The Wisdom of the Crowd? Firm Expectations, Disagreement, and Real GDP Forecast Accuracy
October 15, 2026
Summary
We compare probabilistic real GDP growth forecasts from the CFO Survey with those from the Survey of Professional Forecasters (SPF). Both surveys elicit full probability distributions, so in addition to forecast accuracy, we can measure subjective uncertainty within forecasters and disagreement across them. CFO forecasts are substantially more dispersed than SPF forecasts: cross-sectional disagreement among CFOs is roughly three times as large as among professional forecasters. At the individual level, CFO respondents exhibit larger absolute errors and higher forecast-error variance. Yet this greater individual-level noise does not translate into worse consensus forecasts. Aggregated CFO forecasts track SPF forecasts closely and achieve statistically similar consensus accuracy. Decompositions of ex-ante forecast variance and ex-post mean squared error show that CFO disagreement reflects heterogeneous beliefs and that much of their individual forecast error is idiosyncratic, allowing aggregation to reduce noise. A within-quarter randomization exercise quantifies this wisdom-of-crowd effect directly: the consensus forecast error declines with crowd size with an elasticity of –0.17, evidence of genuine aggregation gains. Dispersed firm-level information can thus complement professional macroeconomic forecasts, though consensus accuracy is ultimately bounded by shared errors rather than crowd size.
View PaperWorking Paper 2026-17
Abstract: We compare probabilistic real GDP growth forecasts from the CFO Survey with those from the Survey of Professional Forecasters (SPF). Both surveys elicit full probability distributions, so in addition to forecast accuracy, we can measure subjective uncertainty within forecasters and disagreement across them. CFO forecasts are substantially more dispersed than SPF forecasts: cross-sectional disagreement among CFOs is roughly three times as large as among professional forecasters. At the individual level, CFO respondents exhibit larger absolute errors and higher forecast-error variance. Yet this greater individual-level noise does not translate into worse consensus forecasts. Aggregated CFO forecasts track SPF forecasts closely and achieve statistically similar consensus accuracy. Decompositions of ex-ante forecast variance and ex-post mean squared error show that CFO disagreement reflects heterogeneous beliefs and that much of their individual forecast error is idiosyncratic, allowing aggregation to reduce noise. A within-quarter randomization exercise quantifies this wisdom-of-crowd effect directly: the consensus forecast error declines with crowd size with an elasticity of −0.17, evidence of genuine aggregation gains. Dispersed firm-level information can thus complement professional macroeconomic forecasts, though consensus accuracy is ultimately bounded by shared errors rather than crowd size.
JEL classification: C53, D84, E37
Key words: firm expectations; professional forecasters; GDP forecasting; wisdom of crowd; CFO Survey; Survey of Professional Forecasters
https://doi.org/10.29338/wp2026-17
Brent Meyer is a vice president and senior economist in the Research Department of the Federal Reserve Bank of Atlanta.
Emil Mihaylov is a quantitative research analysis specialist in the Research Department of the Federal Reserve Bank of Atlanta.
Michael A. Navarrete is an assistant policy adviser and economist in the Research Department of the Federal Reserve Bank of Atlanta.
Daniel Weitz is a survey director in the Research Department of the Federal Reserve Bank of Atlanta.
The views expressed here are those of the authors and do not necessarily reflect the views of the Federal Reserve Bank of Atlanta, the Federal Reserve System, the Board of Governors, or their staff. All errors are our own.
Federal Reserve Bank of Atlanta working papers, including revised versions, are available on the Atlanta Fed's website at www.atlantafed.org. Click "Publications" and then "Working Papers." To receive e-mail notifications about new papers, go to subscribe.