The Power of Alternative Data to Create a More Inclusive Credit Market and Build Household Assets
Today, Prosperity Now is releasing an updated policy brief on the use of alternative data in the credit reporting and scoring process.
Whether you can get an affordable loan to purchase a home, start a business or buy a car, even if you are offered a job or rented a property, is largely dependent on one three-digit number – your credit score.
Credit scores are supposed to determine whether someone is a good credit risk by examining how they pay off certain debt obligations, and lenders rely heavily on this number when deciding who gets a loan and on what terms. Not surprisingly, the credit reporting and scoring industry is a big business dominated by three main credit reporting agencies: Experian, TransUnion and Equifax. The massive fall-out from the data breach at Equifax—where 143 million people had their credit reports hacked—highlights the reach these agencies can have and their potential impact.
Considering the stakes involved, you might think a good deal of time and energy is put into ensuring that these scores do what they are supposed to do, namely, accurately predict a person’s creditworthiness. In truth, the most popular scoring models marketed by these agencies and used by lenders for underwriting result in tens of millions of consumers without scores (so-called “credit invisibles”). It also saddles some consumers with scores that do not adequately gauge their creditworthiness, and this is especially the case for low-income families and households of color.
Part of the problem is the data the agencies use to generate a credit score. They tend to incorporate certain payments into their scoring models and not others. For example, traditional scoring models will consider whether a person routinely pays their mortgage, but not whether someone pays their rent. This means a person paying $1000 a month for their mortgage gets a scoring boost from this behavior, while someone paying the same amount for rent does not, even though both actions illustrate an ability to repay by a consumer. Moreover, since a person with a mortgage is more likely to be wealthier than a person paying rent, these popular scoring models work against the interest of lower income borrowers.
What can be done to address this problem? Incorporate other forms of debt, like rent, into the scoring process. Debt that is not regularly included in scoring models, such as rent, is called “alternative data,” and is a promising way to make the credit reporting industry more inclusive without sacrificing consumer safety or the predictive power of the score.
The policy brief released today highlights the problems that exist in traditional scoring models in more detail and presents evidence to back up the use of certain types of alternative data in scoring, as well as outlining some federal policy solutions that could lead to a fairer and more accurate scoring market. Some key takeaways include:
- Evidence exists to show that some forms of alternative data, namely rent, utility and phone payments, has the potential to boost the scores of many people, as well as create scores for the first time for many more.
- While greater efforts should be made to encourage the use of alternative data like rent, utilities and phone payments, not all forms of alternative data should be part of the scoring process. Indeed, some alternative data raises concerns about transparency, privacy and discriminatory biases.
- Solutions include having agencies like the Consumer Financial Protection Bureau (CFPB) provide more clarity as to how these data can be used by the reporting industry properly and having the Government Sponsored Enterprises (GSEs)—Fannie Mae and Freddie Mac—promote the use of more inclusive scoring models in lender underwriting.
When the consequences of this number could mean the difference between receiving a loan with a 4% interest rate, or receiving one with a rate that is double that or higher, if a loan is extended at all, the financial implications become clear, and it underscores the importance of getting it right. This is especially true for the low-income families and households of color that stand to lose the most, and the sensible use of alternative data can help us achieve this.