Tourism impact modeling for resource extraction regions

Pergamon www.elsevier.com/locate/atoures

Annals of Tourism Research, Vol. 27, No. 1, pp. 188±202, 2000 # 1999 Elsevier Science Ltd. All rights reserved Printed in Great Britain 0160-7383/99/$20.00+0.00

PII: S0160-7383(99)00064-X

TOURISM IMPACT MODELING FOR RESOURCE EXTRACTION REGIONS Janaki R. R. Alavalapati University of Florida, USA Wiktor L. Adamowicz University of Alberta, Canada Abstract: A simple two sector and two factor general equilibrium model is developed to study the interactions among tourism, other economic sectors, and the environment. Tourism is considered an endogenous activity and modeled as a function of prices and environmental damage. Two types of the latter functions are considered. It is assumed that the damage occurs from activities related to the resource sector and that activities from both the resource and the composite tourism sector affect the environment. Simulation results suggest that impacts of policy changes differ with the type of environmental damage function but that the integration of environmental linkages into economic impacts models may reveal signi®cant differences in results. Keywords: computable general equilibrium, environmental damage, regional economy, modeling. # 1999 Elsevier Science Ltd. All rights reserved. Âsume Â: Le modelage de l'impact du tourisme pour les reÂgions d'extraction de ressources. Re Un simple modeÁle d'eÂquilibre geÂneÂral aÁ deux secteurs et aÁ deux facteurs est deÂveloppe pour Âetudier l'interaction parmi le tourisme, d'autres secteurs Âeconomiques et l'environnement. Le tourisme se consideÁre comme une activite endogeÁne et comme une fonction de prix et de deÂteÂrioration environnementale. On examine deux sortes de fonctions de cette deÂteÂrioration. On suppose que les dommages sont causeÂs par des activiteÂs lieÂes au secteur des ressources et que les secteurs du tourisme composite et des ressources ont un effet sur l'environnement. Les reÂsultats de simulation suggeÁrent que les impacts des changements politiques varient selon le type de fonction des dommages environnementaux mais que l'inteÂgration des liens environnementaux dans des modeÁles d'impacts Âeconomiques peuvent Âs: Âequilibre geÂneÂral quanti®reÂveÂleÂr des diffeÂrences signi®catives dans les reÂsultats. Mots-cle able, dommages environnementaux, Âeconomie reÂgionale, modelage. # 1999 Elsevier Science Ltd. All rights reserved.

INTRODUCTION In the recent past, concerns for the environment have prompted many researchers to model the economy-wide effects of environmental policies related to agriculture and other resource activities (Bergman 1995; Merri®eld 1988; Tobey and Reinert 1991; Tsigas,

Janaki Alavalapati is Assistant Professor of School of Forest Resources and Conservation, University of Florida (Gainesville FL 32611, USA. Email ). He earned his Ph.D. from University of Alberta. His research interests are forest economics and policy, impact analysis, trade and environment, and agroforestry. Wiktor Adamowicz is Professor in the Department of Rural Economy, University of Alberta. He received his Ph.D. from University of Minnesota. His research interests are environmental economics and valuation, forest economics, and econometrics.

