The labor market will recover, with or without the Fed

I recently read by a paper by Laurence Ball, which advocated the use of a “high pressure economy” to bring the unemployment rate down to low levels. I’m generally opposed to that sort of policy, as I believe it leads to a procyclical monetary policy—high inflation during booms and low inflation during recessions. That makes the business cycle more unstable.

I initially assumed that it was a recent article, but rereading the piece I saw it was actually from March 2015. That provides slightly more justification for an expansionary monetary policy, but it also raises another interesting question. Consider this prediction:

FOMC members want to accommodate this return to long-run equilibrium while avoiding an overheating of the economy that would push inflation above target. Based on FOMC statements, it appears likely that the Fed will pursue its goals by raising short-term interest rates above their current near-zero levels at some point around the middle of 2015.

This essay argues that a different path for monetary policy would be better for the economy. The Fed should seek to push the unemployment rate well below 5%, at least temporarily. A likely side effect would be a temporary rise in inflation above the Fed’s target, but that outcome is acceptable. To push unemployment down, the Fed should keep interest rates near zero for longer than is currently expected, certainly past the end of 2015.

Notice that Ball thought it would take high inflation to push unemployment down below 5%.  Instead, unemployment has fallen to 4.3% with inflation actually remaining slightly below the Fed’s 2% target (currently it’s closer to 1.5%).

I’ve never believed in “hysteresis” theories that claim unemployment can get stuck at high levels due to a lack of aggregate demand.  I think this theory was based on a misdiagnosis of the European labor market after 1980, where the high unemployment that was assumed to be labor market slack was actually caused by statist labor market regulations.  In the US, the unemployment rate will fall back to the natural rate regardless of whether the Fed pushes inflation above their 2% target.  Thus it’s better for the Fed to focus on stable NGDP growth, and let the labor market take care of itself.

If you want more jobs (and I do), then advocate supply-side labor market reforms.

PS.  A recent piece by Paul Krugman points out that the fall in unemployment to 4.3% also refutes claims made during the recovery period that the high unemployment was partly “structural”.  I was also skeptical of the structural unemployment claim, but even I did not expect unemployment to fall quite this low.

HT:  David Lapidus



46 Responses to “The labor market will recover, with or without the Fed”

  1. Gravatar of Benjamin Cole Benjamin Cole
    12. August 2017 at 22:45

    I think Ball was and is right; inflation is not a concern, keeping as many people working as possible always is.

    Nations with “labor shortages” are happy nations.

    A great way to kill the public’s appetite for “free markets” is chronic high unemployment. Expensive housing will kill desire for putative “free markets” too.

    BTW, the latest BLS report is that unit labor costs in Q2 were down YOY. Down!


    Also, the Fed mysteriously killed off its LMCI just before publication of July data. The LMCI has been trending down since 2012, and looked like it was about to indicate a labor markets deemed “loose” by the Fed.

    Add On:

    Bullard, of St Louis, just gave a speech citing studies that unemployment at 3% would lead to roaring inflation of….1.8% on the core-PCE.

    1.8% is dangerously close to…2%!

    So why is the Fed obsessed with inflation, and fronting Beige Book studies with tales of “spreading labor shortages”?

    The orthodox macroeconomics profession seems badly, and destructively out of step with current observable results.

    See also Japan….

  2. Gravatar of Brett Brett
    13. August 2017 at 00:37

    I think it returns as well. Look at the Panics of 1873 and 1893, which were really awful and happened with no central monetary bank in the US at the time. The economy finally recovered after six years with the first one, and four years with the second.

  3. Gravatar of George Selgin George Selgin
    13. August 2017 at 03:37

    Brett, concerning the 1873 recession, according to Davis’s revised chronology it lasted fewer than 3 years, which further strengthens your point. See

  4. Gravatar of Antischiff Antischiff
    13. August 2017 at 06:05

    Dr. Sumner,

    Data seems to support your concerns about inflation over-shooting being followed by damaging under-shooting, but I think monetary policy is still tight. The Fed is expected to tighten too quickly going forward. That seems an odd claim given the unemployment rate, but we don’t know what the natural rate is and it could fall further than many presently imagine. This is due, in part to lower productivity growth, if it is real, and for to more people choosing to enter the labor market.

