Whassup with those economists who predicted a recession that then didn’t happen?

In a recent column entitled “Recession was inevitable, economists said. Here’s why they were wrong,” Gary Smith writes:

In an August 2022 CNBC interview, Steve H. Hanke, a Johns Hopkins University economics professor, predicted: ‘We’re going to have one whopper of a recession in 2023.’ In April 2023, he repeated the warning: ‘We know the recession is baked in the cake,’ he said. Many other economists also anticipated a recession in 2023. They were wrong.”

I am not an expert on monetary policy or economics. Rather, this story interests me as a political scientist, in that policy recommendations sometimes rely on academic arguments, and also as a student of statistical workflow I am interested in how people revise their models when they learn that they have made a mistake.

Along those lines, I sent an email to Hanke asking if he had written anything addressing his error regarding the recession prediction, and how he had revised his understanding of macroeconomics after the predicted outcome did not come to pass.

Hanke replied:

Allow me to first respond to your query of January 23rd. No, I have not written up why my longtime colleague John Greenwood and I changed our forecast concerning the timing of a likely recession. But, given your question, I now plan to do that. More on that below.

In brief, Greenwood and I employ the quantity theory of money to diagnose and predict the course of the economy (both inflation and real GDP growth). That’s the model, if you will, and we did not change our model prior to changing our forecast. So, why was our timing on the likely onset of a recession off? After the onset of the COVID pandemic, the money supply, broadly measured by M2, exploded at an unprecedented rate, resulting in a large quantity of excess money balances (see Table 2, p. 49 of the attached Greenwood-Hanke paper in the Journal of Applied Corporate Finance). We assumed, given historical patterns, etc., that this excess money would be exhausted and that a recession would commence in late 2023. Note that economic activity is typically affected with a lag of between 6 and 18 months after a significant change in the money supply. The lags are long and variable, sometimes even shorter than 6 months and longer than 18 months.

We monitored the data and realized that the excess money exhaustion was taking longer than we had originally assumed. So, we changed our forecast, but not our model. The attached Hanke-Greenwood article contains our new forecast and the reason why we believe a recession is “baked in the cake” in late 2024.

All this is very much in line with John Maynard Keynes’ quip, which has become somewhat of an adage: “When the facts change, I change my mind. What do you do, sir?”

Now, for a little context. After thinking about your question, I will include a more elaborate answer in a chapter in a book on money and banking that I am under contract to deliver by July. That chapter will include an extensive discussion of why the quantity theory of money allowed for an accurate diagnosis of the course of the economy and inflation during the Great Financial Crisis of 2008. In addition, I will include a discussion of how Greenwood and I ended up being almost the only ones that were able to anticipate the course of inflation in the post-pandemic period. Indeed, in 2021, we predicted that U.S. headline CPI would peak at 9% per year. This turned out to be very close to the 9.1% per year CPI peak in June 2022. Then, the Fed flipped the switch on its monetary printing presses. Since March 2022, the U.S. money supply has been falling like a stone. With that, Greenwood and I forecasted that CPI would end 2023 between 2% and 5% per year. With December’s CPI reading coming in at 3.4% per year, we hit the bullseye again. And, in this chapter, I will also elaborate on the details of why our initial prediction of the onset of a recession was too early, and why the data have essentially dictated that we move the onset forward by roughly a full year. In short, we have moved from the typical short end of the lag for the onset of a recession to the long end.

Again, macroeconomics is not my area of expertise. My last economics class was in 11th grade, and I remember our teacher telling us about challenges such as whether checking accounts count as “money.” I’m sure that everything is a zillion times more complicated now. So I’ll just leave the discussion above as is. Make of it what you will.

P.S. Since writing the above I came across a relevant news article by Jeanna Smialek and Ben Casselman entitled, “Economists Predicted a Recession. So Far They’ve Been Wrong: A widely predicted recession never showed up. Now, economists are assessing what the unexpected resilience tells us about the future.”