188

ALAVALAPATI AND ADAMOWICZ

189

Gray and Krissoff 1996). However, very few studies have been conducted to model the impacts of environmental policies related to tourism activities. One reason is that many people think the industry does not harm the environment (UNEP 1992). But on the contrary, its components include transportation, accommodation, food services, and retail activities (Peter 1990; Smith 1989), and processes related to them may negatively impact the environment (Healy 1994; Inskeep 1987; Mlinari 1985). Many researchers have estimated the economic effects of tourism on regional or national economies (Adams and Parmenter 1995; Archer and Fletcher 1996; Johnson and Moore 1993; Lindberg and Johnson 1997; Wagner 1997; Zhou, Yanagida, Chakravorty and Leung 1997). These studies modeled tourism as an exogenous activity and did not consider the interactions between tourism and environmental quality. In fact, associated environmental impacts may alter the attractive features of a site. For example, large numbers of hikers can compact soil along trails, leave litter, and disturb wildlife (Healy 1994:599); when a wilderness lake becomes more accessible to users, it can no longer offer the same experience of distance and separation from human habitation (Farrell and Runyan 1991); and if people like seashores, development in the form of construction may distract tourists (Raitz and Dakhil 1988). This suggests that environmental changes in the landscape may signi®cantly impact the level of visitation, thereby affecting the economy of a region. In this study, tourism is considered as an endogenous activity which may be responsive to a set of environmental and economic variables. It is very well documented that activities related to resource sectors (agriculture, forestry, or energy) may cause damage to the environment (Duchesne 1991; OECD 1995; Tsigas et al 1996). Therefore, public agencies often impose restrictions such as taxes on inputs or outputs, standards on management practices, and even preservation or closure of areas from production to control environmental damage. For example, the government of Alberta (Canada) in 1995 chose to restrict logging near the Rocky Mountain foothills because of environmental concerns. Similarly, a proposed coal mine near Jasper National Park, also in Canada, attracted much criticism from environmentalists; they thought that the new mine would adversely affect wildlife habitat and thus the park's experiential quality. Again, in the Paci®c Northwest of the United States, large tracts of forest are closed to logging to protect the habitat for spotted owl. Further, arguments that intensive use of pesticides and fertilizers in agriculture cause ecological damage have led the US government to undertake an acreage reduction program in the agricultural sector (Tobey and Reinert 1991). On the other hand, it is generally believed that tourism is a ``green'' sector that does not harm the environment. The industry is also often characterized as an economic activity that employs relatively simple technologies, draws abundant domestic labor, requires limited capital investments, earns substantial net foreign exchange,

190

MODELING TOURISM IMPACTS

and constrains industrial expansion and thus urbanization (Eadington and Redman 1991). As a result, it is not common to see resistance against the expansion of activities related to this sector. However, a close observation of a tourism-dependent region suggests that activities related to it impact the local environment. For example, rapid growth in tourism has caused an increase in forest ®res and extinction of plant species in Mediterranean forests; led to more forest felling and ®res in the Himalayas; and resulted in the damage of coral reefs in Kenya, Madagascar, and the Seychelles (UNEP 1992). Increases in arrivals and construction of tourism facilities have led to declines in the clarity of water, ozone damage to conifers as a result of auto exhaust, and degradation in the scenic quality of Lake Tahoe in the mountains along the California±Nevada border (Healy 1994:599). If environmental costs associated with tourism development are accounted for, this industry may not necessarily bring the highest economic returns or always bene®t the environment and society (Inskeep 1987). The environmental impacts of tourism which were underestimated in the past are now receiving attention. For example, Page, Bayley, Cook, Green and Richie (1996) have noted that if tourist numbers and related development are allowed to continue, serious and irreversible harm to Banff National Park's ecological integrity and its value as a national park may result. Based on their recommendations, the Heritage Minister of the government of Canada announced a series of restrictions on park use. These include establishing limits and reservation systems on the most popular hiking trails, eliminating all sport ®shing, closing some hotels/campgrounds, and capping the permanent population of the town of Banff at 10,000. As another example, automobile access was restricted to the Maroon Bells Wilderness Area in Colorado because fumes from tourists' cars were polluting the mountain environment (Healy 1994:599). The foregoing discussion suggests that failure to incorporate the environmental impacts and feedback effects of tourism in economic analysis may result in overestimating the signi®cance of the industry in regional economies. Therefore, research focusing on the linkages between environmental impacts and tourism activity is sorely needed. This study provides a theoretical framework to model interactions among tourism, other economic sectors, and the environment. First, the model incorporates linkages between the tourism sector and environmental damage and focuses discussion on the implications for the regional economy. Second, the study considers two types of environmental damage functions: it assumes that economic activity related to resource sectors affects the local environment; and it considers that activities of both the resource and tourism sectors affect the environment. Third, the study emphasizes a region where the resource and tourism sectors compete for a scarce resource such as public land. Fourth, it points out a number of areas where future research is needed to extend and apply the model to a particular region. At present, models designed to assess the economic impacts

