Lending, Data, and Civil Rights Conference Highlights Tough Questions
A few notes from a recent gathering of advocates, industry representatives, researchers, and regulators
April 15 2015Last week, The Leadership Conference Education Fund, the Center on Privacy & Technology at Georgetown Law, and Americans for Financial Reform hosted a conference called Data, Lending, and Civil Rights. The gathering attracted privacy, financial, and civil rights advocates, as well as industry representatives and researchers.
Attendees focused their attention on the role of new data in the future of lending. Although there were few easy answers, everyone seemed to agree that new data could both help and hurt communities at the economic margins.
Several key themes and questions emerged throughout the day, highlighting the need for future research:
Will use of new data to make credit eligibility decisions be helpful, or hurtful?
As we wrote last week, FICO is piloting a new credit score that will consider “alternative data,” such as telecommunications and utility bills and property records. And smaller companies are using even more diverse kinds of data, including social network connections and household buying habits.
I gave a presentation explaining that bill repayment behavior might be predictive of loan repayment behavior — and might provide some benefit to many creditworthy individuals who do not currently have access to mainstream credit — but cautioned that we should be skeptical about other nontraditional data.
In short, today’s research suggests that some types of new data might be helpful, but we need to tread carefully and ensure that they are used in ways that are both predictive and fair. Fair lending laws, which prohibit discrimination against a loan applicant on the basis of race, sex, and other factors, were designed for the world of FICO, and it’s not yet clear how they will apply to scores based on new data, especially data gathered from major Internet platforms.
How can we rein in the use of credit data in non-credit contexts, such as in employment?
People’s credit scores — and the underlying credit reports used to generate those scores — are available to employers and landlords. Many attendees worried that credit data (old and new) might block low-income communities from accessing jobs, housing, and insurance. Because of these concerns, there was no clear consensus as to whether a low credit score, even based on accurate data, was better than none.
This is a foundational question that will shape how advocates approach new credit data.
How can we better understand and regulate the financial “lead generation” market, which sells people’s data to lenders?
One major point of concern was the collection and resale of sensitive personal information about financially vulnerable individuals. For example, a person searching for “quick loan” or “fast cash” online is likely to end up on the website of a lead generator, not an online lender. This lead generator would collect that person’s personal data and immediately resell it to multiple lenders, data aggregators, and, potentially, even fraudsters.
These kinds of practices drive much of the online payday loan industry. At the conference, there was widespread consensus that both advocates and regulators needed to know more about how these businesses operate and, potentially, how to better regulate them.