The hidden engineering gender gap

By Joyce Park 01.2.07

(Editor’s note: This essay comes in two parts. Here is the first part.)

Anyone who has met John Doerr knows that he is genuinely passionate about encouraging the participation of women in the venture-creation system: as VCs, as entrepreneurs, and especially as engineers. He was the first person I ever heard clearly articulate something that I think many of us suspect in our hearts but are hesitant to say.

With all love and respect to our sisters in product management, marketing, sales, finance, HR, and G&A, 50 years of Silicon Valley history strongly suggest that technology companies will ever continue to be founded by entrepreneurs from engineering backgrounds; and if women never become engineers in sufficient numbers, they will disproportionately fail to experience the upper end of the range of Silicon Valley outcomes.

Furthermore, it’s impossible to calculate the opportunity cost to Silicon Valley ventures due to insufficient diversity of backgrounds, ideas, and modes of thought; but as the end consumer of our work becomes increasingly female, these costs must be rising.

However, no one can have failed to notice that despite all the efforts of a great many fearsomely brilliant, motivated, and determined individuals and groups to support women — Mr. Doerr and a few other funders, the Anita Borg Institute, Google and Yahoo and HP, Mills College and Stanford and Berkeley — there are still not nearly as many working female engineers as male.

And when it comes to female tech entrepreneurs from an engineering background I believe the ratio is even more skewed, although no one seems to even know a definitive number. Finally, when you count the women who head venture-backed businesses — which is an utterly arbitrary distinction except insofar as it highlights the availability of capital and other resources which can ease the crushing burden of starting a business — it is a lonely little group indeed.

But I’m not here to wring my hands and whine about the status quo. Nor do I plan to propose computer science education as the long-term solution to the gender imbalance problem, because far better-qualified people than I are already working along those lines. Instead I want to focus on a well-known but little-studied phenomenon in the technology industry — particularly in the newer, more experimental, startup-driven subdomains — and in the Open Source movement. This enigma is the lack of self-taught female software engineers. (Throughout this essay, by “self-taught” I mean someone who lacks a degree in computer science and has taken personal responsibility for training him or her self as a software engineer. Some of them may have enjoyed on-the-job training or other instruction, even including non-degree-granting university-level instruction, but in my mind they still count as self-directed.)

Everyone in professional or Open Source software development has worked with countless male colleagues who are essentially self-taught, lacking all or most formal training in computer science, and in not a few cases bereft of any post-secondary degree whatsoever. In the perhaps less glamorous areas of the software development lifecycle — QA, build and release, documentation, i18n, metrics, DBA, etc. — almost everyone I’ve ever worked with has lacked formal qualifications (which in many cases don’t exist anyway). These self-motivated male engineers run the gamut of quality, from the best of the best to the truly sub-par; but it is incontestable that there are a relatively large number of them, and that they form an important part of the technology ecosystem. But everyone seems to agree — and certainly it has been my experience — that there is no correspondingly large pool of female professional and Open Source engineers without formal training.

Why does this gender gap in non-CS-degreed software engineers exist? Could it be bridged through some sort of organized effort? Would such an effort be ethically troubling or bad for the profession in some other way? What opportunities might be lost if we continue to do nothing about the gender disparity in software engineering, or simply wait for long-term educational efforts to take root at some unspecified future date? For those who care about women in the technology industry, I would argue that these questions should be very pressing and our ignorance of their answers should be equally troubling.

(Joyce proposes an answer to these last questions in Part Two: A modest proposal.)