Macroeconomic Effects from Government Purchases and Taxes

"One of the things I’ve been trying to do in my research is to calculate the effect, particularly on Gross Domestic Product, of government expenditure programs. And I’ve been focusing on the US experience, because that’s where I have the information, although it would be good to go beyond that. But the thing you can clearly isolate is the effect of wartime expenditure – particularly World War Two – it is so big that in a statistical sense it gives you a lot of power to figure out what is going on. There’s both the build-up, starting in 1941, and then there’s expenditures coming down after the war, in 1945-6. There’s a lot of evidence there. Sometimes the spending in a year is 20 per cent of GDP, which is absolutely astounding. In comparison, the New Deal programs, particularly in 1934 and 1936, are only two to three per cent of GDP of extra spending. In terms of the stuff that’s not wartime spending – which we’re probably most interested in in the current climate – it’s just hard to know from the history of the data and the time series. The New Deal is part of my research, and it’s bigger than the other non-defence expenditure in terms of stimulus, but it’s not enough to really sort it out. So I don’t think you can reliably say what the effect is. But conceptually you’d expect the wartime spending to have a bigger effect for various reasons on the GDP than the equivalent amount of expenditure in a non-war situation. And the wartime effect you can estimate pretty precisely, and the multiplier is clearly less than one, even in World War Two – it’s in the order of 0.6, 0.7, something like that. Yes, those articles refer to this kind of evidence, and I’ve been working more on it, trying to make it more precise. Some economists have argued that in a time of slack the multiple should be bigger, because there’s more capacity to respond to the extra demand. There’s a little bit of evidence that that’s right. A lot of that comes from the build-up in World War Two, because in 1941 the unemployment rate is still around nine per cent, so you can see what is the effect in that environment, in a high unemployment situation, of having a big expenditure increase. (Later in the war, the unemployment rate is close to nothing, so you don’t have that setting.) There’s a little bit of evidence from that that the multiplier is bigger when there’s more slack. But it doesn’t look like the multiplier gets up to one, even when the unemployment rate is nine per cent. It’s getting closer to that, but even then it is not one. Oh they pulled that out of the air. I have the advantage of having at least a little bit of empirical evidence: as I said, it’s based particularly on military purchases. So even though that evidence is not that great, it’s infinitely better than the alternatives, which are no evidence."