Recently a number of investigations of The Spirit Level - an influential book by Richard Wilkinson and Kate Pickett which claims inequality drives a range of social maladies - have independently come to the conclusion that its empirical claims do not stand up to scrutiny. First, Christopher Snowdon's book The Spirit Level Delusion, then a report we published in English by a team of economists in Sweden - The Spirit Illusion, and finally a report for Policy Exchange by Peter Saunders - Beware False Prophets.
In an article for the latest edition of Prospect I asked how "all those evangelists for The Spirit Level will respond as many of its empirical claims fall apart like an ageing Trabant." So many prominent left wing politicians, activists and thinkers have praised the book to high heaven and are going to find it very inconvenient to admit it is wrong.
We're starting to get an answer. The Guardian recently wrote the critical reports off as polemics and evidence that the right is "spooked". Any of their readers who take the trouble to take the reports will be surprised to find they are actually quite careful dissections of the statistics in the book.
The authors of The Spirit Level published a defence of the book earlier this week. Yesterday we put up a response to that defence from the authors of the report we published. Here are a few of the key points:
- The startling statement in the book that Portugal is as innovative as the United States appears to be the result of not checking statistics from the Nationmaster website against the original source, the World Intellectual Property Organization.
- An article they cite as supporting their work, by Nobel laureate James Heckman, addresses a very different issue. The author told our Swedish team that "[t]his is a misrepresentation of my work".
- Another article they cite for support is about China, while their analysis is focussed on developed countries. That study only has six citations according to Google Scholar. By contrast, there are almost 100 times as many citations for a study in the authoritative Journal of Economic Literature that constitutes the most robust review of the literature on this issue but was not mentioned in The Spirit Level, presumably because it disagreed with their conclusions. This rather undermines the claim repeatedly made by Wilkinson and Pickett that their work is a fair representation of a finding that is well-established in the academic literature.
- The authors are having to back away from the prominent claim that "inequality kills", describing it as having "been made for us, in publishers’ publicity and in translation". Unfortunately, that admission is only coming now after they have used it since before even the publication of The Spirit Level, in an article publicising an earlier book by Wilkinson. It makes this, from the Guardian, a bit rich:
"The titles of the anti-egalitarian studies – which refer variously to The Spirit Level's "delusion", "illusion" and its "false prophesy" – reveal the polemical intent, a telling contrast with the meticulous subtitle of the original book: "Why more equal societies almost always do better."
Not quite meticulous enough to check their press releases evidently.
The evidence that The Spirit Level just isn't reliable is stacking up. The book's many avowed fans need to admit that.