-country skiing. The topography is gentle, there are vast areas of thin forests, and, since the timberline is at only 600 to 900 meters above sea level, the bare mountains are easy to access and experience. The winter season usually lasts between November and April, and there is an extensive system of recreation facilities, including some 8,000 kilometers of managed trails. Some of these trails are open to both snowmobilers and skiers, while others are restricted to skiers only.

external costs into snowmobiler decision-making, so access restrictions and other management interventions are used.


Climatic conditions and the long tradition of non-motorized outdoor recreation contribute to conflict between snowmobilers and skiers. The country's northern location leads to the low treeline, above which much winter recreation occurs. In that setting, snowmobiles are both visible and audible over long distances (Vail and Heldt 2004). In addition, a network of back-country huts and lodges was established by the Swedish Touring Association (STF) at the turn of the 20th century. The mountains have been viewed as a location for health-oriented exercise and contemplation for hikers and cross-country skiers, primarily from the urban areas in the south. Access to both public and private land was, and still is, guaranteed through the Right of Common Access (Allemansrätten). Technical and economic development made snowmobiles available to a broader public in the 1970s. The 1975 Terrain Driving Act provided for snowmobile access to natural areas, though 12 "snowmobile prohibition zones" were established in remote mountain regions. The ensuing conflict between skiers and snowmobilers has been a source of controversy ever since (Abrahamson 1998 provides more detail on the cultural, historical, and legal context). Between 1980 and 2000 the number of visitors who reported snowmobiling in the mountains almost doubled, from 550,000 to 1 million, and at the end of the 1990s one quarter of all Swedish winter tourists traveling to the mountain region went snowmobiling (Heberlein et al. 2002; Fredman and Heberlein 2003). A slightly larger percent went cross-country skiing on day trips, while only two percent went skiing on overnight trips. Skiing has been a popular activity for the past hundred years, but participation has remained stable for the past 20 years. The traditional approach for reducing conflict has been zoning, the spatial segregation and banning of snowmobiling in certain areas. The number of "snowmobile prohibition zones" in the mountain region has grown to 16, though this growth has been motivated more by concerns about protecting fragile environments and endangered species than by recreation conflict. Indeed, within some prohibition zones snowmobiles are required to use designated trails, shared with skiers, rather than being excluded entirely. In other areas, ski and snowmobile trails have been separated to reduce the conflict. Trail separation has not been widely applied to date, but managers increasingly are considering it as a means of dealing with conflict. Snowmobile use is 9

restricted in most of the southern Jämtland mountain area (Södra Jämtlandsfjällen), the geographic focus of this study. Conflicts have been reported on some trails open to both snowmobilers and skiers, and the local management agency (the regional county board) is considering trail separation in order to reduce conflict. 3.2 Data collection Data for this study was collected in the southern Jämtland mountain region9 in central Sweden, bordering Norway to the west. In the area, there are three mountain stations and six lodges managed by the Swedish Touring Club (STF), and they provide overnight facilities, food, and other supplies. There are also several private hotels and cabins located just outside the area, and one of Sweden’s largest downhill ski areas is about ten kilometers away. This is a major recreation area, one that includes the densest network of publicly-managed trails in the Swedish mountains, with about 500 kilometers of marked summer and winter trails and about 200 kilometers of snowmobile trails. While snowmobile use is permitted on several trails in the southern and eastern parts, areas to the north and the interior are not open for snowmobiling. The only exceptions are for maintenance of the STF cabins in the area, reindeer herders, and park managers. Most of the trails in the area are self-made, rather than groomed, and skiers thus may perceive benefit from snowmobile use of the area, as this use may help pack snow and create trails. The overall project involved a survey of both local residents and visitors, with separate surveys for skier visitors and snowmobiler visitors. The analysis discussed here relates only to the skier visitor survey, which was administered in two steps. The first step involved contacting visitors on-site at trailheads and rest-huts. Every visitor encountered by the interviewers was asked to complete a one-page survey, which included recording of names and addresses to receive the longer mail survey. A sub-sample of the surveys was conducted in Storulvån, a part of the area in


The total area is about 2 300 km2, runs approximately 40 km from north to south, and consists mostly of bare mountains and forested mountain valleys. The landscape is diverse, with the highest peaks being more than 1 700 m above sea level. The area contains Sweden’s most southern glacier and is one of the most important habitats for threatened animals like gyrfalcon, golden eagle, great snipe, wolverine, and lynx.


