Examining the Interconnectedness of State High Court Twitter

Abstract

Judges are not the first political officials that come to mind when one considers the role of social media in modern politics. Following in the wake of some prominent judicial personalities adopting Twitter, however, a growing number of state high court judges have adopted and established more public personas on the platform. Judges use Twitter in substantively different ways than traditional elected officials (Curry and Fix 2019); however, little is understood about how the use of such social media platforms affects broader judicial networks. Recognizing that judges, like typical social media users, may aspire to expand their networks to build and appeal to broader audiences, we contend that active participation in judicial Twitterverse could yield personal and professional advantages. Here, we address a currently unexplored question: To what extent have judges formed a distinctive “judicial network,” on Twitter, and what discernible patterns present in these networks? Leveraging the unique structure of social media, we collect comprehensive network data on judging using Twitter and analyze what institutional and social factors impact greater power within the judicial network. We find that early adoption, electoral concerns, and connective links between judges all impact the strength of the judicial network, highlighting the complex motivations driving judicial Twitter engagement, and the significance of network building in judges’ social media strategies and its potential impact on career advancement.

Introduction

State high courts, and the judges who constitute them, could operate with total independence from their sister courts in other states. Yet, we know that they do not. These courts often cite each other’s decisions (Caldeira 1985; Hinkle and Nelson 2016) and interact in a host of other ways building a network of connections through traditional legal avenues. Modern technology has opened new avenues for these judges to build new networks through social media. As more state court judges become active on Twitter and other social media platforms (Curry and Fix 2019), we see a growing network among these judges developing. Understanding the degree to which the emerging judicial Twitterverse is interconnected furthers our knowledge of the types of formal and informal connections across state high courts more broadly. While the development of social media networks differs qualitatively from citation networks, it is important to better understand how and why judges build these networks and the implications thereof. Moreover, this research has broader implications for our understanding of judges as actors who pursue multiple goals simultaneously (see, e.g., Baum 1997). This article contributes to these important areas by providing preliminary answers to a novel research question: Do judges follow each other on Twitter and are there any detectable patterns in these networks?

Judges are not the first political officials that come to mind when one considers the role of social media in modern politics. However, the declared Twitter Laureate of Texas, ¹ former Justice Don Willet of the Texas Supreme Court, had a modest audience of 100,000 followers before his nomination to the Fifth Circuit Court of Appeals, and one of the platform’s most notable jurists and evangelists, Judge Stephen Dillard of Georgia, ² has approximately 29,800 followers as of September 2023. While substantial, these numbers are dwarfed in comparison to follower counts for high-ranking members of Congress such as Speaker Pelosi (8.1 million followers), younger representatives like Rep. Alexandra Ocasio-Cortez (13.2 million followers), or institutional Twitter accounts like those for the White House (8.7 million followers). ³ While the reach of judges is not nearly as wide, judges like Dillard celebrate the platform for its ability to provide outreach to the public. Judge Dillard, in particular, has spoken frequently about his use of the platform as an educational tool and a way of engaging with broader constituencies (Dillard 2017). ⁴

Following in the wake of prominent judicial Twitter users such as Judges Willet and Dillard, a significant number of judges who sit on state high courts have adopted and made use of Twitter accounts to establish a public persona in substantively different ways than traditional elected officials (Curry and Fix 2019). Examining the source tweets of state high court judges, Curry and Fix (2019) find that while social media adoption by judges is almost entirely conditioned by the need for reelection, the behavior judges engage in is dissimilar from traditional elected officials. While judges tweet more during election years, their content is focused on personalization activity instead of electoral appeals or policy pronouncements. You are more likely to learn about a judge’s favorite sports team, musical artist, or BBQ place as opposed to how they feel about a policy issue. While judges do not use Twitter in a uniform manner or in the way traditional politicians do (Curry and Fix 2019; Dillard 2017; Dillard and McCormack 2019), one clear trend is that a significant amount of judicial Twitter behavior centers around personalization activities: activities that increase name recognition and memorability among their constituencies.

