“[The world’s greatest entrepreneurs] all seem to be white, male, nerds who’ve dropped out of Harvard or Stanford and they absolutely have no social life. So when I see that pattern coming in – which was true of Google – it was very easy to decide to invest.”
John Doerr
Abstract
With data from successful founders of high-tech companies, we identify traits common to large majorities of them and any gender differences in those traits. There are few. Further, we identify criteria that might lead to gender imbalance among successful founders by comparing similarities and differences in the gender distribution of these traits among the general population and among successful founders. We find that signature traits of successful founders include: motivation by the desire to build wealth, and not by the inability to find traditional employment, nor because they developed a technology in a lab environment and wanted to see it make an impact; belief that startup success was due to prior industry or work experience, lessons learned from previous successes and failures, the company’s management team, and good fortune, not because of state or regional assistance or alumni networks; access to mentors, and little financial pressure for a steady income. None of these dominant traits appeared to be required unequally of men and women, although some traits were unequally distributed in the general population.
Introduction
Women are severely underrepresented in the high-tech world, and especially among high-tech entrepreneurs. For example, women comprised only about five percent of IT patent awardees in 2005, their highest share to that date (Ashcraft and Breitzman, 2007). This low rate of high-tech creation reflects on women’s low rate of high-tech business creation. Women comprise an estimated five or six percent of high-tech entrepreneurs (Robb and Coleman, 2009), despite being 40 percent of all the self-employed people in the U.S. professional and technical services industry in 2010 (U.S. Department of Labor).
The untapped potential of women’s creativity in the high-tech arena could be promoted if we understood the cultural stereotypes about gender, technology, and entrepreneurship that filter out people who do not fit the expected pattern or template. That template may or may not be legitimate, but it selects people with certain characteristics, and excludes others.
With data collected from 549 founders of successful high-growth, technology-based companies (hereafter referred to as “successful founders”), Cohoon, Wadhwa, and Mitchell (2010) compared the backgrounds, experiences, and motivations of men and women founders. Their results identified very few gender differences among these successful entrepreneurs. Successful founders had no significant gender differences in their levels of education, early interest in starting their own business, desire to build wealth or capitalize on a business idea, nor access to funding, and they largely agreed on the top issues and challenges facing entrepreneurs. They differed in how important they considered prior work or industry experience, although both groups considered it most important. The strong similarities and differences the data documented suggest a pattern for high-tech entrepreneurial success.
This pattern differs from what John Doerr described during his National Venture Capital Association interview in May 2008, when he characterized the most successful founders as: “white, male, nerds who’ve dropped out of Harvard or Stanford and they absolutely have no social life”. Instead, it appears that other characteristics are at play in the selection of successful high-tech entrepreneurs, characteristics that might correlate with race and gender, but could still allow for more diversity among successful founders. We identify these dominant traits in this study.
In addition to identifying dominant or signature traits, we argue here that, based on principles of probability sampling, comparing the distribution of characteristics between successful men and women founders with those in the population of potential founders can help identify the pattern or filter that selects for successful founders. We consider the traits of interest based on the following logic, which is also illustrated in Table 1.
-
When the gender distribution of a dominant successful-founder characteristic (i.e., a characteristic shared by at least 75% of successful founders) matches that characteristic’s gender distribution in the population of potential entrepreneurs (i.e., the general population), this aspect of the selection process for becoming a successful founder is likely to be random. (Random selection processes typically result in representative samples, which are what we see in this case.)
-
When the gender distribution of a dominant successful-founder characteristic differs from its distribution in the population of potential entrepreneurs, the selection process for becoming a successful founder is not likely to be random. It is only these non-random dominant traits that interest us, because they imply a selection process that discriminates on the basis of gender or one that makes use of criteria correlated with gender in the general population.
When the selection process is not random because there is a gender difference on a dominant successful-founder trait that is equal in the general population, it identifies a gendered selection criterion. Dominant traits that are stronger among women than men founders suggest requirements put more on women than on men. The selection process likely requires that women in particular have this characteristic, whereas this trait is less necessary for men becoming successful founders. Correspondingly, traits that are comparatively weaker among successful women founders suggest requirements put more on men than on women.
When the selection process is not random because there is no gender difference on a dominant successful-founder trait that differs in the general population, it identifies a gendered selection criterion. This criterion is likely correlated with gender in the general population.
