A Portrait of Larry Summers

This article gives a good sense of the Larry Summers that I know—particularly how much he loves a good intellectual discussion from almost any starting point. Having Larry as one of my professors in graduate school and watching him during breaks at conferences gave me the view (which I still hold), that the most important economics is what happens in the hallways on those breaks (and often beyond, as those in a good discussion play hooky from the next formal talk). I draw from that premise the conclusion that the purpose of economic models is to train our intuition so that we are well prepared for such free-wheeling discussions.

Chimamanda Ngozi Adichie's Advice on the Benefits of Failure

I ran across this article (“Protect and Value the Truth”) about a wonderful talk by author Chimamanda Ngozi Adichie. Here are my favorite quotations from her that are given in the article. I separated them by added bullets:

  • It is hard to tell ourselves the truth about our failures, our fragilities, our uncertainties. It is hard to tell ourselves that maybe we haven’t done the best that we can. It is hard to tell ourselves the truth of our emotions, that maybe what we feel is hurt rather than anger, that maybe it is time to close the chapter of a relationship and walk away. And yet, when we do, we are the better off for it.

  • Be courageous enough to say ‘I don’t know … Ignorance acknowledged is an opportunity; ignorance denied is a closed door. And it takes courage to admit to the truth of what you do not know.

  • You cannot create anything of value without both self-doubt and self-belief. Without self-doubt, you become complacent; without self-belief, you cannot succeed.

  • Your story does not have to have a traditional arc.

Donald Yacovone: How Ubiquitous History Textbooks Taught White Supremacy →

The details in the book excerpt linked to above are telling.

One very troubling idea, but one that might have too much truth to it, is that an outgroup is necessary to generate unity in an ingroup. Donald Yacovone asserts:

… if no slaves ever existed in the South, Northern white theorists, religious leaders, intellectuals, writers, educators, politicians, and lawyers would have invented a lesser race (which is what happened) to build white democratic solidarity, and in that way make democratic culture and political institutions possible. As one of our greatest authors, Toni Morrison, once explained, in the United States the rights of man were “inevitably yoked to Africanism.” In other words, American democracy depended on Black inequality to sustain white equality.

Jonathan Haidt talks about “groupishness” as one mode of being human. Groupishness makes us act very well towards the ingroup and badly toward the outgroup:

As Rob Henderson says, "A psychopath is a person who treats their ingroup the same way that normal people treat their outgroup."

I hope that having an outgroup is only a shortcut to getting ingroup solidarity and not an absolute necessity. It may require an advance in social cohesion technology to have any better way.

Current Information Provision is Inadequate: Let's Put Salient Warnings on Alcoholic Beverages

Alcohol causes health harms of the same order of magnitude as tobacco, but health warnings on alcohol are much less salient than those on tobacco. It is past time to rectify that. The New England Journal of Medicine “Perspective” piece shown above has useful points to make. Here are some passages, separated by added bullets:

  • … alcohol consumption now accounts for more than 140,000 deaths per year in the United States, or more than 380 deaths per day. The Covid-19 pandemic has exacerbated alcohol-associated harm in the United States, with alcohol-related deaths increasing by 25% during the first year of the pandemic as compared with the previous year.

  • In addition to the data on fatal and nonfatal injuries resulting from acute intoxication (including injuries caused by motor vehicle crashes), mounting research links longer-term alcohol consumption to chronic diseases including hypertensive heart disease, cirrhosis, and several types of cancer.2 Even light or moderate drinking increases the risk of these conditions, particularly cancer.2

  • Warning labels are most effective when they are displayed prominently on the front of product packaging, include pictorial elements such as photographs or icons, and rotate the content of their messages to avoid any one message becoming “stale.” A randomized trial involving 2149 smokers, for example, found that large, front-of-pack, pictorial warning labels for cigarettes increased smoking quit rates by 50% (from 3.8 to 5.7%) over 4 weeks as compared with smaller, side-of-pack, text-only warning labels.4 Similar benefits have been documented in longitudinal studies examining smoking behaviors after countries implemented well-designed cigarette-package warning labels and in quasi-experiments evaluating grocery purchases after implementation of prominent front-of-package warning labels for unhealthy foods and beverages.

  • The current alcohol warning in the United States lacks all the key elements of evidence-based warning design: it uses small text, typically appears on the back or side of product packaging, and doesn’t include any pictorial elements (see figure). The warning message is also static, having remained unchanged since the label was first implemented more than three decades ago.

