Wage flexibility among the unemployed doesn’t help (much)

This is a response to Tyler Cowen.  Take a simple example.  Nominal wages are fixed for the employed.  NGDP falls 5%, and 5% of workers are laid off.  Now the unemployed workers lower their wage demands by 20%.  Why not by even more?  Because of minimum wage laws, unemployment insurance, fear of loss of prestige, etc.

Suppose companies are not worried about workers making invidious comparisons (a big if, but I’ll grant this point to my opponents.)  In the best case scenario firms lay off 4% percent of their workers and hire back the 5% who are unemployed at the same total wage bill.  The excess unemployment is now 4% instead of 5%.  The total unemployment rate falls from 10% to 9% (assuming 5% is the natural rate.)  No big deal, we are still deep in recession.  Thus wage flexibility among the unemployed doesn’t really help very much.  If all employed workers accepted a 5% pay cut (or if the government ordered such a cut) and the Fed kept targeting inflation, we’d experience rapid economic growth.  BTW, I’m not advocating an incomes policy, I favor monetary stimulus.

Now you might claim that this process would keep repeating, and eventually we’d reach full employment.  But that would violate the assumption that wages are sticky for the employed.  I.e. could firms really say “You are all fired; now we’ll hire you back for 20% less?”  I suppose so, but then there’d be no wage stickiness for the employed, and we have lots of evidence that there is wage stickiness for the employed.  Indeed one piece of evidence (this pay increase dispersion graph from a Paul Krugman post, plus quotation) is so overwhelmingly persuasive that it should be considered, by itself, a 100% conclusive refutation of New Classical economics.  There is no possible New Classical explanation for this graph.  None.

So as I see it, we’re in a state of censored wage deflation; underlying forces are trying to make wages fall, but thanks to the combination of dispersion and rigidity actual average wages are still rising slowly.

And that’s all we need, empirical evidence that wages are sticky for the employed.  It is a bit puzzling that wages are so inflexible.  But remember that 2008-09 wasn’t the only adverse nominal shock.  We had another in 2009-10, and then another in 2010-11.  It’s possible that wages have now adjusted to the 2009 shock (although I doubt they’ve fully adjusted) but haven’t adjusted to the more recent adverse shocks.  And of course there are also lots of adverse supply shocks that are slowing the adjustment.  Adverse supply shocks partly caused by the adverse demand shock.  That’s why the sticky wage hypothesis seems implausible when looking at the world from a micro perspective—it is implausible from a micro perspective.  Lots of the problem is supply side, when viewed from a micro perspective.   But macro can’t be understood at a micro level.  For instance there is no way I can explain to someone at the micro level why if the Fed buys a $1,000,000 T-bill from me for $1,000,000 in cash it will create more AD.  But it will.


Tags:

 
 
 

67 Responses to “Wage flexibility among the unemployed doesn’t help (much)”

  1. Gravatar of Master of None Master of None
    15. August 2011 at 08:18

    Scott, I’m looking forward to your response to Krugman’s “The Texas Unmiracle” column today.

    If you read between the lines you can almost hear him arguing for lower real wages and higher nominal wages.

    Also, why aren’t we hearing calls for more immigration as a way to stimulate demand? That should have been a no-brainer for Krugman today and I can’t figure out why he neglected to make a jab at our destructive immigration policy.

  2. Gravatar of Nick Rowe Nick Rowe
    15. August 2011 at 08:24

    Scott: Are you (implicitly) assuming a downward-sloping rectangular hyperbola AD curve (the Fed holds NGDP constant when nominal wages and prices fall)? Because the shape of the AD curve wrt changes in nominal W (and hence P) is very much at issue in this debate. Paul Krugman, for example, would say that the AD curve is (approximately) vertical in a liquidity trap. And Brad DeLong, IIRC, was the author of a long-ago paper showing that price flexibility *could* be destabilising if it caused expected deflation. Roughly, the AD curve slopes the wrong way.

  3. Gravatar of TravisA TravisA
    15. August 2011 at 08:46

    Does anyone know why the graph shows zero density at zero percent change? Is this some artifact of their binning process?

  4. Gravatar of Greg Greg
    15. August 2011 at 08:53

    Scott,
    I highly recommend you read Alchian’s 1969 article “Information Costs, Pricing and Resource Unemployment”. If you can’t find it, maybe I can send it to you. In part of the article Alchian explains why wage stickiness/Keynes is consistent with rational utility maximization (or what you might call new classical econ.) He basically emphasizes information costs, and says that people don’t want wage cuts, because they don’t know at first whether AD is falling or only their own wage is falling due to chance. He also has a discussion of employers I think. It is a fascinating article about unemployment, and it kind of melds Chicago school thinking with Keynes.

  5. Gravatar of Greg Greg
    15. August 2011 at 08:54

    There’s more to the article than what I posted above. But I wanted to be brief.

  6. Gravatar of Ryan Murphy Ryan Murphy
    15. August 2011 at 09:00

    “But macro can’t be understood at a micro level. For instance there is no way I can explain to someone at the micro level why if the Fed buys a $1,000,000 T-bill from me for $1,000,000 in cash it will create more AD. But it will.”

    Are you saying that New Monetarism doesn’t have microfoundations?

    Why can’t you just say, “people want to hold more money, and until the supply of money expands or prices drop, people will sit on money instead of buying goods”?

  7. Gravatar of malavel malavel
    15. August 2011 at 09:02

    “could firms really say “You are all fired; now we’ll hire you back for 20% less?””

    Something similar sometimes happens in Sweden. A company in big trouble might ask its employees to accept a 10% wage reduction so that they wont have to fire anyone.

  8. Gravatar of Scott Sumner Scott Sumner
    15. August 2011 at 09:08

    Master of None, Google my Don’t Mess With Texas post from a few weeks back–I already replied.

    Good points about immigration.