Table 1. Logical Analysis of Traits That May Contribute to Unequal Gender Composition of Successful Founders
In the General Population |
Among Successful Founders |
Implications for Women’s Underrepresentation |
---|---|---|
No gender difference |
No gender difference |
The trait is not likely to make an important contribution to the gender composition of successful founders. The selection process is likely random. |
No gender difference |
Difference between men and women founders |
The trait likely contributes to the gender imbalance among successful founders. The trait is required of one gender more than the other for being a successful founder. For example, when a trait is stronger among successful women founders than among successful men founders, this may lead to underrepresentation because this trait is less frequent in the general population of women, from which successful women founders arise. The selection process is likely not random. |
Difference between men and women |
No gender difference |
The trait likely contributes to the gender imbalance among successful founders. This trait is deemed critical for successful founders, but for whatever reason, is not equally possessed by men and women. The selection process is likely not random. |
Difference between men and women |
Same gender difference as general population |
The trait is not likely to make an important contribution to the gender composition of successful founders. The selection process is likely random. |
Difference between men and women |
Different gender difference than general population |
The trait likely contributes to the gender imbalance among successful founders. As in the situation where there is no gender difference in the general population but one is observed among successful founders, here the trait is required of one gender more than the other for being a successful founder. The selection process is likely not random. |
Dominant Traits of Successful High-Tech Entrepreneurs
As described by Cohoon and Aspray (2007) in their review of literature on gender and entrepreneurship, only a very small amount of scholarly literature exists for the topic of gender and entrepreneurs in the high-tech sector. The literature that does exist considers mostly gender and entrepreneurship, but not specifically in the high-tech sector, and focuses on stereotypes, gender differences in personality traits, access to financial capital, social capital, and more recently, human capital.
Stereotypes
Traits associated with entrepreneurs in the United States are typically considered to be masculine traits. Comparing stereotyped entrepreneurial characteristics with characteristics considered masculine showed a strong correlation, but comparing them with feminine-stereotyped characteristics showed no correlation (Gupta et al, 2009). Further, that research found that both men and women who self-identified with masculine characteristics were more likely to intend entrepreneurial careers.
Similar results were obtained by Baron and colleagues (2001) when comparing stereotypically masculine traits of forcefulness, assertiveness, aggressiveness, confidence, and independence with stereotypes about entrepreneurs. They found that stereotypes about women entrepreneurs are more masculine than stereotypes about women in general. Women entrepreneurs were generally thought of as more decisive, more serious about their career, successful because of specific skills (in particular, social skills), and less feminine when compared with women managers.
In a sense, high-tech entrepreneurship is doubly masculine stereotyped. In addition to the masculine stereotypes associated with entrepreneurship, high-tech interests, activities including academic study, and ability are all stereotyped as in the masculine domain (Wajcman, 1991). So, our successful founders may be similar with respect to exhibiting traits thought of as masculine, but given the “masculine” label, men and women in the general population are likely to differ on these traits.
Personality Characteristics
Research on the actual traits of entrepreneurs has yielded less clear evidence than that on stereotypes. Comparing entrepreneurs with sales representatives showed that openness to experience, risk-taking propensity, and innovativeness were more often self-descriptions of entrepreneurs (Engle and Schmidt, 2011). Findings from other studies vary.
The literature contains contradictions regarding whether risk-taking propensity is a characteristic entrepreneurs have in common and have more than other populations. Some studies found no difference in tendency toward risk-taking between women entrepreneurs and non-entrepreneurs (Sexton and Kent, 1981, as reported in Bowen and Hisrich, 1986). Other studies found that risk-taking distinguished between entrepreneurs and non-entrepreneurs, but not between men and women entrepreneurs (Carland and Carland, 1991). Finally, an Australian study of small business owners found that men and women entrepreneurs exhibited so-called masculine traits to the same degree, except in the case of risk-taking propensity, which was higher in men entrepreneurs (Watson and Newby, 2005). Yet another study found that successful women entrepreneurs in the U.S. were less likely to take risks than their male peers (Sexton and Bowman-Upton, 1990). Thus, even with the personality trait most often believed to distinguish between entrepreneurs and non-entrepreneurs, the evidence is somewhat mixed. It appears that women entrepreneurs may have more risk-taking propensity than women in general, and perhaps less than men entrepreneurs.
Instead of personality characteristics, venture capitalists are more likely to identify experience as a key trait for entrepreneurial success. They considered human capital in the form of industry/domain/market experience and marketing experience to be more important than any personal characteristics (Black et al., 2010).
Motivation and Early Interest
Motivation is necessary for people to engage in voluntary activities, so if men and women are motivated by different values or conditions related to entrepreneurship, this difference should affect whether they become successful founders. Some researchers suggest that gender differences in value systems may affect motivation and lead to women’s underrepresentation among entrepreneurs. For example, Fagenson (1993) found only two significant differences between men and women after testing 30 variables measuring values: women more than men valued equality, while men more than women valued family security. Thus, if desires for equality or family security are common to entrepreneurs, these values could be part of the filtering process we are investigating.
