In a recent column, I made the case that the Federal Reserve’s efforts to get inflation under control will inevitably result in a hard landing for the economy. But that leaves an important question unanswered: When will the recession happen?

My best guess: Probably not this year, more likely in 2023 or 2024. And the later the downturn comes, the worse it will be.

Why no recession this year? For one, the economy has considerable momentum. Ample hiring and wage gains are boosting nominal income. Household finances are in excellent shape: Debt service is at a record low relative to income, and pandemic-related government payments have bolstered savings. All the more reason for consumers to keep spending. And in sectors such as housing and motor vehicles, demand still exceeds supply, which has a lot of catching up to do. The cumulative shortfall of motor vehicle production relative to demands, for example, stands at several million vehicles. So even if demand falters a bit, output will increase as supply-chain constraints ease.

Also, monetary policy and financial conditions remain very accommodative. The federal funds rate is unlikely to reach what officials deem a neutral level—about 2.5%—until late this year, given the Fed’s late start and the near-zero starting point. Moreover, if price inflation remains high and wage inflation continues to climb, the neutral rate will be a lot higher than 2.5%, leaving the Fed with much more work to do.

So if not this year, then when? That will depend on how the Fed responds to economic developments. Consider, for example, the real possibility that year-over-year inflation readings decline quickly from the 8.5% peak they reached in March—as the prices of oil, gasoline and cars come back down, supply chains untangle and demand reverts back to services from goods. Will this good news be enough to keep the Fed from making monetary policy tight?

Such a delay would probably have two main consequences. First, the near-term risk of a recession would decrease. Second, the labor market would remain very tight and this would prevent inflation from falling back to the Fed’s 2% objective. Underlying inflation would actually keep rising, driven by higher wages and expectations of more persistent inflation. Ultimately, the Fed would eventually have to step in with even tighter monetary policy than initially contemplated, precipitating a deeper recession.

This is what happened in the late 1960s and 1970s. Inflation wasn’t tackled forcibly early, so it kept ratcheting upward. Eventually, then-Fed Chair Paul Volcker had to force a severe recession to get inflation back under control.

The Fed’s choice is clear. If it acts sooner, with inflation expectations still well anchored, the cost in terms of foregone output and higher unemployment should be relatively modest. If it waits and allows inflation expectations to get out of hand, the bill will be much higher.

Bill Dudley, a Bloomberg Opinion columnist and senior adviser to Bloomberg Economics, is a senior research scholar at Princeton University’s Center for Economic Policy Studies. He served as president of the Federal Reserve Bank of New York from 2009 to 2018, and as vice chairman of the Federal Open Market Committee. He was previously chief U.S. economist at Goldman Sachs.