The horizon is not so far as we can see, but as far as we can imagine

Lies, Damned Monopolies, and Sex Difference Statistics (UPDATED)


(POST BY MANDOS – was posted early quite by accident, postscripts below basically with how I planned to conclude)

L’affaire Damore

The whole kerfuffle about the Damore manifesto at Google and James Damore’s subsequent firing led to the dredging up of old debates about the sources of underrepresentation particularly in technological professions such as software engineering. The basic idea is quite old: Merely, that the current structure of society, as reflected in by group access to well-paying professional positions (substitute for historical versions: land, money), is obviously the outcome of the “natural,” “biological” difference between those population groups, in this case, gender.

Biological sex has been viewed as one of the intractable differences in terms of career representation in technical fields for some time now, holding especially for programming. The negative reaction to the Damore manifesto (including his firing), which consisted primarily of polite, “geekified” versions of old internet arguments defending sexism, is viewed as denial of the underlying facts that drive the apparent intractability of this difference.

The logic behind the claim is an enormous act of begging the question. Social change is slow and difficult — even the legal architecture that prevented women from living independent lives took a long time to change. Diversity advocates’ underlying claim has always been that rectifying differences in the economic success of identity groups requires change greater than that of mere legal emancipation, but instead deeper changes in employment practices and cultural representation. To claim that women are underrepresented in technology because of biological differences is to dismiss the idea that culture matters: It is using the outcome as the evidence, a fallacy.

Instead, the evidence to justify tech employment differences in biological terms is a red herring, and l’affaire Damore is simply a reflection of the fact that Big Tech is little more than a series of entrenched, possibly natural monopolies.

The reality of programming jobs

Let’s start with the brass tacks. Even in Big Tech like Google, most programming jobs do not really require either very high intelligence, particularly as expressed in mathematical ability as is the implicit argument, or even most of the curriculum of university four-year computer science undergraduate programs to be successful. Competent programming in itself is not a high-intelligence skill. The basic skills can be learned by a committed but average person in the way that singing in a community choir can be learned by a committed but average person. Indeed, the analogy stretches farther: For most large scale projects, beyond basic programming skills, one typically requires good organization, collaboration, architecture, standards, respect for requirements and so on to succeed. A full SATB choir is much the same, it requires some skill, but more importantly, it requires the ability of the signers to integrate the content of different musical scores into a single melody. A singer who is too “primadonna” and does not have the social cognition to fit her voice into the choir can make a near-professional school or community choir sound like beginners. It is no accident that even in the open source/free software community, a great deal of emphasis is placed on version control and design process in order to produce the fairly sophisticated software on your typical Linux desktop.

There are certainly technology jobs that require the higher “puzzle-solving” mathematical ability that people associate with success in IQ-testing frameworks. Areas such as advanced cryptography, compression, and parallel processing sometimes require this kind of reasoning, as well as even the content of graduate-level university algorithms classes. Suffice it to say that these form a small minority of tech employment at companies like Google.  Even advanced computer security work only occasionally involves a small amount of this kind of thinking: What matters far more is the ability to understand and study the human factors and fallibility associated with security regimes.  If gender disparities in programming jobs had been confined to those limited fields, a biological explanation would have held more water.

Many of these kinds of the skills required by the majority of programming jobs are quite easy to associate with stereotypical female characteristics. Many of us may have been raised by a mother who knits. If you have, you may realize that she had been forming fairly complex abstract patterns with her hands even as her attention may be focused elsewhere. Indeed, women have been writing knitting patterns that produce complex structures in an abstract “assembly language” for quite a long time. Numerous other stereotypes pertaining to female creative styles and diligence serve perfectly well to “justify” extensive female employment in software engineering. So the question remains: Why are they not employed in software engineering? Either female ability and inclination really is, in the majority, unsuited to technology jobs, because they are indeed mostly of the category that requires the intelligence that men have marginally more of, or the story I’ve told you so far is probably true, but some other factor intervenes, one that diversity attempts in the tech industry do not address.

The red herring of sex difference in intelligence

Pretty much everyone seriously discussing this topic acknowledges that all human characteristics stem from a combination of factors, including genetic. A starved child with the same genes as a well-nourished child is simply going to look different, although they will also look similar. However, the line between a “genetically-inherited” characteristic and “socially-determined” developmental factor is anything but clear. It’s increasingly clear that some characteristics can be set by the social conditions under which your mother was an unfertilized egg. Which happens when your grandmother was an embryo, the time when ovaries are stocked. Which means that your great-grandmother‘s social conditions may have some (still to be determined) inherited effect on your life today. And there are other potential variations of this situation.

