Keywords

Introduction

We are at a watershed moment in the development of the Internet precipitated by intense and ongoing criticism of a relatively small number of global Internet giants—Google, Apple, Facebook, Amazon, Microsoft, and Netflix (hereafter called GAFAM+). Google and Facebook’s ad-driven business models, in particular, have come under fire for causing a “crisis of journalism”, wrecking the media industries and destroying democracy. These firms also stand accused of remaking the Internet in their image—a centralized Internet ruled by a few search engines, social media services, and digital media content aggregation platforms (Noam, 2016). Their efforts to rewire the Internet for hyper-targeted advertising and messaging, now hijacked by dis/misinformation operations and used to fan the flames of political polarization, promote hate, abuse and violence, have also come under intense scrutiny (Benkler, Faris and Roberts 2018; Ghosh and Scott 2018; Mckelvey 2018).

In response, governments have convened a dizzying number of public policy inquiries into the platforms (Winseck and Puppis 2020). Following one such inquiry, in 2021 the Australian government passed the News Media and Digital Platforms Mandatory Bargaining Code that seeks to have Google and Facebook pay Australian media companies for the news content they use as part of their online search and social media services. While that law makes for a popular tale of a small nation state using its sovereignty to curb the power of the American platform giants, is it really the model for a new generation of Internet regulations for the public interest and democracy that many are holding it up to be? For their part, the Internet giants—especially Google and Facebook in the Australian case—are fighting tooth-and-nail against any such efforts, with their collective spending on lobbying in Washington soaring from $12.2 million in 2010 to $65 million in 2019 to do just that (OpenSecrets.org 2020), yet all the while taking public stances that openly invite governments to create new rules for the Internet.

This chapter agrees that a forceful response to the platforms is long overdue but raises concerns that the case against GAFAM+ has become orthodoxy, anchored in misguided conceptualizations of ‘big tech’, cherry-picked evidence and a tendency to see these firms as the cause of all perceived woes. Working from the cultural industries approach to political economy, I argue that there is a crisis of journalism but argue that GAFAM+ are not to blame for it, and that in fact most media sectors are thriving. I also argue that, while the ongoing attempt to pin the label of media company on the digital platforms may be a politically expedient stepping-stone to justify regulating them based on broadcast standards, this approach rests on superficial analogies between broadcasters/media companies and digital platforms. It ignores the fact that the media industries have developed in close proximity to the vastly larger telecoms, electrical equipment manufacturing, consumer electronics and banking firms since the mid-nineteenth century.

I argue that, too often, media studies scholars neglect this more encompassing history by their focus on the mass media of the twentieth century, while the focus of platform studies on the self-contained history of computing and information systems over the last three- to four decades has a similar affect (Nieborg and Helmond 2019). In contrast, I focus on several rounds of Goliath versus Goliath battles between AT&T, Western Union and “the Radio and Electrical Group” (e.g. GE, RCA, Westinghouse, etc.) from the late 1870s through to the end of the 1930s that had effects on the economy, media and society that were at least as significant for the rest of the twentieth century as the impact that the platformization of the Internet (Nieborg and Helmond 2019) is having today. The last sections of this chapter offer four principles of structural and behavioural regulation drawn from this history as guides for what a new generation of internet regulation could look like in our own times: structural separation (break-ups), line of business restrictions (firewalls), public obligations and public alternatives (Rahman 2018).

The Contours of “Digital Dominance”

By the end of 2018, the combined market capitalization of the GAFAM+ group of internet giants had reached $3.5 trillion.Footnote 1 That level was roughly four times the market capitalization of the big six telecoms and media operators in the US.Footnote 2 The platforms had combined assets worth more than a trillion dollars and revenue of $825.6 billion in 2018 (Company Annual Reports). The upshot, according to Diane Coyle (2018), is that “these giant platforms now go far beyond any other commercial entities in the scale and dominance they have achieved” (57).

