In recognizing Equal Pay Day last week, the Economic Policy Institute (EPI) posted a short article titled "You can't mansplain away the gender pay gap."  The article includes an infographic that "shows some of the most common criticisms of the gender wage gap and rebuts the 'mansplainers' with data."  The EPI's main piece of data is that "in 2016, women were paid 22 percent less than men, after controlling for race and ethnicity, education, experience, and location."

That statistic is grossly misleading.

The important question about the gender wage gap is, how much of it is attributable to discrimination?  In order to accurately answer this question, you need to control for as many variables as possible so you are comparing apples to apples.  (Contrary to the EPI, this approach is not "mainsplaining" away the gap.)  For example, you want to compare men and women who work in the same industry, have the same job title, and have similar experience, to name only a few of the relevant variables.  The wage gap that remains after adjusting for as many variables as possible is a good, though not perfect, indication of discrimination.

Observe that the 22-percent wage gap statistic in the EPI article does adjust for some variables.  The problem is that it leaves important ones out.

"Even after extensive research has been done," the EPI says, "... some skeptics refuse to believe the data."  The article then includes a hyperlink to a lengthy research report (also from the EPI), titled "What is the gender pay gap and is it real? The complete guide to how women are paid less than men and why it can't be explained away."  After presenting the 22-percent statistic, the report says:

Models that control for a much larger set of variables -- such as occupation, industry, or work hours -- are sometimes used to isolate the role of discrimination in setting wages for specific jobs and workers. ... We also run a model with more of these controls, and find that the wage gap shrinks ... to 13.5 percent.

So why did the EPI, in its short article, choose to showcase the misleading 22-percent statistic rather than the more relevant 13.5-percent statistic – the statistic the EPI calculated in its own research report?  One can only speculate.  But the higher wage gap statistic will surely garner a greater emotional reaction among readers.  The EPI did not explicitly mention discrimination in its short article, but it didn't have to.  When people not familiar with the relevant statistics read that women are paid 22 percent less than men, many of those readers will naturally assume that the cause must be discrimination when, in fact, most of that gap is not caused by discrimination.  (And let's be real here.  Most people who read the EPI's short article probably did not read its long, dry research report.)

In the research report, the EPI argues that the fully adjusted 13.5-percent gender wage gap "can radically understate the effect of gender discrimination on women's earnings."  An argument can also be made, however, that the EPI's adjusted wage gap radically over-states discrimination, as it is much higher than what other researchers have found.  For example, a Glassdoor study found a 5.4-percent gap.  A 2009 study commissioned by the Labor Department found a 4.8- to 7.1-percent gap.  And even the EPI points out, "Blau and Kahn (2016) find ... a fully adjusted penalty [wage gap] of 8.4 percent."  Also, one shortcoming of the adjusted wage gap as an indicator of discrimination is that it doesn't include factors that are difficult or impossible to measure, such as the fact, which the EPI acknowledges, that men are more likely than women to negotiate a higher salary.

So why, according to the EPI, might the adjusted gender wage gap understate discrimination?  "Because, for example, by controlling for occupation, this adjusted wage gap no longer includes the discrimination that can influence a woman's occupational choice."  A significant part of the (unadjusted) wage gap is due to more women than men choosing lower-paying industries and occupations.  The report continues, "Women do indeed make choices, but those choices do not occur in a vacuum.  Our society's institutions and norms exert a powerful influence on what choices are available and what form they take."

Both men and women choose their careers based on their strengths, interests, and values, including balancing work and family life.  Arguing that this choice reflects discrimination is dubious.  Furthermore, isn't this insulting to women – to tacitly suggest they are mere puppets to society's "powerful influences"?  Women are fully capable of making sovereign, informed choices about what careers to pursue.

Ending discrimination against women is certainly a worthy goal and a policy that every employer should adopt.  But presenting misleading, cherry-picked data is not the right way to bring attention to the issue.  Moreover, the EPI's short article conveys a snide tone in which its authors think they are schooling all those ignorant "mansplainers" with some hard data.  Yet the misleading data the EPI presents can only perpetuate people's ignorance on this issue.

Michael Dahlen is the author of Ending Big Government: The Essential Case for Capitalism and Freedom (Mill City Press, 2016).  @mike_dahlen