Volcker, 1979 to 1982
Paul Volcker was sworn in as Fed chair on August 6, 1979. CPI inflation was running at 11.8 percent and rising. On October 6, on a Saturday, the FOMC announced a change in operating procedure: the Fed would target nonborrowed reserves rather than the fed funds rate, and tolerate large rate volatility to bring monetary aggregates under control. The fed funds rate, which had been around 11 percent in September, ran above 17 percent within four weeks. The effective fed funds rate peaked at a monthly average of 19.10 percent in June 1981.
The 1981-82 recession that followed was one of the deepest postwar US downturns. Unemployment hit 10.8 percent in November and December 1982. Manufacturing employment fell by about 2 million jobs. The savings-and-loan industry, which had been holding long-duration mortgage assets funded by short-term deposits, was effectively rendered insolvent in aggregate; the FSLIC absorbed the losses over the next decade. Builders mailed two-by-fours to the Eccles Building in protest. Tractors blockaded the Fed garage.
CPI inflation peaked at 14.8 percent in March 1980 and was down to 3.8 percent by December 1982. Long-run inflation expectations became more anchored during the following expansion, and inflation did not return to double digits in the next twenty-five years. The Volcker disinflation is a central case study for the proposition that ending entrenched inflation can require credible willingness to accept real costs. Romer and Romer's narrative-shock dating treats the October 1979 FOMC minutes as a cleanly identified disinflationary intervention in postwar U.S. policy.
Flexible Average Inflation Targeting, August 2020
On August 27, 2020, Jerome Powell delivered the Jackson Hole speech that announced the Fed's new framework. Three things changed in the Statement on Longer-Run Goals. The 2 percent inflation target became an average rather than a point: after a period of inflation below 2 percent, the FOMC would aim for inflation moderately above 2 percent for some time. The shortfalls language replaced the symmetric deviations language: the Fed would target shortfalls of employment from its maximum level, not deviations. The pre-emptive logic was downgraded.
The framework was designed for the world the Fed thought it was in: secular stagnation, a depressed neutral rate, and a Phillips curve that had appeared dead. The 2018-19 episode in which the Fed raised rates against an inflation rate that never reached 2 percent was treated as a cautionary tale. Larry Summers and Olivier Blanchard had been arguing publicly for years that the Fed's symmetric approach was systematically allowing inflation to undershoot. FAIT was the institutional response.
The framework had bad timing. By March 2021 CPI inflation was running at 2.6 percent and rising. The FAIT logic argued for patience: the new framework had been designed to make up for past undershoots, and the Phillips curve seemed flat. The FOMC kept rates near zero through March 2022. By that point CPI was 8.5 percent. The 2025 framework review retired the makeup element altogether: the revised Statement on Longer-Run Goals drops the average-inflation language and simply reaffirms 2 percent, returning the Fed to flexible inflation targeting. It also removed the shortfalls-only wording: the balanced-approach paragraph now speaks of departures from its goals, which cuts both ways. The episode is a useful case study in the cost of regime commitment when the regime turns out to be wrong about the underlying environment.
R*, the rate that nobody can see
The neutral real rate of interest, R*, is the level of the real policy rate consistent with output at potential and inflation at target over the medium term. It cannot be observed directly. It has to be estimated from a model. The two most-cited estimates are Laubach-Williams and Holston-Laubach-Williams, both maintained by the Federal Reserve Bank of New York, both based on a Kalman-filter extraction from observed output, inflation, and the real rate.
The estimate matters because it is the long-run anchor of the Taylor rule and the central organising parameter of the FOMC's Summary of Economic Projections. The Holston-Laubach-Williams estimate of U.S. R* was about 2.5 percent on the eve of the 2008 crisis and at or below about 1 percent through most of the 2010s. The estimate for the second quarter of 2026, published in August 2026, is 1.0 percent. The standard error band is wide, about 1.5 percentage points. The model also struggles around regime breaks: after the pandemic it had to be re-estimated with a COVID adjustment.
The honest reading is that R* is partly an aggregation of forces an econometrician cannot fully control: demographics, productivity growth, the global savings supply, the composition of safe-asset demand, the Treasury's debt-management choices. It moves slowly when those move slowly. It can also move quickly when one of them changes regime, as the post-pandemic period seems to have shown for the U.S. and the euro area. FOMC members put their long-run dot in the SEP partly to communicate where they think R* sits. The cross-sectional dispersion of those dots is now the best available read on how uncertain the FOMC is about the parameter that anchors everything else.