Non-uniform fiscal multipliers

Go and read the esteemed Mr. Wadsworth on the fact that the Government "fiscal multiplier" value actually depends on which bit of Government spending you talk about:

Some of these hundreds of things are of net benefit and some aren't, but as these things are difficult to measure precisely, personal opinion seems to take over and stifle any sort of sensible debate. But surely, it is idiotic to argue that the £ benefit for every £1 the government spends is the same, and that there is always a net benefit (just as idiotic as arguing there is always a net loss), regardless of what the government is spending it on?

Read the whole thing; I'll wait...

This point has irritated me no end whenever increased Government spending is mooted, and its proponents say "oh, it's OK, think of the fiscal multiplier!" The opponents say "yes, it's sub-1.0", and the argument rapidly descends into fighting and backbiting. Some Government spending is clearly 1.0+ (most of the law+order budget, as Mr. Wadsworth notes). Some is well under 0.5 (the Department of Education springs to mind). Other spending is almost impossible to measure (the overall cost of maintaining the Navy or procuring Trident, for instance). But you can't simply say "the Government spending multiplier is 1.2" and use that to argue for paying GPs 20% more or funding overseas aid to India.