The Spirit Level and the imaginary £39 billion (part 1)

On Sunday, the Observer reported that ‘Inequality “costs Britain £39bn a year’. This is based on the belief that ‘a more equal UK would experience less crime and imprisonment, better mental health [and] higher healthy life expectancy’. This sounds very much like hypothesis put forward in The Spirit Level and although neither the Observer article nor the report it is based on mention the authors of that book – Richard Wilkinson and Kate Pickett – by name, the £39 billion claim comes from a report published by a think tank that they founded, the Equality Trust. It is unclear whether Wilkinson and Pickett wrote the Equality Trust report (no authors are listed) but the cost estimates appear to be based on the notoriously shaky scatterplots that appeared in The Spirit Level.

The short response to the claim that billions are lost to inequality each year is that the problems listed (crime, imprisonment, mental health and premature mortality) are simply not caused by inequality and therefore can have no cost. Only one of the issues that the Equality Trust says is caused by inequality (lower life expectancy) has received a significant degree of academic interest and three systematic reviews have rejected the hypothesis. The Oxford Handbook of Inequality (2009), for example, concludes that the ‘preponderance of evidence suggests that the relationship between income inequality and health is either non-existent or too fragile to show up in a robustly estimated panel specification. The best cross-national studies now uniformly fail to find a statistically reliable relationship between economic inequality and longevity.’

In the absence of any serious evidence that inequality affects the problems listed, it seems pointless to go any further, but the Equality Trust report contains several additional errors which are worth noting because they are all too common in ‘cost studies’ of this kind. All its estimates are based on a comparison between the UK today and a hypothetical UK which has lower inequality but, even taken on its own terms, the numbers do not add up.

Firstly, it claims that ‘in a more equal UK, people could expect an extra eight and half months of healthy life’ which, it says, would be worth £12.6 billion a year. This figure is arrived at by multiplying the total number of life years that would be ‘saved’ by greater inequality by £20,000. Why £20,000? Because that it what NICE says it would pay, per person, for a drug that saved a year of life. Like many government departments, NICE devises hypothetical estimates of what they are prepared to pay to prolong lives because resources are limited and they do not have the money to pay for everything. These valuations are used to help them provide the biggest bang for their bucks.

None of this implies that a life year lost ‘costs’ the state £20,000. On the contrary, rising life expectancy in the UK, as in other developed countries, results in higher costs to the state because the elderly are net consumers of tax revenue through pensions, nursing care, free bus passes, medical treatment and so on. If lower inequality led to longer life expectancy (which it doesn’t), it would cost taxpayers a significant sum of money, not save them £12.6 billion.

Secondly, the authors claim that reducing income inequality would result in a 33 per cent decline in murders. This is based on one of the more absurd findings in The Spirit Level, which uses the exceptionally high homicide rate in the USA in the 1990s as evidence that ‘less equal’ countries have a higher murder rate in general. The truth is that the USA is a massive outlier and there is no correlation between inequality and murder amongst the other rich nations. As Tim Harford pointed out on More or Less some time ago, Wilkinson and Pickett’s own evidence shows that the average murder rate of the ‘most equal’ countries is higher than that of Britain.

Nevertheless, the authors claim that a 33 per cent decline in homicides would save £678 million. This is based on a report from the UK Peace Index which estimates the cost of one murder at £3,601,865. Although the Equality Trust authors state that this is an ‘economic cost’, most of the figure refers to estimates of ‘emotional impact’ and other costs which are purely internal and/or hypothetical. The largest economic cost is 'lost ouput', which is mainly an internal cost to the victim (who is, of course, dead). None of these are costs to the taxpayer. The legitimate external costs (to the health care and criminal justice systems) are estimated at less than £200,000 – less than 6 per cent of the ‘cost’ estimate used by the Equality Trust.

Murder is, of course, too serious an issue to be reduced to nickels and dimes, but non-financial and/or internal costs should be presented as such. It is also worth noting that the Equality Trust authors use data from the 1990s to form their estimates. There is no acknowledgement that the UK’s murder rate has fallen – remarkably – by half since then. It should also be noted that the big rise in the post-war homicide rate occurred between 1965 and 1980 when inequality was relatively low.


Part two of this blog post will be published tomorrow.