Job polarization is a big problem. This is one message I take from figures released yesterday by the ONS showing employment by occupation.

These show that in the last ten years we've seen big rises in "professional occupations" and in managers, directors and senior officials - up by 23.3% and 15.5% respectively. But we've seen falls in secretarial and skilled jobs. And within the "unskilled" occupations, there's been a shift from machine workers to care workers; this is partly due to the 3.5% drop in manufacturing output during this period.

These numbers exclude the self-employed, and don't tell us anything about the nature of the employment contracts: some of the rise in caring jobs, I suspect, is in the form of precarious zero-hours contracts.

There's little sign of these trends ending. In the last 12 months, managerial and professional jobs have increased by 5.7% and 4.8% respectively whilst secretarial and skilled jobs have fallen.

This matters, I reckon, for at least six reasons:

- It helps explain why the graduate premium has been stable recently, despite rising numbers of graduates. It's because, as the IFS says, firms have created more graduate jobs. This increased demand for graduates is evident in relative wages: in the last ten years managers' wages have risen by 37.2%, twice the rate for secretarial or care workers.

- It suggests that social mobility will be low. The cliché of a man rising from the post-room to the boardroom will become even rarer, not just because there are no longer any post-rooms but because the middling jobs a post-boy could have risen to are also declining. It's hard to climb a ladder when the rungs are missing.

- It suggests that, in the absence of offsetting policies, we might see increases in household inequality simply because of increased assortative mating (pdf). In the 50s and 60s, inequality was mitigated by professional men marrying their secretaries. Now, they marry their fellow managers and lawyers.

- It poses questions about the feasibility of vocational education. Where will be the good jobs in the future for people who aren't academically inclined? This question is, of course, sharpened further by the possibility that technical change will destroy lots of jobs.

- It deepens the productivity puzzle. We might expect rising numbers of managers and professionals to be accompanied by rising aggregate productivity, if only as a compositional effect: managers and professionals have higher wages and so (in neoclassical theory!) are more productive. That this hasn't happened makes it even stranger that productivity has stagnated.

- It has contributed to social polarization. The EU referendum highlighted a divide between optimistic graduates and less educated people who are pessimists: the Ashcroft poll showed that a majority of leave voters believe that life will be worse for children than it was for their parents. This is consistent with the fact that falls in secretarial and skilled jobs show that there are fewer good jobs for on-graduates. Our socio-political divide has an economic base.

In these ways, job polarization isn't just an economic curiosity. It's a profound political problem - one that might not be cured merely by further expanding the number of graduates. If we had a serious political class, this issue would be high on the agenda. But we don't, so it won't be.