VCs, Founders, and the Performance Gender Gap
Female entrepreneurs have, in the past, shared their negative experiences in operating with venture capital financing. A 2015 Newsweek article described the VC industry as a “boys’ club” that creates a “particularly toxic atmosphere for women in Silicon Valley”.1 A 2018 survey of female founders reported that over half of female founders experienced bias as founders.2 Do these problems systematically impact female founders’ success? Does VCs’ behavior towards them somehow impede female founders’ success?
To further the current debate on these topics, I address two closely related questions regarding female founders’ startups’ performance in my paper. First, do female founders’ VC-financed startups perform worse than male founders’ VC-financed startups? Second, do VCs influence the performance gap by treating female- and male-led startups differently?
Do female founders’ VC-financed startups perform worse? Taking an IPO or acquisition as a measure of success, I document a wide performance gap between female- and male-led startups financed by VCs. In particular, I find that startups with one or more female founders are 20% less likely to go public or be acquired than startups with all male founders. These findings confirm that the negative experiences reported by female founders translate into worse performance for their startups, as well.
Do VCs influence this performance gap between female- and male-led startups? To answer this question, I examine the performance gap for two sets of startups: startups financed by VC syndicates with all male general partners (GPs) in the lead VC and startups financed by syndicates with female GPs in the lead VC. I hypothesize VC financing syndicates with and without female lead GPs should have differing influences on the performance gap. For instance, a female lead GP could favor female-led startups and introduce a bias towards financing more female-led startups, thereby lowering the overall quality of female-led startups and exacerbating the performance gap. Or, if information is transmitted more easily within genders than across them, the presence of a female lead GP may improve the syndicate’s choice and/or advising of female-led startups. For these and a number of other reasons, syndicates with and without female lead GPs are most likely to differ in how they influence the gap. Therefore, to test VC influence on the performance gap, I compare the gap across these two groups of VC syndicates.
I find stark differences in the performance gap across startups financed by syndicates with all male lead GPs and those financed by syndicates with female lead GPs. In particular, the performance gap triples among startups financed by VC syndicates with all male lead GPs. Among startups financed by syndicates with female lead GPs, the performance gap disappears altogether. Additionally, the difference in the gap arises from female-led startups’ improved performance when financed by syndicates with female lead GPs whereas male-led startups’ performance is the same across the two sets of syndicates. This test establishes that there is a clear difference in the performance gap between female- and male-led startups based on who is providing the capital.
The above analysis is not sufficient, however, to establish that VCs influence the gap. In particular, the above results are also consistent with the performance gap arising because higher quality female-led startups preferentially seek financing from VCs with female GPs, which has nothing to do with VCs’ actions. To identify VCs’ role in the performance gap, I have to disentangle the effect of “founder preferences” from that of VCs’ actions.
To isolate VC influence on the gap, I compare the performance gap across VC syndicates that lose female lead GPs immediately before a financing round and those that retain their female lead GPs. It generally takes 3 months for VCs to decide whether to finance a startup, which means that startups submit their financing requests at least 3 months before financing rounds are announced. If any GPs leave in those 90 days, the startup’s founders would not know about it before they request financing from the VC. Therefore, founders’ preferences over VCs should not impact performance gap differences between startups financed by syndicates where female lead GPs left before the financing round and startups financed by syndicates whose female lead GPs did not leave. This disentangles VC influence on the gap from founder preference influence.
Comparing these two sets of startups, I find a much larger performance gap among startups financed by syndicates where female lead GPs left before the financing round. Female-led startups where a female lead GP recently left are 80% less likely to succeed as female-led startups where the female lead GPs did not leave. Again, male-led startups’ performance does not change across the two groups. This test shows that female lead GPs’ departure hurts the performance of female-led startups financed by the syndicate and widens the performance gap. As founder preference cannot play a role in the performance gap difference between these two sets of startups, this test establishes that VCs’ actions impact the performance gap between female- and male-led startups.
Do VCs impact the performance gap through their choice of startups to finance or through their advising of financed startups? Based on a number of analyses presented in the paper, I conclude that VCs with female GPs are better able to choose high quality female-led startups to finance. One of these analyses compares second financing round performance gap differences to those in the initial financing round. It shows that there is no further narrowing of the performance gap in the second financing round for startups financed by syndicates with female lead GPs. As financing choice is most important in the initial round whereas advising remains important through all rounds of VC financing, this result strongly suggests that syndicates with female lead GPs are better able to choose high quality female-led startups.
My paper answers two inter-related questions. First, it shows that female-led startups operating with venture capital perform worse than their male-led counterparts. Second, and more importantly, it presents compelling evidence that VCs impact this performance gap, likely due to differing abilities in choosing which female-led startups to finance. This second finding is important because, if true, it implies that some VCs are wasting investor-provided resources and that who finances a startup matters. A “bad” financier may choose to finance too many low quality startups and not enough high quality ones.