Mineral Rights & Shale Development: A Hedonic Valuation of Drilling in Western Colorado Andrew Boslett PhD Candidate University of Rhode Island Environmental & Natural Resource Economics [email protected]Todd Guilfoos & Corey Lang Assistant Professors University of Rhode Island Environmental & Natural Resource Economics
19
Embed
Mineral Rights & Shale Development: A Hedonic Valuation of Drilling in Western Colorado Andrew Boslett PhD Candidate University of Rhode Island Environmental.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Mineral Rights & Shale Development: A Hedonic Valuation of Drilling in Western Colorado
Our interpretation of the results is reliant on a series of assumptions
1. Close proximity to a horizontal well is exogenous Surface estate is subordinate, no pre-drilling differences in sale price
2. Property buyers and sellers are aware of shale development Ramp-up of planning activity, large size of installations
3. Property buyers and sellers are aware of mineral severance Long history of O&G development, BLM-focused effort at providing more
information
4. Estimates are not impacted by spillover effects Small enough area to not worry about regional effects, location F.E.
5. The financial benefits of local development are negligible for split estate owners Definition of split estate, limited benefits from surface use agreements
Table 1: Summary statisticsFull Sample (N =
47,033)Split Estate (N =
783)
Variable Mean Std. Dev. Mean Std. Dev.
Sale Price ($000s) 250.6 109.9 183.3 81.4Acres 1.4 8.0 6.4 26.7Age at time of sale (years) 18.3 24.1 17.8 16.4Beds 3.2 0.7 3.0 0.7Baths 2.1 0.6 2.0 0.6Finished squared feet (000s) 1.8 0.7 1.6 0.7Distance to municipality 0.5 1.5 1.8 3.4Distance to NPS area 20.1 23.4 28.4 16.9% Agricultural 7.4 23.3 3.0 14.4# of vertical wells < 1 mile 0.1 0.6 0.9 1.1
# of horizontal wells < 1 mile 0.3 3.1 2.1 7.9# of horizontal wells < 2 miles 1.5 12.0 12.1 35.6
% of properties with horizontal well < 1 mile 2.2 14.7 12.5 33.1
% of properties with horizontal well < 2 miles 7.4 26.2 34.6 47.6
% of properties with horizontal well < 3 miles 10.6 30.8 38.2 48.6
Table 2: The effect of unconventional development on the residential property market (N = 47,033), Binary Treatment
R-Squared 0.406 0.493 0.493 0.497 0.562 0.590Notes: Observations represent single family residential properties sold from 2000 to early 2015 in Garfield, Mesa, and Rio Blanco counties. We truncate the data set to exclude the 5 and 95 percentiles of sale price. The dependent variable is the natural log of sale price (CPI-adjusted to 2014 values). Property variables include # of bedrooms and bathrooms, parcel acreage, property finished living area, property age, and squared terms. Location variables include distance to the closest National Park Service Area, distance to the closest municipality, and the percentage of the property in an agricultural use, along with associated squared terms. Census tracts are based on U.S. Census 2010 boundaries. Standard errors are shown in parentheses and are estimated using tract-level cluster-robust inference: *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively.
Table 3: The effect of unconventional development on the split estate properties (N = 783), Binary Treatment
R-Squared 0.412 0.482 0.491 0.523Notes: Observations represent single family residential properties sold from 2000 to early 2015 in Garfield, Mesa, and Rio Blanco counties. We truncate the data set to exclude the 5 and 95 percentiles of sale price. The dependent variable is the natural log of sale price (CPI-adjusted to 2014 values). Property variables include # of bedrooms and bathrooms, parcel acreage, property finished living area, and property age, along with squared terms. Location variables include distance to the closest municipality and National Park Service Area, and the percentage of the property in an agricultural use, along with squared terms. Census tracts are based on U.S. Census 2010 boundaries. Standard errors are shown in parentheses and are estimated using tract-level cluster-robust inference: *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively.
Table 4: The effect of unconventional development on the split estate properties (N = 783), Continuous Treatment
(1) (2) (3) (4)
Variables Property & Location Var. + Year FE Year FE +
R-Squared 0.467 0.550 0.565 0.597Notes: Observations represent single family residential properties sold from 2000 to early 2015 in Garfield, Mesa, and Rio Blanco counties. We truncate the data set to exclude the 5 and 95 percentiles of sale price. The dependent variable is the natural log of sale price (CPI-adjusted to 2014 values). Property variables include # of bedrooms and bathrooms, parcel acreage, property finished living area, and property age, along with squared terms. Location variables include distance to the closest municipality and National Park Service Area, and the percentage of the property in an agricultural use, along with squared terms. Census tracts are based on U.S. Census 2010 boundaries. Standard errors are shown in parentheses and are estimated using tract-level cluster-robust inference: *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively.
# Obs. 1,581 971 919 882 363 783R-Squared 0.468 0.493 0.493 0.484 0.613 0.505Notes: Standard errors are shown in parentheses and are estimated using tract-level cluster-robust inference: *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively.
Table 6: Matching estimates of the effect of unconventional development on split estate properties
Mean Normalized Bias 6.4 6.7 4.1Pseudo R² 0.042 0.033 0.016Likelihood Ratio Test 0.903 0.969 0.999Notes: Property variables include # of bedrooms and bathrooms, parcel acreage, property finished living area, property age, distance to closest municipality, and the percentage of the property in an agricultural use. We also include a count variable of the number of vertical oil and gas wells drilled within a mile of the property from 1980 to 2000. The dependent variable is the natural log of sale price (CPI-adjusted to 2014 values). All statistics are post-matching. Bootstrapped standard errors are shown in parentheses: *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively.
Table 7: Matching robustness checks(1) (2) (3) (4) (5) (6)Alternative P.S. Model Specifications Alternative Datasets
Mean Normalized Bias 6.9 6.5 5.3 16.2 5.4 13.2Pseudo R² 0.065 0.039 0.023 0.291 0.021 0.115Likelihood Ratio Test 0.951 0.901 0.735 < 0.001 0.756 0.793Notes: The dependent variable is the natural log of sale price (CPI-adjusted to 2014 values). All statistics are post-matching. Bootstrapped standard errors are shown in parentheses: *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively.
Discussion & Conclusions
The previous literature that has heretofore focused on net valuations of shale development
We avoid a number of issues by only analyzing split estate properties in western Colorado
12 – 36% decrease, robust across various specifications
~ $60,000 or $3,400 per well
Notes Remote setting of western Colorado? Information issues? No financial benefits?
Acknowledgements
Garfield, Mesa, and Rio Blanco County
Assessment GIS
Bureau of Land Management Colorado office Steven Hall, Martin Hensley,
Deanna Masterson & Courtney Whiteman
Local experts Lois Dunn, real estate agent Cameron Grant, lawyer Local BLM officials