There's one point about the government's proposed benefit cap that's so obvious I'm embarrassed to point it out. It's that spending on non-pension benefits is not so much a fiscal problem as a labour market one, and an international one at that.

What I mean is that since the 1970s there's been a decline in demand for unskilled (pdf) workers across the western world. This has led to a combination of worklessness, economic inactivity and/or poverty wages across Europe and the US. This was the main reason why, even before the crisis, over five million (pdf) working-age Brits were claiming benefits. And it's this that explained why spending on benefits rose as a share of GDP so much between the 1970s and 90s.

In this context, the UK is not obviously an outlier. The OECD estimates that in 2010 (the latest year for which we have figures), only 56% of UK people who left school before 18 were in work. That compares to 85.1% for people with degrees. But these numbers are close to the OECD averages, of 55.5% and 83.1% respectively.

The UK's problem, then, isn't that we have an abnormal number of workshy unemployables. Nor is it - as my chart shows - that working age benefits are unusually high as a share of GDP right now. Instead, it's that we have been victims of a worldwide problem.

This suggests that if we want to cap working age welfare benefits, we should get more people into work.

And herein lies a hope,. It could be that the relative fall in demand for unskilled workers has stopped recently; it has been middlingly-skilled workers who have been displaced by technology recently, not grunt workers. This is consistent with benefit spending falling as a share of GDP in the 00s.

Whether this can continue is, of course, impossible to say; knowledge of the future is a contradiction in terms. My point is simply an obvious one which Keynes made 80 years ago:

Look after the unemployment, and the Budget will look after itself.