This article expands on prior research examining the use of social media by state court judges by recognizing that judges may have another goal motivating their social media activity. The average user does not participate in social media solely for electoral purposes: they join to expand their literal social networks. Judges should also engage in this behavior, especially when one considers the possible personal and career benefits available from being a part of a network of judges. Benefits could range from increased citations across courts to endorsements for promotion to the federal bench. Being central to any network increases visibility, and this should be something judges seek out, as visibility, even on social media, may increase their likelihood for advancement in their career.

By using a unique database of all state high court judges with public Twitter accounts, we examine the degree of interconnectedness of judges on state high courts. We find that judges do have a closely knit associational network on Twitter; however, variation does occur in the centrality of judges, with those retained by partisan elections occupying more central positions than other judges.

The value of social media for judges

Since its creation in 2008, Twitter has grown to be a formidable political tool. While not the only social media platform used by politicians and political candidates, Twitter’s popularity makes it a primary platform for candidates and public officials eager for a low-cost method to gain a following and extend their voice (Amman 2010; Evans, Cordova, and Sipole 2014; Lassen and Brown 2011). A growing body of literature has formed around the examination of Twitter adoption patterns and behavior on the website, mostly focused on its use in electoral campaigns (Evans et al. 2014; Gross and Johnson 2016; Jungherr 2016; Meeks 2016; Stier et al. 2018) and how the platform is leveraged by members of Congress (Evans et al. 2014; Golbeck, Grimes, and Rogers 2010; Lassen and Brown 2011), presidents (Gross and Johnson 2016; Miles and Haider‐Markel 2020), and governors (McGregor, Lawrence, and Cardona 2017) in order to promote public policy and maintain their own personal “brands” in an increasingly online environment.

Notably left out of these investigations has been the growing use of social media platforms by judges and legal actors (but see, Curry and Fix 2019; Mattan, Puddister, and Small 2020). Given the nature of many political actors’ behavior on social media, this oversight may seem intuitive. While the legislators and executives are assumed to be open to the divisiveness that occurs on social media platforms, judges are often viewed as being “above the fray”— seeking to avoid political confrontation and attempting to place themselves “outside of politics.” Yet, as Curry and Fix (2019) note, judges on Twitter largely manage to avoid those controversial aspects of the platform, instead primarily using it to engage in discernably non-political network building. Their findings largely align with research on judicial audiences (Baum 2006; Black et al. 2016; Romano and Curry 2019). According to Baum (2006), judges “court” various audiences in order to garner favorability and maintain the legitimacy of the institution. Curry and Fix (2019) argue that state court judges, likewise, recognize the great degree of variation in their audiences and as such are more likely to innovate by finding new ways to reach out to such audiences.

Twitter in particular (and social media in general) provides judges with a mechanism for achieving professional goals that should increase the likelihood of adoption. Specifically, we theorize that social media use can help judges achieve three distinct, but not mutually exclusive, goals. First, platforms such as Twitter, Facebook, and Instagram can help judges achieve legitimating goals by allowing judges to engage more directly with the public, offering opportunities for both public education and constituency outreach. ⁵ The knowledge of what judges do, and how courts work, is of significant importance to the legitimacy of these institutions. Systematically, as knowledge of judges increases, so does the legitimacy of the courts they sit on (Gibson and Caldeira 2011). Social media offers a low-cost way for individual judges to raise the institutional legitimacy of the judicial system, as well as their personal legitimacy and the legitimacy of their court. Additionally, social media use allows participating judges to magnify and promote other public outreach activities. For example, former Chief Justice Bridget Mary McCormack of the Michigan Supreme Court periodically tweeted announcements regarding live streams of oral arguments. ⁶

For elected judges, Twitter can help them achieve a second and more direct goal. Like other elected officials, winning elections is the primary means to keep themselves in a position of power. These judges, unlike their colleagues in elite appointment systems, gain an individualized benefit from the use of Twitter via increased name recognition and potential electoral benefits by providing a cost-effective mechanism for these judges to connect directly to their constituency. To some degree, this allows incumbent judges and judicial nominees to campaign in much the same way as candidates in other branches of government, albeit decidedly different in some specific ways. While overtly political content is exceedingly rare among judges on Twitter, more modest requests for support in reelection bids are not. Beyond tweets directly related to their reelection, the use of Twitter by elected judges increases generally in years that they are up for reelection (Curry and Fix 2019), making it clear that they are aware of these benefits. Put simply, elected judges on Twitter try to maximize the beneficial electoral gains through name recognition on the platform while minimizing the potential negative consequences that other elected officials face when using social media (for discussion of these consequences, see Evans et al. 2014; McGregor et al. 2017; Meeks 2016). ⁷

