The Use of Text as Data Methods in Public Administration: A Review and an Application to Agency Priorities

• 成果类型: Review
• 署名作者: Hollibaugh, Gary E., Jr.
• 署名单位: Pennsylvania Commonwealth System of Higher Education (PCSHE); University of Pittsburgh
• 刊物名称: JOURNAL OF PUBLIC ADMINISTRATION RESEARCH AND THEORY
• ISSN/ISSBN: 1053-1858
• DOI: 10.1093/jopart/muy045
• 发表日期: 2019
• 页码: 474-490
• 关键词: models

摘要

Due to the large amounts of text generated by government agencies and policymakers, computer-assisted text-as-data methods are starting to become more popular for scholars of public administration, public policy, and political science, as they allow for much faster processing of large amounts of textual data. Here, I review several of the more common text-as-data methods and provide an overview of their applicability to different data structures and substantive questions in public administration. Then, using thousands of documents issued by the Centers for Medicare & Medicaid Services and its predecessor agency-the Health Care Financing Administration-I showcase the utility of topic models by illustrating how they can be used in conjunction with other politically relevant covariates to help explain changes in agency priorities. I then conclude by discussing other possible uses for computational text analysis methods in public administration.