    Nick, This was basically a reply to Tyler Cowen, and since we both assume downward sloping AD, it wasn’t an issue.

    BTW, I DEFINE AD as a hyperbola, so I would interpret the DeLong article as claiming that wage cuts shift AD to the left, but that’s a minor definitional issue.

    BTW, Here is my reply, and Summers and DeLong reply to my critique of their article:

    http://www.cato.org/pubs/journal/cj9n3/cj9n3-11.pdf

    http://www.cato.org/pubs/journal/cj9n3/cj9n3-12.pdf

    In the abstract they say my critique “has some force” which I think is a gross understatement. In my view my critique destroyed their argument, but they ignored the evidence I presented.

    The Fed is either targeting NGDP or inflation. Either way DeLong and Krugman are wrong.

    Travis, I wondered about that too. My hunch is they simply looked at all occasions where actual wage changes occurred, not at specified spans of time.

  9. Gravatar of Scott Sumner Scott Sumner
    15. August 2011 at 09:09

    Malavel, Yes, but that is wage flexibility among the employed–and Cowen stipulated that employed workers have relatively inflexible wages.

  10. Gravatar of Scott Sumner Scott Sumner
    15. August 2011 at 09:18

    Greg, But even if that were true, the graph I presented would still represent a decisive refutation of New Classical economics. Adding uncertainty and lags does not explain a discontinuity at zero, or at any other point in the wage change distribution. Only money illusion does.

    Ryan, Why should swapping one asset for another affect your purchases? People aren’t “not buying” because they have no cash; anyone can get cash any time they want. They aren’t buying because they have no jobs, or wealth. But swapping cash for bonds doesn’t directly affect wealth, so why should people buy more stuff?

  11. Gravatar of Benjamin Cole Benjamin Cole
    15. August 2011 at 09:18

    Excellent blogging.

    BTW, my brother worked for the American branch of a Swedish industrial pump company. When the economy slowed, in Sweden everyone took a 20 percent cut in hours/pay. Here they laid off 20 percent of the workforce (including my hapless brother).

  12. Gravatar of Bryan Willman Bryan Willman
    15. August 2011 at 09:25

    So that’s all fine, but it misses a point I’ve been harping on – wages are the claims an employee earns for working, which that person uses to settle their obligations.

    If wages fall (but you are still employed), essentially everybody will default on or stop renewing some obligation.

    What’s more, once again, if my company only needs 3 people to meet demand, reducing the cost of paying them by 33% does NOT want me to have a 4th person standing around watching. For a whole host of human dynamics reasons you want as few people as possible.

  13. Gravatar of Morgan Warstler Morgan Warstler
    15. August 2011 at 09:49

    Wage flexibility works just fine is the government policy does not support wage floors.

    Why you think we’re going to assume that government policy doesn’t have to change before we eve consider NGDP targeting continues to amaze me.

    I don’t care if you LIKE it, you refuse to imagine that you might have to ACCEPT it.

    The only time the discussion gets interesting or real is when we think about what can be done by you (or the Fed) to quickly FORCE the Administration into fiscal policy favored by the right.

    CLINTON BENT to Greenspan’s will. So we have a recent historical fact staring us in the face. Bending Democrat Presidents is easy.

    And all recent political activity says Obama is the most likely to bend.

    So the shortest distance from here to there is: how do we (Sumner / the Fed) make Obama bend. He is the weakest link. You should be treating him like it.

  14. Gravatar of Morgan Warstler Morgan Warstler
    15. August 2011 at 09:54

    DeKrugman’s post on Texas 100% gives up the goat:

    He cannot imagine that we’d simply relax regulations to increase jobs. He cannot imagine that we don’t care if all the unemployed just get minimum wage jobs.

    And that’s why he’s a tiny minority. Normal people are more than willing to cut back rules by the FDA, EPA, FCC, etc. etc. to make things riskier, but easier to create jobs.

    We don’t care if the unemployed only make minimum wage, we just want them working.

  15. Gravatar of Nick Rowe Nick Rowe
    15. August 2011 at 09:57

    Scott: Neat! I didn’t know you had been in the thick of that debate.

    (So that’s the origin of Laidler’s talking about the Fisherian vs the Phelpsian interpretation of the Phillips Curve, and how Friedman sat on the fence between the two.)

  16. Gravatar of Ryan Murphy Ryan Murphy
    15. August 2011 at 10:09

    Scott, they are forestalling investment and current consumption to build up cash reserves. T-bills are a near substitute for cash, especially for banks with IOR, but it’s not the same thing. I’m just giving the standard monetary disequilibrium story, which was always the “microfoundations” in the back of my mind when reading your justifications of nominal income targeting. I’m a bit perplexed if you are actually rejecting the need to have any microfoundations beyond money illusion + monetary policy causes inflation => more employment.

  17. Gravatar of dwb dwb
    15. August 2011 at 10:18

    nope, spend some time with HR in a big company and you pretty much get why wages are sticky for employed workers. One top goal of HR is “talent retention” so they do studies and benchmarking of pay to other competitors. Those studies often come from one of the oligopolistic consulting firms like Hewitt, which fundamentally rely on the same data set. I think of it somewhat in a game-theoretic framework – for the employed workers, HR wants to make sure that “talent” gets paid at or above some level to make sure that they do not jump to the competition. There is also an element of signaling: at performance review time, your “raise” is driven your performance rating. If total compensation (bonus+raise +…) is less than last year its a sign you did worse performance-wise, or that the company is not doing well. Generally, senior managers, being corporate perma-bulls, are loathe to admit the company is not doing well. Even if revenue is declining, they did better on some metric (market share, something…)

    And so, rather than give everyone, say, a 10%-15% pay cut and retain the entire staff, they reduce the wage growth to 0-2% and cut staff. (again: think of the game-theoretic aspect, if company A cuts 10% of staff and maintains wages for the remaining staff, and company B cuts everyone’s wage 10% but keeps headcount, then when there are job openings at company A, the talent from B will jump to A, and B will lose talent). A comment I frequently hear from my bosses and other top managers is to cut staff somewhat beyond where you want, leaving some openings to offer to the top talent that leave other firms ….

