Are poor people credit worthy? The answer may surprise you.
A few weeks ago, I focused on how data analytics and Analytics4 could help accelerate philanthropy’s move towards a beneficiary voice. Today, we want to share how Analytics4 can help support credit decisions for the lending community.
According to a May 2015 publication released by the Consumer Financial Protection Bureau (CFPB), 26 million adults in the U.S. are “credit invisible,” with an additional 19 million having unscorable credit records. These findings indicate a strong relationship between income and a scored credit record. In our work with low-income families, FII has found that a proportion of those with whom we partner have a relatively low credit score, or do not have any credit rating at all. This report upholds our experience: nearly 30% of consumers in low-income neighborhoods are credit invisible, when compared to only 4% in upper-income neighborhoods. An additional 15% of those in low-income neighborhoods have unscored records, compared to just 5% in upper-income neighborhoods.
For working-poor families, access to a loan via a mainstream lending institution is greatly contingent upon a strong credit score and/or healthy credit history. And, as the CFPB points out, many low-income borrowers do not meet this narrow underwriting criteria. As a result, borrowers are often forced to use payday lenders with high interest rates.
The constricted criteria used for traditional scoring is beginning to be recognized, and there are now a growing number of private sector initiatives exploring alternative data and technological solutions to improve credit risk assessment. For example, companies are using machine-learning techniques to assess credit worthiness, combining financial and non-financial inputs, including information from social media sites such as Facebook, group affiliations, or number of “friends” and contacts. While the use of these alternative data broadens the inputs for credit risk assessment, the criteria used continues to be class-biased. Affiliation with Harvard University and other such institutions can increase one’s credit worthiness. But this continues to limit access for low-income households, as they may not be active in social media, or may not be affiliated with the “right” groups or friends.
At FII, we gather ongoing data from families showing that they will make reliable borrowers. We’ve tested a variety of information. We are capturing data sets such as the participation history in community lending circles, the initiatives parents take to engage in their children’s education, actions they take to help others in their community, and character recommendations from other community members.
A classic example is that of FII family member Mateo, who was continually losing out on job opportunities because the jobs he was applying for required personal transportation. A traditional bank would not make an auto loan to Mateo, as he did not fit the criteria for traditional “credit worthiness.” As a member in good standing, Mateo procured a no-interest loan from FII, and, combined with his own savings, was able to purchase a vehicle. Once he had his own transportation, Mateo was able to land the job that would keep his family afloat, and avoid falling into financial crisis.
Analytics4 has begun using the strength-based data provided by our families to offer access to capital via zero interest loans, endorsements in KivaZip and invitations to participate in Puddle. To date, this represents nearly $150,000 in capital deployed, with a 95% payback experience. While still relatively small, our experience is starting to show that new strength-based information can help inform reliability, bridge the credit gap for the working poor and ultimately support sound credit decisions.