This is the long-run evidence behind the histograms. For every monthly round, we measure how widely the panel disagrees at each of the six forecast horizons. The near-term line sits low — members broadly agree about next quarter — while the far-term lines sit higher, because six quarters out there is far less to anchor on.

Only dispersion is shown. Individual forecasts are aggregated inside the database and never reach the browser, so no panel member's numbers, and no member's identity, appear anywhere on this page.

The history runs from January 2011 and joins two archives at March 2021. Where they overlap, January to March 2021, they agree exactly on every rate and horizon, so the joined series is continuous.

The shaded 2024 stretch is estimated rather than measured: missing answers in those rounds were filled in from similar panel members, which moves the measured disagreement by a few percent on the main rates and by more on thinly forecast ones.

The 3-month series is LIBOR through December 2021 and SOFR from February 2022, when the survey switched. They share a line here but are different rates, so a step at the changeover is the switch, not a change of view.