Impact of Wind Power Projects on Residential Property Values in the - - PowerPoint PPT Presentation

Impact of Wind Power Projects on Residential Property Values in the United States

An Overview of Research Findings

Mark A. Thayer

San Diego State University

PLEASANT RIDGE EXHIBIT 39

Property Values

• Dr. Mark A. Thayer
• Ph.D. in Economics from University of New Mexico, 1979
• Field of expertise is environmental, natural resource, and energy economics
• Professor and Chair in the Department of Economics at San Diego State University
• Nationally known expert in the valuation of environmental commodities
• Thirty years of experience in both university and government service
• Extensive experience integrating environmental and energy related matters into decision

making at the state and federal level

• Published numerous research articles in professional journals such as American Economic

Review, Journal of Political Economy, Journal of Environmental Economics and Management, Land Economics, Natural Resources Journal, Journal of Urban Economics, Economic Inquiry, Journal of Sports Economics, and Journal of Human Resources

• Principal investigator on projects funded by entities such as the California Air Resources

Board, California Energy Commission, U.S. Environmental Protection Agency, U.S. Geological Survey, the South Coast Air Quality Management District, and the National Science Foundation

Presentation

• Primarily based on two revealed preference studies:
• “The Impact of Wind Power Projects on Residential Property Values

in the United States: A Multi-Site Hedonic Analysis” by Ben Hoen (LBNL), Ryan Wiser (LBNL), Peter Cappers (LBNL), Mark Thayer (SDSU), and Gautam Sethi (Bard), 2009

• “A Spatial Hedonic Analysis of the Effects of Wind Energy Facilities
• n Surrounding Property Values in the United States by Ben Hoen

(LBNL), Jason Brown (FRBKC), Thomas Jackson (Texas A&M), Ryan Wiser (LBNL), Mark Thayer (SDSU), and Peter Cappers, 2013

• Studies conducted by Environmental Energy Technologies Division of the Ernest

Orlando Lawrence Berkeley National Laboratory (LBNL), funded by the Office of Energy Efficiency and Renewable Energy (Wind and Hydropower Technologies Program), U.S. Department of Energy

LBNL Wind Studies

LBNL-6362E

A Spatial Hedonic Analysis of the Effects of Wind Energy Facilities on Surrounding Property Values in the United States

Ben Hoen, Jason P. Brown, Thomas Jackson, Ryan Wiser, Mark Thayer and Peter Cappers Environmental Energy Technologies Division

August 2013

• f Energy under Contract No. DE-AC02-05CH1123.

ERNEST ORLANDO LAWRENCE BERKELEY NATIONAL LABORATORY

Conclusion from LBNL Studies

Based on analysis of more than 58,000 single

family home sales before, during, and after wind farm development in the U.S., we concluded that there was NO IMPACT from wind farms on the sale prices of these residential properties

U.S. Literature Developments 2010 - 2014

• Jennifer Hinman (2010), Illinois

– 3,851 home sales

• Jason Carter Study (2011), Illinois

– 1,298 home sales

• Heintzelman and Tuttle (2012), New York

– 11,331 home sales

• Magnusson and Gittell, (2012), New Hampshire

– 2,593 home sales

• Atkinson-Palombo and Hoen (2014), Massachusetts

– 122,198 home sales (6,081 within one mile of a turbine)

• Lang, Opaluch, and Sfinarolakis, (2014), Rhode Island

– 48,554 home sales (3,254 within one mile of a turbine)

Overall Conclusion

• All large-scale, empirical studies of U.S. wind

facilities conclude that, post-construction/

• peration, there is no identifiable effect of wind

power projects on nearby residential property values

• 248,560 home sales evaluated in eight studies

Proximity to and Views of Environmental (Dis)Amenities Can Impact Property Values

• This linkage has been extensively studied

↑$ ↓$

Average Home Superfund Site Landfill/Transfer Station Green Space Ocean Front

↑$

↓ $

Research Relied on Hedonic Pricing Model in Addition to Other Models

• Hedonic Pricing Model

– Used by economists and real estate practitioners for over 40 years – “Method for estimating the implicit price of the characteristics that differentiate closely related products in a product class”

