Polarization and the Dynamics of Media Trust in a Revitalizing Democracy

• In Event: The Drivers and Variations of Media Trust and Distrust

Abstract

In the fall 2023 parliamentary elections, Poland’s right wing populist governing party Law and Justice (PiS) lost to a coalition of more liberal, pro-EU parties led by Donald Tusk. The election represented a major turning point for Poland, given the evidence of democratic backsliding taking place under the PiS government. While in power, the PiS government had implemented policies aimed at further undermining the independence of public media. Public television channels TVP and TVP Info became effectively propaganda outlets for the party. During the 2023 campaign, Tusk promised that one of his coalition’s top priorities if it were elected would be to “return impartiality” to public television. Almost immediately after taking power in December 2023, the new Tusk-led government implemented policies aimed at transforming public television. In this paper, we examine first, how TVP was perceived by audiences prior to this transformation and the effect of exposure to its messaging on audience attitudes. Then we examine how audience perceptions and messaging effects change following the implementation of the new government’s reforms. The paper provides insight into media selection and media effects in the context of a deeply polarized society in which mass media has become a tool of the state. It also reveals how elections and policy changes can alter audience behavior and attitudes.
In December, 2023, in the midst of the government transition and the transformation of TVP, we conducted an online survey and experiment in Poland to measure audience media preferences. In this first study, we ask 1) to what extent do pro-PiS and anti-PiS citizens prefer the content of news outlets aligning with their political views? 2) what effect do source cues associated with partisan viewpoints have on audience trust and 3) what effect does exposure to state media have on audience attitudes related to outsider threat, political discord, and related second order beliefs. Experiment participants (final N = 2,849) evaluate a series of real news excerpts drawn from either TVP or TVN, Poland’s most prominent independent news channel. To create a representative sample of news content from a given source for the stimuli, we constructed excerpts of every politically relevant news story produced by a given source every other day for 10 days. Each participant was asked to separately evaluate eight excerpts. Respondents randomly assigned to the blind condition saw excerpts with no source information while respondents assigned to the revealed condition saw excerpts with their true source. Respondents were additionally assigned to one of three topic conditions: 1) news related to immigration, international organizations and globalization 2) news related to domestic politics and 3) all politically-relevant news. News articles shown to the respondent were randomly sampled from the larger pool of all excerpts within a topic. A pure control group read no articles. With this design, we aimed to maximize realism and were able to separately measure content and source preferences and the effects of exposure on attitudes. Our results from the December study reveal high levels of media polarization in Poland. We find evidence of large source cue effects (which exceed those found in similar studies in other contexts). We additionally show evidence of preferences for the content of aligned sources in the blind condition, depending on the topic. However, we do not find clear evidence of exposure effects on political attitudes.
In the spring, we will conduct a similarly-designed study to assess how perceptions of the media outlets have evolved since the implementation of the Tusk government’s reforms. We will be able to assess 1) whether media attitudes become less polarized 2) whether the reforms cause an increase in trust in TVP among anti-PiS citizens and 3) whether the reforms cause a decrease in trust in TVP among pro-PiS citizens. Importantly, our design allows us to separate changing perceptions of the TVP brand and changing perceptions of its content. This project seeks to shed light on the dynamics of media perception and media effects during periods characterized by democratic retrenchment and renovation.

Authors

• Ashley Blum, Massachusetts Institute of Technology

• Adam J. Berinsky

• David Rand, Massachusetts Institute of Technology

• Gabriela Czarnek