Bryan Caplan  

Reply to Arnold on PSST

Is There an Asymmetry in Benef... Evidence for PSST: A Rejoinde...
My last post critiqued Arnold's PSST ("patterns of sustainable specialization and trade") alternative to the conventional aggregate supply/aggregate demand model:
In Arnold's story, firms and workers violate the First Law of Wing Walking: Never let hold of what you've got until you've got hold of something else.  If you can find a new PSST, great.  But until you do, the smart move is to cut wages so you can stick with what you know.
Arnold replies here.  For the rest of the post, he's in quotes.  In response to my "cut wages" objection, Arnold says:
That is true for small changes. Mathematically, it assumes continuous functions.

It may not be true for large changes. Suppose that somebody brings a tractor to a farm, so that the farmer can do the work himself without the help of hired labor. The marginal product of the worker might now be below the subsistence wage.

Tyler Cowen uses the acronym ZMP to stand for Zero Marginal Product. The point is that new production methods can drive the marginal product of existing workers way down.

Yes, it's conceivable for new production methods to suddenly drive millions of workers' productivity down to zero.  But why on earth should we believe that this has occurred?  This is an extraordinary claim requiring extraordinary evidence.  If lots (say 25%) of old economy firms saw their revenues fall by 80% I might believe it.  If wages fell 50% and the unemployment rate didn't budge, I might believe it.  All I've seen this recession, however, is that nominal GDP sharply fell, and real GDP fell almost proportionally.  That's how recessions in low-inflation economies usually work.

Arnold continues:
Even with discontinuous changes, the phrase "drastically enough" can be invoked to suggest that wage cuts could cure unemployment. But suppose that "drastically enough" means 25 percent or more. If you want to say that the PSST story of unemployment depends on wage stickiness because such a large wage reduction could take care of things, then fine. You have scored a debating point without practical significance.
I heartily disagree.  Suppose low-skilled Americans used to produce $20,000 in value per year, and now produce 25% less - "just" $15,000 per year.  That's not "approximately zero" productivity.  $15,000 per year is enormously high by world and historic standards. 

No practical significance?  It's the difference between (a) a world where much of the healthy population faces a choice of welfare, charity, or starvation, and (b) a world where the unemployed can resume independent, productive lives once workers swallow their resentment and take a pay cut.  It's the difference between (a) a world where 99 weeks of unemployment insurance is a wise adjustment to economic turmoil, and (b) a world where 99 weeks of unemployment insurance is a major cause of the problem it purports to solve. 

Arnold's position genuinely puzzles me.  He's skeptical of stories based on nominal wage rigidity.  But nominal wage rigidity is both introspectively plausible and has decades of empirical research on its side.  What's his substitute?  Discontinuity and/or ZMP - two introspectively incredible stories that, with all due respect to Arnold, have almost no empirical research to back them up.

Question for Arnold: What's the best available "Guide to Discontinuity/ZMP for Skeptics"?  Non-economists have always been quick to believe stories about technological unemployment.  Economists have been ridiculing this popular fear for centuries.  What happened in the last three years that ought to make economists reconsider?

Comments and Sharing

TRACKBACKS (1 to date)
TrackBack URL:
The author at Eli Dourado in a related article titled Caplan on the ZMP Hypothesis writes:
    I’m puzzled by Bryan Caplan’s hostility (1, 2, 3) to the ZMP hypothesis. It is hard to think of another idea that is more Caplanian. This is after all the man who pointed out that “the lower deciles don’t contribute that much to... [Tracked on November 16, 2011 2:52 PM]
COMMENTS (14 to date)
Eli writes:

Nominal wage rigidity is introspectively plausible at most for recessions of 1-2 years (generously). We're pushing 4 years of high unemployment. The longer this goes on, the more silly nominal rigidity sounds to me.

Sonic Charmer writes:

I am clearly missing something. 'Economy struggles to find PSSTs' and 'nominal wages are sticky' do not seem like incompatible, competing explanations to me. Rather, the latter seems like a (not the only) source of friction helping to cause the former.

Mike writes:

PSST certainly fits better with my biases.

Andrew writes:
Yes, it's conceivable for new production methods to suddenly drive millions of workers' productivity down to zero. But why on earth should we believe that this has occurred?

Couldn't agree more. The claim that there has been some sudden pattern or technological change that has reduced worker's productivity to zero is an extreme claim. In fact, claiming that nominal rigidity is not a factor in their unemployment seems equivalent to the above claim, as their product would have to be zero for a wage that would employ not to exist.

