How much structural unemployment was there during the Great Depression?

by on May 15, 2012 at 12:58 am in Books, Economics, History | Permalink

A few times recently Paul Krugman has raised the issue of structural unemployment in the Great Depression, so I thought I would offer a look at what has been written on the topic.  Here is Richard J. Jensen, from a survey article:

Economists agree that Keynesian stimuli would not have helped structural or hard-core unemployment, only cyclical unemployment. As Table 1 suggests, about half of the unemployment was cyclical from 1931 through 1933; it was then that stimulus was needed and might have worked. By 1933, the appearance of a large, new, structural/hard-core element raised the natural level of unemployment from the 5 to 6 percent range to 12 to 15 percent. If a Keynesian stimulus had been tried and it had eliminated cyclical unemployment, the remaining unemployment still would have been io to 15 percent. Further fiscal or monetary stimuli would have resulted in inflation.

Later he moves directly to the key question:

…we need to discover how the war cured hard-core unemployment permanently. On the supply side, the growth of high schools and colleges, the postwar draft, and Social Security retirements removed young and old from the labor force. Wartime training and experience, in industry and in the military, made workers more productive, and upgraded skills so that the supply of unskilled labor was much smaller. In terms of efficiency wages, employers reshaped jobs to suit the skills and increase the productivity of available workers. They had to use men (and women) whom they would not have dreamed of hiring a few years before.

Personnel management became even more important. The number of industrial-relations staff rose from 2.5 per 1ooo employees in 1937 to 8.o in 1948. They were charged with improving productivity despite the extraordinary shortage of manpower, the high quit rates, the government-imposed wage freeze, and the new strength of labor unions. They dropped categorical restrictions against the poorly educated, the unemployed, women, the old, the handicapped, and sometimes, in spite of intense resistance, blacks. Recruitment of new workers became an art form, with sound trucks blaring in the streets beseeching people to come to work and earn big money. Jobs were restructured so that fewer skills were needed. Intensive in-shop and in-school training programs reached millions. Anyone with a modicum of skill was rapidly promoted, even to the status of foreman or instructor. The results further justified the use of efficiency-wage procedures, but this time efforts were made to find the right niches for workers who had been “hopelessly unemployable” in the 1930s.

In other words, the path out of high unemployment involved much more than a mere reflation of nominal values.  (By the way, when it comes to terminology I might not use the phrase “structural unemployment,” but it also is not “simple cyclical unemployment.”  I would say that in some circumstances the traditional distinction between cyclical and structural unemployment breaks down, but note that in terms of its parent literature this piece is using the terms properly, even if they sound somewhat off in a 2012 blogosphere context.)

In any case, history suggests that stimulus policy has to take some very specific forms to reach those “called cyclically unemployed by some, structurally unemployed by others” unemployed workers and that is the practical upshot.

Another practical upshot is that you still can believe in labor market hysteresis, as presented by DeLong and Summers.  Without some analysis like the above, the DeLong/Summers claims are otherwise contradicted by American post-Depression productivity once joblessness lifted.  Where were the long-term scars?  Well, they were fixed but it wasn’t easy.  So the relevance of hysteresis can be saved, but we still are left with proper stimulus being very difficult to do, unemployment being quite sticky, and proper policy requiring lots of structural attention.  The Great Depression is evidence for all of those views, not against them.

Here is one more bit, with a sad sting at the end:

The war, by removing millions of prime men from the labor market, by restructuring the work process, by subsidizing wages, and by massive retraining, finally gave the private sector the methods and the incentives to rehire the hard-core. Never since has hardcore unemployment affected more than one worker in a hundred.

Michael A. Bernstein’s book, The Great Depression: Delayed Recovery and Economic Change in America, 1929-1939, also considers the significant role of structural unemployment (don’t forget my taxonomic caveat) in the Great Depression.

It is important to learn from this literature rather than dismiss it.

