Starting last week, fresh alarm about a pay gap between men and women started percolating, spurred by a new report from the Institute for Women's Policy Research (IWPR). The group suggests that wage disparities between U.S. women and men are more severe than we previously knew. Forget the adage that women only make about 80 cents to every dollar a man does—in actuality, it's more like 49 cents for every dollar, IWPR says.
Most media outlets have been reporting the research credulously and with little context. "Women don't earn 80% of what men earn. The true number is closer to 50%," states Felix Salmon at Axios, for instance. Fortune reports: "In 2016, women reportedly earned 80 cents to the average man's dollar. However […] a woman today earns just 49 cents compared to the average man's dollar."
But whether it's presented as a revision to previous data or some sort of new trend, this statistic is misleading. Like a lot of previous studies on the gender wage gap, IWPR's new work doesn't actually compare whether "equal work" by women and men earns equal pay. It doesn't look at earnings between men and women in the same or substantially similar jobs, take into account number of working hours, or break down wage gaps by industry or age.
In other words, it doesn't do any of the things that would be useful for measuring and interpreting the data. It doesn't tells us anything about the thing that really matters, which is (conscious and unconscious) gender bias in the workplace and the prevalence of discriminatory pay rates. Which is also to say that it doesn't tell us anything about the things people are pretending it does.
Not that you would know that from IWPR's marketing of the report, which it has titled "Still a Man's Labor Market."
"Our analysis finds that we have actually been underestimating the extent of pay inequality in the labor market," said Heidi Hartmann, IWPR's president and co-author of the report, while dismissing wage gap explanations that rely on "occupational differences or so-called 'women's choices.'"
But of course occupational differences matter when measuring pay. Few people think a surgeon should get paid the same as a grocery store cashier, or that someone who works 50 hours a week should earn the same as someone putting in 20 hours at the same job. And outside of certain fashionably progressive circles, most would say the same even when this results in lower lifetime wages for individual women or puts a dent in women's wage averages.
It's certainly worth exploring why women-heavy sectors often offer lower pay rates, why women gravitate toward lower paying occupations, and what other factors may play a role in depressing women's earnings relative to male peers. But we can't consider these things without precise and holistic data, or by looking at the workforce from only one angle.
In this case, Hartmann and co-author Stephen J. Rose looked at a longitudinal dataset from 2001-2015, measure average annual earnings across the period for people who worked any amount during any of these years, and then compared the averages for male and female workers overall as well as different subsets of men and women. Annual earnings were "defined as the personal average over multiple years and count only years with earnings."
Overall, "women workers' earnings were 49 percent—less than half—of men's earnings, a wage gap of 51 percent in 2015," the authors conclude.
It's by lumping together the full spectrum of workers (from those employed full-time, year-around for all 15 years of the study period to those who only worked part-time, or sporadically, throughout the period) that they're able to reach this dramatic conclusion on the gender wage gap. The authors admit that "the earnings gap across the most recent 15 years for those who generally work full-time, year-round in this study is similar to the more commonly used one-year numbers from the same years (23 percent)."
But when you take the data out of its ideological framing, there are some interesting findings in the new IWPR report.
In many ways, the picture has gotten progressively better for women over the past 50 years. During the first period (from 1968 through 1982), only 28 percent of women worked every year—a rate that rose to 57 percent in the latest period (from 2001 through 2015).
Average annual earnings among female workers rose whether they did or didn't take any gap years:
- For women with no work gap years, average annual wages rose 57 percent, from $24,558 in the first period to $38,649 in the latest (all dollar amounts "adjusted to 2015 using the Personal Consumption Expenditure (PCE) price deflator").
- Those who took four or more years off saw average annual earnings rise 51 percent, from $9,001 in the first period to $13,585 in the latest.
- For women overall, annual wages rose from $14,379 in the first period to $28,683 in the latest period (up 99 percent).
And "overall, women's typical hourly wages increased by 45 percent from the first to the third period (from $11.51 per hour to $16.65 per hour)."
Meanwhile, men saw hourly wages remaining "almost flat across the nearly 50 years of this study"—one of many bad indicators for male employment, even as men continue to outpace women in earnings and workforce participation.
Among men, some 77 percent worked every year in the latest study period, down 11 points from the first study period. Most of the drop here happened in the 1983-1997 period, during which the percentage of men who worked every year shrunk to 76 percent.
Average annual earnings among male workers either dropped or rose a modest amount depending on worker category:
- Men employed full-time for most of the study period made an average of $53,863 annually during the 1968-1982 iteration, a figure that rose 7 percent to $57,798 for their counterparts in 2001-2015.
- For men with four or more years off, wages went down 16 percent, from $29,422 to $24,832 per year.
- For all male workers, average annual wages fell two percent, from $51,575 to $50,442.
From where I'm looking, the picture is far from perfect for either men or women. Women were starting from a much lower place with regard to workforce numbers and wages, and remain behind men in both areas. But while women's participation and wages have been rising, men's wages have flattened or decreased over the past five decades in many cases and they're seeing shrinking workforce participation rates.
Looking at men's work is important even when just considering women's workforce gains, as it provides vital perspective on the numbers. Consistent, full-time, year-around labor participation is perhaps lower for both women and men than many realize.
The study defines strong labor force attachment as working full-time year-round for at least 12 of 15 years; moderate attachment as working some every year but fewer than 12 years of full-time year-round employment; and weak attachment as anyone who earned nothing for at least one of those 15 years. Among women, strong attachments rose from 11 to 28 percent since the start of the study period while weak attachments shrunk from 72 to 43 percent. But among men, strong attachment fell from 75 to 59 percent while weak attachment rose from 12 to 23 percent. "Women with weak attachment show the greatest increase in hours of work among women, while their male counterparts show the largest decrease," the study notes. During the latest study period, women with strong attachment worked an average of five fewer hours per week than male counterparts.
And while women have seen an increasing wage penalty for taking time off, men are seeing the same thing. For women in '68-82, a year off during the 15 year period meant about $3,000 less in average annual wages. By the next period, it was closer to a $10,000 difference and by the latest, more than $15,200. For men, a gap year in work meant earning about $8,600 less in the first period and nearly $12,400 less in the second period. For 2001-2015, it meant more than $22,300 less in average annual earnings.