In this article, we argue that social media can help judges achieve more than legitimacy and electoral goals. Specifically, we posit that many state court judges will also use Twitter to achieve a third goal: increasing their own networks by developing a collegial link with other judges both in their states and beyond. Given their unique position within the broader judicial hierarchy, judges may find value in seeking out similar voices to their own to help promote cross-state networks and find mutually beneficial outlets for increasing the legitimacy of their own respective chambers, such as the development of the Lady Justice podcast hosted by Justice Rhonda Wood (@JudgeRhondaWood) of Arkansas, Chief Justice Beth Walker (@bethwalkr) of West Virginia, and former Chief Justice Bridget McCormack (@BridgetMaryMc) of Michigan. These networks can foster learning, sharing, and ultimately improve the administration of justice across states.

Making friends and influencing judicial audiences

Similar to the emergence of scholarly networks (Donelan 2016) as well as #AppellateTwitter ⁸—where legal practitioners have connected to share questions, strategy, and best practices—judges have also begun to develop their own professional network on the platform. While judges are unlikely to use Twitter to directly promote their opinions (Curry and Fix 2019), judges may use it to cultivate a network of other judges, legal scholars, practitioners, and law students, increasing their name recognition and influence among their colleagues, in and out of state. Forming a larger social media network outside their state and beyond the people they represent offers these judges a low-cost way to potentially increase their prestige among fellow judges as well as state and federal politicians.

Beginning with the assumption that judges engage in various behaviors because they wish to cater to certain audiences (Baum 2006; Black et al. 2016; Romano and Curry 2019), we argue that the development of and participation in a distinctive judicial network is a logical creation in that it serves to increase the prestige of individual judges independent of electoral goals or legitimizing effects on their courts. Specifically, we assert that the use of Twitter to achieve network-building goals can help judges acquire three potential individual-level benefits: collegiality, professional recognition, and progressive ambition. While not all judges who use Twitter will use the platform in search of all three of these benefits—indeed some judges may join Twitter without having any specific goals in mind—Twitter offers a possible opportunity for judges with some or all of these motivations. Table 1 provides a summary of the goals judges can achieve from social media use and the specific benefits they can gain from its use.

Table 1. Benefit and goals for judges to use social media

Concerning collegiality, there is an extensive literature on how factors related to intra-court interactions can influence judicial behavior (Epstein and Knight 1997; Maltzman, Spriggs, and Wahlbeck 2000; Romano and Curry 2019), yet significantly less work has been done on inter-court relationship building among judges. There are significant reasons to believe that state high court judges would care about building these networks. Even though state high courts are not bound by the decisions of other state courts, these decisions are typically afforded great respect and deemed authoritative, but not binding (Fix and Kassow 2020). In a shared common law system like the United States where precedents often “migrate” from one jurisdiction to another (Fix 2016), an extensive literature exists showing that judicial decisions, like legislative policies, often diffuse across states (Canon and Baum 1981; Glick and Hays 1991). Additionally, we know that citation patterns among state high courts are governed by a variety of systematic influences including geography and court prestige (Caldeira 1985; Hinkle and Nelson 2016) and that networks of citations exist among state courts (see, e.g. Fowler et al. 2007). However, these studies are all at the level of individual decisions or policies and focus on the behavior of courts rather than individual judges. In other words, we have little systematic evidence regarding whether individual state high court judges work towards building relationships with their colleagues in other states.

