In the 2009 Pre-Budget Report the Labour government proposed a rise in employers' National Insurance Contributions. David Cameron said the measure was a "jobs killer" and George Osborne attacked it as a "tax rise on jobs" that "will stamp out the green shoots and kill the recovery."

Yesterday, Mr Osborne announced his own tax rise on jobs by introducing an Apprenticeship Levy.

It is this, and not just the money the OBR has magically discovered behind the sofa, which is paying for his tax credits U-turn. It will raise the Treasury £2.7bn in 2017-18 - very close to the £2.9bn that the restoration of tax credits will cost it*.

This raises the obvious question: if higher NICs were "a jobs killer", why isn't the Apprenticeship Levy? It's not because of any difference in scale: Labour's NIC increase would have cost employers around £2.4bn a year**.

Instead, it's because higher payroll taxes, such as employer NICs or the Apprenticeship Levy, don't much raise the cost of employment simply because they lead to lower wages instead. The incidence of payroll taxes falls upon workers. Back in 2010 I pointed to experimental evidence from Finland for this, and said: "higher employer NICs are highly likely to be passed onto workers in the form of lower wages than they would otherwise get." The OBR thinks the Apprenticeship Levy will have exactly the same effect:

The additional costs created for firms and workers by the Government's introduction of the apprenticeship levy and ongoing auto-enrolment into workplace pensions - both of which are economically equivalent to payroll taxes - will largely be borne through lower wages (Par 3.70 of this pdf)

Lower wages for whom, though? It can't be the very worst-paid, as these will benefit from the rising living wage during this time. Instead, it is likely to be workers just above the living wage who will suffer: they lack the bargaining power that comes with high skill (or rent-seeking talent!) whilst also lacking protection from the living wage.

In this sense, Mr Osborne's measures will compress pay differences between the worst-paid and those just above them. This might actually exacerbate trends which are already under way because of job polarization.

This might not be a good thing. For one thing, it might increase resentment as those who consider themselves hardworking and mildly skilled suffer a loss relative to those "below" them: we mustn't under-estimate the desire of people to "preserve differentials" as they used to say in the 1970s. And for another, less wage compression at the lower end of the labour market reduces the potential for the less skilled to work their way clear of poverty.

I suppose one might argue that Mr Osborne's measures would tend to increase one form of equality. But I doubt egalitarians will congratulate him for this. And nor should they.

* Lines 1 and 23 of table 3.1 here.

** Table 1.2 of this pdf.