Review of Wilkinson and Picket's "The Spirit Level"

The Spirit Level: Why More Equal Societies Almost Always Do Better by Richard G. Wilkinson
My rating: 4 of 5 stars


This book is easy to read, and highly thought-provoking. But I think the argument it presents is somewhat odder than is commonly appreciated - and a better appreciation of that oddness would probably act as a bit of a dampener to the enthusiasm with which it has been received amongst left-ish people.

For a start, Wilkinson and Pickett make explicit that they are not arguing for bigger government, but for greater income equality. They actually criticise spending on public services (such as health and education) on the grounds that they are "only partially effective", instead pushing redistribution as a cure-all that will improve health and education outcomes.

In UK terms this suggests that they disagree with the last Labour government's emphasis on, e.g. the schools rebuilding program and SureStart, and would have instead preferred the money spent on those things to have been spent on tax credits or welfare benefits, to redistribute it directly to the poor. It is far from obvious to me why this proposition should win automatic cheers from the left - which in turn makes me think that some praise has come from people who made only a superficial reading of the book.

I am also not convinced by the causal mechanisms that are supposed to be operating to ensure that greater income equality tends to result in better health & education outcomes. I have two doubts; the first about the direction of causation; the second about the precise mechanism at work.

The authors make a very strong argument that income inequality is positively correlated with a wide selection of social problems (mainly health and education related). They are of course aware of the old saw that correlation does not necessarily imply causation, but they have an answer, viz. that when one independent variable (income inequality) co-varies with several others (the health & education problems) then it is overwhelmingly likely that the former is the causal factor (unless there is some common cause at work - the authors consider various candidates and reject them all, I think convincingly).

But what they don't consider is that the health and education problems have a tendency to cluster, e.g. that someone with poor mental health is also likely to have poor physical health, and vice versa. If this is the case, then this makes it likely that a large part of the causal connection runs in the opposite direction to that which the authors suggest. This is because, if problems tend to cluster, then an increase in the level of problems (which presumably have a negative impact on people's ability to earn money) will reduce more people from middle-income levels to lower-income levels than from higher-income to middle (and so increase the amount of inequality) - and the opposite changes would to the reverse. This means that spending on health & education projects targeted on people with specific problems will actually reduce inequality - an important, intuitive, conclusion that Wilkinson & Pickett deny. If I am right about this then the policy mix practised by the last Labour government in the UK (some spending on health & education, some on redistribution via tax credits) is defensible after all.

Another thing about their argument against inequality that seems a little odd at first is that it depends on the effect that inequality has on people's psychological states, rather than on any argument about justice or efficiency as such. This leaves it open to the counter-suggestion that any psychological distress caused by greater inequality could be tackled in some other way than the redistributive measures they favour, e.g. therapy, promotion of stoic values by the education system. If it turns out that there is some good independent reason for thinking that toleration of inequality is conducive to social and economic well-being in general, even though it causes specific problems (e.g. stress), then it may turn out that the best all-round solution is to tackle the problems through targeted measures (e.g. therapy) rather than redistribute (and forgo any wider benefits inequality brings). Wilkinson & Pickett might, however, reply that their evidence for correlation between inequality and health problems suggests that no society has ever managed to pull off this alternative solution - as, if they had, then they would show up as an outlier in one of the graphs, with high inequality and low levels of health problems.

But in this case I am moved to wonder why mass therapy route has never worked? Is it that it has never been tried, or does the idea contain some built-in contradiction? This is all pretty speculative, but I think the latter is the more likely explanation. Elites have pretty strong incentives to encourage stoicism (docility) in the mass of the population, so they are not likely to overlook this option. I think the real problem is that as inequality-levels increase, governing elites and ordinary people become more sharply divided into two distinct tribes. As they listen to and trust each other less, so relations between them get worse, causing increased stress levels for both sides (although more so for the ordinary folk). Stoicism proves inadequate as a solution because it is unnatural for any people to submit to alien rule - and the elites begin to seem more like aliens as the gap widens. Rule by aliens is also inefficient - and hence stressful - because the rulers do not understand the ruled, and hence make elementary mistakes even when they are well-intentioned.

