Platform Politics: How Social Media Platforms Shape Political Discourse
In today’s media landscape, platforms such as Twitter (now X), Instagram, TikTok and Meta Platforms’s properties are more than just social tools. They are arenas where political ideas are expressed, contested, amplified or suppressed. The concept of “platform politics” refers to the ways in which these social media platforms influence the political discourse: through algorithms that determine what becomes visible, through moderation policies that decide what gets removed or suppressed, and through manipulation of discourse by organised actors.
Question 1: How do platform algorithms decide which political content becomes visible or “trending,” and what are the
Algorithms on social media serve as gatekeepers: what is shown, prioritized, or hidden from a user’s feed is determined by algorithmic curation (rather than purely chronological or editorial judgment). This means that visibility itself is a form of power.
For example, a major study on Twitter’s home timeline found that in six out of seven countries studied, content associated with mainstream right-leaning political parties enjoyed higher algorithmic amplification than those from the mainstream left. The study measured “linger impressions” (when a tweet is visible onscreen for at least 0.5 seconds) and found the right-leaning accounts had major gains from the personalized timeline compared to a reverse-chronological control.
Furthermore, research finds that algorithmic systems may also amplify low-credibility content, especially where the system is optimized for engagement rather than truth or deliberation. For example, in a 2024 observational study of Twitter, low-credibility content was found to receive amplification, and amplification was especially strong for right-leaning false claims on climate change.
Consequences of algorithmic visibility include:
Certain political voices or ideologies gaining disproportionate reach, not because of quality, but because of algorithmic design.
Less exposure for minority or alternative viewpoints, reinforcing echo-chambers and limiting cross-ideological exposure.
A shift in what “counts” as trending news: content that triggers engagement (emotion, outrage) gets elevated.
Democratic implications: what we see (and thereby what we talk about) is partly determined by invisible algorithmic decisions, which shape public discourse and political mobilisation.
Question 2: What responsibilities do platforms have in moderating political content, and what tensions arise from these responsibilities?
From an ethical standpoint, one researcher argues that platforms have a moral duty to moderate harmful speech: they should not enable or amplify wrongful speech by providing a venue for it or by boosting it. Michigan Publishing At the same time, the exercise of moderation power raises questions about free speech, bias, accountability, and transparency.
For example, content moderation can distort online discourse: a 2024 paper found that removing toxic tweets changed the “topic composition” of what remained visible, shifting means and variances of discourse (i.e., moderation is not simply removal of harmful content, but alters the shape of conversation).
Question 3: In what ways do organised actors (states, campaigns, bots) manipulate discourse on platforms, and how does platform design enable or inhibit this?
Social media platforms are fertile ground for manipulation: campaigns, states, interest groups, or coordinated actors can exploit the architecture of platforms to influence political discussion.
For instance, a widely-reported investigation found thousands of automated accounts on Twitter promoting one political candidate and attacking rivals in the U.S. 2023 context. AP News More broadly, the design of platforms—algorithmic feeds, trending topic mechanics, virality incentives—means that manipulative actors can engineer engagement and thus greater visibility.
Platform design enables manipulation in several ways:
Trending algorithms elevate hashtags or topics that receive bursts of coordinated activity.
Engagement-based ranking rewards content that triggers reaction, often polarising or sensational.
The lack of transparency or auditability allows hidden coordinated networks (bots, sock-puppet accounts) to operate.
Echo-chamber effects reduce exposure to counterarguments, making manipulation easier.
The result: the boundary between user discourse, platform design, and political operation becomes blurred. Political influence is mediated through platform infrastructure.
Question 4: How can users, civil society, and policymakers respond or resist the politics of platforms
Develop digital literacy: understanding how feeds are influenced, how algorithms work, and recognising manipulation or echo-chamber dynamics.
Be critical of what is shown as “trending”: challenge assumptions of platform neutrality.
For civil society and academic actors:
Advocate for transparency and algorithmic accountability: for example, audit studies of recommender systems reveal disparities.
Build alternative platforms or community norms that prioritise deliberation over engagement.
For policymakers and regulators:
Push for governance frameworks that require platforms to publish data about amplification, moderation decisions, and algorithmic impacts.
Consider regulation of recommender systems, especially where political discourse is affected.
In sum: resisting platform politics means both individual awareness and collective institutional action to align platform power with democratic norms rather than simply commercial incentives.
Platforms like Twitter, Instagram, TikTok, and others are not simply neutral conduits of political conversation—they are active shapers of discourse. Algorithms determine what we see, moderation policies determine what stays, and manipulative actors exploit platform design to shift discussions. Recognising the “platform politics” of social media is a critical step. The next step is asking: Who controls the platform? Whose voices are amplified or silenced? And how can we ensure that digital public spheres remain democratic, not just profitable?
Gil de Zúñiga, H., Barnidge, M., & Diehl, T. (2018). Political Persuasion on Social Media: A Moderated Moderation Model of Political Discussion Disagreement and Civil Reasoning. The Information Society, 34(5).
Howard, J. (2024). The Ethics of Social Media: Why Content Moderation is a Moral Duty. Journal of Practical Ethics, 11(2). Michigan Publishing
Corsi, G. (2024). Evaluating Twitter’s Algorithmic Amplification of Low-Credibility Content: An Observational Study. EPJ Data Science, 13, 18.
Audit study. Crowdsourced audit of Twitter’s recommender systems. (2023). Springer Nature.