Claudia Sahm on the risk of recession

A few days ago Claudia Sahm wrote a Bloomberg column entitled:

My Recession Rule Was Meant to Be Broken

Even before I read the article, I suspected that I would sympathize with her argument, which proved to be the case:

The US is not in a recession, despite the indicator bearing my name saying that it is. The Sahm rule, which was triggered with Friday’s weaker-than-expected jobs report, joins a long list of economic tools skewed by the unusual disruptions of the past four and a half years.

That said — and I say this with a mixture of humility and concern — the Sahm rule is still relevant. The risk of a recession is elevated, strengthening the case for the US Federal Reserve to cut interest rates. . . .

But conditions can change quickly, and by the time the NBER has officially designated a recession, it is usually too late to guide policymakers. The purpose of the Sahm rule is to act as a kind of early diagnosis.

I frequently argue that recessions are almost unforecastable, because if there were a reliable technique for forecasting a (demand-side) slump, the Fed would move to try to prevent this from occurring. I’ve previously suggested that the lack of mini-recessions in the US is evidence that the Fed often reacts too slowly to the onset of recession.

Sahm sees her rule not so much as an unconditional forecasting technique, rather as a guide to help policymakers improve on stabilization policy. As an analogy, the beep your car makes backing up near an object is not supposed to predict an accident, it’s supposed to prevent an accident.

It seems to me that Sahm strikes exactly the right note—that her rule should be viewed as useful, not infallible. The recent triggering of the Sahm’s Rule suggests that the risk of recession is substantially elevated and that policymakers need to take this information into account.

The economics/finance community is full of people with big egos. I often hear people brag about how they have been correct time after time, which tends to make me take their views less seriously. At times, it almost seems like they care more about their reputation than about the health of the economy. Sahm seems to understand that if her rule ends up preventing a recession, it would be a vastly greater achievement than if it ends up predicting a recession. Let’s hope that this time her rule fails, and she succeeds.


Tags:

 
 
 

25 Responses to “Claudia Sahm on the risk of recession”

  1. Gravatar of Lizard Man Lizard Man
    10. August 2024 at 03:41

    Is the Fed’s desired inflation rate possible without a recession? It appears like maybe the answer is yes, but if they start cutting rates now, and cut rates fast enough, isn’t the likeliest outcome that they will overshoot the “neutral rate” and inflation will rise again? And then they will try to bring inflation down, and run into the same situation where inflation comes down because the economy slows down, and then they have to try to bring inflation back up again, etc.? I think all of this supports the idea of a NGDP level target, because the Fed would become predictable. I suspect that part of the reason why the US has had no mini-recessions isn’t just that the Fed is slow to act, but because the Fed isn’t predictable.

  2. Gravatar of Sara Sara
    10. August 2024 at 04:44

    “The economics/finance community is full of people with big egos. I often hear people brag about how they have been correct time after time, which tends to make me take their views less seriously.”

    Yes, but this is exactly what you do. You make predictions that have about a 50% success rate, and then claim you’re always right. You’re latest failure was telling everyone inflation would be “transitory”. Your own model predicted that, and the model was proven to be wrong. Every economist — including the Fed, said the same thing at the same time, as if they were afraid to disagree (Groupthink). But nobody on Wall Street actually took it seriously. We just laugh at you. By you — I mean, all economists.

    #pseudoscience.

  3. Gravatar of ssumner ssumner
    10. August 2024 at 05:33

    Lizard, You said, “I suspect that part of the reason why the US has had no mini-recessions isn’t just that the Fed is slow to act, but because the Fed isn’t predictable.”

    Yes!!

    Sara, You said:

    “and then claim you’re always right.”

    LOL, that’s even dumber than usual.

  4. Gravatar of Eharding Eharding
    10. August 2024 at 06:25

    Sumner, I still expect a recession to start prior to February 2025.

  5. Gravatar of spencer spencer
    10. August 2024 at 11:22

    The lags in money flows determine whether money is neutral, robust, or harmful. Contrary to Nobel Laureates Milton Friedman and Anna J. Schwartz: “A Monetary History of the United States”, 1867–1960, the distributed lag effect of monetary flows, the volume and velocity of money, are mathematical constants.