In their respective markets, GAFAM+ paint a picture of dominance. In possibly the most iconic example, Google is responsible for 92.5% of all searches worldwide, 88% in the US, similar levels in Canada and 95% in Australia.Footnote 3 But this story extends far beyond search products. In fact, all core elements of the “platform Internet” are highly concentrated, with concentration levels (using standard economic measures, i.e. CR4 and HHI scores) typically double to triple the threshold above which concentration concerns are registered.Footnote 4

The extent of Google and Facebook’s clout is most obvious in the Internet advertising market.Footnote 5 Online advertising currently accounts for around half of all advertising spending in Australia, Canada and the US. As that figure increases, however, the digital duopoly’s grip on total advertising spending is steadily tightening. However telecommunications giants AT&T, Verizon and Bell Canada are also pursuing new vectors of vertical integration into the “big data” economy through acquisitions of digital advertising and data analytics firms to build their own advertising exchanges to rival Google and Facebook.Footnote 6

The recent Australian Competition and Consumer Commission’s (ACCC) (2019) Digital Platform Report found that Australians spend half of their time online using the services of Google (20.5%), Facebook (18.6%), Microsoft (3.4%), Snapchat (2.4%), Apple (2.1%) and Australian’s main news media outlets (2.3%) (ACCC 2019: 6). If Australia’s figures are a proxy for the US, Americans spend about half of their time on the Internet using these services (Nielsen 2019: 3). What people do with the remaining half of their time online, however, the ACCC does not say.

Google, Facebook and Microsoft have also acquired significant ownership stakes in thirteen trans-oceanic submarine cables, while cloud computing infrastructure is effectively controlled by four companies.Footnote 7 This dominance of the ‘hidden infrastructure of the internet’ underpins the assertion that “Google and FB have direct influence over 70% + of Internet traffic” (Stalz 2017), an implausible but oft repeated claim (Flew, Martin and Suzor 2019: 33; Owen 2019: 3).Footnote 8

Are GAFAM + to Blame for the Crisis of Journalism?

Attempts to legitimate a new era of Internet also often rest on assertions that the GAFAM+ group of Internet giants are the cause of the twin crises in journalism and media. Beginning with journalism, Stoller, Miller and Teachout (2020), for example, state the case bluntly: “The primary cause of the collapse [of] local and independent journalism... is Facebook and Google’s control of digital advertising revenue” (8). The Open Markets Institute also crystallizes the nub of the case:

We cannot understate the threat monopoly poses to the free press…. Google and Facebook are breaking the news. It’s time our policymakers and regulators do their jobs and break up these monopolies, before they destroy our democracy (Open Markets 2019).

The evidence that newspapers are in crisis seems clear. Revenue for the newspaper industry in the US fell by over half between 2006 and 2019. The same is true in Canada and Australia. That Facebook and Google’s revenues in the US and Canada are now double and triple that of the newspaper industry as a whole in both countries, respectively, seems to further prove the point.Footnote 9

However, blaming Google and Facebook as the cause for newspapers’ dire straits is at best a partial explanation. Circulation numbers in the US, Canada and Australia peaked in mid- to late-1980s and have fallen ever since. Circulation revenue in all three countries also began to fall in the early 2000s. The same is true for newspaper advertising revenue: it continued to rise until 2000, fell after the dot.com bubble burst, and truly collapsed after the financial crisis hit of 2008.

In sum, the woes facing ‘legacy’ media sectors started long before Google and Facebook came along. To compound these woes, over the last decade, “legacy media” firms relying on advertising have also been battling Google and Facebook for a stagnant or shrinking pool of ad dollars.Footnote 10 This trend reflects the fact that advertising spending hinges on the state of the economy (see Picard 2011). However, this structural decline in advertising spending has been supplanted by the tendency to focus on Google and Facebook as the exclusive sources of journalism’s crisis, an orthodoxy amongst public inquiries, (many) academics and lobby groups (ACCC 2019; BTLR, 2020; UK 2019a, 2019b; US 2019).

A decade ago, however, when the Internet itself was blamed for killing the press, the charge was refuted by many scholars (Downie and Schudson 2009; McChesney and Nichols, 2010; Picard 2009). Others pointed to self-inflicted wounds brought about by two decades of consolidation, excessive capitalization and bloated debts, hyper-commercialization, and the triumph of corporate values over journalistic norms (McChesney and Nichols 2010; Pickard 2020).

Are GAFAM + Destroying the Media Industries?

Beyond journalism, GAFAM+ are also blamed for upending the media industries writ large. According to Jonathan Taplin (2017), for example, in the US, the Internet giants have diverted “$50 billion per year… from the creators of content to the owners of monopoly platforms” (7). Richard Sturgess (2019), a former president of the CBC, offers a similarly dismal account for Canada. Many media studies scholars seem to agree (Napoli 2019).

