James Forsyth notes a post by Hall and Woodward on the question of whether fiscal stimulus is more effective when enacted through tax cuts or through spending increases. Woodward and Hall suggest that tax cuts are between twice and three times as effective - i.e. for each extra £1 of government debt you take on, you get £1 in extra GDP from a tax cut, but only £1 to £1.4 if you raise spending.
Personally, I'm pretty dubious about the Woodward-Hall methodology here. I'm not really sure that the GDP effects of military spending during wars provides that good a model for fiscal stimulus in 2009. However, the conclusion that tax cuts represent the superior form of fiscal stimulus is certainly one that I share, and have argued for previously, both in the Sixteen Economists' letter and my RSA speech.
There is a very long-running dispute in empirical economics about how much effect there is on GDP of tax cuts or government spending (for nerds, this is called the debate about the "multiplier effect", because it's the question of what multiple of the number of pounds of spending, say, is the increase in GDP). Keynes argued that there might be a rather large multiplier (at least under the conditions on which he focused), whilst Friedman's studies suggested it might be only 1 (if it's only 1, then extra government spending isn't any better than the private spending that it replaces). But these things aren't easy to calculate, and the debate has raged for over half a century.
The point I would emphasize is the following. I do not believe that "government spending" can properly be regarded as a homogenous lump, as if all of it were equally productive. Some things the government spends money on are a good idea and contribute to the General Good; in other cases the government should not be spending money in that area at all and it would be better left to the private sector; in further cases there would be advantages to the right form of extra government activity, but in practice government intervention is sufficiently badly done that it would have been better to leave matters to the private sector; and, finally, there are cases in which government spending is useful, but could have been better. (Incidentally, this is why I disagree with David B. Smith, Allister Heath and others concerning whether there is an "optimal size of the State". In my view there is not: whatever the size of the state, the virtue of each additional government project depends, overwhelmingly, on its own individual merits.)
Thus, when it comes to how much government spending there should be, in my view much the dominant question concerns the proper balance between the State and the Market and how the State should go about intervening when it does. I believe that these balance issues are only very marginally affected by whether the economy is in recession or not. (There are a few technical caveats I'd offer to this, but that slightly crude statement will do for no.)
So, it seems to me that to argue for increased government spending, we need (whether we are in recession or boom) to argue that the activity the government is going to intervene in with its spending is not better left to the Market (and, also, whether the form of government intervention involved is the best one we can come up with). Perhaps, politically, it becomes easier to raise government spending in recession. So if (as I do not) one believes that government spending is generally too low as a proportion of the economy, then recession might be a good time to advance one's case for a larger state. But that is not about "fiscal stimulus" - it is about the role of the State.
In contrast, when I talk of "fiscal stimulus" what I have in mind is government borrowing to replace private borrowing that might not otherwise occur in certain sorts of recession (e.g. when the banks are all bust). Nothing to do with the role of the State. It seems clear to me that the only pure form such fiscal stimulus can take in an economy in which the State already has a large role is that of a tax cut.
In the case of the UK, the window for such "fiscal stimulus" will, I suspect, be rather narrow. In the none too distant future we are likely to move from the government increasing its borrowing to the government funding its debts by printing money. The year 2008/9 was the time for tax-cuts funded by debt (not, in my view, through VAT cuts, but through income tax rebates, but that's another story). The years 2009/10 on will be the years of budget deficits funded by money-printing, so as to prevent deflation from getting out of control. Let's hope it works.
