Research

Working Papers

Brainrot? The Effect of Short-Form Content on Long-Form Attention (Latest Version)

  • Winner, Best Poster, RES 2026
  • Presented at Kings QPE 2026, Warwick PhD Conference 2026, RES 2026, KCL PPE Conference 2026

Does short-form content impact our attention spans? I study viewing patterns for a large set of news channels on YouTube, using attention heatmaps that record within-video click data. Leveraging the staggered rollout of YouTube Shorts across markets, I show that exposure to short-form content reduces continuous spells of attention to videos. Within-video skipping increases immediately and persists at about 40% above the pre-Shorts mean, with larger effects for longer videos. In turn, I document two important findings. First, the likely mechanism is reduced attention span, since I am able to rule out changing user composition and cross- video substitution as alternative explanations. This is supported by a model in which videos are bundles of interdependent slices where skipping trades off single- slice content against multi-slice context. The effect of short-form content then enters by lowering the value viewers place on context. Second, I show that news viewing patterns on YouTube have changed significantly, with viewers effectively reducing the amount of rich, informational context that they consume. Specifically, a structural model indicates that the value of context in news videos fell by at least a quarter with the introduction of Shorts.

Can Hate Speech Be Banned Online? The Effects of Shutting Down Toxic Forums on Reddit with Lily Shevchenko (Latest Version)

  • Presented at NICEP 2025, Warwick PhD Conference 2025, HEC PhD Conference 2025, KCL PPE Conference 2025, CODE @ MIT 2025, LSE-Warwick PolEcon PhD 2026, Doctorissimes 2026, AYEW Political Economy 2026, EAYE 2026, South West Economics PhD Conference 2026, POLECON UK 2026, CESifo Workshop on Digital Platforms 2026, RES 2026,

Is deplatforming effective in reducing toxicity on social media? To answer this question we study a policy change on Reddit in June 2020 which led to a simultaneous ban of thousands of forums containing hateful content, but not the users of these forums. We use data on the near universe of comments left on Reddit to examine the impact of the ban on user behaviour in a differences-in-differences design. We find that the most active users of banned subreddits comment more after the policy change and substitute to new forums in the weeks after the ban. The increase in activity persists in the long run, but is not associated with higher toxicity: instead, the comments left by affected users outside banned subreddits contain 20% fewer instances of hate speech. We do not find evidence that the policy leads to lower quality of engagement, negative spillover effects or recreation of banned subreddits elsewhere on the platform. Overall, the results suggest that moderation targeting toxic digital spaces can be effective in combating hate speech without lowering user engagement, and thus can be aligned with platforms’ incentives.

Social Influence in Online Reviews: Evidence from the Steam Store (Latest Version)

  • Winner of the Best Paper award of the Business and Industry Section of the Royal Statistical Society at NIE Conference 2024
  • Presented at CODE @ MIT 2024, RES 2025

This paper investigates how pre-existing consumer assessments bias online ratings by exploiting a natural experiment on the Steam gaming platform. In 2019, an update exogenously removed the average rating being shown to incoming reviewers by default. I show that exposure to average ratings significantly alters reviewing behavior, driven entirely by compounding negativity. Reviewers exposed to a lower average rating become significantly more negative, but do not respond to a higher rating. Using owner data, I price the policy effect: a one percentage point increase in rating is equivalent to a 2 dollar price reduction.