I’ve been moaning for years, in public forums and on this blog, about the horrible gender imbalance in the software tribe: the women are missing. I’m depressed because, numerically, things haven’t gotten any better. But there are grounds for optimism, just maybe.

It’s the numbers, stupid · They’re horrible. If Wikipedia’s right, less than 20% of university grads entering the profession are female, and the number is falling. So, basically, any employer that can get their female headcount noticeably over 20% is apt to feel smug. Is that pathetic or what?

I don’t want to be all soulless here, but the numbers really matter. Check out The Petrie Multiplier: Man-woman dynamics aside, the experience of sexism is always worse for the under-represented gender.

I don’t think the high-school girls who are massively declining to take up CS are doing those combinatorics, but a lot of them have the (correct) intuition that any environment where there are hardly any women is going to present problems for women.

Which is to say, we’re stuck in a negative-feedback loop: A big reason women don’t go into computing because they notice that there are hardly any women in computing.

Aren’t the problems cultural? · Yes, we have those: casual sexism, too much boozing, and advantaged classes who are oblivious to their privilege. For intelligent focus on this stuff, start reading geekfeminism.org (you should anyhow, it’s good) and if you really want your mind expanded, dip into Model View Culture (warning: some of the opinions there are extreme, not that there’s anything wrong with that).

The fact that these conversations are moving into the mainstream is good; a necessary (if obviously insufficient) condition for starting to right our balance. Further: I think I see progress in beating back overt sexism and related stupidities.

But I don’t think that cleaning up our culture is going to bust us out of that negative-feedback loop any time soon. The best way to get more women to join a community is to have lots of women in the community.

But how? · Self-evidently, a profession that’s 75% male just can’t suddenly have more women, as a whole. But... maybe it can work locally. If a few of our subcultures, conferences, companies, CS departments, really make it a priority to attack their gender imbalance, some of them will succeed. And I think that for some of those, the feedback loop will go positive; women will observe that there are women there and so more of them will want to be there.

But where? · Our sub-tribes tend to cluster around software tools, whether grouped by function (programming languages, databases) or vendors. I can’t see one of the BigCos making gender balance a priority; but who knows, it might be a smart tactic for a scrappy startup. GitHub seems to be trying (and yes, let’s grant that being happy over a 25% ratio is sad, but still).

Also, a community around a programming language or database or GUI platform might have a shot. My personal favorite sub-tribes are the ones clustered around dynamically-typed programming languages: Ruby and Python and so on. And I’m delighted to see things like Record Number Of Women Give Tech Talks At PyCon 2013, and to survey the faces looking out at me from the 2014 Scottish Ruby Conference site.

It occurs to me that if some such community got that positive-feedback loop going, they’d be fishing from a deeper talent pool, bringing more brains to bear, having better events and tools and interfaces and management and, well, results. And that’d get noticed; the competition for talent and attention and love between our sub-tribes is friendly but ferocious.

Back to school · But that’s not good enough. I think that in the big picture we’re still pretty well stuck until universities start graduating better-balanced cohorts of CS majors.

I occasionally give speeches for, and consult with, my alma mater, the University of Guelph, and in particular its School of Computer Science. I was huddled with a bunch of SoCS faculty on a recent visit, kicking these ideas around, and we got ourselves a little bit excited at the possibilities.

Suppose a CS faculty somewhere decided it was going to make gender balance a key focus? It wouldn’t be easy, and it would require concerted outreach into the high schools to touch those girls who are currently getting their impression of computing from, well actually I don’t know where, but I bet those impressions don’t include very many women’s faces.

If it worked it would be a game-changer; competition for the brightest kids is pretty intense, and a school that’s fishing from both halves of the population holds a winning hand, and would find itself in the news.

And the culture too · Of course, this sort of critical-mass-building effort would be wasted if the women who do start showing up find themselves awash in “brogrammer” bullshit.

And anyhow, that culture is bullshit, and we should leave it behind just because it deserves to be left behind.

I’m a greybeard · I’ve seen decades of this gender monoculture. Call me crazy, but I still believe we can do better.


Comment feed for ongoing:Comments feed

From: Paul Hoffman (Jan 29 2014, at 08:29)

Do we know that high school girls are not taking up CS? That number can maybe be hinted at if The College Board has gender numbers for people taking the AP CS test. I don't see that number on their web site, but it might be there.


From: pjm (Jan 29 2014, at 08:47)

"Suppose a CS faculty somewhere decided it was going to make gender balance a key focus?" - I was a grad student at Tufts University almost ten years ago, and the CS department faculty at the time was 50-50 m/f. (I see that this is no longer true; of new faculty since I left, almost all are male.) The chair at the time was a woman and both chairs since then have been as well. I gather that this kind of thing is unheard of.

The grad student m/f ratio was around 2:1, I think (certainly not as high as 3:1; that, also, may have changed, but it's hard to tell) and the undergraduate ratio was dismal. That kind of grad student representation was also exceptional and, I think, a result of explicit emphasis and recruiting. There were several programs at the undergrad level which were less successful.

