Credit Scoring and the Risk of Inclusion
Inequality, credit scoring, and alternative data
In a submission invited by the U.S. House Committee on Financial Services, sociologist Tamara K. Nopper examines the possibilities of alternative data initiatives for mitigating racial inequality in the U.S. credit scoring system. A 2021–2022 Data & Society Fellow, Dr. Nopper addresses the October 27 hearing on “Bringing Consumer Protection Back: A Semi-Annual Review of the Consumer Financial Protection Bureau” and underscores that “inclusion” may benefit companies in the credit scoring industry rather than systemically marginalized groups. The full hearing and additional materials are now available.
I applaud the U.S. House Committee on Financial Services for holding its recent hearing “Bringing Consumer Protection Back: A Semi-Annual Review of the Consumer Financial Protection Bureau” (October 27, 2021). As a sociologist who studies credit scoring and alternative data initiatives, and considers the prospects for addressing racial inequality, I appreciate the invitation to submit my thoughts about consumer protection in terms of the credit scoring industry. Given that the majority of complaints received by the Consumer Financial Protection Bureau (CFPB) regards credit or consumer reporting, better regulation of the credit scoring industry is vital for consumers and for addressing the racial wealth gap.
As underscored by the title of a previous hearing convened by your committee, the current U.S. credit scoring system is “a biased, broken system.” Documented problems include a for-profit credit scoring industry dominated by three credit reporting agencies; a lack of transparency regarding scores and algorithms; credit scores playing an increasingly substantive role in shaping socioeconomic opportunities beyond credit card approvals and small business and mortgage lending, such as applications for automobiles, apartments, or private student loans; credit reporting errors that put a great deal of burden on consumers to try to fix, often without fair resolution; and a relative lack of regulation of the credit scoring industry, particularly regarding transparency, data breaches, and reporting errors.
Additionally, critiques have been raised about how credit scoring disproportionately disadvantages certain racial groups, particularly those who are Black and Latinx, and the implications for the racial wealth gap. Along with credit scores often being significantly lower for African Americans and Latinx compared to white people, there are also millions of people, a significant portion of whom are Black or Latinx and/or working-class or poor, being shut out of the credit scoring system due to being credit unscorable or credit invisible.
In my statement I address alternative data, which is proposed as a solution to some of these issues. I detail some of my concerns, including the risks of alternative data on consumers and how alternative data does not address the for-profit credit scoring system or other issues requiring better regulation.
Alternative data as a proposed solution
Alternative data initiatives are touted as part of a broader agenda of financial inclusion. Regarding credit scores, financial inclusion advocates focus on credit unscorables or credit invisibles. The former have either a “thin” or “stale” credit file and the latter have no credit history with one of the three nationwide credit reporting agencies, Equifax, Experian, and TransUnion. The CFPB estimates 19 million adults as credit unscorable and 26 million as credit invisible. Combined, these 45 million represent almost 20 percent of the adult population, with African Americans and Latinx more likely to be credit unscorable or invisible than white people and Asian Americans.
Being credit unscorable or invisible can have negative consequences as “the default assumption of lenders in that no score equals high risk. Such applicants are almost always rejected.” Financial inclusion advocates suggest that helping groups become “credit visible” will ameliorate some of the racial and class patterns in use of “alternative financial services” (AFS) outside of the banking system, such as “money orders, check cashing, international remittances, payday loans, refund anticipation loans, rent-to-own services, pawn shop loans, or auto title loans.” Some racial wealth gap initiatives, such as those calling for increasing homeownership among African Americans and Latinx, posit including alternative data as a solution.
Broadly, alternative data is understood to be “any data that are not ‘traditional.’” Per the CFPB, “traditional data” is:
data assembled and managed in the core credit files of the nationwide consumer reporting agencies, which includes tradeline information (including certain loan or credit limit information, debt repayment history, and account status), and credit inquiries, as well as information from public records relating to civil judgments, tax liens, and bankruptcies. It also refers to data customarily provided by consumers as part of applications for credit, such as income or length of time in residence.
Alternative data can include payment histories on bank accounts, utility bills, rental payments, and telecommunications. It can also include social media activity.
Potential risks of alternative data
Concerns have been raised regarding the risks of including some alternative data. For example, utility companies might have guidelines regarding late payments that do not impose immediate shut off of services but in alternative data records, delinquent accounts might get flagged negatively, thereby hurting one’s credit score. There are also questions about what impact collecting alternative data might have related to consumer efforts to dispute questionable charges. Finally, the ability to pay bills in a timely manner is related to existing wealth as well as employment and wages and the racial politics of the labor market. As Chi Chi Wu, Staff Attorney at the National Consumer Law Center, notes in her testimony for “A Biased, Broken System,” we should “proceed with caution,” taking into consideration which alternative data is used and how, as well as consumer consent.
The issue of consumer consent raises significant questions regarding issues of privacy. What does it mean to have more of your socioeconomic activity, including possible agonizing decisions regarding what bills to pay late, if at all, versus others, etc., be subjected to a process that could ultimately make your credit profile bad or worse? How might alternative data initiatives subject those already socioeconomically disadvantaged to a potentially unfair higher standard regarding their financial activities and payment behaviors?
