The Open Government Directive of 2009 led to a major change in how federal agencies communicate with the public.  The memorandum advised federal executives to enhance the level of transparency, accountability, and collaboration using a number of strategies and tools.  Many agencies implemented social media engagement as a response.   Several studies have examined the effectiveness, frequency, and content of this new wave of government-to-public communication and found that execution and degree of engagement varied widely across federal agencies.  Cross-agency exchanges on social media are expected to enhance the potential public reach of an agency by orders of magnitude in addition to the indirect benefits of informal collaboration.  Moreover, the interdisciplinary systems of federal agencies networked through social media are expected to compound research and development impacts that address complex policy challenges. However, there is little research that attempts to investigate whether agencies are cross-communicating via social media, and if so, to what degree.

This study contributes to the literature by offering a first look into intra-government federal use of social media in the context of cross-agency communication.  Using Twitter as the communication platform, the first objective is to discern who-follows-whom patterns that exist across agencies.  The study will specifically focus on federal agencies that conduct and disseminate research.  The second objective investigates the posting and sharing of other agencies’ posts, known as tweets and retweets, to determine levels of influence and potential for collaboration, taking into consideration an agency’s overall research mission and organizational characteristics (such as size).  Understanding how agencies leverage Twitter to foster knowledge sharing through intra-government retweets motivates this research.  

This study uses a combination of primary and secondary data.  Focusing on Twitter feeds of 110 federal research-based departments, agencies, and affiliates, I collected 113,746 tweets with meta-data posted from January 2013 to March 2014.  These cross-sectional data were supplemented with publicly available agency characteristics.  Exponential random graph models are used to investigate intra-government federal followership and the propensity to form network ties.  Network autoregressive outcome models are used to assess the propensity to retweet within and across federal institutions. 

The findings of this study offer three-fold benefits.  First, it contributes to the literature on government-to-public communication and knowledge management theory to practice, while assessing the efficacy of social media for transparency, and accountability under the Open Government Directive.  Second, it applies empirical network analysis techniques in a novel way to research relationships within subsections of government, providing evidence of collaboration in new ways.  Third, it presents new perspectives on how technology policies through social media may address government as a communicator of the research it generates.