12.1 Introduction

Fifteen years ago in All Politics Is Global, I developed a typological theory of global economic governance.Footnote ¹ My argument was that globalization had not transformed international relations but merely expanded the arenas of contestation to include policy arenas that had previously been the exclusive province of domestic politics. In my model, global governance structures and international law were intervening variables; what truly mattered was the distribution of preferences among the great powers, defined as jurisdictions with large markets that allowed governments to be rule-makers rather than rule-takers. When great power coordination was achieved, then effective governance would be the outcome. When great power coordination was not, then global governance would exist in name only.

Quite a lot has changed in the last fifteen years. Despite widespread commentary about de-globalization in the wake of Trump, Brexit, and the COVID-19 pandemic, metrics like the DHL Global Connectedness IndexFootnote ² demonstrate that the online exchange of information and data has continued to accelerate. If anything, the work-from-home trend in the advanced industrialized democracies has boosted the data intensity of the modern global economy. Demands for greater content moderation across myriad platforms have concomitantly increased as well. Can any standards be negotiated at the global level?

For reasons I will spell out below, my theory in All Politics Is Global was pessimistic about such an outcome. In reviewing the drafts of the other generative and comparative chapters in this volume, that suspicion has only been reinforced. At the current moment, the likeliest result will be a hypocritical system of “sham governance.” Under this system, a few token agreements might be negotiated at the global level. Even these arrangements, however, will lack enforcement mechanisms and likely be honored only in the breach. The regulatory center of gravity will remain at the national level. Changes at the societal and global levels over the past fifteen years only reinforce the dynamics that lead to that outcome.

The rest of this chapter is divided into four sections. Section 12.2 summarizes the arguments I made in All Politics Is Global and why that model would predict a sham governance outcome. Section 12.3 considers how international relations has changed over the past fifteen years in ways that undercut or reinforce the model’s conclusions. Section 12.4 considers how changes in the political economy of the internet could undercut or reinforce the model’s conclusions. Section 12.5 looks at where we are now and where we might be headed.

12.2 A General Typology of Global Economic Governance

Regulatory coordination across borders is commonly represented as a simple coordination game, sometimes referred to as “battle of the sexes.”Footnote ³ In this game structure, actors are better off if they coordinate to arrive at the same outcome than if they do not. The content of any coordination agreement can have distributional implications, however, which leads to the bargaining problem in world politics. States want to reach agreement, but they would prefer to reach agreement on regulatory standards closest to their preexisting status quo. This creates incentives for governments to obtain a first-mover advantage, writing the rules before any other actor. Of course, such an approach, if pursued by multiple actors, risks an outcome of no coordination. The question regulators and negotiators must consider is how much the added benefit of international coordination exceeds the adjustment costs of any agreement that forces a change in the domestic policy compact.

In All Politics Is Global, I made three addendums to this standard coordination problem. First, not all actors are equal in this standard-setting game. Domestic market size is a powerful source of bargaining power in regulatory coordination. Large markets act as a global attractor for multinational corporations. This endows large-market jurisdictions with greater leverage in translating their domestic standards into global standards.Footnote ⁴ Because multinational corporations tend to prefer policy certainty and uniformity across borders, they will have the incentives to adopt the preferences of their largest customer, thereby acting as an amplifier of that jurisdiction’s preferences.Footnote ⁵

If one market is truly hegemonic and desiring of high standards, the result can be akin to what American political economy scholars call “the California effect” and what Anu Bradford calls “The Brussels Effect” – a race to the top in standards.Footnote ⁶ If the hegemonic actor is more libertarian in its approach to regulation, the result can be akin to what American political economy scholars call “the Delaware effect” and international political economy scholars call the “race to the bottom.” When there is no hegemonic actor, however, the question of regulatory standards is a matter of international bargaining.