  • … when large, pictorial warnings about cancer risk were temporarily added to the front of alcohol containers in some stores in Yukon, Canada, alcohol sales dropped by 6 to 10%.5

The pleasing idea that a little alcohol is good for health has done a lot to maintain a positive image for alcohol. Unfortunately, that idea is false. I lay out one of the most telling pieces of evidence damning alcohol in “Data on Asian Genes that Discourage Alcohol Consumption Explode the Myth that a Little Alcohol is Good for your Health.”

Warnings are likely to substantially reduce alcohol consumption. And we should not stop there. It would help protect people if every instance of alcohol consumption was considered an indulgence and social disapproval of more than minimal alcohol consumption gradually increased.

How a Toolkit Lacking a Full Strength Negative Interest Rate Option Led to the Current Inflationary Surge

In the third part of our April 7, 2022 “The Future of Inflation” trilogy, “The Electronic Money Standard and the Possibility of a Zero Inflation Target,” Ruchir Agarwal and I write:

Despite the surge in inflation worldwide, central banks have been behind the curve in raising rates. One reason is the reliance on forward guidance instead of negative interest rate policy. At the onset of the COVID-19 crisis, most advanced country central banks lowered their rates to the effective lower bound (but not deeper into negative territory) and thus had to deploy ‘forward guidance’—by committing to maintain the near-zero rates until they were confident that the economy was on track to achieve employment targets. However, such forward guidance tied the hands of several central banks when faced with rising inflation—leading to a sluggish interest response to record-high inflation. Tying their hands with a forward guidance promise was especially unfortunate given the unprecedented nature of the shock.

And for Mitra Kalita’s August 9, 2022 newsletter post “We're Asking the Wrong Question About the Recession,” I crafted two unused quotes (quoting from my email to her):

  • Because raising rates too late means it is typically necessary to engineer a recession to bring down inflation, the Fed, in effect, caused the coming recession by failing to take decisive action earlier. What might decisive action earlier have looked like? Instead of waiting to raise rates until April 2022 and starting slowly, the Fed should have raised rates by 3/4 of a percentage point in December 2021 and announced that it would keep raising rates by 3/4 of a percentage point at each meeting until it saw clear signs that inflation was headed back all the way to the 2 % target rate. The Fed had enough information by its December 14-15 meeting to make that call.

  • Thinking falsely that it couldn't fall back on negative rates, the Fed was too slow and timid in raising rates for fear that it didn't have the firepower to quickly reverse a recession if it went too far in raising rates.

Now in his August 26, 2022 Wall Street Journal essay “Can Central Banks Maintain Their Autonomy?” I see that Nick Timiraos, the doyen of monetary policy journalists, takes a similar view of why the Fed and other central banks were behind the curve in raising rates. He writes:

Finally, the pandemic struck just as U.S. and European central bankers were concluding reviews of the policy frameworks they had used to address the problems bedeviling their economies since the 2008 crisis. They wanted to avoid a rut of slow growth and low inflation that would cripple their ability to stimulate the economy in a downturn. They had seen Japan struggle to escape that trap for most of the previous two decades, even after cutting interest rates to zero or below.

At Jackson Hole in 2020, Mr. Powell unveiled an overhauled policy framework. The Fed would aim to return inflation not just to its 2% target but rather to a level a bit above it, so inflation would average 2% over time. For the preceding decade, “We could not get inflation up to 2%,” said Fed governor Christopher Waller. “This sounds crazy. I used to always joke, ‘Go get the guys from Argentina. They know how to do it.’”

Part of the reason Fed officials waited too long to react to the recent surge in inflation is that “they wanted to be sure” that the problem of weak growth and inflation had been vanquished, said Mr. Rajan. “Imagine the hue and cry,” he added, “if they had raised rates immediately: ‘Why are you killing a sound economy?’”

This was in part, a matter of generals fighting the last war rather than the war they are in. But it was more than that. If you thought that fighting the war you are in would destroy all of your hardware for fighting a war like the last war, you could be clear about the war you were in and still hesitate to do what would otherwise be called for.

That is where having a full strength negative interest rate policy in reserve could help. If you know you can cut interest rates as much as necessary to quickly end any recession caused by inadequate aggregate demand, then you can raise rates with the confidence you can get back on track if you make a mistake and overdo it. If you are too afraid to overdo interest rate hikes, you are likely to underdo them. That is what happened.