Nevertheless, we know that some characteristics are inherited, and some are sex-linked. A male and female child raised in exactly the same conditions with the same genetics (other than sex-chromosomes) most probably will develop grossly different visible physical characteristics above and beyond the primary sexual characteristics of genitalia. No one questions this — except, of course, that even the categories of male and female are actually more fraught and complicated and ambiguous than previously considered, because biology is almost never “cut and dried.”

So where does this difference come from? Some biological differences are indeed incidental (rare is the gene that affects strictly one characteristic), but the high degree of physical sexual dimorphism is too systematic to deny the effect of selection pressure, particularly sexual selection. A “just-so” story in which male physical strength is connected to the differential ability of getting female reproductive attention is relatively plausible and seen throughout the animal kingdom even in the present day. Naturally, it applies the other way, and the characteristics that suggest female reproductive fitness are, yes, emphasized in females. You would therefore not be entirely remiss to imagine that this would apply to cognitive characteristics, particularly intelligence.

I will not do a rundown of the evidence on this topic, but instead refer you to, yes, Wikipedia, which has a discussion with references to both sides of this debate. What I will instead emphasize is that while some studies show statistically significant differences between the sexes, the effect sizes tend to be small and apply to (as mentioned many times in this debate) the margins. Furthermore, many of the effects seem to disappear depending on how you account for confounding factors. This is how you should understand situations like this: When, over the course of a number of studies, it becomes apparent that quibbles over confounds are deciding factors in the statistical significances of competing results, you should be skeptical of attempts to apply such results to other areas — they are almost certainly small contributors to the overall effect under discussion, such as women’s employment in technological fields. The area of biggest difference between the sexes seems to be in spatial inference, but as I mention above, the applicability of this to software engineering ability is questionable, and it’s independence from social factors also potentially doubtful, a criticism that applies to quite a lot of intelligence research overall, even if one accepts the existence of “classically” genetically heritable factors of general intelligence.

Postscript I – inclination

It seems remarkable that despite the degree of “gross physical” sexual dimorphism among humans, the evidence does not really show up anywhere nearly as strongly in the characteristic of intelligence, despite its obvious importance to human survival.  I can only suggest that the sex-differentiating pressure of sexual selection fails to apply to intelligence for reasons that have an easy “just-so” story: both males and females for most of human history had much less day-to-day differentiation in “survival labour” as they do in the very recent industrialized society, and they had no way of performing any sort of fine-grained intelligence test on mates, at least for the intelligence that relates to very abstract mathematical abilities.

But if general intelligence and even specific ability fails to explain the magnitude of the gap in software engineering employment, one could still argue, as Damore and his antecedents do, that there is some other tendency away from technological careers for women, and some tendency towards them by men.  Many of this type of claim involves attributing to women personality traits like “neuroticism” and “gregariousness”, which are very old accusations indeed.  However, the same considerations apply as they do to that of intelligence: how much of this gap can be explained by an evolutionary “just-so” story, and how much of an alleged personality gap can be explained instead by, well, the conditions of patriarchy.  Feminists have argued for a long time that the greater presence in women of seemingly self-defeating personality traits is actually the result of very immediate psychological self-defense mechanisms from overt and subtle ills deeply embedded into patriarchal culture — among them, the need to pre-emptively appease a potential oppressor, a kind of psychological insurance policy created by immediate conditions rather than a genetically entrained cognitive difference. The arguments of Damore and his ilk usually fail to take into account these types of explanation for a sex-based personality gap — in fact, feminist arguments about “female” personality are disregarded to an extent that arguments for the social construction of intelligence differences are not.

And precisely how a biological personality or intelligence gap can be used as an explanation for employment and pay gaps in one particular type of industry when other high-status jobs (e.g., like it or not, the aggressively-competitive financial services industry) show greater increases in female employment is rather hard to discern.  It requires that one attribute characteristics such as “status-seeking” as particularly necessary in software engineering employment in a way that is not necessary in other professions, something that really runs counter to the collaborative nature of successful software projects. Instead, feminists have long documented how girls are discouraged from even considering technology careers at a point well before the diversity efforts of large corporations actually take hold.  Instead of engaging with these details, Damore and his antecedents fail even to mention them and the history of work on this type of topic, but yet demand respect for “conservative opinions”.