Second, when there is a great power concert over the content of regulatory standards, there will be effective coordination at the global level. In the first decade of this century, the only large-market jurisdictions of sufficient size were the United States and European Union. When those two jurisdictions reached an agreement on standards, the only variation in the outcome would be the precise global governance process. In a world in which most jurisdictions are on the same page, the harmonized global governance outcome would usually be a mix of private orders (like ICANN) and universal-membership international governmental organizations (IGOs) like the International Telecommunications Union (ITU). When the great powers are in agreement but there is resistance from the Global South or other weaker actors, the outcome is one of club governance. Club-membership IGOs such as the Organization for Economic Cooperation and Development (OECD) or G-20 will set the standards and then use a mix of cajoling and coercion to bring smaller jurisdictions into line. This was how the G-7 countries were able to create the Financial Action Task Force (FATF) and institute a stringent, global anti-money laundering standard.Footnote ⁷

Finally, and most importantly, at the most basic level the battle of the sexes game is not always the correct description of standard-setting. The coordination game always assumes that an outcome of coordination is a Pareto-improvement over an outcome of no coordination. There are some issue areas, however, where the adjustment costs are so high that this condition does not hold. When adjustment costs are prohibitively high, the equilibrium outcome is no agreement.

For example, as an occasional international traveler, I am keenly aware of the potential benefits of coordination for electrical outlets – no more travel plug converters! That said, because electrical grids were created prior to the current era of globalization, the adjustment costs for any country that had to shift those standards would be massive.Footnote ⁸ These adjustment costs vastly outweigh the inconvenience of buying the occasional adapter when crossing borders.

On online content moderation and regulation, I predicted in All Politics Is Global that there would no bargaining core, even between the United States and the European Union. Different cultural histories meant that what was considered taboo in one society would be tolerated in another one. Indeed, as the other chapters in this volume demonstrate, the great powers hold rather divergent preferences on content moderation. As Eric Goldman (Chapter 2) noted in his discussion of the American approach to content regulation, “many categories of speech that are regulated internationally may be Constitutionally protected in the United States and thus subject to little or no government restrictions.” As for the broad immunity provisions contained in the 1996 Communications Decency Act, he observed, “[the immunity of] Section 230(c)(1) is a globally unique policy. No other country has adopted a legal rule like it.”Footnote ⁹ Indeed, this standard stands in marked contrast to the European Union’s E-Commerce Directive or India’s Information Technology Act.Footnote ¹⁰ This is before one factored in the disparate policy preferences of autocratic and democratic governments. I therefore concluded: “Countries have wildly divergent preferences to which Internet content should be regulated …. For this issue, there is no bargaining core among governments. The predicted outcome would be the unilateral use of national regulations to bar undesired content, and the creation of sham standards at the global level.”Footnote ¹¹

12.3 What a Long Strange Fifteen Years It’s Been

It has been nearly twenty years since I developed my model for global governance. That is an appreciable period in the history of international relations and an eon in the history of the internet. What have been the key changes in the world that would affect my prediction?

On the international relations side of the ledger, there have been three significant shifts. The first is that the number of great powers has expanded. China’s economy is now the largest in the world if one measures GDP using purchasing power parity; even if one uses market exchange rates China is projected to eclipse the United States and European Union sometime in the next decade. In 2009, the G-20, which included all the BRICS economies, supplanted the G-7 and was officially designated “the premier forum for our international economic cooperation.”Footnote ¹² China’s ever-increasing market size has had the predicted gravitational effect on multinational corporations. Any truly global economic governance now requires Chinese buy-in.

As Jufang Wang’s chapter demonstrates, China’s regulatory preferences are sharply at odds with those in Europe or the United States. China’s definition of “harmful content” is much more expansive than in Western societies. It includes – but is not limited to – “smearing national heroes and martyrs” and “sabotaging China’s national unity or religion policies.”Footnote ¹³ Chinese preferences for platform responsibility and content moderation are therefore far more restrictive than any standards promulgated by American or European authorities. The probability of harmonized or club governance in this issue area becomes even more remote.

A second significant shift has been the ongoing democratic recession. Whether one looks at Freedom House, Polity, or other metrics, the results are similar: over the past fifteen years the world has become less free.Footnote ¹⁴ This is due to a combination of factors. Authoritarian leaders in places like Russia and Iran have swept aside any countervailing institutions while enhancing the state’s coercive apparatus. In many democracies, populist leaders have hollowed out institutional checks, political opposition, and legal limits on executive power. The reasons for the democratic recession are unimportant for this analysis. What is important is that an increasing number of anocracies and autocracies will possess policy preferences on content moderation and regulation that more closely resemble China than the United States or European Union. India’s penal code, for example, criminalizes “content intended to outrage religious feelings of beliefs” and “content prejudicial to ‘national integration.’” That sounds far more similar to Chinese standards than European standards.

