Computational analysis of 140 years of US political speeches reveals more positive but increasingly polarized framing of immigration

Abstract

We classify and analyze 200,000 US congressional speeches and 5,000 presidential communications related to immigration from 1880 to the present. Despite the salience of antiimmigration rhetoric today, we find that political speech about immigration is now much more positive on average than in the past, with the shift largely taking place between World War II and the passage of the Immigration and Nationality Act in 1965. However, since the late 1970s, political parties have become increasingly polarized in their expressed attitudes toward immigration, such that Republican speeches today are as negative as the average congressional speech was in the 1920s, an era of strict immigration quotas. Using an approach based on contextual embeddings of text, we find that modern Republicans are significantly more likely to use language that is suggestive of metaphors long associated with immigration, such as “animals” and “cargo,” and make greater use of frames like “crime” and “legality.” The tone of speeches also differs strongly based on which nationalities are mentioned, with a striking similarity between how Mexican immigrants are framed today and how Chinese immigrants were framed during the era of Chinese exclusion in the late 19th century. Overall, despite more favorable attitudes toward immigrants and the formal elimination of race-based restrictions, nationality is still a major factor in how immigrants are spoken of in Congress.

Computational Assessment of Hyperpartisanship in News Titles

“…We use mainstream media news as our study sample. It is worth noting that a similar divergence between different politically leaning entities is also identified in studies that are based on social media data (Waller and Anderson 2021) or congressional and presidential speeches (Card et al 2022). Moreover, we will make our new dataset available.…”

“…Pretrained models may introduce unknown bias from the pretraining data (Card et al 2022). To verify that our results (Figure 2 in the main paper) are not significantly influenced by BERT-base, we repeat the hyperpartisanship inference for all the 1.8 million titles using XGBoost classifiers with bagof-words features and replicate Figure 2.…”

“…Similar to Card et al (2022), we train logistic regression models to estimate the predictions of our BERT-base models. The feature importance based on Shapley values (Lundberg and Lee 2017) is used to better understand the language that is suggestive of hyperpartisan or non-hyperpartisan titles.…”

Networks of Beliefs: An Integrative Theory of Individual- and Social-Level Belief Dynamics

“…There are several different ways to measure or estimate the core constructs and predictions specified in the NB theory. First, the state of nodes in belief networks can be measured through survey questionnaires (Dalege et al, 2016;van der Does et al, 2022), by inferences from behavior (e.g., from voting records), or by inferring networks of related beliefs from textual corpora such as books, newspapers, congressional speeches, and social media (Bhatia & Bhatia, 2021;Card et al, 2022;Charlesworth & Banaji, 2019). Edges between belief nodes can be estimated using partial correlations or regression weights based on co-occurrences of beliefs within or between participants at a single time point or over time or by other methods for fitting parameters of network models .…”

Seeded Topic Models in Digital Archives: Analyzing the Swedish Understanding of Immigration, 1945--2019

“…In extension to earlier computational studies of archival text that described national conversations based on sets of political speeches (Rule et al 2015, Barron et al 2018, Fuhse et al 2020, Card et al 2022, the extreme breadth of the newspaper archive (106,000 daily issues in total) permits us to focus on the national conversation about immigration with high granularity. In relation to the newspaper corpora studied in prior immigrationrelated research, our data is vast.…”