    Even if you had perfect information about what other firms would do, if you knew some other firm was going to cut everyone 10% and keep headcount, it would be to your advantage to “cut the chaff” from staff and try to “steal” the most productive employees, thereby raising your average productivity well above theirs.

  18. Gravatar of William William
    15. August 2011 at 10:25

    Greg, But even if that were true, the graph I presented would still represent a decisive refutation of New Classical economics. Adding uncertainty and lags does not explain a discontinuity at zero, or at any other point in the wage change distribution. Only money illusion does.

    I was going to recommend the same Alchian article Greg did, but since I’m too late, I’ll just jump in and mildly disagree with you.

    Lots of prices don’t constantly adjust to match supply and demand. The lowest-level iPad has stayed at the price of $499 for the last 18 months. I don’t expect you’ll say that that’s because of money illusion, too. Fixed prices actually can serve a neoclassical purpose.

    No one really knows whether his current job is his best possible opportunity. The graph is (I assume) showing what happens to the subset of workers who decide to stay at their current job. What’s so irrational about adopting a rule that if your pay raise is at or below 0%, you will quit your current job and employ yourself in the search for job information? Alchian writes that job search is easier when unemployed than employed; it’s the same principle that keeps used cars on lots and apartments vacant, rather than being rented at a low price until a suitable buyer can be found.

    I’m not saying the graph isn’t a great graph, but it’s not quite as much of a slam-dunk as one might think.

  19. Gravatar of Scott Sumner Scott Sumner
    15. August 2011 at 10:51

    Ben, Good point, I think Germany has a similar policy, which explains their lower unemployment rate.

    Bryan, That’s one reason I favor boosting NGDP, rather than cutting wages.

    Morgan, I agree that Krugman is wrong about Texas.

    Nick, Yes, I used to be a Fisherian regarding the Phillips Curve. I suppose I still am, but only if you replace inflation with NGDP growth on the vertical axis. They really got Fisher wrong, however, no way would he have favored artificially raising wages–he thought the NIRA was a big mistake. He thought unemployment was caused by a fall in the equilibrium price level, not the actual price level.

    Ryan, I’m not saying there are no microfoundations for monetarism; it’s a simple application of S&D theory. Supply more of X, and the value of X falls. I’m saying the impact on AD is hard to explain at the micro level.

    dwb, Your explanation may be right, but it explains real wage rigidity, not nominal wage rigidity. The mystery in macro is why nominal wages are rigid. Indeed the mystery is why we even care about nominal wages.

    William, You are missing my point. The point of the graph is not that wages don’t change, every single observation on the graph is a change in nominal wages. The mystery is why tiny increases are much more common than tiny decreases. Why is the discontinuity at zero, and not at the rate of inflation? (Which is where it might be if real wages were sticky.) The graph shows nominal wage rigidity, nominal money illusion.

  20. Gravatar of Benjamin Cole Benjamin Cole
    15. August 2011 at 10:57

    OT, but amazing.

    Here is an OECD chart that says the typical Greek (remember, those self-absorbed loafers) works 2109 hours a year, while the typical upright West German puts in 1409!

    You know, if I worked 600 hours a year more than my German counterpart, but I was told I had to endure “austerity” I too might well be inclined to tell the “establishment” to shove it.

    http://stats.oecd.org/Index.aspx?DataSetCode=ANHRS

  21. Gravatar of cassander cassander
    15. August 2011 at 11:34

    I’ve always felt that the best arguments against deflationary monetary policy were never economic, but psychological. A constant, predictable level of deflation isn’t bad for the economy, but people HATE it, and agitate politically to change it, which invariably leads to all sorts of idiotic government interventions in the economy.

  22. Gravatar of William William
    15. August 2011 at 11:45

    Scott, you are missing my point, too, which is that the graph includes workers implicitly selected on the criterion that they stayed at the same job from one year to the next. A worker offered a pay cut knows that his current position has gotten worse but does not (instantly, costlessly) know whether his overall opportunities have gotten worse, and whether he can improve his situation by specializing in job search (i.e. being unemployed).

    You are saying there should not be a discontinuity, but I am saying that a worker does have to make a discrete decision, whether to retain his current job or search for more job information. You can say that the location of a worker’s “cut-off” for deciding to quit should not be discontinuous at zero, either, to which I would say, “OK, good point.” But I think this graph overstates the case by ignoring the selection effect.

  23. Gravatar of TravisA TravisA
    15. August 2011 at 11:46

    I think there are microfoundations for Fed purchases affecting AD. When someone swaps $1 million in Ts for cash, they are expressing a preference for cash, or something else that cash can buy — either other financial assets or ‘real things’. If it’s the latter, then AD will increase on average, since there will be *someone* who prefers to buy real things rather than hold cash. This only holds if it is the Fed that is buying Ts and not some private party who needs to sell something or use cash to make the purchase.

  24. Gravatar of Tyler Cowen Tyler Cowen
    15. August 2011 at 11:49

    If I understand you correctly (and I am not sure I do), you are taking NGDP to be fixed. But if an unemployed worker lowers his wage and takes a job, NGDP goes up. I thus don’t see that NGDP being a certain size prevents this from happening, unless one is simply assuming it won’t happen. Another way of putting this is that changes in velocity can accommodate available gains from trade.

  25. Gravatar of David Pearson David Pearson
    15. August 2011 at 11:49

    Scott,

    Sorry to beat a dead horse, but here goes. Reading your comments on Friedman, I couldn’t help but think that Friedman say high real rates as evidence of tight policy, not low nominal ones. This excerpt from a 2000 interview with David Laidler supports that view:

    “The Japanese bank has supposedly had, until very recently, a zero interest rate policy. Yet that zero interest rate policy was evidence of an extremely tight monetary policy. Essentially, you had deflation. The real interest rate was positive; it was not negative. What you needed in Japan was more liquidity.”

    Here is another one, from March 2001, on money growth:

    “He noted that in the prior two months, M2 grew at an 11 percent annual rate…But it was a source of concern to him that in trying to thwart a recession, Greenspan was planting the seeds for higher inflation next year. Friedman pointed out that inflation already has risen from 2 percent in 1999 to 3 percent last year, and it could reach 5 percent next year. That would be the largest jump in prices in more than a decade.”

    Our real interest rates are negative, not positive. Our money growth is 10%, not negative. Our inflation rate is has gone, “from 2% to 3%”. In 2011, MWWMFS? (What would Milton Friedman say?).

  26. Gravatar of David Pearson David Pearson
    15. August 2011 at 11:51

    Sorry — WWMFS? (I had an extra M above).

  27. Gravatar of Tyler Cowen Tyler Cowen
    15. August 2011 at 11:55

    p.s. I will commend you for at least attempting an explanation of nominal wage stickiness for the unemployed. No other respondent has done that, at least not so far.

  28. Gravatar of TheNumeraire TheNumeraire
    15. August 2011 at 11:57

    Benjamin,

    Your interpretation of OECD hours worked data is all wrong;

    – in poorer countries hours worked tend to be high, regardless of whether marginal tax rates are low (Mexico, South Korea) or high (Greece, Italy, Portugal) Poorer people value money over leisure.

    – countries like Greece and Italy have fewer two-income households than Germany or the U.S., which eliminates most of the hours disparity when measured by household total work hours
    – Greece has an abnomally high percentage of “self-employed” — these self-employed are not entrepreneurs (as some economists and journalists contend), but rather individuals triying to evade the high marginal tax rates. Greece has been criticized for maintaining a very large employer payroll tax, which is a disincentive to growing a firm and adding new hires and an incentive to self-employment.

    Greece should tell the establishment to shove it — the establishment being the IMF, whose program the Greeks have implemented. Easier money from Brussels or a new Greek drachma will not solve much, if any of the Greek problems. Observe what happened to Hungary in 2008-2009 — it had easier monetary policy than the eurozone but adopted an IMF austerity program. The economy contracted until the newly elected Orban government implented a flat tax regime with sharply lower tax rates, the opposite of what the IMF suggested.

  29. Gravatar of TravisA TravisA
    15. August 2011 at 11:57

    It would be interesting to hear from the futures exchanges why they think there hasn’t been independent private demand for an NGDP futures market. Maybe it isn’t specific enough for hedging. I wonder if it would be better to have NGDP futures for specific sectors of the economy and then the Fed could add up future prices for those sectors to figure out what needs to be done.

  30. Gravatar of DanC DanC
    15. August 2011 at 12:10

    You have a demand and a supply side.

    The wage firms are willing to pay the unemployed must be lower then the wage the unemployed need to give up leisure.

    I think it is like the used car market. Buyers assume that the used car is risky and push the price down. Sellers of used cars see a low price for used cars and withhold better used cars from the market. The markets refuse to clear, in part, because of the cost of information and implied risk.

    In the labor market, employers will only higher workers with extremely strong credentials or at very low wages – because of perceived risk. For example, some of the long term unemployed are older. If an employer hires a bad older worker, the employer risks litigation if they must later fire the worker.

    If an employer is uncertain of the economy they may wait for the perfect hire, which is a function of both wage and risk. Add in the risk of future taxes, government mandates, and uncertain prospects you only hire risk free candidates and actual wage rates can stay rather constant.

    I have seen it reported that many employers prefer to hire workers who are currently employed. (A signal of quality.) The only way they can hire these workers is to offer them something that their current employer doesn’t. (Usually higher pay.)

    I think you get get the above chart from a great deal of uncertainty about the future and a mark that fails to clear due to asymmetric information.

  31. Gravatar of TheNumeraire TheNumeraire
    15. August 2011 at 12:23

    David Pearson,

    Friedman’s reading of the monetary situation from 1999-2002 is exactly backwards. The increase in inflation he refers to was entirely transitory and supply-side induced because of oil prices. Oil prices rose from $10 to almost $40 in less than 24 months from 1999-2000 because of relative supply shortage. In 1999, in the aftermath of the Asian crisis/depression and the strong dollar policy, oil drilling and exploration effective ground to a halt. As the global economy recovered in 1999, the oil price went into backwardation, indicating necessity for immediate supply — with spot price trading at almost $40 in 2000, long-term futures contracts (3-5 yrs. out) were trading in the low $20’s.

    Actual real interest rates during 2000-2002 were higher than Friedman indicates. Breakeven rates on TIPS were trending lower and the gold price was comatose. Sumnerian analysis of NGDP would have suggested money was tight until 2003. By mid-2002, economists became widely worried about deflation which prompted Bernanke’s famous “helicopter” speech.

    Friedman made the same mistake in 1983, focusing on M2 and becoming a very vocal inflation hawk. That’s why I question Prof. Sumner’s assertion that Friedman would instantly recognize today’s nominal bond yields as indicative of tight money. I think Friedman would be torn between Metzler-type arguments and Sumner-style thinking.

  32. Gravatar of Morgan Warstler Morgan Warstler
    15. August 2011 at 12:28

    DanC, this is what I mean….

    It is only AFTER we set government policy so that there are no “external costs” to the hire, that we can even really say wages are sticky.

    If all government aid to an individual is conditioned on their man weeks being auctioned starting at $1 per hour.

    And if companies are able to fire any loser at any time.

    We simply won’t have unemployment.

    If there is a legitimate policy change in front of us that ends unemployment WHY do we let a bunch of egghead economists sit around yammering about wage stickiness?

  33. Gravatar of Morgan Warstler Morgan Warstler
    15. August 2011 at 12:36

    TheNumeraire,

    Friedman would never advocate Sumner style thinking until Obamacare had been repealed, and the public employees were treated like they were in in the mid 1990’s.

    Metzler is exactly where Friedman would be. Any real free market based economist looks at the current structure of things and feels queasy about using monetary.

    If there would be any change about him, Friedman would likely be very impressed with how far right Mundell’s Euro work has driven Greece, Italy, etc.

    Italy just passed a balanced budget amendment fer christ’s sake.

  34. Gravatar of Sticky wages and unemployment — Marginal Revolution Sticky wages and unemployment — Marginal Revolution
    15. August 2011 at 12:51

    […] of the unemployed just isn’t enough. Scott Sumner wasn’t at lunch but nails it with a simple example: Nominal wages are fixed for the employed. NGDP falls 5%, and 5% of workers are laid off. Now the […]

  35. Gravatar of Neal Neal
    15. August 2011 at 13:10

    Scott, could you explain this further?

    “But macro can’t be understood at a micro level. For instance there is no way I can explain to someone at the micro level why if the Fed buys a $1,000,000 T-bill from me for $1,000,000 in cash it will create more AD. But it will.”

  36. Gravatar of Scott Sumner Scott Sumner
    15. August 2011 at 13:12

    I’m running late so I’ll respond to Tyler’s comment now, and the others later.

    Tyler, I’m not all all discounting the possibility that wage flexibility by the unemployed can raise velocity–I’ll even provide a mechanism–higher real interest rates as employment expands. My argument is more subtle–that the actual change in NGDP is a sufficient statistic for understanding the net effect of wage flexibility on NGDP, PLUS monetary policy on NGDP. Remember that NGDP is what it is. If it falls 5% it might be because the Fed caused it to fall 5%, or because worker wage flexibility caused it to rise 5% and the Fed caused it to fall 10%. Whatever the combination of factors, the actual NGDP growth is a sufficient statistic for knowing how much revenue business has for employing workers. And no assumption about the effect of worker wage flexibility on velocity changes that fact. If you are right that worker wage flexibility should (and has?) raised NGDP through some channel, that just means monetary policy was even tighter than we thought.

    Here’s another way of answering your question. I’m far less confident about my overall wage stickiness assumption, than I am about the subsidiary assumption that if wages are highly sticky for those with jobs, then wage flexibility among the unemployed doesn’t actually create many new jobs (in aggregate) when NGDP has fallen far below trend. I’m 60% confident that wage stickiness explains more than half our excess unemployment, and 95% confident about the second assertion.

    However, I do give serious consideration to one thing you said–that wage stickiness might have caused the initial problem, but other factors explain long term unemployment. My response would be that more NGDP can reduce unemployment via the real wage channel, and then this will lead Congress to cut UI back to 26 weeks, which reduces the natural rate of unemployment, especially among the less skilled workers. It will also make the minimum wage a bit less binding.

  37. Gravatar of John John
    15. August 2011 at 13:25

    Scott,

    My eyes kind of jumped out at this quote: “For instance there is no way I can explain to someone at the micro level why if the Fed buys a $1,000,000 T-bill from me for $1,000,000 in cash it will create more AD. But it will.”

    How can you be so sure? The Fed has been buying T-Bills extensively for the past 3 years and yet it isn’t a slam dunk case that AD is up over what it would have been otherwise. What has happened when the Fed bought all those T-bills was the the people they bought them from held cash, essentially not boosting AD. Maybe people who own long term treasury bills are hoping to save, in which case, they’d just save the currency from selling the T-bills, especially since T-bills are fetching quite a bit of money right now.

  38. Gravatar of DanC DanC
    15. August 2011 at 13:54

    Morgan

    I think wages are sticky and that government policy can just make them stickier.

    I think Sumner and Friedman would get along fine. Not sure they would agree in full and Friedman might be closer to John Taylor. But I’m not certain – other then that Friedman would make the answer seem clear and obvious.

    What is a substitute for unemployed workers in the United States? International workers. Unskilled workers in the US may need to adjust their wages to international scales. At that point two income families become single earner families as the value of domestic production (or government subsidy) may exceed potential after tax wages.

    So I see three problems. A poorly working labor market failing to clear (like the used car market). Two, uncertainty driven by government policy. And three, increased international competition

  39. Gravatar of Silas Barta Silas Barta
    15. August 2011 at 14:27

    Does that graph measure total compensation, or just wage compensation? Does it count health care paid for, 401k match, intagibles that affect employee comfort (and are hard to read off a survey), etc etc etc? Because those forms of compensation are trivially easy to cut and happen all the time, making me suspicious that the chart really shows sticky wages.

    As some of the unwashed masses, and you might learn about how their compensation was indirectly cut. They’ll laugh at the idea that their pay + benefits can’t go down.

  40. Gravatar of Trumwill Trumwill
    15. August 2011 at 14:33

    I can’t seem to find the Don’t Mess With Texas post. I searched “Don’t Mess With Texas site:themoneyillusion.com and got only this post. I removed the quotes and got a whole bunch of pages, but none with that title. I also looked for combinations with your name. Help?

  41. Gravatar of Silas Barta Silas Barta
    15. August 2011 at 14:34

    @TravisA: Does anyone know why the graph shows zero density at zero percent change? Is this some artifact of their binning process?

    Nope. It’s due to how no one’s wage ever stays constant. Over any period. Ever. ;-P

  42. Gravatar of bill woolsey bill woolsey
    15. August 2011 at 15:42

    Scott:

    Your story involved firms trying to maximize employment subject to the constraint that they keep their wage bill constant. What?

    Firing current workers and replacing them with new workers that are paid less? Do you really think that will make the remaining workers less unhappy than just cutting their pay?

    What is supposed to happen is that someone buys a bankrupt firm’s assets, hires unemployed workers at lower wages, and undercuts the prices of other firms. Firms that try to keep wages and prices high are driven out of business.

