Digital Incivility in Political Deliberation: The Case of the Chilean Constitutional Convention

In the framework of the Digital Democracy Workshop 2024, held at the University of Zurich, two associate researchers from the Training Data Lab, Bastián González-Bustamante and Sebastián Rivera, presented a relevant study on digital incivility in political deliberation. In their paper entitled, “Toxicity and Digital Incivility during Political Deliberation: The Case of the Constitutional Convention in Chile”, these researchers explore the magnitude and determinants of non-civil comments directed towards politicians, focusing on the Chilean context.

The research highlights that despite the potential of social networks and the internet to enhance and broaden political participation, phenomena such as toxicity and incivility have emerged as prominent features of the digital environment. These manifestations are often associated with behaviours such as the dissemination of hate speech or harassment, which ultimately contribute to the decline of public debate.

To address this issue, the researchers used a groundbreaking dataset of more than nine million interactions between users and members of the Constituent Convention from January 2021 to December 2022. By applying the Google Perspective API, they generated indicators of toxicity and insults to assess the extent of these behaviours in the digital environment. To validate these indicators, they performed classification tasks on a subsample using open-access Large Language Models (LLMs) such as Llama and other Generative Pre-trained Transformers (GPTs), supplemented with human annotations to create a gold standard.

The researchers’ analyses, based on negative binomial regression models (NBRMs), revealed that being male and belonging to an indigenous group increases the likelihood of receiving toxic messages. At the same time, ideological extremism has a moderating effect. This study contributes to existing discussions in two directions: it offers insights into the dynamics of toxicity and incivility in political deliberation and establishes a benchmark for algorithmic and LLM-based classification of incivility in Spanish.

In this context, the Digital Democracy Workshop 2024 becomes a relevant space to discuss and analyse the challenges of the digital environment in political deliberation. This workshop is co-organised by the Democracy Community of the Digital Society Initiative and the Digital Democracy Lab at the University of Zurich, with the aim of facilitating the exchange between scholars from different thematic areas working on these issues.

The presentation of Bastián González-Bustamante and Sebastián Rivera’s work highlights the importance of addressing digital incivility as a relevant concern in political deliberation. Their research offers a significant case study to understand how this issue manifests itself in specific environments, such as the Chilean context. At the same time, their findings contribute to deepening our discussions on possible strategies to promote respectful political participation and avoid the decline of public debate in digital spheres.

In sum, Bastián González-Bustamante and Sebastián Rivera’s research highlights the need to address digital incivility as a relevant challenge to promote healthy political deliberation. Their study offers valuable insights into how the dynamics of toxicity and incivility can affect the quality of public debate in digital environments and highlights the importance of establishing algorithmic and LLM-based benchmarks to identify and mitigate these behaviours.

* AI-generated text
Read more about how we generate our content