The Fed has had the opposite problem in the past. In his August 25, 2012 Slate piece “We Need Inflation-Tolerance, Not Inflation,” Matt Yglesias writes:

Imagine you’re watching an Olympic-quality archer who’s having a very bad day. You notice that not only is his overall score much worse than he normally does, but all of his arrows are falling lower than the bullseye. You ask him about it and he tells you that his daughter’s been kidnapped, and the kidnapper says he’ll kill the girl if he shoots a single arrow above the bullseye. Now it all makes sense. The archer hasn’t lost his skill. He’s deliberately aiming too low because he has enormous aversion to shooting above the bullseye. If he lost that aversion, his score would improve.

The moral of the story isn’t that shooting too high is a good way to win an archery tournament. To win, you need to hit the bullseye—neither too high nor too low. But if you become strongly averse to shooting too high, that’s going to undermine your ability to hit the bullseye.

That’s the situation I think American monetary policy is in. It’s not that three or four percent inflation is such a wonderful goal. It’s that extreme aversion to three or four percent inflation is causing the Federal Reserve to persistently “shoot too low” in terms of aggregate demand. Ben Bernanke’s acting as if someone’s holding his daughter hostage. Specifically, the reigning dogma is that if inflation were to go from 2 percent to 3 or 4 percent that long-term expectations might become “unanchored” and drift higher and higher, undermining the “hard won gains” of the Volcker years. But there’s no empirical evidence that this is true, and no particularly strong theoretical reason to believe than the worst-case scenario if inflation tolerance goes wrong is worse that the current strategy of grinding the recession out by letting America’s long-term productive capacity collapse.

In the current context, it is easier to think about trying to hit the moving target of the “right” interest rate. But Matt Yglesias is correct that being too afraid of aiming high will make you aim too low. And being too afraid of aiming low will make you aim too high. The Fed and other central banks need to have tools they are confident can do the job in either direction. Then they can immediately take rates to the level that is their best guess of what is right for the battle they are in, knowing they can swiftly pivot if another threat emerges.

Coda: In the first part of our “The Future of Inflation” trilogy, “Will Inflation Remain High?” Ruchir and I write:

While advanced economy central banks may continue to dislike inflation, their current apparent plans—according to their current dot plots (or the equivalent)—may be behind the curve on what would be required to bring inflation back down. Standard Taylor Rule calculations suggest that it could easily take interest rates as high as 7 percent in several countries to bring inflation down.

This is above the level currently expected by markets. Matt Grossman’s August 26, 2022 news article “Bond Yields Inch Higher After Powell Says Fed Will Hold the Line on Inflation,” reports:

Earlier in the summer, market-based forecasts showed that many investors were second-guessing whether the Fed would wind up raising interest rates as much as central bankers have projected.

In late July, the pricing of derivatives called overnight index swaps estimated the Fed’s benchmark rate would peak at around 3.3% in early 2023 before moving lower again.

At the time, that was a much different outlook than the one espoused by Fed officials themselves, who have generally said they expect rates will have to hold steady or climb next year to corral inflation.

In recent sessions, traders’ forecasts have moved closer to the Fed’s. On Friday, derivatives traders were now projecting that rates will rise to nearly 3.8% by the end of the Fed’s May 2023 policy meeting.

3.8% is still a lot lower than the 7% or so I think will be necessary to bring inflation back to the 2% per year target. Since I believe the Fed is committed to getting inflation under control, I predict further market surprises as market participants and the Fed itself realize how high rates need to go.

See also:

An Example of Needing to Worry about Reverse Causality: Satisfaction with Aging and Objective Aging Outcomes

Though there are some places where the wording is cautious, there is a strong desire in this journal article, and the associated news articles, to make the interpretation that having a better attitude toward aging will result in better objective aging outcomes. That might be true, and seems likely to at least be true at a low effect size. But the “association” or correlation between aging satisfaction (the measuring of attitude toward aging) and objective aging outcomes could easily be an instance of people having a sense of their own health status that goes beyond easily tallied objective outcomes and having a worse attitude toward aging because they already know their health is poor.

I don’t see how the effect of attitude toward aging on objective aging outcomes can be identified without a randomized controlled trial of an intervention trying to improve attitudes toward aging. Note that getting a placebo here is a little tricky. It requires something like an intervention to change an attitude thought to be irrelevant to aging outcomes.

