This article examines how terrorist attacks affect support for Islamic State (IS) online, analyzing 300,842 observations from 13,321 Twitter accounts during a 127-day period.

Surprise: Our findings show that after major attacks, the number of followers for IS-related accounts actually decreases significantly. We investigate two competing mechanisms driving this outcome:

🔍 Data & Methods 🔍

We use unique Twitter data collected over time to track follower changes following specific attack events, treating these as natural experiments.

💡 Key Findings 💡

• Number of followers drops for IS-linked accounts post-attack
• Evidence points to disengagement rather than deterrence as the primary mechanism
• Followers appear less likely and not more inclined to join despite attacks

🔎 Why It Matters 🔎

This study challenges assumptions about terrorism by showing that violent acts may actually repel potential supporters, offering insights into group dynamics under crisis.