9:58am

Tue June 25, 2013
All Tech Considered

Closing The Tech Industry's Gender Gap Requires Better Data

Editor's Note: As part of our reboot of All Tech Considered, we'll invite contributors to blog about big-picture questions facing tech and society. One theme we're exploring is the lack of women and people of color in tech — a gap so glaring that ridiculously long lines at tech conferences have inspired photo essays and Twitter feeds.

The tech industry faces a dangerous problem — a lack of gender and ethnic diversity that, I think, will lead to increased economic inequality and fewer innovative ideas originating from the Bay Area, the region that produced Google, Tesla, Facebook and more.

The statistics are pretty depressing. Only 3 percent of venture-backed companies were led by all-female teams, while 89 percent were all male. Fewer than 1 percent of those founding teams were led by black founders.

Per capita income among blacks in Silicon Valley dropped by 18 percent between 2009 and 2011, while in the rest of the country income for blacks dropped by 4 percent. Income for whites and Asians went up in Silicon Valley over that same period even though it dropped in the rest of the country. (These numbers are from the Silicon Valley Index published by Joint Venture Silicon Valley and The Silicon Valley Community Foundation.)

According to the Measure of America's California report, women make 49 cents for every dollar a man makes in Silicon Valley. That figure is 77 cents on the dollar for the rest of the country.

For all the attention these statistics got, there's one point I didn't have a chance to make when I presented them at the Personal Democracy Forum last month: These data, especially the data about founder race and gender, aren't very good.

The figures about funding come from a revealing and important report by CB Insights that was published in 2010, but the firm hasn't updated the report since. The methodology it uses for deriving founder race and gender is based on an algorithm that uses census data to predict whether a name belongs to a man or a woman, an African-American or an Asian person. This works pretty well determining gender, but you can imagine how difficult it is to determine whether someone is black based on a last name.

Pete Warden, a developer/entrepreneur and friend of mine, was curious whether we could update the data using the CrunchBase API (CrunchBase is a site that aggregates information about startups and their funders). He used an algorithm similar to the one used by CB Insights that checks against the census data and found that 7 percent of Silicon Valley founders receiving funding in 2012 were African-American. Except when he spot-checked the top 10 likeliest funders to be black based on their names, none of them were. (From inspecting the data, the Hispanic estimate was much more accurate but still hovered in the low single digits, despite that community's making up 17 percent of the population.) More on Pete's methodology can be found here.

Unfortunately, with the data we have, there's just no good way to understand the extent of the problem. We would need more robust data collection methods — surveys, for example — which are expensive. And without good data to understand the scope of the problem we can't have real conversations about why the problem is as bad as it is and what we can do to fix it.

For example, the CB Insights Data tell us that despite all-female teams comprising only 3 percent of funded startups in Silicon Valley, they make up fully 31 percent of teams who get funding in Massachusetts. Some have called that number into question, but in any event, the number of female founders getting funded in Massachusetts is significantly higher than in California. Why is that? Is it an anomaly in the data? Did something happen in Massachusetts in the first half of 2010 that would lead to such an abundance of female-led companies? Or is there truly something systemic at work in Massachusetts that leads to a more gender-diverse industry? We can't answer these questions without investing in better data collection and analysis.

I have deep hopefulness about the future of the tech industry, in part because of the conversations people are starting to have about these issues and the work that organizations like Code2040, Black Girls Code and Girls Who Code are doing to create opportunities in tech for women and minorities. We should be enabling those conversations, and supporting the organizations that are seeking solutions, by doing the hard work to understand the root of the problem. It's in the best interest of the companies growing their businesses in this environment and it's in the interest of all of us who care about a 21st century economy built by Internet technology companies.

Catherine Bracy writes and speaks about the tech industry. During the day she runs Code for All, Code for America's international program.

Copyright 2013 NPR. To see more, visit http://www.npr.org/.