Derek Rose’s meticulously researched data challenges Hannah Rosin’s “The End of Men” and renders much of the demise of men in the workforce to myth.
It’s hard to imagine a more talked-about magazine article in recent years than Hannah Rosin’s 2010 essay in The Atlantic prophesying ““The End of Men.” Guys, she argued, are just not cut out for the New Economy and are being surpassed by women. The proposition has inspired a lot of debate, a forthcoming book by Rosin and even 20 pitches for sitcoms—on CBS alone! (ABC must have receivedquite a few too).
There’s just one problem. Until now, no one has bothered to look at the labor-market statistics that Rosin has used to make her case.
I did—and found many of her claims were misleading or even untrue.
Women aren’t a majority of the workforce, nor are they most of the nation’s managers; 1 in 5 men are not out of work; and women don’t dominate 13 of the 15 job categories expected to grow the most in the next decade.
These aren’t small errors—taken together they form the crux of Rosin’s argument. Hannah Rosin and The Atlantic owe American men everywhere an apology.
Ms. Rosin writes: “According to the Bureau of Labor Statistics, women now hold 51.4 percent of managerial and professional jobs—up from 26.1 percent in 1980.”
In The Atlantic‘s summary of the article—which, in fairness to Ms. Rosin, was probably not written by her—the magazine was more succinct: “Most managers are now women too.”
But as its name implies, “management, professional and related occupations” is a big catch-all mega-category. At that level of aggregation, all of America’s 139 million workers fit in one of these five mega-categories:
You can break the “managers and professionals” category down into, well, managers and professionals:
And once you’ve done so, more men are managers than women. Ladies made up just 38.2 percent of the 15 million people in management occupations in 2010—and just 25.5 percent of the 1.5 million chief executives.
Meanwhile, “professional” females outnumber dudes. The fairer sex made up 57.4 percent of the nation’s 30.8 million professionals in 2010, BLS says. That’s how you get an overall rate of women holding 51.5 of managerial and professional jobs.
Looking at the numbers from the “professional” category, there’s still quite a high degree of gender segregation. Men are not outmanned in “Computer and mathematical occupations” (4:1 male-to-female), “Architecture and engineering occupations” (around 7:1 dudes), or “Life, physical, and social science occupations” (53.5 percent men). Women outnumber men in fields like “Community and social services occupations,” “Education, training, and library occupations,” and “Healthcare practitioner and technical occupations.”
Basically, the reason women held 51.5 percent of the nation’s “managerial and professional” jobs in 2010 is that there a lot more women librarians, social workers and nurses. The Atlantic used that figure to falsely claim there are more women managers! Kind of a crazy example of how statistics can be used to deceive.
Rosin has also claimed women make up “54 percent of all American managers.” I’d love to see her source on this. If any study came out on this from a reputable source, I think it’d be front-page news everywhere—but I can’t find anything on Google that backs her up.
Also overall, according to the statistics, women slightly outnumber men in management and professional positions—but earn 73.5 percent of the salary of men in those fields. This is not a statistic that shows “the deterioration of the male condition.”
Women and the workforce
Not one to shy from grand pronouncements, Rosin also declared at an NPR debate that “in 2010, for the first time ever, women became the majority of the work force. This is kind of an amazing fact.” In “The End of Men,” she writes, “Earlier this year, for the first time in American history, the balance of the workforce tipped toward women, who now hold a majority of the nation’s jobs….”
The women-are-the-majority-of-the-workforce statistic comes from the Bureau of Labor Statistics, but a) it was only briefly true, for four winter months in both 2009 and ’10; and b) it excludes millions of people engaging in paid work in mostly male-dominated professions. As Bill Clinton might say, it depends on what the definition of “workforce” is.
First, here’s a chart that better illuminates things:
See how much the graph wiggles around from year to year? That’s because these are seasonally unadjusted numbers, and a lot of people work in seasonal jobs. For example, many teachers have summers off, while a lot of outdoors work such as construction dries up in the winter months.
But see where the data points wiggle over the 50 percent mark in February, March, November and December of 2009 and January, February, March and April of 2010? That’s why some people are claiming women have become a majority of the workforce. (The percentage peaked at 50.357 in February 2010.)
Of course, when you adjust for seasonality, this little statistical blip goes away.
