During his first year in office, President Donald Trump repeatedly emphasized two metrics as he touted his success in turning the economy around: jobs and stock prices. By those two metrics, his successor Joe Biden is doing a pretty good job himself — even despite employment numbers being consistently adjusted upward after initial news reports.
I will note at the outset that this is not actually a very good way to evaluate a presidency. Presidents have some connection to the health of the economy, certainly, and it’s safe to say that the passage of a major stimulus bill early in Biden’s tenure meant his footprint was slightly larger than normal. Beyond that, though, politicians often like to take more credit for the economy than they deserve. However, these are things on which Trump focused, so it certainly seems fair to evaluate Biden against that same measuring tape.
One of Trump’s trademark moves was figuring out a number that seemed good and then reframing it as the key number that all of the highly respected economists were looking at. So, during his first year in office, he repeatedly pointed to the number of new all-time high prices on the Dow Jones industrial average or the S&P 500. More than a dozen times during his first 10 months in office, Trump hyped how many records the markets had set.
As of Monday, Biden’s doing pretty well on this invented benchmark. The Dow Jones has increased by 16.8 percent during Biden’s administration, a bit under the 18.8 percent expansion seen by the same point under Trump. The Dow has also set fewer record highs, closing at new highs on 41 days of Biden’s presidency compared to 59 during Trump’s.
But the S&P 500 has grown faster under Biden and seen more records: 63 under Biden to 52 under Trump. As Kellyanne Conway might say, “Markets continue to like the policies, action, and vision of @POTUS.”
The Biden team has picked up one habit from Trump, hyping big jumps in employment while ignoring that they followed a massive contraction at the start of the coronavirus pandemic and pointing to absolute numbers in a country that’s seen its population increase steadily. But it is nonetheless the case that the United States has seen the number of people working increase by 5.8 million since last December. That’s a much bigger jump than seen in the first 10 months of Donald “JOBS JOBS JOBS” Trump’s first year in office.
What’s particularly interesting about those jobs numbers, though, is that the initial reports have consistently been adjusted upward in later months. In other words, the Bureau of Labor Statistics (BLS) will announce initial estimates that in following months get tweaked upward, often two months in a row. So the initial news reporting about the economy is based on lower numbers than the BLS eventually determines were actually added.
Again, it’s a weird period. There are always revisions, but generally modest ones and often tweaking the numbers down as often as up. If we look at the evolution of monthly employment data since January 2013, you can see how each month’s numbers are tweaked in the two months that follow, with dark arrows below indicating an upward revision between the third month and the first report and a lighter-colored area indicating an adjustment downward. (April 2020 was such a massive drop that including it would break the scale, so I didn’t.)
Three things stand out about the adjustments this year. First, the reports are a mess. Second, they indicate more jobs added each month than is normally the case. And, third, they are almost all eventually revised upward. The exception was March’s numbers, which started at 916,000 but were eventually estimated to be 785,000.
Consider one way in which jobs numbers are initially reported. Often, news reporters will compare the monthly numbers to expectations from economists. When a report comes in well below that consensus estimate, the jobs report is seen as having significantly underperformed. But since the numbers are often revised upward, it’s consistently been the case that the final jobs numbers are closer to estimates than first reported — or that they are even higher than estimates that they beat. Initially reported jobs numbers since February have cumulatively come in about 829,000 under consensus estimates. But after revisions, the figure is only 119,000 jobs lower.
The consistent adjustment upward has meant that the gap between initial reports and final numbers is, to date, wider than any year in the past nine. (Of course, the numbers for September and October will still see future adjustments that might push them lower.) For those months that have already been adjusted, the increase in 2021 has been 894,000 jobs, a jump of more than 18 percent. The next-largest relative increase after revisions came in 2014.
The rhetorical effects of this are probably minor. After all, it’s not as though the numbers after revision have pushed the new jobs total past the consensus expectation. In other words, it’s not the case in any month that a miss (under expectation) became a beat (over expectation) following revisions. It’s just probably frustrating to the administration that initial reports on the numbers are understating growth. Not that the administration necessarily deserves much credit for increases! It’s happy to take credit, of course, just as Trump was.
Incidentally, in a break with his usual good-faith consistency, Trump has not yet hailed Biden for adding more jobs in 10 months than he did through 31. I’m sure praise for both that and rising stock markets is coming soon.