NPR (Thinks It) Has Solved the Mystery of Declining Female Enrollment in CS

A topic of increasing angst in computer science academic circles has been declining percentages of women in undergraduate CS curricula.  Various programs have sprung up to counter the trend, such as Girls Who Code.  But although these efforts are laudable, they overlook the major cause — economics.

I share the concern about the gender lopsidedness in the profession.  This quarter I’m teaching an advanced undergraduate elective course, and have 13 women out of 59 enrolled.  If that seems passable to you (not to me), note that I had 3 women out of 44 in my course on parallel processing last year, another advanced undergraduate elective course.  Actually, I started voicing this concern to my department chair a bit before the issue became a nationwide topic around 2008.

My theory at the time was that women are more practical than men, and that the well-publicized drastic swings in the CS labor market are offputting to women more than men.  This was confirmed by a 2008 survey in the Communications of the ACM, a professional magazine of the Association for Computing Machinery, which found that in choosing to enter the IT field, women placed significantly more emphasis on job security.

Last Friday NPR ran a piece titled “When Women Stopped Coding.”  It was quite engaging, but was long on Political Correctness, blaming things ranging from boy-oriented toys to sexist institutions, and short on real evidence.  Mind you, I don’t disagree that a sexist element runs through parts of the field, but NPR’s explanations are just wrong.

For example, the piece cited a woman who, as a first-year undergrad, had been intimidated by the presence of a fellow student, Lee Van Dorn, who already had years of programming experience under his belt.  The young woman, who had planned to major in CS, bailed.  NPR’s message was that the Lee Van Dorns of the CS world drive out potential female majors in the field.

Well, it just ain’t so.  I’ve been teaching hundreds of CS students per year, every year since the early 1980s, and at least at my institution, the vast majority of male students are not like Lee Van Dorn in the slightest; some of them have had one course in high school, usually of very weak quality, and nothing else, and many other male students enter with no background at all.  Yes, we do have some Lee Van Dorns too, but they are not typical.

Instead, the reason for the decline in female CS enrollment was literally staring the NPR reporter in the face — his graph of female enrollment (apparently measuring graduations) over time.  The graph matches quite well the ups and downs, and accelerations and decelerations, of the CS job market.  Of course, given the huge PR efforts made by the industry portraying the CS field as a lucrative career choice, it’s no wonder the reporter didn’t think of ups and downs; it’s all up, right?


69 thoughts on “NPR (Thinks It) Has Solved the Mystery of Declining Female Enrollment in CS

  1. Do any of your studies provide an estimate regarding the size of the “internal brain drain” of American citizen technology professionals permanently displaced by employer avidity for “fresh young blood.?”

    Liked by 1 person

  2. Thanks for posting! I recall listening to that NPR piece and it did seem unrealistically heavy on the PC. According to NPR’s graph, medical school was once only 4% women and now it is 47%, I’m sure there have been cultural factors discouraging women from entering medicine.

    You mentioned that women put a greater emphasis on job security according to an ACM study. However, I went to the link and it is behind a pay wall. Do you have the stats?

    Do you think there’s any correlation in the level of women’s enrollment in CS classes and the level of economic demand for the knowledge provided by the CS class?. E.g., is parallel processing less beneficial for getting a job than the other advanced class you taught, given a ratio of 3/41 for parallel processing and a ratio of 13/59 for the other class.

    I think it’s a fascinating idea that economics plays a role, and I wish this idea was presented more in the media.


    • Contact me for that CACM article.

      Yes, I do think that the course I’m teaching now is perceived as of more practical value than parallel processing. It’s a misperception, but probably out there.


    • Adding to it, I have heard (but not seen the original studies themselves) that the disparity is greatest in prosperous first world (more equal!) countries like the US & Norway and lowest is economically challenged and more strongly gender-prescribing countries like China, India and Saudi Arabia.

      To me this strongly points towards culture playing less of an issue, and economics playing a more important one. Have a look at the very interesting series from norway, Brainwash: for a more balanced perspective on the issue than you tend to get in the American press today.


      • Thanks for taking the trouble to do this. As I mentioned with the other graph, it tracks the job market. But the gender gap is growing, and again, I attribute this in part to “practical” women noting the history of ups and downs in the market.


