The big issues in macroeconomics: the fiscal multiplier
The biggest theoretical issue in macroeconomics is “what causes unemployment”. As discussed in the last post, the classical answer, that unemployment is caused by problems in labor markets, is obviously wrong as an explanation of the simultaneous emergence of sustained high unemployment in many different countries. Unemployment is a macroeconomic problem.
The central macroeconomic policy issue, then, is “what, if anything, can macroeconomic policy do to move the economy back to full employment”. If you accept that, under current conditions of zero interest rates, there’s not much positive that can be done with monetary policy, and you stay within the bounds of mainstream policy debate, this question can be restated as “how effective is expansionary fiscal policy” or, in Keynesian terms, “how large is the fiscal multiplier in a depression”.
To restate this in more neutral terms, if the government spends more, say by employing and equipping more firefighters, what happens in the rest of the economy? The answer given by classical economics is that the newly-employed firefighters must be drawn from elsewhere in the economy, presumably in the private sector. Similarly, the production of extra firetrucks means less vehicles can be produced for other purposes. Although the exact way in which resources are reallocated can’t easily be predicted, the classical model gives the clear answer that the overall level of employment and economic activity won’t change
Keynes gave a different answer. The classical solution, he said, was applicable in a situation of full employment, but that was a special case. In general, the economy might be in an equilibrium with high unemployment, with plenty of idle workers who could be hired as firefighters, and factories to produce their equipment. Not only that, but the firefighters would spend at least some of their increased incomes, leading to further growth in demand. So, aggregate employment and production would expand by more than the amount of the initial increase, leading to a ‘multiplier’ effect.
More precisely, the fiscal multiplier is the ratio of the final change in aggregate output (and therefore in employment) to the initial change in government expenditure. The traditional “Old Keynesian” view (implied by the multiplier terminology) is that, provided there are plenty of unemployed resources, the multiplier is greater than 1 (among other things the value depends on the kind of expenditure – low income households are more responsive to income changes, so expenditure that benefits them will have a higher multiplier). That view seems consistent with the evidence of, for example, Romer and Romer and, more recently, the IMF World Economic Outlook (PDF), which concluded
Research reported in previous issues of the WEO finds that fiscal multipliers have been close to 1 in a world in which many countries adjust together; the analysis here suggests that multipliers may recently have been larger than 1
The classical position may be restated as saying that the fiscal multiplier is zero. The core of New Classical economics is the reassertion of this claim. If workers are unemployed it was either because they are unwilling to work at the going wage or because some artificial barrier (unions or minimum wages) stops wages from adjusting to their equilibrium level. To the extent that Keynesian policies worked, New Classical economists like Lucas argued, it was by generating inflation and tricking workers into accepting wages that were higher in nominal terms but lower in real terms.
Before the current crisis, New Keynesians conceded a fair bit of ground to the classical view, but argued for a positive multiplier, though not necessarily greater than 1. One way of putting this is that public expenditure partially “crowds out” private spending. But most NK advocates thought fiscal policy unnecessary, since monetary policy had the same effects and was easier to manage.
I criticized the New Classical view last time, but the dominant idea in European and many US policy circles is even worse. It’s, the theory of expansionary austerity put forward by Alesina and various co-authors that the fiscal multiplier is substantial and negative. That is, cutting public expenditure will increase output.
This claim has no real theoretical basis and almost no empirical support, being based largely on anecdotal evidence and on a few studies that have not stood up to criticism. I took it apart in the paperback edition of Zombie Economics – the relevant section was published here
To sum up, despite the thousands of papers published every year in the field, macroeconomic theory is incapable of giving even a qualitative answer to the most basic questions about fiscal policy; at least, not one that would not elicit dissent from a substantial, and well-credentialled group of leading experts. Worse, public policy decisions to impose austerity policies are being made on the basis of a magical theory, with almost no empirical support.
This is an appalling situation, made worse by the complacency of (the dominant group of) academic macroeconomic specialists, who seem to think that everything in the garden is rosy, or would be if outsiders like Krugman would just shut up. It really is hard for me to see how the economics profession can recover from its current rotten state, at least as regards macro (and, even worse, finance).
fn1. Some advocates of nominal GDP targeting, such as Scott Sumner say that monetary policy could work if central banks showed sufficient resolve to convince people they would tolerate inflation. I disagree, but I’ll have to leave this for another time.
fn2. On this view, welfare would decline, because governments would use resources less efficiently.
fn3. While writing this, I wondered what would happen if you put this question to a group of DSGE theorists as a pop quiz. I suspect most would give some variant of “the question is ill-posed” and the rest would be all over the place. But, if any DSGE theorists are reading, I’d be keen to get their views.