Officials at the U.S. Federal Reserve are grappling with an important issue: how to deal with the possibility that the economy doesn’t act the way they expect it to. Judging from their analyses, different risks will require different responses — and at least one concern isn’t getting enough attention.

Managing uncertainty was among the big topics of last weekend’s annual economic policy symposium in Jackson Hole, Wyoming, hosted by the Kansas City Fed. Jerome Powell, the new chair of the Board of Governors, emphasized the wide range of uncertainty associated with the Fed’s baseline models. In a separate research paper, a number of high-ranking Fed staffers explored what policy makers should do when they’re not sure how worker scarcity will affect inflation.

Their analyses reveal multiple forms of risk, not all of which imply the same policy response. Consider, for example, the Fed’s primary monetary policy tool, the setting of short-term interest rates. At first glance, the central bank’s plans don’t seem consistent with its goals of achieving 2 percent inflation and maximum employment: It intends to gradually raise rates above their long-term average over the next couple years, resulting in abnormally low unemployment and above-target inflation. Why not aim to hit its target by raising rates faster?

One answer is that the Fed doesn’t know how effective its interest-rate hikes will be (a risk originally studied by William Brainard). If they’re more powerful than expected, inflation will be too low and growth will be less than it could be. It they’re ineffective, there’s not much the central bank can do anyway. So from a risk-management perspective, it makes sense to be cautious about raising rates, to avoid going too far.

But now let’s consider another risk — that the relationship between unemployment and inflation, expressed in what economists call the Phillips curve, will prove different than expected. In the Fed’s view, a low unemployment rate spurs inflation by pushing up wages. But there’s a lot of uncertainty about the magnitude of this effect. If it’s stronger than expected (a steep Phillips curve), then the central bank’s planned interest-rate increases will fail to keep inflation in check. If it’s weak (a flat curve), it doesn’t matter much anyway. So in this case, the Fed is better off erring on the side of action, raising rates faster in the hope of hitting its inflation target.

Yet the staff paper downplayed and Powell ignored what I see as the most important risk: that the U.S. economy could face a recession in the next couple years. As then-chair Janet Yellen’s speech at Jackson Hole two years ago revealed, the Fed lacks tools to deal with such a contingency. The best way to prepare is to ensure that the economy is as strong as possible when the downturn hits. And that requires keeping interest rates lower than the Fed is currently planning to do.

Narayana Kocherlakota is a Bloomberg Opinion columnist. He is a professor of economics at the University of Rochester and was president of the Federal Reserve Bank of Minneapolis from 2009 to 2015.