Despite this absence of scholarship on the topic, we have two key reasons to believe that judges may desire stronger connections with other jurists in other states. First, state high court judges “occupy a unique place in the American political system” as they are simultaneously “at the pinnacle of their branch of state government” but “technically bound by U.S. Supreme Court decisions” with respect to federal issues (Fix and Kassow 2020, 63–64). As such, they may seek out relationships with a small number of individuals who can truly understand their position. In these cases, early adopters of new technologies may be key in order to “set an example” for how to effectively use a new platform for professional and political outreach (for more on technological diffusion and its relation to political institutions, see Epstein 2018). Additionally, and more broadly, judges often travel in similar circles as the audiences they wish to cultivate. Social media allows judges a link to professional organizations, colloquia, and conferences, as well as the ability to discover mutual connections that extend from such organizations and events. Similar to its use in other professional networks such as education (see Carpenter and Krutka 2015), judges interested in cultivating inter-court, collegial relationships may see Twitter as an ideal environment to foster greater interaction and dissemination of information about trends and new ideas. For example, there has been a growing number of law school symposia devoted to the use of Twitter by judges and several of the most prominent Twitter-using state court judges have participated, ⁹ as well as collaborative articles by judges on the topic of judicial Twitter use (Dillard and McCormack 2019). In network models, such as judicial Twitter, the seeking out of such preferential networks should lead to increased network homophily ¹⁰ or the tendency of individuals in a network to associate with others of shared attributes, characteristics, or properties. As a result, we should expect that the edges of the network should be reduced and dependent on the centrality a judge has to the network creating a more inclusive open network for association. Put simply, having multiple judges on Twitter gives those judges greater ability to be exclusive when choosing who to follow, decreasing the centrality of a singular judge acting as a linchpin for the entire network. This should also lead to greater tie reciprocity within the network as well, as individual judges form mutual connections and reciprocate follows between each other, reinforcing network clusters.

Hypothesis [1]: As the number of connections between judges (or “edges”) increases in the network, individual judges will become less central to the network.

Hypothesis [2]: As a judge reciprocates following other judges (“tie reciprocity”) more frequently, they will become more central to the network. ¹¹

Second, like most other professionals, we expect that some judges desire external recognition of the quality of their work on bench. As Klein and Morrisroe observe “if one of the things [judges] value is the respect of their colleagues and if they actively seek to earn that respect, then we could better explain and predict the ways in which judges act by identifying the kinds of behaviors that seem to earn prestige” (1999, 372). We assert that network building via Twitter is one such mechanism and that some judges will seek to increase their prestige as this can facilitate greater influence (Klein and Morrisroe 1999). The ability of judges at any level to impact the evolution of the law comes through their written opinions (Fix 2016; Fix and Kassow 2020; Maltzmann, Spriggs, and Wahlbeck 2000; Robinson 2013; Romano and Curry 2019). For state high court judges, this is most clear in terms of their ability to shape law in their own states where their opinions are binding precedent for all lower court judges. Yet, this is not the sole influence their opinions can have. State high courts often cite decisions from other states, especially those that are geographically proximate and high prestige (Caldeira 1985; Hinkle and Nelson 2016). While most prior work on state court prestige focuses on whether certain state high courts are more prestigious vis-à-vis other courts, prior scholarship on the U.S. Courts of Appeals suggests that the prestige of individual judges varies significantly and impacts citation patterns (Klein and Morrisroe 1999; but see, Hinkle 2015) and counter judge success (Hinkle, Nelson, and Hazelton 2020).

Over time, it is likely that some judges will develop a specific reputation among the broader legal community. Building a Twitter network of other judges, appellate attorneys, and legal academics can be helpful with this. While it is rare for state high court judges to directly promote their recent opinions on Twitter (Curry and Fix 2019), it does happen on occasion. ¹² However, it is important to note that while some judges may develop a reputation for crafting especially persuasive, well-written, or scholarly opinions, this is not the only type of positive professional reputation they can seek to attain. Compared to the rare tweet about a specific opinion, general interactions between different judges, between a judge and an appellate attorney, or between a judge and a legal academic are exceedingly common. The bulk of these interactions do not involve a discussion of a specific opinion or even general legal principles but can work to foster a different form of prosocial professional recognition in the legal community. ¹³ Simply put, a discussion over the preferred word processing program for drafting judicial opinions can evolve into a relationship with fellow judges across very disparate courts. We should expect, therefore, that as network transitivity ¹⁴ – or the property wherein like-minded individuals in a network have a higher probability of associating with one another leading to localized clustering or relationships – increases, network centrality should decrease as well, as no singular judge can be said to be vital to the survival of the network.

