December 5, 2020
Vaccinating against: "The labor market is strong; unemployment is way down!”
Welcome to the Weekend Reccs. Today’s world is curious and cacophonous. This newsletter delivers an eclectic sample of some things to read, watch, and ponder over your weekend. There’s a lot of economics and politics, but there is also so much more.
New here? Be sure to subscribe to never miss an issue.
The links marked with asterisks (*) are the recommended reads.
Hi friends,
This is our first week of Vaccinating Against Bad Economics. Much like a medical vaccine, I’m going to pre-expose you to some bad takes so that when you inevitably hear them in 2021 you’ll know better. Given there are four weeks to the end of the year I’ll be covering four bad takes.
As always, you can scroll past The Long Read if you’re just looking for recommended links.
The Long Read: VABE 1, “The labor market is strong; unemployment is way down!”
If we’re lucky, by the end of 2021 this may be a good take. But it is currently, and will remain for the near future, a bad take.
I wanted to start with this one because it is technically accurate but wildly misleading. Let’s take a look at the data the Bureau of Labor Statistics (BLS) released yesterday and discuss.
To start, we need to cover some basics on how we measure “unemployment.” Each month, the BLS reports an unemployment rate that is widely covered in the media. Using a large survey and some statistical magic the BLS calculates the following: the number of people in the US who do not have a job but are actively searching and divides that by all the people who either have a job or are actively searching.
If you plot the unemployment rate over the last 20 years, here’s what it looks like.
We were as low as 3.5% pre-pandemic, then we rocketed up to 14.7%, and as of yesterday are back to 6.7%. This causes some folks to think we’re almost back — we went up 11.2 percentage points and then down 8 percentage points. That’s like, three-fourths of the way back!
But the unemployment rate is a broad measure that misses several very concerning trends happening below the surface. In good times or bad it probably is not your most reliable indicator. I would be happy to devote pages to this topic, but let’s be honest, for most everyone it gets boring really fast. So I’ll stick with two critical consideration for our current situation.
Those leaving the labor force
An important part of how the unemployment rate is calculated is the “actively searching” part of the equation. Many working-age people aren’t searching for a job.
Maybe they’re blissfully retired (great!), maybe they choose to be a full-time parent (good for them if that’s what they want!), but maybe they think they won’t find a job and so searching isn’t worth the effort (really bad!). All of these groups are no longer in the labor force in the BLS’s eyes since they aren’t holding a job or actively actively searching for one. Their reason, good or bad, for not searching doesn’t matter to the BLS.
So, to recap, we’ve got basically three types of people: folks that have jobs (employed), that don’t have jobs but are searching (unemployed), and those that are not currently searching (not in the labor force).
When people leave the labor force they are therefore removed from the denominator of the unemployment rate equation given earlier. This can lead us to believe that the labor market is in better shape that it actually is, especially if a lot of folks want to work but just aren’t searching actively for work (you know if, say, you suddenly have to be a full-time teacher/caretaker/IT professional for your kid).
To drive the point home you can imagine we have three people in the labor force. Avery is unemployed and searching for a job. Brock has a job. Charles is unemployed and actively searching for a job. This population would have an unemployment rate of 66% (1 employed / 3 employed or actively seeking). Let’s say Charles drops out of the labor force (stops actively looking for work) to be a stay-at-home dad because his kids’ school is shut down. If his kids were not at home all day Charles would like to be working. But now, since he is not searching, we ignore him in the oft-quoted unemployment rate, which just dropped to 50% (1 employed / 2 employed or actively seeking) despite things not being any better in the labor market (in fact, one could argue they are worse since it will take Charles longer to restart his job search and get paired with an opening).
Let’s get back to real life and look at what has happened to the percent of adults in the labor force during the pandemic.
To avoid any folks that were, say, going to retire at 65 but were happy to retire early due to the pandemic, we’ll focus only on 25 and 54 years old. We can pretty safely assume most people in this age range that were working in January 2020 would still like to be working if things were normal, and thus it is bad for them to have left the labor force.
Ouch!
We’ve dropped from 83.1% to 80.9% and are on a downward trajectory. That’s a difference of about 2.8 million people. If these people were all in the same city it would be the third largest in the country. And those people aren’t even unemployed! Those are the people that the unemployment rate is missing, and most of them probably would take a job if they were able. And that would-be city could be much larger — we didn’t even add in the 16-24 year olds or 55+ year olds would fall into this same description.
Those who have permanently lost their jobs
A weird quirk of the pandemic has been that a lot of people were temporarily laid off while businesses closed their doors (say, due to state-level restrictions), only to be brought back in a few weeks later.
These folks technically also are unemployed (even if they don’t necessarily search for another job), since they are not working but obviously in the labor force. They are included in the numerator and denominator of the official unemployment rate.
To see how distorting the current layoff situation has been, take a look at the percentage of the unemployment rate attributable to layoffs (which BLS defines someone as laid off if they’re expected to return to their job within six months).
