More on multipliers: why does it matter?
The IMF's reassessment of the "fiscal multiplier" has sparked off multiple reactions in the economics blogosphere both in the US and UK. My initial reaction is here. Meanwhile, Chris Giles at the FT has weighed in, attempting to demonstrate that the IMF's analysis is not robust. I'd like to step back a bit now from the IMF piece (I'll return to it later) and explain why this matters.
As I discuss here, in mid-2010 the international economic policymaking community, led by the IMF, and very much influenced by the new Coalition governnment in the UK, executed what became known as the "pivot" to fiscal consolidation. Pretty much everyone agreed that it was necessary to reduce budget deficits; the question was how quickly, and what the damage, if any, to growth would be. As a reminder for those new to this debate, the "multiplier" measures this: it is the reduction in output resulting from a given reduction in the budget deficit (so if the multiplier is 1, then a reduction in the budget deficit of 1% of GDP reduces output by 1%). On this question, broadly, there were three camps.
First, a small group of economists argued both on theoretical and empirical grounds that fiscal consolidation wouldn't reduce growth at all - indeed it might even enhance growth (so the multiplier would be zero or positive). The doctrine of "expansionary fiscal contraction" argued that tightening fiscal policy could, through exchange rate and confidence effects, actually increase demand and growth; a paper by Alesina and Ardagna was particularly influential in this respect. While this was always a minority view among empirical macroeconomists, this research was quickly picked up on by those politicians who wanted aggressive deficit cuts, in both the UK and EU. For example, Matthew Hancock MP, formerly George Osborne's Chief of Staff (and now Minister for Skills), claimed:
"I discovered that research into dozens of past fiscal tightenings shows that, more often than not, growth doesn't fall but accelerates."
Somewhat more tentatively, the UK Treasury argued (although I doubt any Treasury official believed this for a moment) in the 2010 Emergency Budget that
" These [the wider effects of fiscal consolidation] will tend to boost demand growth, could improve the underlying performance of the economy and could even be sufficiently strong to outweigh the negative effects".
So while this view was never very credible economically, it certainly influenced policy.
The second view was that taken by mainstream economic modellers and forecasters, including most importantly the IMF, but also the UK Office for Budget Responsibility, the Bank of England and indeed us here at NIESR. This was that the negative impact of fiscal consolidation on growth would be significant, but not disastrous. The IMF never believed the Alesina and Ardegna results; in October 2010 the Fund concluded that:
"Fiscal consolidation typically lowers growth in the short term. Using a new data set, we find that after two years, a budget deficit cut of 1 percent of GDP tends to lower output by about ½ percent and raise the unemployment rate by ⅓ percentage point."
These estimates were based on historical experience over the last three decades; using similar data, NIESR's model incorporate similar estimates. And when estimating the impact of the UK fiscal consolidation programme announced in June 2010, the OBR also used very similar estimates. This is hardly surprising: as Duncan Weldon points out in a neat bit of detective work, the OBR's multiplier estimates are based primarily on one IMF paper, as well as two papers from NIESR researchers.
There was, however, a third view. This was advanced most strongly by Paul Krugman and Brad Delong in the US, and here by Martin Wolf (in the columns of the FT) and Simon Wren-Lewis; it was that the experience of the last three decades (except, perhaps, in Japan) was not relevant to that of a world where monetary policy was limited by the zero lower bound on interest rates (or, for those like Scott Sumner who think that monetary policy could have been even more aggressive, by political or institutional constraints). In such a world, multipliers would be significantly higher, and almost certainly greater than one. Simon explains why here, concluding perceptively that this may be "an occasion where thinking about macroeconomic theory can be rather more useful than naively following the evidence of the past." Meanwhile, Antonia Fatas and Ilian Mihov argued on empirical grounds that the Fund and others were consistently underestimating the size of the multiplier, as they explain here.
So what then is the significance of the IMF analysis published this week? For reference, I will repeat the key paragraph:
"In line with these assumptions, earlier analysis by the IMF staff suggests that, on average, fiscal multipliers were near 0.5 in advanced economies during the three decades leading up to 2009. If the multipliers underlying the growth forecasts were about 0.5, as this informal evidence suggests, our results indicate that multipliers have actually been in the 0.9 to 1.7 range since the Great Recession. This finding is consistent with research suggesting that in today’s environment of substantial economic slack, monetary policy constrained by the zero lower bound, and synchronized fiscal adjustment across numerous economies, multipliers may be well above 1"
So, in contrast to the Fund's 2010 view, multipliers are much larger than 0.5 - large enough to have a very substantial, and negative, impact on growth.