Postscript II – monopoly

As I said at the beginning of this post, a part of the instigation for Damore’s essay is justified in that the efforts of Big Tech to encourage diversity, including women’s representation, appear to have had little effect particularly on employment in companies like Google. The truth is, companies like Google have a large incentive for the appearance of encouraging diversity, but little for the actual practice of diversity, even if such would help them avoid debacles such as the failure of Google+, which was in many ways the result of a culture that failed to understand the human factors that apply to populations that are different from those who work at Google. (For example, early on, Google+ attempted to use “true names” filtering that expelled people whose real names look “fake” to someone who works at Google because of cultural difference — it’s very difficult to rectify this kind of error post hoc for a nascent social media entrant competing with an incumbent like Facebook.)

Why would this be so?  Because these companies don’t really exist in competitive markets.  Google’s attempt to “muscle in” on Facebook’s turf with Google+ was never really, despite the hype, about rectifying a gap in social media offerings that really threatened the survival of the company.  Google has an entrenched near-monopoly in key aspects of internet architecture that is very hard to displace.  Consequently, it can leave the potential gains from social diversity on the table without appearing to suffer overmuch.  In the short-to-medium term, even for a company like Google, it is costlier to attempt rebuild internal culture, at the risk of alienating the incumbent constituency whose voice Damore made impolitely audible, than to spend a measly few hundred million on cosmetic diversity placebos.   In reality, the material merits of improving the inclusion of underrepresented groups in “technological power” will only be seen when the vicious cycle that actively keeps them underrepresented is broken through broader cultural change, imposed somehow from without — and over and above the immediate economic incentives afforded to monopolies.

And that is not surprising — whoever imagined that a monopoly- or oligopoly-dominated society would lead to the optimal outcome for the most people?


Nazis Are Not Socialists


Oppressive Precedents Used Against Nazis Will Be Used Against the Left


  1. In my twenty-five years in the business I have found that tech managers don’t necessarily need to technologically skilled, they just need to know how to bully those they manage… who are in fact in need of management, and the most likely to complain. About everything.

  2. tony

    You of course ignored almost every one of his arguments which focus on personality, not intelligence, differences and men’s higher drive for status. Good job.

  3. atcooper

    The assessment is totally dead on with the nature of the work. Very little comp sci needed for the vast majority of computing work. So much of it is rote.

    Women do fine in tech if given room to breathe. I’d point to the little discussed defense funding backbone of silicone valley as one among many contributing factors to the cowboy bullshit.

    It’s really only a big deal because these are among the few decent paying jobs left. If it’s steady of course.

  4. V. Arnold

    Here’s a link to a video interview with James Damore and Stefan Molyneux;

    Very much worth a listen for the real deal.

  5. Hi all, this got posted before I was actually done with it, either because Ian interpreted the length of it as a sign it was good enough 🙂 or because I accidentally pressed “publish” before realizing it. I’m going to leave it up but wrap up the final points I wanted to make and make some edits before the discussion goes too far.

  6. someofparts

    You know, it used to be an enduring hassle at feminist bulletin boards that there would always be some dude who stumbled into the community just to enlighten the ladies. Of course the points he made would invariably be elemental things long ago roundly discussed and decided in the movement.

    I guess my point is that all the yakking from the fellas in that video reminded me of those trolls. They see themselves as debating. The women they interrupt see them as toddlers who think they are entitled to dominate conversations they don’t even dimly understand.

    Also, please spare me the tech worship. Well, I guess Ian gave the idea that being able to code makes you a genius a solid burial already. That said, it still makes me despair when people dismiss all human learning outside of the STEM fields as hokum.

  7. someofparts

    Posted at Angry Bear –

    Various studies do seem to indicate innate preferences. Who knew.

  8. OK, my accidentally-posted draft has been updated with postscripts that cover what I originally wanted to say more fully. Virtual embargo over 🙂

  9. tony: that was an accident, I did indeed want to discuss the topic of non-intelligence-based inclinations, as I do in the first postscript above.

  10. someofparts: do look at the comments to that post for critical evaluation of the problems with studies like that which are very hard to do well, and furthermore, very hard to connect with the requirements of complex industrial professions. For example, I could argue that if such preferences were real and do apply to adult professional choices, it is merely the case that a job like programming needs to be presented in the terms that traditional female occupations, such as weaving, are presented.

    The author Kimel, in comments, also reveals a particularly odd idea about how much time generational cultural change takes at a deep level, particularly when only formal equality is achieved and the rest given over to laissez-faire. Indeed, even if men and women were cognitively identical, there is no reason to believe that the gap couldn’t last decades and decades, if it is expected to be rectified purely voluntarily and only with “free-market (with monopoly)” economic incentives.