    • Yes, part of the hoopla is due to the industry PR people who want to claim a labor shortage. But given that the tech jobs (while they last) are high paying, the decline of women in the field is an issue.


    • The average pay for male nurses is higher than females, male schoolteachers get promoted more than females, don’t know about babysitters – do you have any stats?

      Men may choose to avoid the lower-paying female professions, but it’s not due to discrimination, unless you count pink-collar pay for all as discrimination against men.

      You are correct, though, that there is a much bigger call for women to join the high-tech world than for men to enter the child-care / sick-care field.

      My suspicion is that big employers who run intellectual factories where they need large quantities of detail-oriented, docile workers who will do the mind-numbing work that lots of coding jobs are, for lower wages, are probably wondering how they can get more women into the field.

      That’s my theory about why ALL “Girl” tech events are dumbed-down, focusing more on fashion and family than on platforms, algorithms and dev environments.

      I’ve been to several “women in tech” dinner – discussion panels, and none of them have real techies on the panel. About the most technically-oriented women are testers. Most are program managers, HR managers, or marketers. And I’ve yet to get an even half-way complete answer to even a general technical question like, “what are the major platforms your engineers work on?” or something very general. Deer-in-the-headlights, and then a quick change of topic to work-life balance, and a general assertion from each speaker that she has never encountered any discrimination herself.

      They are clearly not targeting the bright, ambitious path finders, but the plodding mediocrity that leads to predictable delivery dates and replaceable cogs.


      • One common reflection of your experience above is that the industry PR people who are trying to get more students into CS always say, “Computer science is more than just programming.” They mean CS grads can do things like marketing.


      • Margaret Bartley October 22, 2014 at 8:59 am: “[bosses] who run intellectual factories where they need large quantities of detail-oriented, docile workers who will do the mind-numbing work that lots of coding jobs are, for lower wages, are probably wondering how they can get more women into the field.”

        It’s their backup plan 🙂 “What will we do if we can’t get enough highly-exploitable quasi-indentured men?”

        Margaret Bartley October 22, 2014 at 8:59 am: “I’ve been to several “women in tech” dinner – discussion panels, and none of them have real techies on the panel. About the most technically-oriented women are testers. Most are program managers, HR managers, or marketers.”

        I’ve not been to those dinners 🙂 but I was quite active in a buncha user groups, which tended to have quite sparse female attendance. Dunno if women-only groups like pyladies[1] will help to overcome that. (I *am* quite sure that if a *guy* had proposed a name like “pyladies,” said gentleman would have been, umm, “dismembered” in short order 🙂

        I also notice while corpocoding that there seemed to be significantly more women in first-line software *management* than in coding, or even QA. Particularly @ IBM Software Group: OTTOMH it seemed like the female fraction of coders <= 0.2, and |female|/|testers| was about a third, but the |female|/|{team leads, project managers}| seemed closer to half. (Doc–how did we leave that out of the discussion of mind-numbing?–seemed very nearly all women.) ICBW.



    • Actually, in my country (The Netherlands) there are complaints about the lack of male school teachers. In my country, girls tend to outperform boys on school, in the last decade, and the reason seems to be that the large majority of teachers is female.
      In my country, 5% of IT workers is female. In my college class of 1984, 1 out of 40 students was female. By the way, she was the smartest of the 40, she graduated ahead of everyone else.


      • Nursing is well paid, and the last I read the average California teacher makes about $63,000 per year for working only 80 percent of the year.


      • I think a clarification might be, “Some foreigners don’t like working in a mixed male/female shop.” In Middle Eastern countries, it’s quite common for men and women to be somewhat “quarantined” from each other.


      • No, really. Most of the Indians either can’t or won’t work with American women.

        The refuse to hire American women and systematically drive them out of the company.

        This is a much worse problem if the Indian is question came here after college.
        The H1-Bs, as you might expect, are the worst.


        • My girlfriend is a DBA and complains often that in meetings where there are Indian males, she is ignored and has to pass her questions to male members in the meeting to get her questions answered.