ALAVALAPATI AND ADAMOWICZ

191

of alternative policies, be they input-output type models or computable general equilibrium (CGE) ones, tend not to incorporate speci®c linkages with the environment. This paper outlines an approach for constructing this linkage and illustrates the implications of such relations on economic impact assessments. A variety of methodologies have been used to estimate the economic impacts of tourism activities. Some researchers have used input±output (I±O) models to identify the economy-wide impacts of changes in the demand for tourism-related products (Archer and Fletcher 1996; Johnson and Moore 1993). Wagner (1997) used a social account matrix (SAM) to examine the economic effects of tourism. Lindberg and Johnson (1997) examined the effects of tourism using a contingent valuation (CV) approach. Although I±O and SAM models provide a convenient framework to incorporate intersectoral linkages, they suffer from their inability to consider the behavioral responses of producers and consumers with respect to changes in prices. On the other hand the CV approach does not account for intersectoral linkages of the economy. Therefore, some researchers have used a CGE model to examine the economic effects of tourism (Adams and Parmenter 1995; Copeland 1991; Zhou et al 1997). The CGE approach not only accounts for intersectoral linkages but also permits the prices of inputs to vary with respect to changes in output prices and accommodates the indirect effects of a policy change on the overall economy (Dervis, de Melo, and Robinson 1982; Shoven and Whalley 1992). Because of these attractive features, in this study, a CGE approach was chosen to model interactions among tourism, other sectors, and the environment.

MODELING TOURISM IMPACTS The model is kept as simple and general as possible to emphasize the interactions between the economy and environment and their impact on tourism. This model is similar to Jones (1965) but is extended to incorporate environmental damages and tourism. A small regional economy with two productive sectors is considered: the resource sector which includes forestry, agriculture, or energy; and the composite tourism sector. The resource sector is assumed to be a net exporter and price taker in the international market. It is further assumed that all goods and services produced by tourism are consumed in the domestic market and thus treated as a nontraded sector. Tourists from external regions who visit the region also consume these goods and services but within the regional market. This implies that a change in their number causes a shift in the ®nal demand for tourism goods and services. The regional economy consists of two factors of production: a composite input consisting of labor and capital, and land. It is assumed that both inputs are used in each sector and are mobile across these. Furthermore, it is considered that all factor endowments are fully employed between the

192

MODELING TOURISM IMPACTS

two. For the sake of simplicity, factor mobility across regions is ignored. Turning now to the technical description of the model, the ®rst and second parts of equation 1 de®ne the production of resource output and composite tourism output, respectively. It is assumed that production functions exhibit constant returns to scale R ˆ R…L, D†

C ˆ C…L, D†,

…1†

where R is the resource sector; C the composite tourism sector; L, the labor; and D, land. Equations 2 and 3 show that factor inputs are fully employed between the two sectors of the region. Equation 4 indicates that factor proportions depend on relative changes in factor prices. With constant returns to scale, total factor demands are given by the product of aij and levels of output aLR R ‡ aLC C ˆ L,

…2†

aDR R ‡ aDC C ˆ D,

…3†

aij ˆ aij …W, V †,

…4†

where aij is the input i required to produce a unit of output j; i is L, D; j is R, C; W the wage; and V rental rate of land. Equations 5 and 6 show that unit costs are given by the product of aij and factor prices. Since the production functions are homogeneous degree ones, the value of output must be exhausted once factors are paid. In other words, in a competitive equilibrium, unit cost equals market price aLR W ‡ aDR V ˆ PR ,

…5†

aLC W ‡ aDC V ˆ PC ,

…6†

where Pi is the unit price of i; i is R, C. Many authors have speci®ed environmental damage as a function of sectoral output (Merri®eld 1988; Tsigas et al 1996). Tobey and Reinert (1991) speci®ed environmental damage as a function of factors used in the production process. In this study, environmental damage is speci®ed as a function of output and the extent of land use in the production process. This may be appropriate because changes in the environment can be expected from an increase either in the quantity of output (for example, sludge and emissions will increase with pulp output) or in the use of land (soil erosion and disturbance to wildlife may increase with extensive timber harvesting). A conventional notion is that only activities related to resource sectors (such as agriculture, forestry, and energy) cause signi®cant damage to the local environment while tourism causes little. However, as already noted, the tourism sector may also signi®cantly

ALAVALAPATI AND ADAMOWICZ

193

damage the environment. Therefore, two types of damage functions are considered. First, it is considered that land use and the level of output in the resource sector, as well as land use and the level of output in both sectors affect the environment. In simulation experiments, these two parts of equation 7 are mutually exclusive ER ˆ ER …aDR , R†

