Bridging the Language Divide: How Local Governments Implement Housing-Related Language Access for Asian Immigrant Communities with Limited English Proficiency in Metropolitan Washington

Abstract

For the roughly 25 million U.S. residents with limited English proficiency (LEP), navigating housing systems poses uniquely complex challenges: applications, lease agreements, recertifications, and grievance procedures are conducted overwhelmingly in English, and the consequences of misunderstanding a document or missing a deadline can mean the loss of shelter itself. Despite decades of federal mandates requiring meaningful language access, implementation remains uneven and underfunded, and the capacity that has been built is built for Spanish, leaving linguistically diverse communities, particularly Asian immigrant populations, without equivalent provision. This dissertation asks how housing related agencies in the Washington, D.C. metropolitan region construct, implement, and sustain language access for LEP communities, and where in that process Asian LEP communities can be rendered institutionally invisible. The analysis is organized around Institutional Linguistic Capital (ILC), a framework developed in this dissertation that synthesizes Bourdieu's (1991) linguistic capital, street-level bureaucracy theory (Lipsky, 2010), and administrative burden scholarship (Herd and Moynihan, 2018). ILC treats languages as resources that institutions value, accumulate, ration, and convert, holding capacity in three forms: operational, institutional, and embodied. Paper 1 develops the framework and names its default failure trajectory the cascade of invisibility; each empirical paper then examines one form of capacity across the full sequence. Paper 2 develops AI-augmented policy auditing, applying large language model pipelines with critical discourse analysis and human validation to 905 unique documents from three jurisdictions, and finds that inequity lies not in frequency of mention, since languages appear roughly in proportion to LEP population share, but in conversion: Spanish is named alongside staff, departments, and dedicated lines, while other languages receive recognition that commits little to no material capacity. Paper 3 draws on 25 interviews with 29 participants across six jurisdictions to identify administrative home, funded mandates, and point-of-service verification as the structural features producing unequal access, and shows rationing to be self-reproducing, since provision shapes the demand data that later justify provision. Paper 4 examines frontline practice under federal retrenchment, introducing protective discretion and embodied institutional linguistic capital to explain how workers convert personal linguistic and cultural resources into public capacity, and at what cost. Read together, the papers reveal a cascade of invisibility operating through distinct mechanisms at each level, a persistent compliance-access gap, and a sustainability paradox in which the features that make language access work are the most vulnerable to disruption. The dissertation contributes a transferable theoretical framework, an auditing methodology that agencies can apply to their own documents, and documentation of Asian LEP marginalization at a moment when federal rollbacks have shifted responsibility for multilingual inclusion to local jurisdictions.