    The price level falls and real output rises at a given level of Nominal GDP.

    Cowen seems to be thinking about hiring addional new workers at lower wages, so that marginal costs are lower, and so firms can lower prices and sell more.

    It seems to me that it is simple. Get the 10% of workers who are not employed to work for nothing, increase output by 10%, and lower prices by 10%, revenues are the same and costs are the same. Average costs are 90% of current levels, like prices.

    Or, get half of those workers to come work for nothing. Then output is 5% higher. Prices are 5% lower. Revenues are the same, and costs are the same. Average costs (of wages) are 95% of their current level, as are prices.

    So, we know how much we need to cut the wages of the added workers. Zero.

    Or, of course, we follow the usual process of some firms hiring the unemployed workers, driving the high cost firms out of business, hiring their newly unemployed workers at lower wages, etc.

    I think if people really were certain that the new equilibrium level of wages and prices were 10% lower, something like this would happen.

    The problem is that everyone expects that the current conditions are temporary and that the equilibrium value of prices and wages will soon surpass current levels.

    Why go through this price war process when it is all for at best a temporary gain?

  43. Gravatar of Scott Sumner Scott Sumner
    15. August 2011 at 15:47

    Ben, Yes, the idea of the hard-working Germans is a complete myth. They tend to be highly disciplined workers, but do take long vacations.

    Cassander, Good point.

    William. It’s even worse. The discontinuity is at 0% whether the trend rate of inflation is 2% or 4%. There is no rational explanation for that fact. Even if, as you say, workers might look for another job if they aren’t happy with their pay. They should care about the real pay, not the nominal pay.

    Travis, What if the Fed pays you with a check, and you deposit the check in the bank after you receive it? Why would you buy more goods?

    David, I don’t understand your question; where did Friedman say Japan had high real interest rates? Yes, he pointed out they had deflation, and thus real interest rates weren’t as low as they looked–I said the same thing back in the late 1990s.

    But even so, real rates were less than 1% during most recent years in Japan.

    During March 2001 the economy had been booming for 10 years–quite different from today.

    See also my second reply to Numeraire.

    The Numeraire, Greece isn’t really a poor country, or at least wasn’t until the crisis hit. Per capita GDP (PPP) was close to $30,000 in 2008, vs. about $35,000 for Germany. Nothing like Mexico.

    DanC, You said;

    “I think you get get the above chart from a great deal of uncertainty about the future and a mark that fails to clear due to asymmetric information.”

    Sorry, but like everyone else you are confusing real and nominal wages. Your explanation pertains to real wages, not nominal wages.

    The Numeraire, Good reading of 1999-2002. My strong feeling is that Friedman would have viewed IOR as a huge mistake in 2008. It would have been the 1937 RR increase all over again. He would have used that as an “excuse” to ignore the big increase in the monetary base. And he would also have talked about how a banking crisis creates a demand for liquidity. I’m not pulling these ideas out of thin air, they are based on my reading of A Monetary History of the US. He would have recognized the similarities with the 1930s. Then add on the huge increase in real interest rates in late 2008, plus the negative TIPS spreads, and it’s obvious to me he would have blamed the Fed for the recession. It’s also obvious he would have opposed fiscal stimulus, arguing monetary stimulus was better.

    I’m less sure about today. It’s possible he might argue that the natural rate of unemployment rose due to extended UI, etc.

    Neal, I mean that the average person thinks “I gave up one million dollar asset for another. I’m no richer, why should I buy more.”

    John, The Fed hasn’t been buying the T-bills with cash. And in any case, my example referred to normal times, when rates are positive.

    Silas, I don’t know but it hardly matters. New Classical econ doesn’t predict that for either total comp, or wages only.

  44. Gravatar of Scott Sumner Scott Sumner
    15. August 2011 at 15:52

    Trumwill, Here it is:

    http://www.themoneyillusion.com/?p=10093

    Bill, I agree it was far-fetched, but I was trying to make the assumption most favorable to Tyler’s post. In the real world workers would be furious if a firm did that.

    And it’s very costly for modern firms to churn workers–they are no longer just cogs on an assembly line.

  45. Gravatar of Neal Neal
    15. August 2011 at 16:15

    Okay, I see. Thanks. I had thought you were taking a jab at micro-foundations of macroeconomics, but a philosophical point like that didn’t seem to fit the context of the post.

  46. Gravatar of TallDave TallDave
    15. August 2011 at 18:35

    I’d be very cautious of sweeping statements based on that graph as it only covers the late 1990s, a quite different and somewhat unique time frame. Krugman cherrypicks these things now and then, as his own ombudsman pointed out. Does the shape of this curve hold up into TGD? Maybe it does, but I’d want to know for sure.

  47. Gravatar of Silas Barta Silas Barta
    15. August 2011 at 19:13

    @Scott_Sumner: New Classical econ doesn’t predict that for either total comp, or wages only.

    You mean New Classical predicts that wages-only *definitely won’t* be like the chart, or you mean New Classical is *neutral* with respect to which aspect of compensation? The former would involve a strange reading of NC.

  48. Gravatar of james in london james in london
    15. August 2011 at 21:34

    Low to average quality micro would seem to be the judgement on your piece here. Far worse is that it is one helluva poor foundation on which to base one helluva huge macro-economic experiment.

  49. Gravatar of james in london james in london
    15. August 2011 at 21:38

    It is especially a big experiment when M2 SA is running at 44% growth on an annualised basis in the 6 weeks to Aug 1st, and that latest loan officer survey. Meltzer is dead right to be concerned.

  50. Gravatar of james in london james in london
    15. August 2011 at 21:44

    You said “…and then this will lead Congress to cut UI back to 26 weeks, which reduces the natural rate of unemployment, especially among the less skilled workers. It will also make the minimum wage a bit less binding.”

