This study develops a dual-model framework to examine how policy-relevant organizations forge network ties and circulate policy messages within an issue-specific policy arena. Drawing on the core concept of homophily in social network analysis and framing theory, I investigate whether organizational type and policy stance predict dyadic network connections and message-framing strategies. I further incorporate the structural notion of brokerage to assess whether organizational attributes shape actors’ positions in the broader policy network.

In parallel, the study extends existing research by shifting attention to two-way communication and public responsiveness—dimensions often overlooked in extant studies. To address this gap, I analyze how organizational structures and strategic communication choices attract public engagement on Facebook, thereby linking elite interactions with mass-level reactions. Employing a mixed-methods design that integrates exponential random graph models (ERGMs) with content analysis, I examine interactions among 373 organizations and 1,188 policy-related posts in Taiwan’s traffic safety network.

The findings show that organizational homophily strongly predicts inter-organizational ties—particularly among political parties and civil society groups—while the influence of organizational attributes on strategic communication is evident across all types of actors. Moreover, organizational type significantly shapes network positions, which in turn affect levels of public attention. Overall, the proposed model illuminates the structural and communicative dynamics that link diverse policy stakeholders, offering both theoretical insights and practical implications for research on political communication and policy networks.