Unfortunately, academia itself has some significant gender-imbalance issues of its own, and the obstacles are generally additive.


From: Taymon (Jan 29 2014, at 09:04)

Harvey Mudd College seems to have made this kind of concerted effort. Media reports from last year say that their proportion of female CS majors is at 40% and growing.


From: Julia Ferraioli (Jan 29 2014, at 09:05)

Great thoughts, Tim. I can't help but recall when I heard Frances Allen speak about her experiences, and how when she started her career, there wasn't this culture of misogyny and marginalization. She went for years before encountering the behavior that has driven so many women out of technology. That gives me some amount of hope that the trend can be reversed.

Tackling the gender imbalance starts even earlier than college (though efforts at that level would certainly help), it starts in high school, middle school, and elementary school. Programs like Girls Who Code (http://www.girlswhocode.com/) are making great progress in creating curricula that appeals to everyone. One would think that programs like FIRST Robotics (http://www.usfirst.org/roboticsprograms/frc) would also help, but all too often girls are shunted into design roles.

Thanks for adding your voice to the issue, Tim. Allies and mentors mean a great deal.


From: Catheryne Nicholson (Jan 29 2014, at 09:17)

Hi Tim,

I'm a mother of 2 kids, engineer, entrepreneur, and I live and understand this problem on a daily basis. Unfortunately, I believe that by the time HS or college rolls around, it's <almost> too late. If I had to put my weight behind one critical game changer to get more girls into CS and engineering fields, it's this: getting more mothers to encourage their daughters to chose technical fields. And start this effort in elementary school. Most hackers and engineers begin as young kids playing video games, building (and destroying) mechanical devices, and running experiments on all sorts of things around the house. If you get more moms on the bandwagon, you'll get more girls at an earlier age.


From: Harold (Jan 29 2014, at 09:33)

Honest exposure of an elephant in the room, I'd say.

I wonder if this is a second-order effect of both OpenSource and Outsourcing. If I might over-generalize, it would seem women seek stability, and a holistic approach to life.

If I look back at CS and industry since the mid-90's, you've seen the divestment of IT/IS work to "low cost countries," many of whom have a more healty ratio of female-to-male CS workers. The industry itself is inherently unstable...as a woman, why would I endure personal and professional hardship for no result, being "hitting a glass ceiling" when I'm unable or unwilling to pull the 90 hour weeks in a startup.

Restated: Perhaps we're looking at this the wrong way. I ultimately see women as more pragmatic about these matters than men, and women aren't choosing CS. Numbers bear that out. What does that imply about the health/viability of our industry?


From: Rick Levine (Jan 29 2014, at 10:17)

Nicely said. The notion of "fishing from both halves of the population" is appealing; I assume some schools already doing so?

(There is another way to help this along: convince your daughters to acquire a solid background in math and applied math. They might not use it, but the confidence that comes with will help them in whatever they do.)


From: S. Rose (Jan 29 2014, at 11:26)

The fine work being done at Harvey Mudd is being driven by their president, Maria Klawe (who was formerly a dean at UBC after being the head of computer science- she's a local!). It got a fair amount of press coverage a few months ago. For example:


Some very smart and very driven people have thought about, studied, and worked on this problem for decades. A focus of efforts is the annual Grace Hopper Celebration of Women in Computer Science, held in Vancouver several years ago. It gets bigger- and more effective- every year. They are building a persistent, distributed, supportive, excellent community of women in technology. See the Anita Borg Institute web for information on that.

I believe that the tide has been turned.


From: Marcus Brito (Jan 29 2014, at 12:26)

This problem is not unique to CS and I believe its roots are even deeper than hinted by Tim and other commenters.

We live in a culture that that teaches girls from a very early age they're not supposed to go for STEM fields. Girls are supposed to play with dolls while their brothers play with Legos. It might seem silly, but I think that gender exclusion of toys and other past times have a profound and lasting effect. We need to tackle that if we really want the change to be just as profound and lasting.


From: Bill Hovingh (Jan 29 2014, at 12:46)

I was going to mention my alma mater, Harvey Mudd College, but S. Rose beat me to it.

Mudd has been pushing, in a variety of ways, to even the gender balance in all of its majors (all STEM, for what it's worth) for at least 35 years, but the presidency of Maria Klawe has brought additional energy and urgency to the effort.

So, yay! for her, yay! for Mudd, and maybe their experience could give the rest of us some useful pointers.


From: Annon (Jan 29 2014, at 14:35)

Tim - interesting take on moving the gender needle. What is your take on moving the needle on careers traditionally dominated by women, such as early childhood education, nursing and social work; to name a few.

I ask because at the universities I've attended, there were no shortage of workshops and initiatives to get women into STEM majors, but nothing to motivate men into choosing majors or careers that are traditionally monopolized by women.

