It’s not structural, and it wouldn’t matter if it was (pt. 2 of 2)
It didn’t take Andy Harless long to figure out what part 2 of this essay would look like. Here’s his comment to part 1:
At the risk of giving away the punch line: structural unemployment is pretty much the same deal as an oil shock; it’s a reduction in aggregate supply. In both cases, employment eventually readjusts: in the structural case, because workers get retrained, relocated, &c; in the oil shock case, because product prices and productivity rise faster than wages so as to re-establish profit margins. In both cases, the adjustment happens faster if monetary policy accommodates: because there is more incentive to retrain and relocate workers; or because product prices rise more quickly.
The bottom line is that the Fed had been delivering 5% NGDP growth for decades, and no matter what caused the current crisis, they needed to continue delivering 5% NGDP growth. This means that money has been far too tight since August 2008, even if most of the unemployment is structural. I am reacting to statements like this from Narayana Kocherlakota:
Monetary stimulus has provided conditions so that manufacturing plants want to hire new workers. But the Fed does not have a means to transform construction workers into manufacturing workers.
There are so many things wrong with this that I hardly know where to begin:
1. The structural theory is usually based on the fact that the housing bubble caused residential housing to become severely overbuilt by 2006, necessitating a sharp decline in construction jobs regardless of what the Fed did. That’s true, but as this graph shows, almost all the decline in housing occurred BEFORE the severe phase of the recession started in August 2008. (Almost halfway through the blue vertical band that starts in December 2007.) So there was a structural loss of jobs after mid-2006, but it has little to do with the sharp rise in unemployment that only began two years ago.
2. The sharp loss of jobs that did begin in August 2008 was associated with three areas mostly unrelated to sub-prime housing; manufacturing, commercial construction, and services. All three of these turned down sharply precisely when NGDP started falling, i.e. when money got ultra-tight.
3. The problem is not struggling with the re-allocation of construction workers into manufacturing, as manufacturing has also shed lots of jobs.
4. It’s really not that hard to transform construction workers into factory workers. For God’s sake in WWII we put housewives into factories! And the average construction worker is far more skilled with heavy machines than the average housewife.
5. The Fed has not provided the monetary stimulus required so that factories want to hire workers. Volcker provided the monetary stimulus for factories to want to hire workers when he engineered 11% NGDP growth at an annual rate for the first 6 quarters of recovery in 1983-84. We’ve been running 4% NGDP growth in the first 4 quarters, and we are now downshifting to 3%. How can you get the 7.7% RGDP growth of the earlier recovery if NGDP is growing 3%? Are we going to have a minus 4.7% GDP deflator? When has an economy ever boomed with 5% deflation? People who ask why the economy should not yet have adjusted to 1% inflation are asking completely the wrong question. Even if we had adjusted, it would take 8.7% NGDP growth to get a Volcker-esque recovery. It’s simple math. What you are really asking is why isn’t inflation falling even further. That’s a tougher question. The 40% jump in minimum wages, and the 99 week unemployment extension probably made labor markets a bit less flexible that usual. But overall what we are seeing is not that far out of the normal. Both real GDP and inflation fell, with RGDP falling more sharply. That’s pretty normal for steep recessions.
Now let’s return to Andy’s point. Almost everyone agrees that the SRAS is upward sloping, even my critics. After all, my critics claim the Fed blew up the housing bubble with easy money in 2003, and that can only occur if money is non-neutral. This means that even if 100% of the current unemployment problem is structural, it is still true that monetary stimulus will boost employment. And since inflation is still below target, and expected to remain below target, monetary stimulus would also improve the inflation situation. It’s a win-win. So when people say that structural problems argue against monetary stimulus, they aren’t just wrong, they are doubly wrong.
Part 2: Stop searching for the Holy Grail of macro
Since David Hume, every bright young economist seems to want to take a stab at the problem of why nominal shocks have real effects (i.e. why the SRAS slopes upward.) They’ve all failed. It’s not that they haven’t come up with explanations; they’ve come up with plenty. Bennett McCallum once listed ten versions of price stickiness. Then there is wage stickiness. And misperceptions. And money illusion. And that’s ignoring how the supply-side intersects with nominal shocks, as when governments extend UI to 99 weeks during recessions.
They sift through all sorts of micro-level data, develop macro stylized facts, and then try to connect them up with theory. But they never get anywhere. Maybe all theories are true to some extent, and the relative importance of each effect varies from one business cycle to another. Milton Friedman once said that in 200 years we’ve only gone one derivative beyond Hume. (We look at changes in inflation, rather than changes in the price level.)
So I get pretty discouraged when I read economists trying to sift through micro-level data about prices and labor markets, searching for the Holy Grail of the micro-foundations of recessions. Hume explained our recession 200 years ago:
“If the coin be locked up in chests, it is the same thing with regard to prices, as if it were annihilated.” David Hume “” Of Money
That’s right; Hume knew that if the Fed paid interest on bank reserves, encouraging them to lock up excess reserves, it was like a contractionary monetary policy. Fed presidents would be better off spending more time reading Hume, and less time sifting through micro data that supposedly disproves “the” Keynesian model (as if there’s only one!) The brightest minds of the profession have attacked the problem for 200 years, and they’ve all failed. It’s a black box. It is truly the Holy Grail of macroeconomics.
Please stop searching for structural patterns, and start boosting NGDP growth.
(PS: I do agree that Obama’s economic policies are considerably less pro-growth than Reagan’s, but we could still be doing much better than we are. “There’s a great deal of ruin in a nation.” And again, those policies do not excuse the Fed allowing NGDP to suddenly fall 8% below trend.)