“Credit scores can be really hard to understand. You see a number, but what does it mean?” he says.
Many of his Black and Brown clients have never borrowed from a bank before and might not even have credit cards. As a result, their credit files are thin. But that doesn’t necessarily mean they’re a bad bet for a loan. “There are quite a few who’ve paid their rent on time for three years — but that doesn’t show up on their credit report,” he says.
Over the past few years, lenders, regulators, advocates and researchers have been sounding an alarm that the underwriting standards that determine who gets a loan are at best inconsistent and might be downright discriminatory. That cry has grown louder in recent years after George Floyd’s murder in May 2020 and subsequent nationwide protests led many segments of society to examine racism within their ranks.
It’s undeniable that there’s a giant racial homeownership gap: The rate of homeownership among Black families is 30 percentage points lower than among White families. Some of that discrepancy is because of lower incomes among Black households, leading to their decreased ability to purchase a home. And some is because of outright discrimination among lenders, something that’s repeatedly been demonstrated by research and investigative reports.
But another significant reason for Black Americans’ lower homeownership rate is the way underwriting — the process and criteria used to determine who gets a loan — is done. It may not have been designed to intentionally discriminate against Black families, experts say, but that has been the result. And it’s affected other marginalized and low-income groups as well.
For example, traditional credit scores penalize prospective borrowers for having large student loans or a high debt-to-income ratio, both of which tend to be more common among Black and other communities of color. Similarly, underwriters look less favorably on a small down payment, which is more prevalent among non-White borrowers.
And then there are those “credit invisibles” with insufficient credit or no credit score at all, roughly 45 million Americans. According to the Urban Institute, 45 percent of Black consumers compared with 18 percent of White borrowers fall into that group. Without a credit score, most conventional lenders won’t touch them, despite the fact that they may have been making regular payments for rent, utilities and cellphones for years — something that experts say is a much better predictor of whether someone will pay their mortgage every month than credit scores.
“A credit score has almost nothing to do with the performance of people or the quality of their loans,” says George McCarthy, president and CEO of the Lincoln Institute of Land Policy, who has spent years analyzing the outcomes of mortgages made to lower-income people. Defaulting on a loan is usually the result of a larger, unavoidable event such as job loss or illness, he says.
After the 2008 housing crisis, access to credit became much tighter. In part, that’s a good thing: In the years before the meltdown, lenders often provided mortgages without determining borrowers’ ability to repay them. And once the home buyers became delinquent on their loans, they were often cut off without any assistance. “The egregious lending of the housing bubble was combined with a toxic combination of bad due diligence and terrible mortgage servicing,” McCarthy says.
But analysts say the tightening of credit has been an overcorrection. Prospective borrowers with less wealth and little credit history are now deemed riskier by the automated underwriting systems that dominate mortgage lending these days. As a result, they tend to be denied more often or given higher interest rates, cutting them off from one of the nation’s key wealth-building tools — despite the fact that they might well be capable of responsibly making mortgage payments.
This matters in other ways besides equity concerns. With the nation facing an affordable housing crisis, increased homeownership (which can be cheaper than renting in some areas) could be one solution for lower-income people, if only they could get a fair mortgage.
The lending industry has been paying attention. Last year, the Office of the Comptroller of the Currency (OCC), a key bank regulator, began convening leaders from a range of industries to find ways to bring in people without credit histories. The group is examining banks’ potential use of “cash-flow data”: basic records of transactions that indicate balances, routine payments and overdrafts in consumers’ bank accounts. It’s a new concept, one that the research organization FinRegLab has repeatedly shown can effectively predict borrowers’ ability to repay a loan. Even FICO, whose credit scores are most widely used by the industry, is offering a new product utilizing consumers’ account data (with their permission).
And in August, Fannie Mae — which, together with Freddie Mac, has for years relied on a rigid, outdated credit score — announced that it would begin to factor first-time home buyers’ rental-payment histories into its underwriting guidelines.
Pete Mills, senior vice president of residential policy at the Mortgage Bankers Association, says the group supports these new pathways to homeownership for borrowers of color and others with thin credit. “If you have a history of making your rental payment on time, it’s a great indicator for someone who may not have a thick file [that includes] an auto loan or other loans,” he says. But the organization is waiting for the new alternative credit-scoring programs to scale up before putting its weight behind them.
Karan Kaul, a senior research associate at the Urban Institute, acknowledges that few of the changes will come quickly. “Progress is slow because we’re trying to rewire the system of credit underwriting here,” he says. Still, he adds, “There’s tons going on.”
