Abstract
PurposeThe purpose of this study was to explore the concept of distrust of traditional banking institutions as a factor that can explain the choice to remain unbanked in a marketplace that is designed to be financially inclusive.Design/methodology/approachEarning, spending, saving and borrowing data collected between May 2021 and February 2022 from 17,819 consumers living in the United States were used to examine the factors associated with distrust of banks. Using a conceptual framework borrowed from the health services profession, the study was conducted in two stages. At the first stage, distrust among the unbanked and banked was estimated using a Boruta-random forest algorithm. At the second stage of the analysis, a logit regression model was estimated to validate the variables identified in the Boruta-random forest analysis.FindingsResults from the analyses show that distrust of banks is multi-layered where being older, believing the country is heading in the wrong direction and being less confident in one's ability to obtain a personal loan in the amount of $1 to $999 are important factors related to distrust of banks among the unbanked.Research limitations/implicationsThis study shows how an ensemble machine learning technique based on a decision-tree methodology can be used to obtain unique insights into complicated data and large datasets within the bank marketing field.Originality/valueThe paper provides a discussion about ways domains of trust and specific variables can be utilized to address the persistent problem of financial exclusion in the United States. Implications for bankers, researchers, educators and policymakers are provided.