Licht and Siegel (2005) report that the consensus of evidence finds entrepreneurial motivation is non-pecuniary and not due to unemployment. Instead, they summarize research findings as identifying autonomy as the leading motivation, with a secondary characteristic being underestimation of risk masquerading as risk-taking propensity.
Research often finds that teenaged girls express significantly less interest than boys in entrepreneurial careers (Wilson, Marlino, and Kickul, 2004; Kourilsky and Walsh, 1998). Nevertheless, this literature is small and generally lacks longitudinal data to compare adolescent attitudes and motivation with adult entrepreneurship. In addition, some recent studies found no gender difference in the early entrepreneurial intentions of business and economics students (Diaz-Garcia and Jimenez-Moreno, 2010).
Access to Financial Capital
Access to financial capital contributes to successful entrepreneurship. Women entrepreneurs have less access to it, or make less use of financial capital, than do men entrepreneurs, however. This gender difference may stem from women’s propensity to self-finance; the size, age, and industry of women’s businesses; women’s insufficient human or social capital; women’s low participation in the best-funded industries; and perhaps, women’s underrepresentation among funders. In any case, this gender difference has been observed among entrepreneurs, and so, seems a poor candidate for a selection criterion that filters women out.
Social Capital
Social capital is essentially one’s network of useful social connections. These connections grant access to resources, including information, held by members of one’s network. For example, social networks also can make the difference between locating and not locating funding, with success more likely when the entrepreneur’s network intersects with the networks of one or more venture capital firms.
There is a wide literature on how entrepreneurs use social networks to obtain equity funding from venture capitalists: Elfring and Hulsink (2003), Tyebjee and Bruno (1984), Bygrave (1988), Sargent and Young (1991), Freear et al. (1992), and Fiet (1996). Other research also considers the influence industry context has on the usefulness of men’s and women’s social capital, implying for our study that women’s networks in the high-tech arena would be less useful than men’s networks (Loscocco et al., 2009). Regardless, the evidence remains insufficient for drawing conclusions about the effect of social capital on women’s underrepresentation among high-tech entrepreneurs. Gender differences in men and women’s social networks are clear; it is their impact on successful founders that remains ambiguous.
Human Capital
The human capital of a business comprises the education, training, and work experience of its key workers. Educational elements of human capital have to do with both the amount and type of education, such as a liberal arts background versus a technical or business degree, and the level of education specifically focused on entrepreneurship. Work elements of human capital include experience in the same line of business, the amount of management experience the key individuals in the business have acquired, and their experience with startups.
Use of outside equity may be related to human capital. Evidence shows that graduate education increases the likelihood that women seek outside equity (Carter et al., 2003). Other elements of human capital – financial acumen, startup experience, or managerial experience – had no measurable effect on seeking outside equity in that study. Regardless, these elements could affect entrepreneurship in other ways.
The issue of educational degree is particularly important for high-tech entrepreneurship, because women are less likely than men to have degrees in disciplines such as computer science or engineering. For Canadian Fortune 500 companies classified as “professional, scientific, and technical,” women held only 15% of senior officer positions in 2010, according to Catalyst. Likewise, data from Georgia show that women are under-represented as directors (11%) and executive officers (12%) in the state’s Top 50 companies by revenue (Board of Directors Network, 2010). The same data also show that bioscience and non-technology companies are substantially more likely than technology companies to have at least one woman executive or board member (Technology Association of Georgia, 2011). Without the preparatory experiences gained in such positions, potential women entrepreneurs are not likely to become successful founders.
In sum, the scholarly literature to date suggests that signature characteristics of entrepreneurship include:
-
stereotypical masculine characteristics, perhaps including risk-taking propensity
-
early motivation to be an entrepreneur
-
access to financial capital
-
useful professional networks
-
human capital in the form of education and experience
Each of these characteristics, except perhaps access to financial capital, exhibit gender differences in the general population. Women are less likely than men to identify with stereotypically masculine traits. Women appear to be motivated by different values than men, on average, and may express less motivation than men to become entrepreneurs. It is also the case that women generally seem more risk-averse than men (Croson and Gneezy, 2004). Women’s social networks tend to be less utilitarian than men’s. Women earn degrees in science, technology, and engineering at much lower rates than men and are less likely than men to have executive experience in high-tech.