The image of a commercial media system being completely in distress, however, is misleading. Take the music industries, for example. While Taplin points to losses in “recorded music” to buttress his claim, once revenue from the publishing, streaming services, and live concert segments is accounted for, the story changes completely.Footnote 11 Revenue for the “total television market”Footnote 12 in the US also soared from $182.1 billion to $294.3 billion between 2010 and 2019. For digital games, revenue doubled from $13.7 billion to $27.2 billion over the same period. Overall, revenue across the US media content sectors rose from $298 billion in 2008 to $484.3 billion in 2019. This is a net gain of $108.6 billion accrued not only alongside the rise of GAFAM+ but against the headwinds of the global financial crisis.

A similar picture emerges for Canada. Revenue across all segments of the television marketplace rose from $13.4 billion in 2008 to $17.1 billion in 2019. In addition, revenue for the online music, gaming and app stores has sky-rocketed from $718.9 million in 2011 to $5.6 billion in 2019. In total, combined revenue across Canada’s media content sectors grew from $18.8 billion in 2008 to $27.7 billion last year. Television and film production investment in Canada also jumped from $5.4 billion (CDN) in 2008 to a record high of $9.3 billion in 2019 (Nordicity 2020, Exhibit 1–2). In fact, film and television production have been drive to all-time highs in the US, Canada and EU by the vast growth in the television market and massive increase in spending on television and film production by the streaming services (IBIS 2019a; IBIS 2019b; Spangler 2020; Eurostat 2020).

Underpinning these trends is the reality that subscriber fees and direct purchases are now the core of the media economy. In fact, those revenue sources outstrip advertising revenue by a 3:1 ratio in the US, and 5:1 in Canada (reflecting the more commercialized US media system). Given these realities, the fact that so much of the discourse of crisis is centred on lost advertising, while making generalized claims about the state of the media economy is a classic case of misdirection.

To be sure, the IT behemoths have carved out sizeable spots for themselves in this rapidly enlarged and more complex network media economy, but they do not dominate it writ large. For example, while Google and Facebook dominate online advertising, as we saw earlier, they only account for about 6.4% and 3% of the trillion-dollar US media economy, respectively. While Netflix’s US revenue reached $9.5 billion in 2019 (roughly a third of the online video revenue market), its 3.2% stake of the television market is well below that of the “big six” US television giants: AT&T (16%), Comcast (14%), Disney (8.5%), Viacom-CBS (8.2%), Charter (6%), and 21st Century Fox (3.9%). Altogether, the IT giants’ combined domestic media-related revenue accounted for 13.6% of the US media economy in 2019. In contrast, AT&T—the largest communications and media conglomerate in the US—single-handedly had a 15% share of the huge US media economy and its domestic revenue ($143.9 billion) was more than the GAFAM+ group’s US media-related revenues combined (i.e. $139.2 billion). In terms of the 25 biggest firms in the network media economy in the US based on domestic revenue, Google and Facebook rank fourth and eighth, respectively, while Microsoft, Amazon, Netflix and Apple rank twelfth, fifteenth, seventeenth and eighteenth, respectively, on the list. Drawing these comparisons more broadly, also reveals that five biggest telecoms-Internet and media companies in the US—AT&T, Verizon, Comcast, Disney and Charter—employed twice as many people as Google, Apple, Facebook, Microsoft and Netflix in 2018 and held assets worth twice those of the IT giants.

In Canada, similar trends prevail. In 2019, Google, Amazon, Facebook, Apple, Microsoft and Netflix combined revenue reached $9.3 billion. Bell, the biggest communications and media conglomerate in Canada, by contrast, had domestic revenues of $24.9 billion in 2019—nearly three times the Internet giants’ revenue in Canada combined. Compared to the half-dozen global Internet giants’ 10% stake of the media economy, the “big five” Canadian companies—Bell, Telus, Rogers, Shaw and Quebecor—raked in close to three-quarters of all revenue. Figure 12.1, below, rank-orders the biggest players in the media economy based on revenue from their Canadian operations.

Fig. 12.1
figure 1

(Source Canadian Media Concentration Research Project Dataset 2019 [Winseck 2020])

Top Telecom-Internet and Media Companies in Canada, Ranking Based on Revenue in Canada, 2019 (Millions$)

Full size image

Are Digital Platforms Media Companies?