All of which is pretty much in the same ball-park as Wilkinson and Picket are in; so it is not so much that I want to contradict them on this point, but just to say that they could have been a little less simplistic about the causal mechanism at work, which they have running along fairly straight tramlines - from inequality to status anxiety to stress to health problems - whereas I think the interaction might be more complex, involving conflicts of power and interest between groups as stages on the way to the bad outcomes.

All the same, this is really no more than a quibble with a brilliant and important book.

Thoughts on Andrew Rawnsley's "The End of the Party"

The End of the Party: The Rise and Fall of New Labour by Andrew Rawnsley
My rating: 4 of 5 stars


This book certainly isn't short at 800-odd pages, so it is just as well that all blurb on the cover to the effect that it is a real page-turner is justified. Rawnsley creates a real sense of drama, and even tragedy, out of the story of Britain's Labour Government from its re-election in 2001 to its final defeat in 2010.

This book doesn't pretend in the slightest to be a general, or even a political, history of Britain in those years. Its focus is rather on the personal relations and political infighting at the very top of government - in a nutshell, the notorious "TB-GBs". Rawnsley's choice of emphasis is perhaps inevitable given the fame of his weekly political column in the London Observer based on similar material. But do the rest of us need to bother with these old squabbles, or can we safely leave their memory to political nerds?

One reason for making the effort is provided by the contrast between what was supposed to be going on in government according to the Party line swallowed by most UK media at the time, and what actually went on as revealed here (my assumption that the account is essentially accuracy can be justified by the thoroughness of the referencing, the impressive size of Rawnsley's contact book, and the "ring of truth" his story has throughout). Labour's leaders spent extraordinary amounts of their time, energy and politcal and emotional capital not on efforts to improve the public services they professed to love, but in an a secret, bitter struggle with each other for control of the government. In particular, whilst public Health and Education services were boosted by being given more money, efforts to reform the way they were delivered foundered on the factional infighting.

So what matters here is not just that the public's right to know about all this was partly frustrated, but also that the effectiveness of public service delivery was affected (this last point depends on another assumption - viz., that constructive high-level political input into things generally makes them better - but without this assumption I do not see how politics can matter at all).

Some important conclusions are suggested by these concerns. One line of argument might be that the toxic culture of infighting detailed here is stronger in the Labour Party than in its rivals, and so a reason why they should be preferred over Labour. Another line might be that the squalidness of the political process is a reason to minimize the extent to which politics impacts our lives, and hence a reason to reduce the role government plays in our society. Personally I am not persuaded by either of these lines, although I think a serious case could be mounted for either.

My own view is that the sad history of New Labour best supports the conclusion that Britain needs a radical rethink of its political system. Democracy is available in different flavours, and in Britain we have the more adversarial. The first-past-the-post voting system encourages the politically engaged to congregate into two opposing tribes who play up their inter-tribal rivalry whilst each suppressing any internal differences, because open discussion of the latter would undermine the contrast with the enemy that is supposed to motivate one's own side on to the effort needed to win power. The results are brilliantly documented in this book: Labour's heavyweights spent years slugging it out with each other in private about what should be the "dividing lines" between them and the Conservatives without ever coming to a clear conclusion. The policies that actually resulted showed a middle-way caution that, whilst not reprehensible in itself, belies the supposed advantages of the adversarial system (decisive government based on a clear contrast between binary alternatives). The more damning point is that the ability of the politicians to supply constructive input into policy was, if this history is to be believed, seriously limited by the intensity of their struggle with each other - an intensity that might (plausibly) be lessened if political differences were generally out in the open, as they tend to be in countries with more consensual political systems.


View all my reviews