Understanding Best Practices in Managing Financial Constraints Among US Congregations: A Machine Learning Approach

This study investigates factors by integrating resource dependence, institutional, organizational ecology, and entrepreneurship theories to understand factors associated with U.S. congregational income using a data-driven approach. I analyze 76 features across 1,702 congregations from the National Congregations Study (2006–2007, 2012, 2018–2019) using an AutoML approach. The gradient boosted random forest model identifies that successful congregations increase service frequency, utilize technology, and establish endowment funds to secure resources. They adopt professional financial management practices, align demographic composition with local communities, engage in social service activities, and demonstrate entrepreneurial resourcefulness. These actions reflect adaptive strategies, institutional conformity, ecological fit, and entrepreneurial resourcefulness, enhancing the financial viability of congregations.