Book report: The Spirit Level

The Spirit Level puts forth the thesis that more-equal societies are better places for everyone; rich and poor both suffer from inequality. It marshalls many graphs and references to support the thesis. There seem to be two problems with the book. One, the authors are in love with their big idea, and make all the details fall in behind. Two, the data don’t necessarily support the thesis.

The book has been discussed, debated, debunked, and defended extensively, so I do wonder what I can constructively and accurately add. Let me start with some key links:

• The Equality Trust was founded to support the work of the authors to reduce inequality.
• The TaxPayers’ Alliance has taken on the book, its arguments, and its referencing. They found major problems with the research, including misrepresenting the findings of academic papers, cherry-picking of data, and non-replicability of findings.
• The Spirit Level Delusion is a book and blog that takes on the book’s ‘theory of everything’. It has new graphs and good discussions of the datasets, and reveals inconsistencies in the analysis.

Consider yourself fully warned.

Here is one tiny example, a synecdoche for the book’s general sloppiness. In the chapter ‘Building the Future’, the authors say:

The need to provide unrestricted free access in order to maximize the public benefit was offered as an economic explanation of why roads and bridges were in public ownership — until government began to try to recoup the costs of road building by charging tolls.

This says that we used to finance roads publicly, and then government invented toll roads. As it happens, I grew up not far from Little River Turnpike, which was a toll road over 200 years ago. Oh, and Columbia Pike was chartered in 1810. Toll roads aren’t a new idea, and it isn’t the fault of some neoliberal, pro-inequality conspiracy that we suddenly have them. By an accident of geography, I happen to know a bit of toll-road history. So I wonder, what else is wrong with the book? This is where the websites linked above are useful. They have three general charges:

• mis-representing other research
• cherry-picking data
• ignoring contrary findings.

Mis-representing other research is the most serious charge. It is one thing to ignore other research; it is something much worse to cite it to support of your argument when it doesn’t. The critics rightly say that this is exactly what Wilkinson and Pickett do. Take the work of James Heckman. I skimmed this report (pdf) from Cunha and Heckman (or see this similar NBER report). Heckman is talking about the impacts of inequality on people who are disadvantaged. The work isn’t about about how unequal societies perform worse for everyone, which is the thesis of The Spirit Level.

The complaints about cherry-picking data are sometimes important and sometimes not. I wouldn’t tend to be worried about whether Slovenia is in the dataset or not, for example. However, it is fascinating that the significance of the relationship depends on the countries chosen. That suggests that we may have an interesting thesis, but more work needs to be done. The complaints about choosing different series of data are also possibly non-issues. Working with these sorts of international datasets is hard and messy; compromise is inevitable. The two key things are: say what you’ve done (and why), and be prepared to back off your conclusions if new evidence calls them into question.

The complaints about ignoring contrary evidence work at two different levels. If we think just about the book, it isn’t an important criticism. The authors have a thesis and they put forward the evidence. They aren’t under any obligation to show you the contrary evidence. The second level, then, is how you as a reader and we as citizens decide to use this one book. The responsible thing is to look around and see what else is there. Are there alternative theories explaining the same data? Also, is there other research dis-proving the thesis?

One of the authors cited as providing contrary evidence is Angus Deaton. His book Economics and consumer behavior (with John Muellbauer) is an essential reference for microeconomic work, so I’m prepared to take his word/findings over Wilkinson and Pickett’s. I looked at one of Deaton’s papers. He says:

But it is not true that income inequality itself is a major determinant of population health. There is no robust correlation between life expectancy and income inequality among the rich countries, and the correlation across the states and cities of the United States is almost certainly the result of something that is correlated with income inequality, but that is not income inequality itself.

So, yeah, that pretty much says the opposite of The Spirit Level. Again, Wilkinson and Pickett are not required to tell you that this other research is out there. However, caveat lector means that you should inform yourself.

None of that, though, is the central problem with the book. As I said at the beginning, they are in love with their thesis. ‘Inequality’ functions for them as a fetish, in exactly the way that Zizek describes the role of ideology. Ideology can function as fetish or symptom, and in the modern world functions as both at once:

it enables you to fully participate in the frantic capitalist game while sustaining the perception that you are not really in it.

Wilkinson and Pickett propose to reorient the economy around a new fetish/symptom, ‘equality’. But, if inequality is as fundamental to some societies as the data suggest, it can’t be just a case of shifting them a bit or reorienting them. Real change would have to be more fundamental. Instead, the authors have covered over this conclusion with a bit of ideologic paste.