ERC ˆ ERC …aDR , R, aDC , C †

…7†

where E is the environmental damage. For the sake of simplicity, the model does not consider the dynamics of impacts. The ability of the environment to assimilate damage may change over time. The assimilation rates may also vary with the nature of the impact. In this model tourism activity is described in two scenarios. In both cases, it is assumed that tourists are responsive to changes in the price of the composite tourism good. For example, the rise in air fares and hotel costs may cause a decrease in the number of arrivals. Therefore, the price of goods and services is included in both cases. With regard to the environment, it is a basic ``raw material'' of the industry (Mlinari 1985) and it is expected that the level of tourism activity depends on the quality of the environment. Tourists demand that their recreation environment be clean, and some may change their travel patterns if environmental quality expectations are not met (Inskeep 1987). For example, environmental changes such as pollution of coastal waters, forest ®res, and traf®c congestion associated with tourism development in the Mediterranean have been shown to in¯uence the visits of tourists (Mlinari 1985). A decline in water quality (for example, due to sludge disposal from a local pulp mill or due to soil erosion from logging activities) in river systems may have a negative impact on sport ®shing activity. As such, environmental damage is considered one of the determinants of visitation level. However, in the ®rst part of equation 8, it is assumed that only the environmental damage associated with the resource sector's activity in¯uences the level of tourism. In the second part, the industry is shown to depend on the environmental damage associated with the activities of both resource and tourism sectors T ˆ T…ER , PC † T ˆ T…ERC , PC †

…8†

where T is the tourist activity or tourists or visitors; Ei is the environmental damage associated with the production of i; i is the R, RC. Changes in the environment in response to activities in the resource or tourism sectors may affect tourists variously because they have diverse objectives. For example, hunters may appreciate more clearcut areas, while bird watchers and hikers may like to have old growth forests (Personal communication with Peter Boxall, Canadian Forest Service, Edmonton). On the contrary, Bostedt and Mattsson (1995) have found that a decrease in the size and increase in the number of clearcuts have positive impacts on tourism. This

194

MODELING TOURISM IMPACTS

suggests that the relationship between the changes in the environment and the number of tourists may be non-linear. Those complex relationships are not considered in this modeling exercise and it is simply assumed that environmental damage will have a negative impact on tourism. The demand for goods and services is de®ned in equation 9 which shows that the demand for tourism goods and services arises from households within the region and outside visitors. Consistent with consumption and demand theory, both regional income (Y) and the price of the tourism good are included in the demand function C ˆ C…Y, PC , T †:

…9†

Equation 10 indicates that regional income equals the sum of the value of output in both producing sectors Y ˆ …RPR ‡ CPC †:

…10†

The model described in equations 1±10 is illustrated in Figure 1. The prices of outputs and inputs, regional income, outputs, inputs, the environment, and tourism are interlinked. This suggests that a change in policy may impact all the variables simultaneously. A convenient way of determining the comparative static effects of changes in exogenous variables is by transforming the general model into its proportional change form; this requires relatively little data and the results of comparative statics can be interpreted as elasticities. However, this format introduces linear approximation

Figure 1. The Economy of a Tourism-Oriented Resource Dependent Region

ALAVALAPATI AND ADAMOWICZ

195

errors and thus is valid for only small changes in the exogenous variables. Since simulation experiments conducted later in this study deal with only 1% changes in exogenous variables, approximation errors are expected to be small (see Hertel, Horridge and Pearson 1992 for details on linear approximation errors). Linear transformation involves total differentiation, manipulation, and rearranging the equation of the system (Johansen 1960; Jones 1965; Merri®eld 1988). The details of model transformation steps can be obtained from the authors upon request. In solving the model for changes in endogenous variables, a set of values is used for parameters in the model (Table 1). These values are chosen based on economic theory and researchers' judgement and understanding of regional economies. In empirical work, estimation of factor shares is complex (Marcouiller, Lewis and Schreiner 1996; Ruttan and Stout 1960). In particular, factor share calculation for the tourism sector may pose a challenge for modelers because of the existing controversy about the tourism product. However, efforts to grapple with this problem are increasing. One way to handle this problem is to aggregate activities related to tourism (such as transportation, accommodation, and food services) and Table 1. Parameters Values Used In Simulation Experiments Parameter

Fraction of L used in j Fraction of D used in j Cost share of L in j Cost share of D in j Cost share of Z in j Elasticity of substitution in j Elasticity of E wrt D used in j Elasticity of E wrt output in j Share of sector j in income Elasticity of T wrt i Elasticity of demand for C wrt i

