In this study, we explore the factors which contribute to the continued discrimination against racial and ethnic minorities, particularly Black and Hispanic communities, in mortgage loan pricing and approval. We draw on the work of Bartlett et. al. and hypothesize that despite laws against discrimination based on race and ethnicity, financial deserts (areas with low market competition) justify high loan prices, inevitably affecting disadvantaged groups such as racial minorities. We particularly suspect that financial deserts with a larger minority population would see exacerbated effects of low competition on their loan pricing, compared to areas with a smaller minority population. We also explore the effects of financial deserts on loan approval, to support previous findings. Data are sourced from the Home Mortgage Disclosure Act database, Freddie Mac database, and Census data. The data are used to construct models predicting loan interest rates, rate spread, and rejection rates on a zip code level, utilizing Ordinary Least Squares regression, regularization, and regression tree techniques, providing robustness to our findings through drawing on different modeling approaches.