  5. Gravatar of Antischiff Antischiff
    13. August 2017 at 06:07

    Benjamin Cole,

    I think you are generally right, but over selling.

  6. Gravatar of ssumner ssumner
    13. August 2017 at 07:18

    Brett and George, I agree with both comments.

    Antischiff, I think the best way to put this is that Fed policy is not currently a big problem for the economy, but their inability to hit their inflation target may be storing up problems for the future.

  7. Gravatar of Antischiff Antischiff
    13. August 2017 at 07:58

    Dr. Sumner,

    As a thought experiment, if you could look ahead to the top of this economic cycle and see a U4 rate of say, 1.5%, would it change your perspective much?

  8. Gravatar of E. Harding E. Harding
    13. August 2017 at 16:00

    I was not surprised by unemployment falling this low during a state of recovery. I consider full employment to be roughly 4%. We’re almost there. However, I did not expect the recovery to be this robust.

    Though the greater share of unemployment becoming long-term was structural, the high unemployment rate itself was not.

  9. Gravatar of Benjamin Cole Benjamin Cole
    13. August 2017 at 20:24


    Since 2008 I have been haranguing the Fed (with no effect) to follow a growth-oriented monetary policy, even possibly money-financed fiscal programs.

    For nine years the sensible types have advised monetary moderation, if not asphyxiation.

    As a result of tight-money policies, what we have is an economy that has given up trillions of dollars in real output, lost profits and wages—but for what?

    A smudge or two in the measured rate of inflation?

    I am overselling?

    Add on: The last three finalists for the US Presidency were Bernie Sanders (socialist), Hillary Clinton (crony capitalist-socialist-feminist) and Trump, who vowed closed borders, US jobs, no more internationalism and cheap health care.

    The voting public will not embrace “free markets” (as carefully defined and circumscribed by the ruling class) if it means chronic high unemployment and expensive housing.

    Frankly, if the Fed feels it must target a rate of unemployment in which there are at least 1.5 people looking for a job for every job opening (4.8% unemployment rate), then the public has a case in voting for Bernie Sanders.

    Can you blame the public?

    In Japan there are two job openings for every job seeker, and housing in Tokyo is affordable. They have no inflation.

    Egads, worshipping by rote the outdated but hoary totems of monetarists…will bring socialism to America.

  10. Gravatar of dtoh dtoh
    14. August 2017 at 02:17

    So a couple of thoughts/hypotheses.

    1. The correlation between the published unemployment rate U3 and inflation is different when the LFPR is lower. Ball and others aren’t taking this into consideration.

    2. Similarly overlooked is the fact that firms can replace labor with capital more rapidly and more cheaply across a wider variety of jobs and in greater quantity than they could in the past

    3. Ditto for moving jobs overseas.

    4. U3 is perhaps a better measure of how long it takes people to find a job once they start officially looking rather than how many would actually like work. Perhaps technical changes have reduced the amount of time required to find work.

    5. Why do we want lower unemployment? What’s wrong with people choosing leisure over income if the choice is undistorted by by tax policy, regulations, etc.

    6. I thought the hysteresis theory relied more on expectations rather than actual AD.

  11. Gravatar of dtoh dtoh
    14. August 2017 at 02:19

    @Benjamin Cole

    Housing in Japan is affordable?

  12. Gravatar of dtoh dtoh
    14. August 2017 at 02:25

    Any thoughts on the latest GDP numbers from Japan?

  13. Gravatar of Benjamin Cole Benjamin Cole
    14. August 2017 at 03:34


    yes, see

    Tokyo in general allows development of residential land. So housing is affordable to the residents of Japan, if not foreigners.

    The mortgage rate in Japan is 1.25%.

    Rent for one-bedroom apartment in Japan, outside of city center is under $500 a month.

    My estimate is that residents of japan have higher living standards than those of San Jose, Los Angeles, NYC, Boston etc.

    And that is before street safety, crime is taken into consideration.

    The idea that Japan is expensive is dated, and affected by currency exchange rates.

  14. Gravatar of Antischiff Antischiff
    14. August 2017 at 06:30

    Benjamin Cole,

    Yes, I think you’re correct in general, but talking about helicopter drops, for example, when unemployment is approaching 4%, comes across as shrill. While I favor a new monetary regime, I don’t think we need one to maximize employment. We need the Fed to stop credibly promising to tap the brakes during acceleration.

