We need three columns to build a temple! Three policy-oriented reactions to political polarization and online harms

Pic by wal_172619 via Pixabay

The author argues that in order to rebuild and strengthen the columns of this temple that we call democracy, we need to take a holistic approach. Thus, any policy-oriented counterstrategy to current tendencies of political polarization should be built around the pillars of 1) social platform regulation, 2) promoting research, and 3) fostering democratic education and media literacy.

Der Autor argumentiert, dass wir, um die Säulen dieses Tempels, den wir Demokratie nennen, wieder aufzubauen und zu stärken, einen ganzheitlichen Ansatz wählen müssen. So sollte jede politik-orientierte Gegenstrategie zu den aktuellen Tendenzen politischer Polarisierung auf den Säulen 1) Regulierung sozialer Plattformen, 2) Förderung von Forschung und 3) Förderung von demokratischer Bildung und Medienkompetenz aufgebaut werden.

DOI: 10.34879/gesisblog.2021.37

Since liberal philosophers situated the essential roots for Western thought, the Enlightenment, and democratic ideals in antique Greece, the temple is an often-used metaphor for Western democracy. If we think of Western democracies as temples, however, one could get the impression that external and internal forces aim to erode and finally tear down the supporting pillars of this temple. One of the most recent examples may be the storm of the US capitol in January 2021 that should be interpreted as a warning sign of increasing political polarization and extremism. Consequently, I argue that in order to rebuild and strengthen the columns of this temple that we call democracy, we need to take a holistic approach. Thus, any policy-oriented counterstrategy to current tendencies of political polarization should be built around the pillars of 1) social platform regulation, 2) promoting research, and 3) fostering democratic education and media literacy.

While an abundance of articles and policy reports have focused on political and societal polarization in contemporary democracies, their analyses often depend too much on the observants’ political predisposition and stress deep structural imbalances, lack of representation, or the system as the source of phenomena like increasing radicalization, violent protests, and on- and offline hatred. While these analyses may be right for some, they are stained by mono-perspectivity. Thus, I will focus on realizable policy and civil society reactions to phenomena broadly described as political polarization and online harms. Certainly, this way of thinking can be ideologically situated or critiqued and underlies a focus on solutions. However, I believe it definitely represents more viewpoints than a general criticism of the status quo, especially, when this criticism is formulated without offering alternatives or counterstrategies. So to speak, I will present three fields of action against political polarization and online harms.

1 Effectively regulating social platform moderation

Social platforms have been at the center of critique for remaining passive on online harms such as hate speech or radicalization for a long time. However, they are understood increasingly as responsible for the content on their platforms instead of being intermediaries that solely provide a communication platform. While most social platform corporations have taken a more active stance on content moderation in recent years, the massive deletion of content by human moderators and algorithmic systems also weighs democratic rights like freedom of speech and protection of defamation against each other. However, platform corporations are, in a traditional understanding, not democratically legitimated to make these sorts of decisions. Thus, a more effective and precise regulation of content moderation and platform business models is long overdue. Positively, there is the movement and motivation to act on the European level. The European Commission’s Digital Services Act, for instance, seeks to regulate everyday content moderation practices by social platforms. With respect to these practices, I would generally differentiate between positive moderation, such as recommendation algorithms, and negative moderation such as the enforcement of rules by detecting and deleting/acting on content that infringes against these rules (these may be laws, but also the platforms’ community guidelines). In particular, the latter has been at the focus in recent months since Twitter and Facebook started flagging some of Donald Trump’s messages as potential disinformation and eventually even deactivated his accounts after his incitement of the storming of the United States Capitol by his supporters. 

While the European Commission’s proposal for the Digital Services Act and Digital Markets Act also seeks to regulate social platform moderation and redress mechanisms for platform users, recommendation algorithms, e.g., on YouTube or by Google News, remain largely unregulated, although they often reinforce prior beliefs and keep pushing people into so-called rabbit holes with content that becomes increasingly sensationalist, lop-sided, and extreme. Without the regulation of recommendation algorithms, however, social platforms’ business models will remain focused on maximizing peoples’ usage time by showing them emotionalizing content and personalized advertisements between videos or articles, with harmful consequences for social and political cohesion.