Sector j R

C

0.400

0.600

0.600

0.400

0.550

0.750

0.400

0.230

0.005

0.200

1.000

1.000

1.000

1.000

1.000

1.000

0.400

0.600

Regional Income Y

1.000

Tourism T

1.000

Price PC

Envi. damage E

ÿ1.000

ÿ1.000

ÿ1.000

196

MODELING TOURISM IMPACTS

call it tourism (Smith 1989, 1994). West and Roy (1997) discuss details of extending I±O tables to a SAM and developing price-sensitive models to assess interactions between tourism and the environment. If the industry sector is incorporated into regional or national I±O tables, that will form the base to develop SAM and CGE models. In assigning parameter values, it is assumed that the resource sector is land intensive while tourism is labor intensive. A CobbDouglas functional form was assumed for production and utility functions and thus a unit value is assigned to the elasticities of substitution and the elasticities of demand for the tourism good. Similarly unit values are given for the elasticities of environmental damage with respect to each economic activity and the industry with respect to environmental damage and the price of the tourismrelated good. It should be noted that the use of unitary elasticities may not be realistic since tourism, in general, is considered price elastic. Further, the assumption of constant returns to scale associated with Cobb±Douglas functions may not be appropriate if the two sectors show increasing or decreasing returns to scale. In empirical work more ¯exible functional forms like constant elasticity of substitution or translog functions can be used. However, the use of ¯exible functional forms is beyond the scope of this paper. As with any CGE model, the results may be sensitive to the parameter values chosen for simulation experiments. However, it was thought that the use of the same set of parameters in both scenarios of environmental damage may not alter the general results of the analysis (GEMPACK 5.1 software was used to carry out the simulation experiments). Simulation Results One of the policy instruments governments have in regulating the activities of producing sectors is taxation. The concerns for the environment may prompt public agencies to impose additional taxes on factors of production. For example, the government of British Columbia has recently increased its stumpage prices by >60 and 80%, respectively, on the coast and in the interior region. Alternatively, to regulate pollution emissions from the pulp industry, regional governments may impose additional environmental taxes per unit of output. This would cause an increase in the marginal cost of production and thus a decrease in output. In the case of tourism, environmental taxes can be in the form of an increase in entry fees into parks, camping fees, skiing costs, and hunting fees. This study simulated the impact of a 1% environmental tax (Zi where i=R, C) in each of the sectors of the regional economy. In each case, the impacts are investigated under two scenarios. First, environmental damage is assumed to occur from the activity of the resource sector (Resource±Environment Scenario, or RE). Second, environmental damage depends on the activities of both sectors (Integrated Tourism±Environment Scenario, or ITE). In the model,

ALAVALAPATI AND ADAMOWICZ

197

the environmental tax enters the model through the zero pro®t equations 5 and 6. Table 2 reports the impacts of the tax in each of the producing sectors on various variables. The values are percentage changes with respect to a 1% change in the tax. Columns 1 and 2 of Table 2 show the impact of a 1% increase in environmental tax in the resource sector, respectively, under RE and ITE scenarios. Since the parameters used in the simulations are hypothetical, focus is not made on the absolute changes in these variables. Instead, emphasis is on the differences between the two scenarios. The results show that in both cases the output in the resource sector falls and more so under the RE scenario; the demand for the tourism good increases and this is large in the RE scenario; and the returns to land decline and more so in the RE scenario. These differences can be explained by focusing on the source of the environmental damage. In the RE scenario, the reduction in the activity of the resource sector causes an improvement in the environment. This improvement has a positive effect on tourism in the region. Although the increase in the price of the good discourages tourists, the results show that this effect is completely offset by the expansion in the industry associated with an improvement in the environment. Therefore, an overall increase in arrivals is produced. It is shown that the decrease in the output of the resource sector is completely offset by the expansion of tourism and thus an increase in regional income. This stimulus comes from two sources. First, with a contraction in the resource sector, some factors of production are shifted to tourism. Since the latter is assumed to be labor intensive, the increase generates more demand for labor than the demand for land. Consequently, there will be an increase in the wage rate and a fall in the rental rate of land. Second, an improvement in the environment also stimulates tourism activity. Table 2. Impact of 1% increase in Environmental Tax (values in % changes) Variable

