Research

Research lines

1. Protest and collective action
   1. Individual behavior
   2. Elite emotions and catalan independence
   3. Identity formation and media claims
2. Public support for protest
3. Media representations of protest
4. Influence and response of political elites to interest groups
5. Public perceptions of interest representation
6. Interest groups in parliament
7. Agenda dynamics and political representation
8. Presidential agendas in latin america
9. Methodological innovation

Protest and collective action

Individual behavior

Research on individual participation in protests, demonstrations, and the role of social media in mobilization and inequality.

Methodological Strategies:

• News Coverage Experiments
  With Ruud Wouters (2021): Two conjoint experiments with journalists to examine how protest attributes affect media coverage.

• Elite Reactions & Media Frames
  With Silvia Majò and Eva Anduiza: Uses Twitter data to study elite-public interactions during anti-austerity and Catalan nationalist protests.

• Spatial Inequalities in Protest
  With Sheila González: Combines survey data and a geo-located protest dataset in Barcelona to study protest supply.

• 15M/Indignados Movement
  With Eva Anduiza: Surveys and text analysis to study connective action and participant profiles in the Spanish anti-austerity movement.

Elite emotions and catalan independence

Twitter-based analysis of how political elite claims shaped emotional responses during the Catalan independence and counter-movements.

Collective identity in times of moral shocks

With Elena Pavan: Network science applied to demonstrator claims in pro-Palestinian protests across time and space.

Public support for protest

Combines survey data, experiments, LLMs, and text analysis:

• Early work: Twitter analyses of protest events in Spain.
• Comparative survey in 9 European countries during the 2010 economic crisis (with Katrin Uba and Lorenzo Zamponi).
• Latest research explores synthetic data with large language models (with Katrin Uba).

Media representations of protest

Focus on Latin America, using a dataset of 18,000 protest newswire stories (1999–2024):

• Analyzes public support for protest as the extent of the protest paradigm.
• Future work includes:
  1. Narrative framing through the lens of social psychology (e.g., system justification, majority status).
  2. Experimental designs testing individual perceptions of media portrayals of protest.
  3. Algorithmic bias on YouTube via bot-based user simulations.

Influence and response of political elites to interest groups

Rhetorical representation and influence of organized groups in the definition of public debate and in legislative activity

• Tracks rhetorical and agenda influence using datasets covering 2000–2020:
  • Press releases
  • Social media
  • Parliamentary transcripts

Published and in-progress work:

• Influence in emergencies
• Cultural outcomes of women’s orgs
• Narrative adoption by elites (under review)
• Methodological paper on rhetorical influence measurement

Public perceptions of interest representation

• Explores how lobbying success is tied to public approval (With Leire Rincón)
• Survey experiments with a nationally representative sample
• Studies effects of tactics, contact, and issue salience

Interest groups in parliament

• Network and text analysis of 6,500 parliamentary hearings (1996–2023) - With Luz Muñoz
• Maps interactions between 1,491 interest groups and 25 parliamentary parties in Spain
• Empirical approach to corporatism

Agenda dynamics and political representation

• Female MPs in parliamentary discourse
• Mixed-method study of gendered language and issue attention.
• Recent work explores:
  • Classifier performance across MP speeches (France, Italy, Spain)
  • Differences in agenda attention between social media vs parliament (With Laura Chaqués, Marcello Caramia, Federico Russo)

Presidential agendas in latin america

Original dataset of 25,000 everyday presidential speeches (1999-2023):

• Includes everyday speeches and own media
• Sentence-level classification of:
  • Policy issues
  • Ideological position (based on synthetic left/right dataset)

Methodological innovation

1. Frame identification
   • Supervised approach using 2M tweets/news items from 2011–2013 Indignados protests (with Silvia Majò-Vázquez)
2. Protest event extraction
   • Human-annotated Barcelona dataset, LLM-based event features extraction for Europe & Latin America
3. Rhetorical influence measurement
   • Compares similarity across tweets, hearings, press releases for 150 organizations
4. Tweet classification into policy issues
   • Supervised classifier using celebrity/government Twitter to identify topics; applied to Latin American elites
5. Cultural outcomes of movements
   • NLP on manifestos and narratives; published in Participation and Conflict (2021); DIPROMATS challenge (2024)
6. LLMs as survey surrogates
   • Evaluates attitudes toward protest via LLMs; under review in Sociological Methods & Research (2025, with Katrin Uba)
7. Agentic interest groups (simulations)
   • Generative AI in policy actor role-play for teaching (UB 2025); paper on teaching innovation in progress