In multiple ways, this is the most difficult time we have ever seen to make a forecast. “Unprecedented” actions by government – locking down the economy, printing, borrowing, and spending trillions of extra dollars – artificially boosted economic activity. Like giving morphine to an accident victim, printing and borrowing masked the pain of lockdown injuries. As these artificial actions wear off, we expect a recession to appear.
And in the fourth quarter, retail sales, industrial production and other data suggested that the economy was hitting a wall.
Then…January happened. Nonfarm payrolls, retail sales, and manufacturing production all surged.
But we think these reports overstated economic activity. The US had unusually warm January weather. In addition, seasonal adjustment factors played a key role in making the economy look better than usual, as well.
In January, the national average temperature in the “Lower 48” states was 35.2 degrees, the fourth highest for any January in the past thirty years. New home foundations could be dug, fewer plants closed due to weather, and more people could comfortably be out and about.
Normal seasonal adjustment factors also played a role. Before seasonal adjustments, nonfarm payrolls fell 2.5 million in January. After adjusting, they were reported up 517,000 for the month. Before adjusting, retail sales fell 16.2%, but after the government applied normal seasonal factors, sales were reported up 3.0%, the largest gain for any month in almost two years.
In other words, the reason the government reported that jobs and retail sales were up wasn’t that they actually rose in January relative to December, but that they fell less than they normally do.
No one is manipulating the data, nor are we trying to imply that there is anything illegitimate about seasonally adjusting economic activity. Seasonal adjustments are important. For example, agriculture follows weather patterns and holiday shopping is seasonal. If we didn’t adjust for these patterns, the economy would shrink every year in the first quarter – going back pretty much forever – with a big rebound in the second quarter every year and another surge in the fourth quarter.
And now, because of COVID, government shutdowns, and the fiscal and monetary policy response, the normal seasonal patterns of economic activity have been distorted even more. That means we are probably going to experience some months, like November and December, where activity appears unusually weak, and others, like January, where activity appears unusually strong. The best rule of thumb is to wait for at least a few months in a row of unusual strength or weakness to draw any conclusions.
The bottom line is that a yield curve this deeply inverted is a negative sign for future economic growth. Meanwhile, the M2 measure of money has slowed sharply. The growth of the M2 measure of the money supply was unusually fast through January 2022. In the past year, it is reported as falling for the first time since the 1940s, and at the fastest pace since the Great Depression. If M2 affects “real” (inflation-adjusted) economic output with a lag of a year (give or take) then that support for activity likely peaked very early this year and should dwindle sharply by year end.
We obviously hope there is no recession on the way. It’s pretty obvious that the stock market isn’t worried. But January’s economic data aren’t as clear as many might think.
Brian S. Wesbury, Chief Economist
Robert Stein, Deputy Chief Economist
Date: 3/6/2023
