قالب وردپرس درنا توس
Home / US Business / Is an algorithm less racist than a lender?

Is an algorithm less racist than a lender?



Trevor McIntosh, 35, and Brennan Johnson, 31, secured a mortgage on their Wheat Ridge, Colo., Home through Better.com in 2018. "We are both millennials and we must immediately go online for anything," said Mr. Johnson, and data analyst. "It seemed more modern and progressive, especially with the technology behind it."

Previously, the couple had negative home buying experiences. A homeowner, they said, directly refused to sell to them. A lender also put down a bunch of surprise fees just before closing. The couple were not sure if prejudice – unconscious or otherwise – was to blame, but they could not rule it out. "Trevor and I have experienced discrimination in a variety of forms in the past, and it becomes ingrained in the psyche when communicating with any institution," Johnson said. "So starting with digital, it seemed like fewer obstacles, at least the ones we were afraid of, like human bias." (Better.com introduced me to Mrs. Anderson, Mr. McIntosh, and Mr. Johnson, and I interviewed them independently.)

Digital lenders say they assess risk by using the same economic criteria as traditional banks: lender income, assets, credit points, debt, debt, cash reserves and the like. These guidelines were issued by the Consumer Finance Protection Bureau following the recent recession to protect consumers from predatory lending or risky products.

These lenders could theoretically use several variables to assess whether borrowers can repay a loan, such as rent or use history, or even assets owned by the extended family. But in general they do not. To finance the loans, they rely on the secondary mortgage market, which includes the authorities Freddie Mac and Fannie Mae, and which became more conservative after the 2008 crash. With some exceptions, if you do not meet the standard C.F.P.B. criteria, you will probably be considered a risk.

Fair real estate lawyers say it is a problem, because standard financial information puts minorities at a disadvantage. Take credit points – a number between 300 and 850 that assesses how likely a person is to repay a loan on time. Credit points are calculated based on a person's expenses and payment habits. But landlords often do not report rental payments to credit bureaus, even though these are the largest payments that millions of people pay regularly, including more than half of black Americans.

For mortgages, most banks rely on the credit score model invented by Fair Isaac Corporation, or FICO. Newer FICO models may include payment history for rent, but the secondary mortgage market does not require them. Nor does the Federal Housing Administration, which specializes in loans to borrowers with low and moderate incomes. In addition, systemic inequality has created significant wage differences between black and white Americans.

"We know that the wealth gap between white households and colorful households is incredibly large," said Alanna McCargo, vice president of housing finance policy at the Urban Institute. “If you look at income, assets and credit – your three drivers – you exclude millions of potential blacks, Latinos and, in some cases, Asian minorities and immigrants from accessing credit through your system. You perpetuate the wealth gap.


Source link