How We Lost the Women in Computing (2018) | Hacker News Reader

How We Lost the Women in Computing (2018)

cacm.acm.org

58 pointstacon59 comments

> "the industry has self-selected for men: first, antisocial nerds, then, decades later, self-confident and risk-taking bros"

I'm sorry, this is a silly broadside at the nature of male software engineers. In my experience, software engineers tend to be smart, often but not always introverted, and in Silicon Valley, very ambitious. I think the quote captures an inherent contradiction in its own thesis: Are software engineers typically reclusive and antisocial, or outgoing and confident? The answer is... it depends, you'll find plenty of examples of both. We shouldn't blame the software engineers for _both_ their introversion and extroversion here, at the very least. :-)

Another partial explanation I've heard (and that rings somewhat true starting in the early 1980s): early personal computers were very attractive to people on the Autism spectrum, and those who spent their time programming them and hanging out were known as unpopular computer needs/geeks in school.

The social association between computer programming and social ineptitude meant that a lot of folks (men and women) who would have been great programmers didn't want to get involved.

But since Autism disproportionately affects men, the nerd groups were predominately male, and thus the new crop of programmer applicants from the 80s onwards was disproportionately male.

Newer studies consider autism's sex ratio to be 3:1 m:f, which still doesn't explain why even a field that is specifically appealing to autistic people has such low numbers of women.

Sure it does! 3:1 is enormous in terms of creating a social pressure of adhering to gender norms which would likely exaggerate the existing bias.

Found the second-order-effects thinker.

Yeah this is totally at play both in entering STEM as well as women getting autism diagnoses. Regardless of which one we are talking about- if it is more of a "male thing"- women may "conform" to roles as nonprogrammers and nonautistic.

Increasing numbers of women have been getting diagnosed.

Autism may occur in even distributions but doctors have been biased for a long time to view autism as "hyper-male-ness" while girls regardless of functioning are expected to be highly sociable and emotionally aware (traditionally thought of as signs one is not autistic)

Breaking the barrier to STEM is a bit harder than changing preconceptions as you have an industry staffed with "hyper-male" people who have never been pushed to be more sociable or empathetic, and making women who have been taught a lifetime of sociability and empathy compete with them on performance metrics that are largely antisocial.

Anecdotally, the ratio of STEM-interested high school students (even not narrowing to the socially inept) was far tipped towards guys among my peers in high school.

I was in a TIE program (tech, innovation, engineering) with about 15 other students in high school. We had 1 girl in the class... she was a great addition to the class (not at all socially awkward) and we would have liked to have more, but the interest among girls was low. The only requirements to join are pictures of previous projects, teacher recommendations, and answering essay questions to gauge interest.

All the socially awkward girls I knew of had no interest in computers... one of them went into the army, others got liberal arts degrees in college.

It's hard to draw broad themes from one's own personal experience, but I never was aware of forces aligned to tip the scales in any direction (perhaps because of my own lack of social awareness, haha).

I'm not quite sure autism plays out the same way in men and women...

I do feel like a lot of these things are naturally self-perpetuating. To give an artificial example, if you're drawn to green, and I'm drawn to purple, and you outnumber me 3:1 in some domain, that domain will be overwhelmingly green, and I'll go elsewhere. That sort of thing translates into all sorts of choices we've made in many domains.

As a young male gamer I consciously avoided majoring in computer science or applying to software jobs because of this stigma. It took a few years before I became a software engineer anyway and admitted that it is a career I enjoy. Social pressure sucks and must be that much worse for women.

I was heavily involved with programming in the 70's on many of the large systems of the time (including analog computing). If there were women programmers, I never met one. I worked in the aircraft manufacturing industry in engineering. I suspect that women could have been programming in other industries, but I never really made note of it at the time.

Around that time,only on the other side of the world, it was mostly women who operated the first computers.

I tell this to my wife and she finds it hard to believe.

This might help: https://en.m.wikipedia.org/wiki/Women_in_computing#:~:text=T....

Frankly, people who succumb to social pressure are weak. Unless there is an acute barrier to something, if you want to do it, you just should. I relish any opportunity to do things I like that others dislike- making this comment for example.

... unless social acceptance is their goal?

As social creatures, humans derive significant life satisfaction from their place in their social environment. In fact, I'd argue that having good friends has a much greater positive effect on life satisfaction than where you live.

To your point, teenagers are often not wise enough to realize pursuing their career and life goals should take some priority over avoiding social stigma at that stage in life -- since most of their peers will be out of their lives in 3-5 years anyway.

But antisocial behavior for its own sake never yields lasting rewards.

Sounds like a lonely existence.