    Jamie Dimon is right: “JPMorgan Chase & Co. Chief Executive Officer Jamie Dimon said he’s skeptical that inflation will return to the Federal Reserve’s 2% target,”

    We’ll have to have a recession to get to the 2% inflation target.

    To have a recession, the roc in short-term money flows will have to turn negative.

  6. Gravatar of Solon of the East Solon of the East
    10. August 2024 at 16:43

    The clunky Federal Reserve, with its eight times a year meetings and inability to affect fiscal policies, and that all operating within a globalized financial system with globalized interest rates…is going to be late.

    The US needs a system of social security tax holidays, which would almost immediately put more money in the pockets of employees and employers (and not goofy federal spending programs).

    Losses to the social security trust fund could be offset by having the Federal Reserve buy Treasuries and placing them into the trust fund.

  7. Gravatar of kangaroo kangaroo
    11. August 2024 at 03:56

    Uh Scott, bad news about dope over at Marginal Revolution.

    Honestly, I don’t put much stock in most of the claims Tyler says the paper makes. IMO most such research is badly confounded and has a low signal to noise ratio to begin with, so that the results don’t have a lot of muscle. But the suggestion that dope lowers ppls IQ… about that, there is absolutely no question in my mind.

    No, I didn’t read the paper or even the abstract. My priors will suffice.

    I could be a total hypocrite and argue *for* “root causism” regarding dope and against it regarding crime! 🙂 If I was doing such a thing I would argue that one way to combat drug (sans drug war) use would be to dergulate the economy especially related to housing and let people create opportunities for themselves so they don’t need dope escapism.

    Don’t know if you’ve read Chernow’s excellent book “Titan” regarding Rockefeller. Highly recommended. But according to Chernow JR and his wife were very active in the abstinence movement. Would be great to see something like that today WRT drugs: private groups promoting alternatives.

    OH well.

  8. Gravatar of ssumner ssumner
    11. August 2024 at 08:27

    Harding, That’s fascinating. Tell me more!!

    Kangaroo, I already did a post responding to Tyler. You should read it before wasting your time commenting here. By the way, the study was good news for “dope”.

  9. Gravatar of RickG RickG
    11. August 2024 at 10:53

    So if you removed government borrowed money spending from the economy, what would our rate of growth be?

  10. Gravatar of ssumner ssumner
    11. August 2024 at 16:44

    Rick, Maybe a bit higher? Government is a drag.

  11. Gravatar of TF TF
    11. August 2024 at 17:13

    Hey Prof Sumner,
    Claudia Sahm is great, my longstanding problem with her “rule” is whomever called it a rule. It’s obviously not anything of the sort. The probability of a recession is still the same as it was a week ago. In general people who try to predict things based on history should understand statistics (what is n in US recession prediction?)

  12. Gravatar of ssumner ssumner
    11. August 2024 at 19:24

    TF, I don’t think anyone sensible assumed it was a 100% accurate forecasting tool. That wasn’t Sahm’s view, nor was it my view.

    Terms like “rule” and “law” are used in a variety of ways in economics.

  13. Gravatar of Matthias Matthias
    11. August 2024 at 19:34

    Off-topic:

    Scott, you should probably include Astral Codex Ten in the links in your sidebar.

  14. Gravatar of Michael Sandifer Michael Sandifer
    11. August 2024 at 20:39

    In my latest blog post, I introduce a simple, very back of the envelope method for imputing expected NGDP growth for future quarters in S&P 500 data, using this formula:

    Expected NGDP Growth = Baseline Expected NGDP Growth + (Baseline Expected NGDP Growth – Expected S&P 500 Earnings Yield)

    I explain how I choose a seemingly plausible baseline, with the expected earnings yield based on adjusted analysts earnings estimates and S&P 500 futures prices. This approach relies on the longer-run equilibrium relationship between NGDP growth and the S&P 500 earnings yield.

    Interestingly, one implication is that, given a sufficient history of earnings yields, one can estimate past NGDP growth sans government statistics. This is what I meant in the past when I said macroeconomic data is very hard for governments to fake. They often realize this, which is why dishonest governments often choose to stop publishing certain unflattering data, rather than try to fake it.

