Skip to main content

Kennedy School Review

Topic / Science, Technology and Data

When Glass Ceilings Meet Glass Walls


Megan Smith grew up around engineers. Her grandfather helped build roads and bridges throughout the early 20th century and contributed to large portions of the transportation system in Indiana. Yet, when the young Smith enrolled at MIT in the early 1980s to study mechanical engineering, the question her grand- father asked her mother was, “Well, why would she want to do that?”1

Smith, now the vice president of Google[x], the arm of Google responsible for innovations like the driverless car, Google Glass, and Project Loon, shares the vignette to add color to one of the many unconscious biases individuals bring into the tech ecosystem. “We inherit the weight of history, of inherent bias. First, there was conscious bias, now there is unconscious bias. Just things your brain fills in . . . It’s no one’s fault—it’s not like anyone is actively doing this. It’s just that we inherit it, we have it, it’s systemic. So we need to become highly educated. . . . Once you know that, then you can act differently.”2

In Silicon Valley, gender disparities permeate every tier of participation, from entry-level programmers to executive-level positions, from initial pitch competitors to early-stage investors, from hackathon participants to conference speakers. While many claim it’s a pipeline problem— women earned only 18 percent of the computer science (CS) degrees in the United States in 20083—further analysis finds a lengthy negative feedback loop compounding these inequalities throughout the entire cycle. For example, just a year before Smith graduated from MIT in 1986, 37 percent of computer science graduates in the United States were female; by 2008, this number had dropped to 18 percent. In 1991, women held 37 percent of all computing-related occupations; by 2012, this number was 26 percent.4 What is happening in the tech ecosystem, what is driving it, and what can we do to collectively address it?


Women are woefully underrepresented within the technical side of the sector, comprising only:

  • 2 percent of open source developers5
  • 6 percent of mobile application developers6
  • 23 percent of programmers7

These percentages extend to the founding and investing of companies as well, where women comprise just:

  • 5 percent of technology company founders8
  • 11 percent of venture capitalists9
  • 22 percent of angel investors10

Why are these numbers so low, and where do they begin?

According to the National Center for Women & Information Technology (NCWIT), a collective of more than 250 affiliates aimed at increasing the number of girls and underrepresented groups in computer sciences, the average girl gains access to computers at age 14.5 while the average boy does so at age 12.11 Furthermore, girls are more likely to first own a computer at age 19, much later than the average of 15 for boys.12 However, it is more than access and ownership that is affecting girls’ interest in computer science; it is also social constraints.

Young girls lack prominent technical role models in everything from children’s books to toys to media. A study from the Journal of Educational Computing Research found that by high school, boys have “much more sex-stereotypical attitudes to computing than girls.” This, compounded with more experience in computing, means that, “girls not only have to deal with being a minority in computing classes [in high school], but they also face a majority of students who believe men do computing better than women.”13 The study further asserts that when “women are encouraged by role models or a mentor, they . . . are more motivated to study computing.” It appears that female mentors and positive imagery, along with encouragement from parents and teachers, are essential for girls to continue studying computer science. Yet, prominent female computer scientists are difficult to find in the American mainstream media. “Technical women remain largely invisible and behind the scenes despite important and often elite contributions,” says Smith.15 Programming heroines like Ada Lovelace, Grace Hopper, Anita Borg, and the ENIAC Programmers remain largely unknown outside the tech community.16

However, images of women in media are beginning to change. Non-profits like Miss Representation and the Geena Davis Institute on Gender in Media aim to improve the portrayal of women and girls. In addition, a recent partnership between Getty Images and, the foundation established by Facebook CFO Sheryl Sandberg, aims to empower women through increased imagery in traditionally male-dominated professions.17 “When we see images of women and girls and men, they often fall into the stereotypes that we’re trying to overcome, and you can’t be what you can’t see,” says Sandberg.18 Yet, even with increased positive imagery, current norms leave many young girls and boys unable to list even a single female computer scientist.