Men currently hold less the %10 of all jobs in nursing. Women constitute approximately 80% of the veterinary college student population. Social workers are %82 female; %87 of primary school teachers are female.

what do you suggest?


From: John Cowan (Jan 29 2014, at 18:17)

During my time at Google, I heard a presentation from $GOOG_VIP about $MAJOR_EASTERN_GEEK_U, which said that they were able to get their female enrollment of CS majors way, way up by just changing a single factor. They stopped looking at whether entering students already knew how to program. That was it, and why not? We don't, after all, demand that prospective history majors have already written serious works of history in high school or even college before we accept them.

"Men currently hold less than 10% of all jobs in nursing. Women constitute approximately 80% of the veterinary college student population. Social workers are 82% female; 87% of primary school teachers are female."

If you want to change that, change the pay scale, and even more important, start treating these occupations as full professionals with the ability to make autonomous decisions based on their professional judgment. Who'd want to go to school for years and years, pass demanding licensure exams, and then be treated like 20-something code monkeys for the rest of their careers?


From: Joe Clark (Feb 03 2014, at 09:29)

This topic keeps coming up and I keep responding two it with a pair of questions no one bothers to answer: What is your plan to redress the underrepresentation of people with disabilities in the computer sector? *That* is a population that experiences real employment discrimination.

What if most of the women who enter the profession under these new programs are lesbians? (Or let’s try another experiment: Would you be OK with the current gender balance if significantly more gay males joined the profession?)

Women make rational choices in careers. Decades of effort to increase the proportion of females in engineering and computer science have simply plateaued. This is how many females want to join those professions. Sexist attitudes and behaviour must be curtailed or stopped outright, we agree. Yet even after that happens, the numbers won’t budge. They haven’t thus far. Again: This is how many females want to work in computing.

I would be willing to accept computer nerds’ sincerity on this topic if and when they advocate for more men in, say, kindergarten teaching and nursing.

As Susan Pinker puts it: Women decided men have a lot of the prestigious jobs and simply demand half the positions. Women, and supporters like you, never ever demand that half the roofers and miners in the United States and Canada be women.


From: len (Feb 04 2014, at 17:27)

If they have chops, the band says welcome. If they don't, we won't.

Gender balance is less important than professional competence. To claim otherwise is to devalue those who want these gigs, who study for these gigs, and who know how to work the problem instead of working the room. Caveat vendor.


From: John Cowan (Feb 05 2014, at 21:56)

Joe Clark:

"What is your plan to redress the underrepresentation of people with disabilities in the computer sector?"

This is called deflection. Someone complains about a crying injustice, you point to other crying injustices. Meanwhile, nothing is done.

"What if most of the women who enter the profession under these new programs are lesbians?"

What of it?

"Would you be OK with the current gender balance if significantly more gay males joined the profession?"

Certainly not. And just how many gay men are in the profession, anyway? I doubt if you know; I certainly don't.

"Sexist attitudes and behaviour must be curtailed or stopped outright, we agree. Yet even after that happens, the numbers won’t budge."

It would be time to say that if it *had* been stopped or curtailed. It hasn't.

"I would be willing to accept computer nerds’ sincerity on this topic if and when they advocate for more men in, say, kindergarten teaching and nursing."

Right now, men have the privilege of finding better-paying jobs: see my last comment.


"Gender balance is less important than professional competence."

It's a comparison of apples and oranges. Professional competence is about individuals getting what they deserve. Gender balance is not about a female Einstein being promoted to associate professor now and again, but about women being hired, on average, as easily as men and being promoted for work comparable to the work done by men who are similarly promoted.

And oh yes, we're up to 77 cents for every man's dollar instead of 59 cents, as when Fred Small wrote the song back in the early 1980s. It's progress, but as another folk singer says, freedom is merely privilege extended / Unless enjoyed by one and all.


From: len (Feb 06 2014, at 09:37)

You are deflecting, John. This is not about who's zooming who as a form of gender bias because both genders are equally biased in that regard. And this is where you and Tim make a serious mistake. You plead without regard to other issues. As Joe said, and I agree, there are as many women in the trade as want to be in the trade.

This without regard to either gender but both: computer science is a merciless field in the sense that the computer doesn't know you are a dog. It runs or doesn't run; compiles or doesn't compile. Logic is logic. It is the best teacher of the science and the trace and the most unforgiving judge. If someone is culturally or genetically inclined to use other means of judging results, making associations and dominating a decision making process, they are fucked when it comes to computer science by the very nature of the product.

So this advice FWIW: to succeed, work the problem not the room. And if that isn't to the liking of someone, they should choose another field. Many do.


author · Dad · software · colophon · rights
picture of the day
January 28, 2014
· The World (126 fragments)
· · Gender (10 more)
· Technology (86 more)

By .

The opinions expressed here
are my own, and no other party
necessarily agrees with them.

A full disclosure of my
professional interests is
on the author page.