Accounting for Numbers: Group Characteristics and the Choice of Violent and Nonviolent Tactics

The research argues that nonviolent tactics are relatively more effective only when a movement can mobilize a sufficiently large number of participants, creating a 'mobilization advantage'. Conversely, movements with limited audiences or resources—such as ethnic secessionists or those based in rural peripheries—often find violent tactics more feasible. Using data from NAVCO 2.0, UCDP, and SRDP, the authors demonstrate that nonviolent campaigns typically have much higher participation rates and are more frequent in urbanized non-democracies, while violent conflicts are more common in rural areas and among smaller, territorially focused groups.

• Nonviolent tactics provide a strategic advantage only when a movement has a large potential audience and sufficient resources to mobilize a high number of active participants, whereas movements with limited mobilization potential may find violence more feasible.
• There is a stark difference in the scale of active participation between the two strategies: nonviolent campaigns have a median participation rate of 100,000 people, while the median rebel troop estimate for violent armed conflicts is only 4,000.
• The geographic and demographic profiles of the two types of dissent differ significantly; nonviolent mobilization is most frequent in capital cities and urbanized areas, while violent mobilization is typically confined to the periphery.
• Empirical analysis shows that nonviolent campaigns are more likely to occur in non-democracies with large urban populations, whereas the likelihood of violent civil conflict is highest in rural non-democracies.
• The type of incompatibility influences the choice of tactic: governmental incompatibilities (e.g., regime change) are more likely to result in nonviolent mobilization, while territorial incompatibilities (e.g., secession) are more likely to lead to violence.
• For ethnic self-determination groups, the likelihood of employing nonviolent strategies increases more rapidly with the size of the ethnic group than the likelihood of using violence.
• Initial strategic choices are generally stable, with fewer than 10% of campaigns in independent states changing their primary strategy.

APA

Dahl, M., Gates, S., Gleditsch, K. S., & González, B. (n.d.). Accounting for Numbers: Group Characteristics and the Choice of Violent and Nonviolent Tactics. Peace Research Institute Oslo. https://cdn.cloud.prio.org/files/4278c772-992b-491e-87d3-304ceaf190fe/Accounting for numbers Group characteristics and the choice of violent and nonviolent tactics.pdf?inline=true

Chicago

Dahl, Marianne, Scott Gates, Kristian Skrede Gleditsch, and Belén González. Accounting for Numbers: Group Characteristics and the Choice of Violent and Nonviolent Tactics. Peace Research Institute Oslo, n.d. https://cdn.cloud.prio.org/files/4278c772-992b-491e-87d3-304ceaf190fe/Accounting for numbers Group characteristics and the choice of violent and nonviolent tactics.pdf?inline=true.

Wikipedia

{{cite report |last1=Dahl |first1=Marianne |last2=Gates |first2=Scott |last3=Gleditsch |first3=Kristian Skrede |last4=González |first4=Belén |title=Accounting for Numbers: Group Characteristics and the Choice of Violent and Nonviolent Tactics |publisher=Peace Research Institute Oslo |url=https://cdn.cloud.prio.org/files/4278c772-992b-491e-87d3-304ceaf190fe/Accounting for numbers Group characteristics and the choice of violent and nonviolent tactics.pdf?inline=true |access-date=23 October 2026 |via=Climate Insights Directory}}

BibTeX

@techreport{dahlndaccounting, author = {Dahl, Marianne and Gates, Scott and Gleditsch, Kristian Skrede and González, Belén}, title = {{Accounting for Numbers: Group Characteristics and the Choice of Violent and Nonviolent Tactics}}, institution = {Peace Research Institute Oslo}, url = {https://cdn.cloud.prio.org/files/4278c772-992b-491e-87d3-304ceaf190fe/Accounting for numbers Group characteristics and the choice of violent and nonviolent tactics.pdf?inline=true}, urldate = {2026-10-23}, note = {Indexed by Climate Insights Directory} }