David R. Henderson  

Unemployment and Wages: Find the Flaw

More Notes for a Debate... House Price Behavior...

Here's a paragraph from a recent David Leonhardt story in the New York Times:

In Germany and Canada, some companies and workers have averted layoffs by agreeing to cut everyone's hours and, thus, pay. In this country, average wages for the employed have risen faster than inflation since 2007, which is highly unusual for a downturn. Yet unemployment remains terribly high, and almost half of the unemployed have been out of work for at least six months. These are the people bearing the brunt of the downturn.

There's one word in the above that is quite jarring. What's the word? For full credit, give a brief explanation for your choice.

HT to Tyler Cowen for mention of article, although he focused on something else in it.

Comments and Sharing

CATEGORIES: Labor Market

COMMENTS (27 to date)
Blackadder writes:

It's the first word of the third sentence: "yet." It suggests that there is something strange about unemployment being high when wages haven't fallen, when in fact this is exactly what one should expect.

Daniel Kuehn writes:

I imagine the concern is with the word "yet" at the beginning of the second to last sentence. Wages are high, so the "yet" should be a "thus".

Daniel Kuehn writes:

If that's right, how disconcerting should this really be? I think it depends on what exactly "average wages" is. DeLong pointed out this problem with Galloway and Vedder's book years ago. Yes, aggregate wages are countercyclical. But before we read too much into that, we should also note that individual wages are procyclical or at least acyclical. That doesn't mean sticky wage type thinking is wrong - it just means we need to be careful how to interpret these aggregated statistics.

lukas writes:

"Yet?" Mamma mia!

Philo writes:

I don't see the objection to 'yet'. There *is* something surprising in the bifurcation between those who continue to be employed, whose real wages are rising, and the relatively large number who are out of work, with wages of zero. One would have thought that a sharp decrease in unemployment would accompany the rise in wages (rates and total pay per employed worker).

I did wonder whether the "average wages" in the second sentence were per hour (rates) or total pay. Usually, 'wages' means 'wage rates', but the first sentence had mentioned only total pay--nothing about rates.

Andrew writes:

Were the wages of the unemployed (i.e. zero) included when calculating that average wage? Would it really be "unusual" for the wages of those employed to be relatively high and outpace inflation when unemployment is high?

Jody writes:

Not only is the "yet" jarring from an economics perspective, it's jarring from a direct application of the logic of the preceding sentences.

A leads to B
Yet !A did not lead to B.

(where in this example A = wages fall and B = unemployment falls)

Scott Sumner writes:

Isn't that the same newspaper that once reported something like "despite fall in crime, the number of people in prisons soars."

I'm not even a fan of our high incarceration rate, but that made me roll my eyes.

Corey S. writes:

Leonhardt's understanding of economics is poor at best. He links high unemployment to a lack of bargaining power for workers, and laments the fall of unions. This makes no sense. If businesses have all the bargaining power, they should have little reason to fear hiring new employees, since they would be able to fire employees at will.

The main reasons unemployment remains high are 1) uncertainty about future government action and 2) continually extended jobless benefits. But what do I know, I'm an uncaring libertarian who is being controlled by the Koch brothers, or something.

Steve S writes:

After reading the comments I would concur with those who point out the illogical use of the word "yet", though I'll admit that didn't register on my radar during my initial reading of the paragraph.

I was going to go with "agreeing" in the first sentence. You'd have to investigate on a case-by-case basis, but companies imposing furloughs or majority-votes in favor of decreasing hours isn't exactly my idea of an "agreement". More like the loosely used and highly democratic "consensus".

David R. Henderson writes:

The various people above who caught the "yet" are right. It is jarring.

@Daniel Kuehn,
I lost you on your second comment. Could you explain? Oh, and btw, welcome back.

I don't get your first paragraph. What you're saying would make sense only if there were too few people looking for work, in which case an increase in wages would draw some of them into work. I don't think that's the case when we have a 9+ percent unemployment rate.
Your second paragraph, though, does make a good point: I'm assuming Leonhardt meant wage rates per hour, but I don't know that.

Good one! And, like you, I don't advocate putting more people in prison.

The wage data are always of people with jobs. They don't include a zero wage for people without jobs.
As to whether it's highly unusual, wages were high in the 1920-21 recession, the 1929-33 depression, and this recession. In the other ones, they fell, which was a problem for both the Keynesian rigid wage story and the Friedman job search story.

John Thacker writes:

I assume that David Leonhardt is thinking about the causation in the other order: "High unemployment should drive down wages, but wages have been going up and yet we still have high unemployment." In other words, he's thinking of sort of a Marxist "reserve army of labor" concept driving down wages.

Many of us, however, view that as a second order effect, but view as the first order effect "high wages cause high unemployment."

Dewey Munson writes:

Informal economist 89 yrs old.
How come I see "averted" as the word because of the inference of future action?

a question.

Why don't companies (economists) realize that all together their employees are their customers and reducing their income is counter effective?

Silas Barta writes:

I'm not sure I can agree with the "'yet' is jarring" consensus. Isn't there a plausible, market-based story why wages for the employed might reasonably go up at the same time unemployment goes up?

In fact, that meshes well with a recalculation story, in which there is a set of workers with skills that actually match what's in demand, and another set who built up useless skills that put them at effective zero-marginal-productivity. (Real estate workers, lawyers, and liberal arts majors, I'm looking in your general direction here.)