Andre May 15, 2012 at 1:14 am

Did they have any opinion on the idea that long term unemployment and under employment in the earlier period led to the increase in structural unemployment in the later period?

v May 15, 2012 at 3:06 am

I really enjoy the Cowen-Krugman exchanges but does anyone (other than Tyler) think reasoning with Krugman will have any effect?

Krugman has openly stated he views himself as a partisan advocating an ideology (albeit one that will result in the greater good in his eyes). He has repeatedly stated that this is not an intellectual debate to him but an attempt to convince the masses of his preferred policy prescriptions. Given this context, he is basically a DNC-inspired hack with more impressive credentials (e.g., the Nobel prize) than his competitors like the Daily Kos, etc.

On the other hand, Tyler is treating this as a real intellectual debate where understanding the other side’s points and analyzing/responding to them is important. This is a mistake. Krugman will not act rationally

Asleep at the Keyboard May 15, 2012 at 4:23 am

But isn’t there another debate implied by this? A disagreement about how discussions themselves should occur.

JWatts May 15, 2012 at 11:09 am

“On the other hand, Tyler is treating this as a real intellectual debate where understanding the other side’s points and analyzing/responding to them is important. This is a mistake. Krugman will not act rationally”

Shouting past each other would do little good. Tyler’s approach is a sound one for the situation. I think Krugman is unlikely to influence anyone not already inclined to lean his way, while Tyler’s less partisan approach is liable to influence a lot of people across the ideological spectrum. At the very least it’s likely to convince people to think about the issues and not just react to a political stance.

TallDave May 15, 2012 at 6:58 pm

One the plus side, that means Krugman generally loses, at least in the eyes of people who care about truth more than ideology.

I May 15, 2012 at 4:28 am

So the longer we sit on our hands and do nothing, the greater the numbers of the unemployed who are now structurally unemployed?

Great, that means if we sit on our hands long enough then we don’t have to feel guilty about sitting on our hands at all as there’s simply nothing that can be done like we always said!

And in case you’re wondering, no I don’t care about the ruined lives this represents. Excelsior!

mpowell May 16, 2012 at 11:14 am

This is the first thing that came to my mind. I wonder why TC didn’t mention it. Plenty of people have been talking about this effect over the last few years.

Kaleb May 15, 2012 at 4:45 am

With sufficient AD, wouldn’t businesses have an incentive to train/retrain workers themselves, thus solving some of the “structural” problem?

Pat May 15, 2012 at 10:47 am

Kaleb gets the point. The quoted material in Cowen’s piece describe how a sufficient boost in nominal demand resulted in the so-called “structural” problems resolving themselves, without particular effort. Astonishingly, those long-term unemployed whom no one would hire were suddenly eminently hireable, now that enough rivets needed riveting to push the economy to full employment. Miracle of miracles, those “hopelessly unemployed” weren’t actually that hopeless at all, once nominal GDP got inflated. (By the way, have you ever looked at GDP growth and unemployment rates during the first four years of the New Deal and compared them to deficits in those same years? If so, you wouldn’t find any of this to be surprising….)

Cowen then misinterprets this drastically: “the path out of high unemployment involved much more than a mere reflation of nominal values.” Sorry, that’s precisely wrong. What was believed to be a structural problem, when nominal GDP was too low, turned out not to be a problem of any sort when those nominal values got reflated.

I’m really not sure if it’s kinder to suggest that misinterpretation is deliberate or accidental.

Bob Knaus May 15, 2012 at 4:00 pm

Read the linked paper. It’s a classic, well worth your time. The change was not driven by nominal GDP. It was driven by the end of skill-killing WPA makework and a shift in to skill-building wartime production.

