Krugman on Schumpeter
Krugman wrote the other day that Schumpeter’s macroeconomics falls apart because:
[Schumpeter says] mass unemployment is necessary, because you have to shift resources away from sectors that got too big, stimulus is a bad thing because it slows the necessary adjustment. And now as then, the whole notion falls apart when you ask why, say, a housing boom — which requires shifting resources into housing — doesn’t produce the same kind of unemployment as a housing bust that shifts resources out of housing.
Thus urged, I did ask myself why there would be unemployment in a housing bust, but not a boom. The answer, Mr. Krugman, seems pretty obvious:
- Investment bubbles collapse much faster than they inflate. In the real world, labor can only redeploy so fast. If capital reallocation exceeds that rate, you get unemployment. So on the way up, capital redeploys slowly enough for labor to react smoothly. No unemployment. On the way down, capital redeploys much faster than labor can. Presto, unemployment.
- (more speculatively) In the short run, bubbles increase the blended rate of return on capital for the whole economy. Higher ROI permits overall employment to rise above the rate that would have prevailed without a bubble. When it bursts, ROI falls, so unemployment rises.
October 8th, 2009 at 10:24 am
Krugman was surprisingly weak on this one. Lots of good replies in comments. For instance, #3 (saw this firsthand in tech bubble: SF cabbies became highly-payed Java programmers, then delayed returning to cab-driving), #8, #25, #45, #48, #51, #58, #60.
Also, I suspect a serious bubble funds itself by borrowing against the future, not by sector capital reallocation. The housing bubble was essentially financed by the current bailout money. So those housing bubble jobs (+ HELOC spending) were financed by the present bailout money (also borrowed). Krugman wants to keep the borrowing/bailout merry-go-round running. Interesting to see how Latvia versus Sweden works out:
http://www.nakedcapitalism.com/2009/10/latvia-in-crisis-threatens-to-stiff-swedish-banks-with-mini-jubilee.html
Looked at this way, Krugman’s premise is wrong: a housing boom doesn’t require shifting capital resources into housing. Instead prices are rising, you HELOC borrow this magic capital and reinvest it in housing, so prices rise more. At the end of the boom the capital is magically gone. No shift of pre-existing capital, but lots of debt left behind. Probably similar in the tech bubble to buying stocks on margin using other rising stocks as collateral.
What am I missing?
October 8th, 2009 at 12:07 pm
I see my argument as more general than the idea of bubble/non-bubble.
I believe the Great Depression, and the current one, were both driven by an abrupt, unexpected explosion in labor productivity. In the 1920s, it was factory electrification and farm automation. In 1995-2005, it was labor market globalization. Both vastly increased potential output per US dollar almost instantly. This happened faster than people and capital could be rebalanced, so the result was bubbles (capital sloshing around, with nothing to do, will find a home somewhere, sensible or not) — followed by a crash and piles of people abruptly out of work.
Note how government policy doesn’t figure in — it’s bigger than that, a tectonic shift, driven by technological revolution.
But back to Krugman. When otherwise smart Keynesians blow a fuse, the common factor seems to be a failure to appreciate the difference between 2009 and 1931. It is unequivocally logical to follow Keynes when you are a net creditor with a trade surplus, high savings and low public debt. But it’s less clear what to do when everyone is already leveraged to the limit, especially as history suggests hyperinflation is more destabilizing than deflation.
This is yet another special case of the canonical reason for any individual or group to maintain low debt in “normal” times: greater flexibility in unexpected circumstances.