    Isn’t that just a case of wishful thinking? What if Congress didn’t act this “rationally” (in your view) on UI or increased the minimum wage yet further? What then? We need a social contract for fiscal prudence and against further socialistic reforms in the follow up to any pick-up in NGDP growth otherwise the risk of your experiment failingis very high. You are too trusting.

  51. Gravatar of TravisA TravisA
    16. August 2011 at 05:16

    Travis, What if the Fed pays you with a check, and you deposit the check in the bank after you receive it? Why would you buy more goods?

    Basically, the relative price of bonds (or the expected price of bonds in the future) is changed by the Fed action. So some people will want to move from bonds to purchase good X, where good X seems relatively more attractive than bonds OR cash. It’s very common for portfolio managers to change their asset allocation from bonds to stocks, because stocks seem relatively more attractive. They sell the bonds for cash and then exchange the cash for stock. Cash is simply an intermediate point in the transaction process.

  52. Gravatar of Robert Bell Robert Bell
    16. August 2011 at 06:09

    Scott: Regarding your comment:

    “For instance there is no way I can explain to someone at the micro level why if the Fed buys a $1,000,000 T-bill from me for $1,000,000 in cash it will create more AD. But it will.”

    This makes the macro view sound a little bit like deus ex machina, whereas presumably there is a parsimonious explanation of what, if anything, is special about the T-Bill, the cash, the Fed, the Treasury etc that explains the additional AD. I.e. what kinds of transactions boost AD and what kinds of transactions don’t.

  53. Gravatar of Scott Sumner Scott Sumner
    16. August 2011 at 06:15

    TallDavem, I’ve seen similar graphs for other time periods–it’s a well established proposition.

    Silas, I’m saying this data (either way) violates the predictions of New Classical econ.

    James. What experiment? I am calling for a continuation of the Fed’s policy of 1990-2007, stable 5% NGDP growth. Meltzer supports the radical experiment of slowing NGDP growth sharply. If my policy had been followed, the monetary base today would be less than 1/2 its current level.

    You said;

    “Isn’t that just a case of wishful thinking?”

    No, Congress always cuts UI back to 26 weeks after a recession, always, every time. With the GOP controlling the House, it’s a lock.

    Travis, What if interest rates don’t change (they might not change) or what if interest rates rise with easy money? Then how do you get the extra AD?

  54. Gravatar of Scott Sumner Scott Sumner
    16. August 2011 at 06:25

    Robert, The transmission mechanism involves the fallacy of composition, hard to see at the individual level. Let’s try this:

    1. The extra money raises future prices by a tiny amount (by one millionth.)

    2. Higher future prices raise current prices by an even smaller amount.

    3. Higher current prices (and sticky wages), lead to tiny increases in each firm’s output (say by one penny)
    This leads to a few more jobs some in the economy, but no one knows where.

    4. The few extra jobs leads to more shopping by the newly employed workers. More AD.

    I suppose that’s a micro explanation, but I’d wager it’s hard for most to follow. And even with all that work, I still haven’t explained why the extra money raises future expected prices by a tiny amount. That involves the hot potato effect, which involves the fallacy of composition.

  55. Gravatar of James in London James in London
    16. August 2011 at 07:08

    Is the average American better off than she was in 1900? By how much?

  56. Gravatar of Silas Barta Silas Barta
    16. August 2011 at 07:10

    @Scott_Sumner: Silas, I’m saying this data (either way) violates the predictions of New Classical econ.

    Then you’d be wrong. New Classical does not require that reductions in worker compensation are flexible downward specifically in the money-component; it only requires that total compensation be downward-flexible. Certainly it allows for different marginal benefit from cutting different kinds of compensation?

    And if you look at the fortunes of workers that held their jobs through the past few years, rather than ultra-aggregated data, I think you will get an earful from such “constant-wage” workers.

    So, I still think you’re confusing, “NC concentrates high probability mass on the money component of compensation being more downwardly-flexible than the other parts” and “NC is agnostic on which form a reduction in compensation will take”.

  57. Gravatar of James in London James in London
    16. August 2011 at 07:11

    I mean 1990, even I think she is better off than 1900!

  58. Gravatar of TravisA TravisA
    16. August 2011 at 07:40

    Scott, if unbeknown to everyone I had a printing press in my basement and I used it to print dollars anonymously to buy treasuries, my guess is that the price of treasuries would always rise initially, at least until I stopped buying. As the cash circulated through the economy resulting in higher prices for most goods, the price of treasuries might fall back (with a lag), depending on whether credit was previously tight, adequate or loose.

    When the *Fed* makes purchases, anonymity is no longer possible. Market players look at the overall size of the Fed’s purchases and make judgements about what the state of the world will look like several steps ahead when the cash has circulated through the economy. If the Fed announces a $1 trillion debt purchase, market players might think, ‘Given the current state of the economy, that increase in cash will result in R% more purchases, that will result in prices rising X%, a greater demand for credit of Y%, a greater supply of credit of Z%. On balance, the price of credit will rise/fall K%.’ No one person knows enough to make a judgement about all the variables, but in a distributed fashion, the (imperfect) extrapolations are made.

    (Obviously, market players try to look several steps ahead in order to take advantage of current prices. As an extreme example, if my neighbor knew I had a printing press in my basement and I told him that I was going to buy $10 trillion worth of treasuries, he would probably think, “Hmmmm, maybe I should use my current cash now to buy some real goods ahead of time, before everyone else has the opportunity to spend their increased cash balances.” Similar types of decisions are made by traders and cause prices to change greatly at Fed announcements even before the Fed has made any purchases. Essentially, there is a profit motivated rush by market players to anticipate what the state of the world will be like in the future.)

  59. Gravatar of Morgan Warstler Morgan Warstler
    16. August 2011 at 08:18

    DanC,

    “So I see three problems. A poorly working labor market failing to clear (like the used car market). Two, uncertainty driven by government policy. And three, increased international competition.”