• Other models Used in Analysis

– Repeat Sales and Sales Volume Models

Hedonic Pricing Model

• v. Appraisal Model
• Appraisal Model

– Designed to determine the estimated selling price of an individual home – Uses a small # of home sales (comps) – Controls (holds constant) a small # of variables (square footage, home style, pool) – Uses data from a very restricted area (e.g., close to the subject home)

• Hedonic Pricing Model

– Designed to place an economic value on specific characteristics of a home (e.g., value of an additional bathroom, a pool, or view of wind turbines) – Uses a large # of home sales (many thousands) – Controls (holds constant) a large number of possibly confounding variables (everything under the sun) – Uses data from a large area to

• btain enough variation in all

characteristics

Hedonic Pricing Model

• v. Appraisal Model, continued
• Appraisal Model

– Uses data from a very restricted time period (e.g., previous six months) – Cannot be used effectively to evaluate the contributory value

• f a specific home

characteristic unless sufficient controls are in place – “Paired Sales” analysis is an attempt to evaluate a specific home characteristic but suffers if adequate controls are not in place

• Hedonic Pricing Model

– Can use data from a restricted period of time (cross-sectional analysis) or an extended period of time (time-series analysis) – note that this latter case requires adjustment to constant dollars – Can be used effectively to appraise homes due to extensive data set – however, constantly updating the data set is expensive and time consuming – Hedonic pricing is essentially a very large “Paired Sales” analysis with sufficient home sales and controls

Location Attributes - Relative Values

Location Characteristics Crematory Agee and Crocker (2008) Rawlings, WY

• 2% to -16%

Within a mile Superfund Gayer, et al (2000) Grand Rapids, MI

• 4% to -6%

Within a mile Superfund Kiel and Zabel (2001) Woburn, MA

• 15%

Within a mile Groundwater Pre-Remediation Case, et al (2006) Scottsdale, AZ and Tempe, AZ

• 7%

Currently Contaminated Groundwater Post-Remediation Case, et al (2006) Scottsdale, AZ and Tempe, AZ No difference Previously contaminated Waste Transfer Station Eshet, et al (2007) Israel

• 12%

Within a mile Industrial – Superfund Carroll, et al (1996) Henderson, NV

• 7%

Within a mile Lead Smelter Dale, et al (1999) Dallas, TX

• 0.8% to -4%

Within a mile Power Plant Davis (2008) Assorted

• 3% to -5%

Within 2 miles Earthquake Special Studies Zone Brookshire, et al (1985) Los Angeles & San Francisco,

• 3.3% to 5.6%

Inside Zone Distance to Beach Brookshire, et al (1982) Los Angeles, CA

• 1.4%

Per Mile from Beach Direct Water Access Thayer, et al (1992) Baltimore, MD 25.3% Water or Pier Access Total Suspended Particulates Brookshire, et al (1982) Los Angeles, CA

• 1.6%

1000 ug/m3 Foreclosures Lin, Rosenblatt, and Yao (2009) Chicago, IL

• 1.2% to -1.7%

0.9 kilometers Sex Offender Linden and Rockoff, 2006 North Carolina

• 4%

One-tenth mile Landfill – High Volume Ready (2005) Assorted

• 13%

Adjacent to landfill Landfill – Low Volume Ready (2005) Assorted 0% to -3% Adjacent to landfill Landfill Reichert, et al (1992) Cleveland, OH

• 5% to -7%

Within a few blocks Landfill Thayer, et al (1992) Baltimore, MD

• 1.3% to -5%

Within a mile Landfill Atkinson-Palombo and Hoen (2014) Massachusetts

• 12.2%

Within one-half mile School Quality Brookshire, et al (1982) Los Angeles, CA 0.2% Standardized Scores Transmission Lines Atkinson-Palombo and Hoen (2014) Massachusetts

• 9.3%

Within 500 feet Highways Atkinson-Palombo and Hoen (2014) Massachusetts

• 5.3%

Within 500 feet Beachfront Atkinson-Palombo and Hoen (2014) Massachusetts 25.9% Within 500 feet

Property Value Concerns for Wind Energy Fall Into Three Categories

• 1. Area Stigma: Concern that surrounding

areas will appear more developed

• 2. Scenic Vista Stigma: Concern over

decrease in quality of scenic vistas from homes

• 3. Nuisance Stigma: Concern that factors

that occur in close proximity will have unique impacts

Each of these effects could impact property values; the effects are not mutually exclusive

No one will move here! It will ruin my view! I won’t be able to live in my home!