Tax in Resource Sector

Tax in Tourism Sector

RE Scenarioa ITE Scenariob RE Scenario ITE Scenario Regional income (Y) Price of composite (PC) Resource output (R) Composite output (C) Wage (W) Land rent (D) Envi. Damage (E) Tourism (T) a

0.0137 0.0023 ÿ0.1022 0.0885 0.012 ÿ0.0291 ÿ0.0796 0.0772

ÿ0.0021 ÿ0.0047 ÿ0.0058 0.0051 ÿ0.0043 ÿ0.0066 0.0023 0.0025

ÿ0.053 ÿ0.0091 0.3962 ÿ0.3433 ÿ0.0671 0.0923 0.3086 ÿ0.2994

0.0083 0.0183 0.0226 ÿ0.0196 ÿ0.0038 0.0053 ÿ0.0088 ÿ0.0095

Environmental damage is assumed from the resource sector. Environmental damage is assumed from both the resource and composite tourism sectors. b

198

MODELING TOURISM IMPACTS

In the ITE scenario, the effect of the tax in the resource sector is shown to have a negative effect on the environment. The improvement in it generated by the contraction of this sector is offset by the environmental damage associated with the expansion of the tourism sector. This may have a depressing effect on the number of arrivals and thus on the demand for the tourism good. Furthermore, the expansion in tourism is not shown to offset the contraction of the resource sector. Therefore, there is a decline in returns for factors of production and an overall reduction in regional income. In short, results suggest that the tax policy in the resource sector bene®ts the regional economy under the RE scenario and hurts in ITE scenario. Simulation experiments were also conducted by imposing an environmental tax in the tourism sector. Again the impact of the tax policy was investigated under the two scenarios as explained above. In the RE scenario, the environmental tax causes a signi®cant decline in tourism and an expansion in the resource sector. Unlike the case of the tax in the latter, this tax causes a decline in the wage rate and an increase in the rental rate of land. The results also show that under the RE scenario, the expansion of a polluting sector, R, causes signi®cant damage to the environment and thereby a reduction in tourism activity. Furthermore, the expansion in the resource sector is not shown to offset the contraction of the composite sector. As a result we notice a decline in regional income. On the other hand, under scenario B, the impact of a tax in the composite sector actually improves the environment. This suggests that this upgrade due to the contraction of tourism offsets the damage associated with the expansion of the resource sector. Thus, an overall improvement in regional income is noticed. In sum, the results suggest that the tax policy in the tourism sector hurts the regional economy under the RE scenario and bene®ts it in the ITE scenario. CONCLUSION Recently, the environmental impacts of tourism have been receiving more attention from researchers, the media, and the general public. However, few studies have been conducted to model the interactions between tourism and environmental damage. This study provides a framework to model the interactions among tourism, other economic sectors, and the environment in the context of a region where resource extraction is also present. In particular, the model considers both environmental damage and tourism as endogenous variables. Interactions are investigated under two types of assumptions. First, it is assumed that environmental damage in the region is largely due to the activity related to the resource sector. Second, it is considered that activities from both the resource sector and tourism related-activities affect the local environment. Using a simple computable general equilibrium model and conducting simulation experiments, it was shown that the effect of a policy change is not the same under these two assumptions. The results indicate

