Thoughts about labor markets

Today, the number of layoffs and discharges is very low — in fact, layoffs are at their lowest level since the Labor Department began collecting this data in 2000. Today’s problem instead is the very slow pace of job creation.

Yet unemployment remains high.  File under “increasing polarization of labor market outcomes.”

Here is Steve Pearlstein channeling A. Michael Spence:

…what ails the U.S. economy is primarily a structural problem, not a cyclical one that can be effectively dealt with through the magic of short-term Keynesian stimulus.

Or let’s consider David Leonhardt’s very good piece on the labor market:

Lawrence Katz, a Harvard labor economist, calls the full [labor market] picture “genuinely puzzling.”

That is Larry Katz, who was chief economist at the Department of Labor under Clinton and who is arguably the most knowledgeable labor economist in the world.  Larry Katz, who is renowned for how many literatures he holds at his fingertips.  (Here are some papers by Larry Katz, including a good, short piece on unemployment in the great recession.)  Larry Katz, who wrote the most effective critique of the Lilien “sectoral shift” hypothesis.  Larry Katz, force of nature.

I am not asking you to agree with Spence or Katz, only to keep them in mind.  What is striking about the current generation of popular Keynesian models is how little effort they make to integrate cyclical and structural phenomena.  This is very much like some of the original Keynesian models of the 1930s, but it is an out of date approach and it will lead to excess optimism about the ability of fiscal stimulus to set things right.


From the schemes I have come across a negative income tax seems like the only way to make a structural economic shift without a huge amount of pain. Are there any other realistic methods?

Perhaps I am missing something, but could you clarify why anemic job creation rates are necessarily a structural problem and not a demand problem? I'm not suggesting demand approaches are proven by this concern - I'm just curious why you're under the impression they're ruled out by it.

The word 'yet' seems to imply conflicting facts. Not sure why that would be the case. If the economy grows slowly, wouldn't you expect layoffs to ebb and unemployment rates to flatten?

Perhaps I am missing something, but could you clarify why anemic job creation rates are necessarily a structural problem and not a demand problem?

I think that is the proposed interpretation of of the rate of 'layoffs and discharges' being relatively low. Workers who are currently employed are valued by their employers AND those employers are profitable enough to keep those valued employees on staff. In a generally weak economy with insufficient overall demand, you'd expect the rate of layoffs and discharges to be higher. But workers who do not have jobs have a great deal of trouble finding them (which we see in the growing numbers of the long-term unemployed). Contrast this with a non-structural situation where unemployment may be high but the length of unemployment is lower and workers move back and forth between jobs and periods of unemployment.

In a period of weak demand I would have thought you would expect employment to be lower. I'm not clear on why you would expect that to express itself in one gross flow vs. another. Initially, if the marginal efficiency of capital or labor drops, the concomitant drop in demand would be expected to express itself through layoffs. This is what we see in the JC and JD data - an initial job destruction spike. If demand continues to be weak, you would imagine that normal hiring rates would be dampened (why hire more?). Why would you continue to have layoffs if you've presumably reached something like a stable underemployment equilibrium predicted in these sorts of Keynesian models?

This sort of depressed-with-no-recovery state could by a symptom of a lot of different underlying problems: a Keynesian underemployment equilibrium, a structural shift in the natural rate of unemployment, technological unemployment (I'm guessing Tyler discounts this one given his recent tome), or a combination of the three. But I don't see how any of this would be unexpected from a Keynesian story. You explain the structural story. I understand the structural story and think there's likely something to it. I don't see how we have anything here that rules out a Keynesian story.

Keynesian models don't predict an "underemployment equilibrium, they predict a death spiral. Low demand -> reduced employment -> lower demand -> further reduced employment -> etc. -> we're all dead.

A death spiral? How melodramatic! I hadn't realized that was what was called for.

You're not thinking of the long run, are you? Because we are all dead then.

In a period of weak demand I would have thought you would expect employment to be lower.

You would. But would also expect layoffs to be higher. And they aren't. Even during a period of weak demand, the labor market is not static -- workers are hired and laid off, business grow and shrink, some firms go out of business and others are started, etc. If the problem was low overall demand, you'd expect the number of workers 'on the bench' waiting for a chance to work to be higher than usual. But you wouldn't expect some workers to go to the bench and never get off again. Not unless their skills were no longer valuable (think of a bunch of 350 lb offensive lineman trying to get off the bench and onto the pitch in a soccer match after the bursting of a huge bubble in American football). Construction workers = offensive lineman.

"You would. But would also expect layoffs to be higher. And they aren’t."

This is exactly what I'm wanting you (or Tyler) to explain - why would you expect layoffs to be higher? Once you reach an underemployment equilibrium, why would you expect either rate to be any different from usual? And if you did expect employment to continue to drop why are you insisting that higher layoff rates are the way the Keynesian model predicts it will happen? What's wrong with it happening through lower hiring rates? They both produce the same result.

"Not unless their skills were no longer valuable"

Isn't this the very heart of the Keynesian model? Low marginal efficiency of capital and labor?

When the economy was good, it was fast. Companies needed to retain all this excess labor capacity because it made them agile. An idea could be churned into a product very quickly which would give a company an advantage over its competition. Credit was cheap, and thus made risk-taking cheap. Cheap credit also accelerated demand in a sort of feedback loop which boosted AD.

when the credit market collapsed, companies shed their excess labor (ZMP) as risk-taking became too expensive. Risk-taking was always expensive, it's just the new credit market that emerged in the 2000's made it artificially cheap. and now, the real cost of that cheap credit is realized and the economy has basically reset. All that excess labor came at the cost of real wages increases, and now that layoffs has leveled off, I think real wages will increase before unemployment decreases. which means that the current unemployment is structural as real wages have to "catch-up" before ZMP is no longer zero.

