Abstract
A rigorous investigation of the past 50 years of scholarship in the Journal of Health Policy, Policy and Law requires the application of large-scale computational text tools. We utilized four strategies to analyze research articles (N = 1,532): keyword searching, named entity recognition, unsupervised topic modeling, and use of large language modeling. We examined geographies, main topics and sub-topics of articles over time; attention to equity/diversity, racism, women's health, and LGBTQ+ people; and authors' methodological approaches. Articles have examined a diversity of health policy topics, although most have concerned health care access and health insurance, and the majority are U.S.-based. Authors have consistently (more than 50% of articles) mentioned issues of health equity and health disparities, across the 50 years. They have applied a range of methodological approaches, with empirical legal and policy analyses the most prominent; qualitative approaches have been consistent, while quantitative articles have seen an increase in the past decade. Findings demonstrate the utility of computational methods in future applications for health policy and politics scholarship.
