Association for Public Policy Analysis & Management

Local governments are increasingly part of online networks, with the vast majority of them using multiple social media platforms such as Facebook, Twitter, or Instagram (Mossberger et al. 2013; Fusi & Feeney, 2016). The growth of social media has attracted great attention from public management researchers, who have widely examined how public organizations use social media to connect with their constituency, access or disseminate information, and engage with citizens (Campbell, Lambright, & Wells, 2014; Oliveira & Welch, 2013).

While most public management literature has focused on how social media tools can influence the relationship between citizens and government, fewer studies have investigated online engagement among government agencies (Wukich et al. 2017). Online connections among government agencies are important because they offer easy access to information and opportunities to share resources among members of the network. Online networks are also gaining salience due to their facilitation role for collaboration and coordination in emergency situations (Yates & Paquette, 2011; Kavanaugh et al, 2012).

This paper investigates which micro-level mechanisms guide the formation of the global structure of Twitter network among US small and medium sized cities as well as their connections with other key institutional actors, such as government agencies (i.e. US departments) and membership organizations (i.e. ICMA, League of Cities). In particular, we are interested in how city governments send, receive, or reciprocate ties on social media and we examine relational mechanisms such as reciprocity and centrality, and node attributes such as capacity, size, information needs and homophily.

We suggest that cities tend to reciprocate ties from other cities and send ties to institutional and policy actors when they face greater information need. Cities with bigger technical capacity are also more likely to form ties with government management association such as ICMA. We further hypothesize that key network organizations such as ICMA and Leagues of Cities will serve as broker to facilitate formation of ties between city governments. Finally, we expect to find homophily effect for state and size.

We draw from extensive Twitter data collected in 2017 from the accounts of 500 US cities with a population between 25,000 to 250,000. Using both ‘followers’ and ‘following’ ties we are able to investigate the full network of cities and institutional actors. Public management researchers have called for a more nuanced approach to social media use which considers the “articulated network of confirmed contact or follower relationships” (Mergel, 2014, p. 4). This approach recognizes the network structure of social media tools and relies on social network analysis (SNA) to understand patterns of behavior online. We integrate relational data with node attribute data such as demographics from the US Census, disaster declaration from the Federal Emergency Management Agency and Twitter account information. Our results shed lights on the antecedents of online network formation and explain how cities engage on social media for information access, dissemination and engagement purposes. Finally, we provide first evidence on organization centrality and brokerage in online networks.