Lawrence Berkeley National Laboratory (Berkeley Lab) along with University of Connecticut analyzed more than 122,198 home sales near 26 wind facilities (with over 1,500 within a mile of operating turbines) between 1998 and 2012 in densely populated Massachusetts, yet was unable to uncover any impacts to nearby home property values.
"This is the third of three major studies we have conducted on this topic [the first was published in 2009, and the second last August], and in all studies [using three different datasets] we find no statistical evidence that operating wind turbines have had any measureable impact on home sales prices," says Ben Hoen, a co-author of the new report and a researcher in the Environmental Energy Technologies Division of Berkeley Lab.
One of the unique contributions of this most recent study is that impacts from turbines as well as a suite of other environmental amenities and disamenities were investigated. The study found strong evidence that highways, major roads, electricity transmission lines, open space and beaches impact property values, but no similar evidence was uncovered for turbines.
"When we find our model so accurately predicts impacts from other amenities and disamenities, we are considerably more confident of our findings for turbines", says lead author Carol Atkinson-Palombo, Assistant Professor in the Department of Geography of the University of Connecticut.
This study, the most comprehensive to-date, in terms of numbers of transactions, builds on both the previous U.S.-wide Berkeley Lab studies as well as a number of other academic and published U.S. studies, which also generally find no measureable impacts near operating turbines.
Download the new 2014 UConn / LBNL report "Relationship between Wind Turbines and Residential Property Values in Massachusetts"
To register for a related webinar on the New 2014 LBNL / UConn Report at 12:30 PM Eastern Time, January 22nd, 2014 go here: Webinar
Download the 2013 LBNL report "A Spatial Hedonic Analysis of the Effects of Wind Energy Facilities on Surrounding Property Values in the United States"
Download the 2009 LBNL Report "The Impact of Wind Power Projects on Residential Property Values in the United States: A Multi-Site Hedonic Analysis"
The home sales analyzed in this study occurred in one of four periods based on the development schedule of the nearby turbines,
Despite the consistency of statistical significance with the controlling variables, statistically significant results for the variables focusing on proximity to operating turbines are either too small or too sporadic to be apparent. Post-construction home prices within a half mile of a wind facility are 0.5% higher than they were more than 2 years before the facility was announced (after controlling for
Post-announcement, pre-construction home prices within a half mile are 2.3% lower than their pre-announcement levels (after controlling for inflation/deflation), which is also a non-significant difference, though one of the robustness models suggests weak evidence that wind-facility announcement reduced home prices.
An additional tangential, yet important, result of the analysis is the finding of a statistically significant “pre-existing price differential”: prices of homes that sold more than 2 years before a future nearby wind facility was announced were 5.1% lower than the prices of comparable homes farther away from the future wind location. This indicates that wind facilities in Massachusetts are associated with areas where land values are lower than the surrounding areas, and, importantly, this “pre-existing price differential” needs to be accounted for in order to correctly measure the “post construction” impact of the turbines.
Finally, [the] analysis finds no evidence of a lower rate (i.e., frequency) of home sales near the turbines.
.. The effects of wind turbines may be somewhat context specific. Nevertheless, the stability of the results across models and across subsets of the data, and the fact that they agree with the results of existing literature, suggests that the results may be generalizable to other U.S. communities, especially where wind facilities are located in more urban settings with relatively high-priced homes.
Discussion of Findings in Relation to Research Questions
Q1) Have wind facilities in Massachusetts been located in areas where average home prices were lower than prices in surrounding areas (i.e., a “pre-existing price differential”)?
To test for this, [the authors] examined the coefficient in the prioranc period, in which sales occurred more than 2 years before a nearby wind facility was announced. The -5.1% coefficient for the prioranc period (for home sales within a half mile of a turbine compared to the average prices of all homes between a half and 5 miles) is highly statistically significant (p-value < 0.000). This clearly indicates that houses near where turbines eventually are located are depressed in value relative to their comparables further away. Other studies have also uncovered this phenomenon (Hoen et al., 2009; Hinman, 2010; Hoen et al., 2011). If the wind development is not responsible for these lower values, what is?