It is increasingly common to hear arguments that the global Internet giants are not only affecting the media but that they, themselves, have become media companies (broadcasters), and should be regulated accordingly. According to this perspective, rather than being neutral intermediaries, these companies’ decisions deeply influence the terms of search, the visibility of content in social media spaces, and how the media in general operate. They are said to be media companies because they are competing with traditional media firms for advertising dollars. They are also ‘pathways to the news’, with roughly two-thirds of Americans, Australians and Canadians getting some news from social media and two-out-of-five from Facebook (Reuters Institute 2020). Furthermore, the platforms are becoming more involved in original content creation (Napoli 2019). As Flew (2019) surmises from these observations, “the digital platform companies are becoming de facto media companies, insofar as they are critical gatekeepers around the circulation of digital media content in all of its forms” (13), but without the usual obligations that come along with that status.

From this angle, regulating the platforms like media/broadcasters is also said to be the best option because the current alternative—competition policy—and the old line between carriage and content upon which the platforms’ status as “mere conduits” allegedly rests are of recent vintage, quintessentially American in origin and based on narrow economic and technical conditions that no longer hold (Flew, Martin and Suzor 2019). In contrast, defining digital communications platforms as media firms would bring them under the authority and broader policy remit of broadcasting regulators such as the ACMA, CRTC, FCC, Ofcom, etc. (ACCC 2019; BTLR, 2020; Napoli 2019; UK 2019a, 2019b; US 2020).

Arguments that Platforms Are Media Companies Are Sociologically Misleading

While accepting the case for Netflix or Hulu, two companies whose operations and identities are centered on media production and distribution, I agree with Hesmondhalgh (2019) that, while “conflating the cultural industries with the IT sector might assist in a worthwhile battle for greater regulation of online content, it would also be sociologically and historically inaccurate and misleading, downplaying important tensions between the different sets of corporations and their varying interests” (472). Instead, the GAFAM+ group of companies are better seen as giant IT conglomerates with significant media subsidiaries.

To be sure, Amazon, Apple, Facebook and Google have upped their spending on television and film production,Footnote 13 but such spending still accounts for a tiny fraction of their total operating expenses (e.g. less than one percent and just over five percent of spending at Google and Facebook, respectively, versus one-half to two-thirds at Disney and Viacom-CBS). Simply put, creating and commissioning original content is not central to the businesses of GAFAM+ or their corporate culture and identity.

Second, it is also a stretch to see millions-upon-millions of automated decisions made by machines at great scale and speed as comparable to the human decision-making and thicket of social relationships that go into commissioning, publishing, scheduling, and creating a catalogue of book, movie, television, music and digital game titles. Drawing an equivalency between the digital platforms’ content moderation practices and the work of media professionals ignores a valuable history of approaches to the sociology of news/media/production (Banks, Connor and Mayer 2015; Hesmondhalgh 2019).

A recent report by an inter-agency group of French regulators based on their six-month stint embedded at Facebook’s offices in Paris fleshes out this point. The report differentiates between the machine-based “ordering” of content versus editorialization and, in so doing, distinguishes between the platforms and media companies on this basis. The result is that while there is no doubt that Facebook is intervening extensively in the flow of content made available on its platform, it is not really involved in the creation and editorial selection of content on the basis of journalistic norms or those of other media professionals. As the report also states, “[u]nlike traditional media, social networks do not select each item of content published on the service. This is a defining characteristic of such services” (France 2019: 14). The report also stresses that even if “this function of ordering content constitutes a form of de facto editorialization, this cannot question the legal status of the operators or lead to legal requalification of hosting providers as publishers, since the majority of social network services do not carry out any selection prior to the publication of content” (France 2019: 9). Given all this, the report concludes that digital platforms are “information organization systems” not media companies (9).

Claims that the IT Giants (Digital Platforms) Are Media Companies Are Historically Misleading

The argument that platforms are media companies also ignores the extent to which the modern press, recorded music, film, radio and television industries, computing, and the Internet have all developed in proximity to vastly larger telecoms, electrical equipment, consumer electronics and banking sectors for the past 170 years (Miege 2011; Hesmondhalgh, 2019). While those industrial giants played a leading role in the creation of the media industries since the mid-nineteenth century, at no point were they ever understood to be media companies. Rather than shoehorning them into a politically expedient definition, policymakers recognized the unique roles they played and regulated them accordingly.

