The IFS reported this week that income inequality, as measured by the Gini coefficient, hasn't much changed since the early 1990s. This does not, however, mean that it is not a problem. There are several reasons why it is.

One is purely statistical. Survey-based measures of income might understate top incomes and hence inequality. Stephen Jenkins and colleagues point out that tax-based measure of inequality show a bigger increase since the late 90s than survey-based ones.

That, though, isn't my main point.

First, a stable Gini coefficient is consistent with very different types of society. A big reason why the Gini coefficient hasn't changed much since the early 90s is that a rising share of income going to the top 1% has been offset by a decline in the relative incomes of the 90^th percentile - and recent years, the latter's absolute incomes have fallen. This means that the UK shifted from being a "bourgeois society" - with a big middle class, to a "winner take all" society in which there are high top incomes but increased support for the worst off.

This is not obviously a healthy development. It might be a symptom of the degradation of erstwhile "middle-class" jobs, which might generate increasing discontent in future among people who have incurred massive student debt in an effort to get such jobs. And if you believe Deidre McCloskey's work showing that a big middle-class has healthy cultural effects - by promoting virtues such as prudence and honesty - it is potentially corrosive of a healthy society.

Which brings me to a another point. Even if inequality has stopped rising, the adverse effects of its earlier rise are still with us. To adapt Joseph Schumpter's insightful if tasteless metaphor, if a man has been hit by a truck, he does not return to health when the truck stops. We have strong evidence that inequality generates distrust. The hostility to elites which contributed to Brexit - and to a debased political discourse - is in part a product of the rise in inequality (in the sense of higher top incomes and increasing social distance) we saw in the 90s and early 00s.

Also, it's quite possible that higher inequality has other long-term cultural consequences - such as a Messiah complex, decreased job satisfaction or increased insecurity - which depress growth. Those who seek to defend inequality must answer the question: if increased inequality doesn't help explain worsening macroeconomic performance since around 2000 - lower investment and slower growth in GDP and productivity - what does?

Focussing upon inequality statistics, however, misses an important point. What matters is not just the level of income inequality, but how that inequality arose. A free market society in which high incomes arise from the free choices of consenting adults - as in Robert Nozick's Wilt Chamberlain parable - might have the same Gini coefficient as a crony capitalist society. But they are two different things. A good reason to be worried about current inequality - even if it hasn't changed - is that it is a symptom of market failures such as corporate welfare, regulatory capture or the implicit subsidy to banks.

In this context, what matters is not just inequalities of income but inequalities of power. Top footballers and top bankers might be earning similar sums, but one's salary is the product of market forces and the other of a tax-payer subsidy. The freelancer on £30,000 who's worrying where his next contract is coming from has similar income to the bullying middle managers who created intolerable working conditions at (for example) Sports Direct. But they have very different degrees of economic power. And the low income that results from having to take a lousy job where your wages are topped up by tax credits gives you much less power than the same income that would come from a basic income and the freer choice to take or leave a low wage job.

My point here is a simple one. There are very good reasons why we should worry about inequality - not just leftists but also rightists who want freer markets and "bourgeois" virtues. Focusing only upon the stability of the Gini coefficient is a form of statistical fetishism which overlooks important questions.