Adapted from “Funding Black High-Growth Startups,” by Lisa D. Cook of Michigan State University’s Department of Economics and James Madison College, Matt Marx of Cornell University’s SC Johnson College of Business, and Emmanuel Yimfor of Columbia Business School.
Key Takeaways:
- Only 3.47 percent of founders seeking funding from VC firms are Black, suggesting there are systemic barriers for Black founders to get these opportunities. Additionally, Black-founded startups raised only a third of the amount raised by non-Black-founded startups.
- More than a third of this gap can be accounted for by looking at what VC investors consider when making funding decisions. Typically, these factors involve very narrow criteria, like an Ivy League education or history of running successful startups.
- However, if a Black person heads the investment team, the funding gap narrows by nearly 50 percentage points.
When the Black Lives Matter protests kicked off around the country in May 2020, Emmanuel Yimfor, an assistant professor in Columbia Business School’s Finance Division, was inspired to use his own unique expertise to contribute to the cultural moment. “There was a lot of discussion about the drivers of wealth inequality, and I've always been interested in capital formation and allocating capital to its most productive uses,” he says.
So Yimfor turned that lens on the racial wealth gap. “If we think about the allocation of human capital by race and by gender, we would expect to see that the share of entrepreneurs getting funding would match their population share,” he says. “If we see any divergence between those two things, that reflects underlying frictions in access to funding.”
Up to that point, there had been little work using large-scale data to understand barriers and access to funding by race for high-growth startups, Yimfor says. “So I wanted to be the first to shine light on this important topic.”
How the research was done: Yimfor’s first hurdle was finding a usable data set. “Ideally, you would want self-reported race information,” Yimfor says. But in the absence of that, he hoped the Consumer Financial Protection Bureau (CFPB) would be a simple, central source of this information. Unfortunately, he discovered a problem in its data collection methodology: Because not all companies self-report their demographics, CFPB uses last names and geography to triangulate a best guess about founders’ race and ethnicity — a method that was implicitly biased and leaned on stereotypes.
The impact of this faulty data set could be sweeping, so instead, Yimfor and his team collected the data themselves, largely by pulling more than 100,000 images from social media company websites. Using this method, the team ultimately ended up with a dataset that represented about 92 percent of all the founders in PitchBook, a leading resource for investment and venture capital data.
Then, to sort them by race, they let AI do the first pass, sorting the images into folders — which was about 85 percent accurate but revealed its own issues of bias. “The mistakes that the algorithm made were not random,” Yimfor says. “For example, if you're a Black woman and you happen to have long, straight hair, the algorithm is going to put you into the white folder.” So the researchers then combed through each image to confirm they were sorted correctly.
What the researcher found: Yimfor and his team found that only 3.47 percent of the founders seeking funding from VC firms were Black, suggesting there are systemic barriers for Black founders to get these opportunities. But even among those that have made it into the VC pipeline, startups founded by Black entrepreneurs raised only a third of the amount raised by non-Black-founded startups.
The next step was to ask why. “We asked ourselves, ‘How much of this gap can we explain by differences in observable characteristics?’” Yimfor says. “Black founders, for example, are less likely to have the type of signals that investors are looking for — how well you’re networked, whether or not you have a patent, if you’ve founded a startup before or attended one of the selective Ivy League schools.” By isolating those types of factors alone, Yimfor and his team were able to explain about 35 percent of the funding gap.
Why it matters: Yimfor’s findings suggest that there are immediate actions that VC firms can take to help close this gap. His team investigated investments made immediately after the George Floyd protests in 2020 — a moment that presented unique pressure to invest in Black founders — and assessed the decisions of Black and non-Black venture investors. The team’s review found that if a Black person heads the investment team, the funding gap narrows by nearly 50 percentage points.
“We found that Black partners disproportionately fund Black founders, and they rely less on the observable markers that we identified, like Ivy League schools or patents,” Yimfor says. Black investors are more likely to have overlapping networks and backgrounds with Black founders, which can help contextualize a founder’s potential for success. As a result, the research found, Black investors are more likely to identify successful Black founders than their non-Black counterparts.
To help close that gap, Yimfor’s research suggests that VC firms invite Black investors into the room or, when that’s not possible, interrogate the rubric that white investors are using to assess Black founders. “Think about the model that you're using to map founder characteristics to the likelihood of success,” he says, like recognizing competitive founders from historically Black colleges and universities rather than Ivy League schools. “Those things might be more context-dependent than you realize.”