In other words, don't commit the "lump of labor fallacy" because labor is heterogenous, and you can't reach full employment by just cutting back on the labor of good workers and "spreading it around".

Joey Donuts writes:

To help you get over your jarring sensation, I will attempt to clarify what the author tried to convey.

In Canada and Germany some companies have shared the pain of the down turn by sharing jobs among employees. Here in the U.S. real wages have risen during the down turn. YET companies and workers have not shared the pain of the recession by sharing hours and jobs.

Norman Pfyster writes:

Econ 101: When demand declines and supply remains constant, prices should drop.

Econ 201: Sticky wages, heterogenous labor, etc.

Leonhardt's "yet" is the Econ 101 question: wages should fall, "yet" they haven't (although by distinguishing between wages of those employed and aggregate wages, he has already muddied his argument). Econ 201 provides all kinds of nuances for why Econ 101 is simplistic. The "yet" is less jarring if you step back through all the hoops to get to the explanation of the observed phenomena.

David R. Henderson writes:

@Dewey Munson asks:
Why don't companies (economists) realize that all together their employees are their customers and reducing their income is counter effective?

My answer: Check out my post on this:

Chris Koresko writes:

Silas Barta: Isn't there a plausible, market-based story why wages for the employed might reasonably go up at the same time unemployment goes up?

In fact, that meshes well with a recalculation story, in which there is a set of workers with skills that actually match what's in demand, and another set who built up useless skills that put them at effective zero-marginal-productivity.

Or perhaps when times are tough, the first workers to be laid off are the low-paid ones, as companies seek to protect their organizational capital (which resides mainly in the high-wage workers). Then the average wage of the remaining workers should rise.

I like your point about the fallacy of aggregation. Micro laws are only valid when applied to homogeneous goods/services. Simply averaging together a bunch of heterogeneous workers, goods, or services doesn't convert them into a homogeneous whole, and attempting to treat them as such leads to false conclusions.

joecushing writes:

I'm going with "thus"

Just because hours fall doesn't mean pay falls. If wages or benifits go up, total pay can rise or stay the same as hrs fall. Also, the story is talking about hours in Canada and Germany, while it talks about wages here. These are two different sides of the total pay coin.

Daniel Kuehn writes:

David - on the second comment... I'm still new to the literature, but my understanding is that average aggregate wages for most of the post-war period have been counter-cyclical like this. They seem to rise a lot more during downturns than they should. This has lead people like Galloway and Vedder to conclude that its government policies and unions keeping wages up that causes unemployment, and it has lead New Keynesians to make a big deal about sticky wages after several decades of that particular piece of theory being out of the limelight.

However, a lot of recent research on real wage cyclicality has suggested that wages are actually weakly procyclical. The problem is, there is a composition effect - low wage workers get laid off during downturns at higher rates, pushing average aggregate wage rates up.

So the author sees average wage rates go up and unemployment go up. Is this unexpected? Well, it all depends on whether that wage cyclicality holds up at the micro-level, controlling for skill level etc. I don't know what wage cyclicality has been doing in this downturn. But if its followed other post-war downturns, it's quite likely that while aggregate wages are counter-cyclical, disaggregated wages are weakly pro-cyclical and the "yet" becomes far less disconcerting.

Philo writes:

I still don't get it. A trend of rising wages (with the supply of workers or potential workers holding more-or-less steady) suggests increasing demand for labor; that makes it surprising that unusually many potential workers are out of work. Not impossible, of course, but surprising, justifying the term 'yet'.

I do see a minor stylistic objection to the passage: the author means to say "rising wages *yet* high unemployment," which I think is OK. But he actually says something closer to "unusual occurrence in a downturn *yet* high unemployment," which is not what he meant.

andy writes:

How "highly" unusual are rising real wages during recessions?

Charles R. Williams writes:

It would be good to disaggregate the data. Two important things have happened this recession that might cause aggregate wages to rise as employment falls. First, there have been huge increases in the minimum wage in the last few years and this would affect workers at any wage close to the minimum. Second, government programs have drastically reduced the marginal incentives to work for low income people by extending unemployment benefits, allowing people to go to school while on unemployment, and covering medical care for some low income unemployed.

I know people who have effectively ceased looking for work or who have turned down jobs for the second reason.

ben writes:

'yet' kind of makes sense. a higher price than normal is sometimes associated with a shortage. so it is odd that we have a surplus and a high price.

David R. Henderson writes:

No. My reasoning is that it's the high price that leads to the surplus.
@Daniel Kuehn,
Thanks. Now I get it.

Eric Hosemann writes:

If at first a thing you don't understand, look to the laws of supply and demand!

Philo writes:

I finally understand your point about the Leonhardt quote. Your thought is: high wage-rates imply that less labor will be demanded. But if there are lots of unemployed people--lots of potential labor supply--why haven't wage-rates been bid down? If wage-rates are going up, you would expect (_ceteris paribus_) that supply was being constricted; YET we see all these people looking for work.

Maybe if we had New-Deal-type programs to raise wage-rates artificially, and everybody knew about them, the word _yet_ would be inappropriate. But we don't. So I am still siding with Leonhardt: what is going on--the radical disequilibrium in the labor market--is anomalous, 'yet'-provoking and 'yet'-justifying.

Comments for this entry have been closed
Return to top