Also, a lot of guys were laid off in the recession, obviously. But we’ve bounced back. Since May 2010, women have been from 49.9 percent to 48.8 percent of the “workforce.” The most-recent data, released Friday, shows women at 49.4 percent for December 2011.
Here are the raw numbers:
The same numbers, presented a little differently—so you don’t see population growth, just women’s share of the workforce:
But more importantly—this very statistic is deceptive, or at least incomplete. What journalists call the “workforce” comes from the Current Employment Statistics program, a large survey conducted monthly by the Labor Department’s Bureau of Labor Statistics. Also called the payroll survey or establishment survey, this program “surveys about 140,000 businesses and government agencies, representing approximately 440,000 individual worksites, in order to provide detailed industry data on employment, hours, and earnings of workers on nonfarm payrolls.”
When you hear on the news that first Friday of the month Labor Department statistics that the nation gained or lost X number of jobs the previous month, that’s where the data is coming from—the payroll survey. (That was also the source source on all these graphs)
But there’s an obvious limit to this data. Because it’s a survey of businesses and governments aimed at payrolls, it doesn’t include:
- Management. Yup, the “workforce” doesn’t include managers … and, as we’ve already covered, about 60 percent of the nation’s 15 million managers are men. (In the BLS’ words, the payroll survey includes “production employees in mining and logging and manufacturing, construction employees in construction, and nonsupervisory employees in the service-providing industries. These groups account for approximately four-fifths of the total employment on private nonfarm payrolls.” In other words, it excludes more than 20 percent of total nonfarm payroll employment.)
- The unincorporated self-employed. In 2009, 62.4 percent of the 9.8 million unincorporated self-employed were men, according to BLS.
- Farmers. Men made up 70 percent of the 3.3 million farm operators in 2007, according to the Department of Agriculture. Among the 754,000 hired farm laborers, 84 percent were male in 2010.
- Active-duty members of the military. There were 1.43 million servicemembers on active duty as of September 30, 2011, and 85.5 percent were male. About 1 million to 1.2 million were stationed in the U.S. as of June 30, depending on how you count.
- Domestic workers hired directly by families. (This would be the one category where women dominate). There were about 247,000 nannies employed by individual families in 2008, according to BLS, and there are obviously other occupations where workers are hired directly.
Still, the workers excluded from this statistical definition of the workforce are generally men.
Now—I know your objection here. You’re saying, “Hey, it’s a big country, Rose. No statistic is ever going to measure everyone. Don’t be a nit-picker unless you have something better to offer.”
Well… as a matter of fact… I do have something better to offer. The BLS actually conducts another survey every month—the Current Population Survey, aka the household survey. This is a monthly survey of 60,000 households that is used to derive the unemployment figures you hear on the news. As the BLS says, “the household survey has a more expansive scope than the establishment survey because it includes the self-employed, unpaid family workers, agricultural workers, and private household workers, who are excluded by the establishment survey.” (It still doesn’t include the 30 to 35 percent of Department of Defense personnel living stateside on military bases, but what can you do).
When you look at the household survey, in December 2011, 73.1 million men over the age of 20 and 63.3 million women over the age 20 said they were employed. On a (seasonally adjusted) percentage basis, 73.4 percent of men were classified as “participating in the labor force”—which is to say, they were either employed (full-time or part-time) or were actively looking for work. Women’s “participation rate” was 59.5 percent. Fully 67.5 percent of men over 20 were actually employed, compared to 54.8 percent of women.
So basically, according to BLS, in December there were 9.8 million more men working than women. Claims that women make up a majority of the workforce are misleading at best.
(This is not a typical way in which you’d look at these labor statistics. Generally you’d look at what percentage of women vs. men are working. This is a chart that illustrates, out of all the people who told the U.S. Census Bureau they were working, what percentage were women. I’ve broken down the numbers by full-time workers (35+ hours/week) and by total employment, as well as by seasonality).
One in five men are not out of work
Rosin said in the NPR debate, “Now, how are men doing these days? They’re doing very, very badly. … Right now, one in five men are out of work, which is the highest percentage that’s ever existed.”
That does sound apocalyptic, right? I half-expect hordes of unemployed men to start setting up tent cities in parks across America! (Oh… wait).