ALAVALAPATI AND ADAMOWICZ

199

that an increase in an environmental tax in the resource sector bene®ts the regional economy if environmental damage is assumed to occur only from the resource sector activity. The converse holds under the assumption that this occurs from both resource and tourism sector activities. On the other hand, while an increase in an environmental tax in tourism hurts the regional economy under the ®rst assumption, the tax policy bene®ts the regional economy under the second one. It should be noted that these results cannot be generalized since regions generally differ in their economic structure. Furthermore, application of CGE models to a small region requires signi®cant modi®cations to the assumptions made in this study. For example, in small regional economies, it may be unreasonable to assume that the supply of labor is ®xed. Nevertheless, these ®ndings have important implications for tourism management decisions. The model developed in this paper is simple and very general in nature. Depending upon the region, time frame, and issue under investigation, several extensions can be suggested. One, efforts are being made to incorporate the tourism industry into I±O tables and to develop a SAM. Successful development of a SAM allows modelers to consider multisector and multifactor CGE models. For example, the composite tourism sector can be subdivided into transportation, manufacturing, and other service sectors. Resource sectors can also be disaggregated into agriculture, forestry, mining, and oil and gas. Similarly, combined labor inputs can be disaggregated into capital and labor, with the latter further divided into skilled and unskilled because of the differences in their mobility across sectors. Two, depending on the short-run or long-run scenario, factor markets can be modeled differently. For example, in the short-run, capital can be sector speci®c but in the long-run it is mobile across sectors. In the very long-run, capital can be mobile across the regional boundaries. Therefore, in the very long-run, issues such as international migration and foreign ¯ow of capital may have to be considered. Three, if the resource dependent region has market power in the international market for its exports, it is appropriate to consider the price of the resource sector endogenous. For example, in modeling the economy of British Columbia, several researchers have considered the price of the wood products sector as endogenous (Percy 1986). Four, a variety of ¯exible functional forms can be used to specify production technology and household preferences. These include Constant Elasticity of Substitution (CES), Constant Ratio of Elasticity of Substitution Homothetic (CRESH), Generalized Leontief, and Translog functions. Furthermore, ®ve, savings, investment, trade balance, and various types of social accounts can be included in the model along with the government sector. In some cases it may be appropriate to model these issues in a dynamic setting. GEMPACK 5.1 or MPSGE/GAMS software are now available to handle dynamic analysis. Finally, the relationship between tourism and environmental damage may not be linear as shown in the model. Depending upon the availability of

200

MODELING TOURISM IMPACTS

data, complex non-linear relationships can be examined. Moreover, modeling the relationship between these two variables requires input from various disciplines. For example, information on marginal damage to the environment due to additional logging, mining, or oil drilling have to be obtained from experts in those areas. Modeling marginal emissions from various production activities, pollution assimilation rates, and threshold limits of the environment also involves interdisciplinary research. In addition, issues such as an increase in the probability of forest ®re associated with tourists and extent of damage under alternate production activities can also be incorporated in extended models. There is signi®cant interest in understanding the linkages between the environment and the economy. In regions where extraction of resources (such as forestry and mining) has historically supported the economy, residents are concerned about the degree to which such activities affect new industries like tourism. Indeed, the relationship between the environmental impact of resource extraction and tourism and the local economy is frequently debated in policy analysis and discussion. Models such as the one presented here provide a framework for analysis of such issues, and illustrate the importance of modeling these relationships in order to accurately assess the economic and environmental impact of policy alternatives. A similar call for such models is arising in the literature on lifecycle analysis (Hendrickson, Howath, Joshi and Lave 1998) in which linkages between physical ¯ow models and ones of economic activity are necessary for accurate analysis of process options. Models structured to identify and illustrate the importance of environment±economy linkages will likely be of increasing importance in the future as policymakers are called to assess alternatives with economic and environmental implications.& Acknowledgments ÐFinancial support from the Sustainable Forest Management Network of Centers of Excellence and the Florida Agricultural Experiment Station is gratefully acknowledged. Florida Agricultural Experiment Station Journal Series R-06568.

REFERENCES Adams, P. D., and B. R. Parmenter 1995 An Applied General Equilibrium Analysis of the Economic Effects of Tourism in a Quite Small, Quite Open Economy. Applied Economics 27:985± 994. Archer, B., and J. Fletcher 1996 The Economic Impact of Tourism in the Seychelles. Annals of Tourism Research 23:32±47. Bergman, L. 1995 Environment-Economy Interactions in a Computable General Equilibrium Model: A Case Study of Sweden. In Current Issues in Environmental Economics, P. Johansson, B. KristroÈm and K. MaÈler, eds., pp. 153±170. Manchester: Manchester University Press.

