Apart from the hefty gain of 517,000 jobs added to the economy in January, there was another surprise buried deep in the report the Bureau of Labor Statistics released last Friday. → www.bls.gov/...
Every year, when the Bureau releases the jobs numbers for January, it also revises the numbers reported in earlier months. The yearly revision is extended backward over a number of years which smooths out the effect. Usually, the revisions are small and hardly worth mentioning. But this year and last year, there was something unusual. The revisions were bigger than anything the BLS reported since it started tracking them in 1979.
In fact, the number of jobs gained in 2021 and 2022 was 12,081,000, or 2,131,000 more than what was originally reported.
Imagine the press and the media on the morning of September 3, 2021, rushing to report the new jobs numbers, as they do on the first Friday of every month.
From the New York Times: “We saw disappointing job growth in August. Employers added just 235,000 jobs, far below economists’ expectations.” www.nytimes.com/...
A month later, that number was revised upward to +366,000. Two months later, it was revised upward again to +483,000. In the yearly revision reported in February 2022, it was revised upward to +517,000. And in the revision reported last Friday, it was revised upward, again, to +663,000.
A similar pattern repeated quite a few times in 2021 and 2022.
For example: in October 2021, the Washington Post reported:
“The US added just 194,000 jobs in September, disappointing results that raised fresh concerns about the economic recovery.” www.washingtonpost.com/...
A month later, that number was revised upward to +312,000. Two months later, it was revised upward again to +379,000. In the yearly revision reported in February 2022, it was revised upward to +424,000. And in the revision reported last Friday, it was revised upward, again, to +557,000.
In December 2021, the media reported, once again, on a “disappointing” jobs report.
CNBC, for example, said:
“Job growth disappoints in November, with a gain of just 210,000.” www.cnbc.com/...
A month later, that number was revised upward to +249,000. The usual revision expected two months later coincided with the February 2022 revision so instead of two revisions, there was one, which was +647,000. And in the revision reported last Friday, it was revised downward, to +614,000.
Even with the downward revision, the number released last Friday is still almost triple the number originally reported.
In January 2022, the media reported again on a bad jobs report.
ABC News said:
“Employers add dismal 199,000 jobs in December.” abcnews.go.com/...
The revision usually expected a month later coincided with the February 2022 revision, so instead of two revisions, there was one, which was +510,000. Two months after the original report, it was revised upward again to +588,000. In the yearly revision reported in February 2023, the number was revised downward, slightly to +569,000.
Even with the downward revision, the number released last Friday is still far above the number originally reported.
What to make of this striking pattern in the revised jobs numbers? In November 2021, the Washington Post ran a story about it: The Government Dramatically Underestimated Job Growth This Summer www.washingtonpost.com/… Last Friday’s revisions come on top of the numbers that were already off by a record amount when WaPo looked at them 15 months ago.
The experts WaPo quoted explained the trend. “These waves of revisions in the same direction tend to happen at turning points in the labor market.” In other words, when the economy is rising, the BLS tends to underestimate the jobs number in its initial report. Conversely, when the economy is falling, the jobs number tends to be overestimated.
Ironically, during the first three Trump years, when the economy was mainly flat, the jobs numbers were only revised a little. But looking back on 2020 when the pandemic hit, the revised numbers show that the job loss was 400,000 worse than what was reported at the time.
The Post aside, it doesn’t seem that the press and the media pay much attention, if any, to revisions in the jobs numbers. They never did in earlier years when the revisions were small. And they don’t now, either, when the revisions are significant. Think of all the pundits who appeared with their explanations and analyses of the “disappointing” jobs numbers and how they shaped public opinion, based on a number that wasn’t final. Remember that when they chatter today about the public perception which doesn’t agree with data that shows a strong economy. They helped to create that perception. And they aren’t doing much, now, to inform the public more accurately.
Of course, on every first Friday of the month, journalists and economic experts can only comment on the jobs numbers that are reported. They would say that’s all they did. But anyone who reports on economic statistics is expected to know that numbers are revised. And just as the Bureau of Labor Statistics checks and doublechecks its work, and revises it when new information becomes available, journalists should follow suit, checking and upgrading the quality of their work, too.