Examination of turbine locations reveals possible explanations for the lower home prices. Six of the turbines are located at wastewater treatment plants, and another eight are located on industrial sites. Some of these locations (for example, Charlestown) have facilities that generate large amounts of hazardous waste regulated by Massachusetts and/or the U.S. Environmental Protection Agency and use large amounts of toxic substances that must be reported to the Massachusetts Department of Environmental Protection. Regardless of the reason for this “pre-existing price differential” in Massachusetts, the effect must be factored into estimates of impacts due to the turbines’ eventual announcement and construction, as this analysis does.
Q2) Are post-construction (i.e., after wind-facility construction) home price impacts evident in Massachusetts, and how do Massachusetts results contrast with previous results estimated for more rural settings?
To test for these effects, [the authors] examine the “net” postcon effects (postcon effects minus prioranc effects), which account for the “pre-existing price differential” discussed above. In the base model, with a prioranc effect of -5.1% and a postcon effect of -4.6%, the “net” effect is 0.5% and not statistically significant. Similarly, none of the robustness models reveal a statistically significant “net” effect, and the range of estimates from those models is -2.6% to 2.8%, effectively bounding the results from the base model....
These postcon results conform to previous analyses (Hoen, 2006; Sims et al., 2008; Hoen et al., 2009; Hinman, 2010; Carter, 2011; Hoen et al., 2011). [Their] study differed from previous analyses because it examined sales near turbines in more urban settings than had been studied previously. Contrary to what might have been expected, there do not seem to be substantive differences between [the] results and those found by others in more rural settings, thus it seems possible that turbines, on average, are viewed similarly (i.e., with only small differences) across these urban and rural settings.
Q3) Is there evidence of a post-announcement/pre-construction effect (i.e., an “anticipation effect”)?
To answer this question, [the authors] examine the “net” postancprecon effect (postancprecon effect of -7.4% minus prioranc effect of -5.1%), which is -2.3% and not statistically significant. This base model result is bounded by robustness-model postancprecon effects ranging from -4.6% to 1.6%. One of the robustness
models reveals a weakly statistically significant effect of -4.6% (p-value 0.07) when the set of data screens is relaxed. It is unclear, however, whether these statistically significant findings result from spurious data or multi-collinear parameters, examination of which is outside the scope of this research. Still, it is reasonable to say that these post and precon results, which find some effects, might conform to effects found by others (Hinman, 2010), and, to that extent, they might lend credence to the “anticipation effect” put forward by Hinman and others (e.g., Wolsink, 2007; Sims et al., 2008; Hoen et al., 2011), especially if future studies also find such an effect. For now, [the authors] can only conclude that there is weak and sporadic evidence of a postancprecon effect in our sample.
Q4) How do impacts near turbines compare to the impacts of amenities and disamenities also located in the study area, and how do they compare with previous findings?
The effects on house prices of [their] amenity and disamenity variables are remarkably consistent with a priori expectations and stable throughout our various specifications. The results clearly show that home buyers and sellers accounted for the surrounding environment when establishing home prices. Beaches (adding 20% to 30% to price when within 500 feet, and adding 5% to 13% to price when within a half mile), highways (reducing price 4% to 8% when within 500 feet), and major roads (reducing price 2% to 3% when within 500 feet) affected home prices consistently in all models. Open space (adding 0.6%-0.9% to price when within a half mile), prisons (reducing price 6% when within a half mile), landfills (reducing price 13% when within a half mile) and electricity transmission lines (reducing price 3%-9% when within 500 feet) affected home prices in some models.