For the young women who defy these stereotypes and go on to pursue computer science and engineering degrees, the classroom presents another challenge. In the United States, only 0.3 percent of females list computer science as their intended major upon entering college.19 When beginning their first computer science course, young girls find themselves outnumbered and report feeling the need to prove themselves or endure sexist remarks from male classmates or professors.20 They describe struggling silently alongside male peers who, often enjoying more support from their families and society, have spent summers attending coding camps or hours working on side projects.21 Yet, when universities revamp their curriculum, positive trends emerge. In 2006, Harvey Mudd College, an elite math, science, and engineering university, conducted extensive research to determine how to attract more female CS students. Its findings encouraged the university to adopt a three-prong approach that overhauled its CS curricula.22 The result? By 2010, female representation in computer science went from 10 percent of the class to almost 50 percent.23 While Harvey Mudd College may naturally attract more young women interested in science, engineering, and technology (SET), the basic reform strategies of tailoring incoming CS courses to student backgrounds, increasing collaborative projects, and updating its programming curricula should be noted by other universities across the country.

Increasing the number of women and girls that pursue degrees in SET is not the only challenge, however. Once in the profession, women leave these fields at almost twice the rate of men—over 52 percent leave the profession—with the attrition rate spiking ten years into their career.24 The 2008 study “The Athena Factor: Reversing the Brain Drain in Science, Engineering, and Technology” found the top five reasons women left their professions were due to “unconscious bias, isolation, supervisory relationships, [unclear] promotion processes, and competing life responsibilities.”25 Women reported feeling isolated as the only female and often “stuck” in positions with unclear promotional paths. One reason for this is the lack of female mentors and (often male) sponsors. While a mentor can provide guidance and feedback, a sponsor is usually a senior person within the organization who will advocate for the individual and potentially risk his or her reputation on the person’s performance. At least 45 percent of the women in the study lacked formal mentors and more than 80 percent lacked sponsors.26 However, the masculine culture itself often prevents male-female sponsorship, especially amongst younger women. Recent findings from “The Sponsor Effect: Breaking Through the Last Glass Ceiling” finds “the majority of senior men (64 percent) at the level of vice president and above are reluctant to have a one-on-one meeting with junior women—and half of junior women likewise avoid seeking out such contact.”27 Overall, the perception of an illicit romance is enough to keep both parties distant, which also means that a man is 46 percent more likely to have a sponsor than a woman.28


While issues of representation, unconscious bias, and engrained culture might perpetuate the glass ceiling, several other constraints appear to be impacting women’s technological trajectory. According to the Kauffman Foundation, many high-growth firms—the ones creating the most jobs and generating the highest revenues—are startups in science and technology. While women comprise 35 percent of startup founders, they account for only about 5 percent of technology startups.29 Why is this?

Most research agrees that, at least in the sciences, the pipeline is not the issue. According to the report “Overcoming the Gender Gap: Women Entrepreneurs As Economic Drivers,” “[Women] have earned about 60 percent of the bachelor’s degrees in the biological sciences, 41 percent of those in the physical sciences, and 35 percent in chemical engineering.”30 One argument, according to the Kauffman Foundation, is that, “What it takes to succeed in business is not necessarily the same as what it takes to succeed in starting a business. While women have made great strides in breaking through the proverbial ‘glass ceiling’ to advance to high rank within corporations, few have made similar strides in breaking out laterally—through what might be called the ‘glass walls’—to start their own high-growth firms.”31 This could be the desire to take academic leave from leading institutions to found high-growth companies or leaving high-profile positions within a firm “to start something big.”32

Furthermore, a breakthrough study entitled “Gender Differences in Patenting in the Academic Life Sciences” followed more than 4,000 life science researchers for a thirty-year period and found several prominent trends. First, women patent their research at about 40 percent the rate of men, which presents “a very significant narrowing of the field at one of the first major steps along the road to creating a startup company from one’s research.”33 Second, of the women in the study, only 6.5 percent were on the science advisory board of a high-tech company, compared to more than 93 percent of the men.34 What this implies is that “women were less likely to have the connections—or make the connections—that can help scientists recognize the commercial potential of their research in the first place and then help them to commercialize it effectively.”35 It found that men had networks outside traditional academic channels and were also more likely to use these networks informally to discuss potential ideas and opportunities.