War is hell, but it does give workers purpose.

mark May 16, 2012 at 7:04 pm

Theoretically yes. but note:
1) The premise “with sufficient AD ..” supports an infinite number of corollaries. E.g., with sufficient AD, unemployed persons would form their own small businesses, etc.
2) Demographics are not favorable to the premise being true
3) There are structural impediments to hiring – excessive regulation of employment, physical space constraints and regulation of expansion, etc.
4) There are risk factors outside of the simple demand -> hiring thesis. Is the financial system safe for holding profits. Is the marginal tax rate so high it skews the risk/reward tradeoff (heads I lose all my investment, tails the government takes 40% of the profit).
5) As well laid out in the literature, if employers believe the “sufficient AD” is merely a short-term effect of an unsustainable stimulus, they are less likely to invest and hire in response.
6) So the question really becomes, how does sustainable sufficient AD arise organically and does it translate that into employment. Much more difficult in the real world than the theory suggests.

ThomasH May 15, 2012 at 5:33 am

I do not understand why this continues to be cast as either/or. Macroeconomic policy needs to keep AD from becoming a constraint and AS policy keeps structural rigities from becoming a constraint. The WWII story to me sounds like “structural” problems melt like butter when employers have an incentive to employ. I’m not aware of any structural rigities that were eliminated just before the War.

Mike May 15, 2012 at 6:51 am

Among other items, most of the National Industrial Recovery Act (basically, mandatory cartel pricing and unionization) was ruled unconstitutional in 1935, which falls after the rise of structural unemployment. Other New Deal laws were also overturned in the mid- and late 1930s.

spencer May 15, 2012 at 11:28 am

I love how people blame the NIRA, that was in effect for only about one year, for all of the problem over about a ten year period –almost half before the NIRA .came into being

Ray Lopez May 15, 2012 at 9:42 pm

Or how people claim Keynesism had an effect when in fact any stimulus was half hearted at best. Let’s face it: the Great Depression lasted so long due to corporatism and the same reasons Japan’s Lost Decades have lasted 20+ years, and that’s that–Keynesianism had no effect, and nor did the Federal Reserve for that matter. Likely both schools (Friedman / Keynes) were wrong. As for this below book, it’s out of print at Amazon.com — RL

*Michael A. Bernstein’s book, The Great Depression: Delayed Recovery and Economic Change in America, 1929-1939, also considers the significant role of structural unemployment (don’t forget my taxonomic caveat) in the Great Depression.

AC May 15, 2012 at 5:52 am

“On the supply side, the growth of high schools and colleges, the postwar draft, and Social Security retirements removed young and old from the labor force”
These things sound like Keynesian stimuli to me… If that’s the supply side reform you want, I doubt you’d get much disagreement from Krugman frankly… More money to schools, workfare projects (draft) and more benefits sounds like basic left-wing proposals…

Bill May 15, 2012 at 11:33 am

AC, they sound like, and are, Keynesian!

If some of the commenters, and even Tyler, had bothered to read the criticism of the book he is quoting from, you would have seen this!

Go to the link and read the critical review.

The author was a follower of a New Deal economist and planner who believed that there should be job retraining programs because he believed the Depression was due to increased mechanization and new industries replacing old ones, hence the need for central planning!

The book criticisms was that it was a jumble of facts used to support a theory. Ask Tyler if he read the entire book and if he disagrees with the cricism.

On the brighter side of the discussion, Tyler did mention Krugman, which is always good for some ad hominems, which also increases comment counts.

Itchy May 15, 2012 at 6:36 am

“Jobs were restructured so that fewer skills were needed. Intensive in-shop and in-school training programs reached millions. Anyone with a modicum of skill was rapidly promoted, even to the status of foreman or instructor. ”

Isn’t this really a big part of the current long term unemployment problem? I’ve heard numerous times that companies can’t find workers with the right skills to fill opening for jobs that are “blue collar”, but technical in nature. In the professional world, it is understood that the fresh out of college new hire is basically useless and needs a lot of on the job training. We call this mentoring or professional development, but it’s accepted that your college degree gets you in the door, but there is much to learn. This does not seem to be as accepted in the blue collar world.

