Where are the women?
Yassamine Mather looks at one area where gender inequality looks set to continue
These days it is rare for a day to go by without hearing about capitalism’s efforts in support of gender equality. Business programmes on radio and TV tell us about the drive to have more women on the board of major companies. Political parties compete with each other over the number of female MPs. In most cases quotas are expected to deliver equality and - if you are not concerned about that unfashionable notion, class - at least on the surface women in the middle and upper-middle classes seem to be achieving equality on many fronts.
However, this hardly tells the full picture. First of all, the main concern remains achieving gender equality amongst the elite. Working class women are told, day in, day out, that the success of women on the board of the FT’s 500 top companies demonstrates equality. Nothing could be further from the truth. Women at the top of capitalist transnationals are just as ruthless as men, when it comes to the exploitation of the working class - why else would they want to be on board in the first place? And, rather than increase women’s pay, they reduce that of men to the often already casualised, contracted female employees. In the real world (beyond the media bubble) women’s wages remain lower than men’s. As everyone knows, the much heralded pay increase demanded by the BBC’s China editor, Carrie Gracie, would not have translated into a decent wage for women cleaners, working for unscrupulous contractors in that august institution.
However, even as far as gender equality amongst the educated and the elite is concerned, women are disadvantaged in one area where no amount of positive discrimination or use of quotas can make a difference: that is computing science and in particular in areas that will affect our future: parallel processing, high-performance and quantum computing, and code writing associated with these sectors.
Recent advances in computing are based on the use of newer hardware - in particular graphics processing units (GPUs) and the evolution of quantum science. Mastering GPUs and the parallel programming of GPU nodes (servers), the understanding of algorithms in high-performance clusters - these things require a good background in mathematics. Such machines are now used extensively in all aspects of scientific computing, including the ‘machine learning’ associated with artificial intelligence and programming computers to perform specific repetitive tasks.
Such programs progressively improve machine performance, the idea being that this will eventually pave the way for the automation of mundane tasks. For example, whereas current automated tills fail to work eight times out of 10, the latest models have ‘learned’ how to deal with almost any situation, including customer errors.
Quantum computing, which will probably become the dominant form in the future, is based on the intersection of maths, physics and computer science. Using quantum-mechanical phenomena, such as superposition and entanglement, a new generation of devices are coming into existence that will outperform existing machines. Unlike existing digital computing, where the data is coded into ones and zeroes (binary), quantum computation uses quantum bits or ‘qubits’, and these can be in several states, which can be ‘superposed’ with each other! While no-one fully understands the physical laws involved in these effects, the implication is that complex problems with many potential solutions - climate modelling, orbital mechanics, or protein folding in cells - can be worked on very quickly in comparison with digital computers.
What does all this mean for women? MIT’s minimum requirements for entry to a course on quantum computing lists high qualifications or expertise in quantum mechanics, linear algebra and probability, random matrix theory and asymptotic analysis, to name just a few.
And here lies the problem. In all of such subjects women account for only 10%-12% of the undergraduate population, and that percentage drops further at the postgraduate and post-doctoral levels. No amount of artificial quotas or positive discrimination can make up for that. In most spheres the proportion has not improved over the last three to four decades - in some cases it has actually got worse.
Yet women have been among the pioneers in some fields of computing science, such as code writing - where Ada Lovelace in Bletchley Park and Grace Hopper in Harvard became prominent during World War II. In fact the number of women studying computing science was growing until the mid-1980s, when the proportion got as high as 37% - before dropping back to 11% in 2017.
The sad reality is that for decades generations of young girls have been discouraged from pursuing maths and physics. According to one study, published in 2016, the problem starts at a very young age:
… there is a tiny gender gap when kids start school (albeit larger among the very top performers) and that it widens, across all ability levels, through third grade. That’s a critical time frame, as past research shows that early math achievement determines a child’s interest and confidence in the subject during elementary and middle school, and strongly predicts how good at math she’ll be later on.1
At secondary school those who do well at maths or physics are often encouraged to choose other subjects at degree level. That is because teachers often underestimate girls’ mathematical abilities - even when they have similar or identical capabilities to male pupils.
In a Planet Money podcast, Caitlin Kenney and Steve Hen explain what has happened:
it was in the early 1980s that the narrative first emerged that computers are for boys. The first personal computers weren’t much more than toys and they were marketed almost exclusively to boys and men. Computer geek culture also began to emerge during this period and TV shows, movies and video games all reaffirmed that computers were the domain of boys.2
So we have ended up with a situation where women make up a very small percentage of software developers - a mere 11.2%, according to one 2013 survey. The women who choose to work or research in this sector face discrimination in an industry that remains male-dominated. Once they start a career as code writers, that discrimination intensifies.
In February 2018, a story about GitHub - the giant repository used by more than 12 million code writers - made the news headlines. Researchers at a US university found: “GitHub approved code written by women at a higher rate than code written by men, but only if the gender was not disclosed” (my emphasis). Out of three million ‘pull requests’ submitted on GitHub, “code written by women was approved at a higher rate (78.6%) than code written by men (74.6%)”.3 So clearly, once given the opportunity, women are capable of writing good code, yet they account for less than 12% of code writers.
Why does all this matter ? It matters because of the developments I mentioned earlier. We are moving towards a period when mundane tasks will be automated, once machine learning and artificial intelligence using new hardware allows developers to iron out the current deficiencies. Once that happens, repetitive jobs, including many in the clerical and service sectors, will be automated. Most such jobs employ a high proportion of women, which means their economic future will be bleak under capitalism.
Of course, those code writers who determine aspects of this automation - within the confines of global capital’s requirements - will prosper, but, for all the talk of gender equality, very few of them will be women.