However, the importance of a network goes beyond the advisory board. Large and diverse networks play a critical role in being able to capitalize on opportunities, secure initial funding, and accelerate professionally. Often these networks are honed through boardroom meetings, speaking engagements, and professional sponsors. Yet, women comprise only 16.9 percent of board memberships in the Fortune 500,36 19 percent of CIOs in the Fortune 250,37 and an estimated 10 percent to 14 percent of speaker positions at tech conferences.38 Roughly 49 percent of all publicly traded information technology companies do not have a single woman on their board, compared to 36 percent of the largest public companies in the country.39 As board members are usually former CEOs of other tech companies or venture capitalists, this pipeline proves challenging.40


Women are also less likely to secure funding compared to their male peers. According to the Center for Venture Research, in the first six months of 2013, only 24 percent of female-led businesses received angel investment and a mere 7 percent received venture capital.41 Investors often invest by “pattern matching” or by looking for startups that have historically proved successful. In Silicon Valley, this often plays out with investors picking entrepreneurs from a traditional cohort where “founders had held a senior position at a big technology firm, worked at a well-connected smaller one, started a successful company already, or attended one of just three universities—Stanford, Harvard and Massachusetts Institute of Technology.”42 In fact, Reuters recently analyzed “Series A” funding given by the top five Silicon Valley venture firms and found that almost 80 percent fit the above description.43

Sharon Vosmek, CEO of Astia, a nonprofit focused on getting women into high-growth ventures says, “VCs [venture capitalists] hold clear stereo- types of successful CEOs. [T]hey call it pattern recognition, but in other industries they call it profiling or stereotyping.”44 Yet, pattern matching does more than promote the traditional cohort. “The consequence of starting a pattern of white males-only rooms is that patterns tend to repeat themselves—after patterns and norms are well-established, it would take a lot of imagination for some of the boardrooms and lunchrooms in Silicon Valley to try to look or feel drastically different,” observes tech reporter Elise Hu.45

One extreme of this is the “brogramming” culture that is beginning to rise in Silicon Valley. Derived from a mashup of the fraternity greeting “bro” and the traditional “programmer,” this culture attracts a more “testosterone-fueled breed of coder,”46 or one that parties like Sean Parker’s character in the movie The Social Network. This culture increasingly alienates women. “Anything that encourages the perception of tech as being male-dominated is likely to contribute to this decline,” says Sara Chipps, who cofounded Girl Develop It, an organization offering software development programs to women.47

Additionally, investors are likely to miss opportunities based on personal and institutional biases. Debbie Sterling is the founder of GoldieBlox, a toy company established to tackle gender disparities within engineering. She remembers, “My very first investor pitch was a perfect example of this. I presented to two men and one woman. The woman totally got it, and the men didn’t get it at all. They just couldn’t relate to my idea in the way that the woman could.”48 Kathryn Minshew, founder of a site dedicated to professional women called The Muse, had a similar experience. “I went into one meeting with a venture partner and he said, ‘yeah, I pulled up the site yesterday, but honestly, it just isn’t compelling to me at all.’ . . . [I would like to have replied], ‘It doesn’t matter that it’s not compelling to you. It’s not built for you.’”49

This reality is reflected in data as well, in that, “firms with at least one woman investment partner are 70 percent more likely to lead investments in a woman entrepreneur than those with only males.”50 In general, less diversity in founders means both fewer ideas coming from women and underrepresented groups as well as fewer channels for these groups to become angel investors or venture capitalists themselves, thereby perpetuating the cycle. It appears that although Silicon Valley prides itself on its ability to laud great ideas and hard work, many have begun to call out the systemic biases for what they really are: the Myth of Meritocracy. After all, with all this glass, it’s almost as if women are the proverbial mimes— stuck in a box.