Hypothesis [3]: As judges become more interconnected within the network (“network transitivity”), individual judges will become less central to the network.

Finally, while we are primarily concerned with the first two benefits judges can achieve though the use of Twitter for network building discussed above, it is also reasonable to assume that one other benefit for network building on Twitter may be the goal of progressive ambition. ¹⁵ In the context of more traditional politicians, scholars have found that those with progressive ambition, or the desire to seek higher office, will engage in behaviors distinctive from those without such goals (Black 1972; Maestas et al. 2006; Maisel and Stone 1997; Rohde 1979). Central to any sort of progressive ambition is name recognition, and name recognition is something that social media sites like Twitter can provide. In much the same way that Twitter can foster interaction between judges and the broader legal community, it can cultivate relationships between judges and other government officials. By using Twitter to increase their professional recognition, ¹⁶ judges will organically increase their likelihood of being recommended by members of the legal community from whom advice may be sought when there is a federal judgeship vacancy. For these judges, Twitter offers a low-cost way to get their name in front of those actors influential in the appointment process. Anecdotally, it is worth noting that the first three years of the Trump presidency witnessed the appointment of four Twitter using state high court judges to the federal bench. ¹⁷

Cultivating a judicial network

Thus far, our main argument has been that judges may use sites like Twitter to increase inter-court collegiality, broader professional recognition from their peers, and career advancement. While there are many ways to acquire these things, Twitter can provide them in a comparatively low-cost, low-effort medium. Furthermore, as more judges join social media, and the network of judges grows, the cost of network maintenance reduces significantly for all involved. The use of Twitter by judges is largely self-promotional in nature, even though the individualized goals of each judge may be different. Whether the specific benefits they seek are related to legitimating goals, electoral goals, network-building goals, or some combination, judges will attempt to increase the size of their audience. Our examination focuses directly on how judges cultivate associations using Twitter in order to develop a larger public profile centered on their office. While this directly ties to the third goal of network building, it can indirectly help advance all three goals, as building a more prolific profile on social media platforms will bring them to the attention of additional members of their reelection constituency and—assuming their social media activity is positive—can increase the public’s goodwill towards their institution, increasing its legitimacy.

Judges at the state level contend with multiple audiences when engaging in any public activity (Baum 2006; Romano and Curry 2019), and this is likely to be reflected in how they use Twitter. Those judges concerned with external professional audiences may emphasize the development of links within their own institutional and professional networks to gain greater acclaim and respect amongst their peers. Thus, we are primarily interested here in how judicial association networks function as a small, but important, part of judicial engagement on social media and what factors lead to increased or decreased centrality beyond those features natural to the creation of social networks.

To develop a better understanding of such networks, we examine data on 49 state high court judges who created public accounts on Twitter by 2017. We choose to focus here on how judges use Twitter to follow other judges specifically as part of the broader network of the legal community, as we suspect that these associations will be seen as the most outwardly beneficial to the development of a judge’s online persona. In order to evaluate the interconnected networks of judges that use Twitter, we created an original data set that included each state high court judge that had a public account on the platform in July 2017, when the data were collected. Modifying the publicly available code provided by Tweepy, we scrapped judges’ accounts of all tweets, media (pictures and video), the posts they like, and central to our analysis here, who they followed and who followed them. Because of the limits of Twitter’s public API, we were only able to scour each user’s previous 3,200 tweets, their first 50,000 followers and who they follow, and their latest 15,000 likes. While this does not reach the totality of the content generated by superusers like Judge Don Willett (who had 25.9k tweets and over 90k followers), it is more than enough for the overwhelming majority of the justices in our data. ¹⁸

Since our interest lies in the examination of association networks, we employ a social network analysis approach as a way to descriptively represent the data and conduct our empirical analysis. Political science has seen substantial growth in interest in social networks as a key means by which the political world operates and develops (Victor, Montgomery, and Lubell 2017). Network analysis relies on the development of relational connections between individuals (or nodes) (Wasserman and Faust 1994; Knoke and Yang 2007) and tests the strength of the bonds that form between each node. One of the primary benefits of our approach is the ability to visually represent social networks in a way that is not only appealing but also informative. Figure 1 provides our first glance at the directed relationship between state high court judges on Twitter. We find that most judges tend to be fairly clustered in their connections, indicating that judges exist in a relatively small and closely linked cluster within their broader social networks.