  15. Gravatar of ChargerCarl ChargerCarl
    14. August 2017 at 09:06

    Tim Duy is upset at Matt Yglesias for referring to the hawkish forces at the Fed as “the deep state”, but I think it’s a somewhat appropriate analogy:

  16. Gravatar of Benjamin Cole Benjamin Cole
    14. August 2017 at 12:25

    Charger Carl: What is the reason for the Fed’s policy of targeting a rate of unemployment at which there will be at least 1.5 people looking for a job for every job opening?

    Deep state? Or does vulgar Marxist analysis present a more clear conceptual lens?

    I often say vulgar Marxist analysis explains much, even as Marxist medicine is poison.

  17. Gravatar of Major.freedom Major.freedom
    14. August 2017 at 12:34


    “I’m generally opposed to that sort of policy, as I believe it leads to a procyclical monetary policy—high inflation during booms and low inflation during recessions. That makes the business cycle more unstable.”

    The Fed CAUSES the business cycle, and it is causing them even when NGDP is “stable”. This is the case when market forces would otherwise reduce NGDP as part of economy wide relative coordination corrections that cannot occur with NGDP being increased on account of relative capital structure distorting inflation.

    The Fed is not a savior of business cycles. The Fed is not a moderator of business cycles. The Fed is not an exacerbator of business cycles. The Fed is not merely an institution that makes allegedly endogenous business cycles better or worse.

    You cannot know how to fix economic problems unless you know how markets actually work. Since you don’t understand how markets work, you keep recommending destructive and counter-productive responses that only make the Fed caused problems worse.

    Only markets can fix monetary problems. This is the case even in a world without a market in money. Here, the only difference is that the problems cannot be fixed, permanently. Markets can only find out after the damage is done, because it is impossible to know in advance how exactly people will react to being misled and hampered, knowingly or unknowingly.


    “A recent piece by Paul Krugman points out that the fall in unemployment to 4.3% also refutes claims made during the recovery period that the high unemployment was partly “structural”

    No it doesn’t. The structural problems referred to here are, contrary to Krugman’s Keynesian doctrine, most pronounced during the “boom” times. The overall level of unemployment on a given day is irrelevant to whether or nor there are structural problems. For example, too many people can be building houses for a time, and during this time unemployment overall may be low. But just because the unemployment is low now, it doesn’t mean there are no structural problems.

    One flaw among many in the Keynesian doctrine is that it considers singular aggregate statistics as the main statistics with real world impact and meaning, when in reality the aggregates just cover up the meaningful relationships in the market process, and cause confusion in the minds of statists like Krugman who data mine all day looking for confirmation biases when it is suitable, and being absolutely silent other times when the theory is falsified for the umpteenth time, to claiming empirical validation no matter what the data says.

    Structural problems do not necessarily show up as permanent aggregate unemployment declines.

  18. Gravatar of major.freedom major.freedom
    14. August 2017 at 15:55

    “Thus it’s better for the Fed to focus on stable NGDP growth, and let the labor market take care of itself.”

    It is not possible for the labor market to “take care of itself” when it is not legally allowed to do so on account of one half of all labor contracts being composed of socialist controlled currency.

    In order for the labor market to “take care of itself”, labor needs to be able to take care of producing and distributing money.

    Sumner is just claiming the labor market can take care of itself because he wants to divert attention away from his own anti-labor market advocacy. Pay no attention here, look over there!

  19. Gravatar of Benjamin Cole Benjamin Cole
    14. August 2017 at 19:18

    In some ways, one of the most flabbergasting speeches in Fed history….

    PS I think I disagree with the idea the labor market will fix itself despite the Fed.

    We saw in the Great Depression long-term consequences of too-toght money (worsened by certain NLRB actions). The lesson of the Great Depression is that U.S. central bankers would have never ended it.

    And we saw in pre-war Japan the successful application of money-fnanced fiscal programs. Korekiyo Takahashi steered japan out of the Great Depression, alone among developed nations. (There’s a history short-shrifted in orthodox macroeconomic textbooks).

    If the Fed targets 4.8% unemployment, then who is to say labor markets will ever recover?

    And a labor market recovery, by whose standards? Real weekly earnings in the U.S. for adult males are lower now than in 1979.

