Monetary Policy

John Taylor's 1993 Carnegie-Rochester paper ^[1] looked at nine years of Fed behavior and proposed a four-parameter rule that fit the data: take the equilibrium real rate, add current inflation, add half a point for every point inflation runs above target, and add half a point for every point real GDP runs above potential. The rule was descriptive at the start, then prescriptive at industry conferences, then aspirational once Federal Reserve research staff started running it as a benchmark every quarter.

The coefficients are secondary. Greenspan testified to Congress in 1994 that the FOMC was not using the rule and never would ^[2]. The structure is what matters: it forces a policymaker to declare two unobservable parameters before announcing a rate. The neutral real rate \(r^*\) is the level at which monetary policy is neither tight nor loose. Potential output \(y^*\) is the level the economy can sustain without generating inflation pressure. Both move slowly. Both are estimated, not observed. Both can be wrong by enough to invert the rule's verdict on whether policy is tight or accommodative.

The 2018-2019 case is the canonical example. Many published Taylor rule estimates suggested the Fed was at or above neutral. Inflation refused to reach target. Powell paused, then cut ^[2]. The post-2020 framework review accepted that the rule's verdict had been wrong because the inputs had been wrong ^[3]. \(r^*\) had drifted lower than the consensus estimate. The rule itself was fine; the parameter set was off. That is the failure mode any serious user of the rule has to plan for.

The Fisher identity says \(i = r + \mathbb{E}[\pi]\). It is an accounting identity, not a behavioral claim. It tells you nothing about which side moves when the central bank announces a hike. The empirical question is whether a 100 basis-point increase in the policy rate moves \(r\) by 100 basis points (the rule of thumb in the textbook), or whether part of it bleeds into expected inflation through the credibility channel.

Volcker's 1979-1982 disinflation is the cleanest case for the credibility channel ^[4]. Goodfriend and King document that the federal funds rate rose to 19 percent in late 1980 as a result of restrictive monetary policy ^[5]. The ten-year Treasury yield moved with it but resisted the disinflation: while inflation fell from over 10 percent in early 1981 to under 6 percent by mid-1982, the same paper reports that the ten-year bond rate actually increased from around 13 percent to over 14 percent ^[5]. Most of the nominal-rate increase was rebuilding a positive risk premium for holding nominal bonds, not raising the real cost of capital by the same amount. The disinflation worked partly because expectations adjusted before realised inflation did.

The post-2020 case is the contrast. By the time the Fed lifted off in March 2022 ^[6], expectations had already begun to drift. Survey one-year inflation expectations had risen well above the 2 percent target ^[7]. The Fed had to tighten enough to move both the real rate and to cap expectations. The fed funds rate rose by 525 basis points between March 2022 and July 2023 ^[8], and the ten-year real rate moved roughly with the policy rate for the first time in decades. Credibility was being tested, not assumed.

Olivier Blanchard's 2016 paper for the Peterson Institute ^[9], titled "The US Phillips Curve: Back to the 60s?", argued that the estimated slope had collapsed to roughly a quarter of the 1960s value. Janet Yellen used a similar slope estimate in her September 2017 speech to argue the Fed could let the labor market run hot ^[10]. Christopher Waller and Jared Bernstein argued the opposite at different times. The disagreement was always about \(\kappa\).

The slope is hard to identify because the demand shocks are not random. A central bank that targets inflation suppresses the variation in inflation that comes from demand shocks, leaving mostly cost shocks in the residual. Cost shocks move output and inflation in opposite directions; demand shocks move them together. If you regress observed inflation on observed unemployment in a credible regime, the slope you recover is biased toward zero. The curve looks flatter than it is because the regime has been doing its job.

The 2021-2022 inflation surge was a real-time test. Wage Phillips curves, sectoral price Phillips curves, and aggregate Phillips curves all picked up. The slope, properly estimated with instrumental-variables identification of demand shocks (Hazell, Herreno, Nakamura, and Steinsson 2022) ^[11], is closer to the textbook value than the post-2008 reduced-form regressions had implied. The implication for policy is that demand-side inflation can come back faster than the flat-curve view predicts, and that a tightly anchored expectations regime is what lets the curve look flat between shocks.

The federal funds market is now small ^[12]. Daily trading volumes fell sharply after the 2008 crisis as excess reserves flooded the system. A 2022 Federal Reserve FEDS paper documents average daily trading volume of $72 billion when the effective funds rate sat at or below interest on reserves and $67 billion when it sat above, during April 2018 to February 2020, with Federal Home Loan Banks and a handful of other institutions as the active lenders ^[13]. The market that matters for transmission is overnight repo. A 2025 Federal Reserve FEDS Note estimates the gross size of the U.S. repo market at $11.9 trillion in 2024, dwarfing the fed funds market ^[14]. The Fed's policy rate is still defined as the fed funds target range. The market that enforces it is the wider repo complex.

The Open Market Desk publishes the Standing Repo Facility (SRF) rate at the top of the corridor and the ON RRP rate at the bottom ^[15]. Between those two posted rates, the secured overnight rate floats. The Reserve Balances on the Fed's balance sheet act as the buffer; when reserves fall toward what the Desk and the market consider the lowest comfortable level, balance-sheet runoff is paused ^[16]. The 2024 announcement of the QT taper was an explicit attempt to avoid a repeat of September 2019 ^[17].

Forward guidance has become the highest-bandwidth instrument. Each FOMC meeting now produces a statement, the dot plot in March, June, September, and December, the Summary of Economic Projections, and a press conference ^[18]. Markets price the next eighteen months of policy from the dots and the press conference more than from the actual change in the target range. That makes the credibility of central-bank communication a first-order policy variable. Yellen called the SEP a forecast rather than a commitment. Markets read it as both.