On the Gender Wage Gap
Today is “Equal Pay Day”, which is the day on which the median earnings of female workers starting from the first day of the previous year (January 1, 2013 in this case) to today equals the median earnings of male workers for only the previous year. The date is based on Census Bureau statistics which show that the median earnings of full-time female workers is 77% of the median earnings of full-time male workers. The gender pay gap is allegedly caused by sexism, but in order to make such a claim the Census Bureau’s gender wage gap must be completely misinterpreted.
In claiming that the gender pay gap is proof of sexism, feminists and other Equal Pay Day promoters assume that women with equal qualifications doing equal work as men are paid approximately 23% less in the median. The putative cause of this sexist behavior is, unsurprisingly, The Patriarchy — a “good ol’ boys club” of higher level managers (mostly male) paying and promoting younger men preferentially over female workers. The problem is that feminists fail to control for confounding factors which contribute to a worker’s pay. Such factors include work experience, education, hours worked, and job hazards — generally, an increase in these factors results in higher pay for a particular job. In order to claim that the gender wage gap is due to sexism, leftists and feminists falsely assume that a woman at the median (who by definition earns more than half of all other women and less than the other half of all other women) has equal qualifications and does equal work as a man at the median.
It is not difficult to understand how this assumption is plainly false. Indeed, the leftist/feminist scapegoat for the gender wage gap — a “good ol’ boys club” — assumes that men generally have more experience and dominate higher-level managing jobs, which unsurprisingly pay higher wages. Given the fact that women have only started matching the number of male college graduates and workers fairly recently, it is entirely unsurprising and expected that women at the median have less experience, education, etc. than men at the median and therefore deserve lower pay. There are many additional factors which result in higher pay for men — men tend to work more dangerous jobs (the fatal occupational injury rate for men was nearly ten times the rate for women in 2009), men tend to work in more difficult and higher paying occupations and specializations (e.g. STEM jobs for men vs. social work for women, or surgical specializations for men vs. pediatrician specializations for women), men tend to work in less desirable locations, men tend to take jobs which require overtime or work on weekends (so that men work more hours on average than women), etc. The evidence that these confounding factors contribute to most of the gender pay gap is so compelling that even some feminists like Hanna Rosin admit that it is a lie to say that the gender pay gap is due mostly or entirely to sexism. Unfortunately, there are very few feminists who admit that the gender pay gap is not due to sexism — most feminists use the 77% statistic and blame it all on sexism.
Re-framing the gender wage gap
Although leftists and feminists like to frame the gender wage gap as an example of enduring sexism against women getting their “fair share” of wages, it is entirely possible to completely re-frame the gap as example of women not working their “fair share”. For example, instead of saying that women are only paid 77% of what men are paid (at the median) one could just as easily remark that women are only doing 77% of what men are doing (at the median). Instead of complaining that it takes until April of the following year for women to be paid as much as men were paid in just the previous year, one could just as easily say that it takes until April of the following year for women to finally complete as much work as men did in only the previous year. Instead of asking why women are paid less one could ask why women generally don’t work as much, don’t work in as difficult jobs, choose less difficult jobs and specializations, etc. Re-framing the gender pay gap (gender work gap?) in such a way is just as valid as the leftist/feminist interpretation, but it suggests that women have it easier than men rather than harder as a result of unfair discrimination. Moreover, the corrective action suggested by such an interpretation is for women to work harder and longer, choose more difficult and dangerous occupations, etc. — rather than government action to correct alleged sexism against women.
Ultimately both interpretations are flawed because they are based on the median earnings of men vs. women, but the woman who has median earnings for women is not the same as the man who has median earnings for men — statistically the woman at the median works fewer hours, has a less difficult and dangerous job, etc. than the man at the median. In other words, both interpretations fail to control for confounding factors which contribute to an individual’s pay. Unfortunately, Census Bureau statistics do not provide a way to control for such confounding factors.