Luckily, one thing that continues to influence Silicon Valley is numbers, and recent data suggests promising financial and social trends for women:

  • Compared to peers, companies with women in upper management positions reported 35 percent higher return on equity and 34 percent better overall return to shareholders.51
  • Women-led startups are shown to have higher revenue returns, around 12 percent, and they use one-third less capital than their male peers.52
  • The odds of a venture-backed company’s success increase with more female executives at director and vice president levels.53
  • Venture-backed startups that went public, were acquired,or turned profitable were found to have twice as many women in senior roles than their peers, at 7.1 and 3.1 percent respectively.54

While these financial trends provide cues to venture capitalists and angel investors to seek out women-led startups, they are also being met by positive social trends. Vosmek, in her ten years at Astia, senses the cultural shift as well. “I am excited to see that the conversation has changed dramatically from when I came to Astia [in 2004],” Vosmek affirms. “I find that everyone is ask- ing themselves why they don’t see more women in their deal flow [and], at Astia, we are now leading to the next level in the conversation—one that focuses the attention on what we see the opportunity for change to be—rethinking and redesigning the ecosystem. Time to stop fixing the women and start fixing the innovation ecosystem.”55 Vosmek encourages intentional changes that can be measured and replicated throughout the entire ecosystem.

One example of a recent disruption to this ecosystem was the departure of two prominent female Silicon Valley venture capitalists in early 2014. Theresia Gouw and Jennifer Fonstad left their respective firms to found Aspect Ventures, an investment firm specifically targeting mobile startups and diverse teams.56 “There’s tons of data showing that diversity on boards, both public and private, and in management teams, leads to better financial returns,” highlights Gouw.57 “If you’re a consumer-facing service, half of your audience is going to be male and half of your audience is going to be female.”58 The creation of Aspect Ventures sends a signal to others in the space: diversity is an asset.

Another disruption to this ecosystem is crowd-funding platforms like Indiegogo and Kickstarter. Danae Ringlemann, the founder of Indiegogo, says the company’s goal is to “democratize finance” by giving the power to fund back to the people.59 One way they do this is by utilizing social networks, an area where women display a clear advantage over men. Women spend roughly 30 percent more time on Facebook, drive 62 percent of the sharing, and have a network that is 8 percent larger than men.60 These networks may prove helpful for crowd-funding. “Communities are really important—almost 30 percent of the founder’s overall fundraising goal will come directly from the individual’s personal network,” says Breanna DiGiammarino, director of causes for Indiegogo.61 By removing traditional financial barriers, women increase their funding options. The data reinforces this: 47 percent of fully funded Indiegogo campaigns are run by women, compared to only 13 percent of venture-backed companies.62 Furthermore, for large-scale and successful campaigns, “Angels and VCs can look to crowd-funding as proof of demand for concepts, and founders can show demand through the success of their campaign,” high- lights DiGiammarino.63

An additional promising social trend is the number of groups dedicated to women and girls in technology. Dozens of groups around the country like Girls Who Code, Women 2.0, and Black Girls Code aim at increasing coders, building networks, and providing sponsors and mentors. Angie Chang, cofounder of Women 2.0, recalls, “One of the reasons we started Women 2.0 in 2006 was, almost selfishly, to have more women in our networks. Of course, we also wanted to highlight women as creators and allow this conversation to become more mainstream.”64 Women 2.0 is now in seven countries and twenty cities, offering monthly online and offline events for female founders to network, learn, and connect. As these groups work to increase the number of coders and networks, dozens of accelerators’ programs like Women Innovate Mobile, NewMe, and Astia aim to increase the number of female entrepreneurs.

For women who are already established in the tech scene, advocating internally proves fruitful as well. After identifying two main constraints around women in tech—unconscious bias and visibility—Megan Smith began to institute a training for 46,000 Googlers in unconscious biases and began to highlight prominent women in the sector through Women Techmakers.65,66 Sheryl Sandberg, in writing Lean In, sparked an extensive and necessary dialogue around women in the workplace, leading to everything from “Lean In Circles” at the office to nuanced cultural critiques around childcare and patriarchy.

This cultural element may be more deeply rooted than we want to admit. Parents and society encourage young boys to pursue computer science, whereas research finds that girls view it as “boring” or they “don’t think they’d be good at it.”67 Chang, a longtime resident of San Francisco, says, “Even in Silicon Valley, parents are still only enrolling their boys in extracurricular coding programs.”68 Most recently, in 2011 the affluent school district of Los Altos outside of Silicon Valley began a program to teach over five hundred sixth-graders how to code. At that age, the feedback from students showed girls had the same level of interest as boys and that “special girls-only programs [were] unnecessary. . . because the stereotypes may not have yet set in.”69 It appears that if we really want to tackle the gender divide, it may start with us.


