The University of Tennessee
Natural Resource Policy Center
About the Center
Scientific Analysis
Policy Process
Capacity Building
Center Personnel
Paper and Presentations
A Feasibility Study for Utilizing NVUM Survey Results and Us Census Bureau Data to Estimate and Forecast Recreation Demand on National Forests
NRPC Lead: Seong-Hoon Cho

Research Assistant: Seung Gyu Kim

Funding Source: USDA Forest Service

Project Description
The national visitor use monitoring (NVUM) program is the “corporate” recreation data collection system of the USDA Forest Service. NVUM collects both use and user information, including basic user characteristics, attitudes and preferences, and economic expenditures. Currently, NVUM focuses on reporting use behavior for the periods wherein sampling occurred, resulting in a somewhat backward-looking approach. Forecasts of visitation must rely on the assumption that the estimates of past use are reasonable approximations of future use. Such forecasts by their nature include an incomplete accounting for the factors influencing change in use, and add little to managers’ understanding of the likely future for visitation patterns. However, the NVUM data hold as yet untapped opportunities for forecasting recreation demand if those data could be used to create valid and reliable demand estimation functions based upon (among other factors) basic demographic characteristics of recreation users and their place of residence (or origin).

The analytic approach will be to develop county-level aggregate visitation models, building on work done to develop demand projections for Region 10 (Bowker 2001, Bowker and Harvard 2006). Preliminary work for Region 10 completed at the Southern Research Station suggests that NVUM data could be used to develop statistically rigorous demand functions (by activity group) suitable for forecasting purposes when the data is paired with available secondary demographic characteristics of a geographically defined population. One of the areas that can be improved from the preliminary work is incorporating spatial econometrics in the existing county-level aggregate visitation models.

Two types of spatial econometric models, spatial error and lagged model and locally weighted regression model, will be considered. The both models will represent the complexity of spatial heterogeneity and/or spatial dependence of county-level aggregate visitation models. The spatial heterogeneity refers to structural relations that vary over space. Spatial dependence is a systematic spatial variation that results in observable clusters or a systematic spatial pattern (Florax and Nijkamp, 2003). The development of consistent and efficient estimators for capturing the spatial dependence and spatial heterogeneity has been an important part of the spatial econometric/statistics literature for the last few decades. To our knowledge, visitation models of this kind have not been modeled in a spatially-explicit context incorporating both spatial dependence and spatial heterogeneity.

Research Objective:
The main objective of this study is to improve county-level aggregate visitation models by incorporating spatial econometric models. By doing so, it will help national forest managers’ understanding of the likely future for spatial structures of visitation patterns.

Research Approach:
This study integrates the spatial effects of spatial heterogeneity and spatial dependence of visitations of national forests within the framework of spatial error and lagged model and locally weighted regression model. Spatial dependence and spatial error models developed by Anselin (1988) will be used to detect and accommodate the spatial effect. A locally weighted regression approach, as first proposed by Cleveland and Devlin (1988), will be adopted to deal with the spatial heterogeneity and allow estimates of the marginal effects vary across to the space.

Project Links
Seong-Hoon Cho
University of Tennessee
Agricultural Economics
Seung Gyu Kim
USDA Forest Service
Copyright © 2007
~To enhance policy making relative to the sustainable management
of natural resources in Tennessee and the Southeastern Region~
The University of Tennessee