JWatts May 15, 2012 at 11:37 am

“I’ve heard numerous times that companies can’t find workers with the right skills to fill opening for jobs that are “blue collar”, but technical in nature. … his does not seem to be as accepted in the blue collar world.”

There are a couple of caveats to this:
First, when companies say they can’t fine workers with the right skills, the natural assumption is that they mean at a certain (competitive) wage. Obviously, if they can afford to pay more then they can either provide training or raise the wage outright. Businesses hiring professional positions can obviously pay for the higher wage level and still remain competitive.

Second, many “blue collar” positions are highly technical in modern American factories. Factory automation (including advanced controls i.e. computers) have created a dichotomy whereby it has vastly increased productivity (decreasing manpower per product requirements) and also increased the intellectual requirements for the average worker. Many of the simple jobs, feeding machines, basic welding, product delivery, etc have been replaced by automation. So industry needs both fewer workers (per unit of production) and more intelligent workers.

Itchy May 15, 2012 at 9:30 pm

“So industry needs both fewer workers (per unit of production) and more intelligent workers.”

Yes, but I’ve heard the CEO of Caterpillar claim, on multiple occasions on CNBC, that they cannot find people to work in their manufacturing lines or work as technicians. Fewer workers or not, the claim is that there are not enough skilled workers to fill the jobs [at a given wage]

If there really is a skills gap, (we know there are many people looking for jobs) why aren’t some creative companies trying to find competent people and train them for these skilled jobs by offering lowering initial pay in exchange for the training (or something similar) It’s hard to imagine that most fresh out of high school or trade school potential hires posses the skills to operate the equipment or work as technicians.

Either the jobs aren’t really available, or the employers can’t really make a financial case for adding more workers. With 4 unemployed people for each job available, employers should have their pick of workers.

Kyle May 15, 2012 at 6:53 am

I don’t buy that up to 15% of people in 1933 were structurally unemployed. They just decided not to work anymore because the wanted more leisure. During the war buildup they decided to stop being lazy and went back to work.

UnlearningEcon May 15, 2012 at 7:14 am

An excellent example of Poe’s Law in action.

B May 15, 2012 at 11:17 am

Kids these days haven’t heard of Poe’s law, but they have an excellent meme for such a situation.
http://i.qkme.me/x39.jpg

Bill May 15, 2012 at 2:10 pm

Kyle,

You don’t have buy it, you just have to BELIEVE it.

Steven Kopits May 15, 2012 at 8:12 am

Keynesian stimulus will inhibit the resolution of structural problems. It facilitates the perpetuation of non-viable economic activities.

Anyone notice that the Canadian central bank has created an impressive housing bubble north of the border?

DW May 15, 2012 at 8:40 am

So absent the motivating forces of scarcity, etc from WW2, ZMP workers are real and not getting retrained any time soon?

Tom T. May 15, 2012 at 8:56 am

“…we need to discover how the war cured hard-core unemployment permanently.”

Wrecking the productive capacity of the rest of the world ought to figure in there somewhere, but I’m not sure how useful that is as a policy prescription.

Dan May 15, 2012 at 11:16 am

Nearly all industrialized countries increased their output in the early years of the war. The wrecking phase of the conflict came late, long after full employment had been restored.

The Original D May 15, 2012 at 12:19 pm

Well… yeah. R&D to build airplanes, aircraft carriers and bombs costs a lot of money. Just getting an aircraft carrier out of port is a win for GDP. Using it to launch attacks on Germany was a “bonus.”

Benny Lava May 15, 2012 at 8:57 am

Hold your hearses, isn’t this a self defeating argument? Stimulus would have worked in 31 but a mere 2 years later it wouldn’t work? Or maybe it would work but it would cause inflation? Well isn’t that what happened in the post war era? The deflation of the 30s turned into moderate inflation?

