The following post first appeared on FactCheck.org.
The Republican National Committee chairman says Hillary Clinton paid women in her Senate office less than men. But annual salary data provided by the Clinton campaign show median salaries for men and women in Clinton’s office were virtually identical.
What gives? The answer may be unsatisfying, but it boils down to methodology.
RNC chairman Reince Priebus based his claim on a report by the Washington Free Beacon of publicly available expense reports submitted biannually to the secretary of the Senate. Looking at median salaries among full-time, year-round employees, the Free Beacon concluded that women working in Clinton’s Senate office were paid 72 cents for each dollar paid to men.
Pushing back against that analysis, the Clinton campaign provided FactCheck.org a list of the names, titles and annual salaries of every full-time person employed in Clinton’s Senate office between 2002 and 2008. Those data show the median salary for men and women to be the same at $40,000. The data also show Clinton hired roughly twice as many women as men.
The Clinton list of salaries included full-time workers who may have worked only part of the year, or who took brief unpaid leaves of absence. The Free Beacon list excluded anyone who did not work for an entire fiscal year. Left off the Free Beacon list, for example, was a male assistant to the chief of staff earning a salary of $35,000, because he took a two-week unpaid leave of absence to work on a House campaign.
“There are many different ways to measure these things and you will get slightly different answers,” Eileen Patten, a research analyst at the Pew Research Center told us in a phone interview. “It’s not that either data set is flawed. They just show different things.”
American Enterprise Institute scholar Norman Ornstein, who regularly sifts through disbursement reports from the secretary of the Senate while doing research for the annual Vital Statistics on Congress report, said the data are difficult to use to track salaries because Senate staffers often toggle between Senate and campaign work. That churn was particularly true on Clinton’s staff, he said, because she was running for president in 2007 and 2008. For that reason, he believes the Clinton campaign methodology provides a more accurate measure of her record on pay equity.
We take no position on which may be the superior methodology — as Patten told us, both have benefits and tradeoffs. But we think it’s instructive to consider those benefits and tradeoffs.
Pay in Clinton’s Senate office figures to be an issue because Clinton has made pay inequality, and gender discrimination, a focus of her campaign for president.
On the day Clinton formally announced her candidacy for president, Priebus went on CBS’ “Face the Nation” and attacked Clinton on one of her signature causes — equal pay for women — claiming that she paid women in her office less than men.
“[She] can’t have it both ways,” Priebus said. “She can’t pay women less in her Senate office and claim that she is for equal pay.”
“We don’t know she did that,” host Bob Schieffer interrupted.
Said Priebus: “Well, the facts don’t bear that out, the facts show that she didn’t pay women an equal amount of money in her Senate office.”
As we said, Priebus’ claim is based on an analysis by the Washington Free Beacon, which concluded that women in Clinton’s Senate office were paid 72 cents for each dollar paid to men. Using publicly available disbursement reports, the Free Beacon based its conclusion on the median salary for men and women — regardless of position — among employees who worked full-time for an entire fiscal year from 2002 to 2008.
“Salaries of employees who were not part of Clinton’s office for a full fiscal year were not included,” the Free Beacon report states.
Using that methodology, the Free Beacon found the median annual salary for women working in Clinton’s office was $40,791, and it was $56,500 for men. The Free Beacon reporter who prepared the report, Brent Scher, declined to provide us with the raw data from his analysis to compare with the data from the Clinton campaign. But he said the Free Beacon stands by its report and its methodology, and his methodology was transparent enough to see how he arrived at his numbers.
The Clinton campaign doesn’t dispute the accuracy of the Free Beacon data, but it argues the data and methodology lead to a misleading conclusion.
“The Free Beacon based their analysis off an incomplete, and therefore inaccurate set of numbers,” said Josh Schwerin, a spokesman for the Clinton campaign. “The fact is, Hillary paid full-time men and women equally.”
Schwerin provided FactCheck.org a list of the name, gender, title and annual salary of every full-time person employed in Clinton’s Senate office between 2002 and 2008. Notably, the Clinton campaign’s figures show the annual salaries of employees regardless of how long they worked in any given year. So if a woman was hired at an annual salary of $50,000 but only worked part of the year (and therefore earned some fraction of that $50,000), the Clinton data would include that salary in the women’s salary column. The Free Beacon report would not have included that employee at all. The Clinton campaign data also include employees who may have taken a brief leave of absence (sometimes to work for Clinton’s 2008 presidential campaign). Because they did not work the entire fiscal year, they were not included in the Free Beacon report.
