Summary

During political contention, states and social media platforms often reach for measures that sit outside their everyday moderation rules: blocking orders, throttling, registration pressure, emergency takedowns. My research asks why. Why do states and platforms implement exceptional content moderation measures during political contention? To answer this, I map the moderation stack (the operational, institutional, and infrastructural layers through which online content can be governed) and compare moderation as codified in national regulations and platform policies with moderation as applied during contentious episodes.

The pilot study annotated content moderation policy documents from nine countries and the historical community guidelines of three platforms, each with more than a billion users. It also built a dataset that joins records of contention episodes with records of tensions between states and platforms. The thesis adds depth through a case study of Indonesia, drawing on elite interviews, to examine how control over the moderation stack shapes what is decided during political contention.

Findings to date

The pilot study reached two working conclusions, both provisional pending its full analysis:

  1. Moderation instruments follow the ordering of the stack. Each layer is traditionally controlled by states or by platforms, but rarely by both at the same layer.
  2. During political contention, both states and platforms tend to reach for exceptional instruments, formal or discretionary, that target something other than the content itself.

From these, the thesis takes its working hypothesis: beyond the characteristics of the contention or the content, states and platforms choose an exceptional measure because that measure is already in their control.

Further details