Nothing written on this site regarding high unemployment has made logical sense nor been remotely accurate (see: Ireland). Stick with micro and leave macro to the men.

tt May 15, 2012 at 9:11 am

hmmm, , this sounds like mood affiliation

Wimivo May 15, 2012 at 11:54 am

Yeah, I too am getting a bit sick of seeing Tyler accuse anyone and everyone of committing “the fallacy of mood affiliation”. At this point it’s mostly just a repackaged “you’re biased and irrational, but I’m not!”

Rico May 15, 2012 at 9:28 am

I would rather see the question asked this way:

What would be the effect on the unemployment rate of killing roughly 400,000 young American men?

Willitts May 15, 2012 at 11:06 am

400,000 of the most able bodied and productive men, along with hundreds of thousands more who were permanently disabled.

In the 1940s, it meant tapping into the human capital of women and minorities. So the effect on the unemployment rate is ambiguous because we had a large entry into the labour force along with a large decline- soldiers are not included in the labor force, nor are the deceased and disabled.

Today, we don’t have that store of value.

During the recent boom, people were drawn into the labor force or retained by strong opportunities. With those opportunities now gone, they went back to whatever non labour force thing they were doing or now wanted to do.

The Original D May 15, 2012 at 12:20 pm

+100. Not to mention that we didn’t outsource any of these jobs back then. Jobs HAD to be replaced by internal resources. Fortunately we had excess stock due to racism and sexism.

Ted Craig May 15, 2012 at 10:10 am

How much did this play into the issue?

http://en.wikipedia.org/wiki/Dust_Bowl

B May 15, 2012 at 11:20 am

An honest-to-god technology shock. We don’t see those very often.

Eric H May 15, 2012 at 9:07 pm

According to Hugh Bennett’s report, “The Federal homestead policy, which kept land allotments low and required that a portion of each should be plowed, is now seen to have caused immeasurable harm. The Homestead Act of 1862, limiting an individual to 160 acres, was on the western plains almost an obligatory act of poverty.” Many people seem to think that federal conservation programs fixed the problem, but Hansen and Libecap (“Small Farms, Externalities and the Dust Bowl of the 1930s”, NBER) find that local programs were more successful.

Lord May 15, 2012 at 10:37 am

This minimizes what are probably the largest contributors, the plunge in population growth during the depression, the constraints and rationing during the war, the destruction of economies abroad, the redistribution of wealth through wages during the war, and the large Keynesian efforts after it including the GI Bill, free college education for returning draftees,and subsidized mortgages rates.

orionorbit May 15, 2012 at 11:50 am

Tyler, we DID get a natural experiment in WW2 and unemployment did indeed come down. Now, you offer absolutely no explanation about what would make the NAIRU dramatically rise in the early 20s (a 10 pct points jump is as far as i know something that has never happened anytime, anyplace). Additionally, your explanation for an equally dramatic and unprecedented drop in the NAIRU after WW2 doesn’t hold water either, given that the first half of the 20th century was one of the worst post industrial revolution periods for TFP growth (source: http://www.nber.org/chapters/c10122.pdf ). Surely, if veterans acquired such strong skills to lower the NAIRU by the dramatic amount you propose, such a structural swift would have left a very noticeable imprint on TFP growth; and it simply didn’t.

B.B. May 15, 2012 at 1:50 pm

People like Krugman look at WW2 and see only a large rise in government spending.

Is that all we need? The government handing out big checks?

It is very incomplete. What we saw in WW2 was millions of men conscripted, regimented, organized, disciplined, trained, moved around. Such actions change culture and psychology; handing out checks does not. The federal government created the Organization Man. The vets, returning to a civilian economy, were prepared to join a union, pay dues, and work on an assembly line, or they were prepared to work in cubicles in Big Corporation, or to be obedient civil servants. And be good spouses and church members. The government created “the 1950s.”

It is ironic that Krugman celebrates a decade of conformity.

It had its advantages: booming stock market and productivity, low inflation, balanced budgets, low long term unemployment.