    Tyler Cowen suggests wages are lower now than in the 1960s.

    This is a labor market recovery?

  20. Gravatar of Pyrmonter Pyrmonter
    14. August 2017 at 21:03

    Is the unemployment rate the right measure? Male Labour Force participation is still well below pre-GFC levels; while some of the explanation is no doubt more time spent in education, it’s surprising; compare it with the comparable position in the UK:

    I’m neither a labour economist nor a sociologist; but intuit that the lower rate of “male breadwinners” matters.

  21. Gravatar of dtoh dtoh
    14. August 2017 at 22:08

    @Benjamin Cole

    “Rent for one-bedroom apartment in Japan, outside of city center is under $500 a month. “

    It depends on whether you’re talking Tokyo or other cities, and it also depends on how you define “one-bedroom” and “city center.” If you mean a 25 sqm apartment that’s a one hour standing commute on a crowded train, then yes you can get an “affordable” apartment for under $500 a month.

    Average rental rates in Tokyo are $80/sqm/month, which is what you would pay for at the very high end of the market in Manahattan. Compared to Brooklyn and Queens, it’s 2x or 3x. Smaller cities in Japan are cheaper but so are equivalently smaller cities in the U.S. Housing expenditures are not that much higher in Japan than the U.S., but it’s only because Japanese live in tiny houses and apartments. I’d guess on average housing cost per sqm is 3 or 4 times higher in Japan than the U.S.

    Tokyo IMHO has by far the best living standard of any big city on the planet, but affordable housing is not part of the equation. (Outside of the big cities, Japan is now like a 3rd world country.)

  22. Gravatar of Mark Mark
    15. August 2017 at 02:13


    Tangential: Did you see John Tamny’s article on Market Monetarism from last month?

    It seems that Tamny and Paul Krugman have something they agree on: market monetarism is a form of economic interventionism.

    But the way it seems to me, just by virtue of having a monopoly on currency creation, the government has *already* intervened egregiously. Once the state has already asserted control over a market, it is absurd to argue that it is somehow ‘less interventionist’ for the state to produce less of that good, or to produce a fixed quantity, even though demand is changing.

    If the government took over the potato industry and committed to producing only as many potatoes as were consumed each year so as to keep the number of potatoes out there fixed, and demand for potatoes increases, not increasing potato production isn’t really non-interventionist. The intervention ship sailed when the government asserted control over the potato market.

    Does this make sense (to anyone?)? Sorry for the random question.

  23. Gravatar of Benjamin Cole Benjamin Cole
    15. August 2017 at 15:37

    Dtoh: I provided a cite for my info. Please provide a cite for your info.

  24. Gravatar of Benjamin Cole Benjamin Cole
    15. August 2017 at 17:02

    A studio apt. in Tokyo, on average, about $715 a month.

    Outside Tokyo, about $300 to $500 a month.

    Maybe the units are small. But add low-cost health care and low-cost housing together, it strikes me that Japanese have higher living standards than Americans, on average.

    If Japan outside Tokyo is “3rd World,” I wonder how we define large swaths of Detroit, the Texas border, or West Virginia?

    I live now in rural Thailand, which is perhaps 3rd World. I have yet to see anything as depraved as parts of Detroit.

    Probably this reflects the behavior of inhabitants, but no doubt in some measure public policies.

  25. Gravatar of dtoh dtoh
    15. August 2017 at 17:15

    @Benjamin Cole

    I’m personally somewhat knowledgeable on this subject. I was a Director of one the largest publicly TSE traded residential REITs in Japan. I’m a condominium owner in central Tokyo. I’ve been responsible for a $350 joint use development project in Tokyo. I’ve successfully sued the Tokyo Metropolitan Government to block a residential development project. And I have detailed historical (but private) monthly rental rate data for all major Japanese cities for a broad class of residential rental properties.

    If you want more info, CBRE sometimes publicly releases data on the Japanese real estate market in English. If you speak Japanese, you can check one of the many apartment rental brokerage web sites like Suumo or AtHome.

  26. Gravatar of dtoh dtoh
    15. August 2017 at 17:19


    I not arguing about Japanese standards of living. I pretty much agree with you on that. I’m just saying housing is very expensive on a per sqm basis.

