Forget all that guff about a growing gap between rich and poor, says Philip Hammond. Pay no heed to Jeremy Corbyn when he talks of how he will deliver for the many not the few, Theresa May says. The chancellor and the prime minister might not always see eye to eye, but on this at least they are in unison: income inequality is at its lowest since the mid-1980s.
May regularly makes this point when she is fending off attacks on inequality from Corbyn at prime minister’s questions. Buried away in the depths of the budget speech, Hammond said exactly the same. It’s not hard to see why. The Conservatives are uncomfortable with the idea that they are the party of the 1%, so evidence to the contrary is mighty useful.
For many in the world beyond Westminster, the notion that there is less income inequality in Britain since Margaret Thatcher was in her pomp sounds unbelievable. It doesn’t seem to square with the idea that wages are being squeezed for the gig economy workers at the bottom and going up for those at the top.
Look at the official data, the prime minister and the chancellor would say if challenged. It is not us saying that income inequality is at its lowest for three decades, it is the Office for National Statistics. Point proven.
Except that it isn’t. The government has two ways of calculating income inequality, and Hammond and May are using the one that best supports their argument. The ONS measure does indeed show inequality at a 30-year low but only because it doesn’t take account of what’s happening to the incomes of the top 1%. An alternative measure provided by the Department for Work and Pensions, which takes more account of what is going on at the very top of the income distribution comes up with a different conclusion: inequality is not at a 30-year-low. At best it is flat, but it appears to be gently rising. Disraeli’s famous dictum about there being lies, damned lies and statistics springs to mind.
Here’s what’s going on. Inequality is measured by the Gini co-efficient, which looks at how income is divided up across the nation. A Gini coefficient of zero would mean that everybody had the same income; a Gini coefficient of 1 would mean a single person took everything. In reality, every country’s Gini co-efficient falls somewhere between zero and 1: the UK’s moved sharply higher in the 1980s when tax cuts at the top coincided with widespread job losses in manufacturing.
The ONS’s estimate of the Gini coefficient comes from its annual publication The Effects of Taxes and Benefits on Household Income (ETB). This shows that inequality peaked at around 0.37 at the end of the 1980s, was still at around 0.35 in the mid-2000s, and has fallen to around 0.32, according to the latest available data.
Turn to the DWP and it is a different story. The Gini coefficient is higher than it is according to the ONS (0.35 before housing costs) and on an upward trend.
The discrepancy is the subject of a new report from the Resolution Foundation, a thinktank that focuses on the living standards of those on middle incomes and below. Research by the foundation’s Adam Corlett shows that both the ONS and DWP are failing fully to take account of the incomes of the very rich. Neither, for example, takes account of inheritance or capital gains, both of which are relevant when it comes to how income is distributed. But while neither are perfect, the ONS approach is particularly bad at capturing what is happening at the top.
“In the case of the top 0.1%, often only around half of income is captured. An attempt is made to correct this in Households Below Average Income (the DWP measure) using tax data from HMRC. No such attempt is currently made in the ONS’s ETB and so this data greatly underestimates the scale of top incomes. This has a large effect on inequality trends too, as the share of income going to the top has increased.”
The divergence between the two measures, Corlett adds, is due to the fact that the DWP is correcting – albeit in a crude way – for the lack of full data about the incomes of the rich. So despite often coming out first, the ONS is not the best guide to what is happening to inequality.
“Using the ETB data it would be possible to say that the top 1%’s income share has been lower over the last three years than at any time since the mid-1980s. However, sources that use HMRC data show that to be clearly incorrect.”
Does any of this actually matter? In an era of “fake news” it is clearly important that official statistics are viewed by an increasingly sceptical public as unbiased and reliable. That is especially true when it comes to politically sensitive areas such as inequality, which has become a much more contentious issue in the decade since the financial crisis.
What’s more, many of the tax changes made over the past decade – by both Labour and Conservative governments – have benefited the better off, and these would be far less politically defensible if those responsible for them had to reveal what is really happening to inequality. Taken together, the tax changes are not really compatible with the UK’s commitment under the 2030 UN sustainable development goals to reduce inequality within and among countries.
During Gordon Brown’s time as prime minister, Labour made the capital gains tax regime a lot more generous. It also introduced entrepreneurs relief, under which there was a 10% rate for the first £1m of qualifying gains. Subsequently that rose to £2m, £5m and now £10m. Dividend income, the preserve of the seriously well-off, is tax free up to £5,000 a year. To be sure, there have been moves in the other direction. Labour pushed the top rate of income tax up to 50% (later reduced to 45% by the Conservatives) and the personal allowance has been withdrawn for those earning more than £100,000 a year. But if inequality is rising (which it is) and the government wants to do something about (which it says it does), the tax system should be far less skewed in favour of the top 1%.
The Resolution Foundation has recommended 17 steps that would improve the accuracy, consistency, breadth and reliability of household income and inequality statistics. It says the problem is not really the ONS, which is doing its best to improve, but with the government.
This is not just a matter of resources, although that is certainly an issue. It is also that ministers are using statistics they know (or ought to know) are unreliable in order to make a political point. Voters deserve better than that.
• Follow Guardian Business on Twitter at @BusinessDesk, or sign up to the daily Business Today email here.