The author gives the impression that women were well represented in the beginning, but were pushed out. We can look at photos of programming teams in industry, going back as far as you like, even to the 50s, to demonstrate that that's a fantasy. You would typically see one or two women and fifteen men.

There are graphs of "women in computing" over the years. It did indeed peak around 1982 and begin a long decline, but it was never as high as 50%. [1] shows it peaking around 35% and this is college majors not women actually working in computing.

[1] https://i.pinimg.com/originals/3f/52/65/3f5265aeeb87fd2f9675...

Interestingly my country (France) shows the same trend, but with a peak almost reaching 50%. It also shows that after that peak, the number of women in IT kept increasing then stagnated, while men exploded: https://imgur.com/a/gNHHIxm (rehosted, original website is dead)

This might be anecdotal, but I did an engineering school, CS/IT section. It was one of the very first school of my country to have a dedicated CS/IT program and diploma (the section opened over 50 years ago, 40 at the time I was there). A few hundred students each year. One of my teachers was one of the students dating from a few years after the section opened. He told us a few times that when he was studying, each promotion of the section was usually 50% men 50% women, sometimes more women even. When I graduated, it was close to 80% men. The nearby CS focused private school, where one of my friends was going, had 2 women for hundreds of men.

> The author gives the impression that women were well represented in the beginning, but were pushed out. We can look at photos of programming teams in industry, going back as far as you like, even to the 50s, to demonstrate that that's a fantasy. You would typically see one or two women and fifteen men.

It also gives the impression that "computer programmer" was a good job.

It wasn't, by and large.

There were a bunch of "keypunch operators" that were mostly women because of typewriter training. They were directed by a much smaller number of men who were the "programmers".

There really wasn't any advancement for the women--it was just a slightly better secretarial job. As for the men, it required a 4 year degree (and associated monetary outlay for college) back when working in the steel mill or the automobile line paid really well.

Men only flocked into CS/computer programming once the mill jobs got destroyed.

If you're curious about what evidence people have contrary to your assertion, you could search the web. If yuo did you might find links like these:

https://text.npr.org/357629765

https://mitpress.mit.edu/books/recoding-gender

https://flowingdata.com/2014/10/30/decline-of-women-in-compu...

Funny that one of your sources (the NPR one) is exactly the same as my graph. Other than that, it's full of qualitative statements that give an impression the data doesn't support ("For decades, the number of women in computer science grew faster than the number of men, but then we get to 1984.").

Yes, it was growing faster, but it never got close to parity.

In fact, the graph in your flowingdata.com reference IS my graph.

> So how did we lose the women in computing? They did not just leave; they were pushed out.

The problem is how we define "computing" in the first place.

The movie "hidden figures" [0] comes to mind. The three main characters are all "women in computing" and work as "computers"... meaning they perform computations manually [1] alongside other female computers (the term used at the time).

Computers before the 80's required a lot of support staff to operate. Programmers and Engineers didn't directly interact with the machine; typically programs were written on paper and then transcribed into punched cards by secretaries (who at some places would be referred to as "programmers" or "computer operators") and then organized to be batch processed. The results would then be retrieved once the computation actually happened as they tried maintaining the machines close to 100% usage.

That's where most of the confusion stems from. These were all accounted as "computer related occupations" back in the 50's through 80's. The personal computers and minicomputers made most of these positions obsolete, since the engineers could simply use the machine interactively.

But years after years people find this "collapse" of "women in computing" and interpret it as woman getting pushed out of engineering or proper programming positions.

[0] https://en.wikipedia.org/wiki/Hidden_Figures

[1] With the advent of the first numerical computers, they eventually learn programming, but it is implied that the human computers won't be needed once the machine can be directly programmed by them or the engineers.

That's all true but you still see a peak in women enrolling in computer science around 1980 or so. Those weren't women as "computers."

I'm not certain how they aggregate these numbers.

I assume that includes a lot of technical schools that might have had “computer operators” and “programming” shorter degrees. But these would not match to a modern CS education at a serious school. Back then a lot of corporate divisions had “analysts” in charge of writing the program and “programmers” in charge of typing and running the program. You could still see remnants of that around the 90’s, especially with offshoring (keep the analysts in-house, offshore the “programming” abroad).

Knuth was considered an oddball in the 60’s and 70’s for instance because he would keypunch his programs himself (and was apparently quite fast!).

Most professional devs/software engineers who graduated around that time (early 80’s) did so from EE programs. These were not balanced in terms of gender. It’s also around that time that “modern” CS programs started to emerge (either from math departments or EE).

> In July 2017, Google engineer James Damore distributed a memorandum [...] > In essence, argued Damore, women are less interested in computing then men.

It still baffles me how that was even controversial. It's not rocket science that men tend be more interested in computers than women (anyone not lying to themselves will see that in high school or campuses).

