Imaginary conversation

We’re in a recession!

I don’t think so.

Yes we are; GDP fell two straight quarters.

That doesn’t mean we’re in a recession

Yes, it does.

OK, let’s say it does. Why should I care if we are in a recession?

Recessions are bad.

Why are they bad?

Because lots of people lose their jobs.

But lots of people are getting jobs right now; we have one of the strongest labor markets ever. So why are recessions bad?

Everyone knows recessions are bad, that’s why they’re called recessions.

Yes, but why are they bad? What’s the big problem?

Well, productivity has recently declined.

Does productivity usually decline in recessions?

No, but it did this time.

But that’s just a coincidence, as you yourself have just admitted. So why are recessions bad?

We have high inflation.

Is inflation usually high during recessions?

No, but it is this time.

But that’s just a coincidence, as you yourself have just admitted. So why are recessions bad? I need an answer.

You just don’t want to admit that we are in a recession!

Sigh . . .

PS. Perhaps were are experiencing a banana.



22 Responses to “Imaginary conversation”

  1. Gravatar of Carl Carl
    6. August 2022 at 10:49

    One reason recessions are bad is they give pretext to all our Keynesian leaders to run their fiscal multiplier experiments on us again.

  2. Gravatar of James Alexander James Alexander
    6. August 2022 at 13:14

    Very funny but … microscopic falls in QoQ RGDP, c’mon. YoY rate of change still well positive.

    And massive volatility of last three years makes errors highly likely.

  3. Gravatar of ssumner ssumner
    6. August 2022 at 14:11

    Carl, Great point.

    James, Obviously, I agree.

  4. Gravatar of artifex artifex
    6. August 2022 at 14:16

    Well, a decline in real output means a decline in consumption. Why focus on employment, separate from its effect on consumption (for example, when less surplus gets created with lower employment)? I agree with your conclusion but don’t like the argument made here. If we defined a recession as a fall in GDP (whether in two consecutive quarters or not), the immediate reason to think recessions are bad would be that, due to less things being produced, less things would be consumed. I’m not saying that’s the case here, but that would be the natural response, not “Because lots of people lose their jobs.”

  5. Gravatar of ssumner ssumner
    6. August 2022 at 14:25

    artifex, Consumption rose:

  6. Gravatar of agrippa postumus agrippa postumus
    6. August 2022 at 15:44

    economist manque sumner forgets the wisdom of zhou enlai: when ask of his thoughts on the french revolution, answered “its too early to tell”. calm down sumner, have a cream soda, the recession will sneak up on you while your eating your crow sandwich.

  7. Gravatar of ssumner ssumner
    6. August 2022 at 16:19

    agrippa, LOL. Congratulations on making a fool of yourself twice in one short post. I expect a recession, and Zhou was referring to the 1968 Paris revolt.

  8. Gravatar of artifex artifex
    6. August 2022 at 18:04

    The biggest contributors to GDP decline over Q1 and Q2 are net exports, change in private inventories, and government consumption and investment, in that order. The two other components of GDP are up: private fixed investment ( and personal consumption expenditures.

    I don’t think the source data is accurate enough that we can trust increases or declines that small, and GDI is less reliable. I’m very confident that there hasn’t been something like what we usually think of as a recession.

  9. Gravatar of ssumner ssumner
    6. August 2022 at 18:30

    artifex, You said:

    “I’m very confident that there hasn’t been something like what we usually think of as a recession.”

    Exactly. We are in a boom, with 3.5% unemployment.

    Will we have a recession? That’s an entirely different question. I suspect the answer is yes.

  10. Gravatar of Larry Larry
    6. August 2022 at 19:44

    I thought recessions were bad because output went down.

  11. Gravatar of David S David S
    6. August 2022 at 21:30

    Real estate prices in the Metaverse are collapsing, which means we’re in a recession. Could the Zuck create a Metaverse Fed with NGDP targeting and avoid the destructive volatility and wanton carelessness of Jerome Powell?

    Anyone who wants to hedge against recession should buy my personalized digital currency. Details to follow.

  12. Gravatar of Michael Sandifer Michael Sandifer
    7. August 2022 at 04:14

    Which measure of productivity are you using when you say productivity doesn’t usually decline during recessions? Even TFP often declines during recessions.

    I agree with the general point that we’re not in recession, and think a recession is unlikely at this point.