    When this approach is used to determine the mean expected NGDP growth rate, not specific to any particularly point in time, it could be useful in targeting NGDP. The baseline in that case might be the mean NGDP growth rate of the prior economic cycle, for example.

  15. Gravatar of Michael Sandifer Michael Sandifer
    11. August 2024 at 20:42

    It’s important to mention that the earnings yields use concurrent quarter earnings, not those of the trailing 12 months.

  16. Gravatar of Michael Sandifer Michael Sandifer
    11. August 2024 at 20:59

    I should also point out that if one wants to impute past NGDP growth in historical S&P 500 earnings yield data, one must use long run averages or smooth earnings. I have a method for adjusting for the short-run hypervolatility of earnings that I’ll elaborate upon when I have time. It involves a model that substitutes earnings for dividends and expected NGDP growth for r – g in the Gordon growth model. That model simply doesn’t apply to stock markets as a whole, as the case of Japan clearly demonstrates. Falling real interest rates won’t boost stock prices, for example, if due to falling NGDP growth expectations.

  17. Gravatar of ssumner ssumner
    12. August 2024 at 08:50

    Matthias, At some point before the end of the year, I hope to do a major upgrade.

    Michael, Do you have a link for that blog post?

  18. Gravatar of Michael Sandifer Michael Sandifer
    12. August 2024 at 13:53

    Scott,

    Here’s a link to that post:

    https://exactmacro.substack.com/p/stock-and-gdp-outlook-for-week-ending-0f2

    Clicking on my name above my comments also takes one to my blog.

  19. Gravatar of bill bill
    13. August 2024 at 05:44

    We shouldn’t ignore that her rule had a second part. That once the threshold was breached, Congress would automatically have a fiscal response (I think it was to mail out checks, but I’m not sure). Clearly, to the extent we respond, it should be monetary. I’d hope that even non-monetarists would prefer that at least until IOR was at zero.

  20. Gravatar of bb bb
    13. August 2024 at 07:52

    Scott,
    I listened to an interview of Sahm back in 2020 (during the days of podcasts an covid walks). At the time I remember thinking she was very smart but probably would benefit from a few more years of experience before she achieved notoriety. In that interview she fairly confidently pitched the idea that recessions could be averted with early/automatic fiscal stabilizers. She was also dismissive of her former colleagues at the Fed, which I didn’t love. Didn’t get past the paywall, but it seems she now favors monetary action over fiscal in her article. Glad to hear that she’s landed in a more pragmatic place, because I did find her very smart and compelling.
    Good post.

  21. Gravatar of ssumner ssumner
    13. August 2024 at 09:29

    Michael, I notice you have 3.8% for Q3. The Atlanta Fed has 2.9% RGDP, implying (probably) at least 5% for NGDP. It will be an interesting test.

  22. Gravatar of steve steve
    13. August 2024 at 11:52

    The dismal science indeed. Full of dismal, petty people with no concept of reality.

  23. Gravatar of Randomize Randomize
    13. August 2024 at 12:44

    Solon, you said: “Losses to the social security trust fund could be offset by having the Federal Reserve buy Treasuries and placing them into the trust fund.”

    That sure sounds like the Fed printing money to deposit directly in social security? The line between monetary stimulus and funding the government with printed money is a fine line but it does exist for good reason and I believe your proposal would cross it.

  24. Gravatar of Michael Sandifer Michael Sandifer
    13. August 2024 at 14:24

    Scott,

    Yes, I’m skeptical about that GDPNow nowcast and am increasingly wondering whether Q2 GDP will be revised downward, as such hot figures seem to conflict with other data. For example, why would markets so strongly predict rate cuts this year if NGDP growth will continue running at 5%+? Why did we see the VIX spike to crisis levels last week if there’s confidence NGDP growth will remain high? Sure, Japan and South Korea suddenly going into recession would represent a real shock for us, but not large enough to make such a difference. Most of the US reaction last week was due to concerns about nominal shocks.

    It’s true that inflation expectations are now well-below the Fed’s target, but real GDP’s been booming due to immigration and increased AI investment. Also, recent inflation has remained above target.

  25. Gravatar of Eharding Eharding
    14. August 2024 at 09:51

    Sumner, will you give me any plaudits if I’m right?

Leave a Reply