I don't know and shouldn't attempt to guess your gender, but were you provided with toys and games that were typically considered non-typical for you as a child?

I've still got pretty mad respect for the Care Bears having watched a number of their adventures repeatedly.

Evidently Professor Coldheart has a Hacker News account.

(and thank you, to a kind restorative stranger)

I think it’s pretty clear that woman are interested in using computers for social purposes. But perhaps not so focused on the computer itself. It’s a tool for a purpose rather than a hobby.

I don't work in tech, but I'd like to mention something as somebody who fits that slice of gal + highly interested in tech as a hobby(1):

Personally have had a lifelong passion for computers since the Win95 days; I don't code per se, but I've built personal sites (I find a lot of my friends who are also women get into coding via the HTML/CSS side and making personal sites a la neocities), and am the general family IT support.

That being said even despite that passion, I'd never work in "the tech industry" as a dayjob for a few reasons. Some I'm not going to mention here, but others are like ... I find there to be a lot more abusive companies on average just in terms of work/life balance and people treating others with no dignity ... Too many companies are just plain unethical (The one time I considered joining an adjacent tech company to my current industry, a week later there was a scandal of data harvesting), and I'm not fond of the constant "must know the hot trends" treadmill. Design's got enough of that already.

Just my two cents.

(1) my main computer's a linux machine from S76, if that's enough for surface level creds.

"coding via the HTML/CSS side", sorry but HTML or CSS are not coding, they are description languages. a male software engineer here!

Oh please. Enough with this tiresome pedantry.

Troll

This article is bad directed. If there were about 40% of women doing IT-related studies but only 10% working in IT every year it would be clearly explained by the bad culture exposed that some of those IT companies have, but that's not the reality.

I don't think it matters a good or a bad working culture for the amount of people studying something, and if this were true, there would be a lot less persons studying bachelors in law or English, for example.

On the other hand let's look other places, like the North of Europe, where university education is free, there are a high level of work culture quality, there are pretty progressive about men and women having same kind of jobs, and they lack those nerd and bro cultures the article talk: Why don't they have a 50/50 (or a 60/40) rate of men/women?

With that I don't want to belittle the importance of those questions, and how bad are for the women, for the industry, and for the society in general, but also think this isn't the only thing that avoid women entering in that world.

I'm not a woman and there are for sure details I don't know about all of this, and asking those who don't picked an IT career and are interested in tech is surely will give us more answers.

Women were never in computing in the sense imagined by the author.

The manual "computers" and early patch-cord "programmers" he mentions carried out instructions by wrote. Instructions created by men. Women left the field when these jobs became obsolete.

Hopper is a notable exception. Lovelace, not so much.[1]

Women did a lot of QA through the aughties. Again, this work was automated.

[1] https://www.youtube.com/watch?v=7K5p_tBcrd0&t=36m29s

In the 80's and 90's I worked for/on companies/teams that had female devs. >4 times. Computing (at least the parts I observed) didn't seem especially hostile to women. I distinctly recall the sense that women just made the team nicer, more complete, and we'd better not screw that up.

Having lived through it, it's clear to me that something changed for the worse between then and now. Tech bros, gamer aggression, greed, competition, eternal September, I don't know. But it changed.

My wife and I have > 40 years of industry experience between us in Silicon Valley, including two FAANGs, five unicorns, and several less exciting companies. Anecdotally, we've observed an increase in the percentage of female engineers here since the late '90s. And I'd be surprised if it was less in the '90s than in prior decades. The culture certainly has changed, but I think that's more a consequence of the insane growth in the field than anything to do with gender.

This post doesn’t explain how we lost women in computing. And I’m not sure how this describes evidence of “pushing out.”

I think programming is awesome and that any discrimination limits the value that comes out of programming. It’s bad and should be designed out.

I’ve only been doing this for 30 years or so and there have never been many women. That’s not to say that I didn’t work with some smart women, but there weren’t that many. Maybe 1-10%.

Maybe there was once a time when there were higher ratios, but I’m not aware of them. And I missed it in this article. There were (and are) great women in technology. But just because there’s not an Ada Lovelace today doesn’t mean she was pushed out. Just that she was a once in a century genius that hasn’t repeated in any gender.

Obviously it would be great if we had more Lovelaces and Hoppers, but to expect people of this caliber to reliably repeat or think that’s some signal of bias requires some more thought than this article presents.

> Maybe there was once a time when there were higher ratios, but I’m not aware of them.

https://www.npr.org/sections/money/2014/10/21/357629765/when...

