Paul Krugman has an excellent post today titled “Is the Great Recession Holding Down Wages? (Wonkish)“, New York Times, May 4, 2018. It appears the same day that the Bureau of Labor Statistics reports an unemployment rate that has broken below 4 percent for the first time this century.
Paul didn’t need to add the word “Wonkish.” It’s not really that wonkish. (And I wish my spell checker wouldn’t keep changing it to “Monkish.”)
Here, in his words, is the puzzle:
the unemployment rate is now very low by historical standards. In fact, it’s back down to 2000 levels; those were the days when employment was so full that people used to joke about the “mirror test” for employment: if your breath would fog a mirror, that is, if you were alive, you could get hired. (Sorry, zombies.) Yet wage growth remains restrained, well below pre-crisis levels:
And here’s key background for his tentative explanation:
The notion that firms are very reluctant to cut wages has a long history, for a very good reason: it’s true. That truth has been obvious to many observers; Truman Bewley made a systematic survey to confirm the point. Employers believe that actual wage cuts, as opposed to, say, letting real wages erode via inflation, are demoralizing and perceived as unfair. So there tends to be a zero lower bound on wage changes, except in the face of very high unemployment.
Co-blogger Bryan Caplan has in the past highly recommended Bewley’s book.
Now to his tentative hypothesis:
OK, here’s my theory about the brontosaurus, I mean, about wages. What employers learned during the long slump is that you can’t cut wages even when people are desperate for jobs; they also learned that extended periods in which you would cut wages if you could are a lot more likely than they used to believe. This makes them reluctant to grant wage increases even in good times, because they know they’ll be stuck with those wages if the economy turns bad again.
And, as a bonus, another datum that fits his hypothesis:
This hypothesis also explains something else that’s been puzzling me: widespread anecdotes about employers trying to attract workers with signing bonuses rather than higher wages. A signing bonus is a one-time cost; a higher wage, we now know, is more or less forever.
I had been puzzled about this too. But Krugman’s explanation gave me an “ah-hah” moment.
Here’s one thing I’m wondering about those bonuses, though. Let’s say you give an employee a $2,000 signing bonus. What’s to stop him from taking it, showing up for a couple of months (or even less), and then quitting? What I would like to see is how the payment of the bonus is structured. Is it pro-rated over, say, 6 months, so that the bonus per month is $333 and the moral hazard is avoided? Does anyone reading this know the answer?
By the way, Krugman’s line in the first quote about the “mirror test” reminds me of two things. First, my Hoover colleague Richard Epstein had a great line at about the 18:00 point of this interview: “Now you see signs ‘Criminals, please apply.'” LOL.
Second, I’ve been having trouble having my print Wall Street Journal delivered in the last few weeks. They’ve missed days and sometimes got to me as late as the afternoon. (I like to read it with my morning coffee.) When I told the woman at the Journal on the phone the other day that I had been so happy with the deliverer until recently and, as a result, had given him a $20 as a late Christmas bonus, she confessed that with the unemployment rate as low as it is, it is hard keeping good people.
HT2 Tyler Cowen.
READER COMMENTS
Michael Watts
May 4 2018 at 1:50pm
The signing bonus I was offered in 2016 was paid entirely up front (net of taxes; bonuses are subject to tax withholding as if they were part of your regular salary, or in other words subject to a pretty extravagant withholding that is known to all involved parties to be incorrect).
It was subject to clawback if I left the company within 12 months.
David R Henderson
May 4 2018 at 1:53pm
@Michael Watts,
Thanks.
It was subject to clawback if I left the company within 12 months.
How would they claw it back? Out of your last pay check?
rtd
May 4 2018 at 2:22pm
It’s pretty simple – there’s often a contract that states if the employee voluntarily leaves the company before a certain amount of time, they must reimburse the employer within a certain amount of time. Who is going to fight that?
DJ Campbell
May 4 2018 at 2:24pm
The signing bonus I was offered in 2015 was structured the same as Michael’s. I have no clue what they would have done to try and recover it if I had left within 12 months. There was probably some fine print somewhere I didn’t read.
Oscar Cunningham
May 4 2018 at 3:26pm
If wages are now upward sticky as well as downward, should central bank inflation targets be lowered? To 0%?
