Issue Attention, Media Narratives, and Immigration Shocks: An LLM Classification Approach
Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
How do media narratives shift in response to immigration shocks? I develop a dynamic model of political communications that predicts more pro-immigrant coverage soon after immigration shocks, which I call the humanitarian phase due to prominent themes of sympathy and vulnerability. After a period of time, I predict more anti-immigrant coverage—which I call the threat phase due to themes of material and symbolic threat—will become more popular as the media reflects changing sentiments and conditions on the ground. I test my theory using the cases of Colombia, Germany, and Poland in response to their respective immigration shocks. Empirically, I use a few-shot ChatGPT 4o model to classify articles from national newspapers in each of these countries and then illustrate changes in the proportion of immigration-focused media narratives over time. I find evidence for humanitarian phases across these contexts, inundating the information environment with sympathetic immigrant narratives. I also find long-term secular trends toward threat coverage. These findings systematize elite discourse and its changing nature on salient issues, contributing to our understandings of the relationship between the media, its responses to shocks, and its influence on democratic societies.