Yeah. I agree with the article there almost certainly isn't a singular reason. But the timeline correlation to personal computers (and associated gaming) is hard to shrug off. And it makes sense. Another aspect of the same correlation is that, at the university level, computers increasingly became something you were sort of expected to have some familiarity with if you were going to major in computer science.

Today, look at the intro computer science classes at most of the better schools and imagine take them never having really done any programming or serious computer use before.

(And of course gender imbalance can beget gender imbalance for all sorts of reasons.)

This article needs more work. Comparing Computer Science degrees to progressional programs like medical school or law school shouldn’t be done because they are very different.

It seems like labor stats should be used but I can’t find BLS stats breaking out gender.

These degree stats could be explained by other science degrees becoming less sexist and allowing more women (ie, computer science goes down because other science goes up). Lots of potential reasons.

1) Once mainframe electronic computers started to be routinely manufactured computing jobs became engineering jobs which were primarily male.

2) When I was at Stanford and MIT in the 1970s, there was a debate whether Computer Science was a legitimate STEM topic and deserved a separate department (took MIT until 2018 to recognize such). Part of this was programming was considered a trade-school subject, feminine and secretarial, not unlike typing.

3) My first industrial science job I was assigned a (female) clerk who would keypunch my coding sheets, submit job decks and retrieve lineprinter output. The senior scientists considered it beneath their dignity to sully their hands with mundane programming tasks. I was considered an upstart for doing my programming personally and code rapidly outcode the senior staff.

re: #2.

I'm probably missing some nuance to your statement, but in the late 70s CS was part of CS/EE at MIT (within the school of engineering) with course 6.3 corresponding most closely to was people would think of as a CS degree. What was certainly true is that CS was a bit of an odd duck generally with it being at offspring of the Math department at some schools (e.g. Dartmouth) and Engineering at others (e.g. MIT). But it was always under the STEM umbrealla at those schools.

Every team I have been a part of in the last 10 years had been at least 6/4 split male/female. What is it with press and academics summoning problems out of thin air? Also Damore story is more about corporate abuse of people in the spectrum as opposed to some corporate breakthrough in diversity

Counterpoint: over the 18 years that I've been in the field, we've always had very male-heavy teams. Most teams have been all men. I can only think of one team that had more than one woman on it.

I'm US east-coast, suburban, and definitely not inside the SV culture bubble.

I am a contractor/consultant. I have roughly 3 contracts per year. The only all men team I have seen in the last 5 years was ironically lead and hired by a female manager. Now that I think about it is possible that industries also matter. The team I mentioned worked at heavy machinery manufacturer, teams in health-it, financial, e-commerce, and startups are all super diverse. In health m/f ratio is probably around 3/7. I have no experience with academia

Is that really going to solve any problem?

70% of college graduates are now women, and there are still complains of income disparity etc.. Now we even discover there are not enough "eligible bachelor" to marry..

The article seems to be a survey of some recent pieces on this topic. The titular question, however, cannot be answered "how?" without answering "when?". I, too, heard it is the early personal computers. I suppose that it did not reduce the interest of girls in computing, but rather overinflated the interest of boys.

It’s interesting that in Scandinavia (Denmark, Norway, Sweden, Finland) where everything possible has been done by the government to remove barriers of entry for woman to traditional male jobs, the ratios are still 80% male engineers and 80% female nurses. The only thing that seems to explain that is differences in interests (things vs. people) or social pressure. But the social pressure explanation doesn’t seem to add up given the massive effort by the governments and media to make traditional male jobs interesting to woman. But interestingly enough the majority of students studying chemistry at my University were females. I have no idea why Chemistry was popular among females and not (say) programming.

I wonder how the wide-scale move to work-from-home will change this. Having more flexible work environments while also making caregiving more possible (because, let's be honest, women still do the vast majority of caregiving, even if they work full-time) AND not spending as much time in a physical office reduces the exposure to these "traditionally male cultures" might make software development more attractive to women.

Just some random thoughts, it will be a long time before we know the effect.

Have you ever tried to work from home while simultaneously caring for children? The children demand ~120% of one sleep-deprived adult’s total available attention.

How did we lose the men caring for children?

My impression (as a stay-at-home father) is that there are more men caring for children than ever before in the US. Primarily due to better access to careers for women and less (though still non-zero) social stigma.

Yes, I did it at the beginning of the pandemic. I'm not implying that they would be working from home simultaneously with caring for children. Just that their schedules could be more flexible to deal with the various things that come up in parenting easier than when working from an office.

Source: myself being a parent that has benefited from this very thing.

Content-free.

TLDR: "So how did we lose the women in computing? They did not just leave; they were pushed out. There is hard work ahead of us to start to undo the damage."

This was published in 2018 it seems; has there been any change, additional information, or improvement since then?

Reply on news.ycombinator.com