[Their] disamenity findings are in the range of findings in previous studies. For example, Des Rosiers (2002) found price reduction impacts ranging from 5% to 20% near electricity transmission lines; although those impacts faded quickly with distance. Similarly, the price reduction impacts [they] found near highways and major roads appear to be reasonable, with others finding impacts of 0.4% to 4% for homes near “noisy” roads (Bateman et al., 2001; Andersson et al., 2010; Blanco and Flindell, 2011; Brandt and Maennig, 2011). Further, although sporadic, the large price reduction impact we found for homes near a landfill is within the range of impacts in the literature (Ready, 2010), although this range is categorized by volume: an approximately 14% home-price reduction effect for large-volume landfills and a 3% effect for small-volume landfills. The sample of landfills in [the] study does not include information on volume, thus [they] cannot compare the results directly.
[The authors] amenity results are also consistent with previous findings. For example, Anderson and West (2006b) found that proximity to open space increased home values by 2.6% per mile and ranged from 0.1% to 5%. Others have found effects from being on the waterfront, often with large value increases, but none have estimated effects for being within 500 feet or outside of 500 feet and within a half mile of a beach, as [they] did, and therefore [they] cannot compare results directly.
Q5) Is there evidence that houses that sold during the post-announcement and post-construction periods did so at lower rates than during the pre-announcement period?
To test for this sales-volume effect, [the authors] examine the differences in sales rate in fixed distances from the turbines over the various development periods. Approximately 0.29% percent of all homes in [their] sample (i.e., inside of 10 miles from a turbine) that sold in the prioranc period were within a half mile of a turbine. That percentage increases to 0.50% in the postancprecon period and then drops to 0.39% in the postcon period for homes within a half mile of a turbine. Similarly, homes located between a half mile and 1 mile sold, as a percentage of all sales out to 10 miles, at 1.9% in the prioranc period, 1.8% in the postancprecon period, and 2.2% in the postcon period (and similar results are apparent for those few homes within a quarter mile). Neither of these observations indicates that the rate of sales near the turbines is affected by the announcement and eventual construction of the turbines, thus we can conclude that there is an absence of evidence to support the claim that sales rate was affected by the turbines.
Previous studies using this hedonic modeling method largely have agreed that post-construction home-price effects (i.e., changes in home prices after the construction of nearby wind turbines) are either relatively small or sporadic. A few studies that have used hedonic modeling, however, have suggested significant reductions in home prices after a nearby wind facility is announced but before it is built (i.e., post-announcement, pre-construction) owing to an “anticipation effect.” Previous research in this area has focused on relatively rural residential areas and larger wind facilities with significantly greater numbers of turbines.
Four of the most commonly cited previous studies (Carter, 2011; Heintzelman and Tuttle, 2012; Hinman, 2010; and Hoen et al., 2011) analyzed a combined total of 23,977 transactions, whereas the present study analyzes more than five times that number.
January 9, 2014
by Carol Atkinson-Palombo 1 and Ben Hoen 2
1. Assistant Professor, Department of Geography, University of Connecticut
2. Staff Research Associate, Lawrence Berkeley National Laboratory
A Joint Report of University of Connecticut and Lawrence Berkeley National Laboratory
Lawrence Berkeley National Laboratory addresses the world's most urgent scientific challenges by advancing sustainable energy, protecting human health, creating new materials, and revealing the origin and fate of the universe. Founded in 1931, Berkeley Lab's scientific expertise has been recognized with 13 Nobel prizes. The University of California manages Berkeley Lab for the U.S. Department of Energy's Office of Science. For more, visit www.lbl.gov. Created by the Green Jobs Act of 2008, the Massachusetts Clean Energy Center (MassCEC) is dedicated to accelerating the success of clean energy technologies, companies and projects in the Commonwealth—while creating high-quality jobs and long-term economic growth for the people of Massachusetts. Since its inception in 2009, MassCEC has helped clean energy companies grow, supported municipal clean energy projects and invested in residential and commercial renewable energy installations creating a robust marketplace for innovative clean technology companies and service providers. The research was supported by the U.S. Department of Energy's Office of Energy Efficiency and Renewable Energy and by Massachusetts Clean Energy Center.