Black small business owners had more difficulty getting loans from banks, new data suggests
The report’s findings are being made amid increasing scrutiny of algorithmic systems and how they can inadvertently perpetuate biases.
It was evident that Black entrepreneurs and Black business owners had more difficulty finding a lender than white borrowers since the Paycheck Protection Program's inception last year. A new study has shown that this problem is more prevalent at smaller banks. Human bias may be the primary reason.
According to an economic paper, the majority of Black borrowers who were eligible for assistance under the $800 billion relief program received their loans from financial technology companies and not banks. Black borrowers were more inclined to use these so-called fintechs than any other race.
"I was stunned by the striking disparity -- It was a surprising, unexpected fact and we wanted to find out why," stated Sabrina T. Howell, assistant professor of finance at New York University Stern School of Business. She was also the lead author of this paper.
Researchers found that automated loan processing and vetting systems, such as those used by fintechs and some of the country's largest banks, greatly improved approval rates for Black borrowers. These gaps were not found for other racial groups, such as Asian and Hispanic applicants.
These findings are being made amid increasing scrutiny of algorithmic systems and how they can inadvertently perpetuate biases. The Consumer Financial Protection Bureau is examining whether such systems are in violation of fair-lending laws, even if they are accidentally.
Dr. Howell stated that her research showed how technology can also be used to level the playing field.
She said that the human brain is more frightening than any machine-learning algorithm. You can make an algorithm conform to fair lending standards, and ensure that the data it uses isn't biased. Although it may seem difficult, this is a possible objective outcome. You can't do that if you don't have a human loan officer to make a decision for you.
Every weekday, receive daily business updates. Get the latest news on business, markets, and economy via email. It will be delivered to your email.
Independent Community Bankers of America (a trade group representing small banks) defended its members by stating that community lenders had "outperformed" the rest of the banking industry in serving veteran-owned, minority-owned and women-owned businesses.
The group was particularly critical of the steps taken by researchers to determine applicants' race. Lenders were not required to collect data about borrowers' ethnicity. Dr. Howell and her coworkers used Census Bureau data to determine the likely race of applicants. According to the banking group, these methods made research "unreliable and guessable."
Sergey Chernenko is an associate professor of finance at Purdue University’s Krannert School of Management. He was not involved with Dr. Howell's research. However, he said that the paper matches his own findings about race-based gaps for Paycheck Protection Program lending. He will present a paper at the next economic conference. It concluded that Black-owned businesses were disproportionately excluded from the relief program.