Too often, the debate over student loans omits a key problem affecting millions of Americans — a lack of access to fair credit that will enable them to be approved for financing. For those looking to receive education to improve their career opportunities, the key to entry into these opportunities is financing to help pay for career training programs like coding bootcamps, welding programs, healthcare technician programs, and more.
That’s why Climb works hard to identify these types of programs that provide positive outcomes for their learners, and why it’s our mission to increase access to them. For too many of their applicants, a thin credit history or a low FICO score ends up barring them from the very thing that would improve their trajectory. According to 2020 data from LendEDU, the average approved private student loan borrower had a 748 credit score, almost 20% higher than the average applicant score.
And the impact of this discrepancy is even greater on communities of color. A recent Brookings Institute paper showed that credit scores are “deeply correlated . . . with race.” It cites the conclusions of other fair lending stakeholders who note that “[o]ur current credit-scoring systems have a disparate impact on people and communities of color” in the form of lower approvals and higher-cost credit.
It’s because of this that we’ve teamed up with Zest AI to address the lack of access to fair credit and break down the current system’s dependence on traditional credit scoring. They’re working to avoid the pitfalls that come with traditional credit scoring by using transparent and explainable machine learning (ML) lending models. These models may effectively analyze up to ten times more data points from the credit bureaus, which allows them to paint a more holistic picture of borrowers beyond just their credit history.
With a more comprehensive look at applicants, ML models can expand credit access safely — they generate a more accurate risk assessment, while reducing the weight of variables that have been shown to correlate with race, gender, and ethnicity. Removing the biases from a model, especially one as long-standing as credit scoring, requires more tedious and intricate calculations than a human can do. But Zest AI allows lenders to generate a fairer model (measured by approval rate disparities between races) that is just as if not more accurate than the one they had before, all while providing more clarity and transparency on how the underwriting data specifically drives accuracy and fairness in the model.
Teaming up with Zest AI is just one way Climb is working to ensure more fairness and accessibility in education financing. If you’d like to learn more about Zest AI, and other steps that can be taken to address this lack of access, click the link below!