Association for Public Policy Analysis & Management

Objectives. As education advocacy organizations become increasingly influential in shaping public policy, researchers have begun looking to the ways these organizations use evidence to substantiate their agendas and build coalitions. Prior research suggests that building advocacy coalitions involves engaging in discursive debates over core policy beliefs that shape how the public and policymakers think about policy issues. Building from this work, the present study uses policy network analysis to analyze how members of the Policy Innovators in Education (PIE) Network use different forms of evidence to shape public perceptions about education policy and leverage their policy agendas.

Data Sources. Established in 2007, the PIE Network is now comprised of over 80 advocacy organizations operating in 35 states and D.C., and 22 National Partners consisting of think tanks, research centers, and national-level advocacy organizations. The primary source of data for this study includes the three most recent publicly available policy briefs retrieved from n=60 PIE members’ websites (i.e., a total of 180 policy briefs). From each member’s and National Partner’s website, we also collected mission statements, advocacy summaries, and press releases. All documents were uploaded to the Discourse Network Analyzer (DNA) for coding and formal social network analysis. Finally, the latter datasets were linked to IPPSR’s Correlates of State Policy dataset to capture state-level policy attributes where PIE members do their advocacy work.

Methods. The analysis began with a formal content coding of the 180 policy briefs to categorize the types of evidence used (e.g., academic article, think tank brief, governmental report) in direct relation to the policy topics being advocated (e.g., performance-based accountability). We also coded instances in which PIE Network Members cited documents from other members or National Partners. This process yielded a total of 4,628 citations linked to 48 unique policy topics. Second, matrices were exported from DNA and uploaded to R for the social network analysis. To describe the networks, we used two-mode network analysis to explore the centrality of the different sources of evidence that PIE members used to establish claims about education policy. Two-mode network graphs, multidimensional scaling, and correspondence analysiswere also used to explore variation in the types of evidence by the substantive topics being cited. Finally, exponential random graph models (ERGMs) were run to test the probability of the network structure given a set of relevant covariates.

Results & Implications. Preliminary results suggest that PIE members tended to selectively reference other PIE members when advocating for teacher-based reforms (e.g., merit pay, evaluation), thus creating an echo chamber when advocating for the most central issues in the PIE Network. On the other hand, members tended to cite academic articles only when referencing work related to equity-based issues, such as wrap-around services, the achievement gap, and early childhood education. The ERGMs suggest that these patterns can be linked to preference popularity within the network. Taken together, the results provide policy makers with a first-hand look at the types and quality of evidence currently shaping public advocacy work in education.