Chapter 1:
Financial Inclusion

Furthering Financial Inclusion with “Alternative Data”

A lack of high-quality, individualized financial data can exclude a person from the mainstream financial system.

Credit is often extended on the basis of an individual’s credit score. ^[12] Today, most credit scores are generated from credit reports, which are maintained by national credit bureaus. Credit reports contain a somewhat limited set of financial indicators, including data about existing credit cards and loans. Traditional credit scores have been shown to be accurate in predicting consumers’ creditworthiness (that is, the chance that the consumer will repay credit in accordance with its terms). ^[13] But not all individuals have a credit report with enough data to generate a credit score. Thus, in some cases, a lack of high-quality, individualized financial data can exclude an individual from the mainstream financial system.

According to the National Credit Reporting Association, as many as 70 million Americans do not have a credit score, or have a lower score than their full financial history would warrant. ^[14] Because many of these so-called “no file” or “thin-file” individuals regularly pay their utility and phone bills, some groups have argued that this payment data (which is currently not included in most credit files) should be routinely reported to credit bureaus. The major credit bureaus agree, and have developed scoring algorithms that can consider this so-called “alternative” data when it is included in a credit report. ^[15]

The Policy and Economic Research Council (PERC), a non-profit think tank focused on economic policy issues, claims that there is “overwhelming and incontrovertible” evidence that including bill repayment data in credit scores would help low-income individuals. ^[16] It argues that most people will benefit when such data is included, particularly low-income individuals. This is true, the group continues, “whether the metric is credit score changes, credit score tier changes, or changes in portfolio acceptance given a target default rate.” ^[17] PERC thus urges advocates to make the financial system “more inclusive by making credit files more inclusive.” ^[18]

But the National Consumer Law Center (NCLC) has arrived at different conclusions. It claims that the industry is motivated in part by a desire to force utility bills to the “top of [consumers’] payment pile,” where such bills might go if they became a factor in access to credit. ^[19] It also emphasizes that if short-term delinquent payments become part of a credit file, “many low-income customers would receive negative credit reporting marks.” ^[20] Finally, it worries that reporting of utility payments would conflict with established state regulatory policies designed to protect low-income individuals, who may “sometimes defer full payment of utility bills, knowing they are protected from shutoff.” ^[21] In short, concluded NCLC, “[f]ull utility credit reporting will cause disproportionate harm to low-income consumers.” ^[22]

Complicating matters, credit reports are also used to evaluate individuals for jobs, screen applicants for apartment rentals, and generate “marketing scores” for use in marketing consumer products. The impacts of these uses have not been tested or evaluated with the same rigor or transparency as the central use case of consumer credit underwriting, and there are risks that such non-credit uses of credit scores may have a disproportionate adverse impact on protected status groups. Some protections are in place: for example, most states now have some rules in place to regulate the use of credit information for insurance underwriting. ^[23] But as the use of credit data continues to expand, so too must the regulatory scrutiny as to the accuracy, fairness, and aggregate impact of such uses. Even if new data would be helpful in the specific context of credit, a broader debate that encompasses the other regulated uses of credit scores is needed.

Alternative data represents both an opportunity and a challenge for the civil rights community.

Alternative data represents both an opportunity and a challenge for the civil rights community. There is some strong evidence suggesting that alternative data could benefit marginalized groups, but much of the data underlying PERC’s studies remains proprietary. Greater transparency regarding the impacts of including new data have important work to do in making sure that none of these changes harm vulnerable groups. ^[24]