You should contrast the spike in 2020 to the lack of one in 2009. The unemployment rate has whipsawed around so much because at one point there were 78% of unemployed folks on layoff, all of them expecting to get called back to work within a half year. Those that ultimately did reprieve their jobs have been the reason for that big, misleading bounce back in the official unemployment rate.
I say it is misleading because when we talk about the unemployment rate we generally conjure images of people not working and without a new job lined up. For these temporary layoffs, however, that wasn’t the case. They were just having to wait (often on their own dime, so it was a good thing we had sent out stimulus checks and expanded unemployment benefits). This sort of a thing was unprecedented, thus we didn’t have a better way of talking about it.
It would have been a really, really, capital-b Bad thing if these people ultimately did not get called back into work. But that doesn’t mean that everything is fine just because most of them were called back into work.
So how can we get a better sense of the labor market’s health during the pandemic? How can we look under the hood and ignore this whipsaw from temporary layoffs?
We can look at jobs that are permanently lost. That is to say those jobs that folks have no chance of going back to. This could be everything from a business closing for good to one worker just being fired from a position. The point is not that the job will never exist again, but that the person who lost it has no chance of getting it back.
This next graph should be our primary gauge on the employment situation during COVID. Instead of looking at all unemployed (including temporary layoffs) in the numerator like the unemployment rate does, it only uses the number of permanent job losers. This allows us to ignore all those folks that were expecting to reprieve their jobs. Let’s call it the permanent losers rate.
As of this week’s BLS report, we’re looking at a permanent losers rate of 2.3%. That’s about where it was in November 2008.
But the trajectory is more concerning than the level. Since the pandemic hit, a ton of jobs have been permanently lost each month, and the pace at which they’ve been lost has rivaled the worst months of the Great Recession. We’re now starting to possibly flatten that trajectory, but we certainly haven’t seen an improvement, much less a booming recovery.
We are in the deepest depths of the real recession.
We’ve got a long way to go back.
As we continue into this year, this should be our primary gauge but not our only one. We should also check in on measures of labor force entry/exit, as mentioned earlier. We should pay attention to sector-specific unemployment rates (tourism vs construction, etc.). We should pay attention to discrepancies between demographic groups (age, gender, race, education attainment, etc.). We should care about the number of multiple job holders, and those working part-time because they can’t find full time work.
But if you’re only going to look at one graph, make it this one.
Wrapping up
The unemployment rate has declined, but a lot of those gains have been from people who were called back from temporary layoff. We expected these folks to get their jobs back.
Under the hood the number of people who are permanently laid off has grown tremendously and is not showing signs of real improvement. The labor market is not in a good place.
Finally, I would be remiss to not address the fact that there is no recovery without handling the pandemic. If people are afraid to work, shop, or go to school, we’ll not get back to pre-pandemic levels any time soon. These numbers suggest that more stimulus would be helpful for the economy, but it wouldn’t be a cure-all. We need to effectively distribute real-life vaccines, too.
Please pass this “vaccine” along. There’s no herd immunity for bad economics.
Preview for next week’s Vaccinating Against Bad Economics: “Time to raise rates!”
The Quick Links
People Watching* But it is everywhere, and on Youtube. This is one of my favorite web-projects I’ve seen in a long while.
Advent Calendars* The season of Advent is upon us. One of my favorite family traditions as a child was moving small Christmastime characters from their numbered pouches into the house on our Advent calendar. The linked article gives a fascinating history of such calendars.
Superstar Exemptions* What do you do if you’re a rising global powerhouse and your #1 source of soft power is about to be broken up due to compulsory military service? You just change the law.
Back in 2018 a similar situation raised the temperature at the Asian Games — Son Heung-min, one of the top soccer players in the world, needed the South Korean soccer team to win at least a gold medal to be exempted from service. Obviously two years of military service is not just inconvenient, it is career-ending for pop stars and pro-athletes. Luckily for Son, South Korea won.
JetBlues* Business travel could be reduced permanently as folks realize there’s no need for many in-person meetings. This has been lauded by many as a good thing for economic efficiency. But if it is, it begs the question, was it really just cultural norms that kept us bouncing around so frequently for business?
Lagniappe
This week I watched the 2012 comedy Sleepwalk with Me, starring/written by/directed by Mike Birbiglia. It was unique in that it was sort of a movie, sort of a series of sketches, sort of a stand-up routine. It was the exact type of thing you would expect from a film co-produced by Ira Glass, the host of This American Life. I enjoyed it and you should watch it if you have an hour to kill.
Graph(s) of the week
[Segal & Feldman] YLS and HLS grads dominate the law school professor market.
[The Economist] Caroline and I finished The Queen’s Gambit last week. Despite the meh ending it made me want to get back into playing chess. As it turns out, each succeeding generation gets better and better at chess.
Your friend,
Harrison