David Boaz
May 4 2018 at 4:03pm
Maybe this is obvious. But Bewley says
So they don’t want to increase wages now. But the theory also implies that they didn’t cut wages during the recession? So that would mean there’s no “lower wage” that should obviously be increasing now.
David R Henderson
May 4 2018 at 4:33pm
@David Boaz,
So they don’t want to increase wages now.
No. They are increasing wages now. Look at the graph. It’s a think-on-the-margin thing. Paul’s perplexion, and mine, was about why wages aren’t increasing as much as we would have expected.
Justin
May 4 2018 at 5:03pm
I don’t think there’s much of a wage puzzle at all.
The average ratio of labor compensation to national income from 1947-1973 (which I think is fair to consider as a golden economic age from Krugman’s perspective) was 62.5%.
It was 62.1% in 2017, and averaged 62.5% from 2007-2017. It’s currently a little depressed compared to the Clinton years and most other business cycle peaks, but not drastically so. A 4.4% across the board compensation increase would raise it to 64.8%, the same as the average for Clinton’s 2nd term and the 1973 business cycle peak.
Now a 4.4% raise is always welcome, but a gap that small makes it hard to make the case that something is utterly broken.
XVO
May 4 2018 at 5:11pm
The signing bonuses at my company, an auto parts manufacturer, stated that you have to stay for 2 years or you have to pay the entire sum back when you quit.
As soon as my first 2 year signing bonus expired I locked myself in for another 2 years by relocating and the relocation and relocation bonus would have to be repaid prorated.
Another benefit for the employer of these signing bonuses is the leverage it gives them over you in the future. It weighs heavily on my mind that I have to pay a big sum back if I want to leave.
Justin
May 4 2018 at 5:12pm
Wider use of bonuses make sense. It avoids the sticky wage problem, and from my understanding have been used often in Japan in recent decades.
I found a paper published back in 2002 in the Japanese Economic Review titled “Why has the Unemployment Rate Been so Low in Japan? An Explanation by Twoâ€Part Wage Bargaining” mentions the use of bonuses in the abstract:
–“The Japanese wage payment system is considered from a perspective of twoâ€part tariff pricing. Using the “amusement park†analogy, Shunto wages can be regarded as an “entrance feeâ€, whereas bonuses are a “variable chargeâ€. Empirical investigation showed that a qualitative difference exists between these two types of wage: Shunto sets the coordinated wage rate by focusing on the whole labour market condition, while bonuses respond to idiosyncratic shock. Based on the standard prediction of twoâ€part wage tariff pricing, such a unique combination is the ultimate source of Japan’s low unemployment.”–
Maximum Liberty
May 4 2018 at 5:13pm
I’m on the other side of the signing bonus — the person who writes the offer letter that includes it.
I see two cases.
In the first, we include a clawback provision. Ideally, we’ll include an attorneys-fees provision, too. Typically, an employee will believe (correctly) that we can effectively bankrupt them in court if they refuse to pay it back. We would do so in order to maintain credibility, including for other programs (e.g., tuition reimbursement, relocation) where there is clawback.
In the second, we don’t get a clawback because the position is an executive position where we are competing for the target and the target would take a significant reputational hit by taking the signing bonus and quitting.
Hana
May 4 2018 at 7:34pm
To add words to your Word dictionary, right click the underlined word and then select “Add to dictionary”.
Joe R
May 4 2018 at 8:55pm
I think there are several features of having a sign-on bonus.
1.) As mentioned, it provides the company flexibility to lower wages later in case of economic downturn.
2.) It also provides flexibility in case the employee ended up underperforming. There’s always uncertainty about how well an employee will work out, and this provides options other than overpaying or firing.
3.) Companies routinely differentiate between one-time costs and recurring costs. Sign-on bonus is better than salary from that perspective.
4.) The clawback if they leave early is a powerful incentive to keep people on. They may spend it immediately, so they won’t want to leave.
5.) For some, the hiring bonus is better than salary because you get it all at once.
On the receiving side, sign-on bonus is potentially worse because:
1.) You have to pay it back if you leave.
2.) Your salary at your next job is often based on your current salary. So salary is better than sign-on bonus for long term salary.
