Oswald, L., Schulz, W., Hertwig, R., Lazer, D., & Stier, S. (2025). The tip of the iceberg: How the social media production-consumption gap distorts public opinion for citizens and researchers. SocArXiv. https://doi.org/10.31235/osf.io/frcv5_v2
Summary
This conceptual synthesis examines the production-consumption gap on social media: a small minority of highly active users generates the bulk of political content, while the majority consume passively and remain silent. The authors argue that this skew makes visible online discourse the “tip of an iceberg” that obscures the far larger, quieter public beneath. The distortion has consequences for two audiences at once — ordinary citizens who form impressions of collective opinion from what they see online, and researchers who treat social media traces as proxies for public attitudes. In both cases, inferences drawn from observable content risk systematically misrepresenting what the broader public actually thinks.
Key Contributions
- Articulates the “tip of the iceberg” framing for how social media distorts perceptions of public opinion.
- Synthesizes evidence that the production-consumption gap is a consistent empirical regularity across platforms, time, and cultures.
- Highlights the dual risk of misinference — for lay citizens gauging public opinion and for researchers using social media as a data source.
Methods
A conceptual and synthesis piece rather than a primary empirical study. It draws together prior findings on user activity distributions across social media platforms to establish the robustness of the production-consumption asymmetry. Specific analytic methods are not detailed in the available abstract.
Findings
- A small minority of users produces the majority of political content, a pattern robust across platforms, time periods, and cultural contexts.
- Most users consume passively and remain silent, so observable discourse is unrepresentative of broader opinion.
- Visible online political content systematically misrepresents the views of the silent majority, biasing inferences by both citizens and researchers.
Connections
This paper speaks directly to methodological debates about whether social media traces can stand in for public attitudes, connecting to work on measuring opinion and polarization from platform data such as Le-Mens2025-qz and Green2025-ap, and to critical assessments of the representativeness and construction of social media data as an object of study like Vertesi2026-lv and Munger2025-cz. Its concern with how skewed visible discourse shapes perceptions of collective opinion also links to accounts of engagement asymmetries and amplification dynamics such as Bak-Coleman2026-mk.