  11. FriarTuck

    Preface: I am a development manager of a small team (former programmer) without a formal degree in CompSci.

    I agree with you that programming in of itself doesn’t take any great degree of intelligence. Anyone with a basic text editor can come up with web-oriented markup or scripting that does something useful. Even compiled programs aren’t too difficult when you come down to it. The most demanding math that would be required is basic algebra until you get down towards the metal.

    What is difficult (but doesn’t necessarily require intelligence of the type you’re expounding on) is exercising good judgement of the different tools and standards available to plan and construct something that doesn’t fall apart under its own weight, that is usable to end-users, and is extendable and maintainable.

    What does require high intelligence IMO is the creation of the tools and toolsets. The unfortunate thing, especially in web programming (server or client side), is that there are an awful lot of people who think very highly of their own intelligence. In that way, there’s always a new fad framework that is sometimes better but usually worse in some way, with a great big compromise that whacks you, that a great many people latch on to as the “next great thing” (cough Angular.js cough) that lasts a while until someone else comes up with “the next great thing.” And unfortunately, a lot of these toolsets appear to be monuments to their creator’s idea of their own intelligence. They’re difficult to use, perhaps powerful in one specific way, and basically require you to build yourself into being one of the “club of understanding.”

    In my experience it takes a unique kind of person, man or woman, to do the more complex tasks in CompSci. Yes men might more commonly have outliers that exhibit these traits – and they ARE outliers -, but I’ve met women that are just as introverted, just as… odd, and just as genius as their male counterparts. It may also be that men with these traits are more encouraged, more accepted to live with these traits and not conform than women.

    What I find is funny (but particularly sad) is that in CompSci history, women and men occupy various roles of achievement in the development of our modern idea of computers, without equal acclaim. In computer gaming, yes, John Carmack may be a wizard of the algo and co-creator Doom, but Dona Bailey was the co-creator of Centipede which arguably touched more people.

    Typically when searching for frameworks, I look for something that will last a long while, that is simple and more importantly understandable to the people I supervise. If something claims to be faster and better, yet is not human readable unless you twist your brain sideways to deal with compromises baked into it, it’s next to useless. I don’t care who built it.

  12. Steeleweed

    “Competent programming in itself is not a high-intelligence skill. “

         I strongly disagree. Most programming done today is ‘close-enough-for-government-work’ and requires only high-level languages and a cavalier attitude towards quality. Competent programming, on the other hand, does require significant intelligence, as well as some less common qualities.

         I spent 50 years in IT, most of that time doing ‘bleeding edge’ work, building systems that were state-of-the-art. A handful of us invented the ‘profession’ of Systems Programming. For about 20 of those years I was recognized as one of the top 15-20 people in the world at what I did and every couple of weeks I got job offers from major corporations. I have written code that boggled the mind of experts in the business. Even today, being long retired, there are people out there who think I can walk on water. I know whereof I speak when it comes to designing and programming systems – and what is required to do it right.

         I around 1965, a Senior VP told me that “by 1970 we will need 50,000 programmers in the US”. My response was that there weren’t 50,000 people in the country capable of being programmers. We were both right: we got a handful of geniuses, a few hundred very good programmers and 49000+ crappy programmers. We’ve got more people writing code today but the proportions haven’t changed much, as the crappy software out there demonstrates. And part of the problem is that programmers (and managers) do not recognize that programmers cannot necessarily design systems. (Hell, for that matter, most programmers cannot really program).

         As far as the people, I found that the person with an CS degree had about a 6-month headstart over a complete neophyte. What needed to be learned had to be learned OJT. My hiring criteria were not gender-based. I hired on Integrity, Intelligence and Enthusiasm and never had reason to change that approach.

         Women were about 15% of the staff, on average, as programmers and managers. I found them better than average in both positions. (I can only recall only one female programmer-cum-manager who was not up to snuff, while at least half of the males’ work quality was well below their pay-grade).

         It is quite possible – even probable – that only women who were exceptionally competent would be successful. And that fact pretty much demonstrates the sexism that existed – and still exists. Reminds me of the old joke that “For a woman to be considered half as competent as a man, she must do everything twice as well. Fortunately, this is not difficult.”

  13. Max Osman

    I don’t agree with a lot of what you say but this essay is epic. I like how it ties in your argument
    about oligopoly at the end.

  14. DMC

    2 words: Institutional Bias. Geeky white dudes hire more geeky white dudes. Put of by POC’s , women and other aliens.