The third trend is the blowback from US security allies with respect to Edward Snowden’s revelations in 2013 about US covert surveillance. Snowden revealed that the United States had taken advantage of its network centrality to surveil key allies’ email and phone communications. Although this monitoring did not extend to social media, the revelations of US hypocrisy forced many key allies to adopt countermeasures that did have ramifications for social media.Footnote ¹⁵ This US surveillance highlighted the ways in which the United States could exploit weaponized interdependence in online networks.Footnote ¹⁶ In particular, the United States could exploit the “panopticon effect” of network centrality to collect data and metadata about social media users. As the chapter by Artur Pericles L. Monteiro demonstrates, key US allies seized on the Snowden revelations to call for greater data sovereignty – that is, for social media firms like Google, Facebook, and Twitter to keep personal user data within a jurisdiction’s national boundaries.Footnote ¹⁷ Such a move enables these jurisdictions to legislate questions of data privacy with minimal concerns about great power responses.

12.4 From Web 1.0 to Web 2.0

On the political economy side of the ledger, the most significant shift has been from a Web 1.0 to Web 2.0 architecture. Web 1.0, which roughly characterizes the first fifteen years of internet commercialization, was a “read-only” venue, in which a small number of providers generated content for a much larger audience. The consumption of internet content was relatively passive. Even the small number of content providers are fragmented, however. The industrial organization of Web 1.0 limited the market power of any internet service provider relative to its competitors. Therefore, conversations about platform moderation and regulation made little sense, since there were no platforms of any appreciable size beyond, perhaps, AOL.

The transition to Web 2.0 led to a much more interactive online experience. The introduction of social media enabled a more participatory experience for users. The economics of social media, however, made Web 2.0 radically different from Web 1.0.Footnote ¹⁸ The quality of social media is primarily a function of quantity; as the size of the network increases, so does the consumer utility. This creates a structure of natural monopolies, in which giants like Apple, Facebook, Google, or TikTok dominate certain spheres of Web 2.0. Also-rans like Friendster, MySpace, Yahoo, or Vine quickly receded from view. Content regulation and moderation shifted from an internet-wide discussion to more focused questions about corporate culture within the firms that hosted the dominant platforms.

At the same time, the primary revenue source for Web platforms raised thorny questions about data privacy. As one review essay famously noted, the users of social media provide the value-added for these platforms in supplying the raw data that these platforms can exploit for commercial purposes.Footnote ¹⁹ This gives Web 2.0 winners a powerful incentive to resist any official restrictions to be placed on their platforms. Because these companies prove to be natural monopolies, they also have the resources to wage political fights against any such regulations. This does not mean that Web 2.0 monopolies can always thwart regulatory outcomes; the actions of the Chinese censorship regime, the European Commission’s Competition Directorate, and even the US Federal Bureau of Investigation falsify this claim. It is more accurate to state that these firms can use the threats of political voice and market exit to apply some constraints on these regulatory efforts. As Google’s litigation strategy in India reveals, “the nature of litigation in India coupled with the legal resources of platforms may render intermediary liability (i.e., the risk of liability enforced through private lawsuits) a weak regulatory tool to regulate platform behavior.”Footnote ²⁰

Beyond the lobbying advantage, Web 2.0 providers also possess a massive informational advantage over potential regulatory actors. Simply put, the financial remuneration for working at a large platform firm like Facebook or Google vastly outweighs working for the state. This means that the allocation of human capital will be heavily skewed toward the platform provider in any negotiation or litigation with either legislative or regulatory actors. The imbalance in skilled personnel makes it harder for the elected officials to pass the requisite laws and regulations. In some instances, members of legislatures have demonstrated such radical ignorance about the nature of online platforms that they lack the requisite knowledge to ask informed questions.

Even if elected officials were more tech-savvy, it is far from clear whether they could craft effective regulation of online platforms. Ordinary models of regulation assume imperfect information between the producer and the regulator. The growing complexity of the online sector, however, exacerbates that information asymmetry even more. Political science models demonstrate that as the complexity of the relevant business sector increases, the regulatory outcomes will be increasingly biased toward those preferred by the firms.Footnote ²¹ This holds for sectors like pharmaceuticals and commercial aviation; it likely holds for Web 2.0 as well.

12.5 The Present and the Future

All of the trends discussed in Sections 12.3–4 reinforce the likelihood of a sham governance outcome for content moderation and platform regulation. The rise of China makes the possibility of great power agreement on any substantive content moderation extremely unlikely. China’s treatment of Google and other US-based platforms demonstrate its hostility to Western modes of content regulation. US resistance to a Code of Conduct for Cyberspace, as promoted by Russia and China, speaks to this lack of consensus between the great powers.Footnote ²² In recent years, China has also grown more comfortable in using its market power to pressure Western multinationals.Footnote ²³ Indeed, Chinese officials may well view their “Great Wall” approach to online content as an implicit industrial policy for Chinese internet providers, allowing the likes of TikTok, Huawei, and Weibo to compete with Facebook, Samsung, and Google on the global stage. This only increases the incentive for Chinese public policy preferences to diverge from the West. Rising geopolitical tensions over Taiwan and the South China Sea will further decrease the likelihood of cooperation. A great power concert over platform moderation is a nonstarter.

The growth of more illiberal autocracies increasing the number of states that might share Chinese regulatory preferences hints at a possible rival standards outcome. China and the OECD economies could promulgate two competing sets of content moderation standards and try to cajole or coerce the rest of the world to their standard. The alienation of many US allies to Washington’s hypocrisy, however, makes even this kind of agreement unlikely. The United States and European Union already have disparate policy preferences over content regulation; the festering lack of trust from the Snowden affair will make agreement that much more difficult to achieve. Similarly, China’s myriad overreaching and “wolf warrior” diplomacy will make even autocratic leaders wary of permitting too much Chinese control over cyberspace.Footnote ²⁴

Finally, over time, the dominant platforms have acquired sufficient lobbying resources and political influence to dramatically raise the costs of additional regulation. The libertarian origins of Silicon Valley permeate the entire sector, causing these firms to be reflexively hostile to additional regulation. Some platforms have ratcheted up self-regulation in an effort to forestall additional government intrusion. In this political environment, the only possible remaining outcome is one of sham governance, in which content-free nostrums are agreed upon at the global level and barely implemented at the national level.

Indeed, this largely matches current descriptions of internet governance. A recent Council on Foreign Relations task force report concluded that while the United States aspired to create an open, global internet, that ideal no longer bore any resemblance to reality:Footnote ²⁵

Could this change in the future? Further technological innovation could rejigger the distribution of policy preferences. The claimed evolution from Web 2.0 to Web 3.0 – in which, as with Substack, the user could exercise more autonomy and ownership over content creation – could lead to a redistribution of benefits and responsibilities that makes content regulation easier. That said, most of the salient trends suggest an even further fragmentation of platform regulation. In theory, a massive scandal could create a policy window that would prompt additional regulation. In practice, Facebook’s myriad scandals have not led to any meaningful change in content regulation. This suggests that Web 2.0 monopolies still possess the requisite political power to act as effective veto players. On the international relations side of the ledger, rising geopolitical tensions will act as a further deterrent to effective global standards.

The rise of populist nationalism within the advanced industrialized democracies will also have debilitating effects on the ability to regulate. As Goldman warns in his memo, “online speech freedoms have become inextricably intertwined with partisan politics, which creates irreconcilable conflicts.”Footnote ²⁶ In many instances, populists exploited disinformation campaigns via online platforms to obtain greater political influence; they will stoutly resist efforts to regulate that political tool. Furthermore, by definition, populists dislike technocratic methods of regulation, perceiving them to be a tool of elites that constrain political power. Populism and political polarization both erode the rule of law, enervating effective regulation. Populism also emboldens private actors to litigate against the state, hopeful that a new government will countermand the orders of the previous government.Footnote ²⁷

The increase in data localization will enable a more heterogenous set of platform moderation policies to emerge. Multinational firms will find it harder to claim that they are beyond the scope of national regulation. Given ongoing trends in international relations and the industrial organization of the internet, however, there will not be any substantive global governance on this issue for the next generation.