Methods: We reviewed the extant research literature on credit scoring as a step toward applying this framework (Denyer and Tranfield 2009; Paul and Criado 2020; Tranfield et al. 2003). Key intentions of this approach to the literature were to enhance knowledge and inform practice (Paul and Criado 2020; Tranfield et al. 2003), and our review combined structured and theory-based approaches (Green et al. 2006; Paul and Criado 2020). Our review included 51 studies that met specific inclusion criteria out of an initial sample of 126 published between 1971 and 2021. A majority (70%) of studies were published in business, economics, and finance journals and were published since 2008.
Results: Overall, studies deployed theoretical frameworks and assumptions that were consistent with neoclassical economics and favored industry practices. A majority of studies analyzed proprietary data from credit bureaus or lenders, meaning that industry representatives determined early data structures in ways that undoubtedly influenced research. Finally, many studies failed to control for both credit scores and racial demographics in predicting outcomes like loan originations and defaults. When scores were predictive, the strength of their associations were often smaller and less significant (even nonsignificant) when racial demographics were otherwise controlled.
Conclusions/Implications: Existing research provides an alibi for anti-Black racism embedded in credit scores. Normative research adopts rationales for credit scoring that bear some similarity—even if distant—to how accounting tools and economic principles were deployed against Black people during chattel slavery. Based on our framework and literature review findings, we conclude with implications for policy and a call to abolish the practice of credit scoring while imagining new, abolitionist alternatives for people to live safely and with dignity.