January 28, 2009. The Federal Reserve, facing the worst financial crisis in generations, announces that interest rates will remain near zero "for some time." By market close, federal funds futures price a different outlook: immediate tightening ahead. Same day, same news: complete disagreement on where monetary policy is going. My job market paper refers to this gap as "Fed–Market interest rate disagreement" and studies how it systematically shapes the transmission of monetary policy.
Gaps between Fed and financial markets' expectations are not an anomaly. And yet, workhorse macroeconomic models operate under the assumption that markets and the Fed share information and form expectations similarly, implying a common policy outlook. I show that this is rarely true. Disagreement about interest rates is a pervasive, persistent feature of the US economy—and, I argue, a very consequential one. It impairs the transmission of monetary policy to the economy: conventional effects on output attenuate, and in states of high Fed-Market disagreement, they can even reverse.
The Gap is Large and Persistent
To study this, I construct a real-time measure of Fed-Market disagreement at each FOMC meeting from 1990 to 2019. For the Fed's expectations, I use the Tealbooks—rich reports prepared by the Fed Board staff before each policy decision. For financial markets, I use federal funds futures contracts, whose prices embed expectations about the future policy path.
The gaps are substantial. Throughout the sample, absolute disagreement averages around 40 basis points and is highly persistent. At extremes, the difference exceeded 250 basis points. These are not small measurement errors. They reflect fundamentally different views about where policy is heading.
Disagreement Weakens Policy
Here is the central finding: monetary policy transmission depends on this disagreement measure.
Using lag-augmented local projections (Montiel Olea and Plagborg-Møller, 2021), I show that when Fed and markets agree, a surprise interest rate hike reduces industrial production with the familiar hump-shaped dynamics that standard theory predicts. But as disagreement rises, this response attenuates. At high levels of disagreement, the effect can flip sign entirely: a tightening shock may increase output rather than reduce it, providing a belief-based explanation for known anomalies in macroeconomics, like the "real activity puzzle."
Quantitatively, moving from median to 75th percentile disagreement—a shift of about 25 basis points—reduces the output response at 12 months by roughly 50 percent. These results hold up even after accounting for uncertainty, adjusting futures prices for risk premia, and dropping the zero-lower-bound years.
This is not noise. It is a systematic pattern that any theory of monetary transmission must confront.
Is this just another information channel story?
A natural reaction is to invoke the "information channel" (Nakamura and Steinsson, 2018): perhaps markets read Fed actions as signals about fundamentals rather than pure policy moves. If the Fed raises rates, maybe it knows something positive about future demand. Under this view, the Fed's information advantage "explains" the puzzling responses. The central empirical corollary is that such superior information should translate into superior forecasts.
I test this directly. Using carefully synchronized data—selecting market expectations at the exact Tealbook cutoff date to align information sets—I conduct accuracy comparisons and information-advantage tests (Hoesch, Rossi, and Sekhposyan, 2023).
The results are stark. Markets outperform the Fed in forecasting the policy rate at horizons of 1 to 8 quarters. Information-advantage tests—essentially asking whether combining Fed and market forecasts improves on either alone—reveal no systematic Fed edge throughout the sample, with only episodic exceptions around the zero lower bound.
The Fed does not forecast better than markets. It doesn't even hold relevant private information to predict the policy rate. The information channel cannot be the whole story.
The Mechanism: A Wedge in Beliefs
If not information, then what? The answer lies in how expectations feed into spending decisions.
In standard macroeconomic models, monetary policy works through intertemporal substitution: higher interest rates make saving more attractive, reducing consumption and investment today. But this mechanism requires households and firms to agree with the Fed about where rates are going.
I show that when we relax the full information assumption – i.e. we allow Fed and markets to have different views on interest rates –, disagreement emerges as a "wedge" in the Euler equation—the fundamental condition governing consumers’ intertemporal choice. This wedge curbs how much aggregate demand responds to policy shocks: when the Fed hikes rates but markets expect a reversal, the perceived path of real interest rates mitigates transmission. In fact, consumption and investment decisions depend on the expected path, not the overnight rate alone: if markets discount the hike as temporary, demand doesn't adjust.
In other words, disagreement acts as an intertemporal friction to monetary propagation.
This mechanism, in theory, is compatible with both attenuation and amplification of policy effects, but it generates predictions about how expectations revise differently in high- versus low-disagreement states. And that's where the data confirm the model, ruling out amplification and validating state-dependent attenuation.
Where Does Disagreement Come From?
Understanding that disagreement matters is one thing. Understanding its sources is another. But—as often in macroeconomics—things are connected.
I propose a way to think about the learning problem of financial markets. A state-space model decomposes disagreement into three primitive channels: (1) disagreement about fundamentals—different views on the state of the economy; (2) disagreement about the policy rule—different beliefs about how aggressively the Fed responds to conditions; and (3) disagreement about forward guidance—how credible the Fed's announced intentions are.
Using an extended Kalman filter, the model estimates the relative importance of each source and how financial markets learn and update on all three dimensions. When disagreement is high, markets discount incoming information more heavily, and policy surprises move expectations less. This creates a vicious cycle: low credibility begets weak transmission, which may further erode credibility. Methodologically, this framework unifies shocks, beliefs, and transmission in one system, illustrating markets' state-contingent expectation formation process.
What This Means for Policy
The Fed's own 2025 framework review identified communication and belief management as first-order challenges. The FOMC itself has noted that when market expectations diverge from the Committee's, questions arise about the effectiveness and credibility of communication (e.g., Mester, 2024; Fischer, 2017).
My findings provide empirical grounding for these concerns. Disagreement is not just noise or measurement error. It is a measurable state variable that affects monetary transmission. Treating it as such—monitoring it in real time, understanding its sources, and designing policy with belief coordination and credibility in mind—is essential for preserving the Fed's clout on the economy.
In short: it is not enough to move rates. Markets must believe it.
To see more of the empirical results and play with different shocks, controls and much more, visit: https://www.amodeo-mindthegap.com/
About the Author
Francesco Amodeo is a Ph.D. candidate in Economics at UC San Diego. He is currently on the economics and finance 2025/2026 job market. To learn more about his research, visit his website: https://sites.google.com/view/framodeo
Social Media Handles:
