The Economic Benefits of Elk Viewing at the Jewell Meadows Wildlife Area in Oregon

Abstract

In this study a travel cost model is used to estimate the value of elk viewing at the Jewell Meadows Wildlife Area in Oregon. Jewell Meadows was originally established to provide winter browse and supplemental feeding for elk to reduce damage to nearby agricultural and forest land. However, because visitors are virtually guaranteed the opportunity to see large numbers of elk at close range, Jewell Meadows has also become a popular wildlife viewing site. We estimated total access value for the 2007 winter feeding season to be $6.5 million, which exceeds Jewell Meadows' annual operating budget of approximately $200,000. Results suggest that visitors highly value the assured wildlife viewing that Jewell Meadows offers.

Keywords:

• elk
• wildlife viewing
• Jewell Meadows
• travel cost method

Notes

1. An example of an economic impact of wildlife viewing study would be Loomis and Caughlan (2004). They estimate the economic impacts of different elk and bison management strategies in Grand Teton National Park and the adjacent National Elk Refuge and found that a no supplemental feeding management option would reduce visitation by 20% and would reduce local employment by 11%.

2. The feeding of wildlife to reduce agricultural damage is not unique to Jewell Meadows (there are two other programs in Oregon alone), so understanding the benefits associated with wildlife viewing would be useful to other wildlife management agencies, which operate, or may begin to operate, supplemental feeding programs.

3. Copies of the survey are available from the corresponding author.

4. Poisson models assume that the mean and the variance of the dependent variable are the same. In our data, the mean of number of trips is 2.6, whereas the variance is 11.4. Estimating a Poisson model when the data are over dispersed will result in the standard errors being underestimated (Haab & McConnell, 2002).

5. One final point on model estimation: we had some trouble getting the zero-truncated negative binomial model with endogenous stratification to converge. In addition, estimates of the intercept and the dispersion parameter, alpha, were sensitive to the starting values specified in the maximum likelihood routine, and, in some cases, alpha took on an extremely high value. Hilbe (2007, p. 166) experienced similar problems with this model but does not comment. The convergence problems we experienced persisted with a mean-only model, which implies that the problem may stem from the fundamental formulation of the model adjusting for both zero-truncation and endogenous stratification and not ill-structured data as Hilbe (2007) suggests. Although the estimation problems we experienced were somewhat troubling, the parameters of interest were very stable and varied little from the zero-truncated negative binomial model. Therefore, we feel confident using the coefficient estimates from the zero truncated and endogenously stratified negative binomial model to calculate access value.