What Congress Talks About When They’re Not Talking Policy

• In Event: Elites, Media, and Public Opinion

Abstract

Decades of work shows that US Members of Congress’ (MCs) individual characteristics influence their public communications. Scholars find that gender shapes the diversity of policy-related topics that MCs discuss in public, most recently on Twitter (Atkinson and Windett 2019; Evans 2016; Huddy and Terkildsen 1993). Different political parties use Twitter to communicate differently (Golbeck et al. 2018; Hemphill, Otterbacher, and Shapiro 2013; Russell 2018), and tend to discuss different policy-related topics (Egan 2013). An MC’s chamber also shapes MCs' behavior on Twitter (Hemphill, Otterbacher, and Shapiro 2013) and the policy topics they address (Hemphill, Russell, and Schöpke‐Gonzalez 2020).

To date, we know little about how individual MCs' characteristics influence how they talk about topics other than policy. However, these types of communication are quite common and make up as much as half of Congress’ tweets (Hemphill, Russell, and Schöpke‐Gonzalez 2020). Studying what MCs talk about besides policy is important because these discussions too have effects on democratic processes–e.g., how constituents access government services, how emergency resources are allocated, how politicians and constituents build and maintain connections.

We investigate how individual characteristics shape MCs’ not-policy-related behavior using statistical analyses of all tweets made by the 115th US Congress. First, we identified topics in each of the 1,447,180 tweets using an unsupervised topic model (Hemphill and Schöpke-Gonzalez 2020). Then, using mixed-effects logistic regression (MELR), we learn that party, chamber, and gender do impact the likelihood of a tweet by an MC being about policy or not. To explain why we see these effects, we use multinomial logistic regression (MLR) to assess the predictive effects of party, chamber, and gender on each most probable not-policy topic - Veterans, Sports, District Affairs, Holidays, Awards, Politicking, Self Promotion, Sympathy, Emergency Response, Legislative Process, Constituent Relations, and Power Relations. This model shows that women MCs are more likely to tweet about Sports, Holidays, and Emergency Response topics than men MCs - the opposite of what our MELR model led us to expect. To understand what might explain this unexpected pattern, we do an event history (EH) analysis of a time period where the Emergency Response topic saw a lot of tweets: during Hurricane Harvey. Our analysis shows that the only independent variable assessed whose effect is statistically significant over the course of the Hurricane is geographic proximity to where the hurricane landed, not gender as we would have expected.

Policy-related communications research finds that party, gender, and chamber often explain patterns that we see in MCs' communications. Our research shows that MCs' communications about topics other than policy cannot be explained by these characteristics alone. For instance, in the case of Emergency Response, geography plays a significant role. Our findings complicate assumptions about how and why MCs communicate. We argue that explaining MCs’ not policy discussions and their implications for democratic processes require us to consider additional characteristics.

Atkinson, Mary Layton, and Jason Harold Windett. 2019. “Gender Stereotypes and the Policy Priorities of Women in Congress.” Political Behavior 41 (3): 769–89.
Egan, Patrick J. 2013. Partisan Priorities: How Issue Ownership Drives and Distorts American Politics. Cambridge University Press.
Evans, Heather. 2016. “Do Women Only Talk about ‘female Issues’? Gender and Issue Discussion on Twitter.” Online Information Review 40 (5): 660–72.
Golbeck, Jennifer, Brooke Auxier, Abigail Bickford, Lautaro Cabrera, Meaghan Conte McHugh, Stephani Moore, Jacquelyn Hart, Justin Resti, Anthony Rogers, and Jenna Zimmerman. 2018. “Congressional Twitter Use Revisited on the Platform’s 10-Year Anniversary.” Journal of the Association for Information Science and Technology 69 (8): 1067–70.
Hemphill, Libby, Jahna Otterbacher, and Matthew Shapiro. 2013. “What’s Congress Doing on Twitter?” In Proceedings of the 2013 Conference on Computer Supported Cooperative Work, 877–86. CSCW ’13. New York, NY, USA: Association for Computing Machinery.
Hemphill, Libby, Annelise Russell, and Angela M. Schöpke‐Gonzalez. 2020. “What Drives U.S. Congressional Members’ Policy Attention on Twitter?” Policy & Internet 41 (June): 769.
Hemphill, Libby, and Angela M. Schöpke-Gonzalez. 2020. “Two Computational Models for Analyzing Political Attention in Social Media.” In Proceedings of the International AAAI Conference on Web and Social Media, 14:260–71.
Huddy, Leonie, and Nayda Terkildsen. 1993. “Gender Stereotypes and the Perception of Male and Female Candidates.” American Journal of Political Science. https://doi.org/10.2307/2111526.
Russell, Annelise. 2018. “The Politics of Prioritization: Senators’ Attention in 140 Characters.” In The Forum, 16:331–56. De Gruyter.

Authors

• Angela M. Schöpke-Gonzalez, University of Michigan

• Libby Hemphill, University of Michigan