3.) It can increase your taxes in a single year vs. salary would be taxed over a longer time period.
sk
May 4 2018 at 9:41pm
What Krugman leaves out of the discussion is that there may be more payment of performance bonuses than in the past. Maybe business is better trying to tie outcomes (worker marginal productivity) with compensation? Worker comp. includes more than just a wage payment.
How about the granting of stock options or stock rewards, deferred compensation via profit sharing? I know many who are no where near the C-Suite whose company makes yearly payments of deferred compensation to retirement plans that are over and above any of the matching 401K plans out there. How about subsidized health care?
Thomas
May 4 2018 at 11:23pm
Why would you have a prior about the rate of wage increases? I understand why Krugman would have one: It gives him another stick (or strawman) with which to beat free markets.
TF
May 5 2018 at 1:24am
Isn’t Krugman just talking about sticky wages? Sumner has been writing about this for years, decades even. Prof. Sumner please comment.
Bruno Duarte
May 5 2018 at 6:53am
CPI tracks unemployment but it is forecast by industrial production by about a quarter – except for the past five years. There’s a stickiness – but Paul Krugman knows it doesn’t resist churning past a few quarters.
Churning is returning while CPI recoupled with industrial output only since 2016.
Alan Goldhammer
May 5 2018 at 7:49am
I wonder if contract work is part of this. In some industries, short term projects are now being outsourced where the payment of benefits and concern about increasing the permanent workforce become a non-issue. I don’t know if any of the extant data sources break this out.
John P Palmer
May 5 2018 at 8:08am
During the oil booms in Alberta, numerous places were offering very large bonuses to new employees, but only if the employees stayed for 6 months or a year (depending on the firm). I don’t know what protections were in place to keep employers from letting someone go one day short of the bonus-qualification date.
For example: http://www.eclectecon.net/2008/01/slop-over-effec.html
Michael Byrnes
May 5 2018 at 8:23am
@ Michael Watts
Same deal for me – I got a sign on bonus at my last hiring, payable immediately, that was contingent on my staying on for 12 months.
jccooper
May 5 2018 at 9:57am
The unemployment rate has recovered to 2000 levels but the workforce participation rate is still at Great Recession levels. It’s possible, I suppose, that the diminished labor pool is permanent, but I doubt it. The cause seems to have been largely “lack of jobs”, rather than age or other structural changes, and so should be reversible (and reversing, though there’s no visible uptick yet). I would expect we won’t get back to “mirror test” conditions until that missing part of the labor pool finishes returning from exile.
Michael Sandifer
May 5 2018 at 10:12am
I’m curious as to why I’m not seeing low labor productivity growth being mentioned in these discussions.
dmk
May 5 2018 at 11:09am
Your WSJ is likely printed at/by your local newspaper’s printing plant, and delivered by said newspaper’s delivery person. So delivery problems may be due to local errors not directly controllable by WSJ.
David R Henderson
May 5 2018 at 11:10am
@Justin,
It was 62.1% in 2017, and averaged 62.5% from 2007-2017. It’s currently a little depressed compared to the Clinton years and most other business cycle peaks, but not drastically so. A 4.4% across the board compensation increase would raise it to 64.8%, the same as the average for Clinton’s 2nd term and the 1973 business cycle peak.
Interesting numbers, and thank you for that. Those aren’t the right numbers, though, to address the issue. The issue is hourly wages, not wages as a percent of national income.
@Justin,
Wider use of bonuses make sense. It avoids the sticky wage problem, and from my understanding have been used often in Japan in recent decades.
Correct. I wrote about this in my “The Myth of MITI” article in Fortune, August 8, 1983. But the issue Krugman and I are addressing is signing bonuses, not bonuses in general.
@Maximum Liberty,
Very interesting. Thanks.
@JoeR and sk,
Good points.
David R Henderson
May 5 2018 at 11:19am
@Thomas,
Why would you have a prior about the rate of wage increases?
Because of the tight labor market and all of the employers I hear about who can’t find employees.
I understand why Krugman would have one: It gives him another stick (or strawman) with which to beat free markets.
It’s true that he often looks for such sticks. I don’t think that’s what he’s doing here.
