Flatten the curve: Effects of geotagging on user engagement in Weibo hashtags – Song Yuan

Sprecher

Song Yuan (Zhejiang University)

Abstract

This paper evaluates the impact of mandatory geotagging on user engagement, exploiting a platform-wide policy on Sina Weibo in April 2022 that discloses users’ IP-based locations. Using a novel dataset of top-trending hashtags, we examine how geotagging affects both passive exposure (reads) and active participation (interactions). A quasi-difference-in-differences design, using reads as a control for interactions, shows a significant decline in engagement, driven primarily by reduced user interactions. To address feedback between reads and interactions, we develop and estimate a structural SIR-style diffusion model that disentangles interaction probability from amplification effects. We find that geotagging substantially lowers users’ likelihood of interacting after exposure, while leaving reading dynamics largely unchanged, resulting in flatter engagement trajectories. Effects are heterogeneous across topics, with stronger declines for salient or sensitive content. These findings highlight a key tradeoff: while geotagging may enhance accountability and curb harmful content, it also reduces participation and limits information diffusion in online platforms.

Autoren

Ao Wang (University of Warwick) — Jianghao Wang (IGSNRR, Chinese Academy of Sciences) — Song Yuan (Zhejiang University) 

Raum

Media Room (A-211)