Article | Regulations
The New Data of Student Debt
by Christopher K. Odinet*
From Vol. 92, No. 6 (September 2019)
92 S. Cal. L. Rev. 1617 (2019)
Keywords: Student Loan, Education-Based Data Lending, Financial Technology (Fintech)
Where you go to college and what you choose to study has always been important, but, with the help of data science, it may now determine whether you get a student loan. Silicon Valley is increasingly setting its sights on student lending. Financial technology (“fintech”) firms such as SoFi, CommonBond, and Upstart are ever-expanding their online lending activities to help students finance or refinance educational expenses. These online companies are using a wide array of alternative, education-based data points—ranging from applicants’ chosen majors, assessment scores, the college or university they attend, job history, and cohort default rates—to determine creditworthiness. Fintech firms argue that through their low overhead and innovative approaches to lending they are able to widen access to credit for underserved Americans. Indeed, there is much to recommend regarding the use of different kinds of information about young consumers in order assess their financial ability. Student borrowers are notoriously disadvantaged by the extant scoring system that heavily favors having a past credit history. Yet there are also downsides to the use of education-based, alternative data by private lenders. This Article critiques the use of this education-based information, arguing that while it can have a positive effect in promoting social mobility, it could also have significant downsides. Chief among these are reifying existing credit barriers along lines of wealth and class and further contributing to discriminatory lending practices that harm women, black and Latino Americans, and other minority groups. The discrimination issue is particularly salient because of the novel and opaque underwriting algorithms that facilitate these online loans. This Article concludes by proposing three-pillared regulatory guidance for private student lenders to use in designing, implementing, and monitoring their education-based data lending programs.
*. Associate Professor of Law and Affiliate Associate Professor in Entrepreneurship, University of Oklahoma, Norman, OK. The Author thanks Aryn Bussey, Seth Frotman, Michael Pierce, Tianna Gibbs, Avlana Eisenberg, Richard C. Chen, Kaiponanea Matsumara, Sarah Dadush, Jeremy McClane, Emily Berman, Donald Kochan, Erin Sheley, Melissa Mortazavi, Roger Michalski, Kit Johnson, Eric Johnson, Sarah Burstein, Brian Larson, John P. Ropiequet, the participants and the editorial board of the Loyola Consumer Law Review Symposium on the “Future of the CFPB,” the participants of the Central States Law Schools Association Conference, the faculty at the University of Iowa College of Law, and Kate Sablosky Elengold for their helpful comments and critiques on earlier drafts, either in writing or in conversation. This Article is the second in a series of works under the auspices of the Fintech Finance Project, which looks to study the development of law and innovation in lending. As always, the Author thanks the University of Oklahoma College of Law’s library staff for their skillful research support. All errors and views are the Author’s alone.