THE 77 PERCENT CONFUSION: Are those figures “adjusted for relevant factors?”

THURSDAY, APRIL 17, 2014

Part 3—Or are you just happy to see us: According to one familiar statistic, American women earn, on average, only 77 percent as much as men.

This statistic may well represent a real societal problem. But no one who actually works in this field claims that this is a measure of how much women get paid, as compared to men, for doing the same work.

That said, partisans constantly make that claim. When they do, we the liberal rubes get conned, just as our conservative brethren get conned when they watch Sean on Fox.

How much less do women get paid for doing the same or equal work? It isn’t easy to answer that question.

It’s a little bit like Heraclitus’ river. You’ll never see that important question answered the same way twice!

That said, no one claims that the shortfall is as large as 23 cents on the dollar. No one claims that women get paid 77 percent as much “for the same or equal work.”

You may hear President Obama imply that women get shortchanged that much. He’s trying to fire the rubes.

You may hear Rachel Maddow imply that. She’s displaying her substantial, increasing skill as a Sean Hannity type.

Repeat: No expert claims that women only get paid 77 percent as much for doing the same work. That 77 cent statistic isn’t a measure of discrimination. It’s a measure of average income before adjustments have been made for a range of “relevant factors.”

Everyone knows this, except the people who keep getting it wrong. Often, this includes our ranking journalists, who can’t seem to stay on point.

Can we talk? If a journalist wants to discuss the wage gap, he or she should probably observe a bone simple distinction:

He should observe the distinction between (1) the earning gap which does exist between men and women, on average, and (2) the percentage of the income gap which results from discrimination or unequal treatment.

That may seem like a bone simple distinction. But our journalists have long been expert at fudging such simple points up.

Consider three recent examples. Simply put, no distinction is so simple that our scribes can’t botch it:

Katie McDonough: In this recent piece for Salon, Katie McDonough correctly wrote that “the pay gap is real...women do get paid less than men for doing the same work as their male peers.”

On average, that statement is almost certainly true. But McDonough cited large gaps in pay in various occupations—and her data almost surely don’t reflect what women get paid for the same or equal work.

Here’s the way it went down:

MCDONOUGH (4/8/14): Women, on average, earn less than their male peers. How much less depends on a number of factors.

The first rebuttal one hears when trying to discuss pay discrimination is, “But are we taking about equal work?” And the answer is yes, women do get paid less for doing the same work as their male peers. It’s why the Lilly Ledbetter Fair Pay Act exists. Ledbetter worked as an overnight supervisor at the Goodyear Tire and Rubber Co. for nearly two decades. Just as she was about to retire, she received an anonymous note alerting her to the fact that she was making $3,727 per month, while men doing the same job—the same job— were being paid between $4,286 to $5,236 per month.

Ledbetter isn’t some anomaly. She is the face of the insidious operations of pay discrimination. It’s why there is now a law named after her. (A law that people like Rick Perry do not want to enforce.)

Here are some other examples of pay inequity within a single job, according to a breakdown of median weekly salaries from the Bureau of Labor Statistics:

*A male education administrator makes, on average, $1,566 a week. His female colleague earns, on average, 67 percent of that salary.
*A male high school teacher makes, on average, $1,050 a week. His female colleague earns, on average, 93 percent of that.
*A male physician makes, on average, $2,099 a week. His female colleague earns, on average, 67 percent of that.

This trend continues across fields.

In that passage, McDonough moves directly from a discussion of unequal pay for the same work to a set of statistics which almost surely aren’t intended to measure that.

As best we can tell, those statistics from the BLS are not intended as measures of pay for equal work. In the original source materials, we find no claim that they have not been “adjusted for relevant factors.”

This is why we say that:

For her source, McDonough links to this piece for the Pew Charitable Trusts by Susan Milligan. In the text of her piece, Milligan never says that she is discussing pay for equal work.

There is one graphic whose title may give the impression that pay for equal work is being measured, although even that isn’t clear. (Milligan probably didn’t create the graphic.) But Milligan links to this BLS report, which seems to make no claim to be measuring pay for equal work.

That BLS report presents “median usual weekly earnings” within a range of occupations. We see no claim that the data have been “adjusted for relevant factors”—for such factors as hours worked, years of seniority or type of job within the given occupation.

Almost surely, that isn’t pay for equal work. But neither the BLS report nor Milligan ever make the point clear. McDonough said it was.

Rachel Maddow: McDonough’s piece appeared in Salon on April 8. That night, Rachel Maddow interviewed Professor Heidi Hartmann (again) about the gender wage gap.

The two had staged a famously incoherent discussion of the gap in April 2012. Now, they tried it again.

Maddow blustered and misstated a bit, then introduced Hartmann. Right out of the gate, this exchange occurred:

MADDOW (4/8/14): Joining us now is Heidi Hartmann. She’s president of the Institute for Women’s Policy Research and she is a professor at George Washington University. She is also at the White House today for the equal pay event. Professor Hartmann, thank you very much for being with us.