Figure 1. Judicial Association Networks on Twitter.

While visualization of social networks is beneficial to get a clear sense of the relational network of any community, it does not provide information on the value of nodes in that community by itself. To understand how particular judges find value in their Twitterverse, we must determine their centrality, or the social power of an individual node or link between nodes, to the broader network. Centrality can be measured in several ways; however, we begin by focusing on two distinct measures to begin our examination of the network for broader hypothesis generation by emphasizing the importance of individual judges within the network. First, we emphasize an individual judge’s importance to the network by focusing on degree centrality or the number of links that a given judge has with others in their network. Degree centrality is one of the more common units of analysis in social networks and can be best understood as a simple count of the number of connections one individual has within the network. In our data, degree centrality ranges from 0 (judges with no links within the broader judicial network such as Rebecca G. Bradley and Ann Walsh Bradley of Wisconsin) ¹⁹ to a maximum of 14 (Shawn Womack of Arkansas in our data), with a median number of links of 2.

As our interest is focused on the dyadic relationship between judges on Twitter, and how these relationships provide judges with broader connections, we can also examine the distance between two judges, and how one judge may act as an indirect “link” to others in the network. “Between” centrality focuses on the pathway from one judge to another, where shorter paths are considered “stronger” compared to those that must go through other nodes to connect with others in the network. Distance is measured specifically by counting how many nodes a particular judge must travel through in order to see or be seen by another judge in the network. Looking back at Figure 1, we can get a better visual understanding of the importance of betweenness. If we consider distance as the number of steps (or nodes) one has to travel in order to get from one judge to another in the network, we can see that some judges have more direct ties than others. For example, the distance between Judge Lidia Stiglich (@lidia_stiglich) and Judge Jess Dickinson (@JessHDickinson) is connected in the broader judicial network; however, the link is not direct, being mitigated by Josiah Coleman (@Coleman4MS_SC) who acts as an indirect link between the two. Thus, the distance between the two is greater compared to others in their direct networks. Judges in our data generally exist in short paths with one another, with a median betweenness score of 0.019 and a range between 0.000 and 0.050. This indicates that the current judicial network is fairly “tight knit,” with members having mutual followers and followings on Twitter, providing multiple “pathways” by which one judge may interact with others on the platform, providing some cursory evidence for hypotheses 2 and 3.

What does this mean for better understanding what influences the centrality of judges in the broader judicial network, however? To better understand the links between these two measures and how they work together to cultivate a collegial audience on social media, Figure 2 restructures the Twitterverse network so that we can better visualize the strength of both degree centrality (represented by the size of an individual node) and betweenness centrality (represented by the thickness of a given link between nodes). We find there is a minimal but positive correlation between degree and betweenness centrality (correlation coefficient of 0.116), but we do not find this correlation to be significant (p = 0.3). Given this, we can presume that no single individual or group of individuals explicitly control the network, acting as a gatekeeper by funneling connections through their account or by acting as the primary focus of the network, partly confirming hypothesis 1 and strengthening our theory that these networks exist as broader extensions of collegiality networks. Interesting for us here, nodes with greater degree centrality tend to come from states that use contested elections (either partisan or nonpartisan) as their method of retention (such as Arkansas Supreme Court Justices Shawn Womack, @judgewomack, and Rhonda Wood, @JudgeRhondaWood and North Carolina Supreme Court Justices Sam Ervin, IV, @ervin4supreme, and Barbara Jackson, @JudgeJackson), as well as notable Twitter “personalities” in the judicial network (such as Justice Don Willett, @JusticeWillett, in Texas and Justice Bridget Mary McCormick, @BridgetMaryMc, from Michigan). ²⁰ Given the distinct electoral pressures of contestable elections, we believe that judges from states that use partisan or nonpartisan elections gain greater centrality due to their need to promote themselves in an apolitical fashion, something that judges tend to do more frequently on Twitter compared to other actors engaging in electoral politics (see Curry and Fix 2019). We therefore expect that these will exhibit stronger ties within the model, as judges in these systems will find greater benefit in having a broader association network with other judges.