          Liked by 1 person

          • I also have had this experience. Although I will not accuse all Indian males of this – it does happen, and I see it more where these folks are not native to the US or were not educated here.
            The sad thing is that we aren’t always sending the right people to diversity training; it needs to go both ways…


  3. The sad thing is that as most of us with 30 years of experience are forced out of the industry, even though we do not tell our kids not to pursue software or hardware engineering, they will get the message loud and clear as they watch us struggling to survive in storage sheds when we can no longer buy an interview for any job other than a convenience store.


    • Not only that, studies show that age discrimination is even worse for women than men, thus further highlighting the economics aspect of the trend described in the article.


    • Actually, Harvey Mudd College, and extremely selective engineering school, took actions a few years ago that might be viewed in the manner you describe, which I agree is the wrong way to go.


      • The Bill Gates-inspired Common Core testing software has the dumbing-down effect already built in. Watch the video at:

        The software has been designed to mimic the PTSD and epileptic seizure effects of the Japanese Pokemon sickness episode a few years ago:

        Interesting that there are two state-led Common Core testing consortia, but they BOTH are using test software written by Pearson.

        The man in charge of the Dept. of Education CC testing regime is a Mr. Shelton, who worked with Gates for 20 years.


        • I don’t have time to look at those links, unfortunately.

          I approve of the goal of Common Core to require more critical thinking, and explanation of one’s math solutions. The sad part is that, as far as I know, they are dropping (or maybe greatly de-emphasizing geometry, continuing an existing slide. To my knowledge, most high school geometry classes, at least in California, no longer emphasize proofs, a key part of mathematical thinking.

          Liked by 1 person

          • The “explanation of math solutions” involves the use of:

            1. number lines instead of memorization of times/sums tables
            2. long, convoluted “new math” methods of grammar school math solutions, to slow down the smartest kids to the level of the slowest kids
            3. new definitions for old terms (“number sentence” instead of “equation”) so that parents won’t be able to help kids with their homework & studying.

            The CC standards papers use the term “mastery”. What they really mean by that is, “Keep teaching basic math and forget about advanced algebra.” They want to keep advanced math like Algebra II and Calculus out of the schools, so that the HS grads will all be “equal.”


  4. So there’s a purported downward curve since 1984?! But the trends in the field were not established by then – the FIELD was not established by then! The real problems in the field started only a few years after H-1B began in 1991. Let me propose a much simpler explanation: we ran out of women with good math skills.

    It seems a genetic fact that more men than women have good mechanical and math skills. There are probably some minor elements in preference as well, not all women who DO have the skills will want to enter the field, but the pool is also much smaller. That rather explains the chart – we exhausted the women’s pool in 1984 while the number of men kept increasing from a much larger pool.

    The kicker is you don’t really need great mechanical or math skills to do most computer work, that’s mostly a false filter the field puts on itself. So addressing just social elements might increase the number of women in the field somewhat. But to explain the baseline, just differential pool size explains it pretty darn well.


    • At UCD, our first CS program began in 1979, in our College of Letters and Science. Then in 1981 a second one was established in the College of Engineering. Universities across the nation were doing similar things at that time.

      The first big software boom occurred around 1978.

      According to a researcher I talked to recently, the data show that the math gap between male and female students is shrinking and has almost completely disappeared.

      Liked by 1 person

      • The real question is this: Is the gap shrinking caused solely by increased opportunities for women, OR, are the brains of the men being dumbed down by computer games, videos, etc.? I suspect both factors are involved, but nobody seems to be interested in the problems caused by the second of these.


    • A female family was awarded an academic scholarship to UCD, and was going to study math. Very bright young lady. Math was a lot of work, esp. as she got to upper division, and she found herself taking english classes for fun, and getting straight As. She also commented that she didn’t find many of her fellow math students very cute or interesting. (I’m being honest here.) She soon thereafter switched to an english major with a math minor. Her sister did end up studying a hard science, and did very well. She comes off non serious and whimsical, but can ace any test run before her. Kinda scary.


  5. matloff: “Instead, the reason for the decline in female CS enrollment was literally staring the NPR reporter in the face — his graph of female enrollment (apparently measuring graduations) over time. The graph matches quite well the ups and downs, and accelerations and decelerations, of the CS job market.”

    I agree that NPR’s economic reporting is increasingly pathetic–the same line as the rest of the corporate-funded media, just fewer/different commercials. (“Planet Money” is especially offensive–straining to convince 20-somethings that neoliberalism is really hip and cool.)