which snowmobiling is not allowed. Some of the following results are differentiated across the Storulvån and main samples. The surveys were pre-tested in early March, 2001 and administered on-site during the Easter period, from 1 April to 11 April. The mail surveys were sent on 8 May. A reminder letter was mailed 17 May, and a second letter and replacement survey was mailed 29 May, in each case only to those who had not returned completed surveys by those dates. Some visitors were missed because the interviewer was busy approaching other visitors. However, of the 565 persons contacted on-site, only 24 (4%) refused to participate in the survey. An additional 36 (6%) completed the on-site survey but refused the mail survey. Of the 505 mail surveys sent, six (1%) were returned as undeliverable. A total of 374 mail surveys were completed, which represents 75% of the mail surveys delivered and 66% of on-site contacts. The CE scenarios were contained in the mail survey. A fractional factorial design was used to create 16 scenarios (out of 256 possible combinations). Each survey containing four scenarios, so four survey versions were created and administered on an alternating basis. The five returned surveys with no or only one CE scenario completed were removed from the dataset, leaving 1,476 scenarios (369 * 4)10. Of these, there were eight individual scenarios that were not completed, leaving 1,468 observations for analysis. In the scenarios, two trails were described in terms of distance from residence (home or lodging), the cost of using the trail (in the form of a daily pass fee), the presence or absence of shelters, the quality of the scenery, and the degree of interaction with snowmobiles. Descriptions of the CE attributes and their levels are given in Table 3.1.11 10

In the analysis we follow current practice and treat the 1476 choices as independent observations. However, multiple responses from an individual in valuation exercises has shown to be correlated, which leads to overstating the efficiency in welfare estimates. Although recent, not unquestioned (Vossler and Poe 2005) findings suggest that multiple responses can be treated as independent draws (Alberini et al. 2003). One alternative would have been to estimate with a random effects probit model. However, such a model imposes the assumption of equal variance across choices for the same individual. In practice this implies no learning or fatigue effects and preference stability over choice sets.


The kronor-dollar exchange rate has varied widely during recent years, but 8 kronor (kr) to 1 US dollar is indicative.


Table 3.1 – Scenario attributes and levels (the base level for ordinal variables is bolded) Attribute Levels Distance from residence 4 levels: 0 km (direct access), 3 km, 10 km, 25 km Daily fee (use pass) 4 levels: 0 kr (free), 20 kr, 50 kr, 75 kr Wind shelters 2 levels: none, several Scenery 2 levels: average, nicer than average Snowmobile presence 4 levels: 1) share trail 2) hear, but don't see or smell snowmobiles 3) don't see, hear or smell snowmobiles, but snowmobiling allowed in area 4) snowmobiling not allowed in area

For each scenario, respondents were asked to consider the two trail options and report which of the trails they would prefer. If respondents would prefer not to go skiing rather than choose one of the two presented trails, they would select "don't go skiing". Sample scenario wording, including the introduction to the CE task, is provided in Appendix 1. Note that the scenarios did not refer to a specific location, with associated features of proximity and previous experience. This may increase respondent willingness to substitute away from a scenario site relative to an actual site. 3.3 Analytical model Conjoint analysis has long been used in the transport and market research, but it has also been used within leisure and tourism (see Louviere and Timmermans 1990 for an early overview; Hanley et al. 2002), environmental valuation (see Hanley et al. 2001; Layton and Brown 2000), and forest management (Boxall and Macnab 2000; Dennis 1998; Xu et al. 2003). Comparisons across SP methods (including CE) and between SP and revealed preference (RP) methods suggest a common underlying preference structure and estimates of willingness to pay (Cameron et al. 2002; Carlsson and Martinsson 2001). Hanley et al. (2001) discuss CE's potential advantages and disadvantages relative to the more common CV approach (see also Bennett and Blamey 2001; Boxall et al. 1996). CE is well suited for situations in which changes, and trade-offs between them, are multidimensional. For example, the results from the present study can be used not only to estimate the marginal 12

monetary value of reduced snowmobile presence (across multiple levels), but also of distance, presence of shelters, and quality of natural scenery. In addition, the value of reduced snowmobile presence can be traded not just with a monetary measure but also with the other attributes. Relatedly, CE provides more information on preferences across payment amounts, as respondents typically answer multiple choice sets and each set has multiple payment values; this can reduce the required sample size and thus survey cost relative to CV. Moreover, CE generally avoids explicit elicitation of willingness to pay, which may reduce CV problems such as protest no votes, strategic behavior, and yea saying. CE's main weakness is its cognitive difficulty, as respondents are asked to make multiple choices, each of which is fairly complex. CV may also be better suited to deriving values for a sequence of elements and for a whole program comprising multiple parts. CE also shares some of CV's weaknesses, including the potential for sensitivity to study design and hypothetical bias (Mitchell and Carson 1989). The basic CE model structure uses random utility theory (Small and Rosen 1981; Ben-Akiva and Lerman 1985) to relate the probability that a certain alternative is chosen to i) the characteristics of the alternative (e.g., price and snowmobile presence), ii) competing options, and iii) characteristics of the individual (e.g., income). It is assumed that respondent preferences or utilities for an alternative can be represented as a weighted sum of their preferences associated with each characteristic of the alternative. To allow for errors in the choice process as well as in the modeling process, utility in the model consists of i) the systematic utility the respondent derives from the alternative, explained by the characteristics of the alternative and the individual, and ii) a random error component. The utility of alternative i out of a choice set with I alternatives is given by (subscripts for specific individuals are omitted):