The discussion around “women in tech” is ongoing; however, several prominent themes and recommendations emerge. While this list is not exhaustive, many would agree that in order to increase the representation of women in technology, we must do the following:

  • Recognize prominent female figures in technology at all stages, from historical figures to entrepreneurs to investors
  • Introduce coding at an early age through school programs and other educational opportunities
  • Encourage universities to update their CS curricula to increase and support the number of young women interested in pursuing the degree
  • Expand programs aimed at increasing professional networks, female conference speakers, and board memberships
  • Address unconscious biases and identify institutionalized norms that prevent diversity from naturally emerging, including male-female sponsorships
  • Increase the number of female venture capitalists and angel investors continue; and,
  • Join the dialogue around gender norms to create an environment open to ongoing analysis and critique.
While the three decades of trends around women in technology might appear to display a negative feedback loop, the beauty of this engineering term is that it can be, by definition, self-regulating. In fact, just the right amount of intervention can often stabilize the circuit.
Melissa Sandgren is a 2013 graduate of the John F. Kennedy School of Government at Harvard University. She previously worked for UN Women, Congresswoman Nancy Pelosi, and the World Health Organization. Sandgren currently lives in San Francisco, where she researches how tech can disrupt social norms and promote gender equality.


1 Malter, Jordan. “Why Google Wants Women.” CNNMoney video.
2 Ibid.
3 National Center for Women & Information Technology. By the Numbers. National Center for Women & Information Technology, 28 February, 2014.
4 Ashcraft, Catherine, and Sarah Blithe. Women in IT: The Facts. National Center for Women & Information Technology, 2009.
5 Holliger, Andrea. The Culture of Open Source Computing. National Center for Women & Information Technology, 1 November 2007.
6 Cravens, Amy. “A Demographic and Business Model Analysis of Today’s App Developer.” GigaOM Pro, September 2012.
7  Ashcraft and Blithe, Women in IT.
8  Tam, Beverly. “Why Women Continue to Lag Behind Men in the Startup Community—And What We Can Do About It.” Forbes, 9 April 2012.
9 Gage, Deborah. “Battery Ventures Names First Female General Partner.” Wall Street Journal, 18 February 2014.
10 Stengel, Geri. “How Women An- gels And Entrepreneurs Are Beating Investment Odds.” Forbes, 22 May 2013.
11 Women Who Tech Web site. Resources.
12 Ibid.
13 Dryburgh, Heather. “Underrepresentation of Girls and Women in Computer Science: Classification of 1990s Research.” Journal of Educational Computing Research 23(2): 181-202, 2000.
14 Ibid
15 Smith, Megan. “’Passion, Adventure and Heroic Engineering’ . . . and Talent Inclusion.” Huffington Post, 11 October 2013.
16 Ada Lovelace was an English mathematician and is recognized for creating the first algorithm run by
a machine; Grace Hopper was an American computer scientist and one of the first programmers at Harvard University; Anita Borg was an American computer scientist who founded the Institute for Women and Technology; and the ENIAC Programmers were a group of six women who were programmers for the first electronic digital computer calculating ballistic missile projections for World War II.
17 Miller, Claire Cain. “ and Getty Aim to Change Women’s Portrayal in Stock Photos.” New York Times, 9 February 2014.
18 Ibid.
19 Miller, Claire Cain. “Opening a Gateway for Girls to Enter the Computer Field.” New York Times, 2 April 2013.
20 Pollack, Eileen. “Why Are There Still So Few Women in Science?” New York Times Magazine, 3 October 2013.
