NRPLUS MEMBER ARTICLE I t wasn’t so long ago that the Federal Reserve provided almost no real-time information to investors, with market participants left to themselves to try to figure out what it was up to. For example, a 2006 academic article written by a Fed economist dug into the controversy over which year the Fed switched from targeting the money supply to targeting the federal funds rate, a policy change it had never announced. After careful study of transcripts, the paper concluded that the Fed had switched its policy target in . . . 1982. It took more than 25 years for academics to officially confirm the switch.
Nowadays, of course, the Fed is as transparent as can be. Its policy targets are announced out loud, and its future policy for the next few meetings is signaled well ahead of time. The information published by the Fed does not just focus on policy. On a quarterly basis, Federal Open Market Committee members report their forecasts for real GDP, the unemployment rate, the inflation rate, and the path of the policy rate.
Transparency only has value and meaning if the signals given by the Fed about the future have information content. If the Fed’s economic and policy forecasts foretell the future, then they do serve transparency and help citizens make better economic decisions. If the Fed is going to keep rates low for a long time, for example, then there is no rush to get a mortgage today. If it thinks rates are going up steeply, then you better hurry up and buy the house. But if the Fed’s outlook is highly inaccurate or biased, it could significantly harm consumers and investors who take its “transparency” seriously.
A new paper by economist Mickey Levy documents how truly awful Fed signaling has been in recent years. Levy assembles a history of Fed quarterly projections and compares these to the actual outcomes. The tables are, to say the least, about as ugly as economic results could be.
In September 2020, the Fed expected 2021 inflation (as measured by the PCE) to be 1.7 percent, with inflation climbing to 1.8 percent in 2022 and 2.0 percent in 2023. Given that the Fed’s target is 2, one can say that the Fed had a high confidence it could achieve its target inflation.
By June 2021, it was apparent to all that deficit spending was spinning out of control, and inflation was picking up as well, having climbed all the way to 4.4 percent. How did the Fed respond to this spike in inflation? It told everyone it was “transitory” and worked that observation into its forecast. Remarkably, at that June meeting, the Fed predicted that 2022 inflation would drop all the way down to 2.2 percent in 2023, and stay there in 2024.
By September 2022, inflation had climbed to 6.4 percent, meaning that the Fed’s forecast for 2022 was wildly inaccurate. But even then, the Fed expected inflation to drop to 2.8 percent in 2023, and all the way to 2.0 in 2024.
These terrible forecasts, Levy documents, had a huge effect on monetary policy, with the Fed pulling the trigger on tightening far too late. Levy concludes, “These forecasting mistakes contributed to the biggest monetary policy error and the highest inflation since the 1970s.”
In a comment on Levy’s paper, my Hoover colleague Steve Davis has wondered if the problem with the Fed might not be a good deal more complicated than just having lousy Keynesian models. Davis takes Levy’s discussion to its logical conclusion and finds an internal conflict in Fed communications that may well give “transparency” a fatal blow.
The Fed, it seems, believes that consumer expectations about future inflation have a big effect on whether inflation takes off in a sustained way. If expectations stay “grounded” at 2 percent, then workers won’t ask for wage increases, and a wage-price spiral will never take off. But that focus on expectations gives the Fed an incentive to use its forecasts of future inflation to attempt to engineer a self-fulfilling prophecy. If the Fed says inflation will be low, and people believe it, then inflation might stay low. If the Fed’s models say inflation is lifting off, the Fed could make the liftoff more serious if it says it believes it will happen. Perhaps it is better to just lie and hope for the best.
To be clear, Davis does not accuse Fed officials of dishonest behavior in this episode and chalks the “transitory” debacle down to confirmation bias. But the conflict between managing expectations and telling the truth is an inherent conflict in Fed communications, one without an easy resolution. And Davis makes the convincing case that the temptation to “shade” forecasts in the desired direction could negatively affect policy itself, as the Fed both loses credibility and struggles to achieve its improbable outlook.
Baseball would never consider a rule that required the pitcher to announce the type of pitch before each throw. The Fed should stop publishing its forecasts, and go back behind the curtain.