Moira Donegan writes in the Guardian today about the exit of workers from the labor force:
The fact of the matter is that when we speak of the Great Resignation, we are really referring to a great resignation of women. During the pandemic, women have exited the labor force at twice the rate that men have; their participation in the paid labor force is now the lowest it has been in more than 30 years.
This is just not true. If you follow the link, it's nothing more than an NPR host asserting it with no backup at all. Here are the real numbers:
Since the start of the pandemic, 2.31 million men have left the workforce compared to 2.38 million women. The numbers are nearly identical.
Now, there are different ways you can look at this. There are fewer women than men in the labor force, so 2.38 million women represents a larger percentage of the whole than 2.31 million men. And if you look solely at prime-age workers, about 1.4 million women have dropped out compared to 1.0 million men.
But no matter how you measure it, you'll never get women dropping out at twice the rate of men. And the absolute numbers, which are what most people probably think of when they hear about this, are very close regardless of which figures you use.
There are many ways in which women have had a harder time during the pandemic than men. The NPR segment linked above does a good job of going through them. But that hasn't resulted in a huge disparity of lost jobs between men and women. It just hasn't. Everyone needs to stop mindlessly repeating this.
POSTSCRIPT: In the realm of misleading statistics, I'd also include the factoid that women's "participation in the paid labor force is now the lowest it has been in more than 30 years." Technically this is true, but only because labor force participation for both men and women has been declining since 2000. Nearly any year represents the lowest level in the past couple of decades, and any year after a recession is automatically going to represent the lowest level going back even further. This is really a senseless statistic.
Employment level for all age groups:
Percent drop women: (1-42.4/44.8)*100 = 5.357%
Percent drop men: (1-58/60.3)*100 = 3.814%
5.357%/3.814% = 1.4
Source: Kevin's numbers above
Prime age:
Percent drop women: (1-46.045/47.598)*100 = 3.263%
Percent drop men: (1-52.700/53.825)*100 = 2.090%
3.263% / 2.090% = 1.56
Source:
https://fred.stlouisfed.org/series/LNS12000062
https://fred.stlouisfed.org/series/LNS12000061
Comparing Jan 2020 numbers with Oct 2021 numbers
Kevin is correct that the claim that twice the is an inaccurate exaggeration. The fact that the rate (percent) decline in employment for women was 1.4 or 1.56 times that of men should catch policy makers and the media's attention. The fact that Kevin doesn't address the fact that fewer women were 15 million fewer were employed prior to the pandemic is itself misleading, and is one of the reasons comparing raw numbers of jobs lost will not convey the entire picture.
Thank you, great post.
Don't forget that the % drop may have been different at different times during the pandemic.
Reporters were promised there wouldn’t be any math if they went to journalism school.
Well yes. US population growth in terms of speed has been flat since 1990. With waves up and down.
KD: “And the absolute numbers, which are what most people probably think of when they hear about this …”
Only true if most people are innumerate. ‘Rate’ properly never means raw or absolute number, rate always requires a ratio of two numbers. E.g. a rate of travel in km/hr, a fuel-consumption rate in miles/gallon.
And thanks to Joshua Curtis for doing the math for us, and drawing the correct conclusion, that women’s employment has been impacted more, not just differently, by the pandemic.
Somebody's got to look after the kids after state daycare was suspended, doncha know.
Thank you for using "factoid" correctly.
I want to give Mr. Drum some credit here. Whenever I encounter some journalist writing or talking about economics, I now hear all the BS that used to just go over my head. All the clichés and exaggerations are now obvious. So hats off to Mr. Drum for continually pointing this out. I see it now.
On the topic of "workforce participation"...
It's great that people can retire or stay home with kids if that is happening more. Why should we care about that? It represents a mature and successful economy where some people have made enough to support themselves without the daily grind.
But if that retiring person or spouse staying home with the kids was in a vital job for which there is a shortage of workers that is bad.
I suppose. There are 330 million plus people in the US and almost 8 billion in the world. While I’m sure my job / work is useful and probably even essential, there are still plenty of people who can take my place if I retire.
This shortage has some effect. My local fast food places aren’t open 24 hours anymore. I might have a harder time getting a doctor appointment. Ultimately though, I just don’t see the impact.
Forcing people to work when they don’t want to is about the worst effect of this debate.
Kevin says: "Technically this is true, but only because labor force participation for both men and women has been declining since 2000. "
Something that is "technically true" is true. This use of the phrase "Technically this is true..." to discount something true and give the impression that it is not true at all, is itself misleading. A technically true statement is true, whether Kevin likes it or not.
Dismissing the significance of a fact is not the same as showing that the fact is untrue. When I see a dismissal like that, I wonder what the speaker's motives are for discounting it. This is opinion being cloaked as analysis.
I can understand that, not being female and not having school-age kids, Kevin Drum might not appreciate the extra difficulties women have had during the pandemic, but saying that these don't matter because a graph shows that there hasn't been a huge dip compared to men, is just bigotry.
I've been noticing that Kevin Drum seems to have an interest in showing that whatever change someone is claiming is far less than claimed, that things are going on as before, that nothing much is happening. He does this across many topics. Unless you are familiar with techniques for displaying data, you may not recognize how he has produced results minimizing change and difference. While his graphs are technically true, it is also true that his preference is for debunking whatever someone else has shown in their data. That is something to be suspicious of, especially when it addresses claims that are important to subgroups in our society, in this case working women.
Not at all. 'Statistically significant' results doesn't mean that the results are significant, for example. An oldie but goodie from my first stats class waaay back in the day.