    However, I’m not sure that your practicality hypothesis works in this case. Consider the alternative of the “life sciences” (e.g., biomedical research): IIUC, job prospects (not to mention compensation and work conditions) in life sciences are increasingly poor, yet women continue to dominate life-science enrollments at all levels (undergraduate through postdoctorate).

    Liked by 1 person

    • The field whose labor market I know well is CS. I don’t know the life sciences field, and could only speculate, e.g. that part of this is women who wish to go into the health care field.


  6. In my opinion a woman who measured herself against a much more accomplished fellow student would be more likely to feel inadequate and drop out of the class, whereas a guy wouldn’t take it so hard. There’s a lot of academic research on this subject. It’s called the imposter syndrome. Both men and women feel it, but women more so.


    • Yes, I definitely agree.

      Universities make a lot of mistakes, but one thing they’ve done right in recent years, in my opinion, is warn students about the Impostor Syndrome. New students are certainly told about it these days, though I’m sure that many students feel like “impostors” in spite of the warnings.


  7. Life, in the real world, has countless options, and I think I know of a few that impact these observations.

    As someone who has done business with hundreds of hi-tech companies, and attended dozens of trade shows, the full-time programmers that were female were between 0 and 15 percent (depending upon the company and / or department). However, related fields appeared to have a higher percentage women – for example, testing / QA and project management. It appeared that some women would work in programming (development), and then often proactively move into these closely related fields.

    If we take this as plausible, we would then have to ask, why? I have asked this question of female testers, QA professionals, and industry veterans. Here is our unscientific “take”.

    1. These jobs often pay as much as programming jobs, but they also involve “soft” qualitative skills, which many women enjoy. These soft skills pertain to the actual program (is it intuitive and user-friendly when I go from screen to screen); as well as having to interact with fellow developers, testers, product managers, salespeople, customers, beta testers, “usability labs”, focus groups, etc.

    2. While these jobs may be demanding, they seem to require less hours than the often brutal schedules some hi-tech companies employ, which include frequent late nights, and numerous weekends. It is a constant grind. These individuals are typically (not always) less verbal than testers.

    3. Many women find coding boring and repetitive, and they’d gladly take 10% or 20% less pay for more human interaction. (See 1 and 2.)

    Has anyone every studied this angle?

    Liked by 1 person

    • Coding is “boring and repetitive”? After doing it for several decades, I still love it, and find it a challenge, just like working a puzzle. And I can’t think of anything more boring than QA.

      I do believe it’s true that QA has a disproportionately large number of women.

      Liked by 1 person

      • Some projects are challenging and like a puzzle. But many many coding jobs are on the order of, “We’ve just changed the definition of FOB – we now are including a 5% surcharge for goods that sit on the docks for more than a week. Go through the 366 inventory and P & L reports and update all that use an FOB price,”

        As a long-time coder, believe me, this scenario is more common than the the challenging puzzle work, if you are looking at the millions of technical workers, and not just the advanced developers.


        • Yes, some coding jobs are not very exciting either, just maintenance rather than developing new code. But I do submit that the majority of software developers do more than that. As another poster here said, most people I know would strongly prefer programming to QA.


    • I got out of corporate coding and into academia a few years ago, but still know folks in big and small shops in RTP. Based on my experience and theirs (and noting that the plural of anecdote is not data)

      ML1999 October 21, 2014 at 8:46 pm: “1. [QA] jobs often pay as much as programming jobs,”

      In those shops where QA and coding are separated (which is A Really Bad Idea, but that’s another issue) of which I’m aware, I don’t know of any with parity between QA and code. Senior testers sometimes make as much as junior coders.

      ML1999 October 21, 2014 at 8:46 pm: “2. [QA jobs] seem to require less hours than the often brutal schedules some hi-tech companies employ, which include frequent late nights, and numerous weekends.”

      No, testers just peak workload at a different “part of the cycle” than coders, esp in waterfall shops (Another Really Bad Idea). All too often, coders have one deadline, QA has another, and both are unrealistic.

      ML1999 October 21, 2014 at 8:46 pm: “3. Many women find coding boring and repetitive[.]” Compated to QA? Ever done GUI acceptance testing (still done manually in most shops) ?-) Every tester I’ve known who had the opportunity to go to code *jumped*, immediately.

