Wind farms and abuse of statistics: bird edition
When "wind farms are dangerous" really just means "there are a lot of birds".
When dealing with large numbers, it helps to have an idea of the expected order of magnitude. That way, you can know whether it is merely the number which is large, or the thing it's describing as well.
For instance, if I tell you of a country with 140,000 people in long-term unemployment, it's rather important for you to know if I'm talking about the US (population 280 million) or Luxembourg (population 525,000).
That's a test a Spectator article, Wind farms vs wildlife, failed quite badly this week.
The author, Clive Hambler, is a lecturer in biological and human sciences at Oxford university, and quotes a number of statistics to demonstrate how dangerous wind farms are to wildlife. For instance:
Every year in Spain alone — according to research by the conservation group SEO/Birdlife — between 6 and 18 million birds and bats are killed by wind farms. They kill roughly twice as many bats as birds. This breaks down as approximately 110–330 birds per turbine per year and 200–670 bats per year. And these figures may be conservative if you compare them to statistics published in December 2002 by the California Energy Commission: ‘In a summary of avian impacts at wind turbines by Benner et al (1993) bird deaths per turbine per year were as high as 309 in Germany and 895 in Sweden.’
Similar claims are made throughout. Apparently the annual death toll of bats in the US and Canada is "up to three million", "Norwegian wind farms kill over ten white-tailed eagles per year", and so on.
What is missing is any context through which we can examine these numbers. It might change our interpretation of the figures to know that:
Domestic and feral cats have also been considered a major source of anthropogenic-caused mortality with estimates near 100 million annual bird deaths [in the US].
Or that, on roads near wetlands in Canada:
223 birds were killed per mile per year.
Power lines in the US are estimated to kill:
…approximately 130 million birds per year.
While we're banning things, we may want to keep an eye out for that scourge of the avian world, windows:
97.6 to 976 million bird deaths per year in the U.S. due to collisions with windows… based on an estimated 1 to 10 bird deaths per structure per year from a fatality study in New York.
All those figures come from a 2005 paper by the US Department of Agriculture.
In other words, even with the massive figures from Spain – figures which show deaths per turbine per year two orders of magnitude higher than equivalent figures cited in the above paper, which are based on an assumption that for every confirmed death, there's nineteen uncomfirmed, and which come from a set of guidelines which explicitly concludes wind farms are OK for birds if built correctly (pdf) – wind farms kill fewer birds than cats, power lines, roads or windows. That comparison would have been rather useful to include in the original piece. With that in mind, the numbers in the piece become less a demonstration of the awesome mortality of wind farms, and more a confirmation that yes, there are a lot of birds in the world.
It's been pointed out on Twitter that I'm not comparing like to like. Spain is smaller than the US, of course. Thankfully, the USDA also estimates the number of birds killed by wind farms in America: between 20,000 and 37,000 a year. I let you draw your own conclusions from the discrepancy.