Tweeting in echo chambers? Analyzing Twitter discourse between American Jewish interest groups

ABSTRACT

Recent research of interest groups’ use of Twitter focuses on its use for intra-group member mobilization. Less has been said about social media’s effects on the discourse within an issue community. Do interest groups that are part of a larger community, but whose positions divert from each other, use Twitter to engage each other? Our research explores whether groups within a defined issue community – American-Jewish organizations advocating on U.S.-Israel relations – use Twitter for dialog, or whether their Twitter networks represent echo chambers. Based on the results of our social network analysis we find the use of Twitter is more complex than the black-and-white narrative of echo chamber vs. public sphere. We find groups interact with both supporters and detractors, using Twitter’s communication tools (we focus here on @mentions) strategically – but on balance the results of our analysis support a “sphericules” interpretation of social media.

KEYWORDS:

• social media
• interst groups
• social network analysis
• public sphere

Supplemental data

Supplemental data for this article can be accessed on the publisher’s website.

Notes

1. For framing process more generally see prospect theory: (Kahneman & Tversky, 2013; Quattrone & Tversky, 1988).

2. Common social media platforms include Twitter, Facebook and more recently Instagram. A substantial amount of research has focused on Twitter because of its public character and availability for research. We focus on Twitter because of the accessibility of its API as well as its relevance to contemporary politics. For example, journalists are increasingly reliant on Twitter for information from the office of the President as the office has moved away from formal press conferences.

3. Democratizing of interest group access to the polity may be wishful thinking as scholars have found positive correlations between interest group resources and social media use. See (Goldkind & McNutt, 2014; Hong & Nadler, 2016; Van der Graaf et al., 2016).

4. Direct messages would be the strongest form of discourse, but such messages are private between users, while we are interested in the larger, public group discourse. Further they cannot be scraped and the majority of the organizations under study place limitations on the capacity for users or followers to send direct messages.

5. The most common form of connection on Twitter is via the follower-followee relationship Bruns and Moe (2014) refer to as the meso-level of Twitter communication. This is the fundamental level of connection because information in the default feed consists of tweets from those entities one is following. While all tweets from non-private accounts are public, the primary intended audience of a user is generally understood to be their followers.

6. see Rubenzer (2016) for an exception.

7. The assertion that there is one shared pool of potential members of all organizations might be seen as problematic, considering the wide range of opinions within the American Jewry and the fragmented nature of the interest community around U.S.-Israel relations. The American Jewish demographic has increasingly been divided into categories based on religious denominations, primarily between Orthodox and non-Orthodox (Reform, Conservative, nonaffiliated) Jews: Some political issues, for example the question of Jewish settlements in the West Bank, divide American Jews along denominational lines, with Orthodox Jews overwhelmingly supporting settlement expansion (Pew Research Center, 2013, in Waxman, 2016). But despite the heterogeneity of opinions, there remains a lot of overlap, as the American Jewish Committee Survey of American Jewish Opinion shows. A survey from August 2016 asked 1,002 respondents about their political and religious opinions. 73% agreed with the statement that caring about Israel is an important part of my being Jewish This supports our claim that interest groups addressing U.S. policy toward Israel will draw support from the Jewish community in the United States – how individuals define caring for Israel,” and what they believe is the correct” U.S. policy toward the state of Israel then determines which interest group they will join. Within the larger constituency, the different interest groups attract different segments of the American Jewry, based on their political leanings in regard to Israel-U.S. relations.

8. Supporters of BDS want to end the occupation of Palestinian territories by putting economic pressure on Israel. They point to the anti-apartheid movement in South Africa as a historical precedent (e.g. Di Stefano & Henaway, 2014).

9. For a more detailed description of the example cases, please refer to the online appendix. The data set is available on the authors’ websites.

10. www.conferenceofpresidents.org/about/members.

11. The 2014 edition was the most recent electronically accessible version of the yearbook.

12. A growing literature exists examining the differences between Twitter scraping software with differing levels of access via the API versus firehose” access via corporations such as GNIP (eg Driscoll & Walker, 2014; Mazon, Morer, Um Amel, & Lotan, 2012). This literature focuses on the streaming sprinkler” versus firehose” access, both of which scrapes tweets in real time. We are interested in a small community and sought to include historical tweets rather than only tweets in real time.

13. Domestic instances of anti-Semitism are included. The issue of anti-Semitism, and protection from it, is deeply intertwined with Israel’s identity as a state and often brought up in discussions of the threat of BDS.

14. Coding agreement = 94%; Krippendorf’s α = 0.918.

15. Fruchterman-Reingold is a force-directed layout algorithm that bases nodes’ proximity to one another on minimizing the energy” equilibrium if ties between nodes are treated as springs – relocating any organization from their current community to another would increase the energy level.

16. Formally, reciprocity occurs in a network when an edge exists directed from node A to node B and an edge exists directed from node B to A.

17. In-degrees are the number of links to the organization from other groups; out-degrees the connections to other groups originating from this node.

18. 131 of these tweets are from Open Hillel and directed toward Hillel.

Additional information

Notes on contributors

Megan Osterbur

Megan Osterbur is an independent scholar based in Austin, Texas. Her research focuses on LGBTQ social movements and public policy relevant to gender and sexuality movements.

Christina Kiel

Christina Kiel is Professor of Practice at Tulane University in New Orleans ([email protected]). She received her Ph.D. in political science at the University of New Orleans. Her research focus is international conflict studies, mediation, and transnational advocacy movements.