  15. Synoia


    We meet again.

    Systems integration of disparate vendors was the most fun. Call centers were my most extreme example.

    I discovered that most people do not have the discipline to follow the scientific method, and prefer to guess as to cause, as opposed to think through the symptoms and possible causes.

    I never met many people doing complex systems integration, and cannot speak to the male/female ration of that work.

    I’m a little puzzled in the male programmer vs female programmer debate. When I started in the business, the ratio appeared about 50:50, because programming was seen as a new skill and training provided by employers. Anyone with a pulse and working brain cell could apply (The difficult hurdle was the working brain cells).

    In working with programmers I remember working with more women than men, but that might be my subjective bias.

  16. FriarTuck: indeed. And as you seem to suggest, there’s no reason why the intelligence required to create well-designed toolsets should correspond to the traits generally considered to be the relevant sex-linked ones that allegedly give men an advantage.

    Steeleweed: we agree in principle, aside from the quibble on what it means to be a “competent programmer”. My argument holds as long as women are underrepresented at the levels of competence at which most programming jobs are already filled, regardless of how low or high that appears to be, except possibly the extreme ends of certain very specific areas that might engage certain forms of cognition that require the marginal skills that some studies show may be better developed in men. We are very far from the point at which those are the only significant employment differences relative to representation in the workforce at large.

  17. Peter

    Mandos’ manifesto seems a tweaker like rant attempting to cover all related topics while avoiding the core of the Google story. Deflection and misrepresenting what people have written is an old tactic used to muddy the waters and keep people from addressing the bad behavior at Google.

    Too many people, management and workers, men and women, were involved in an attempt at political cleansing that may have included actual crimes but certainly was unethical. This conflict wasn’t about which side of the debate was the most correct but about how one side used its power to force others into submission or banishment.

  18. Peter: I like that you think of the above as a “rant”.

    Damore’s firing was an inevitability, because he would thereafter have been seen as creating a hostile workplace. Certainly his career was dead-ended by that point: who would trust him to supervise any team which had women or minority members in it, or to participate fairly in hiring decisions with an eye towards eliminating discrimination?

    People who express discriminatory ideas should not be surprised when their employer assesses their suitability for the position they’re in. Smaller offenses would also get you kicked out of Google.

  19. nihil obstet


    When I started in the business, the ratio appeared about 50:50, because programming was seen as a new skill and training provided by employers.

    In the 60s, IBM actively recruited “FLAG”s, Female Liberal Arts Graduates, to train as programmers. The thinking appeared to be that most of the time spent in programming was sitting in front of a keyboard (you know, like a typewriter) and that was women’s work. For women, the entry into business was to go to an upscale business school for a year after college to become a top manager’s assistant, so IBM management was used to very bright women doing jobs that men would require a different title and higher salary for. As programming became more in demand and salaries rose, higher status followed as did the determination that men were more suited for this higher status career.

  20. The Stephen Miller Band

    Corporations are Private Tyrannies. Pretending that Liberation & Diversification & Egalitarianism can be achieved via Corporations is nonsensical. If we seriously want to address the most pressing & urgent issues facing Humanity & The Planet, the Corporation has to go, and go fast. Defending it in any way as a tool of Diversification & Liberation & Egalitarianism is insane.

    The People’s Party will end Corporate Personhood and Too-Big-To-Fail. Organizations will be horizontal and employee-owned & operated. There will be no Monopolies because there will never again be Too-Big-To-Fail. Oh, and of course, Wall Street will be no more.

    You think it’s too complex to accomplish? Not according to Colonel Mustard it’s not. For once I agree with George Peppard.

    You should not think that complex situations are necessarily complex in fact. Often they are merely matters of personality. ~ Pat Lang

    A Person Who Knows A Thing Or Two

  21. Matt

    Women choose largely NOT to study computer science. 5 men graduate with a CS degree for every woman.

    Start THERE.

    There is no magic after that to all of the sudden even the score in the workplace.

  22. Steeleweed

    @nihil obstet
    In the ’60’, IBM doubled in size in a few years, mostly staffing for the /360 that came out in 1964. They hired pretty much anyone with a degree in any subject. To them, a degree indicated some minimal level of intelligence, maturity and self-discipline. They knew they’d have to train the employees (and there were few if any CS curricula in place in academia at the time). They also didn’t want to see an uneducated person climb the ranks and end up dealing with high-level customers. I knew a HS grad who started as an operator, then became a salesman. He was smart and ambitious, a good operator and a good salesman but I think they strongly pressured him to get a degree just to make him a more ‘suitable’ representative of the company.
    As computers began to replace clerical jobs, which were mostly done by women, many companies simply converted their clerks to programmers. That process was problematic until PL1 came along and a 5-year-old could produce useful work. Not necessarily great code but at good enough for their purposes.