U i = Vi + ε i



Where Ui is the utility of alternative i, Vi is the systematic component of the utility function, and εi is the random error component. V is characterized as:

Vi = β k X i + ε i


Where β is a vector of coefficients and Xi is a vector of attributes associated with alternative i. The probability of choosing i is:

P{i I } = P{Vi + ε i ≥ V j + ε j ; ∀j ∈ I }


Assuming that the error terms are independently and identically distributed extreme value type I with scale parameter μ, the probability of choosing alternative i is the conditional logit model:

P{i | I } =

exp( μVi ) ∑ exp( μV j )



With single data sets, as in this analysis, the scale parameter μ is typically assumed to equal one.

3.4 Data analysis The CE data were analyzed using the conditional logit routine in Limdep (version 8.0). Distance and cost were specified in linear form. The remaining attributes are ordinal and were specified using effects codes, with the highest attribute level serving as the base and coded as “-1” (see Boxall et al. 1996:249 or Louviere et al. 2000:86-87 for details on this approach). Utility functions were specified for the two trails and the "don't go" options. Alternative specific constants (ASCs) commonly are specified in utility functions to measure "brand effects" – the preference for a specific option beyond its characteristics as described by the attribute levels. Because the trails used in this study were generic, there is no brand effect, and the same ASC is 14

used for both. This intercept measures respondent preferences for going skiing rather than not going.

4. Results In addition to the CE scenarios, survey responses provide data on the magnitude of encounters with, and disturbance by, snowmobilers. In the on-site survey, only five percent of respondents in the main sample (excluding Storulvån) reported not seeing or hearing snowmobilers during their visits to the area in the past three years. Forty-three percent had seen or heard snowmobilers a few times, 39% fairly often, and 10% very often. Though respondents in the Storulvån sample reported fewer encounters, many did see or hear snowmobilers, presumably "official" snowmobiles used by park managers or to service huts in the area. Thirty-one percent of those in the main sample reported that their experience was not disturbed by snowmobilers or their behavior; that is, that they did not experience conflict. Thirty-three percent reported that they were a little disturbed, 17% fairly disturbed, and 7% very disturbed. Respondents in the Storulvån reported slightly greater disturbance.12 Respondents that reported being at least a little disturbed were asked how the disturbance would affect their future visits to the region. Almost a third (29%) would visit the region but select a different trail, while five percent would visit a different region. Skiers were also asked to evaluate various aspects of snowmobiling, on a 1 to 5 scale, with 1 being very negative and 5 being very positive. Table 4.1 shows results separately for the main and Storulvån samples, with responses of 1 or 2 combined as negative, 3 as neutral, and 4 or 5 as positive. Responses are similar across the samples, with only one of the six aspects being rated differently, based on significance at p = .10 using a χ2 test. The first three items in Table 4.1 measure direct contact while the fourth item measures indirect contact. Results indicate that snowmobile exhaust and noise are the most disturbing aspects. There is recognition that snowmobiling in the area provides some benefits, including assistance in the case of accidents 12

Similar questions were asked in the mail survey, with the most recent trip used as a reference point rather than all visits within the past three years. Those results, as well as others from the survey, are available from the authors.


and breaking of trails that skiers can use. Respondents were largely neutral with respect to the indirect contact measure "knowing that snowmobiles are in the area…". The statistically significant difference between the two groups suggests that the item did measure indirect contact, insofar as skiers in Storulvån rated that item more negatively than did other skiers (the absence of snowmobiles may be a reason for people to go to Storulvån). However, the fairly large percentage rating this item positively may indicate that it is confounded with other aspects, such as safety or the benefits of having the cabins supplied by snowmobiles.

Table 4.1 – Skier evaluation of various aspects of snowmobiling (percent of sample)

Main sample Seeing snowmobiles Hearing snowmobiles Smelling snowmobiles Knowing that snowmobiles are in the area, even if one doesn't see, hear, or smell them a Being able to use snowmobile tracks Knowing that there are "private" snowmobiles in the area in case of accidents a Chi-square: p