Data and Methods
The data for the study reported here were collected by Wadhwa and colleagues in 2008-2009 from 549 respondents. (Five hundred and forty-two were eligible for the current analysis.) The response rate was about 40% of the founders from randomly selected high-tech companies who were invited to participate. In this study group, eight percent of the founders were women. This representation was on the high end of estimates for women entrepreneurs in the high-tech sector.
Respondents founded companies in the following industries that we consider high-tech: aerospace and defense; audio and video equipment; computer hardware, networks, peripherals, services, and storage devices; electronic instruments and controls; scientific and technical instruments; semiconductors; software and programming; biotechnology and drugs; health care facilities; medical equipment and supplies; computer services; engineering consultants; software and programming services.
The people included in this sample were all successful entrepreneurs, 59% of whom had founded two or more companies. The data made it possible to compare similarly positioned men and women entrepreneurs as few, if any, studies of entrepreneurs have done before. The survey participants were well matched in key respects. Because of the sampling methodology, they were in the same types of industries: more than half the respondents of each gender classified themselves as working in computing or some other highly technical field. The study subjects also had founded their current companies at about the same age and at around the same time.
The primary data source for this study is a subset of an existing data set of corporate records included in the OneSource Information Services company database. To construct the sampling frame, records were extracted for companies in selected industries. Company records were then stratified by geographic region and were selected randomly. Visits to the selected companies’ web sites ensured that they were still in business and provided the names and contact information for founders. Founders (defined as very early employees, typically having joined the company before the products or business model were fully developed) were contacted by email as many as four times and were invited to complete an online survey. In some cases, email invitations were followed up with phone calls. For more information about the data and its collection, see Wadhwa et al., 2009.
Despite their relatively high representation in our sample, women still comprise only 41 of the 542 useable cases. This small number is not surprising given women’s scarcity among high-tech entrepreneurs, but it could affect the generalizability of the results reported here. For the population of 1,373 founders identified as eligible participants for this study, with women’s true representation between five and six percent of that population, the number of women eligible to participate in the study would be between 69 and 82. In order to meet the standard criterion of 95% certainty (plus or minus four points) that our results accurately represent the population, we needed responses from more women than we have in our sample – between 62 and 72 women. With the data we have, our confidence interval is between eight and nine points. This means that if there is less than a 16 or 18 point difference between the true percents of men’s and women’s responses, our analyses may fail to detect it.
Findings and Discussion
We tested a large number of respondents’ reported characteristics and experiences for gender differences. These variables included: highest degree earned, interest while a college student in becoming an entrepreneur, family educational and entrepreneurial history, motivations for starting business(es), sources of funding, importance of various factors for startup success, challenges faced, and professional networks. Most of these variables showed no statistically significant gender differences. Nevertheless, it is important to recognize that differences could exist that our methodology was unable to uncover.
Strong majorities (75% or more) of our successful entrepreneurs exhibited the following traits:
-
They were motivated by the desire to build wealth (76%), contrary to longstanding views in the literature. They were not motivated by the inability to find traditional employment (5%), nor because they developed a technology in a lab environment and wanted to see it make an impact (23%).
-
They believed the success of their startup was due to their prior industry or work experience (96%), lessons learned from their previous successes (93%) and failures (87%), company’s management team (86%), and good fortune (77%), not because of state or regional assistance (8%) or alumni networks (21%).
-
They found little challenge in either accessing mentors (83%) or in overcoming family or financial pressures to keep a traditional steady job (75%).
There were other similarities, but none that such a substantial portion of these successful founders shared.
Only one statistically significant gender difference was found among these dominant traits of successful founders: family or financial pressure to keep a traditional steady job, which women reported far less than men (12% of women were challenged by this pressure, versus 27% of men). This trait is likely to reflect a gender difference also present in the general population, so it is not likely to be part of a biased selection process.
We discuss the results in more detail below. Overall, they suggest that criteria for successful founders are evenly applied to men and women, although gender inequities in the general population may affect who becomes a successful founder.
Motivations
Gender differences in motivations are commonly invoked for explaining gender segregation across occupations. In particular, women’s concentration in low-paying, low-prestige occupations suggests that they are less motivated than men by desire for wealth. Beyond this observation and what appears to be a widespread assumption of a gender difference on pecuniary motivation, however, there appears to be little evidence for the general population. Without better evidence for a general gender difference in motivation by the desire for wealth, this defining characteristic of successful founders does not appear to filter women out. If instead, a gender difference in monetary motivation does exist in the population, then this criterion could filter women out because fewer of them than men possess this trait. The two non-motivations – unemployment and desire to see impact from one’s invention –offer little explanation for gendered filtering.