But the way she said it is misleading. What Rosin wrote in “The End of Men” is actually right: “In 1950, roughly one in 20 men of prime working age, like Henderson, was not working; today that ratio is about one in five, the highest ever recorded.”
There’s a difference, obviously, between being “out of work” and not working. My 34-year-old brother would count as one of those men not working—that’s because he’s going to law school at the University of Chicago, with a job already lined up for the summer at a prestigious San Francisco firm. But technically, right now he is one of those men of prime working age who isn’t “working.”
Let’s look at how Rosin came up with this statistic. In January 2010 the employment-to-population ratio for (civilian, noninstitutionalized) men aged 25 to 54 hit a seasonally adjusted low of 80.4 percent, according to BLS. That is indeed the lowest it’s ever been since figures were kept in 1948, when the rate was 94.1 percent.
As you can see, dudes were hammered in the recession—but are starting to bounce back a little. In November 2011, 81.6 percent of guys 25-54 were working, although that’s still a good way off the 86-88 percent levels found around 2001-05.
The percentage of women working, meanwhile, has been falling slightly for the past decade.
In 1948 just 33.7 percent of women 25-54 worked outside the home. That number rose gradually over the years, peaking at 74.2 percent in 2000. But then it started dipping—to 71.8 by ’04, actually rose during the recession (72.5 percent in ’07), but has now been falling again. In November 2011, 68.7 of women aged 25 to 54 worked outside the home—the same proportion that were back in 1988.
It’s worth noting, though, that of the 7 million men and 15.6 million women aged 25-54 and classified as “not in the labor pool,” only 1 million men and 1.5 million women say they want a job. (These were people who weren’t actively searching for one, though. There were another 4.2 million men and 3.8 million women who were classified as unemployed, meaning they had made some effort to find a job in the last four weeks).
Some of these not-working guys were, like my brother, going to school, or enjoying early retirement, or taking care of their children. (Isn’t that last what feminists wanted—more dudes to be house-husbands?).
Here’s a snapshot of what the 61.6 million men and 63.1 women aged 25-54 in the U.S. were up to in September 2011:
Here’s the same data except for the folks that were working, so you can more easily see the other numbers:
(These figures don’t quite add up—I ended up with 639,302 “extra” men and 665,413 “extra” women among the ranks of the not-employed. I think some of the retirees and disabled people are being counted twice, which makes a certain amount of sense: someone can be a disabled homemaker, for example).
As you can see, aside from the the women taking care of their family responsibilities, there’s not a huge gender gap in any of these numbers.
Here’s a graph looking at all the percentage of 25- to 54-year-olds who were either working, or say they weren’t working because they were in school or “taking care of house or family.” (That is how the BLS groups people, folks that are taking time off from the labor force to reinstall their vinyl siding get placed in the same category as full-time parents, for whatever reason). (Also, note that these numbers only go back until 1994.)
Aside from the last few years of the recession, there’s not much gender difference, is there? Basically from 1994 through 2008, about 90 percent of both men and women aged 25-54 were either working, in school or raising kids. That changed with millions of men thrown out of work with the recession, but here’s to hoping the economy rights itself in the next year or two.
Just to be clear, it’s not that sociologists, urban planners and politicians shouldn’t be concerned about the unemployed, the rise in disability cases or people being layabouts. But the implication that dudes make up more of the “idle unemployed” than women—perhaps popularized by this New York Times article—is not really supported by the evidence, or at least by this statistic.
Rosin writes that, “Of the 15 job categories projected to grow the most in the next decade in the U.S., all but two are occupied primarily by women.” She continues:
Men dominate just two of the 15 job categories projected to grow the most over the next decade: janitor and computer engineer. Women have everything else—nursing, home health assistance, child care, food preparation. Many of the new jobs, says Heather Boushey of the Center for American Progress, “replace the things that women used to do in the home for free.” None is especially high-paying. But the steady accumulation of these jobs adds up to an economy that, for the working class, has become more amenable to women than to men.
This is a statistic that has echoed around the web—googling it I got nearly 7,000 hits—but it’s not true.
There are two ways to look at job growth, of course: by percent or numerically. But in either case, the Bureau of Labor Statistics figures don’t show what Rosin says they do. The 15 professions expected to grow the most from 2008-’18 by total openings include the male-dominated fields of construction laborers, heavy truck drivers and landscaping and grounds keeping workers. Men also dominate several fields in the top 15 by percentage growth, including biomedical engineers, network systems and data communications analysts and computer software engineers.