    In short bits, my Guaranteed Income plan is this:

    1. Use clones of Paypal / Ebay software.
    2. EVERYONE unemployed who wants to work registers and automatically starts receiving at least $200 in their acct every week. It could be $240, etc.
    3. EVERYONE who wants to register must create a real name CV and their work weeks will be auctioned off starting at $1 per hour.

    Your wife registers and tomorrow the neighbor lady has her cleaning her house for $40 per week, another greedy neighbor lady who saw your wife outside gardening, goes in and bids $80 ($2 per week), suddenly client #1 has to bid more or loses your wife.

    All of the concerns ANYONE could raise have been cleared by the tech industry long ago. Cred / Reputation, data mining, bad buyer / seller, before after pictures, video, all of it… so please skip the smallish problems in your reply.

    1. The market clears instantly – ask yourself one question: for $40 what would you have someone do for you all week long.

    2. Uncertainty: this plan instantly tells America we are in the business of business, there is no free lunch, all labor is for profit. SMB markets will soar.

    3. See #1, what would in fact happen is that you’d never hear someone in India answer the phone again. You’d likely see a good chunk of small batch manufacturing move back here.

    You can see a sample graduated pay schedule here:

    http://biggovernment.com/mwarstler/2011/01/31/guaranteed-income-part-ii-a-real-end-to-illegal-immigration/

    Something like this is coming. There is no way around it.

  60. Gravatar of Robert Bell Robert Bell
    16. August 2011 at 09:49

    Scott: Thanks for your reply. While steps 2-4 are pretty clear to me, it is step 1 that really seems to involve the “miracle of composition” as it were. E.g. why does the t-bill seller move from indifference/equilibrium to deciding to sell, and (following TravisA’s discussion of basement printing presses) when do sellers of products that the t-bill seller might buy realizes that there is a tiny bit of inflation and decide to raise their prices etc?

  61. Gravatar of TravisA TravisA
    16. August 2011 at 10:20

    @Robert Bell: Everyone has some price that they are willing to sell anything they own. If you own a t-bill that gives you $1000 in 10 days, would you accept $1000 + $x for that t-bill now, where $x ranges from $1 to infinity? At some $x, you would. There’s some price that everyone would sell.

    Sellers of products will raise their prices when doing so increases their profits. If I have a shop and my products are being bought as fast as I can put the products on the shelves, I might think, “Hmmmm, maybe if I increase my price X%, volume will fall less than X% and I will make an increased profit. Let’s try that.”

    I think it’s helpful to think about the extremes. Then realize that at the margin, the same process is going on.

  62. Gravatar of Martin Martin
    16. August 2011 at 11:09

    “by itself, a 100% conclusive refutation of New Classical economics. There is no possible New Classical explanation for this graph. None”

    Scott,

    1. How about an unanticipated nominal shock and expectations based on previous nominal growth and expectations that are slow to adapt?

    2. If a wage cut would restore growth, and wage contracts are based on expectations of future growth, why would you accept a wage cut? The result of which will be growth?

  63. Gravatar of Scott Sumner on sticky wages for the unemployed — Marginal Revolution Scott Sumner on sticky wages for the unemployed — Marginal Revolution
    16. August 2011 at 12:32

    […] writes (and Alex seconds): Nominal wages are fixed for the employed. NGDP falls 5%, and 5% of workers are […]

  64. Gravatar of Scott Sumner Scott Sumner
    16. August 2011 at 19:45

    John, I don’t know.

    Silas, I don’t agree. There must be models explaining the breakdown between the two components, and they can’t be nominal models.

    Travis, I agree that with QE the Fed buys enough of certain maturities to affect their price. But during normal times they don’t buy enough to have a significant direct effect on the price. It’s mainly the indirect effect that matters.

    Robert, It’s the same as in any perfectly competitive market. Suppose there were a billion people holding gold coins. They are all happy with their coins. There is a going rate for gold coins in the free market. I come in to buy one. If it’s a competitive market I can buy one for the going rate plus epsilon (a tiny, tiny amount, too small to see.) That’s how any competitive market works. After I do so, then the gold is worth a tiny bit more in the market place. Now assume a gold producer adds a single gold coin to the market. They must drop the price an epsilon below the market price, maybe one millionth of a cent, to get one more person in the world to be willing to hold gold coins. The effects are so small you don’t see it, but we know it’s there, because we observe competitive markets where individual people are essentially price takers. When millions of extra people want more gold coins, then the price rises by several dollars. This is no different for Fed policy, except replace gold with cash and T-bills.

    Once you’ve figured out how this works, you just have to understand the the nominal price of a dollar bill is always one, so it’s value can only change via changes in the price level, or real value of currency. So an extra few dollars leads to a tiny rise in prices (unless demand also rises, which is especially likely to be true if you have interest bearing reserves.)

    Martin, In that case the sharp break point would depend where the previous rate of inflation was trending. It would not always be zero.

  65. Gravatar of Martin Martin
    17. August 2011 at 08:59

    Scott, I see it differently.

    If you expect nominal income to grow, then it would be very odd not to see a break at zero. The only disagreement would be about how fast you’d expect nominal income to grow and this is what you see in the distribution.

  66. Gravatar of Status update on the macroeconomy « azmytheconomics Status update on the macroeconomy « azmytheconomics
    1. September 2011 at 04:54

    […] more quickly than others. For example, commodities adjust nearly instantaneously, whereas wages and houses fall the slowest. Which areas of the economy are the most out of equilibrium? Why, wages […]

  67. Gravatar of I’m glad Caplan is finally saying it « Entitled to an Opinion I’m glad Caplan is finally saying it « Entitled to an Opinion
    14. November 2011 at 22:36

    […] the view I want to push”. In contrast, the evidence for downward rigidity in nominal wages is overwhelming. And now his co-blogger finally says that to him. Since Arnold seems to delight in persisting with […]

Leave a Reply