Polarized Online Discourse on Abortion: Frames and Hostile Expressions among Liberals and Conservatives

TL;DR

The paper analyzes national-scale Twitter discourse on abortion from January 2022 to January 2023, examining hostility (anger, toxicity, obscenity, insults, hate speech) and five frames (Religion, Bodily Autonomy, Fetal Rights, Women's Health, Exceptions) across liberals and conservatives around four pivotal events, including the Dobbs decision. Using a two-stage approach to infer ideology and state-of-the-art classifiers for hostility and framing, it conducts an interrupted time series to quantify immediate and trend-level shifts in response to events. The key findings show symmetric increases in hostility across both groups after major events, but substantial ideological differences in framing and the contexts in which frames are used; importantly, frames favored by one side provoke hostility from the other, underscoring deep affective polarization. The work contributes a large-scale, longitudinal portrait of US abortion discourse, demonstrates methodological integration of ideology inference, frame analysis, and ITS, and discusses implications for democratic discourse and policy debates in digital environments.

Abstract

Abortion has been one of the most divisive issues in the United States. Yet, missing is comprehensive longitudinal evidence on how political divides on abortion are reflected in public discourse over time, on a national scale, and in response to key events before and after the overturn of Roe v Wade. We analyze a corpus of over 3.5M tweets related to abortion over the span of one year (January 2022 to January 2023) from over 1.1M users. We estimate users' ideology and rely on state-of-the-art transformer-based classifiers to identify expressions of hostility and extract five prominent frames surrounding abortion. We use those data to examine (a) how prevalent were expressions of hostility (i.e., anger, toxic speech, insults, obscenities, and hate speech), (b) what frames liberals and conservatives used to articulate their positions on abortion, and (c) the prevalence of hostile expressions in liberals and conservative discussions of these frames. We show that liberals and conservatives largely mirrored each other's use of hostile expressions: as liberals used more hostile rhetoric, so did conservatives, especially in response to key events. In addition, the two groups used distinct frames and discussed them in vastly distinct contexts, suggesting that liberals and conservatives have differing perspectives on abortion. Lastly, frames favored by one side provoked hostile reactions from the other: liberals use more hostile expressions when addressing religion, fetal personhood, and exceptions to abortion bans, whereas conservatives use more hostile language when addressing bodily autonomy and women's health. This signals disrespect and derogation, which may further preclude understanding and exacerbate polarization.

Figures (12)

• Figure 1: Use of hostility and frames by ideology (a)Boxplots show the distribution of the daily share of original tweets using hostile expressions for liberals and conservatives. (b)Boxplots compare the distribution of the daily share of original tweets employing different frames by liberals and conservatives. * indicates significance at $p<0.05$, ** - $p<0.01$, *** - $p<0.001$ and, **** - $p<0.0001$ (Mann-Whitney U Test with Bonferroni correction).
• Figure 2: (I) Daily number of tweets by user ideology. (II) The fraction of liberal and conservative original tweets expressing (a) anger, (b) toxicity, (c) obscenities, (d) insults, and (e) hate speech per day. (III) Use of frames over time. Daily share of tweets about (a) religion, (b) fetal rights, (c) exceptions, (d) bodily autonomy, and (e) women's health by liberals and conservatives. Vertical lines indicate major events: the Supreme Court leak (May 3, 2022), the official Dobbs verdict (Jun. 24, 2022), the Kansas referendum (Aug. 2, 2022), and the midterm elections (Nov. 8, 2022).
• Figure 3: (a)-(d) Heatmaps showing the immediate change in anger, hateful and toxic speech, insults and obscenities for liberals and conservatives. (e)-(h) Heatmaps showing the immediate change in the use of five frames: religion, fetal rights, exceptions, bodily autonomy and women's health, among liberals and conservatives. Change is quantified using Interrupted Time Series Analysis. The values in the heatmap are coefficient of the treatment variable (i.e., share of tweets containing a particular hostile expression or frame) for liberals and conservatives. * indicates significance at $p<0.05$, ** - $p<0.01$, *** - $p<0.001$, **** - $p<0.0001$.
• Figure 4: Word clouds highlight the semantic contexts in which (a) Religion, (b) Fetal Rights, (c) Exceptions, (d) Bodily Autonomy and (e) Women's Rights frames are discussed by liberals and conservatives. The intensity of color denotes which group is more likely to use a phrase (shades of red for conservatives and shades of blue for liberals) and text size denotes the likelihood of usage.
• Figure 5: Dot and whisker plots compare the prevalence of (a) anger, (b) toxicity, (c) obscenities, (d) insults and (e) hate speech in the daily tweets of liberals (blue) and conservatives (red) when addressing different issues such as religion, fetal rights, exceptions, bodily autonomy, and women's rights. The circles represent the mean daily share of the hostile expressions in tweets and the whiskers show the standard deviation. * indicates significance at $p<0.05$, ** - $p<0.01$, *** - $p<0.001$ and, **** - $p<0.0001$ (Mann-Whitney U Test with Bonferroni correction).