  13. Gravatar of Spencer Bradley Hall Spencer Bradley Hall
    7. August 2022 at 05:07

    NBER’s Business Cycle Dating Procedure: Frequently Asked Questions

  14. Gravatar of David David
    7. August 2022 at 06:51

    Output is down over the past 6 months, which has presaged a “recession” every time it’s occurred for the past 70 years. Say the NBER declared that we’re in a recession. One could have nearly the exact same fictional conversation. I think that’s the problem.

  15. Gravatar of ssumner ssumner
    7. August 2022 at 07:55

    Larry, Why is that bad?

    Michael, Labor productivity has been countercyclical since the 1980s:

    David, Yes, and industrial production has risen sharply in the past six months, which has never occurred during a recession in the past 100 years. My 100 years beats your 70 years. 🙂

  16. Gravatar of Michael Sandifer Michael Sandifer
    7. August 2022 at 17:06


    Thanks for that link. Looks like a nice deep dive.

    Superficially, meaning at first glance, labor productivity appears somewhat procyclical, but those dips associated with recessions since 1982 may not be statistically significant.

  17. Gravatar of David David
    7. August 2022 at 17:32

    Scott, the two quarters of negative GDP growth thing wasn’t my main point. My primary point is that relying on a definition of a recession that is essentially, “when this group of a few economists says so,” is imprecise and ultimately unhelpful. What even *is* a recession? It’s the declaration of a few people! OK, but what does that mean? Nothing, really.

  18. Gravatar of ssumner ssumner
    8. August 2022 at 08:00

    Michael, Certainly there is no RELIABLE relationship.

    David, If there were lots of borderline cases, I’d agree with you. But in the US there aren’t ANY borderline cases. Recessions are radically different from expansions. It’s a very useful term to describe the contraction phase of the business cycle.

    Look at the Fed graph of the unemployment rate—of course economists are going to need a term for those periods where unemployment shoots up.

  19. Gravatar of David David
    8. August 2022 at 10:15

    What is a “borderline case” given that there’s no definition besides NBER fiat?

    I think Q2-Q4 of 1991 can reasonably be considered a borderline case. The recession was officially over in April while unemployment was still rising. So, is GDP or unemployment rate more important? I don’t think we can say for sure, because the actual definition of a recession has nothing directly to do with either one. That’s my point.

  20. Gravatar of ssumner ssumner
    8. August 2022 at 10:43

    David, AFAIK, the 1991 recession is almost universally viewed as a recession by economists, by the public, by the media, etc. Look at the graph of unemployment:

    It either goes up by more than 2%, or almost not at all. There’s no case of unemployment going up between 1% and 2% and then starting down again.

    I’d consider a rise by 1% to 2% (trough to peak in unemployment) to be a borderline recession

  21. Gravatar of Sarah Sarah
    9. August 2022 at 08:36

    First of all, how do you have job growth when there is less production? Lower production and lower output is never a good sign. Either they are blowing through cash on hand, are borrowing cash, or the numbers are fudged. How many people left the labor force? How many people quit working? Unemployment in the United States is a lot higher than the numbers say.

    Secondly, the companies that are growing are the biggest companies who are leaning on overseas growth, not domestic growth. Industry continues to consolidate which is not a sign of good health.

    And finally, the whole point of science is to create a model that predicts something of value, and to clearly define your model.

    It’s fine to change a bad model, but it’s never been okay during election season to arrest a former president on trumped up third world charges, then change the definition of recession because both are politically harmful.

    You are so caught up in the politics that you are willing to destroy everything to get what you want.

    And there is nothing more totalitarian than that. It reminds me of the CCP!

  22. Gravatar of Ricardo Ricardo
    9. August 2022 at 08:48

    I was looking at Sumner’s older posts, and not once did he ever mention changing the definition of a recession…until now.

    This is precisely what would happen in the old Soviet Union. The party would change their mind, and all economists would follow suit.

    So why is this so important to Sumner right now? And then it occurred to me the party changed their mind. The democrats, now, almost 24/7, on CNN and CNBC, propagate a more “holistic approach” — whatever that means — to determining whether a country is in a recession.

    So it makes sense he would write about it. Party changes their mind, Sumner changes his mind.
    In other words, what is important to the party, is important to Sumner. If the party had not mentioned it on CNN, Sumner wouldn’t have written about it.

Leave a Reply