The Goliath versus Goliath battle between Western Union and the Bell telephone system in the late 1870s that led to major advances in telegraphy and telephony and in recorded music, broadcasting and the motion picture industries illustrates this point well. This bout of rivalry ended in a truce in 1879 wherein Western Union and the National Bell Telephone Company agreed to segment the fields of telegraphy and telephony and stay out of one another’s main line of business. The intense battle before this truce also turned scientific research and development into an industrial enterprise, exemplified by the creation of Edison Labs and eventually the renowned Bell Labs by 1925—the centre of American telecoms, electronics and computing supremacy for most of the rest of the twentieth century (Danielian 1939: 92–96; John 2010: 156–170, 209).

The 1879 truce settled matters between the two corporate titans for the next three decades, but in 1908 AT&T took-over Western Union in a bid to create a universal telegraph and telephone system. Five years later, however, AT&T was forced to unwind that acquisition in return for the US Department of Justice (DOJ) settling the case accusing the company of monopolization (John 2010: 352–361). The alternative was a high-profile court case that could have risked AT&T being broken up into regional units, as the Wilson Administration Attorney General, James Clark McReynolds, had originally sought.

This outcome is highly relevant to how we think about platform regulation and the potential benefits of breaking-up some of the Internet behemoths today, with four aspects of the McReynolds Settlement (also referred to as the “Kingsbury Commitment”) standing-out:

  1. 1.

    AT&T would divest Western Union;

  2. 2.

    Henceforth, AT&T would need government approval before buying-out competitive independent telephone companies;

  3. 3.

    AT&T would interconnect with independent telephone providers using standardized technical interfaces and contracts;

  4. 4.

    AT&T’s interconnection agreements, technical interfaces, interoperability standards, and retail prices would be subject to federal regulatory review (Danielian 1939: 92–110; John 2010: 352–361).

Despite these relatively tough regulatory interventions, AT&T’s interests continued to expand far beyond telephony and it confronted other corporate titans with their own ambitions. In fact, this was a period when two opposing industrial groups often engaged in bouts of fierce competition mixed with moments of cooperation at the frontiers of industry, electrical equipment manufacturing and communications: the “Telephone Group” (AT&T and Western Electric) and the “Radio and Electrical Group” (GE, Westinghouse Electric and Manufacturing Company, RCA, Wireless Specialty Apparatus Company. United Fruit and Tropical Radio).

By the mid-1920s, both groups had established a sprawling melange of interests that reached across the fields of telegraphy, telephony, phonographs, electrical circuit arrangements, sound recording, radio broadcasting, domestic and trans-Atlantic wireless communication, submarine telegraph and telephone communication, pictures by wire, telephoto (wireless pictures), photoelectric cells, television, carrier current on power lines, communication with trains, train dispatching, railway and traffic switching control equipment—collectively constituting the infrastructure of twentieth century industrial capitalism. The name of the game, especially at AT&T, was not to innovate on the frontiers of technology but rather to conduct some modest research while buying patents whenever they could and defending the patents they did have to the hilt (Danielian 1939: 109–124)—much like the contemporary strategy of major platforms, which have bought out small innovators and their intellectual property.

Despite creating a “patent pool” under the watchful eyes of the US government in 1920 intended to limit conflict between them, the advent of radio broadcasting became a “no man’s land” where the “Telephone Group” and “Radio Group” continuously clashed between 1922 and 1926. During this time, AT&T controlled many of the patents upon which the new field was being built, and in addition it owned the WEAF radio station in New York. The station was the linchpin in a chain of seventeen broadcasting stations in the northeastern US that formed the centrepiece of AT&T’s Broadcasting Corporation of America and its ambitious plan to create a monopoly broadcasting network operated by the “Bell System” across the country.Footnote 14 This battle, however, was put to rest in 1926 when AT&T agreed to exit the broadcasting business while the Radio Group was left to develop radio broadcasting and all areas of wireless communications on its own but also to rely exclusively on AT&T for wire services when needed (Danielian 1939: 126).Footnote 15

These examples from the US were not exceptional. Similar relationships involved communications and electrical equipment manufacturers in Germany (Siemens), France, the UK and other countries. In each case, industrial manufacturing enterprises not only built up the technological side of radio broadcasting but in some cases, as in the UK, the “big six” equipment manufacturing companies—Marconi, the Radio Corporation of America, Metropolitan-Vickers, British Thomson-Houston, GE, and Western Electric—created the British Broadcasting Company in 1922. Within four years, however, they exited broadcasting after the British Government refashioned that company into the public service British Broadcasting Corporation. In the US, General Electric and Westinghouse were shoved aside from their lead role in NBC several years later based on the recommendations of a Federal Trade Commission inquiry in 1931.

