DOE study says wind farms don’t affect property values—but…
Human impacts, Science, Wind turbines Add commentsA detailed statistical analysis of 5000 homes sold within ten miles of wind farms has failed to find any clear relationship between sales price and proximity to, or views of, industrial wind farms. However, close reading of the results raises some questions about trends within a mile of turbines, and the authors recommend more detailed study of the closest homes as a top priority for future research. Co-author Ryan Wiser affirmed that “It is possible that individual homes have been impacted, and frankly, I think it would be a bit silly to suggest otherwise. Human development impacts property values.”
As I look closely at the result, it seems likely that the apparent trend toward some property value effect largely mirrors surveys of residents near wind farms. The “problem” in interpreting this data and the surveys is that there is NOT a universal increase in annoyance or sleeplessness or dropping property values as you move closer to turbines; rather, there is an increasing MINORITY of neighbors who are negatively impacted. Several earlier posts address this factor: from 10% of neighbors at around a mile or so, increasing to 25% or so at a half mile, and perhaps 40% at a quarter mile, survey results reflect the well-known individual variability in susceptibility to noise. It seems clear that the “small and infrequent” numbers of homes negatively affected in the study addressed here, are mostly concentrated near turbines, and may represent a similar percentage of landowners as report sleep disturbance and other noise impacts. If so, the 5% average price hit is apt to represent a much larger valuation drop in a quarter or so of the homes that are within a mile.
The study, funded by the US Department of Energy at the Lawrence Berkeley National Lab, took into account size and other features of the homes, and looked particularly for any trends in prices resulting from views of wind turbines, having turbines in the vicinity (1-5 miles), and nuisance factors like shadow flicker and noise. This third focus is the one of most interest to AEI, and we took a close look at what they found. We were not surprised that merely seeing wind turbines had no significant effect on property values, despite a vocal minority of people who dislike seeing them. We did expect, however, that being very close to turbines, close enough that noise and vibration issues may arise, would have an impact. While the study found no statistically significant relationship, a closer look at the data reveals some troubling trends.
To assess whether nuisance factors affected home prices, the researchers compared sales within a mile of turbines to sales five miles or more from turbines. Their distance classifications included 0-3000 feet, 3000-5280 feet, 1-3 miles, 3-5 miles, and over 5 miles. While all zones beyond a mile showed virtually no differences in sales prices, both of the classes under a mile had modest 5% declines in value (interestingly, the decline was slightly more extreme between 3000 feet and a mile). However, there were not enough sales of homes this close (125 under a mile, compared to roughly 4000 at 1-5 miles, and 870 beyond 5 miles) to provide statistical significance; that is, the margin of error is greater than the 5% change found, which means that with a larger number of sales, the average change in value might move to zero (or, just as likely, to -10%). Despite the limitations of the small sample size, it is striking to look at the data charts presented in the paper, and to see that of all the wind farm-related factors that were being considered, only proximity nudged the values away from the baseline average. In the context of the rest of the findings, the change co-efficients (minus.05 and .06) actually jump out of the data.
At the risk of making a mountain out of a statistical molehill, I would note that these results correspond very closely to survey results of residents around wind farms, which suggest that for a small but noticeable minority of residents within a mile, and especially within a half mile, noise and vibration from turbines can be very noticeable, to the point of significant sleep disruption and physical and psychological stress. These homes, which can include up to 40% of those within a quarter mile, a quarter of those within a half mile, and a small number more out to a mile, are likely to account for nearly all of the reduced property values found in the DOE study. It seems quite probable that the 5% lower average value found in homes within a mile is concentrated in no more than a quarter of these properties, suggesting that a significant minority of homes may experience a dramatic loss in value. Missing in the DOE analysis, as in nearly all commentary and reports regarding wind farm noise (both people with problems and those for whom it’s no big deal), is information on whether the turbines are upwind or downwind of the properties being analyzed. Again, properties that are downwind are more likely to be affected with either noise issues or lost property value, so that any averaging of results will give a brighter picture than may be experienced by nearby down-winders. It is also worth remembering that in the US, it is common to allow turbine siting as close as 1000-1500 feet. Future studies should concentrate on homes within a mile or so, perhaps with subsets of 0-1500 feet, 1500-3000 feet, and 3000-4500 feet, and 4500-6000 feet; it may be difficult to get enough data points for so many subsets, but given the noise annoyance reports to date, zeroing in like this is more likely to find a connection, if there is one.
A fascinating and important trend is worth noting: values of homes under a mile from PROPOSED wind farms was dramatically depressed (15%) after announcement of the wind farm and prior to construction, with values rebounding to within 5% of expected norm once they were up and running. This is in line with general attitude/opinion surveys as well, where dislike of wind farms is often most pronounced prior to construction, and moderates after people have lived with them for a while. Again, though, this is looking at average results, and there is of course a significant minority of wind farm neighbors whose initial opinion actually declines once they are running, generally due to unexpected noise issues.
December 3rd, 2010 at 5:05 pm
[…] in my reading the results are subtler than the overall averages suggest, just as they were with the big DOE-funded study that came out about a year […]
December 16th, 2010 at 8:02 pm
[…] large studies of property values have found no clear correlation between distance and price, they may not be as definitive as the industry would like to believe — there have also been widespread reports of unsellable […]
January 17th, 2011 at 2:12 am
Thank you. The fact that the analysis is not predictive when actual values are reviewed, as with actual assessments for Adams Co. Illinois, indicates that the model is flawed. As in any scientific endeavor, when a model turns out to be impeached by the facts, then it is the model that is scrapped.
July 13th, 2011 at 3:35 pm
[…] significant change. This study uses a hedonic analysis methodology similar to two previous studies (Hoen and Hinman) that found no significant price change. This new study, by Martin Heintzelman and […]
April 16th, 2012 at 4:34 pm
[…] noted in AEI’s previous coverage of property values research, while there is little evidence of decreased property values due to seeing turbines in the […]