City AM recently published an article about the gender gap in venture capital. They cover a new study from the Vienna University of Business which claims that female entrepreneurs are held back by a false “prototype” of what a real start-up founder looks like. The article then mentions the research that we did with Barclays, where 72% of high growth female founders agreed that it would have been easier to raise cash if they were men and 59% said they had been discriminated against because of their gender.
The Vienna study identified 107 blockchain-related start ups from Crunchbase and found that 15 of the teams were all-male, 92 were mixed gender, and none were all-female. They controlled for team size, education level, industry experience, managerial experience and start-up experience and found that all-male teams were less successful than the teams with some women on them. And majority female-founded teams were again less successful than the mixed teams.
This is similar to what we have found in our own research. Female founders are just as likely to receive follow-on funding and female-founded companies have the same return on investment as male-founded companies, but female founders struggle more when raising their first rounds of equity investment.
However, the press release of this study claims much more than this . According to the study’s author “while female founders need to be as different as possible in order to stand out from the competition, the study suggests that being female already deviates too much from the normative standard. As a result, female founders are not able to prove themselves in the first place because they are simply denied the chance, or the investors’ funding, to do so, regardless of their education or experience.”
Their evidence does not support this broad claim. For one, blockchain businesses are not going to be representative of startups in general. A lot of the interest in blockchain has come out of online sub-cultures interested in libertarianism and crypto-currency. These cultures skew very heavily male, and are often accused of being quite misogynistic. Data gathered based on blockchain cannot be generalised to represent the culture in, say, healthtech or green technology.
The hypothesis being tested by the study is that the more deviation from the “prototype” of a typical entrepreneur (who is assumed to be male), the less investors will support the company. They then tested this hypothesis with the method described above, and were surprised to find that companies with some female founders outperformed the all-male companies. Despite finding evidence that contradicted the initial hypothesis, the rest of the study and the press output continues with the same argument as before. With little, if any, alteration based on their results.
From my own research, their hypothesis that women find it harder to be entrepreneurs because people are not used to female entrepreneurs seems plausible but their method of investigating this question does not give much evidence either way.
This represents a broader trend in academia (and, to be fair, in think tanks) where the pressure to publish eye-catching research means twisting the results of a study into something more sensational. Very few people will read the paper, so where’s the harm, they think, especially if what you’re saying may be broadly true? But it comes at a very real cost. We need good quality research about the tough problems that we face collectively, and this means being clear with your audience about what you found.
I would have been perfectly happy to see a news article saying “Researchers find that blockchain firms with a female founder are more successful than all-male blockchain companies.”