Is Resident Valuation of Ranch Open Space Robust in a Growing Rural Community? Evidence From the Rocky Mountains

Abstract

The conversion of ranch open space to development uses is common in rural areas experiencing rapid growth, often due to in-migration. There is concern that a growing and changing population could alter community preferences regarding development, permanently changing the physical and social landscape. This study investigates the nonmarket value of ranch open space to residents of an amenity-rich county in the Rocky Mountains (USA) using survey data from 1994 and 2004. We find that on that, on average, there was no difference in what residents were willing to pay for ranch open space conservation between the two years, and that income has a positive effect on willingness to pay. Because the population of the county is growing and average willingness to pay has stayed constant, aggregate willingness to pay for public ranch open space has increased, leading to greater total social benefits of public conservation initiatives.

Keywords:

• conservation easements
• contingent valuation
• nonmarket value
• ranch open space
• rural landscape

Acknowledgments

This article was derived from the first author's MS thesis completed at Colorado State University. We acknowledge, without implication, Randall Rosenberger for his comments, advice, and data. We also thank Robert Kling, Stephen Koontz, C. J. Mucklow, and Karen Thome for their valuable input and Jeffie Duncan for data entry. Three anonymous referees contributed greatly toward improving the quality of this article. All remaining errors are our own. This project was funded by the Colorado State University Department of Agriculture and Resource Economics, Routt County Cooperative Extension, Colorado State University Cooperative Extension, the City of Steamboat Springs, and the Colorado Conservation Trust.

Notes

Note. Differences in means between years are: *significant at 5%, **significant at 1%. Differences in means for continuous variables are tested using both a t test and a Mann–Whitney–Wilcoxon test due to nonnormality. Differences in categorical variables are tested using a Pearson chi-squared test. The number of observations listed in the table is the number of total respondents in each survey year; due to missing responses, the number of observations for individual variables can be less.

Note. Robust standard errors in parentheses; *significant at 5%, **significant at 1%. A chow test fails to reject that there is no structural difference in the model between years (p = .42).

Note. Robust standard errors in parentheses; *significant at 5%, **significant at 1%. A chow test fails to reject that there is no structural difference in the model between years (p = .76).

Note. Robust standard errors in parentheses; *significant at 5, **significant at 1%. A chow test fails to reject that there is no structural difference in the model between years (p = .55).

Active or working ranches are privately held livestock grazing areas. Ranch open space refers to land used for these purposes free of development.

We thank an anonymous referee for articulating this point.

More land may be held under easements at present.

The 1994 response rate is likely higher because a second, abbreviated survey was sent to initial nonresponders.

The 2004 study had greater financial resources, allowing for a greater sample size.

Likert scales are ordinal scales of degree of importance or degree of agreement/disagreement with a particular statement (Babbie 2010).

For comparisons of categorical variables, including concern for ranch open space, we use Pearson's chi-squared test. For continuous variables we use a two-tailed t-test and a Mann–Whitney–Wilcoxon test as a robustness check due to nonnormality of some variables (Mann and Whitney 1947). In no instance did the choice of test affect the confidence level of a significant difference between years (using .01 and .05 as cutoffs).

The area of ranch open space reported in 1997 is less than one-third the amount reported in the other three years, so we presume that the 1997 value is erroneous.

All data reported in dollars are adjusted to their 2004 value.

The interested reader can refer to McConnell and Walls (2005) and Racevskis et al. (2000) for more details on some of these studies.

An ordered logit specification is used when the categorical regressand is ordinal, with changes between categories that are not clearly equivalent (see Greene 2003).

The odds ratio is ratio of the probability of being in one group compared to the probability of being in another. For the Likert scale variables used in this article, this is the equivalent to the odds of choosing 5 instead of 4, 4 instead of 3, etc.

As a robustness check to the concern that the comparison of WTP across years is biased by oversampling respondents with an agricultural background, we compared WTP only for those not reporting an agricultural background, and found a similar insignificant difference across years ($26 as opposed to $35).