Undercover Economist

Stimulus spending might not be as stimulating as we think

Few things annoy me more than rhetoric that implies government spending is funded by the generosity of ministers rather than by taxpayers. (Alistair Darling’s pre-Budget report speech included lines such as, “Mr Speaker, we chose not to let people sink when they lost their jobs but to intervene to help them stay afloat.” No, Mr Darling, you didn’t – the taxpayer did.)

Such quibbles aside, it seems only sensible that when unemployment rises and companies stumble, the taxpayer should take up the slack. And yet the economic case for government stimulus is far from clear cut. Stimulus spending can erode private spending. My wife, for example, is a portrait photographer. Recently she secured a contract from a local council that kept her busy for weeks. While she was working on it she kept her head down, actively avoiding work in the private sector. A company looking for a photographer would have had to go elsewhere, perhaps paying more for an inferior snapper, perhaps giving up on the whole business.

The pro-stimulus view is that the government hires otherwise-unemployed workers, who spend money, which is used to hire other otherwise-unemployed workers, who go on to spend more money, and so on. No wonder such government spending is said to have a “multiplier”. But the example of my wife suggests that the multiplier could also be zero. Rather than reducing unemployment, the government may be shifting workers from the private to the public sector.

There is nothing absurd about assuming a multiplier of zero. It is implicit in the traditional cost-benefit analysis of government projects, photographic or otherwise, which simply asks whether the projects should go ahead on their own merits, rather than speculating on all the jobs that might be multiplied into existence. If the multiplier is zero and you want to spend a billion dollars on bridges, then make sure you think the bridges are worth a billion dollars.

If government spending snarls up the economy, the long-run multiplier might well be negative (look up “Soviet Union” in any encyclopedia). But the assumption has tended to be that it is positive, at least in times of recession. In his General Theory, Keynes outlines an example with a multiplier of 10. President Obama’s Council of Economic Advisers puts forward a multiplier of 1.6, which seems modest in comparison. But even a multiplier of 1.6 would be impressive. It means that if the government spends a billion dollars building a few bridges, the knock-on effects will be to increase the size of the private sector by $600m. We get the bridges, and we get more of everything else, too. With a multiplier of 1.6, paying people to bury money and dig it up again is sound policy. Even a multiplier of 0.5 would mean we could get a billion dollars of bridges while losing only $500m of private sector activity – a pretty good deal.

This analysis helps to clarify what we’re talking about. But it does not tell us what the multiplier is. I have seen estimates based on careful work by respected economists that range from zero to 1.5 – perhaps higher still if we think the world economy was on the brink of depression in 2009.

It should be no surprise that there is disagreement, sometimes ill-tempered disagreement. Government spending varies because the economy is in flux; it is almost impossible to say what causes what, and the tantrums tend to be about whose methodology is the least absurd.

My own conclusion: government projects probably do enjoy a multiplier-related discount in straitened times. But it is still worth asking whether the projects themselves are worth doing.

Also published at ft.com.