Financial infrastructure outreach to indigenous peoples: a three-approaches exploration for Colombia, Ecuador, and Peru

Objective: assess and contrast disparities in the extent and location of financial infrastructures of branches and correspondent agents for outreach in areas with high incidence of indigenous peoples and explore correlations with territorial covariates, in Colombia, Ecuador and Peru. Methodology: compare three approaches: a descriptive study explores key distributions, econometric analyses identify spatial disparities and heterogeneous associations, and machine learning techniques (regression trees, SHAP values) uncover complex, non-linear relationships, not revealed by traditional methods. Results: high indigenous peoples incidence, above 50% of the population, and location in rural areas and Amazon regions are associated with lower financial outreach, reflecting universal and idiosyncratic barriers. Low population density is associated with low financial outreach in different regions. Limitations: data availability on a few variables have limited model specifications. Secondary sources (census, official financial infrastructure data) are used. Results are for three specific countries. Originality: this paper covers an unexplored topic; there is no other three-country, three-methods study on this topic. Conclusion: financial inclusion policies must address the need for convenient physical, institutional and digital meeting places and close existing gaps, to respond to the unique socioeconomic and cultural context of indigenous peoples.