Updated 2022-02-04
Think back to January 2020. Whispers were floating across the Pacific about a flu-like disease affecting Chinese folks in a town unknown to most Americans. The world seemed fairly stable by historical norms.
Fast-forward three years. Do words like stable or calm or predictable seem appropriate to describe the world in which we live? Put aside the antics of despots, lying politicians, and bigots. Think just about the influencers in your life (and we don’t mean Tik-Tokkers). We mean inflation, housing costs, wages, and the like. Do you get a stable vibe from the media or your colleagues or friends? By comparison to January 2020, we think not.
In times of instability and uncertainty, reliable forecasts for the economy would be a welcome comfort. None of us would expect a perfect track record for any forecaster. Better than even odds of being correct would be helpful.
At least one economist has the courage to acknowledge that the average of forecasts, much less individual forecasts, is not better than a coin-toss. Charles Gascon, a senior economist at the St. Lous Federal Reserve Bank, shared his honesty in a recent paper entitled “Professional Forecasters’ Outlook for 2023 and Caveats Based on Past Performance”. The paper refers to the projections made by approximately 50 economists for the performance of the U.S. economy in 2023.
Gascon observes that the “consensus” (a.k.a., the average) of the predictions of economics professionals is “historically the most accurate”. He summarizes that the consensus GDP growth estimate for 2023 is a meager 0.3%, which is lower than any consensus estimate since 1993. The consensus estimates for inflation, unemployment, and the 10-year Treasury Note yield are 4.0%, 4.4%, and 3.8%, respectively. At the writing of this post, inflation using the CPI had just rung in at 6.5%, unemployment stood at 3.5%, and the 10-year closed at 3.505%. (Regarding the consensus of economists, see also this Quartz article.)
To assess the accuracy of consensus forecasts, Gascon calculated the average of the top 10 and bottom 10 predictions for each economic metric. For example for GDP, the top 10 estimates average 1.3% growth and the bottom 10 average -0.5%. He then did the same calculation for 1993 to 2022 and asked how often the actual GDP fell within the range between the top-10 and bottom-10.
Actual values of real GDP growth were within this range of forecasts less than half the time (46.7%). . . . Based on historical performance, it is essentially a coin toss as to whether actual real GDP in 2023 will fall within th[e] relatively narrow range of –0.5% to 1.3% [of economists’ predictions].
“Professional Forecasters’ Outlook for 2023 and Caveats Based on Past Performance“, Charles S. Gascon in Economic Synopses, 2023 No. 1 (St. Louis Federal Reserve Bank)
In fewer than half of the 30 years, the actual GDP fell within the range of the consensus of economists’ forecasts. Gascon also observed that “forecasters have consistently missed the long-run downward trend in interest rates since 1993”.
We suggest looking for comfort for the future in a community other than forecasting economists. Very wonderful people they may be, but accurate they are not.
Update
We neglected to mention one of the most famous forecasters in American history — Punxsutawney Phil. Unfortunately, the famous groundhog is no better than economists, according to National Centers for Environmental Information. Phil’s accuracy over the past 10 years — 40%. Our search continues . . . .
