Is a recession looming?

The financial press is full of speculation over the prospects for a recession in late 2022 or 2023. While it’s not easy to predict turning points in the business cycle, two factors have at least some predictive power. First, certain economic fundamentals, such as an overheating economy, can make a recession more likely. Second, asset prices in financial markets often provide useful signals before the data show up in slower-moving macroeconomic indicators. Let’s start with the fundamentals.

It is often said that if the Federal Reserve could predict a recession then it could prevent the recession. Strictly speaking, this is only true of recessions that occur when the economy is relatively stable, rather than in an “overheating” economy in need of a contractionary monetary policy to slow inflation. When the economy is not overheated, the Fed could prevent an anticipated recession with enough monetary stimulus (rate cuts and/or quantitative easing) to keep spending growing at a moderate rate.

Today, however, we’re experiencing the latter. The economy is severely overheated. While some of the inflation problem is due to supply constraints, aggregate demand has also been growing at an excessive rate and pushing prices too high, too fast. Nominal GDP is up more than 10 percent over the past 12 months, a rate of growth in spending that is not consistent with the Fed’s 2 percent inflation target. The Fed needs to slow that growth to no higher than 4 percent to get inflation under control.

Here’s where it gets even more complicated. During normal times (such as the late 2010s), a 4 percent nominal GDP growth rate would be consistent with a healthy job market. These are not normal times. The Fed’s overly expansionary monetary policy in late 2021, which included huge infusions of extra money into the economy, has pushed wage inflation up over 6 percent. Nominal GDP growth ultimately determines the revenue available to firms to pay wages, and today, 4 percent nominal GDP growth might not provide enough revenue to maintain current employment levels at a time when wages are rising fast.

If the Fed’s contractionary monetary policy does succeed in reducing nominal GDP growth to roughly 4 percent, one of two things might happen. The best outcome would be for wage growth to slow sharply from current levels, which would allow firms to avoid large layoffs. But if wages continue growing at 6 percent while nominal GDP growth slows sharply, higher unemployment is almost inevitable.

In the past, the Fed has not succeeded in achieving a so-called “soft landing.” Any time the unemployment rate has risen by at least one percentage point, it has always gone on to rise much further — at least two points above the low during the previous boom. America has not had any borderline recessions since the government began collecting unemployment data in the 1940s. Either we’ve clearly had a recession, or clearly avoided one.

Today, market indicators are presenting a mixed picture of the risk of recession, with the market consensus viewing one as increasingly likely but not certain. For instance, while stock prices are down sharply, if there actually were a recession, they would probably fall even further. And while interest rate futures markets show rates declining slightly during 2023, if there were a recession, interest rates would probably fall much more sharply — perhaps to zero.

These facts are certainly no reason for complacency. The patterns we see in the markets, including soaring oil prices, falling stock prices and a flattening yield curve, often occur right before an economic contraction.

The yield curve is particularly interesting. Today, two-year Treasury bond yields are still a bit lower than 10-year T-bond yields. But the Fed’s policy of raising rates is likely to “invert” the yield curve soon — that is, to cause short-term interest rates to rise above longer-term rates. Contrary to popular belief, this is not a foolproof recession indicator, but recessions have often occurred within a year of a yield curve inversion.

One reason to be humble is that economists do not have a good track record in predicting recessions. Indeed, during the 1990, 2001 and 2008 recessions, the consensus forecast did not predict a recession until several months after it was underway. There is even a possibility that we are already in recession today, as the first quarter of 2022 showed negative growth in output. I suspect that first-quarter GDP data will later be revised upward, however, as job growth has been extremely strong.

The inflation data going forward may end up being worse than many pundits expect, while the unemployment rate will do somewhat better than in previous recessions, despite slow growth in real output. In other words, while the jury is still out on a recession, we may face a few years of stagflation.

Scott Sumner is the Ralph G. Hawtrey Chair of Monetary Policy with the Mercatus Center at George Mason University and a professor emeritus at Bentley University.