The 'Balancer' Trap for Commentators; or, How I Got Inflation Wrong
Jerusalem Demsas joins the Very Serious podcast to talk about bad economic prediction-making, including mine.
Jerusalem Demsas, who covers economics for The Atlantic, wrote a really interesting article earlier this month about all the early pandemic economic predictions that turned out to be totally wrong. You may remember some of them: home prices were going to crash, as many as 10% of Americans were going to be evicted from their homes, there would be crises in state and local government budgets, and we’d see a “she-cession” of women getting pushed out of the workforce due to child care and other care needs.
None of these forecasts panned out.
As Jerusalem notes in her article, we can attribute some of that to policymakers recognizing those risks coming down the road and acting to stop them. (This is the usual story told about Y2K: It wasn’t a fake crisis, but a real one that we successfully prevented.) A major reason state governments are now in surplus rather than deficit is that the federal government gave them a ton of money to keep them out of deficit (too much money, as I’ll discuss in a moment.)
But most of the story is less favorable to the prognosticators.
As Jerusalem and I discussed on this week’s Very Serious podcast (and I have discussed frequently in this newsletter), one key problem was people with real expertise looking at the crisis before them and seeing what they wanted to see — or rather, seeing what would make a strong argument for the policies they already wanted to argue for. She thinks (and I agree) that this was more self deception than conscious deception: For example, if you worry about the social cost of evictions, you’re likely to look at a situation and see it crying out for policies to reduce evictions — and that could cause you to expect a lot more evictions than were ever actually likely.
And because a lot of the relevant experts — knowledgeable journalists, and the academics and policymakers they talk with — tend to be left of center, a lot of the predictions ended up going wrong in a left-of-center way.
For all those dogs that didn’t bark, there was an unexpected one that did: Inflation, which has soared in a way few commentators expected. Even the people with a reputation for having been “right” in their warnings about inflation last year were really just less wrong — as I discussed with Jason Furman back in February, he was publicly worried about inflation in early 2021, but he wasn’t anywhere close to expecting it to go over 8%.
When I asked Jerusalem about what she’s learned from all the wrong predictions over the last two years to help her better assess the predictions she’s hearing now, she said something that hit close to home for me:
When people view themselves as “balancers” in a dialogue versus trying to find the correct answer, it can create a lot of problems… If your model is ‘Hey, I’m really worried that most macroeconomists are just way too concerned about inflation. Maybe a lot of rich people care more about inflation than they do about unemployment because they’re not going to be affected by it, so in general, my goal in the public discourse is to push against this.” If you see yourself as a ‘balancer’ in that way, which I think a lot of activists do — and potentially that makes sense for their role… — it can often make it hard to change your mind.
I’ve certainly been guilty of that over… well, I guess my entire career in media. And up until the last 18 months or so, it made me right about inflation. For years coming out of the Great Recession, the Federal Reserve had been tending to under-shoot its inflation target. We’d gone through a long period when the labor market had usually been too slack, suppressing wage growth, making it too hard for workers to get ahead, and not doing economic growth any favors along the way. Worrying too much about inflation had been translating directly into not having enough jobs.
Funnily enough, these conditions seemed to be finally improving under President Trump. However craven his reasoning, the “low interest rate guy” president urged the Fed in the same dovish direction I wanted it to go, labor markets tightened, and wage growth picked up. I don’t know how responsive the Fed was to his tweets, but he did make personnel choices for the Fed — chiefly, Jay Powell — who took it in a direction that more heavily emphasized job creation.
Then, the pandemic hit, unemployment soared, and the Fed responded with a huge (and appropriate) intervention to stabilize financial markets and stimulate the economy. The Fed also codified a policy change that had been in the works before the virus emerged — “average inflation targeting,” a policy that gives the bank more flexibility to let inflation run over the 2% target for significant periods, in pursuit of 2% as an average over a longer time horizon. As the economy rapidly recovered and inflation crept up, a key question was how quickly the Fed should react — was this inflation just “transitory”? Was it just being driven by supply-side factors outside the Fed’s purview, such as port bottlenecks? Or was this a sign of economic overheating calling for a traditional response of significant monetary tightening, through higher interest rates and a cessation of asset purchases?
The Fed took a soft touch, even in its rhetoric, until late 2021. And I thought that was right. But in retrospect, it wasn’t.
I was so used to the idea that “the conversation” was too hawkish on inflation that I was too slow to take inflation concerns seriously when inflation became seriously concerning. Like a lot of people more qualified than I am, I was too interested in granular, supply chain-related explanations for inflation in 2021 when I should have been focused on the bigger picture: nominal GDP was growing too fast, a phenomenon that had to be driven by demand-side forces related to people having too much money chasing too few goods and services. The economy was overstimulated because the American Rescue Plan was too big (including those too-large payments to states that put them into surplus) and because the Fed waited too long to start tightening policy. And I probably would have seen that faster if I wasn’t emotionally invested in being a dovish guy on inflation.
So, first of all, sorry. I will try to do better! But I also think Jerusalem is right that this behavior is very common — and pretty understandable, since there very often are systematic errors in policy conversations that one might notice, object to, and try to correct. Still, that frame can trap you in the wrong answer in times like the ones we’ve recently been through, leading to highly consequential errors. You’ll do better at understanding what’s going on if you can decouple your normative and emotional views on what sort of society we should have from your positive assessments of what’s actually going on.
I hope you enjoy the podcast. If you have responses, please leave them in the comments.
It's worth also noting that being an expert in a field doesn't make you a good forecaster! This is most famously written about in Tetlock's "Superforecasting" -- who would make a fascinating guest!
I buy that that the ARP exacerbated the inflation problem, but it seems like inflation is becoming a problem all over the world (including in countries with far less generous fiscal responses to COVID) while US inflation is still mostly being blamed on US stimulus. Has anyone studied how much inflation can be attributed to the ARP?