We could do the same. We could conscript the poor and unemployed and hammer them into being soldiers, creating down the culture of poverty and helplessness. The penalty for insubordination is the firing squad. We could teach them to be conformists and to be conscientious and to stop questioning authority. We could give them basic training and schooling. Then they could get jobs. Is that the price we want to pay? If you want fiscal stimulus, this is the way to do it. If this approach offends you, try Scott Sumner and look for monetary solutions.

TallDave May 15, 2012 at 3:48 pm

They also generally seem to forget that WW II saw the kind of rationing that is nigh unthinkable today. You literally could not buy hundreds of the most common household goods without a ration book.

TallDave May 15, 2012 at 3:47 pm

The elephant in the room here is the massive wage controls and labor market restrictions FDR undertook. Is that “structural?” Either way, I think they should not be ignored.

Donald A. Coffin May 15, 2012 at 4:46 pm

I’m going to read the entire thing, but there is certainly one thing in what Tyler quotes that I have to disagree with: “…we need to discover how the war cured hard-core unemployment permanently…Social Security retirements…[the] old from the labor force…”

This cannot refer to the 1940-1942 decline in the unemployment rate from the mid-teens to less than 2%, because no retirement benefirts were paid until several years after WWII ended. So it must refer to post-WWII develoments that *prevented* the recurrence of one component of structural unemployment–among those age 65 and older–after the war. (Incidentally, the same passeage refers to the decline in structural employment among younger age cohorts as a consequence of expanded higher education–also a post-WWII phenomenon).

So this part, at least, of his explanation leaves the question of how, if indeed a large percentage of the late-1930s/early 1940s unemployment was structural, the unemployment rate fell so far and so fast.

Steve Sailer May 15, 2012 at 6:22 pm

Evidently, contrary to fashionable thought, patriotism is good for the economy.

Lord May 15, 2012 at 10:11 pm

People hold veterans in higher regard than the unemployed.

Jason May 16, 2012 at 1:12 am

“It is important to learn from this literature rather than dismiss it.”

Table 1 of Jensen shows a 4 point decline in “hard core/structural” unemployment going from 1940 to 1941 before any of the shooting had begun.

458,365 active duty military personnel in 1940, up to 1.8 million in 1941. That is a change from 0.35% to 1.35% of the population. Therefore even assuming everyone who joined up didn’t have a job beforehand that leaves at least 75% of the “hard core/structural” unemployment decrease to other factors besides joining the army.

A maximum of a year of job training got rid of half the hard core? How hard could this core be?

“The war, by removing millions of prime men from the labor market, by restructuring the work process, by subsidizing wages, and by massive retraining, finally gave the private sector the methods and the incentives to rehire the hard-core.”

Again, already half the hard core in Table 1 was gone by 1941. The unemployment rate in 1942 was 4.7%, so assuming ~ 4.5 to 5% frictional unemployment as seems typical in Table 1, this implies all the hard core was gone by 1942. It happened too quickly for it to be some massive retraining or restructuring of processes. Only 1.5% more of the population went into the army between 1941 and 1942, so basically another 75% of the reduction in hard core unemployment was going into industry.

Removing prime men from the labor market doesn’t make some hard core unemployed person suddenly have new skills — he only looks like he has skills relative to the applicant pool. Any tightening of the labor market should decrease the hard core unemployed. But a tightening of the labor market *is* a reduction in the number of unemployed period — so it seems like an artificial distinction and the statement about removing prime men (or tightening the labor market) reducing the “hard core” unemployment is basically a meaningless tautology.

Merijn Knibbe May 16, 2012 at 8:01 am

As far as i’m concerned this all sounds soooo Post-Keynesian… We have to invest in people and to invest in the future, instead of cutting wages and safe for the future.

However – in 1945/1946 the backward bending supply curve of labor did the trick: the participation rate went down with about 4 to 5% in the USA, while the number of hours worked per year also declined with about 400. Which of course was a ‘free market’ phenomenon.

Comments on this entry are closed.

Previous post:

Next post: