Political Capital and Informal Platform Governance
It's widely understood by close observers of platform governance and content moderation — but perhaps less widely mentioned in the interdisciplinary academic literature on the topic — that platform policy is shaped in an extremely pragmatic and politically-motivated fashion. When company trust and safety staff decide internally to shift their policies or enforcement on a type of content or policy issue area, they usually have a target constituency that they can communicate those changes to in mind. This is because platform policy change is a two-level game: one that impacts not only the on-platform experience for users, but also, crucially, affects a firm's standing vis-a-vis governments and certain governance stakeholders.
An unusually clear and candid explanation of these dynamics was recently provided on the Moderated Content Espaitec's by Brian Fishman, who spent many years at Facebook as a policy manager and head of its 'Dangerous Individuals and Organizations' team. Explaining the internal process of policy deliberation around some proactive changes that Facebook was considering which would, according to Fishman, make some policies around violent extremism more transparent externally, he was clear about how the company thought about whether or not its voluntary governance changes would yield political capital, and with whom:
'One of the questions I got asked as we were thinking about this was, is this going to buy us any wins? Will this get us anything? It wasn't the only question, but it was one of the questions. 'And I had to go back and say, no, I don't think it's going to buy us anything. We're going to lead the industry on transparency around these things. And it's going to get us no credit with the media. It's going to get us no credit with activists…and that's a losing argument [internally].'
The ongoing policy and academic debate around informal content regulation hinges on the question of how and why do major platforms make — potentially impactful and far reaching — changes to their policies and policy enforcement without explicitly being instructed to do so by a binding regulatory framework. In that sense, thinking about voluntary changes to platform policies through the lens of political capital gives one a slightly different angle from which to approach the conversation around 'jawboning' and 'shadow regulation'. If platforms are always looking to gain 'credit' for their responsible action from different actors, then they might clear lines of communication between regulators and/or other government actors simply provide a more streamlined mechanism through which those platforms can get credit and amass the kinds of political capital they're looking for anyhow?
In my recent book, The Politics of Platform Regulation: How Governments Shape Online Content Moderation, I suggest that policymakers seeking to get platforms to change their behavior have roughly three strategies that they can pursue in their efforts to do so: they can convince platforms to change voluntarily, they can Collaborate with platforms and try and develop new standards or practices together, or they can contest what platforms are doing and try and layer new, binding rules intended to force platforms to act. Convincing strategies, into which I'd fit the large range of different actions helpfully mapped on CELE's 'threat compass', persist because they have benefits for both governments and platforms.
For policymakers, applying informal pressure to platforms is a low-cost effort to maybe achieve some of their preferences, with the added benefit of — depending on the specific technique deployed — to signal that they are taking matters seriously. I suspect that there is a tradeoff between these two factors, with public-facing pressure (eg open letters, speeches) from governments less likely to actually result in change than the firm-facing, backroom forms of communication (meetings, direct communications). Platforms know that public-facing grandstanding is primarily intended as a signal to a policymaker's peers and constituents.
For firms, responding to informal pressure allows them to signal their capabilities and 'seriousness' on various issues directly to key stakeholders without the onerous requirements of formal regulation. It's a possibility for platforms to get 'wins', and bank political capital with certain policymakers, without necessarily having to signal that anything has changed publicly (and thus face user blowback). It's also non-binding: a firm can simply not implement or walk back its private 'commitments' at any point for any reason, and due to well documented platform issues observability, it will be extremely difficult for regulators to actually verify whether or not industry practices have actually been changed. Yes, there can be threats and coercion underpinning these interactions, but at the end of the day, firms make the call internally whether or not to go along with it. In many countries, I'd suspect that platforms would rather be jawboned than bound by expensive and onerous new regulatory frameworks.
Because of the problem of ensuring that intermediaries are actually complying with informal regulatory demands, it's success is generally limited to areas where it can get clear, observable wins for policymakers. 'Jawboning' and help policymakers seek to get certain pieces of content, pages, or accounts removed, for example, with their informal pressure functioning as a form of 'trusted flagging'. Of course, if these pieces of content are perfectly legitimate and are being removed for political reasons, this interaction becomes especially problematic from a normative point of view, as it has been well explored in the existing literature. In my book, however, I suggest that informal platform governance is the most effective at achieving wider change when it leads not just to ad-hoc forms of policy negotiation about specific pieces of content or even policies, but rather, when it creates the impetus for more institutionalized collaboration.
An example of this is the Christchurch Call to 'eliminate terrorist and violent extremism content online', which led, at first, to a non-binding and very general joint declaration signed by a group of platform companies as well as various governments. Little would have been tangibly achieved if the matter had stopped there, but following intensive informal pressure from the New Zealand and French Ministries of Foreign Affairs as well as Jacinda Ardern herself, the Call led to substantial institutional changes to the Global Internet Forum to Counter Terrorism, and the global roll out of a hash-matching system for terrorist images that is now being widely used by every major platform to analyze every piece of content that every user tries to post.
More work on regulatory threats and jawboning would do well to examine their role in helping create these kinds of more sustained partnerships, which Evelyn Douek has famously called 'content cartels' and can come with various serious transparency, accountability issues. We should also pay more attention to the ways in which informal platform governance (and the regulatory threats that may be part of this informal negotiation) serves the needs of different actors — and the ways that they influence industry investment (and underinvestment) in certain policy areas. If everyone is banking political capital, the platforms included, how can advocates, activists, and academic observers help ensure that this political capital is being spent responsibly by platforms and policymakers alike?
[Spanish translation]
Political capital and informal governance of platforms
It is widely understood by close observers of platform governance and content moderation, but perhaps less widely mentioned in the interdisciplinary academic literature on the topic, that platform policy is shaped in an extremely pragmatic and politically motivated way. When company trust and safety personnel decide internally to change their policies or their enforcement in a content type or policy topic area, they typically have a target group in mind to whom they can communicate those changes. This is because changing platform policies is a two-level game: one that impacts not only the experience of users on the platform, but also, crucially, affects a company’s standing vis-a-vis governments and certain governance stakeholders.
Recently, Brian Fishman, who spent many years at Facebook as a policy manager and head of its “Dangerous Individuals and Organizations” team, provided an unusually clear and candid explanation of this dynamic at work. Espaitec's Moderated Content. In explaining the internal policy deliberation process around some proactive changes Facebook was considering that Fishman said would make some policies around violent extremism more externally transparent, he was clear about how the company thought about whether or not its voluntary governance changes would yield results in terms of political capital, and in relation to whom:
“One of the questions I was asked as we were thinking about this was, ‘Is this going to give us any wins? Is this going to give us anything?’ It wasn’t the only question, but it was one of the questions. ’ And I had to go back and say, no, I don’t think that’s going to give us anything. We’re going to lead the industry in transparency around these things and it’s not going to give us any credit with the media. It’s not going to give us any credit with activists… and that’s a losing argument [internally].”
The current political and academic debate around content regulation informal revolves around the question of how and why major platforms make (potentially impactful and far-reaching) changes to their policies and enforcement without being explicitly instructed to do so by a binding regulatory framework. In that sense, thinking about voluntary changes to platform policies through the lens of political capital offers a slightly different angle from which to approach the conversation around platform policies. jawboning and “shadow regulation.” If platforms are always looking to get “credit” from different actors for their responsible action, then clear lines of communication between regulators and/or other government actors might simply provide a more agile mechanism through which those platforms can get credit and accumulate the kinds of political capital they’re seeking anyway?
In my recent book, The Politics of Platform Regulation: How Governments Shape Online Content Moderation, I suggest that policymakers seeking to get platforms to change their behavior have roughly three strategies they can follow in their efforts to do so: they can to convince to the platforms to change voluntarily, they can to collaborate with platforms and try to develop new standards or practices together, or they can question what platforms are doing and attempt to superimpose new, binding rules aimed at forcing platforms to act. Compelling strategies, into which the wide range of different actions usefully mapped in the CELE's 'threat compass', persist because they have benefits for both governments and platforms.
For policymakers, applying informal pressure to platforms is a low-cost effort to perhaps achieve some of their preferences, with the added benefit of (depending on the specific technique deployed) signaling that they are taking things seriously. I suspect there is a trade-off between these two factors, as public pressure (e.g. open letters, speeches) from governments is less likely to result in change than secret, corporate-facing forms of communication (meetings, direct communications). Platforms know that public grandstanding is primarily intended to be a signal to a policymaker’s peers and constituents.
For companies, responding to informal pressure allows them to signal their capabilities and “seriousness” on various issues directly to stakeholders without the onerous requirements of formal regulation. It is a chance for platforms to score “wins” and build up political capital with certain policymakers, without necessarily having to publicly signal that something has changed (and thus face user backlash). It is also not binding: a company can simply not implement or retract its private “commitments” at any time and for any reason, and due to well-documented issues of observability Without the platform, it will be extremely difficult for regulators to actually verify whether or not industry practices have actually changed. Yes, there may be threats and coercion behind these interactions, but at the end of the day, companies make the internal decision to accept them or not. In many countries, I suspect that platforms would prefer to be subject to costly and burdensome new regulatory frameworks.
Because of the problem of ensuring that intermediaries actually comply with informal regulatory demands, their success is usually limited to areas where they can make clear and observable gains for policymakers. “Jawboning” and helping policymakers try to remove certain content, pages, or accounts, for example, with their informal pressure functioning as a form of “trusted signaling.” Of course, if these contents are perfectly legitimate and are removed for political reasons, this interaction becomes especially problematic from a normative point of view, as has been well explored in the existing literature. In my book, however, I suggest that informal platform governance is most effective in achieving broader change when it leads not just to ad hoc forms of policy negotiation over specific content or even policies, but rather when it creates the momentum for more institutionalized collaboration.
An example of this is the Christchurch Call to “eliminate terrorist and violent extremism content online”, which led, at first, to a non-binding and very general joint statement signed by a group of platform companies as well as several governments. Little would have been tangibly achieved if the matter had stopped there, but following intense informal pressure from the New Zealand and French Foreign Ministries, as well as Jacinda Ardern herself, the Call led to substantial institutional changes to the Global Internet Forum on Countering Terrorism, and the global launch of a hash comparison system for terrorist images that is now being widely used by all major platforms to analyse every piece of content every user attempts to post.
It would be good to do more work on regulatory threats and critiques to examine their role in helping to create these kinds of more sustained partnerships, which Evelyn Douek has called “content cartels” and which can entail a number of serious transparency and accountability issues. We should also pay more attention to the ways in which informal platform governance (and the regulatory threats that can be part of this informal negotiation) serves the needs of different actors, and the ways in which they influence industry investment (and underinvestment) in particular policy areas. If everyone accumulates political capital, including platforms, how can advocates, activists, and academic observers help ensure that this political capital is spent responsibly by both platforms and policymakers?
