WEAI/AERE 2012 - Individual Paper Abstract
Title: The Role of Forests as Natural Amenities: A Seemingly Unrelated Regression Model with Two Spatial Processes
Author(s): German IZON, Department of Economics, Eastern Washington University, 013 Hargreaves Hall, Cheney, WA 99004, USA, 509-359-2838, 509-359-6983, gmuchnikizon@ewu.edu; Michael Hand, USDA Economic Research Service; Jennifer Thacher, University of New Mexico; Daniel McCollum, US Forest Service; Robert Berrens, University of New Mexico [Photo credit: adapted from Wikimedia Commons, Peaks-Inner Basin.JPG]
Abstract:
The objective of this paper is to examine spatial variations in wage and housing prices in the presence of forest amenities in a study of Arizona. Undeveloped, open-space lands, such as Inventoried Roadless Areas (IRAs) and Congressionally-designated Wilderness Areas (WAs) provide a number of non-market benefits to society, which may not be fully accounted for in land management decisions. Since natural amenities generate multiple beneficial end uses, there have been competing allocation schemes for these resources. This has clearly been the case in the policy debate about WAs and IRAs that centers on the question of how these public lands should be managed and allocated. In light of petitions filed by various U.S. states to maintain the status of IRAs as roadless lands, compensating differentials are estimated based on a seemingly unrelated regression model with two spatial processes (e.g., spatial lag and spatial error autocorrelation) using a matched sample of wage-earner housing units at the household level. The underlying model used for this empirical analysis follows Roback's (1982) theoretical framework and the econometric approach outlined in Kelejian and Prucha (2004). Hedonic regressions of housing prices and wages indicate that the average total implicit price for forest areas is $980 per mile compared to $696 for WAs, annually. The presence of compensating differentials suggests that care must be taken when applying the travel cost method to value regionally-delineated characteristics.
The contributions of this study to the current literature are threefold. Firstly, the possibility of spatially-dependent relationships is addressed by estimating spatial regression models. The underlying spatial relationships among observations were determined by applying Lagrange Multiplier (LM) tests for the co-existence of spatial lag and spatial error processes. Secondly, the effect on housing and labor markets of site-specific characteristics, such as forest areas, wilderness, superfund sites, outstanding waters, is estimated based on a Geographic Information System (GIS) road network. These variables represent the road mile distance from house i to its closest natural amenity within each category. While a growing number of studies have looked at forest amenities to account for persistent differences in wages and housing prices, measurement of such variables have been limited to percentage of forest areas within a predefined administrative boundary or a straight line distance. Thirdly, to the degree that preferences for forests are related to recreation travel, observed travel cost would be endogenous. This would suggest that the travel cost would not reflect the full price of recreation site access, and may lead to underestimates of the value of recreation access (Hand et al., 2008; Schmidt and Courant, 2006).