ABSTRACT
CONTENTS 14.1 Introduction ............................................................................................. 313 14.2 Methods.................................................................................................... 315
14.2.1 Data ............................................................................................. 315 14.2.2 Random Utility Model ............................................................. 317
14.2.2.1 Specification of the Representative Utility Function ........................................................ 318
14.2.2.2 Model for Total Demand........................................ 319 14.2.2.3 Assessment of Visits to New Sites ........................ 320
14.2.3 Accessibility-Based Model....................................................... 320 14.3 Results....................................................................................................... 321
14.3.1 Estimated Dependence of Distance and Site Amenities .... 321 14.3.2 Specification of Total Demand for the Discrete-
Choice Analysis......................................................................... 323 14.3.3 Estimated Potential of Reafforestation Areas ...................... 323
14.4 Discussion ................................................................................................ 324 14.5 Conclusion ............................................................................................... 326 Acknowledgment............................................................................................... 327 References ........................................................................................................... 327
The environment’s ability to provide goods and services beyond those with traditional and well-studied markets, such as agriculture and forestry, has received increased attention over the last decades from environmental economists (Haab and McConnell, 2002). Policy makers, too, are now
widely discussing these goods and services as they attempt to balance economic developments with environmental concerns (Forest Commission, 2001). One such environmental good is the benefit people gain from recreational use of forest areas. The value gained by recreationists using forests needs to be taken into account, along with other benefits from forests such as watershed protection and wildlife diversity, when policy makers assess land-use decisions (Bateman et al., 1997). As a step toward understanding the recreational benefits offered by forest areas, policy makers are likely to require information about the patterns of recreational use for existing forest areas as well as information about where potential reafforestation projects might be best located. This inherently spatial information is well suited to study using GIS applications. The recreational use made of forest areas is clearly dependent on the spatial distribution of those areas and the potential recreationists. Spatial analysis of policy initiatives, such as reafforestation schemes, using GIS and aimed at improving the recreational opportunities for local populations, would potentially be informative in the development of forest planning and policy proposals. Researchers have approached modeling of recreational use from a variety
of perspectives. Economists have been interested in estimating recreational demand functions to assist in the economic valuation of forest resources (McConnell, 1985; Hanley et al., 2003). The studies that fall in this category have been interested in the relationship between recreational demand and socioeconomic characteristics but have put less emphasis on the spatial pattern of recreational demand. Land-use planners on the other hand have mainly been interested in understanding the recreational activity and obtaining estimates of the spatial patterns of the use or potential use of recreational areas (De Vries and Ghossen, 2001). As expected, the different perspectives have resulted in different modeling approaches. The use of GIS, for example, has a long history in land-use planning but has been little used in economics, except the work on travel cost analysis (Bateman et al., 1999, 2002; Brainard et al., 1999). The statistically based approaches usually used in economics can be divi-
ded into two main categories of models: continuous regression models and discrete-choice models. Continuous regression models have been adopted when focus is on estimating recreational demand for a specific site. Discretechoice models have been used where the number of alternative sites is large and substitution between sites is important (Haab and McConnell, 2002). The continuous regression models have been combined successfully with
GIS (e.g., Brainard et al., 1999). However, little has been done to combine discrete-choice models with GIS. Furthermore, little has been done to assess the relative merits of purely GIS-based approaches and the economics-based approaches for policy analysis. In this paper, we address these issues by using the same data set to model
forest recreation in Denmark using a discrete-choice model built on GISprocessed spatial data and an accessibility-based model using methods entirely available within most GIS software. In both approaches visitor
numbers are modeled as a function of distance between visitors and forest areas and forest site amenities. We assess the underlying assumptions in the two approaches and compare the two methodologies and the data requirements for the two analyses. In particular, we assess the applicability of the two approaches for identification of suitable reafforestation areas. We illustrate this using information on an existing reafforestation project, initiated by the Danish Forest and Nature Agency, Department of Environment. The project is located in a forestry poor region close to one of the main population centers in Denmark.