LBNL Study Methods Built And Improved On Past Work

• Multiple U.S. wind project locations
• Valid residential sales values – not assessed values
• Large sample size of sales transactions (e.g., over 400)

near wind farm area

• Field visits to homes
• Hedonic pricing model
• Tested for all three potential effects
• Rigorously analyzed data and had results peer-reviewed

Research Questions

• Do views of turbines measurably affect home sales prices?
• Does proximity to turbines measurably affect home sales

prices?

• Are the results stable over time?
• Are there other observable impacts?

Collected Sales Data from 10 Study Areas Surrounding 24 Wind Farms in 9 States

3 Adjoining Counties Washington & Oregon 7 Facilities: 582 WTG, 790 Sales Howard Cnty, TX 46 WTG, 1,311 Sales Custer Cnty, OK 2 Facilities: 98 WTG, 1,113 Sales Lee Cnty, IL 103 WTG, 412 Sales Buena Vista Cnty, IA 5 Facilities: 381 WTG, 822 Sales Kewaunee Cnty, WI 2 Facilities: 31 WTG, 810 Sales Wayne Cnty, PA 43 WTG, 551 Sales Somerset Cnty, PA 3 Facilities: 34 WTG, 494 Sales Madison Cnty, NY Area 1: Madison 7 WTG, 463 Sales Madison Cnty, NY Area 2: Fenner 20 WTG, 693 Sales

7,459 Residential Sales Transactions

1,754 Pre-Announcement, 4,937 Post-Construction, and 768 Post-Announcement-Pre-Construction

Comparative Data

County Population Population/mi2 Median Age Median Income Median Home Value Benton, WA 159,414 94 34.4 $ 51,464 $ 162,700 Walla Walla, WA 57,709 45 34.9 $ 43,597 $ 206,631 Umatilla, OR 73,491 23 34.6 $ 39,361 $ 138,200 Howard, TX 32,295 36 36.4 $ 36,684 $ 60,658 Custer, OK 26,111 26 32.7 $ 35,498 $ 98,949 Buena Vista, IA 19,776 36 36.4 $ 42,296 $ 95,437 Lee, IL 35,450 49 37.9 $ 47,591 $ 136,778 Kewaunee, WI 20,533 60 37.5 $ 50,616 $ 148,344 Door, WI 27,811 58 42.9 $ 44,828 $ 193,540 Somerset, PA 77,861 72 40.2 $ 35,293 $ 94,500 Wayne, PA 51,708 71 40.8 $ 41,279 $ 163,060 Madison, NY 68,829 106 36.1 $ 53,600 $ 109,000 Oneida, NY 232,304 192 38.2 $ 44,636 $ 102,300 Livingston, IL 38,647 37 40.0 $ 47,887 $ 99,788/149,488

Hedonic Pricing Model

• Requires information on large number of sales (e.g.,

7,459) and home characteristics

• Home characteristics include

– Quantity Measures (e.g., square feet of living area, lot size, # of bathrooms, bedrooms, etc.) – Quality Measures (e.g., # of fireplaces, condition of home, presence of pool, air conditioning, scenic vista, etc.) – Location Specific Variables (e.g., local school quality, demographics, socioeconomic status, distance to important activities, environmental quality measures, etc.) – Variables of Interest (e.g., view of wind turbines, distance to wind turbines)