Calculating the gap with statistical controls
Some studies have attempted to calculate the gender wage gap after controlling for confounding factors — i.e. calculate what the gap would be for a man vs. a woman who are otherwise identical. One example is from the American Association of University Women (AAUW), which issued a report stating that
after we controlled for all the factors included in our analysis that we found to affect earnings, college-educated women working full time earned an unexplained 7 percent less than their male peers did one year out of college (p. 21)
To control for confounding factors this study compared
men and women with the same major, who attended the same type of institution and worked the same hours in the same job in the same economic sector (p. 20)
Since this study controlled for these confounding factors, the 7% gap it found is far more accurate than the 23% gap from the Census Bureau which does not have any statistical controls. If leftists and feminists were intellectually honest they would cite a study which used statistical controls like the AAUW one rather than the Census Bureau’s uncontrolled data, but the vast majority of leftists and feminists claim a 23% gap instead. Even so, the 7% gap unexplained by the AAUW study’s confounding factors does not necessarily imply that sexism causes the gap — there may be other confounding factors not accounted for by the AAUW study which contribute to some or all of this gap (job hazards, flexibility of work schedule, etc.).
Even if bias against women does cause a gender pay gap, such a bias may not be the conventional “good ol’ boys’ club” sexism asserted by feminists. For example, another study asked male and female science faculty to evaluate hypothetical job applications which were identical in all ways except the the applicant’s name (some had a typically male name, others female); this study found that the male applicant was generally rated as more competent and offered higher pay, but also that female faculty were biased against the female applicant just as male faculty:
female faculty participants did not rate the female student as more competent…or hireable…than did male faculty. Female faculty also did not offer more mentoring…or a higher salary…to the female student than did their male colleagues.
It makes no sense to claim that the female faculty members’ bias against the female applicant is due to sexism/misogyny, and the fact that both male and female faculty were biased against the female applicant suggests that sexism/misogyny may not be the cause of the male faculty members’ bias, either. Perhaps the female faculty members’ bias was simply due to some sort of intrasexual competition with the female applicant while the male faculty members’ bias was due to sexism, but more likely the bias for both male and female faculty was due to some common cause(s).
What that potentially common cause could be is anyone’s guess. One odd study found that women tend to be absent from work more than men (due in part to menstruation!) which would result in a true economic reason to pay a man more than a woman. Perhaps the science faculty judging the otherwise identical male/female job applicants had previous experience working with both male and female workers which gave them an expectation that the male applicant would be absent less and thus deserving of higher pay. On a related note, the hypothetical female applicant in the science faculty study would be at her child-bearing age and would have a risk of going on maternity leave (to rule out this possible explanation a similar study could be conducted in which the applicant’s age was known to the reviewer to be past the female child-bearing years). Reasonable people may reach different conclusions on what is fair here: on the one hand it isn’t fair for a woman to receive lower pay on the possibility that she will become pregnant and take maternity leave, but on the other hand it isn’t fair to the employer to pay a woman on maternity leave for no work when that employer could have instead hired a man whose work would not be disrupted for months by maternity leave. Either way, biological differences in the sexes can lead to true economic reasons for employers to pay men and women differently.
Correcting the gap
Given the variety of confounding factors which have been found to contribute to the gender wage gap, it can seem an impossible task to correct it. Unless a study can find a set of confounding factors which explain all of the gender wage gap we must consider the possibility that sexism causes women to be unfairly paid less than men. On the other hand, if we assume that the gender wage gap is caused by sexism then we risk using the force of law to unfairly pay women more than men in the pursuit of correcting for sexism that does not exist.
Fortunately, eliminating the portion of the gender wage gap caused by sexism (if any such portion exists) may be as simple as allowing the free market to do its work. Economist and Nobel Laureate Milton Friedman explains:
Friedman’s main points are:
- If men and women do not have equal productivity then the only weapon of the less productive sex in seeking a job is to offer to work for less. By mandating that they be paid the same for less work prices them out of a job in much the same way as a minimum wage.
- In the absence of an equal pay law, sexist employers bear a cost by discriminating against women who are truly equally productive since they are paying men more for no additional productivity. Moreover, competitors can hire women at a lower wage and would have an advantage over the sexist employer. Over time this would either force the sexist employer out of business or force him to hire women for the same wages as men (in order to negate his competitors’ advantage and stay in business). Either way, the market would tend to eliminate the portion of the gender wage gap caused by sexism so that any remaining gap would be due to relevant economic reasons — which may already be the case today.
The lesson is that statistics which show a gender wage gap do not necessarily indicate the presence of sexism. Such statistics at a minimum must control for confounding factors, and even then it may not be possible to account for all such factors. Consequently, it may not be possible to quantify the portion of the gender wage gap that is due to sexism (if any), in which case an equal pay law will just as likely overcorrect and discriminate against men. A better way to make sure women are not unfairly discriminated against is to promote a free market which tends to punish sexist employers.