    What part of rural Thailand? IMHO rural Thailand is positively prosperous. Certainly more so than what’s happening in rural Japan… and definitely a lot more so than Detroit and other parts of the U.S. that have been blighted.

  27. Gravatar of dtoh dtoh
    15. August 2017 at 17:22


    Also, I would not be so complimentary of Japanese health care (or at least medical care). It’s cheap but not necessarily great. Depending on the type of medical care required, BKK is often a better choice.

  28. Gravatar of Benjamin Cole Benjamin Cole
    16. August 2017 at 03:49

    Near Pak Chong, which is not so far from Korat.

    I don’t know if “prosperous” describes my part of rural Thailand.

    Surely, many prosperous people, but also the scooter instead of a car, small cinder block house family is very common. There is national health insurance, very cheap, but many major illnesses not covered…

    My limited experience is that Thai hospitals are first-rate and can be 10 cents on the dollar to US facilities….

    I used to have access to CBRE data, knew them better back when I lived in L.A….

    Congrats on having a r/e career in Tokyo, sounds interesting

  29. Gravatar of dtoh dtoh
    16. August 2017 at 05:30

    @Benajamin Cole

    Actually I didn’t really have a career in real estate. It was more of a distraction I got sucked into as a volunteer or as a favor to friends.

    Never been to Pak Chong but have spent some time further east around Ubon Ratchathani. I’d guess scooter ridership there is down by 60% over the past 10 years. Last time I was there, they’d just put in a Mercedes dealership. These days I mostly hang out in VN.

  30. Gravatar of ssumner ssumner
    16. August 2017 at 09:28

    Antischiff, Can you be more specific?

    Harding, This may well be the weakest recovery in American history. 2% GDP growth is really slow for a recovery.

    dtoh, I’m not sure I agree with point #2. If that were true, you’d expect faster productivity growth. But it’s very slow.

    You said:

    “Why do we want lower unemployment? What’s wrong with people choosing leisure over income if the choice is undistorted by by tax policy, regulations, etc.”

    People who choose not to work are not counted as unemployed. The unemployed are people who want to work. (Plus measurement error.)

    The past year has been strong, perhaps due to the BOJ’s bond pegging program, and/or the strong dollar. We’ll have to see if the dollar’s recent weakness affects the numbers going forwards.

    Can you elaborate on point #6?

    Pyrmonter, Unemployment is a better business cycle indicator: the LFPR reflects longer term (supply-side) issues.

    Mark, Yes, I agree (as does Nick Rowe), no specific Fed monetary policy is more “interventionist” than any another.

  31. Gravatar of Antischiff Antischiff
    16. August 2017 at 13:33

    Dr. Sumner,

    If you could know now that the natural rate of unemployment was 1.5%, would you favor looser monetary policy? How about 2.5%?

  32. Gravatar of dtoh dtoh
    16. August 2017 at 15:17

    You said, “The unemployed are people who want to work.”

    I think that depends on your point of view. I have a friend who is no longer working. He says he’s retired. His wife insists that he’s unemployed.

  33. Gravatar of ssumner ssumner
    17. August 2017 at 07:54

    Antischiff, Maybe, but the point is that we don’t know what the natural rate of unemployment is. That’s one reason why I favor a NGDP rule.

    dtoh, I thought you were talking about the people that the government calls unemployed. If you mean people not working, then my 91 year old mom is “unemployed”.

  34. Gravatar of dtoh dtoh
    17. August 2017 at 16:27

    I was being tongue in cheek, but now that you mention it, U3 does not measure whether people want to work, it measures whether they are looking for work.

    I may have mentioned before but I had previously seen a study where Japanese male unemployment doubles and female unemployment goes up 60% when the question is changed from “have you looked for work in the last two weeks” to six weeks (same as U.S.).

    Increasingly, I think that the LFPR is a much better (although still highly imperfect) measure of the state of the employment market.

  35. Gravatar of ssumner ssumner
    19. August 2017 at 09:18

    dtoh, I don’t agree–it is affected by societal trends like women entering the labor force, or the rise of SSDI, or teens no longer preferring to work, or aging of the population.