Meanwhile, some groups are more ambitious about the changes they’d like to see. Beneficial State Foundation, the philanthropic wing of California-based Beneficial State Bank, has convened a year-long “Underwriting for Racial Justice” working group. Like the OCC, the foundation has gathered leaders from across the industry to examine promising practices that might help underserved borrowers.
Unlike the OCC, however, this cohort includes representatives from less-mainstream lenders like community development financial institutions (CDFIs), credit unions and community banks — institutions that are deeply familiar with the needs of low-income and minority communities and are already successfully lending to them.
“Many of the folks in this group are saying, ‘We’re not using credit scores at all,’ ” says Erin Kilmer-Neel, Beneficial State Foundation’s executive director. Instead of trying to incorporate new data into credit scores, or raising rates for those deemed particularly risky, these lenders focus on potential borrowers’ true ability to repay and try to make it easier for them. And much of that comes from interactions. “In this group, it’s about relationships,” she says.
The smaller banks scrutinize information that’s similar to what the credit agencies and bigger lenders are examining: whether the client has a steady source of income, a record of consistently paying monthly expenses and a history of repaying smaller loans on time. But CDFIs and community banks do some of their underwriting manually, rather than using an automated system. That hands-on process is slower, but allows them to include less-conventional borrowers without penalizing them for being “risky” or charging predatory fees, and to flag elements that might not otherwise show up in an automated system.
Molina-Brantley, who leads the financial literacy courses at Berkshire Bank, can testify to that. While Berkshire has branches across New York and New England, it functions like a community bank. So when Molina-Brantley hears that clients with no credit history have been faithfully paying their bills, he’s able to get that information incorporated into their file — without complex algorithms. “We can work with our underwriters and let them know that the rental history is available,” he says.
Similarly, Come Dream Come Build (CDCB), a CDFI in Brownsville, Tex., has a loan fund it uses to grant mortgages. While the organization, located in one of the nation’s poorest cities, does use credit scores, they’re not front and center. “Just because someone’s daughter broke her arm and they can’t pay the hospital bill doesn’t mean they won’t pay their mortgage,” says Nick Mitchell-Bennett, CDCB’s executive director. “We look at other factors, like utilities, and we’ll not look at some, like medical debt.”
Because CDCB services the loan fund itself, these loans have slightly higher interest rates than conventional mortgages, but the increase isn’t because borrowers are perceived as a bad bet. And they have held up, even during hard times. “We haven’t had a foreclosure in 25 years, not on that [loan] product,” he says. “We’ve gotten close. But we’re very tied into our community, and we know people.” So if a payment is late, the organization’s staff might call the borrower to learn what’s going on — and then help them figure out how to pay at least some of it.
In New Mexico, the CDFI Homewise serves largely low-income borrowers of color, three-quarters of whom wouldn’t traditionally qualify for a “good” mortgage. But utilizing a one-on-one process, Homewise has been able to get them into affordable loans at the same terms as conventional borrowers, and to keep delinquency rates below 2 percent, outperforming the market.
Around the country, CDFIs and other community lenders have been reporting similar success stories for at least two decades, even throughout the housing crisis. As a 2015 Treasury Department report wrote: “Despite serving [predominantly] low-income markets, CDFI banks and credit unions had virtually the same level of performance as mainstream financial institutions.” And that has continued over the past few years.
“It tells you that there’s ways to do homeownership and lending right. “We’re talking about expanding the market ... and doing it responsibly,” says McCarthy at the Lincoln Center for Land Policy.
But the process takes more time and effort, and the loans tend to be smaller. Scaling up these “high-touch” transactions remains a challenge.
Meanwhile, online lenders, or fintechs, have come to dominate the mortgage market. Those companies have also recognized the need — and buying power — of consumers with less-than-stellar credit or no credit at all. They’re trying to bridge that gap with more data, analyzing disparate aspects of consumers’ lives to determine their ability to repay a loan. But it’s unlikely technology alone will allow them to reach all of the consumers who could successfully handle a mortgage.
Over time, the two approaches to lending may meet in the middle. For example, the Center for NYC Neighborhoods, a CDFI, is developing a financial technology platform to gain a holistic, but scalable, picture of consumers’ borrowing capacity, and overcome some of the mortgage market’s racial inequities.
And CDFI advocates say they’d like to see more conventional lenders incorporate their hands-on style. It may well save them money through reduced foreclosures — but there’s more to it than that, they say.
“It’s pretty unfair of us to say that we can’t afford to spend time and money to have high-touch relationships,” says Kilmer-Neel of the Beneficial State Foundation. In an industry that’s perpetuated so much racial inequity over the decades, she adds, it’s also the right thing to do.