Taking out Hillary Clinton’s salary — we didn’t think it was fair to include her since she didn’t hire herself — the median annual salary for both men and women, regardless of how much of the year they worked, was identical: $40,000.
(We spot checked dozens of the salaries provided by the Clinton campaign against the expense reports filed with the secretary of the Senate. Direct comparisons were not possible because the Clinton salary data was based on calendar years, while the public disbursement records are based on fiscal years. The annual salary numbers also do not take into consideration any bonuses an employee might have earned. But pro-rated for the amount of the year worked by the employee, the figures we checked generally matched up.)
The 77-Cent Figure
The Free Beacon notes that its methodology more closely mirrors the methodology used by the Census Bureau to arrive at the oft-cited statistic that women earn 77 cents for every dollar earned by men in the U.S. Like the Free Beacon, the Census Bureau only considered full time, year-round employees. And so, the Free Beacon argues, Clinton leaves herself vulnerable to this kind of attack because she has, in the past, repeatedly cited that same 77-cent figure.
For example, on Clinton’s Senate Web page just before she left the Senate (accessed via the Internet Archive Wayback Machine), it stated, “More than forty years after the Equal Pay Act was signed into law by President Kennedy, women still earn only $.76 cents for every dollar men earn for doing the same work.”
More recently, Clinton tweeted this last year:
@HillaryClinton, April 8, 2014: 20 years ago, women made 72 cents on the dollar to men. Today it’s still just 77 cents. More work to do. #EqualPay #NoCeilings
We at FactCheck.org have been critical of this statistic in the past when it is portrayed as the pay disparity “for doing the same work.” That’s not what it represents.
As we noted when Obama cited the statistic in a campaign ad, the Census Bureau figure is the median (midpoint) for all women in all jobs, not for women doing “the same work” or even necessarily working the same number of hours as men. In fact, women on average work fewer hours than men and are generally under-represented in jobs that pay more. In other words, it is inaccurate to blame the entirety of that wage gap on discrimination against women doing the same jobs as men for the same number of hours. Furthermore, the raw gap for all women is not quite as large when looking at weekly earnings rather than yearly earnings.
The Pew Research Center, for example, did estimates based on hourly earnings of both full- and part-time workers and found that women earn 84 percent of what men earn. Why? According to Pew’s surveys, women were more likely to take career interruptions to care for their family, which can hurt long-term earnings. In addition, Pew noted, “women as a whole continue to work in lower-paying occupations than men do.” And last, Pew noted “some part of the pay gap may also be due to gender discrimination.” Women were nearly twice as likely as men to report that they had been discriminated against at work because of their gender.
In a recent speech at the United Nations Conference on Women on March 10, Clinton did not cite the 77-cent figure, and she noted that in addition to fighting for equal pay for equal work, closing the pay gap will require “encouraging more women to pursue [higher-paying] careers in science, technology, engineering or mathematics” (about the 11:35 mark).
But the Clinton campaign isn’t arguing that the Free Beacon report is skewed because it is not a comparison of similar-level positions. It says the data show there was no pay disparity in Clinton’s office when looking at the median salaries of men and women regardless of job title. For that reason, we would caution that neither methodology — neither the Free Beacon‘s nor the Clinton campaign’s — purports to compare the salaries of men and women who were doing the same jobs.
Using the salary data supplied by the Clinton campaign, we looked at median and average salaries for men and women in Clinton’s office year by year and found relatively minor differences. In five out of the seven years, the median salaries were slightly lower for women without Clinton’s salary included. But when all the years were combined, the median salary was $40,000 for both groups. The average salary — again, taking out Clinton’s salary — was nearly identical, $50,398 for men and $49,336 for women. And again, Clinton hired nearly twice as many women as men.
So what accounts for the difference between the two sets of findings? Is it just because one includes employees who worked only part of the year (or had a leave of absence)? The example of 2008 is instructive.
According to the 2008 salaries provided by the Clinton campaign — which, again, includes anyone who even worked part of the year – the median salary for women was $39,500, while the median for men was $43,000. That works out roughly to women making 92 cents for every dollar earned by men. (In other years, it was the opposite — but as we noted earlier, the median for all seven years combined showed median salaries to be the same.)