21 Kaufman, Wendy. “Addressing the Shortage of Women in Silicon Valley.” NPR, 11 November 2011.
22 Cohoon, J. McGrath. Harvey Mudd College’s Successful Systemic Approach (Case Study 2): Attracing Students Through an Engaging Introductory Computing Curriculum. National Center for Women & Information Technology, 2010.
23 Barker, Lecia, and J. McGrath Cohoon. How Does Engaging Curriculum Attract Students to Computing? (with Case Study 2). National Center for Women & Information Technology, 2010.
24 Hewlett, Sylvia Ann et al. The Athena Factor: Reversing the Brain Drain in Science, Engineering, and Technology. Harvard Business Re- view Research Report, June 2008.
25 Ibid.
26 Ibid.
27 Hewlett, Sylvia Ann et al. The Sponsor Effect: Breaking Through the Last Glass Ceiling. Harvard Business Review Research Report, December 2010.
28 Ibid.
29 Tam, “Why Women Continue to Lag Behind Men.”
30 Mitchell, Lesa. “Overcoming the Gender Gap: Women Entrepreneurs as Economic Drivers.” Ewing Marion Kauffman Foundation, September 2011.
31 Ibid.
32 Ibid.
33 Ibid.
34 Ibid.
35 Ibid.
36 Catalyst. 2013 Catalyst Census: Fortune 500 Women Board Directors. Catalyst, 10 December 2013.
37 Demby, Lee. “F-250 Female Chief Information Officers Approaching 20%.” Boardroom Insiders, 6 February 2012.
38 Geek Feminism Wiki. “Women speakers.”
39 Bajaj, Vikas. “A Striking Absence of Women.” New York Times, 12 October 2013.
40 McBride, Sarah, and Poornima Gupta. “Insight: Tech Start-Ups Show Little Imagination on Board Gender Diversity.” Reuters, 6 December 2013.
41 Slade, Hollie. “Why Is It So Hard for Female Entrepreneurs to Get VC Funding? Could Crowdfunding Be the Answer?” Forbes, 29 November 2013.
42 McBride, Sarah. “Insight: In Silicon Valley Start-Up World, Pedigree Counts.” Reuters, 12 September 2013.
43 Ibid.
44 Contextly. “Silicon Valley Isn’t a Meritocracy. And It’s Dangerous to Hero-Worship Entrepreneurs.” Wired, 25 November 2013.
45 Hu, Elise. “How the Meritocra- cy Myth Affects Women in Technology.” NPR, 7 February 2014.
46 MacMillan, Douglas. “The Rise of the ‘Brogrammer.’” Bloomberg Businessweek, 1 March 2012.
47 Ibid.
48 Sterling, Debbie. “Debbie Sterling: It’s About Finding the Right Fit.” Wall Street Journal, 11 October 2013.
49 Davis, Noah. “Money Matters: Why Women Founders Struggle in Silicon Valley.” The Verge, 6 March 2013.
50 Padnos, Cindy. High Performance Entrepreneurs: Women in High-Tech. Illuminate Ventures, February 2010.
51 Ibid.
52 Ibid.
53 Canning, Jessica, Maryam Haque, and Yimeng Wang. “Women at the Wheel: Do Female Executives Drive Start-Up Success?” Dow Jones & Company, Inc., 2012.
54 Ibid.
55 Vosmek, Sharon. Interview by author via e-mail, 9 March 2014.
56 Miller, Claire Cain. “Two of Venture Capital’s Senior Women Start a New Firm.” New York Times, 5 February 2014.
57 Gustin, Sam. “Powerful Female Venture Capitalists Aim to Broaden Silicon Valleys View.” Time, 24 Feb- ruary 2014.
58 Ibid.
59 Farr, Christina. “Indiegogo Founder Danae Ringelmann: ‘We Will Never Lose Sight of Our Vision to Democratize Finance.’” Venture-Beat, 21 February 2014.
60 Goudreau, Jenna. “What Men And Women Are Doing on Facebook.” Forbes, 26 April 2010.
61 DiGiammarino, Breanna. Phone interview by author, 14 March 2006. 62 Thirteen percent of companies have one female on their executive team. DiGiammarino, Breanna. Phone interview by author, 14 March 2006.
63 Ibid.
64 Chang, Angie. Phone interview by author, 11 February 2014.
65 Malter, “Why Google Wants Women.”
66 Lublin, Joann S. “Bringing Hidden Biases into the Light.” Wall Street Journal, 9 January 2014.
67 Taylor, Colleen. “How Harvey Mudd Transformed Its Com- puter Science Program—And Nearly Closed Its Gender Gap.” TechCrunch, 10 October 2013.
68 Chang, Angie. Phone interview by author, 11 February 2014.
69 Vaidyanathan, Sheena. “Should Kids Learn to Code in Grade School?” MindShift, 26 September 2012.