Now, the IMF analysis, in isolation, is clearly not definitive "proof" that multipliers are now 0.9 to 1.7 - and even if it was, that would not "prove" anything about multipliers in a specific country. I won't attempt to arbitrate between the Fund and Chris Giles on econometrics, except to say that his detailed analysis confirms my view, which he also reports, that cross-country regressions are typically not very robust, and in general can be used to make pretty much any argument you like (indeed, this is precisely the same reason I never believed the Alesina and Ardegna result either). So while I think the new Fund analysis does broadly support the view that in general terms one of the reasons the Fund's forecasts (in common with pretty much everyone else's) have been too optimistic is that they underestimated the negative impact of fiscal consolidation, I wouldn't place much weight on them in isolation.
But what is clear - particularly in the last sentence I quote above- is that the Fund has now accepted that the balance of the argument, both theoretical and empirical, has tilted decisively in favour of the third group of economists above. It's not just about one set of regressions; these are simply a further piece of supportive and confirmatory evidence supporting those of us who argued that aggressive fiscal consolidation was an unnecessary and dangerous gamble, with very serious downsides. The Fund is now squarely in this camp. This is a major intellectual shift - as Isabella Kaminska writes, no wonder Paul Krugman is feeling "smuggish". But leaving aside the economists' debate, how should this affect policy? In the UK, I can think of two key implications:
- the first relates to the current debate about how large the UK "output gap" is and hence how much scope there is for expansionary policy (both fiscal and monetary). The UK economy has essentially seen zero growth for the past two years. Some analysts - Chris Giles being the most credible, but the OBR has also taken this line - have argued that given the sort of multipliers assumed by the OBR and IMF, fiscal consolidation can't explain much of this growth shortfall, so it must be something else: supply side weakness, commodity prices, and so on, meaning that changing fiscal policy might not do much good. If, however, multipliers were in fact much higher, then fiscal consolidation is indeed the main reason for weak growth; and correspondingly, the scope for boosting growth through expansionary policy is much greater;
- the second relates very specifically to the OBR. As Duncan pointed out, the OBR's excessively optimistic forecasts were explicitly based on multipliers derived from IMF research. The IMF has now explicitly changed its mind; the OBR's position is no longer tenable. If it wants to retain its credibility as an economic forecaster independent of government, it needs to examine its assumptions and methodology, both retrospectively and prospectively, on the impact of fiscal consolidation on growth. The December OBR forecast should include at a minimum both a reassessment of its forecast record, in the light of the Fund's change of view, and an assessment going forward of the impact of different multiplier assumptions on growth.
Arguably, however, far more important than the UK debate- and far more central to the concerns of the IMF - are the implication for the eurozone, and in particular for the current adjustment programmes in Greece, Spain, Italy, Ireland and Portugal. Several months ago, I argued:
Clearly long-run solvency is also essential. But, in Spain and Italy, trying to hit arbitrary short-run deficit targets, as proposed by the European Commission, is likely if anything to be counterproductive to the objective of long-run sustainability. Spain’s long-term fiscal position, for example, is relatively strong; what it needs to ensure that remains the case is decent levels of economic growth, and what it needs for that is structural reform, especially labour market reform. Both politically and economically, such reforms will be both less painful and more effective if fiscal consolidation is much slower, as I argue here. These arguments on timing hold good even if multipliers and hysteresis effects are relatively small; if such effects are large – and there is every reason to believe that in European labour markets hysteresis effects are of profound macroeconomic importance – then they are even more compelling,
The IMF clearly now agrees with this, as Christine Lagarde has made clear in the case of Greece. They need now to point out to the European Commission and the German government as forcefully as possible that if they do not belatedly come to their senses, they will run the economies of Southern Europe - and possibly the euro itself - into the ground on the basis of an economic analysis that has now been discredited both theoretically and empirically.
Finally, what about us here at NIESR? Well, we did produce this, examining why the multiplier might be larger in current circumstances, and examining the implications; precisely what the OBR should have done. But, more broadly, when presenting NIESR forecasts in 2011, I was frequently asked why we were rather pessimistic relative to most other forecasters, and certainly the OBR. My response was often that what I worried about most was not that our model's predictions looked rather gloomy; it was that the economists I took most seriously - those listed above, who don't use quantitative models - thought our model was far too optimistic. And so it proved.