ALAVALAPATI AND ADAMOWICZ

201

Bostedt, G., and L. Mattsson 1995 The Value of Forests for Tourism in Sweden. Annals of Tourism Research 22:671±680. Copeland, B. R. 1991 Tourism, Welfare and De-industrialization in a Small Open Economy. Economica 58:515±529. Dervis, K., J. de Melo, and S. Robinson 1982 General Equilibrium Models for Development Policy. Cambridge: Cambridge University Press. Duchesne, I. 1991 The Use of Tax Policies to Achieve Environmental Objectives: A General Equilibrium Case Study of British Columbia. Unpublished M.A. thesis, University of Alberta, Edmonton. Eadington, W. R., and M. Redman 1991 Economics and Tourism. Annals of Tourism Research 18:41±56. Farrell, B. H., and D. Runyan 1991 Ecology and Tourism. Annals of Tourism Research 18:26±40. Healy, R. G. 1994 The ``Common Pool'' Problem in Tourism Landscapes. Annals of Tourism Research 21:596±611. Hertel, T. W., J. M. Horridge, and K. R. Pearson 1992 Mending the Family Tree: A Reconciliation of the Linearization and Levels Schools of AGE Modelling. Economic Modelling 9:385±407. Hendrickson, C., A. Horvath, S. Joshi, and L. Lave 1998 Economic Input-Output models for Environmental Life-cycle Assessment. Environmental Science and Technology 32(7):184±191. Inskeep, E. 1987 Environmental Planning for Tourism. Annals of Tourism Research 14:118± 135. Johansen, L. 1960 A Multi-Sectoral Study of Economic Growth. Amsterdam: North-Holland Publishing Company. Johnson, R. L., and E. Moore 1993 Tourism Impact Estimation. Annals of Tourism Research 20:279±283. Jones, R. W. 1965 The Structure of Simple General Equilibrium Models. The Journal of Political Economy 73(6):557±572. Lindberg, K., and R. L. Johnson 1997 The Economic Value of Tourism's Social Impacts. Annals of Tourism Research 24:90±116. Marcouiller, D. W., D. K. Lewis, and D. F. Schreiner 1996 Timber Production Factor Shares by Forest Tenancy Group. Land Economics 72:358±369. Merri®eld, J. N. 1988 The Impact of Selected Abatement Strategies on Transnational Pollution, the Terms of Trade, and Factor Rewards: A General Equilibrium Approach. Journal of Environmental Economics and Management 15:259±284. Mlinari, I. B. 1985 Tourism and the Environment: A Case for Mediterranean Cooperation. International Journal of Environmental Studies 25:239±245. OECD 1995 Environmental Taxes in OECD Countries. Paris: Organization for Economic Co-operation and Development. Page, R., S. Bayley, D. Cook, J. E. Green, and J. R. Brent Ritchie 1996 Banff±Bow Valley: At the Crossroads. Summary Report of the Banff±Bow Valley Task Force. Prepared for the Honorable Minister of Canadian Heritage, Ottawa. Percy, M. B. 1986 Forest Management and Economic Growth in British Columbia. A Study Prepared for the Economic Council of Canada, Ottawa.

202

MODELING TOURISM IMPACTS

Peter, M. 1990 Tourism: Environment and Development Perspectives. The Worldwide Fund for Nature. Raitz, K., and M. Dakhil 1988 Recreational Choices and Environmental Preference. Annals of Tourism Research 15:357±370. Ruttan, V. W., and T. T. Stout 1960 Regional Differences in Factor Shares in American Agriculture: 1925± 1957. Journal of Farm Economics 42:52±58. Smith, S. L. J. 1989 Tourism Analysis: A Handbook. Harlow: Longman Scienti®c. 1994 The Tourism Product. Annals of Tourism Research 21:582±595. Shoven, J. B., and J. Whalley 1992 Applying General-Equilibrium. New York: Cambridge University Press. Tobey, J. A., and K. A. Reinert 1991 The Effects of Domestic Agricultural Policy Reform on Environmental Quality. The Journal of Agricultural Economics Research 43:20±28. Tsigas, M. E., D. Gray, and B. Krissoff 1996 Trade and Environmental Impacts of Economic Integration in the Western Hemisphere. Paper Presented at the Annual Meeting of the American Agricultural Economics Association, July, San Antonio, TX. UNEP 1992 Tourism and the Environment: Facts and Figures. Industry and Environment 15(34):3±5. Wagner, J. E. 1997 Estimating the Economic Impacts of Tourism. Annals of Tourism Research 24:592±608. West, G. R., and J. R. Roy 1997 Towards an Interregional Framework for Joint Assessment of Tourism and the Environment. A Working Paper Presented to the International Tourism Research Conference, Manly NSW, July. Zhou, D., J. F. Yanagida, U. Chakravorty, and Pingsun Leung 1997 Estimating Economic Impacts from Tourism. Annals of Tourism Research 24:76±89. Submitted 26 March 1997. Resubmitted 23 October 1997. Resubmitted 17 July 1998. Resubmitted 29 December 1998. Accepted 29 December 1998. Final version 12 April 1999. Refereed anonymously. Coordinating Editor: Robert W. McLellan