Who gets to have privacy in this society is often a race, gender, and class issue and Black people have historically been the most targeted for a lack of privacy. While we are all, regardless of our social status, exposed to a lack of privacy given the datafication of social life, a commitment to racial, economic, and gender justice encourages us to consider how a lack of privacy can also be a form of compulsory visibility, which can also be a form of surveillance. How might this surveillance come with penalties that could negatively impact consumers? And given the long history of racism, and specifically anti-Blackness, in shaping who is disproportionately punished and how punishment operates, how might what is touted as inclusion simply be compulsory visibility, where increased visibility is imposed upon you to participate in existing systems? How might the inclusion of alternative data increase the visibility of those deemed credit unscorable or invisible — disproportionately Black and Latinx and/or poor and working-class — and possibly subject them to more scrutiny or a visibility penalty — where becoming more visible within existing discriminatory logics and systems increases the vulnerability to being penalized or victims of structural disadvantage? Terms such as inclusion and visibility can be powerful, but what are the risks associated with these processes and how can we ensure that those who are credit unscorable or invisible do not incur more penalties than they already do from existing discrimination and monitoring of socioeconomic activity?
Additionally, as Chi Chi Wu underscores, “feeding more data to the credit bureaus is not the solution. Feeding them more data only increases the oligopoly power of these three companies, giving them even more power over our information and our financial lives.” Simply, including alternative data will not address the current for-profit structure of the U.S. credit scoring system and does little to regulate the credit scoring industry.
Alternative data and the for-profit credit scoring system
I want us to consider how alternative data is part of this for-profit landscape. In some cases, alternative data initiatives are about competition among for-profit companies in the credit scoring marketplace. For example, in 2006 credit reporting agencies Experian, Equifax, and TransUnion, who control the credit reports of over 200 million U.S. residents, established the company VantageScore Solutions to challenge the dominance of Fair Isaac Corporation and its FICO score. VantageScore Solutions supported the Credit Score Competition Act of 2017, which was introduced by Senators Tim Scott and Mark Warner. As the former announced the bill on his website, “Our goal in introducing this legislation is to encourage Fannie Mae and Freddie Mac to consider more inclusive methodologies in determining a borrower’s creditworthiness. Alternate scoring models have the potential to make homeownership a reality for more qualified borrowers who lack access to traditional forms of credit.” In a statement applauding the Credit Score Competition Act of 2017, president and CEO of VantageScore Solutions Barrett Burns stated, “No single company should have a government-sanctioned monopoly, especially when there are millions of consumers that are negatively impacted. Infusing competition into this integral area will improve fairness, transparency, and inclusiveness without compromising on standards.”
“Inclusive methodologies” and “inclusiveness” may be appealing language but the references to inclusion here may be more about competition between for-profit companies than a genuine solution to the racial wealth gap.
And “inclusiveness” certainly does not address the need to regulate the credit scoring industry and to make credit scoring companies more transparent and accountable. While there may be some stated differences between scores, such as how data is collected and how long of a credit history is required, these differences do little to challenge the for-profit nature of the credit scoring industry. Inclusion, then, is often depicted as good for the systematically discriminated against group when in reality it may primarily benefit for-profit companies competing against each other in the credit scoring marketplace. What is treated as good for marketplace competitors is not always good for consumers, especially when the for-profit credit scoring industry does not involve fairness or transparency. As Amy M. Traub, Associate Director of Policy and Research at Dēmos, in her testimony for “A Biased, Broken System” emphasizes,
Since consumers are not the customers of the private credit reporting agencies, they have no market mechanism to demand accountability or fairness: Consumers cannot opt out of the system or choose to work with a competing company.
These credit reporting agencies might band together to create a new score to compete with the FICO score and claim to care about fairness, transparency, and inclusiveness, but ultimately they “collect lending and payment data on 220 million Americans without consumers’ permission or approval, and there is no way for consumers to opt out from having personal financial data collected.” Simply put, “the aim of the credit reporting 3 agencies is to generate profit, which they do by extracting, packaging, and selling data about consumers’ personal borrowing and payment activity.”
The credit scoring industry benefits from weak regulation. Regarding transparency:
Behind the three-digit score…is a process that cannot be fully understood, challenged, or audited by the individuals scored or even by the regulators charged with protecting them. Credit bureaus routinely deny requests for details on their scoring systems. No one outside the scoring entity can conduct an audit of the underlying predictive algorithms. Algorithms, and even the median and average scores, remain secret. The lack of transparency of credit-scoring systems leaves consumers confounded by how and why their scores change.
Instead of stronger regulation of the credit scoring industry, which is needed, onerous burdens are imposed on consumers to deal with errors and the unscrupulous practices of credit reporting agencies — often with frustrating results and a lack of accountability.
The proposed Comprehensive Consumer Credit Reporting Reform Act and National Credit Reporting Agency Act are steps in the right direction. Related to the latter, a public credit registry, particularly that outlined by Traub in her testimony on behalf of Dēmos for “A Biased, Broken System,” would be a powerful change from the current for-profit system. Nevertheless there is still a great deal of research that needs to be done to determine what, if any, alternative data might be included or how it might be used in a public credit registry to ensure it does not subject people to a visibility penalty.
Finally, there is also another possibility to consider, which is the elimination of the credit scoring system altogether. Ultimately, alternative data initiatives, whether proposed as part of the current for-profit credit scoring system or for a public credit registry, require a scoring system to have relevance. We spend a lot of time and energy trying to figure out how to get people scored rather than on reconstructing our economic system to be more equitable and not rely on the datafication and scoring of our socioeconomic activity. It would be bold, generative, and generous to consider what would it mean to design a socioeconomic system that does not require a “scored society,” and in this case a credit score, determining our socioeconomic opportunities and cost of living.
