The Data Revolutionizing Credit Scoring

Font Size:

Scholar argues that new data sources will expand credit access and raise issues of privacy and equity.

Font Size:

You may be surprised to learn that your next internet search could affect your credit score.

Lenders use credit scores to assess whether they will extend credit to consumers for homes, cars, education, and many other aspects of life. Traditionally, the credit reporting agencies that construct these scores have relied on consumers’ borrowing history. Now, however, they are making use of new and varied “alternative data” sources.

The use of alternative data in credit reporting holds both promise and peril, according to a recent article by Sahiba Chopra, a student at Vanderbilt law at the time of publication. Chopra calls for reforms to the regulatory framework governing credit scores to ensure fairness and equity in the use of alternative data.

Alternative data promise to expand access to loans for the more than one in seven adult Americans who lack an established credit score, Chopra argues. These individuals, a majority of whom are Black or Hispanic, are effectively locked out of applying for loans because they do not have any prior borrowing history that credit reporting agencies can analyze.

But today credit bureaus can use alternative data sources that go beyond borrowing history to assess whether these individuals will repay their loans. Sources of information can include consumers’ utility payments, educational attainment, or even internet browsing history.

Although it holds promise for extending access to credit, the use of alternative data to calculate credit scores also raises several concerns, Chopra warns. Data may contain inaccuracies. The collection of data can intrude into consumer privacy. And the complexity of algorithms used to process data also creates the potential for unintentional discrimination based on factors such as race and sex.

Chopra views existing statutes and regulation as insufficient to address these concerns.

Two statutes govern credit scores. The first—the Fair Credit Reporting Act (FCRA)—imposes duties on individuals or businesses that provide information for credit reports or that use these reports. The FCRA requires furnishers of credit information to take steps to ensure its accuracy. It also requires that consumers receive notification whenever a business or individual takes adverse action against them based on their credit report, such as by denying them a loan.

Chopra raises several concerns about the FCRA’s treatment of alternative data. The law imposes no requirement on the level of detail the adverse action notices must contain, meaning that consumers may have no idea which data sources factored into a decision to deny them credit. It also contains no limits on the types of information that credit reporting agencies use to construct their scores, which raises issues of privacy and equity when the agencies incorporate data such as browsing history.

A second federal law—the Equal Credit Opportunity Act (ECOA)— also governs credit scores and focuses on preventing discrimination in lending and borrowing against “protected classes” such as race, sex, or religion. The ECOA allows borrowers to sue lenders for intentional discrimination whenever a lender’s actions cause a “disproportionately negative” discriminatory impact.

Chopra claims that the ECOA fails to prevent discrimination in the age of alternative data because of the high hurdles it imposes on consumers seeking to prove a lender discriminated against them. Individuals already face difficulty showing that a business acted in a discriminatory fashion. Proving that an algorithm violates the ECOA is even more difficult, as consumers have little to no access to the underlying methodology used to grant or deny them access to credit.

To remedy these issues, several legal scholars have proposed a new Model Fairness and Transparency in Credit Scoring Act (FaTCSA). FaTCSA seeks to solve transparency issues by requiring credit reporting agencies to provide state attorneys generals and the public with detailed explanations of their methodologies. Proponents hope that increased disclosure will prompt the reporting agencies to comply with anti-discrimination requirements and increase the accuracy of their scores.

Chopra identifies several proposals that would enhance FaTCSA or other regulatory updates. New legislation or proposed regulations should allow consumers to trace data used in their credit report to its ultimate source, she argues. The law should also require lenders to define their standards for creditworthiness and inform consumers of concrete steps they can take to improve their scores. In addition, Chopra favors banning credit reporting agencies from using consumers’ browsing data as part of their credit score.

Chopra also proposes several steps that regulatory agencies can take under their existing authority. The Consumer Financial Protection Bureau, for example, could issue no-action letters to lenders appearing to use consumer data in a discriminatory manner. Such letters can condition the further use of these data on increased disclosures and compliance with anti-discrimination protections.

Furthermore, Chopra argues that agencies could operate on the presumption that the use of alternative data in credit reporting is discriminatory, shifting the burden to credit reporting agencies to prove their compliance with the law.

Chopra emphasizes the importance of a unified federal response to prevent regulatory fragmentation that will leave consumers in different states with varying levels of protection.

Based on current trends, the use of alternative data in credit reporting seems likely only to continue to increase. Chopra argues that Congress and federal regulators must work to ensure it increases access to credit while protecting consumer privacy.