The spatial distribution of alternative financial service providers in Alabama: a multi-scale analysis

Alabama is a high-poverty state with permissive small-dollar credit regulation and one of the largest licenced alternative financial service provider (AFSP) markets in the United States. Using Alabama State Banking Department licencing data for 2011, 2014, 2015, 2016, and 2020, we estimate Poisson pseudo-maximum likelihood (PPML) models with high-dimensional fixed effects to examine conditional spatial associations between traditional financial institutions (TFIs), socioeconomic and demographic characteristics, and counts of deferred-presentment providers (payday lenders), small-loan providers, and pawnshops. Models are estimated at census-tract, ZIP-code, and county levels, and two-part hurdle models separate AFSP presence from conditional intensity. The results are more consistent with co-location than with a spatial-void account. Lagged credit-union intensity shows the most consistent positive association with AFSP counts across specifications, especially at the ZIP-code level, while lagged bank intensity is positive in most specifications, with significance concentrated on the ZIP-code intensity and tract-level entry margins. Population size and an inverted-U income gradient are the strongest predictors. Demographic associations are strongest at the tract level, especially Hispanic share, share aged 65 and over, and lower educational attainment, while Black share is not significant in the main PPML models. The findings underscore the importance of spatial scale for interpreting AFSP location.