  23. Synoia

    @nihil obstet

    The thinking appeared to be that most of the time spent in programming was sitting in front of a keyboard . In the ‘6os?

    In front of keyboards? Not a chance. It was coding sheets, Data Entry Departments, and print output.

    I was nowhere near US cultural rigidity in the ’60s & ’70s. Almost half the globe away.

    27 deg S and about 30 deg E.

  24. highrpm

    so who says pursuing diversity is a worthy and good goal for a society? where’s the scientific proof?

    its simply the religion du jour, pushed by media high priests. “gimme that old time religion…” rings in my ears. religionists have not delivered on their promises of making societies better. more often than not, “hardened beliefs in imagined truths” polarize and divide and fuel wars.

    utter nonsense. do something good today. stick to picking up cigarette butts on your daily walks and litter that our sordid underclassers leave their neighborhoods graced with daily. while the public education system does not teach normal acceptable behaviors for living in community as a contributing member. at a minimum, members should be reminded of the backpacker’s mantra, “leave no trace” and “minimize your presence.”

    and me old whitie boy, i live in the black neigborhood. and it sucks.

  25. highrpm


    “Competent programming in itself is not a high-intelligence skill. ” similar to the orbits of local competition and the professional who make their living in the national level of competition, it’s know and accepted in programming circles that you cannot call yourself a profession until you’ve internalized/ mastered several — 3 to 5 — programming languages. (dreyfus model of skill acquisition.) and that, good author, takes both 10,000 hours of focused practice and higher than average I.Q. i programmed in C in corporate for 15 years and never internalized or mastered it as my single language. ain’t got the requisite neuron count. for whatever reason. i supported old line products. and performed acceptably. and the top level eliters dispised us lowlies, but accepted that we were custodial programmers and necessary.

  26. Synoia


    I believe those metrics nonsense because the phrase “internalized/ mastered several — 3 to 5 — programming languages” cannot be measured.

    Results count. How well do you deliver code and how many errors per kloc remain in your code.

    The IT sector is so much more than code. That’s why good salespeople and sales engineers make so much more than programmers.

  27. Steeleweed

        In my view, a computer ‘system’ entails hardware, software application design, coding, testing, documentation, operational procedures, backup & disaster recovery, ongoing support (changes, updates, additions), education of both technical personnel and end users. All these aspects must be in my mind when I design a system.
        As far as languages go, I learned Cobol, Fortran & PL1 but seldom used them. Over the years, I used ‘scripting languages’ like VM’s Exec & Exec2 & REXX. When PC’s came out, I programmed in assembler and Basic on TI994A and ‘IBM’ PCs. I’ve used Window scripting and coded several ASP-based sites as well as some PHP work on WP sites.
        However, most of my programming has been IBM Assembler, which is immensely powerful but requires a thorough understanding of the hardware, in addition to programming principles and logic. The IBM instruction set is so much more powerful than Intel that programs are much more concise and less complicated (as well as faster and significantly smaller). The programs are much easier to debug and less risky, simply because a mainframe Assembler program might contain 2000 lines of code to be debugged while the same work done in Assembler on a PC might contain 20,000 lines of code – which is why most coding is done in high-level languages, at the price of high memory and cpu requirements.
        Companies throw memory & CPU at a problem because it’s cheaper than paying for better programming. The resulting code is inefficient. Try running a server beyond 50% CPU and you’re likely to crash. A mainframe can run at 95-100% CPU (for years) without crashing. .

  28. brian

    Women just aren’t as interested in computer science as men. 20% to 10% of applicants, depending on how prestigious your company is, are women. A similar percentage of those are actually good as men. It’s pretty simple why women prefer communications majors, health care majors over computer science. Computer science lowers your chances of having and attracting a mate, more so for women. It makes you sedentary and get really involved with a screen all day. The most attractive people are represented in computer science. The real famous coders of the world were and still are, probably around a 4-5 on the attractiveness scales. For women to be unattractive much strongly effects her biological success of finding the highest scale mate, where for men the hit to attractiveness is much lower since financial security and status are much more attractive to women, while men are still mostly attracted to looks and caretaking ability. Ignoring the complexity and trying to reduce it to simple variables will not work. Women and men are fundamentally different and will be as long as women have wombs and give birth and men have seed and don’t give birth no matter how much all else changes. You can take the human out of nature, but you can’t take the nature out of humans (yet). Making it shameful to talk about differences is … insane.