More importantly—why look at just the top 15 professions? The BLS lists the top 30. And those top 15 professions (by number) are expected to generate less than 5 million new jobs in an labor force expected to generate 15 million new jobs by 2018. It’s as if she’s cherry-picking statistics!
Looking at industries, rather than specific occupations, construction is expected to generate the most new jobs from 2008-’18, 1.3 million—completely offsetting the expected declines in manufacturing (-1.2 million) and mining (-104,000).
It’s also worth noting that according to the BLS, “Replacement needs are projected to account for 67 percent of the approximately 50.9 million job openings between 2008 and 2018. Thus, even occupations that are projected to experience slower-than-average growth or to decline in employment still may offer many job openings.”
For example, mining is an industry in decline—yet it still offers some really phenomenal job opportunities.
Basically this is a not a statistic that should worry men or said to presage our downfall.
About those childless young urban women outearning men
There’s another somewhat misleading labor-market statistic that came out after “The End of Men” was published and has gotten a lot of attention. A 2010 study report by Slingerlands, N.Y.-based consumer-research firm Reach Advisors found that in 2008 (the height of the “mancession”) childless, single, urban women aged 22 to 30 outnumber their male counterparts by about 8 percent.
The big reason for this isn’t really that urban men are being “outcompeted” by women—it’s that married men make MUCH more than single guys. (This is actually “common knowledge among empirical labor economists”—who knew?)
Red: Male ever-married, no kids
Dark blue: Male ever-married, kids present
Purple: Female ever-married, no kids present
Orange: Male never-married, no kids
Light blue: Female, never-married, no kids
Green: Female, married, kids present)
Married, childless women also make more than single childless women, although it’s not as big a gap. Here’s another way of looking at the data … see how the pale colors start out lower than their darker counterparts, and then exceed them as you move right, along the income scale?
(This chart is a little deceptive, though, because the average age in these groups is certainly not the same. Still, illustrative I think).
Even when you control for education and age, ever-married guys make significantly more than never-married dudes. For example, nationally the average 27-year-old guy with a bachelor’s degree (only) in 2008-’10 made $45,976 if he was (or had been) married, $39,604 if he wasn’t. (Ever-married childless 27-year-old women with bachelor’s made just slightly more than their never-married female counterparts, $35,574 vs. $35,122).
This type of wage survey can’t tell, of course, whether the association between wages and marriage is a cause or an effect. Does financial success lead to dating success? Or does marriage and having a supportive spouse lead to job success? Or is a third factor the cause of both? Perhaps certain individuals possess something (drive, maturity, attractiveness, intelligence, charisma, dependability) that allows them to do well in both the employment and marriage markets. Or perhaps, as two German economists posited, married men are less satisfied with their incomes, and therefore work harder. Theories abound, but none seem to be widely accepted—in a 2002 paper for the St. Louis Fed, two researchers said the reason for the gap “might remain an enigma.”
I wanted to dig through the Reach Advisors survey for myself, but it’s not available on their website and a representative told me via email, “At the time of the study, we released our findings to a small number of media outlets but we are no longer circulating it.” Having made a media splash, it seems like the company doesn’t want its work scrutinized … but after digging through the numbers myself quite a bit, I’m quite satisfied that this is what is happening.
Here’s a table I constructed looking at the earnings of 178,750 American Community Survey respondents aged 22-40 representing 7 million people in counties from this list of where single women are said to outearn men the most. As you can see, the never-married childless women in this subset do outearn similar men well into their 30s… but that slight gap is dwarfed by the one between the never-married men and men who are or have ever been married. (Also, note now the salaries for never-married men and women and childless married women stagnate or even decline in their 30s—that’s because there’s less and less people in those subgroups).
Basically, that never-married women outearn never-married men in certain areas is somewhat interesting, but it’s hard to say whether that would still be true if all the married men had been somehow prevented from marrying—or that guys are getting “outcompeted” in the workplace.
Having examined labor-market statistics here, in part two of this post I’ll look at the education statistics used to predict that guys are being surpassed.
Originally appeared at DerekRose.com.
–Photo Seattle Municipal Archives/Flickr