Today's labour market figures show that we're seeing job polarization - a relative decline of middling jobs.

My chart shows the point. It shows changes in employment by occupation since 2010, when employment troughed*. You can see there's been a fall in the number of administrative and secretarial jobs, and stagnation in the number of skilled tradespeople, but increase in the number of professional and "associate professional" jobs, and in the number of care workers. These trends would look stronger if we looked at changes in the last ten years; in this time, the number of skilled tradespeople has fallen.

Looking at wages, there's a more complicated picture. The clearest story here since the trough of the recession has been the relative rise in managerial pay. This isn't simply because this is more cyclical than other pay: their wages matched average pay during the 2007-10 downturn. What is perhaps cyclical, though, is the wages of skilled trades. These have outstripped average pay, but this doesn't recoup their fall in relative wages during the recession.

Despite rising demand and rises in the minimum wage the pay of care workers has declined relative to the average, perhaps because of an abundant supply.

We've also seen a relative decline in secretarial pay, which is consistent with declining demand for the occupation.

However, professional and associate professional pay has declined in relative terms. This is consistent with such jobs being less good than they used to be. Or maybe there's an element of job rebadging here: previously humble occupations have been renamed, dragging average pay down. Or perhaps what we've seen is demand for such occupations respond to a decline in relative price.

A big question here is: what does this mean for social change generally? I'd suggest (tentatively) two things.

One is that all this is ambiguous for equality.

One the one hand, there's an equalizing tendency here. A couple comprising a skilled tradesman and a secretary would have an above-average household income. But the gap between them and the worst-off has fallen, because of their declining relative pay.

But on the other hand, these changes are increasing inequality not just by raising managerial pay but also by encouraging assortative mating. Once upon a time, well-paid men married their modestly paid secretaries, and this tended to equalize household incomes. Today, there are fewer secretaries to marry. The equivalently placed man is therefore more likely to marry a wellish-paid female. And this exacerbates household inequality.

Perhaps, though, there's something else. These changes might justify the expansion of universities. Job opportunities for diligent but non-academic people - secretarial and craft work - have declined. Such folk need therefore to go to university to have a chance of a good job. Having a degree doesn't guarantee getting such a job. But it increases your odds.

There is, though, a more general point here. You don't have to believe in economic determinism to think that the labour market shapes our culture, with perhaps longish lags. It's plausible, therefore, that these occupational changes will lead to cultural change perhaps of sorts which we cannot yet predict. It's not at all clear to me that these changes will be wholly welcome.

* The numbers aren't seasonally adjusted, so I'm comparing 2017Q2 with 2010Q2.