The SuperFreakonomics chapter on global cooling is still being kicked from one end of the blogosphere to another, with error after tired delusionist error being pointed out. Most of the time, it’s just sloppy contrarianism of the type you might expect from people who hang around with rightwingers a lot and are in a rush to produce a controversial book. But there is one point that, coming from Steve Levitt, I find unforgivable. Before pointing it out, I’ll quote what I said about Freakonomics when it came out, in a post entitled “Getting the data to talk”
what Levitt has taken from the economics profession is not so much a body of theory to be applied, as a set of tools for empirical analysis and an unflinching willingness to look at social and policy issues without regard to social norms or received wisdom. More importantly, he’s combined all this with creative flair and an impressive capacity to see the right way of teasing compelling conclusions out of refractory data.
Looking back, I still think this judgement stands up as regards Freakonomics, which makes the tragedy of Superfreakonomics all the greater.
Given Levitt’s justified reputation as someone who knows more about data than I ever will, how could he put his name to this?
Then there’s this little-discussed fact about global warming: While the drumbeat of doom has grown louder over the past several years, the average global temperature during that time has in fact decreased.
and defend it by saying, as reported by AP that
he did not do any statistical analysis of temperatures, but “eyeballed” the numbers and noticed 2005 was hotter than the last couple of years.
Someone ignorant of such complex statistical concepts as variance might indeed be tempted to include that temperatures have flattened out over various periods in this data set (from 1983 to 1995, for example). Any competent social or natural scientist should be aware that a trend picked out by a selective choice of start and end dates is meaningless (the fact that a number of people with social or natural science qualifications, including a few who were, at least in the past, notable, have made such claims is a regrettable instance of how standards slip when ideology and wishful thinking get in the way). But Levitt, of all people, can’t claim ignorance as an excuse. And he doesn’t even try to pretend there is a structural break that would justify his cherrypicking. Unsurprisingly, real statisticians (the group to which Levitt formerly belonged) are crying foul.
Granted, this book is just a silly exercise in contrarianism. Still, it’s being presented as a serious piece of work, trading on Levitt’s well-justified reputation. After something like this, it’s hard to see how Levitt can distinguish himself from his former adversary John Lott.