The Fiscal Multiplier; A Real Shape Shifter

by John Mauldin, Mauldin Economics

The chief economist for the International Monetary Fund, Olivier Blanchard, and his associate Daniel Leigh gave us an eye-opening three-page paper, buried in a 250-page World Economic Outlook release last week. They studied an economic concept called the fiscal multiplier, which is usually defined as the change in real GDP that is produced by a shift in fiscal policy equal to 1% of GDP. In simple terms, if the fiscal multiplier is assumed to be 1.0 then a change in government spending by 1% (either an increase or decrease) would produce a corresponding change of 1% of GDP.

Most institutional economists prior to this paper assumed the fiscal multiplier to be about 0.5. Again in simple terms, this would mean that government spending cuts equal to 1% of GDP would reduce actual GDP in the coming year by about 0.5%. The fall in GDP would of course reduce tax revenues, which means that you would have less than a 1% actual cut in the deficit. If the tax rate is 30% in this example, the deficit will be reduced by only 0.85%. That may be an acceptable outcome when an economy is growing nicely or the deficit and total debt are too high and the bond market is forcing the government to cut back.

While Blanchard and Leigh agree that in the past the fiscal multiplier was generally about 0.5%, they suggest that in the recent fiscal crisis the fiscal multiplier has been much higher. Their study suggests that it has been at least 0.9% and perhaps as much as 1.7%. This certainly seems to be the case in Greece and Spain, as their austerity measures appear to be working in reverse.

Gavyn Davies of the Financial Times views the IMF research finding from the perspective of England, where the government is taking 40%. The results are less than pleasing:

“If, however, the multiplier is 1.7, then the same initial public spending cut of 1 per cent of GDP would reduce real output by 1.7 per cent. The second round effects of this reduction in output would reduce tax or raise transfers by 0.68 per cent. The net overall improvement in the budget deficit would therefore be only 0.32 per cent. The economy would be in recession, and the budget deficit would hardly improve at all. Even if this were acceptable to governments, it would not be acceptable for very long to their electorates.

“This pessimistic arithmetic is not that far away from describing what has actually happened in some countries, like the UK, in the past two years. Furthermore, if we take this arithmetic as a given, there is more bad news to come. The major four advanced economies are now all planning to tighten fiscal policy in the years ahead by an average of 1 per cent of GDP per annum.

“…With a fiscal multiplier anywhere near the upper end of the Blanchard/Leigh suggested range, the effects of these policy changes would eliminate any chance of a rebound to normal growth rates in the advanced economies for some time to come. Interestingly, the planned fiscal tightening in the troubled economies of the eurozone is no longer any greater than it is for the major economies, because of the recent relaxation of some budget targets. Even so, it is hard to see how these plans could be sustained if the fiscal multiplier is at the upper end of the possible range.”

But there is a problem with this analysis, as Davies and others point out. It assumes that the results in one country will pretty closely match those in any other country. However, if you take out Greece and Germany, as an example, you pretty much remove the increase in the fiscal multiplier. Not that that would stop Paul Krugman and his fellow Keynesians from trumpeting this analysis as a reason to eschew all forms of austerity. In any case, the research does call attention to the dangers of creating economic models that are used to guide public policy. Davies continues:

“…The decline [in the IMF model of the fiscal multiplier] occurred mainly because economists became much more aware of the need to make assumptions about monetary policy when making the estimates. If the central bank is assumed to hold monetary growth or inflation at a given target rate when fiscal policy is tightened, then interest rates will decline and this will offset some of the negative effects of the fiscal change on output. The multiplier will be lower.

“The opposite is also true. Now that interest rates are stuck at the zero lower bound, central banks cannot reduce policy rates when fiscal policy is tightened, and the multiplier is correspondingly increased.” (emphasis mine)

I have mentioned several times a paper by the powerhouse economic couple Christina and Paul Romer (she was the chairperson of Obama’s Council of Economic Advisers), which showed that the multiplier for tax cuts or tax increases is around 3 in the US. Accordingly, a tax increase or cut equivalent to 1% of GDP should affect the US economy by 3%.

Davies notes that Larry Summers and Brad DeLong have argued that the multiplier should be assumed to be a minimum of 1.0 under present circumstances. [Their work] explains very clearly why the multiplier should be much higher than normal when the economy is stuck in a recession with interest rates at the zero lower bound. Another noteworthy paper, by Auerbach and Gorodnichenko, says that “the multiplier during recessions might be around 1.5-2, while in expansions it drops to zero.” Additionally, Robert Barro of Harvard has suggested that the tax multiplier is 1.

In another article, Professor Carlos Vegh of the University of Maryland said lots of evidence suggested that multipliers would differ greatly from country to country; and

“the whole exercise of trying to forecast growth for many different countries using essentially a single multiplier, whatever the value may be, is, in and of itself, an exercise in futility”. (FT)

You can pretty much pick a fiscal multiplier that works for your desired outcomes and find an academic study that will support it. And that is the point. Economists want to create models. It is in their DNA (perhaps defectively, I admit). And sometimes they have to make assumptions in order to make the models look like something that might be useful. The problem is that politicians, in particular, don’t look at the underlying assumptions but use the parts of the studies that most closely reflect their particular biases. The IMF report above uses data from many countries that are, indeed – as our quotes at the top suggest – unlike each other. Comparing Greece to Germany and then using that data to suggest policies for Spain and Ireland is a dubious practice. There are just not enough data points for such conclusions to be statistically valid – but that won’t stop the politicians from using the IMF study if it supports the policy outcome they prefer. The IMF, for what might be very good reasons explained way down in the footnotes, excluded countries such as New Zealand and the Baltics that had better outcomes from austerity policies. Their inclusion would alter the study.

Editor’s note: John Mauldin is co-author of the insightful book about the global debt crisis “End Game” which was reviewed by Econintersect. From that review:

This book is highly recommended as a plain language dissertation on the nature of the global debt bubble and some of the resolutions that may obtain. It does not give a prescription for how the crisis will be worked out; but it does outline some of the paths that may be followed.   The authors go out of their way to be non-judgmental and discuss the need for political decisions to be made, not just economic choices.

This is a book that can be understood by those who are quite unschooled in the world of political economy and how the financial world works. At the same time it is a worthwhile reading for those who are much more knowledgeable in the arcane intricacies of economic theory and national/global government politics. There are some technical details that are not quite correctly presented in this reviewer’s opinion and others that are very correct and at the same time presented succinctly, a difficult task for the subject.

And, as I stated at the beginning, if there ever was a time to read Endgame, it is now.

Go to the Amazon page to read further reviews and order the book by clicking on the book image.