Yet, policymakers and researchers as informants of civil society and policy decision processes have little insight into algorithmic content moderation systems by social platforms. While the hesitance to regulate speech from governments is understandable, it underlines that legislation can only be one column when building resilient democracies. Thus, the second column of building a more resilient democratic society and countering polarization should increase access to information/data by research institutions.

2 Enabling and promoting research policy-focused computational social science research

After the Cambridge Analytica Scandal, Facebook has largely blocked data access for researchers through its API (application programming interface). While it has put the Crowd Tangle platform in place to allow researchers to monitor groups and debates on Facebook, the data gathering capabilities are quite limited. While only a handful of selected research groups are permitted extended access to data; these limitations make it very hard to research Facebook’s recommendation and illicit content detection algorithms’ functionalities. Consequently, this kind of access is insufficient to close the information gap between platforms and policymakers that colleagues and I have recently outlined in a policy brief on content moderation. One way to establish better-informed research and an oversight could be algorithmic audits, such as the auditing of algorithmic systems by algorithms supervised by human auditors similar to other fields, where government agencies directly monitor the operational business, e.g., by controlling production standards or demanding reporting. This would certainly be a challenging task, but it would allow insight into the complex algorithmic content moderation systems and would empower policymakers in Brussels to effectively monitor, supervise, and regulate these systems. 

While extended data access and support for policy-oriented computational social science research would help in enabling effective regulation, it would not solve the problem of polarization itself but instead may reduce one of its push factors. Another factor, however, is that many people feel like they do not understand technological changes and how they affect our ways of communicating, consuming, loving, working, or basically all aspects of our everyday lives. Thus, promoting digital skills and media literacy and strengthening a participative understanding of democratic politics needs to be an additional column of counterstrategies to polarization.

3 Fostering digital skills, media literacy, and democratic knowledge

All societal and political answers to polarization must also be based on open and constructive debates, decision-making, and lifelong education. In many European countries, an active civil society consisting of clubs, foundations, and NGOs created various programs for young people, providing course materials to teachers and organizing workshops on online harms, media literacy, and democratic reasoning in schools. 

However, for adults, workshops on online harms and media literacy training are rare in Germany. As for young people, strengthening media literacy can, for instance, contain the training of source checking or practicing image search with tools that are easy to use and accessible to everyone. Recently, a new initiative by the Institute for Strategic Dialogue, the Hertie Foundation and the Robert Bosch Foundation in Germany called the Business Council for Democracy which I am involved in as a research consultant started organizing workshops with similar training for adults as a broad offer to foster digital skills and the understanding of online harms such as hate speech, disinformation, and conspiracy narratives. Without a doubt, increasing media literacy of all ages needs time, and in particular, reaching pensioned people will be difficult. Therefore, programs that organize training in small groups are vital to ensure people of all ages have a chance to participate and feel represented in a changing democratic society. This approach to life-long learning and education on political and technological developments should provide a third column of policy answers to refurbish the temple of democracy.


How can we strengthen the resilience of democratic societies in times of rapid globalization and digital technological change? In this blog post, I have argued that policy reactions to increasing political polarization should be based on (at least) three columns which I have outlined from my perspective as a young researcher of these issues. Moreover, I emphasized that we should generally aim to design and implement holistic approaches when countering complex socio-political phenomena such as political polarization or online harms. Concludingly, only if these policy reactions are developed in more open and participative processes can we renovate this beautiful temple that we call democracy.

Follow the author

Philipp is a PhD Candidate at the Hertie School in Berlin and an affiliated researcher with the DiCED project. As a guest researcher at the GESIS EUROLAB he and Dr. Sebastian Stier investigate the use of social media by political candidates during the German federal elections in 2013 and 2017 based on the GLES candidate study and social media data.


Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.