Hypothesis [4]: Judges that are retained through competitive elections will be more central to the network than judges retained via elite reappointment or retention elections.

Figure 2. Judicial Association Networks by Degree Centrality (Nodes) and Betweenness Centrality (Links).

Generating ties between judges on Twitter

We are interested in what institutional and social-demographic factors influence the creation of judicial networks on Twitter. Particularly, we are interested in how such a network can be used to cultivate a broader, collegial environment for judges and how methods of retention (among other factors) may influence who joins. To better examine how such features affect the likelihood of a tie being formed between two judges, we use an exponential random graph model (ERGM). ²¹ ERGM is a common method for analyzing social networks and allows us to determine whether edges between judges have a substantial impact on the determination of the network itself and whether other features similarly affect the network composition. Table 2 presents the results of our model.

Table 2. Exponential random graph model of network ties between judges on Twitter

Note: *p < 0.05.

Before moving to a discussion of how covariates and institutional variation affect the network structure, we start by examining how features of the network itself impact the possibility of a tie between two judges. As a baseline, a tie between judges indicates a unidirectional link between two judges in the network, or more specifically for analyses of Twitter, the instance that a judge follows another judge in the network (a “following”). Following Cranmer and Desmarais (2011), we interpret the strength of an edge as a measure of closeness between any two judges on Twitter. A negative edges coefficient therefore indicates that ties formed in the network are not random. The probability that any given judge in the network is tied to another is approximately 1.19%, based on the results in Table 2 and confirming hypothesis 1. While not necessarily the case that simply being a judge on Twitter leads other judges to follow, we do find however that forming a link (in the form of “following” another judge) does lead to significant levels of reciprocity. The probability that a judge follows other judges in the network (j -> i), as measured by tie reciprocity, is 97.69%. Since reciprocity is the measure of how likely it is for one actor to “choose” another in the network given that the latter actor chooses the first (Wasserman and Faust 1994), it makes sense that simply following judge A on Twitter does not necessarily mean that judge A will follow back. Finally, we find significant levels of transitivity in the network based on an analysis of edgewise shared partners. Transitivity, best understood as being a “friend of a friend” (Wasserman and Faust 1994), appears to lower the density of the network, indicating that judges are selective in the creation of their network creating localized clusters and confirming hypothesis 3. Based on the coefficient values in Table 2, we find that transitivity reduces the network density by −9.47%. Taken together, these findings are indicative of a rather loose network of association, with judges finding each other through common linkages (similar follows and followings) to form a tighter, more inclusive network over time.

Moving to an analysis of the individual and institutional covariates in the model, based on our findings in Table 2, we find interesting new evidence concerning the use of Twitter for campaigning and electoral politics. Specifically, judges who use Twitter to engage in electoral politics (through the creation of a campaign Twitter account) tend to have a significantly greater role in the network, increasing the probability of a tie by 2.04%. We contend that this increased significance is likely due to the transientness of electoral politics and an interest in keeping tabs on electoral outcomes. ²² As a result, other judges in the network appear to find value in forming links with these accounts, despite their continued relevance after election day. This finding modifies the theoretical arguments originally made by Curry and Fix (2019), as it appears that to some degree judges do take an interest in electoral politics, even if their professional accounts are not directly used for such purposes. This is further backed by our time-based covariate measuring the distance between a judge joining Twitter and their most recent retention. Judges may join Twitter leading up to their closest electoral period (a minimum of 8 years before their next retention period) or after their most recent electoral period (to a maximum of 5 years after their most recent retention). On average, most judges joined Twitter within one year of their most recent retention period. Accounting for whether the account is “professional” or “campaign” based, we find that judges who joined Twitter in the average one year timeframe increased the probability of a tie between other accounts by 1.60%. As the time between joining Twitter and their next retention period increases, the probability of a tie increases by 0.121 points. Overall, judges with the most time between joining Twitter and their next retention have a probability of a tie of approximately 0.68%, which increases to a maximum of 2.63% as their next retention year approaches.

