In recent years, the world has experienced an unprecedented number of mass protest events. Yet despite valuable research endeavours, one area that has not been systematically addressed is the conditions under which such protests turn violent. This is a significant limitation, as violent protests have serious detrimental effects: They harm people as well as infrastructure, erode trust between citizens and government, and polarize societies. To address this shortcoming, the Dynamics of Mass Protest (DMP) project will systematically study the heterogeneous and dynamic nature of protests.
DFG, 2024-2026

Kooperationspartner

Universität Konstanz German Institute for Global and Area Studies (GIGA)

Leitung

Prof. Dr. Christian von Soest

Lead Research Fellow / Head of Policy Exchange and Berlin Office

Dr. Belén González

Universität Zürich

Prof. Dr. Nils B. Weidmann

Universität Konstanz

Forschungsfragen

The project’s main objective is to answer: Why do some protests turn violent while others do not?

Beitrag zu internationaler Forschung

Despite the significant progress made in the expanding research on protest, two strands of literature - on political mobilization and violence escalation on the one hand and state repression and violence escalation on the other - predominantly focus on explaining the incidence and success rate of protest events. As a result, less is known on a systematic level about how within-protest features and the fluid nature of protest events affect violent outcomes.
The DMP project therefore sets out to tackle three main research gaps in current assessments of violent and non-violent protest:
1. Predominant focus on structural factors
2. Conception of protest as unitary events
3. Limited coverage of disaggregated cases
We offer three contributions to existing research: We will focus on protest-specific features, on the dynamics and heterogeneity of protest features, and on complementary quantitative and qualitative data.

Forschungsdesign und Methoden

The DMP project uses a multidisciplinary methodology; we collect and analyse original visual data using state-of-the-art computer vision tools and complement this with qualitative case studies.
To understand the conditions under which protest events turn violent, we collect information on within-protest features (composition, demands, and tactics) at the event level. To collect and code data for all protest events within identified protest campaigns, we draw on visual sources. Images are a unique data source for protest analysis since they provide us with information that is typically not contained in textual sources such as news reports. We will complement the statistical visual analysis with qualitative case studies to investigate how particular protest events turn violent.