We then compared the annual salary data provided by the Clinton campaign with disbursement data available from the secretary of the Senate for fiscal year 2008 (Oct. 1, 2007, to Sept. 30, 2008). That doesn’t perfectly match up with the Clinton campaign’s calendar year figures, but it’s close.
Of the 44 women listed in the annual salary data provided by the campaign, 26 of them worked only a portion of the year. And 10 of 24 men worked only part of the year. That means they either started or ended their employment sometime during the fiscal year, or, as was often the case, they took unpaid leaves of absence at some point during the fiscal year. Those would be the people not included in the Free Beacon analysis. If those part-year employees are excluded, the median gap widened to $42,500 for women and $59,000 for men. That translates to women earning just 72 cents for every dollar earned by men.
In other words, the Clinton campaign has a good point: Not counting those who worked only part of the year results in a wider pay gap for women in Clinton’s office.
A comparison of both data sets shows that those who only worked part of the year represent a little over half of the men and women who worked in Clinton’s Senate office that year. Among those who only worked part of the fiscal year, and would not have counted in the Free Beacon analysis, the average and median salaries were higher for women. The median annual salary for women who worked only part of the year was $38,000, compared with $35,000 for men, our analysis of the Clinton salary database showed. The Clinton campaign argues that including those who only worked part of the year makes more sense, because it shows that women and men were offered comparable salaries.
The Clinton campaign also argues that any analysis ought to consider the salaries paid to Senate staffers who also worked for any of Clinton’s three political entities: Hill PAC, Friends of Hillary or Hillary Clinton for President. Often, employees were splitting their time between the Senate and political entities and earning significant salaries from those campaign entities, sometimes more than their work for the Senate office.
For example, Huma Abedin, Clinton’s longtime assistant/senior adviser, was making a modest salary in Clinton’s Senate office ($14,000 in 2002 to $20,000 a year in 2008), but in the latter years of that time period, she was making significantly more money working for Clinton’s political entities (Friends of Hillary, Hill PAC and then the presidential campaign beginning in 2007). Public records filed with the Federal Election Commission show in 2008 that she was paid a total of nearly $97,000 in wages from Friends of Hillary, Hill PAC and Hillary Clinton for President.
Another employee, Sarah Gegenheimer, was being paid a $20,000 salary as deputy communications director for Clinton’s Senate office in 2007, but she was also making $40,000 a year in the communications office of the Democratic Leadership Offices — Office of Senate Majority Leader and Office of the Democratic Whip, the Clinton campaign says. In addition, FEC records show she was paid another $24,000 in wages for work provided to Hillary Clinton for President and Friends of Hillary.
In other words, both of those employees would have been counted in the Free Beacon tally, and both were paid less than the median in Clinton’s Senate office, even though their combined salaries were much higher than the median.
On the other hand, Dan Schwerin, a system administrator/assistant to the chief of staff, was not counted in the Free Beacon report, Scher said, because disbursement records show he was not on the payroll from Nov. 2 to Nov. 15, 2007 — even though his salary for the first half of the fiscal year was $15,349 and $20,333 for the second. The Clinton campaign said Schwerin took a brief unpaid leave of absence to help out on a House campaign.
Ornstein said this kind of movement is typical in Senate offices, particularly if the senator is running for reelection or higher office. Some full-time employees are permanently on the payroll year to year, but others bounce back and forth. The better way to make pay comparisons, he said, would be to look at the annual salaries adjusted for the amount of the year someone worked.
“You have to try to compare apples to apples and that is difficult to do, but there is more sense in the way the Clinton people said to do this,” Ornstein said.
LegiStorm, a nonpartisan group that tracks congressional salaries, warns on its website that the disbursement figures in the reports filed with the secretary of the Senate do not represent annual salary figures. On its FAQ page, LegiStorm explains, “Because of fluctuations associated with things like holiday bonuses or leaves of absence to work on political campaigns, annual salaries must be calculated with great caution. Some staffers receive additional non-taxpayer-paid income for political work they perform in their free time.”
According to the Hatch Act, federal employees like those in Clinton’s Senate office are prohibited from engaging in partisan political activities while they are working on government time. However, as the Congressional Research Service explains, the law allows “most federal employees to engage in a wide range of voluntary, partisan political activities on their own off-duty time and away from the federal workplace.” Indeed, as the New York Times noted in 2001, “Virtually every member of Congress enlists government employees to do some campaign work.”
As the data show, heavy turnover in the office together with movement between Senate and campaign staffs can make a big difference when comparing salaries in Clinton’s Senate office.