HARTMANN: Pleasure to be with you again, Rachel.

MADDOW: I just want to ask you a very basic but specific question. Do American women make less money than men when they are doing the same work?

HARTMANN: Well, it’s that “doing the same work” that is tricky. There are many companies in which a man and a woman in the same job would get equal pay. But it turns out there are many companies where they don’t. And you know, we took a look just recently at the twenty largest occupations for women, these are occupations that sound the same like let’s say customer service representative, retail worker. And the wage gaps ranged, the wage ratios ranged from like 68 percent, women making only 68 percent of what men make in retail, to 94 percent, women making 94 percent of what men make in customer service work.

Maddow asked how much less women get paid for doing the same work. After saying the “same work” part of the question was tricky, Hartmann cited several statistics which don’t seem designed to answer that question.

(To see the IWPR report in question, click here. We see no claim that the earning data have been “adjusted for relevant factors”—that the statistics represent pay “for the same work.”)

Did Professor Hartmann’s statistics represent pay “for the same work?” Viewers probably got that impression, but Maddow never clarified the point.

Maddow and Hartmann are getting to be a bit like a vaudeville team. Tomorrow, we’ll offer more detail from their two performances.

The New York Times editorial board: Two days later, the New York Times editorial board cleared its throat, then discussed the gender wage gap.

Midway through their piece, the editors acknowledged the fact that the famous 77 cent statistic is not a measure of pay for equal work. (“It is not a comparison of what men and women are paid for performing the same or comparable jobs.”)

We’ll bite! How much do men and women get paid “for performing the same or comparable jobs?” In this fairly lengthy chunk, note the way the editors fail to answer that question.

The editors supply a trio of statistics which haven’t been “adjusted for relevant factors.” But they omit the AAUW statistic which has been so adjusted:

NEW YORK TIMES (4/10/14): Threaded through the political fight over pay fairness is a continuing debate about the size of the pay gap. Mr. Obama and others often cite 77 cents as what women make on average for every $1 earned by men—a figure that critics say is an exaggeration.

In fact, it is a rough, but important, measure of overall workplace inequality. It is not a comparison of what men and women are paid for performing the same or comparable jobs. But, in representing the full-time wages of a working woman against that of a full-time working man, it reflects overt discrimination as well as more nuanced gender-based factors, like the fact that women are disproportionately concentrated in the lowest-paying fields and not well-represented in higher-paying fields. Of course, 77 cents is not the only measure. But there is no doubt that the pay gap is real.

The Pew Research Center last year found that women earned 84 percent of what men earned in its study of the hourly wages of all workers, including those who work part time. Similarly, a 2013 review by the Economic Policy Institute of annual hourly wages for men and women with college degrees, including salaried and hourly workers, found that the men earned on average $33.71 per hour and the women just $25.35 an hour.

Even controlling for hours, occupations, marital status, and other relevant factors, college-educated women earn less than their male counterparts, according to a recent study by the American Association of University Women. And a study issued this month by the Institute for Women’s Policy Research reported that women’s median earnings are lower than men’s in nearly every occupation, including the most common occupations for full-time working women, like elementary- and middle-school teaching and nursing.

The editors cite three different statistics which haven’t been “adjusted for relevant factors.” These statistics show significant gaps in average incomes between women and men.

But how odd! The editors report that another study—a study by the AAUW—shows that “college-educated women earn less than their male counterparts” even after adjusting for relevant factors. But they fail to report how large that adjusted wage gap is!

This seems to be the report in question. It says the gap is 7 cents after controlling for relevant factors.

If that gap results from discrimination, it should of course be addressed, where possible. But was that measure of pay for equal work too small to suit the editors?

They specifically cited three larger pay gaps, using statistics which haven’t been adjusted for relevant factors. But they didn’t include the smaller pay gap, the statistic which has been designed to approximate pay for equal work.

So it endlessly goes.

How much less do women get paid for doing the same or equal work? It’s an important question.

We’d like to see the news division at the Times report on this important topic. We’d like to see them clarify the various statistics which float around when this topic gets discussed.

The Times won’t likely do any such thing concerning this fraught topic. As we’ve often told you, elementary facts play almost no role in our clownish discourse.

Instead, the proselytizers often take over when this topic gets discussed. As they teach you what to think about the wage gap, they may routinely fail to observe a basic, bone-simple distinction:

They may report the gap in earnings without “adjusting for relevant factors.” As they do, they may pretend or suggest that they’re reporting the difference in pay “for the same or equal work.”

In our view, Rachel Maddow’s work on this topic is becoming a modern classic. Tomorrow, we’ll show why we increasingly think of Sean when we see Rachel at work.

Tomorrow: Maddow and Costello

Just to be clear about the lack of clarity: Look at Milligan’s report. Look at the BLS report from which she worked.

Look at the IWPR report, the one which Hartmann cited.

Do those reports present statistics concerning pay for equal work? Do those statistics represent pay for women after “adjusting for relevant factors?”

We aren't sure—but go ahead, waste your day! Just try to puzzle that out!