  29. brian

    computer science have the most unattractive physically people in all professions.
    it’s more important to women then men to look good.
    therefore women shun CS.

  30. realitychecker

    “Simple” can also describe a mind.

    Just sayin’ . . .

  31. nihil obstet


    It’s a minor point, but yes, keyboards. Not electronic. Punch cards.

  32. nihil obstet

    This whole discussion is going to convince me to agree with that wretch Margaret Thatcher when she said there’s no such thing as society!!! Is it better for “society” to reserve more access to power, influence, and resources to some groups over others? The only reason to care has to do with whether you can live your life as you choose. Members of the powerful group pontificate on what’s good for society to protect their power, and they experience anything that challenges their position as wrong and harmful.

  33. The thing is, Thatcher would have been right if she said “There is no such thing as a society.”

    What there is, is either society, as in an inchoate indivisible mass that includes all individuals capable of participating socially, or societies, shifting, overlapping subsets of that mass. But there is no such thing as a society, as in an easily separable organic unit with a life and reality of its own.

  34. Programming may not require higher math, but it does require being methodical and disciplined – things that a science-y college education tries to bring out. However, in my generation, growing up with the first wave of widely affordable computers, being a programmer meant – crudely stereotyping here:

    (1) having middle class parents who bought you a computer to play with as a kid
    (2) being told it’s ok to live the nerd life

    Such a background produces a combination of enough enthusiasm and a head start that you get on the path to tech jobs in computing (or science and engineering, or if you’re a special kind of person, business and leadership).

    I think those factors were more male, and so it’s largely a matter of the male culture that evolved. Women face a lot of distractions in such a world (just as in science), but there’s absolutely no physical reason they can’t do as well or better. It’s a matter of holding one’s nose wrt/ the culture.

    PS- I agree 1000% w/ Steeleweed about Systems with a capital S. Although I continue to be more and more impressed by the power of large teams wielding blunt tools (e.g., Java) – I’m sure you’ll agree that is the way of the corporate world. Most code isn’t systems code.

  35. roadrider

    ” Competent programming in itself is not a high-intelligence skill.”

    I have > 30 years of experience in software development. I’m not a CS graduate (life sciences) and am mostly self taught (with some coursework) but had the opportunity to work in some advanced development environments (financial, bioinformatics) and some not so advanced. I think you’re really misrepresenting the intellectual qualities required to do “competent” programming. And no, its clearly not a male/female thing – the gender disparity is mostly a social phenomenon not one of ability or intellectual capacity.

    One not need to be a genius to be a competent programmer but to say its not a “high-intelligence skill” is bullshit. There are different kinds of intelligence. I’ve know and worked with many very intelligent people who were poor programmers because they could not think abstractly, could not see the big picture in terms of design and code reuse and were flummoxed when the problems scaled from average to more than average complexity. All of those people could write trivial programs as can many non-tech people who dabble in writing small programs or web sites for their own purposes. The thing is, its easy to write code that “works” without realizing the limitations in design, bugs that weren’t revealed by the simple use cases they tested, limitations in scalability and many other issues that most inexperienced programmers fail to consider.

    Developing software is more of a craft than a science. It does require above average intellectual capacity but in specific types of intelligence that are related to the job. I should also mention that being able to translate business requirements into well-designed and carefully implemented programs also requires above average intelligence, perhaps more than is required to implement a design created by others.

  36. Adrena


  37. Adrena


    For every ugly woman there is an ugly man

  38. nihil obstet

    Recent interesting short by the BBC: Girl toys vs. boy toys: The experiment. It’s 3m25s long.

  39. Billikin

    Steeleweed: “Women were about 15% of the staff, on average, as programmers and managers. I found them better than average in both positions.”

    As has been pointed out in a number of places, that is indicative of anti-female prejudice, whether internal or external.

    Synoia: “I’m a little puzzled in the male programmer vs female programmer debate. When I started in the business, the ratio appeared about 50:50”

    That was my impression, too. I suspect that there is a good sociological study or two there.

  40. Will

    While differences in cognitive ability are not very significant, differences in interest are. I represents unique challenges involved trying to get women interested in tech. Besides, there are plenty of female hacker/computer geek figures on TV shows and movies now, yet the rates of women in computer science continue to decline. This study is a great read, as it took in ALL of the relevant meta-analyses…the exact opposite of cherry picking.