Robin Hanson has been pointing out for some time that advances in tech and production methods in recent years has been incremental and expected (see his interview with Martin Ford) and can't account for the short term noise we see in unemployment and income. Again, that requires some nominal mechanism; would be interesting to see Robin's take here.

matt writes:

Imagine that I build houses. Housing collapses and the value of my labor just dropped significantly and suddenly. I start making solar panels. Some guys in China will make them much more cheaply. The value of my labor drops.

"$15,000 per year is enormously high by world and historic standards."
But many people here won't work hard for that much. They have other ways of getting by.

Alex Godofsky writes:

What exactly is PSST beyond "hey guys you know that economy thing? guess what, it's really complicated"? Are there any kind of predictions it can make, or policy implications it has?

andy writes:

I never understood why wage rigidity of the employed cannot be compensated by wage flexibility of the unemployed. Unless the unemplyed suffer from wage rigidity too.

david writes:

@andy: there is the matter of magnitudes (what level of wage reduction on the part of new employees would be necessary to justify employment given wage rigidity of the employed? Something near ~80%?), and a rational refusal on the part of employers to easily substitute in lower-wage labour and dismissing existing employees.

jb writes:

I misread your initial post, but my misunderstanding led to some thoughts:

Technological change doesn't change someone's productivity - Fred is still Fred, and he can accomplish the same amount he did yesterday. His productivity hasn't dropped from $20k to $15k.

What has happened is that, this morning, the manager turned on a new piece of software, who can do the exact same job Fred can do, but at a fully loaded cost of $15k/year.

And next year, the software will be even cheaper.

MG writes:

Before we use the current duration of US high unemployment as evidence that something different (must be tech) is at play; let's consider that similar (and longer) periods have been observed in many European Countries (and most of the causes were accepted to have been centered on rigidities.

Second, rigidities exacerbating unemployment need not only be about price (wages), but also about occupation choice. Flexibility in finding areas where one's output is valued higher should also be an option. Human needs are limited; human wants unlimited. Some of this may require new skills, but often all it requires is imagination -- and freedom to move and work where and in what you want.

Daublin writes:

A few comments, all of which I think have been posted here before:

1. Bryan, you seem to propose standard macroeconomics as a good explanation for recent events, but this emperor has no clothes. While it appears to hold during steady times where only minor changes are happening, large shocks around the world have more often led to the opposite of the predicted effect. Examples are the end of World War II, the stagflation of the 70s, and Japan's lost decade. Arguably the recent years in America are another example: like in Japan, we spent large amounts of money on public projects, but we didn't observe an obvious jolt to the economy.

2. There are suggestions that the changeover from most labor being in agriculture to most labor being in manufacturing was not smooth. While it didn't happen at any one point in time, it roughly overlapped with the Great Depression. Continuing that example, the friction of changing from an agriculture job to a factory job is not due to the wages being wrong. It's a large change in lifestyle, and it probably involves physically moving into the city.

3. In recent years, America went from a construction boom to a construction bust. That would handily explain why a large chunk of the labor force has suddenly found their skills unmarketable.

4. In recent years, unemployment has been concentrated in lower-skilled, lower-educated labor. People with higher levels of education are not finding it an especially bad job market. People in computers wouldn't even know there's a recession except by reading the news. It's the low-tech end of the job market that is being hammered.

Any or all of these might be wrong, but they seem at least plausible. It would make for an enlightening discussion if you could follow on on any of these arguments, rather than simply launching a broadside and pretending they haven't been posted.

Peter writes:

The United States has bottomed out its wages as it is. Low-skilled labor jobs that pay a living wage are almost non-existent. These people can only afford to eat because of heavy corn subsidies and require extensive government help elsewhere. Wage cuts are not the answer.

Steve Roth writes:

"What happened in the last three years that ought to make economists reconsider?"

I think there's a misconception here. The conditions could have been growing for a long time, but their effects were masked by AD-goosing government deficit spending and (much larger) private debt issuance.

The debt deflation of the last three years, and its effects on AD, finally enforced the effects of a much longer-term trend.

floccina writes:

Bryan, I agree mostly with you but it is interesting that food and petroleum prices are high but you do not see farmers looking for employees, petroleum is booming ND but it is not enough to push unemployment down in the rest of the country.

Comments for this entry have been closed
Return to top