      Liked by 1 person

  8. There’s been a spate of articles this summer on women and minority employment in Silicon Valley.

    Age discrimination was hardly touched on, though that was often mentioned in the replies posted, along with H1-B.

    This Atlantic article is kind of an overview of the phenomenon.

    ‘A litany of tech companies followed Google’s lead, using the formula it had established. And that formula has now become the de facto way to share (and apologize for) diversity data in Silicon Valley. It goes something like this:

    1. Write a blog post about the importance of transparency, acknowledging how your company has a long way to go and outlining a few diversity-related initiatives

    2. Include a sleek graph showing how few women and minorities you employ

    3. When asked to talk about the issue, decline interview requests and redirect people back to the original blog post’

    Liked by 1 person

    • They should just point to the stats on % of women earning C.S. degrees from top institutions and, if the % of women in technical positions at their company is in roughly the same ball park, leave it at that. For all institutions I think the number is somewhere around 18%.


  9. I remember an H1-B debate maybe 15 years ago, where the head of prominent tech industry lobbying group ITAA, Harris Miller, kept repeating like a mantra how the tech industry was focused on hiring ‘women, minorities, and the handicapped.’ No one pointed out that the industry had an abominable record hiring in those categories.

    Of course since the beginning of H1-B the industry has also pushed the idea that they need to hire foreign workers for the ‘temporary’ shortage, until more Americans can be trained for the job, and they’re not ashamed to keep making that same claim almost 25 years later.

    In the face of some bad publicity we’re getting some more industry PR on hiring women and [non-Asian] minorities. ‘La plus ca change.’ At least some PR flacks are getting work, if not American technologists.

    [Incidentally Harris Miller lost the 2006 Senate Democratic primary in Virginia to Jim Webb, who painted Miller as ‘the Antichrist of outsourcing.’]


    • Actually, on the contrary, the point made at the time by Miller, and his then-ally, the Dept. of Commerce, was “If only we can get more women and minorities into IT, we’d have no labor shortage. But meanwhile we need H-1Bs.” In other words, they were using the lack of women as an excuse for having more H-1Bs. Just “temporarily,” of course, as you point out.

      And yes, “The more things change, the more they stay the same.” The industry PR came up with effective slogans early on, and those slogans work so well that there is no need to change them. 🙂


  10. I don’t see my favorite hypothesis for this phenomena. But it is closely connected with your hypothesis that there are gender differences in sensitivity to economic risk. Coursework in the fields in which women are under-represented (CS, physics, electrical engineering, etc) are quite abstract and disconnected from an obvious social application. I think gender differences in tolerance of social irrelevance are a major factor. A different way to phrase the same hypothesis is that there are gender differences in people’s tendency to get lost puzzling over technical puzzles when there is no obvious use for the solution. ( ) Of course these fields do have social relevance, but the undergraduate curriculum usually requires temporary suspension of expectation of seeing any social relevance for a few years. Take a look at the very different prevalence and expression of autism in males and females and then try to sincerely believe that social cognition does not have major gender differences.


      • Any explanation for why women opt out of C.S. can be made to explain the declining trend if one claims that the explanation started (or became more intense) around the same time the % of women started decreasing. Now, it could be totally false that prior to ~1980 women saw C.S. as having more “social application” but, starting around 1980, started to perceive it as more “abstract puzzle solving”. In fact it probably is false. But that’s one way his explanation could “work” with respect to explaining the trend.


  11. “My theory at the time was that women are more practical than men”

    Now who’s being politically correct? You could just have easily have written, “Men are more courageous in their willingness to pursue a career that they like, despite the risks.”


      • I wasn’t referring specifically to “practicality.”

        What I was referring to is the broader political correctness with respect to gender: when comparing men and women, bend over backwards to portray women as superior whenever possible.

        Here is a case where we have a fairly neutral subject: the level of risk to adopt when choosing a career. I think it’s fair to generalize (at least within the bounds of common professions, such as computer science) that people who take more/less risk with their career are making neither a better nor worse decision than people who play it comparatively safe.

        But saying that “women are more practical than men” does convey (at least to me) a judgment. After all, who wouldn’t want to be “more practical” than someone else?