Where do you get your data?

Take a look at initial claims:

There's no sense in which this is back to 2000 levels

Isn't the claim that new hires are down? Layoffs can be the same as 200 levels, but if fewer new openings exist in 2011 than 200, then yes, initial claims will be higher than 2000.

From the same site, isn't this the better graphic (though it unfortunately only goes back to 2001):

Layoffs are just below the 2001 level, while hires have been steadily declining over the ten-year period.

hm, those two 200s should be 2000s. Though an 1800 year employment history would be interesting . . . .

Judging by the "began collecting the data in 2000" language, I think he's refering to the JOLTS survey, which does show the lowest levels of layoffs since 2000. You're looking at unemployment insurance claims. Layoff statistics have been collected long before 2000, though, in the CPS (I'm not sure if they have an easy wizard for this on the BLS website, though). I'm not sure how this compares to the pre-2000 CPS data.

Tyler, ignoring everything in your post except the last paragraph, you're wrong about Keynesians' optimism about the ability of fiscal stimulus to set thing right. Actually Keynesians are confused and pessimists about macro policy. You should read what O. Blanchard learnt from last week IMF conference on the future of macro policy. See
I don't have the time now to go through the details of Blanchard's conclusions, but I'm glad that his assessment amounts to recognize the fiasco of macroeconomic theory and policy in the past 50 years. I hope to have time later today to go through his 9 points. In the meantime hope you write about them.

the current generation of popular Keynesian models

Yep, everywhere you look there are powerful political figures just relentlessly pushing
the importance of spending! And when do you ever hear about the deficit? Oh, wait...

I don't think anyone has been proved to be "over-optimistic" yet. This Scott Sumner post seems to make a lot of sense to me:

...once NGDP growth decouples from the job market, then I'll be surprised. As of now, nothing has been surprising to me, taking a strictly monetary perspective.

Prof. Cowen, the fixation on structural factors in discussing what ails us is really rather weird. And my guess would be that Larry Katz doesn't share your emphasis. Channeling another Larry might be helpful: there are potholes, look around.

Larry Kudlow?

Now that's funny! (And even funnier in the context in the unmodified quote.)

We should indeed try to integrate cyclical and structural models, and the link to Katz's short testimony is very helpful in thinking about that. But look again at that piece. Katz emphasizes that for now by far the biggest problem is a cyclical drop in demand. The main immediate priority is to fix that. That won't cure anything, and more needs to be done to address long-run structural problems. But the tone and emphasis of this post strikes me as very different from that in Katz.

What Fiscal stimulus? Also, why did these structural factors just appear suddenly in 2008-09?

Keynesian economics was a lovely idea for a previous simpler generation. We have outgrown the model where the state replaces demand because we have overused the credit of the state under the auspice of Keynesian/Regan/Voodoo/Bush economics and now the state no longer has the wherewithal to inject enough stimulus to effectively course correct a state ship of this size. Perhaps we need a bigger rudder...

I am not asking you to agree with Spence or Katz, only to keep them in mind. What is striking about the current generation of popular Keynesian models is how little effort they make to integrate cyclical and structural pneumatic represents tools and compressors that are engineered for high performance. In addition to pioneering design, high-impact technology, and durability, our Pneumatic means customer value.

I read Katz's short paper about unemployment during the so-called "Great Recession" and noted that he dismissed the idea that extending unemployment benefits increased the length of unemployment. This conclusion came from a paper that he co-authored which is not available for free, but I did find an abstract that indicated that each week of extended benefits was associated with an increase of 0.16-0.20 weeks of unemployment. I may be mistaken, but isn't this equivalent to one out of six to one out of five workers remaining unemployed another week? That doesn't seem insignificant to me.

I am also somewhat puzzled about discouraged workers. How do they survive without working? Do they subsist on savings or do they sell some of their possessions? Do they depend on a working spouse? I know that I would (at least when I was younger) be willing to take any job just to keep money coming in.

I am also skeptical about training programs - I doubt that they really help the unemployed that much but do give them the feeling that they are doing something when they are not working.

What would be useful (but very time consuming and difficult) is to follow a number of the umemployed to see what is really happening from the micro level.

BTW, the blog does look much cleaner and more attractive. Not that it was unattractive before.

I might add that I would be less likely to take any job now only because my most desireable alternative would be retirement. I'm enjoying my work, which keeps me working, but if my job ended involuntarily, I'd call it a day (so to speak). Five years ago, I would have taken the best job available.

I dunno about your models, but its a common view in California among small businesses that you do not hire unless you absolutely must, because each hire causes excessive headaches beyond just training, real estate, morale: increase in UI, prolonged, expensive termination process and inevitable law suit, HR and legal costs, etc., Certain employees are experts at the game, interview well and then immediately jump on disability or start building their "file". Plus onerous regulation often kicks (and subsdies shut down) when total employments exceeds headcounts (e.g., 10, 25, or 50 employees).

That, plus negative income taxes, probably make 5% unemployment a historical relic and 8% the new 5%.

won his freedom; not because his confession was recanted or fresh evidence was presented, but because of a wad of cash. He paid the victim’s family $9,400 and walked out of prison a free man. The slain man’s relatives said they would use the

Comments for this post are closed