Focusing our primary hypothesis concerning tie dependency on the method of retention, we find contradictory evidence compared to our initial hypothesis. As the primary distinction on Twitter tends to be based on whether a judge is required to face a challenger in their next election, or simply is required to survive a confidence vote, we include a binary variable for judges whose retention includes a contested election, which we find does significantly affect the probability of a tie in the network, however not in the direction we had expected. Our initial hypothesis had been that judges in contested election states were the ones most likely to form ties in the network; however, we find that being in a contested election state decreases the probability of a tie by −0.875. Accordingly, being in a contested election state and controlling for other variables in the model, the probability of a tie is 0.754%, and this tie increases by 1.808% for non-contested election states. It is noteworthy that followings on Twitter, in the broader sense, do not necessarily impact the likelihood of a tie between judges. Both variables for “number of followers” and “number of accounts followed” on Twitter do not reach statistical significance in the model, which indicates that while major personalities in the judicial Twittersphere may have farther reaching influence with other audiences, within the judge-specific network, personality may be outweighed by collegiality. This is further backed up by our findings that Chief Justices and years on the court have significant though countervailing effects. Chief Justices were found to be more prominent in the network – indicative of the leadership role these individuals take on including performing more public-facing work for the court. In contrast, judges with longer tenures on the court tend to have a less prominent role in the network, with a year increase in tenure diminishing the prominence in the network by 0.030. Future research will need to tease apart this difference.

Conclusions

Judges, even those who are retained via elections, behave differently than other elected officials. Yet, that does not mean that they cannot benefit from the same tools as other elected officials, they just use those tools in different ways. Twitter, and other social media platforms, are one such tool. The use of social media can help judges achieve three distinct goals: legitimating, electoral, and network building. While this article focuses specifically on the third of these goals, it is important to note that building a large social media network indirectly makes it easier to achieve the other two goals as well.

The judicial Twitter network is quite interconnected, but those who are most likely to be central are judges who are selected and retained by partisan elections. This mirrors findings of Curry and Fix (2019). They find that partisan judges are much more likely to use Twitter than are judges selected and retained by alternative methods. Of all the methods used to retain judges in the states, partisan elections are perceived to be the most political. Not only are judges in these systems first movers comparatively on the adoption of Twitter, but they are also most likely to be central in the maintenance of the network. Romano and Curry (2019) find that judges retained by contested elections, especially partisan judges, are most likely to consider their audience when writing salient opinions. Our findings confirm that while all judges engage in largely the same behavior, by using a method of selection that prioritizes public politics, partisan elected systems condition the behavior of judges differently.

The judicial Twitter network is also quite isolated. Despite how interconnected the Twitterverse may appear, one of the most notable findings of this analysis is that the network has not been formed based upon reciprocity. This means that when one judge follows another, that judge is not statistically likely to follow the first judge back. Judges on Twitter are selective. Judges who create accounts for campaigning purposes rather than professional networking are much less likely to be central to the network, pushed to the fringes of the network. Centrality in the judicial network is most assuredly for possible electoral benefit, as judges in states with contested elections exhibit much stronger ties in the network along with early adoption of the platform. These judges have some other distinct features. They were early adopters, used their personal accounts to educate about judges and judging more generally, and cultivated a public persona that increased name recognition. As judges and other institutional actors begin to migrate from Twitter to other social media platforms, future research ought to continue examining how these actors engage with the public and others within their networks in novel, innovative ways in order to disseminate information and provide an educative role to the public.

Clearly state high court judges are but one part of the legal community. It is likely that other actors, including attorneys, law professors, and even political scientists, will play an important role in these networks for judges who desire to build greater collegiality, achieve more professional recognition, and facilitate their progressive ambition. Future research should explore these broader networks. Specifically, by examining the connections between judges and state and federal government officials, it will be possible to empirically evaluate the theoretical mechanisms for progressive ambition we posit in this article.

While state supreme courts are jurisprudentially distinct, with no real need to interact that is likely to be sufficient to outweigh the cost of building such relationships, the judicial Twitterverse has reduced these barriers. Judges in this network follow, like, and share each other, but they do so in predictable ways. And while most judges have joined social media for electoral benefits, those who are most important to the network have pioneered how the format can be used for increasing collegiality and leadership, further making the use of Twitter by judges distinct from the way other politicians behave on the format.