Despite the 1926 agreements, the introduction of sound technology into the motion picture industry opened a new front in the battles between AT&T and Western Electric, on the one side, and GE, RCA and Westinghouse on the other. Western Electric (AT&T) gained the upper hand when it signed a series of long-term deals between 1928 and 1930 with Columbia Pictures, First National Pictures, the Fox Hearst Corporation, Metro-Goldwyn Pictures, Paramount Famous Lasky Corporation, United Artists and Universal Pictures, etc.. Under the terms of these deals, the Hollywood studios could only distribute their films to theaters with Western Electric equipment while theaters could only show pictures produced on such equipment. Those who dared use rival technology faced punishing penalties for doing so (Danielian 1939: 142–150). According to AT&T, such arrangements were needed to maintain the quality technical standards that it was known for. AT&T also parlayed this experience into becoming one of the largest investors in Hollywood films during the 1930s, which gave it a direct role in commissioning and editing motion pictures. Ultimately, AT&T exited the movie business, however, after coming under pressure from an emboldened FCC that was investigating the “monopoly problem” in the communication and broadcasting industries in the late 1930s (Danielian 1939: 142–150). To put all this in language familiar to platform and media studies scholars today, AT&T used its extensive control over technical interfaces, interconnection and interoperability to buttress its dominance in telephony and to extend its influence into two burgeoning new fields of the entertainment and the cultural industries: broadcasting and film.

Monopoly busting regulatory approaches were again applied when the FCC later sought to break up AT&T and Western Electric’s emerging control over the motion picture industry, to put an end to the Associated Press’s exclusive licensing arrangements driving consolidation of local newspapers in 1945, and to force the break-up of the vertically-integrated Hollywood Studio system (the 1948 Paramount Decision). The Department of Justice used Consent Decrees, for example, to prevent AT&T from entering the computer hardware, software, and processing industries in 1956 and again in 1984 to break-up the company. Meanwhile from the 1960‒1980s, the FCC’s Computer Inquiries used line of business restrictions to keep AT&T out of the nascent information and computer services industries before allowing it to enter these markets via separate subsidiaries and on non-discriminatory, common carrier terms.

Cannon (2003) sees these steps as key to the success of the open Internet because they limited AT&T and, after its break-up, the “Baby Bells” control of the interconnecting network of networks. This conceptual framework was subsequently exported around the world, first through a series of bilateral deals between the US, UK, Japan, Canada, and the EU in the 1980s and 1990s and, ultimately, as the template for the WTO’s Basic Telecommunications Agreement in 1997. As markets were liberalized, however, concerns about the need to control dominant market power fell and antitrust principles fell out of favour, thereby yielding the Internet governance regime that has bequeathed to us the globe-spanning platform giants of the twenty-first century.

In sum, the industrial giants that constituted the “Telephone” and “Radio” groups, respectively, played leading roles in the creation of the early-twentieth century media industries. However, at no time did this mean that they were seen as broadcasters/media companies, or regulated as such. Instead, governments fashioned new regulatory tools fit for such realities. They broke apart the fusion of AT&T-Western Union. They imposed line of business restrictions that limited firms’ ability to expand into adjacent industries. And they regulated technical interfaces, interconnection, interoperability, service pricing and universal service mandates. They adopted common carrier and cross-ownership rules that prevented carriers from controlling media content companies. In the UK and Europe, the public service broadcasting alternative to the commercial model also emerged from this context.

The Promise of a New Generation of Internet Regulation for the Public Interest and Democracy

Instead of resting the case for platform regulation on superficial analogies to media companies, a more compelling case can be made for using the following four principles drawn from history of telecoms, the media and antitrust regulation: structural separation, firewalls, public obligations, and public alternatives (Rahman 2018: 1623). These starting points shift the focus away from the content-centric concerns typical of the media policy and platform governance literature to “structural remedies [that] seek to eliminate the incentives that would make that conduct possible or likely in the first place” and “behavioral remedies [that] seek to prevent firms from engaging in specific types of conduct” (emphasis added, Khan 2019: 980).

Structural separation and firewalls: When talk turns to “breaking-up” Google and Facebook, a key focus is on their ownership and control of online advertising exchanges, data, audiences, terms-of-trade and the other hidden levers of power that echo historical telecommunications tactics of using technical control to buttress market power. Ghosh and Scott (2018), for example, see control over these resources as the taproot of Google and Facebook’s growing dominance of online advertising and a threat to democracy. To address the far-reaching public interests at stake on both counts, we need to start with a prime target for platform regulation: breaking up the vertically-integrated online advertising stack.

