None So Blind, Macroeconomics Division

NYTimes.com – The Opinion Pages / The Conscience of a Liberal
January 25, 2014.   Paul Krugman

Noah Smith reminds us of a quarter-century-old diatribe by Robert Barroagainst the New Keynesian economics, which is notable for its bitterness. You can sort of see why: people like Barro had triumphantly declared Keynes dead a decade earlier, and were evidently horrified to see even a mild version of Keynesian ideas making a comeback.

But what’s also striking is Barro’s evident inability to understand why these ideas didn’t shrivel up and die the way they were supposed to. His only answer was politics – leftists looking for rationales for government intervention – which is kind of funny when you realize that in its early days New Keynesian economics included among its leading lights Greg Mankiw and John Taylor. Also, the kind of intervention under discussion – monetary policy – was hardly the stuff of Big Government.

So why did so many macroeconomists feel that they needed to resurrect something Keynesian in feel? Don’t tell anyone, but they were looking at this thing called evidence.

Just an aside – Noah seems to me to have a blind spot here, an urge toward nihilism on the question of evidence in macro. I don’t really see why. The evidence that, at the very least, we don’t live in a classical world is very strong, and in any normal science would long have been considered conclusive.

Let’s ask what the debate in the 80s and 90s was about. It wasn’t about fiscal policy, which only came back into central focus after we hit the zero lower bound. Instead, it was about monetary policy: whether actions by central banks could drive economic fluctuations, or whether these were all real shocks of some kind. Keynesian models (including the monetarist subclass, which isn’t really very distinct in content as opposed to attitude) argue that money has real effects because some wages and/or prices are sticky in nominal terms. (This is not the same thing as arguing that increased price flexibility would help end recessions – I’ve written about that before, but just leave that on the side right now.)

So, what evidence might you look for on the proposition that monetary policy can drive the real economy? You could look for direct evidence of the asserted real effect – preferably in the form of natural experiments, where there was a clear change in policy and you could track the outcome. Alternatively or additionally, you could look for evidence of nominal stickiness.

And by the mid-1980s there was already overwhelming evidence of both kinds. Romer and Romer hadn’t yet published their classic event-study demonstration that money matters, but as they said, their methodology was largely based on the work of a guy named Milton Friedman. And the mother of all Romer-Romer natural experiments took place when Paul Volcker first tightened policy to break the back of inflation, then loosened it when he thought we had suffered enough; the results – the worst recession since the 30s, followed by a roaring recovery – were pretty decisive. I remember, during our time at the CEA in 1982-3, someone making a new classical argument, and Larry Summers saying something like “aren’t 12 million unemployed enough reason to stop listening to this nonsense?”

Meanwhile, on nominal stickiness: there was the evidence from real exchange rate behavior, where nominal and real rates not only moved in tandem, but real-rate behavior changed totally when the exchange rate regime changed. And there was clear evidence from surveys both that there was a spike in the distribution of wage changes at zero and that employers believed that nominal wage cuts were very costly for morale.

All of this evidence has, of course, only grown stronger since.

Seriously: if this were a normal scholarly field, can you imagine a large part of the profession not only ignoring this evidence but doing all it could to excommunicate anyone trying to face reality? And no, I’m not engaged in hyperbole: remember, it was Ken Rogoff, not me, who wrote about bearing the scars of “new neoclassical repression.”

So my take on macro is that we have plenty of evidence about what kind of approach works – and that approach really does work, giving lots of useful guidance. In fact, the empirical evidence for basic macro propositions is better than that for most of micro! The problem is that so many macroeconomists refuse to see the obvious.

< http://krugman.blogs.nytimes.com/2014/01/25/none-so-blind-macroeconomics-division/ >

Leave a Reply

Your email address will not be published. Required fields are marked *