    1. Gender differences in math/science ability, achievement, and performance are small or nil. (See especially the studies by Hyde; see also this review paper by Spelke, 2005). There are two exceptions to this statement:
    A) Men (on average) score higher than women on some tests of spatial abilities, such as the ability to rotate 3-dimensional objects in one’s mind. This ability may be relevant in some areas of engineering, but it’s not clear why it would matter for coding. Furthermore, because women get better grades at all levels of schooling and score higher on a few abilities that are relevant to success in any job (e.g., reading comprehension), we assume that this one area of male superiority is not likely to outweigh areas of male inferiority to become a major source of differential outcomes.
    B) There is some evidence that men are more variable on a variety of traits, meaning that they are over-represented at both tails of the distribution (i.e., more men at the very bottom, and at the very top), even though there is no gender difference on average. There is an ongoing debate about whether or not this is true across nations and decades; We are currently reviewing this literature, and will post our conclusions and links to studies next week.
    2. Gender differences in interest and enjoyment of math, coding, and highly “systemizing” activities are large. The difference on traits related to preferences for “people vs. things” is found consistently and is very large, with some effect sizes exceeding 1.0. (See especially the meta-analyses by Su and her colleagues, and also see this review paper by Ceci & Williams, 2015).
    3. Culture and context matter, in complicated ways. Some gender differences have decreased over time as women have achieved greater equality, showing that these differences are responsive to changes in culture and environment. But the cross-national findings sometimes show “paradoxical” effects: progress toward gender equality in rights and opportunities sometimes leads to larger gender differences in some traits and career choices. Nonetheless, it seems that actions taken today by parents, teachers, politicians, and designers of tech products may increase the likelihood that girls will grow up to pursue careers in tech, and this is true whether or not biology plays a role in producing any particular population difference. (See this review paper by Eagly and Wood, 2013).

    As far as biology playing a role in interest, in very well may. Diane Halpurn was the president of the American psychological Association and a very important figure in the field.

    In her preface to the first edition, Halpern wrote: “At the time, it seemed clear to me that any between-sex differences in thinking abilities were due to socialization practices, artifacts and mistakes in the research, and bias and prejudice. … After reviewing a pile of journal articles that stood several feet high and numerous books and book chapters that dwarfed the stack of journal articles … I changed my mind.”

    Why? There was too much data pointing to the biological basis of sex-based cognitive differences to ignore, Halpern says. For one thing, the animal-research findings resonated with sex-based differences ascribed to people. These findings continue to accrue. In a study of 34 rhesus monkeys, for example, males strongly preferred toys with wheels over plush toys, whereas females found plush toys likable. It would be tough to argue that the monkeys’ parents bought them sex-typed toys or that simian society encourages its male offspring to play more with trucks. A much more recent study established that boys and girls 9 to 17 months old — an age when children show few if any signs of recognizing either their own or other children’s sex — nonetheless show marked differences in their preference for stereotypically male versus stereotypically female toys.

    Halpern and others have cataloged plenty of human behavioral differences. “These findings have all been replicated,” she says. Women excel in several measures of verbal ability — pretty much all of them, except for verbal analogies. Women’s reading comprehension and writing ability consistently exceed that of men, on average. They out­perform men in tests of fine-motor coordination and perceptual speed. They’re more adept at retrieving information from long-term memory.

    Men, on average, can more easily juggle items in working memory. They have superior visuospatial skills: They’re better at visualizing what happens when a complicated two- or three-dimensional shape is rotated in space, at correctly determining angles from the horizontal, at tracking moving objects and at aiming projectiles.

    Navigation studies in both humans and rats show that females of both species tend to rely on landmarks, while males more typically rely on “dead reckoning”: calculating one’s position by estimating the direction and distance traveled rather than using landmarks.

    The Stanford Article is a great read.

  41. Except, even if that all panned out, you still need a theory of why computer science needs some of those inclinations and not other fields. There’s lots of reasons why a “landmark-finding” ability is perfectly cromulent with the practical skills needed for e.g. debugging code. I’d trust a good landmark-finder rather than a dead-reckoner for that — why wouldn’t I? It doesn’t explain the sudden change in gender balance.

    The persistence of the CS gap for women is still much more easily explained by social factors, such as the way it became male dominated in the first place, how it became in some quarters an idolized career for young men with hostility to women (and it did), how a lot of “tech bros” still do view it as a bastion despite corporate diversity bafflegab, etc, etc.

Powered by WordPress & Theme by Anders Norén