post by Mike Kimel
This post looks at differences between the hours worked, by gender, and the degree to which it could explain the gender gap. To lay the groundwork, the table below shows the number of full time employed (but not self-employed) men, women, their weekly wages, and hours worked.
If this data is to be believed, on average, full time working men have a weekly pay about 21% higher than full time working women, and men work about 6.3% more hours per week. This only includes people working full time during the job, so people on leave are not included.
The difference may be greater when only managers and non-hourly professionals are included, and these tend to be higher paid positions. This is a bit data, but it is the most recent example I could find – in 1999, male managers put in 14% more hours per week than their female equivalents, and male non-manager non-hourly professionals worked almost 17% more hours than women.
Are these numbers accurate? Well, our cousins across the pond report a similar pattern:
For example, full-time men worked on average 44 hours per week whilst full-time women worked 40 hours per week.
So it seems safe to note that a) on average, men are paid more per hour than women, b) on average, men work more hours than women, and c) the word “whilst” is by no means extinct in the wild, as a specimen was observed in its natural habitat less than three years ago.
It should be noted that women also seem to take more sick leave than men, though that won’t be covered in this post.
Now, there is the possibility that men work more hours because they are paid more, and while I believe that is part of the story, I don’t think it is the whole story. And perhaps the numbers can give us more information.
Figure 2 below is similar to Figure 1, but it includes some straight line extrapolation from one week to fifty two.
Notice that the extra 6.37% hours per week extra that men, on average, work relative to women implies that in general, men work 135 extra hours per year than women. Dividing 135 extra hours by 40.8 hours a week (i.e., the amount of hours in a typical woman’s full-time week) gives us 3.3 weeks. Thus, the average woman would need to work an extra 3.3 weeks a year to spend the same time at the job as the average man. These figures, it should be noted, assume no time taken off for leave.
So how do we determine of an extra 3.3 weeks a year on a person’s career? Let us start by assuming two individuals working at the same company, with identical skills, temperaments, experience, etc. In short, assume two identical employees, except for the fact that one works an extra 6.37% hours per year. What is the effect of that extra time on one’s earnings?
If two employees are identical except for the fact that one works 6.37% more hours per year, the employee who works more hours will typically accomplish a little more than the one who works fewer hours. Come performance evaluation time, all else being equal, the employee who puts in more hours will usually get a better evaluation, leading to a slightly higher raise. How much higher? 6.37% is probably a reasonable floor, given how corporate America allocates raises among employees on the basis of a pre-determined, fixed pool of money set aside for the that purpose.
Let’s put some numbers on this. Assume two identical people with identical skills, background, education, etc., who are creatively named Employee 1 and Employee 2. If Employee 1 works 40.8 hours per week, Employee 2 works 43.4 hours per week, and Employee 1 gets a 3% annual raise (which seems to be in the ballpark for 2016 based on a quick internet search), we would expect Employee 2 to get a raise of 6.37% greater than 3%, or 3.19%. If we assume that carries on for the length of their respective careers, and both are in the workplace for from age 22 to 67, here’s how their earnings (weekly and hourly) will evolve:
Figure 3. Wage Trajectory Over Time, With Different Annual Wage Increases
But annual wage increases by themselves aren’t the entire story. Throughout a person’s career, promotions come. And we can expect that Employee 2 will get promoted more rapidly than Employee 1. For simplicity, assume Employee 1 is promoted every six years, and Employee 2 is promoted every ten years, and in the years that either employee gets promoted, she receives a 10% raise (instead of 3% or 3.19%). Here’s how that wage trajectory evolves:
Figure 4. Wage Trajectory Over Time, With Different Annual Wage Increases and Different Promotion Schedule
Interestingly enough, with these simple but (I believe) reasonable assumptions, at the end of their careers, Employee 2 is making 21.7% more a year and 14.4% more an hour than Employee 1, which happens to be almost precisely the difference we observed between Male and Female pay in 2014 according to the Department of Labor, which is where we got the difference in hours worked.
Now, the model this is not evidence that women aren’t being discriminated against in the workplace. For one thing, the assumptions in the example above may be wrong. As an example, the 3% pay raise per year may seem on target for this year, but most people in the job market today can remember periods where raises tended to be higher. Higher pay raises imply a greater percentage differential after a few years. Another reason this model doesn’t necessarily show there is no discrimination is that the pay differential noted by the Department of Labor is an average for all (sampled) people in the workplace, not just of people at the end of their career. Finally, we assumed identical people with identical skills working at identical jobs.
Still, it seems that things are changing. Speak to women in their 60s and it is clear that there were many career avenues that were completely closed off to them. Today, HR departments in companies, universities, and government put a lot of resources into fighting discrimination against women. Separate and lower career paths for women, due to the fact they are women, are not tolerated. (That isn’t to say there aren’t some troglodyte entities out there, but there are proportionately fewer over time, and they tend to run afoul of the legal system and/or the court of public opinion, not to mention driving themselves out of business by chasing off employees and customers.)
In fact, today disparate impact tests sometimes mean holding women to lower standards than men, such as when it comes to physical tests for joining and remaining in the military or various firefighting units. Additionally, there are separate leagues or tournaments for women that aren’t open to men in activities such as soccer and basketball. Interestingly, “women only” events exist in non-physical activities such as poker and chess, and affirmative recruitment projects to steer women into careers where they have been under-represented numerically, such as computer programming or mathematics are mainstream, if not ubiquitous.
All of this ensures that in America today, women have a lot more opportunity than they did in the past, and this is a good thing. But the difference in hours worked may make it hard to come up with an outcome everyone deems fair. For example, in Figure 4, while Employee 2 ends up being paid far more than Employee 1, Employee 2 actually gets paid less per hour for many years, despite accomplishing more throughout the year. (Would Employee 2 choose to remain at that particular job under those circumstances?) On the other hand, if both received the same hourly pay in year 1, weekly and annual disparities in pay over time would only increase.
Therefore, assuming all else is equal, for the wage gap to truly become zero, so will the hours gap. And based on the table below (with data from this table produced by the Department of Labor), perhaps that will happen (although the figures for the youngest demographic may point to something else).
Figure 5. Hours Worked, by Gender & Age Group
Another thing that may have to change for there to be full gender equality in wages is for the two genders to select the same types of jobs. This table, which shows women as a percentage of the total workplace plus women as a percentage of various occupations, makes it clear that women tend to be under-represented in many of the higher paying blue collar occupations, such as working on oil rigs, pipelines or electricity transmission lines. (These are also typically more dangerous jobs, explaining in part the fact that a man at work over 10 times more likely to die from a job related injury than is a woman at work.) Additionally, men and women also have very different rates of self-employment.
In any case, I’m happy to share my spreadsheet. If you want it, please email me at my first name (mike), dot, my last name (kimel) at gmail, dot, com. Because I’ve been getting a spate of requests for spreadsheets associated with posts that are more than five years old lately, and since I’ve changed my computer twice since that time, I’ll add a caveat: I reserve the right not to go digging for the spreadsheet if the request comes too long after this post is published.