In the case of Google, for example, this would require the company to spin-off (structural separation) or, less ambitiously, to build a firewall between its suite of services, Android operating system and its online advertising exchange, respectively. It would also need to impose the dual requirement that it not unjustly discriminate between any user for any one of these services, and conversely that it not provide unduly preferential terms to favoured parties (especially including, but not limited to, itself). Following either path would also require the vast trove of first-party data that Google has acquired on billions of Gmail, YouTube, Android and Chrome users be separated from the operation of its online advertising system. The current complaint against Google by the DoJ and eleven states (2020) in the US, in fact, requests that the reviewing court “[e]nter structural relief as needed to cure any anticompetitive harm” (57).

The ultimate sword hanging over Facebook today is the potential that it could be required to spin-off Instagram and WhatsApp (US 2020b). While Facebook complains that these cases seek a do-over on deals that were approved years ago and that they are based on a revisionist history of how antitrust law works in the US (Newstead 2020), as we saw in the forced spin-off of Western Union by AT&T a century ago, such actions are neither novel nor a break with past antitrust conventions.

In Germany, rather than pushing for the break-up of Facebook, a Federal Cartel Office ruling in 2019 imposed functional separation rules that require the company to erect a firewall between its flagship service, Instagram and WhatsApp, respectively (Bundeskartellamt 2019). This action in Germany took place on top of a trilogy of recent decisions by the European Commission that have penalized Google for abusing its dominant market power in online search and shopping services in 2017 (€2.3 billion fine billion), its Android mobile operating system in 2018 (€4.34 billion fine), and the online advertising market in 2019 (€1.5 billion) (EC 2017; EC 2018; EC 2019) as well as the coming into effect of the EU’s General Data Protection Regulations in 2018.

It is indeed heartening to see structural remedies that have been out of favour for decades now contemplated in the recent volley of antitrust complaints and abuse of dominance cases lodged by governments on both sides of the Atlantic against both Google and Facebook, and that two of the leading intellectual figures behind the revival of the US anti-trust movement—Lina Khan and Tim Wu—are being tapped to help lead the FTC.

Public Obligations: The French communications regulator, ARCEP, is building on this momentum by focusing on neutrality issues across the internet stack—internet access, platforms, app stores and devices (France 2018). This is a promising development that draws lessons from the history of telecoms regulation but without treating the platforms as common carriers, not least because they do not serve as gateways to the whole Internet.

ARCEP’s focus suggests opportunities to put key principles of common carriage at the heart of platform regulation. At the top of the list is what Pasquale (2016) calls the “presumption of inclusion” for all legal content, applications and services in the search, social media and apps stores of the large platform companies (498). Moreover, as he argues, “massive internet platforms must take the bitter with the sweet: if they want to continue avoiding liability for intellectual property infringement and defamation, they should welcome categorization as a conduit for speech, rather than speaker status itself” (543).

In Germany, a nascent “fair carriage” obligation is based on the legal and political premise that citizens have a positive right to express themselves, and it is the government’s role to ensure that private actors that offer public communication services must respect such rights. This approach allows platforms to moderate their services but limits their scope to do as they please insofar that citizen’s lawful expressions and interactions must stay up, unless a proper and just explanation of why it has been removed, and will stay down, is offered and defensible in court (Ketteman and Tiedeke 2020: 9–11). Germany has also proposed “platform neutrality” rules for large commercial audiovisual platforms (e.g. Netflix and Hulu, but not YouTube or those used for private ends) and for the ranking and sorting algorithms of the biggest social media services (e.g. Facebook). This effort will no doubt encounter hard cases where, instead of wanting the platforms to be neutral, some will want them to actively discriminate against, for instance, disinformation in favour of “quality journalism” (Helberger, Leerssen and van Drunen 2019). However, the outcome in such cases will turn on whether the activity/expression at issue is legal and the specific exceptions to the rule made by policy-makers on public interest grounds rather than by private fiat, and also if the action a platform takes toward such expressions is a form of just or unjust discrimination. In difficult cases where sources have been blocked, or “de-platformed”, courts can determine whether the speech is legal or not and if the platform’s action are just and reasonable.