  36. Gravatar of W. Peden W. Peden
    19. August 2017 at 13:10

    The hysteresis hypothesis was also partly based on bad figures for UK unemployment. The raw stats suggested a rise in the UK natural rate during the high unemployment of the early 1980s and a jobless recovery in 1983-1986. However, if you control for changes in labour participation, then you can see that both the natural rate and the actual rate rose more sharply in 1980 than people thought as hidden unemployment was removed and demographics underwent a big shift. Similarly, employment started recovering in 1983, just as you’d expect if the hysteresis hypothesis was false.

    There are also problems with identifying causes of changes in the natural rate without sufficient controls. I remember reading a paper that suggested that hysteresis was undetectable once you controlled for things like exchange rate shifts. In the UK at least, a lot of trends in the natural rate can be explained in terms of regional deindustrialisation effects and insufficient labour force mobility, because the labour force participation rates are much more regionally divergent than the raw unemployment figures.

    There might well be some hysteresis effects (it’s plausible that skills erosion can play SOME role in determing the naturl rate) but they don’t seem to be large or predictable, as far as I can tell.

  37. Gravatar of dtoh dtoh
    19. August 2017 at 18:10

    I agree. Longer term, you definitely need to look at societal trends, but the 12 million person drop in the labor force after 2007 didn’t have anything to do with women working and little to do with the aging work force or other societal trends. It was mostly because of poor policy and it’s not reflected in the U3 number.

    IMHO, the focus on U3 is one of the reasons the Fed is allowed to get away with such poor performance. LFPR would be a better target…. or total wages or NGDP growth rate.

    And yes you would need to adjust a LFPR over time to accomodate societal trends, but that’s true for an NGDP target also.

  38. Gravatar of ssumner ssumner
    22. August 2017 at 20:07

    W. Peden, I agree.

    dtoh, If there had been a 12 million person drop in the labor force after 2007, I’d definitely agree with you. But there wasn’t, it was basically flat:

  39. Gravatar of dtoh dtoh
    23. August 2017 at 16:10

  40. Gravatar of dtoh dtoh
    23. August 2017 at 22:34

    Sorry for not being clearer, but what I meant was the 12 million drop after adjusting for population growth. I think since we were talking about LFPR that should have been obvious even if what I was wrote was imprecise.

  41. Gravatar of dtoh dtoh
    23. August 2017 at 22:37

    And adjusting for the shift to part time work. I.e had we had the same LFPR and mix of full time to part time job that we had at the beginning of 2007, there would be 12 million more FTE jobs.

  42. Gravatar of Jose Jose
    24. August 2017 at 12:19

    Sorry for the external link, but the graph helps a lot. My engineer eye tells me that probably the low and stable inflation of the current environment has produced the STEEPEST reduction in unemployment ever in US economic history. Don’t know it that is being noted …

  43. Gravatar of dtoh dtoh
    24. August 2017 at 16:51

    Another reason the Fed shouldn’t be looking at U3.

  44. Gravatar of ssumner ssumner
    24. August 2017 at 22:09

    dtoh, It’s not “obvious” unless you say it. In any case, that graph in no way supports your argument. Where do you get the 12 million?

    And why adjust for population growth but not the retirement of boomers? Or the rise in disability? You can’t just grab some data off a Fred graph, you need a careful study of the evidence.

    Jose, Your eye deceives you—the decline in unemployment during 1983-84 was much steeper.

    dtoh, I agree the Fed should not be looking at U3, or any other unemployment rate.

    They should focus on NGDP.

  45. Gravatar of dtoh dtoh
    25. August 2017 at 00:43


    “You can’t just grab some data off a Fed graph.”

    It was a BLS graph. 🙂 and I was merely responding to your just grabbing data off a Fed graph.

    You need a careful study of the evidence.

    Agree. And the 12 million number did come from what seemed like a careful study. I just can’t remember where. I do remember though that it took into account the retirement of the boomers, which is why in my original comment, I said that it had “little” (rather than nothing) to do with the aging population.

  46. Gravatar of Jose Jose
    28. August 2017 at 04:42

    Prof. Sumner
    Actually I was expecting that comment, but if you take the entire leg of the cycle the results are actually pretty similar, 5.5% reduction in unemployment over a time period of 7 years. The thing is, 2010-1017 has been with almost no variance in inflation, and the rate of fall is pretty regular, with low variance too. I the period you mention, 82-89, we can see three different periods of unemployment change, and unstable inflation. I would contend that a period like 2010-17 is preferrable to a period like 82-89.

Leave a Reply