        I doubt you start off your semesters by telling your students how impractical Computer Science is as a career, and discouraging them from pursuing it. I’m guessing you probably think CS and coding are rewarding in many dimensions. So it came across to me as political correctness when you chose to elevate women, even at the expense of essentially denigrating the discipline that you teach.

        So it came across to me like a case of bending over backwards to portray women in a favorable light in a situation where it was unnecessary to judge any gender as better or worse at all.


          • I will admit to being sensitive about this, since this is such a familiar pattern in the media. But I would wager a paycheck that most people would judge “more practical” as a compliment. And so might some dictionaries:

            Practical: “Level-headed, efficient, and unspeculative.” (American Heritage Dictionary and The Free Dictionary)
            Practicality: “a sensible attitude toward making decisions and plans” (MacMillan Dictionary)
            Practical: “mindful of the results, usefulness, advantages or disadvantages, etc., of action or procedure. ” (


  12. We started with this thesis, more or less: “I share the concern about the gender lopsidedness in the [CS] profession. This quarter I’m teaching an advanced undergraduate elective course, and have 13 women out of 59 enrolled. If that seems passable to you (not to me)…”

    So your elective course is roughly 22% female, and that is a cause for concern to you.

    From what I recall, for several decades there has been a huge push to get more women into STEM majors, as well as all academic areas. There has been “bring your daughter to work day”, special enrichment classes for female students, affirmative action, various Presidential commissions on the status of young girls, and more. Many colleges now enroll 60 percent or more female students, and there are still women-only colleges.

    Interestingly, there are other college majors which have very “lopsided” enrollment data in well-paid, highly respected professions, and there appears to be little fanfare about the causes.

    Nursing. A quick search gave figures of men being between 6 to 10 percent of the registered nurses in the United States, and despite this “lopsided” data, there appear to be no programs to address this imbalance. (I’m open to contrary data.) Men also appear to drop out of the profession at a higher rate than women.

    Veterinary medicine. Women today are approximately 80% of the graduate veterinary student population.

    If there are unfair barriers to entry for female students, they should be removed. But where are the movements for male nurses, veterinarians, and simply college students? The President of the United States has a commission on the status of young women (not the exact name), but the preponderance of citizens in jail are men, and men have fallen behind in college admissions. If there are cultural or gender differences in CS or STEM majors, how far do we go to achieve gender neutrality? And how far does the social engineering go? From my layman’s view, a large chunk of American CS majors appear to be Asian American. If this is true, should we address these “lopsided” enrollment figures, or simply congratulate those who work hard, make free choices, and succeed?


    • Several people have brought up with me the issue of vet schools, making the assertion that no one worries about the fact that most students in vet schools are women. Not true. My late friend who was a professor in the UCD Vet School told me that the relevant lack of male students is a very big concern there.

      Liked by 1 person

      • Would you happen to know if the vet school has done any special outreach or education to attract male students?

        Real-world experience shows me that when I have gone into young boys rooms, they typically have sports posters on their walls… while young girls have posters of horses and such.


        • Wasn’t the same true about kids’ rooms 50 years ago, when vet schools were mostly male?

          My friend indicated that the UCD Vet School was taking actions to increase the number of male students, but I don’t know the specifics.


    • This is exactly right. The double-standard is quite pronounced, even alarming.

      I believe a great deal of the problem is rooted in the country’s fascination with wealth and power, somehow equating the latter with success. In truth, a human’s value and happiness have little to do with wealth or power, beyond a certain point.


  13. I’m unconvinced that that women are more deterred by career obstacles in STEM than men fully explains the striking evolution in the sex ratio for computer science since the late 1970s, when departments were first being established, but there’s no doubt that you aren’t going to fool all women that a career won’t be problematic by talking about severe and desperate shortages.

    I met a girl still in high school who said she was doing some programming for the fun of it, who did seem at least fairly conversant with the field, presumably taught by her father who was an embedded systems, real-time programmer. But she also said she and her family had known real hardship from periods of unemployment and underemployment for her father, and appeared bright enough that I would expect her to think twice about her own career.

    It’s purely anecdotal, but I think I’ve noticed in Silicon Valley that Indian women tend to be more involved with hardware, and men with software, and there may be a jati aspect to it.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s