Our work published last week suggests that public sector employment is good for local services, bad for local manufacturing. In the short run, these effects balance out so that total private sector employment is left unchanged. In the longer run, the effect on local manufacturing appears to be stronger, the effect on local services weaker and total private sector employment decreases offsetting any additional employment from the public sector.
What does this tell us about the impact of public sector pay? Not much, I'm afraid, contrary to the case made by some commentators. Public sector pay differentials are ONE of the channels that might explain our results (particularly the negative effect on manufacturing). But public sector employment could have the same sort of effects even in the absence of any public sector pay differential (because of general pressures on wages and house prices that come from higher labour demand in the area).
My best guess, based on our results, is that removing public sector pay would be bad for local services (who benefit from the extra demand that comes with higher public sector pay) but would remove some of the distortionary effect on manufacturing. I suspect you could figure out how large the first effect might be (looking at how much of their income people spend locally and what sort of changes to pay are planned - although getting to employment trickier). No one knows how large the second effect would be nor whether the benefits to manufacturing would offset the loses to local manufacturing.
In short, no one knows whether moving to more localised public sector pay would be a significant step in rebalancing the economy. Our results suggest that it might help with sectorial rebalancing (towards tradables away from non-tradables). We have no idea what it would do to with geographical rebalancing.
But, I have to say, that all this argument around rebalancing rather misses the point. The fundamental reason to care about local pay is because of the implications for the quality of public good provision in high cost areas. And on that front, what evidence we have, strongly favours more localised pay.