John Rentoul has been causing trouble. "The gap between rich and poor has not changed significantly for about 20 years" he says - which is "at variance with the accepted story of food banks and greedy bankers".
In one important sense, he's right. According to the ONS, the share of post-tax, post-benefit incomes of the top 20% rose from 37 to 45% between 1977 and 1990, but has not changed much since; it was 44% in 2010-11 (table 26 here). And the share of the bottom 20% fell from 9 to 6% between 1977 and 1990, but is now also 6%.
However, these data - and measures of Gini coefficients - are not incompatible with stories about food banks and greedy bankers, for three reasons:
1. The story of inequality in recent years is not mainly about the top fifth or tenth, but about the growing incomes of the super-rich. According to the world top incomes database, the share of pre-tax income going to the best-off 1% rose from 9.8% in 1990 to 13.9% in 2009.
2. Similarly, simple measures of inequality can disguise what's happening to the very worst off. A given Gini cofficient can be compatible with either quite acute poverty or not, depending upon what's happening elsewhere in the income distribution. There is no necessary inconsistency between a stable Gini coefficient and the spread of foodbanks - because as Paul points out, the poor are less resilient to wealth shocks than the rich. Popular measures of inequality such as Gini coefficients or 90/10 ratios do not suffice to answer the Rawlsian question, of whether we are maximizing the position of the worst off.
3. Measures of inequality don't tell us how the rich get rich. The Gini coefficient is indifferent between the income of a Russian oligarch and Warren Buffett, or between that of Lionel Messi and Fred Goodwin. But as Acemoglu and Robinson stress, it matters enormously for the health of an economy (and society) whether the rich owe their wealth to extraction or production. Those of us who are unhappy with the wealth of the rich worry about its origin, not just - or even mainly - its size.
It's good that John is drawing our attention to statistics. But let's remember what they do and do not measure.