Increased inequality does not necessarily bring demands for redistribution.

I say this because of a recent paper by Elvire Guillard and Michael Zemmour. They studied the political attitudes of the well-off but not super-rich in 19 developed countries between the 80s and 00s - those in the 80^th to 99^th percentile of the income distribution. Such people's preferences matter because they might well have disproportionate political influence.

They found that increased inequality has ambiguous effects upon their demands for redistribution.

One the one hand, a growing gap between the top quintile and those immediately below them increases demands for equality. This is because a bigger gap tells the well-off that the cost of falling down the income scale is large, and people want insurance against this, such as a higher social wage - better health and education.

On the other hand, though, a bigger gap between the super-rich and the well-off actually reduces demand for equality. One reason for this might lie in something suggested (pdf) by Albert Hirschman in 1973: people tolerate rising inequality, at least initially, because they expect to join the super-rich - an expectation exacerbated by over-confidence. Alternatively, say Guillard and Zemmour, it might be that the well-off know that the super-rich won't pay their fair share of taxes and so fear it will be they who 'll bear the burden of higher redistributive taxation.

The point here is that what matters is not something as simple as the Gini coefficient, but rather a more precise structure of inequality.

Does this help explain the UK's odd political divide? I'm not sure. On the one hand, it does. Maybe Labour's popularity among apparently "middle class" workers owes something to the fact that job polarization has destroyed middling jobs which has increased insecurity and so created an insurance demand for equality. Also, because many pensioners have securish incomes, they don't fear a loss and so don't have such insurance demand. Hence their Tory preferences.

On the other, though, pensioners have even less hope of joining the super-rich than do well-paid workers. This should dispose them towards more leftist preferences - which is not what we see.

Guillard and Zemmour's work is, however, consistent with laboratory evidence which shows that increased inequality does not call forth demands for redistribution. One reason for this, says Kris-Stella Trump, is a form of anchoring effect: our belief in what's fair is shaped by the existing income distribution:

Public ideas of what constitutes fair income inequality are influenced by actual inequality: when inequality changes, opinions regarding what is acceptable change in the same direction.

Jimmy Charite, Raymond Fisman, and Ilyana Kuziemko have suggested (pdf) another mechanism - that inequality changes people's expectations, their reference points:

If voters tend to respect others' reference points, then if a country experiences a shock that increases income inequality, voters may be reluctant to tax away those gains.

Klaus Abbink, David Masclet and Daniel Mirza demonstrate a different mechanism - resignation. As inequality becomes extreme, they show, people simply give up fighting it*.

US politics is, I fear, consistent with all this; high inequality has given us a kleptocratic billionaire.

It's also consistent with world history as described by Walter Scheidel. He shows that significant falls in inequality have generally been brought about not by gentle redistributive policies but by wars, revolution, disease and state collapse.

Perhaps there is no stabilizing negative feedback loop from increased inequality towards demands for redistribution. If so, a sustained** increase in equality is far harder to achieve than social democrats would like to believe.

* Plus, of course, there's the fact that the richer the rich are, the more they can spend on entrenching their position by buying the media and lobbying.

** How much could a one-term Corbyn government do to permanently increase equality?