Four Qualitative Ratings Were Used for Dominance of View of Wind Turbines

Each home was given a view of turbines dominance rating, based on field visits

Minor Moderate Extreme Substantial

Area and Nuisance Stigmas

Distance to Nearest Turbine at Time of Sale Was Determined

Five Distance Bands

Nuisance Stigma

• Inside of 3000 Feet
• Between 3000 Feet

and 1 Mile Area Stigma

• Between 1 and 3

Miles

• Between 3 and 5

Miles

• Outside of 5 Miles

“Sold Homes” include all homes sold both before and after construction of the wind facility

Conclusions Based on the Research Sample

• Risks of property value impacts are often expected but
• ur research suggests that property value impacts

related to view and distance are not significantly different from zero. Specifically,

• Scenic Vista Stigma –no statistical evidence that

sales prices of homes with a view of the turbines were significantly affected (i.e., stigmatized) even if the view was “dramatic”

Base Hedonic Model Results

A Lack of Statistical Evidence that Views of Turbines Affect Sales Prices

• 1.2%

1.7%

• 0.5%

2.1%

• 25%
• 20%
• 15%
• 10%
• 5%

0% 5% 10% 15% 20% 25%

No View of Turbines (n=4207) Minor View (n=561) Moderate View (n=106) Substantial View (n=35) Extreme View (n=28)

Average Percentage Differences

The reference category consists of transactions for homes without a view of the turbines, and that occured after construction began on the wind facility

Average Percentage Differences In Sales Prices As Compared To Reference Category

Reference Category

No differences are statistically significant at the 10% level

Conclusions Based on the Research Sample, Continued

• Area Stigma – no statistical evidence that sales

prices of homes near wind facilities were significantly affected by those facilities as compared to other homes in the region

• Nuisance Stigma – no statistical evidence that

sales prices of homes within a mile of the nearest wind turbine were significantly affected by those facilities as compared to other homes in the region

• Timing – no statistical evidence of a trend in sales

prices of homes near turbines that is consistent with scenic vista, area, or nuisance stigma

Results from Alternative Models

• Repeat Sales Model – appreciation rates for homes

near the wind farms were not significantly different than appreciation rates for homes located farther from the wind farms

• Sales Volume Analysis – no statistical evidence

that the sales volume of homes near wind farms was different than the sales volume of home located farther from the wind farms

2013 LBNL Study

• Why do a second study?

– Initial LBNL study was most comprehensive, data rich analysis conducted

• Additional home Sales needed to find relatively small effects

– Small Number of Sales in Close Proximity – Dilution?

• Relative geometry
• Treatment / Control Groups

– Consistency/stability of results

• Massachusetts, Rhode Island

– Advanced Econometrics

• Spatial Analysis
• Difference/Difference Analysis

2013 LBNL Study

• 51,276 Home sales from 27 U.S. counties

related to 67 wind facilities

• 1,198 home sales within one mile of a wind

turbine

2013 LBNL Study Results

• Regardless of the dataset or specification,

there is no evidence that homes near

• perating or announced wind turbines are

impacted in a statistically significant fashion

Conclusion Relative to Proposed Project

• Based on the LBNL reports and the larger

literature, the Pleasant Ridge wind project in Livingston County, Illinois will not significantly reduce the sales prices of properties in the neighborhood of the wind facilities.

Additional Information

• Independence / Objectivity
• Analysis of Improved Residential Sales

– Appropriate focus – Use established scientific protocols – Methodology has been used for over 40 years – Evaluate a wide range of variables – Conduct extensive sensitivity analysis

• Use Assessor Data

– Electronic data – Actual Sales transactions and characteristics – Home Visits

Additional Information, Continued

• Multi-site Analysis, Pooling

– Transferability of results – Effect on Statistical Significance

• Appraisal/assessor background, guidelines

– Objective of analysis, methods employed

Alternative “Literature”

• Small, unrepresentative, non-transparent samples

– Selection process undefined

• Anecdotal information

– Plural of anecdote is not data

• Combination of sales, appraisals, assessments

– Lansink (twelve home sales, two areas)

• Vacant land rather than residential homes

– Kielisch; Gardner; Sunak and Madlener

• Insufficient controls for important influences
• Inappropriate analytical methods