The Times' headline writers have given us a nice example of journalists' habit of equating the economy and the deficit. They've titled a piece by Ed Conway on how low wage growth has depressed tax revenues: "The UK is paying the price of its job miracle" - as if the UK is the same as the government.
But what, exactly, is the problem here? My charts, taken from the ONS, might shed some light.
This shows that there has been strong growth in jobs in two low-wage occupations: personal services and elementary occupations. But it is wrong to say that there's been a general shift to low-paid work. Sales workers and plan operatives have little jobs growth - and indeed contraction in the former case. And highly-paid management jobs have grown faster than other occupations.
My second chart shows nominal wage growth. This has been relatively strong - I stress relatively - for lower-wage occupations. However, managers have also seen relatively high wage growth.
All of this isn't obviously consistent with the idea that falling real wages mean that workers have priced themselves into jobs*. If this were the case, you'd expect to see the biggest rises in employment in those occupations where relative wages have fallen. Instead, some sectors which have seen big employment rises have also seen rises in relative wages: this has been true for personal services and management.
These data are more consistent with job polarization than with a general shift to less skilled work or workers pricing themselves into jobs. There's been increased demand for managers and personal services workers - hence the combination of rising jobs and relative wages - but declining demand for routine white collar administrative jobs. (Quite why demand for managers seems to have risen is another story.)
I don't say this to reject the idea that employment growth has been unfavourable for the Exchequer; I suspect the problem has been decorporatization - a shift to self-employment - more than shifts within the composition of employment.
* I'm not saying the hypothesis is wrong - just that it doesn't jump out of these data.