If one considers underlying health status to be distinct from easily tallied health outcomes, then one could see this as “Cousin Causality” from underlying [health status] affecting both [aging satisfaction] and [later easily tallied bad health outcomes]. A virtue of seeing this as an instance of cousin causality (sometimes called “third-factor causality”) as opposed to reverse causality is that it makes clearer that the aging satisfaction being measured earlier in time than the easily tallied health outcomes doesn’t mean the aging satisfaction caused the health outcomes. The ancestor factor affecting both can easily be earlier in time than both. Then on one branch we are saying that poor underlying health begets bad easily tallied health outcomes, which seems like a truism.

In the conclusion, the authors Julia S. Nakamura, Joanna H. Hong, and Jacqui Smith write:

… there is potential for confounding by third variables. However, we addressed this concern by implementing a longitudinal study design, robust covariate adjustment, and E-value analyses.

“Longitudinal study design” could mean two things: one, having aging satisfaction measured before easily tallied health outcomes. That doesn’t solve a problem of “confounding by third variables,” which is another way of describing cousin causality, for the reason I said above: the third ancestor variable can easily be before both aging satisfaction and easily tallied health outcomes. The other thing “longitudinal study design alludes to is a focus on changes in aging satisfaction. But changes in aging satisfaction can easily be due to changes in underlying health status, and everything I say above goes through. “Robust covariate adjustment” doesn’t solve the problem of confounding because subjective health measures contain information that goes beyond all of the other health measures in the Health and Retirement Study (HRS); research shows that subjective health measures add predictive power for later easily tallied health outcomes beyond other covariates in the HRS. I am suggesting that aging satisfaction has to be considered as being analogous to the subjective health measures in the HRS. Note that doesn’t mean controlling for subjective health would solve the problem either, because subjective health is surely measured with error. (See “Adding a Variable Measured with Error to a Regression Only Partially Controls for that Variable.” Having aging satisfaction as a second measure of subjective health ought to increase predictive power over only one measure of subjective health. Finally, the “E-value analyses” simply say that the story for why there might be confounding would have to be one that makes a lot of confounding plausible. I think that is satisfied by the story I give above.

A Linear Model for the Effects of Diet and Exercise on Health is a Big Advance over Popular Thinking

It is easy to take for granted what one is good at. (On this, see “How to Find Your Comparative Advantage.”) Economists often don’t appreciate the sophistication of the models they deal with in comparison to common ways of thinking. I point out to my undergraduate students that, relative to “P affects Q and Q affects P; it is the seamless web of history that can’t be untangled,” supply and demand is a big advance. And I point out to my graduate students how holding a co-state variable fixed as a step in the analysis of a phase diagram even when it won’t be fixed in the end is fully analogous to holding price fixed to develop supply and demand curves, that then make it possible to cleanly analyze equilibrium effects in which price is not fixed.

Another key example is this: relative to the idea that there are only two possibilities, A or B, that C is the one cause of D, or that either E or F are enough to get from one of two states to another, a model in which two different things have a linear effect on an outcome is a big advance. The two articles flagged above show the value of even a simple linear model for understanding, diet, health and exercise. Both eating right and exercising are valuable for health, and doing one doesn’t reduce the marginal product of the other that much. And the curvature in the effect of exercise on health isn’t big enough to keep more than the usually recommended amount of exercising from having a substantial positive effect.

Of course, the linear model, while almost always a good approximation locally, almost always breaks down eventually due to curvature. On that, see “The Golden Mean as Concavity of Objective Functions.” But note that a model with gradual curvature is strictly more complex than a linear model. A linear model is a key stepping-stone on the way to understanding a model with gradual curvature.

When I previewed the content of this post for one of my friends, the reaction was that this sounded like looking down on others. I don’t view it that way. We need to see what we are good at and what is hard for others in order to do the most good in the world. And conversely, we need to see what others are good at that is hard for us. And of course, we need to see what we and others are relatively good at. That is the essence of comparative advantage. It is a matter of mathematics that everyone has some comparative advantage! (At least weakly.)

Recognizing your own technical advantages makes it clearer what you have to offer, and the efforts you will need to make to explain an idea clearly given where your listeners are coming from. One of the best pieces of advice I ever got was one of the excerpts I was given from letters of evaluation when I came up for tenure. One writer said “Miles sometimes overestimates his audiences.” The translation is that what I was saying was obscure because I hadn’t given the appropriate background or had gone too fast.

People can be offended by being treated as if they have less background than they do, and on the other hand, they can be confused by being treated as if they have background that they don’t. There is no way around the need to get it right.