@TF,
Isn’t Krugman just talking about sticky wages?
No. If he were just talking about sticky wages, I wouldn’t have posted. He’s talking about an implication of sticky wages that I hadn’t seen talked about.
Sumner has been writing about this for years, decades even.
About sticky wages, yes. About this issue, I don’t think so. Scott can comment if I’m wrong.
@John P. Palmer and Michael Byrnes,
Thanks.
@jcooper,
The unemployment rate has recovered to 2000 levels but the workforce participation rate is still at Great Recession levels. It’s possible, I suppose, that the diminished labor pool is permanent, but I doubt it. The cause seems to have been largely “lack of jobs”, rather than age or other structural changes, and so should be reversible (and reversing, though there’s no visible uptick yet). I would expect we won’t get back to “mirror test” conditions until that missing part of the labor pool finishes returning from exile.
Interesting thoughts. I do think that, with many people, even relatively young people, receiving substantial disability payments from the feds, a few million of them won’t be back.
@Michael Sandifer,
I’m curious as to why I’m not seeing low labor productivity growth being mentioned in these discussions.
Good point.
@dmk,
Thanks.
Anonymous
May 5 2018 at 12:23pm
In violation of the normal rules again anonymous posting …
Let me answer Davids question about the bonus clawback.
– Yes they will take it out of your last pay if possible but usually that is too little money.
– As a manager, I’ve been instructed by HR to go to former employee’s house and demand payment on the spot. I’ve been instructed to resist leaving and make multiple attempts and phone calls until the money is paid back.
– They will eventually sue you for large enough amounts of money.
Signing bonus are often at least ten% of pay and I’ve seen as high as 100% of annual base pay.
Matthias Görgens
May 6 2018 at 7:29am
I just paid the claw back for a hiring bonus that was pro rated over two years (I stayed only one year).
They took from my last paycheck plus I had to wire the rest.
Matthias Görgens
May 6 2018 at 7:29am
I just paid the claw back for a hiring bonus that was pro rated over two years (I stayed only one year).
They took from my last paycheck plus I had to wire the rest.
David R Henderson
May 6 2018 at 12:19pm
Thanks, Anonymous and Matthias Gorgens. Interesting.
Testing
May 6 2018 at 10:27pm
Krugman for once making sense.
But would he also agree that creating protected classes makes those classes rationally more costly to hire?
Of course he wouldn’t, which is why he fails.
Mark Z
May 6 2018 at 10:40pm
Has anyone looked into whether wage stickiness, either upward or downward, is less acute in polities with hire/fire-at-will regimes, as opposed to more regulated ones?
John Fembup
May 6 2018 at 11:58pm
I’ve received two signing bonuses, one cash, the other stock options. Neither, so far as I was ever aware, was subject to clawback.
However, for another job I received a relocation expense payment in an advance lump sum with no subsequent true-up. That advance was subject to a 100% clawback if I left voluntarily within 2 years.
Justin
May 7 2018 at 9:36am
–“Interesting numbers, and thank you for that. Those aren’t the right numbers, though, to address the issue. The issue is hourly wages, not wages as a percent of national income.”–
It’s a different metric, yes, but if the ratio of labor compensation to national income is within historical norms then compensation is following the overall productivity of the economy and so I don’t see the problem.
1965 had roughly the same unemployment level and trajectory as 2017, and 2017’s ratio of labor compensation to national income (62.1%) was a bit higher than 1965 (61.5%). Yet I don’t see anyone puzzling over why wage growth in the several years leading up to 1965 was too slow.
Here’s another way to look at it if we want to focus on hourly metrics.
In the 5 years to 2007Q4, productivity growth averaged 2.4% and inflation (per the GDP deflator) averaged 2.7%, so nominal hourly productivity increased at a compounded 5.2% annualized pace. In the 5 years to 2018Q1, those numbers were 0.9%, 1.5% and 2.5%, respectively. And that’s really all we need to see to know why nominal wage growth isn’t as fast as last cycle. Low inflation and low productivity growth means low wage growth.
One other thing to note: bonuses won’t explain the slow growth of the series Krugman is using as average hourly earnings is a measure of total compensation per hour rather than just wages.
Comments are closed.