Perceived Success Factors
Research often notes gender differences in attributions for success. Men are more likely than women to attribute their success to themselves and failures to external factors, whereas, women are more likely to attribute success to luck and failure to their own shortcomings (Deaux and Farris, 1977). With these gender differences in mind, it is somewhat surprising that our successful founders exhibit no measurable gender differences in this area. It appears that successful founders must recognize the value of learning from experience, assistance from colleagues, and luck, all of which fit well with stereotyped attributions to women. In this way, it seems that successful founders attribute success in much the same way women in the general population do. Therefore, these traits would not filter women out.
If instead, these factors reported as necessary for startup success are viewed as truly necessary for success, and not as only attributions, some of them may indeed disadvantage women. From this perspective, there is a gender difference in the population, with women less likely than men to have the necessary characteristics. Women’s position in the educational system and workforce give them less opportunity than men to obtain the human capital that leads to founder success. They are less likely than men to study science, engineering, and technology, and less likely to hold high-level positions in high-tech industry. Therefore, equal application of these criteria for founder success would filter out women because of their relative lack of access to experiences deemed necessary for entrepreneurial success.
Resources and Pressures
Our findings indicate that successful founders have good access to mentors and advisors, and they have little pressure to keep a steady job. In general, women may have more need than men for mentors because standard social practice gives women less informal access than men to mentors. Given this observation of a population gender difference, but none among successful founders, it appears that access to mentors and advisors is a filtering characteristic. People can only become entrepreneurs if they have adequate access to the mentoring so important for entrepreneurial success, but fewer women than men have social networks that include people who can mentor them on entrepreneurship.
Pressure to keep a steady job is a gender difference that disadvantages men in our society, but not successful founders, according to our study. This observation suggests that it is difficult to be a successful founder if one experiences this pressure. It acts as a filter, but one that would inhibit men more than women.
In sum, the necessity of becoming an entrepreneur because one desires wealth may contribute to filtering women out, but there is insufficient evidence to determine if this trait differs by gender in the general population. Human capital offers much more likely suspects for filtering traits. Because of their position in education and the workforce, women have less opportunity to acquire the backgrounds required of successful founders. Finally, the necessity of being mentored and the likelihood that women’s networks will include people who can advise them may also act as a filter deterring women from becoming successful founders.
Implications for Gender Differences that Affect Successful Founders
We found that signature traits of successful founders include: motivation by the desire to build wealth, and not by the inability to find traditional employment, nor because they developed a technology in a lab environment and wanted to see it make an impact; belief that startup success was due to prior industry or work experience, lessons learned from previous successes and failures, the company’s management team, and good fortune, not because of state or regional assistance or alumni networks; access to mentors, and little financial pressure for a steady income.
These findings corroborate reports in the literature in some cases, but contradict them in others. The largest contradiction comes from our finding that a strong majority of successful founders were motivated by pecuniary rewards. Both men and women were motivated by this factor, in contrast to a long history of theory and research that argues entrepreneurs are motivated primarily by non-pecuniary rewards. The difference in our findings might have to do either with the high-tech startup environment and recent popular stories of massive wealth accumulation, or perhaps with varying definitions for the population of study. Our population included only successful founders, not all new entrepreneurs, as some other studies include.
The two clearest points of agreement between our findings of successful founders’ signature traits and those already in the literature are professional networks and human capital. Our study shows that successful founders themselves view their professional networks and prior experiences as playing substantial positive roles in their high-tech startup’s success.
Comparing the gender distribution of dominant traits in the population and in our sample of successful founders showed that none of the traits appeared to be required unequally of men and women, although some traits were unequally distributed in the general population. In other words, the filter that impedes women from becoming successful high-tech entrepreneurs likely works through the unequal distribution of traits in the general population, not through discriminatory application of requirements for men and women founders.
Our findings should be taken as suggestive, rather than with a high degree of confidence that they accurately reflect conditions for women founders. The small number of women in our study limits our ability to make generalizable claims. Further, the filter we seek to describe could operate through combinations of traits, rather than the single traits we investigated.
Conclusion
With data collected by Wadhwa and colleagues (2009), traits common to a large majority of founders of successful high-tech companies were identified. These traits were then examined for significant gender differences among the founders and in the general population. When the gender distribution of a trait was similar in both groups, it was not likely to be a cause of women’s underrepresentation among successful high-tech entrepreneurs. If instead, there was a difference in the gender distribution of a trait in the two groups, this trait could be filtering women out of entrepreneurship. We found that motivation by desire for wealth, importance of knowledge gained from experience, and access to social networks including mentors and advisors could all contribute to the gender imbalance among successful high-tech entrepreneurs.