Search for: All records

Abstract Affordable housing that incorporates sustainability goals into its design has the potential to address both health and economic disparities via enhanced energy‐efficiency, structural durability and indoor environmental quality. Despite the potential for these win‐win advances, survey data of U.S. local governments indicate these types of equity investments remain rare. This study explores barriers and pathways to distributional equity via energy‐efficient housing. Using archival city sustainability survey data collected during a period of heightened U.S. federal investment in local government energy‐efficiency programs, we combine machine learning (ML) and process‐tracing approaches for modeling the complex drivers and barriers underlying these decisions. First, we ask, how do characteristics of a city's organizational learning methods—its administrative structure, past experience with housing programs, resources, stakeholder engagement and planning—predict policy commitments to green affordable housing? Using ensemble ML methods, we find that three specific modes of organizational learning—past experience with affordable housing programs, seeking assistance from neighborhood groups and the technical expertise of professional green organizations—are the most impactful features in determining city commitments to constructing green affordable housing. Our second stage uses process‐tracing within a specific case identified by the ML models to determine the ordering of these factors and to provide more nuance on green‐housing policy implementation.