The public obligations dimension of this approach to Internet regulation could also focus on “opening the black box” of the platforms and other online service providers by imposing regulatory oversight over, for example, terms of interconnection and interoperability and setting common technical standards so as to reduce operating and switching costs for both end users and third parties who rely on the platforms to offer their services to others. Seen from this angle, similar to the auditing and reporting requirements that banks and publicly-traded firms must meet, something like a Digital Platform Commission could oversee a certified annual audit of these firms’ blackboxes.

To its credit, the BTLR (2020) report in Canada also proposes such public obligations in relation to the concept of electronic communications services that it proposes to cover services such as WhatsApp, Skype, Facetime and Wechat that are functionally similar to telecoms services. To this end, the BTLR report recommends that such services would be required to carry all legal content and prohibited from controlling or influencing the content or meaning of messages and/or unjustly discriminating between different classes of speakers and users. The report also proposes that such services be required to adopt non-discriminatory interoperability and interconnection practices and to protect the privacy of individual users based on the standards set down by common carrier rules and privacy law. Similar to the common carrier principle, these recommendations do not ignore the reality that all technologies are socio-political artefacts but distinguishes between what is reasonably necessary to offer such services while preventing unjust discrimination. All of this would be overseen by a revamped CRTC, which it dubs the Canadian Communications Commission (CCC) to reflect the proposed expansion of the regulator’s remit to cover digital platforms.

Australia’s just passed News Media and Digital Platforms Mandatory Bargaining Code can also be read through the lens of public obligations. While its opponents have tried to cast the code as imposing a “link tax” on Google and Facebook that makes them pay news companies each time they—or their users—display a link to a news story, this is misleading. The ‘crown jewel’ of the code is that it enables the ACCC to compel Google and Facebook to carry designated Australian news services for a to-be-negotiated fee (i.e. it is a limited “must carry regime”). It also requires them to give Australian news outlets fourteen days advance-notice of algorithmic changes that could up-end their operations. It also obligates them to share more monetizable audience data with national news providers. The code will also let the ACCC peek inside how Google and Facebook’s business models and algorithms work, and expand its remit to cover other platform players, with hints that Apple’s App Store may be next in line. It also puts the ACMA in charge of overseeing the Code’s application in practice.

If it is successful, the Australian Code has the potential to lessen news media companies’ dependence on the platforms and to break through the “attention trap”: namely the condition where media companies invest substantial resources to make their services “platform ready” only to see those efforts frustrated by frequent, unannounced changes to the platforms’ operating logic and technical code, unreliable audience metrics and valuation models and, crucially, only a few new subscriptions and a trickle of new revenue despite increased traffic to their own websites (Myllylahti 2018).

Public alternatives: While other chapters elaborate on this possibility, a few broad-brush strokes will help to sketch out the role that public alternatives might play in new approaches to Internet regulation.

The first aspect entails the creation of a public data trust based on the idea that privacy and personal data are public goods. In this view, while personal data (at least in non-EU contexts) is currently harvested without any effective regulatory limits, based on the “surveillance capitalism” model (Zuboff 2019), a public alternative might start with the premise that the amount of data collected, retained and traded should be minimized to that necessary to provide a functional service. As such, instead of casting data in terms of individual consent, public repositories of personal and social-environmental and meta data could then be created. Access to that data could be governed by personal and privacy data protection standards that take the EU’s General Data Protection Regulation as the baseline applicable to all service providers—including telecoms operators, media companies, data brokers, political parties. They should also abide by information fiduciary obligations similar to those that health care providers and doctors, lawyers, teachers and other professions follow with respect to their “clients” (Kerr 2002).

These standards would replace the current situation, based on an inscrutable “dirty web” rooted in rival proprietary technical standards, unbridled data harvesting and fraudulent representations of audiences that even one of the world’s largest advertisers, Proctor and Gamble, condemned in 2019 “for its lack of transparency, fraud, privacy breaches as well as violent and harmful content placed next to ads” (De Vynck and Frier 2019). This same “dirty web” has also been hijacked for dis/misinformation operations in recent years, with damaging consequences for the standards of trust and truth upon which the fate of democracy itself depends (Ghosh and Scott 2018; Mckelvey 2018). On all these issues, broadcasting regulation and media policy have little to say.

Those creating a new generation of Internet regulation must also wrestle with the reality that the same domestic communication and media groups who are in many ways driving the policy agenda have also taken a miserly view of public goods. The same can also be said for platform companies, where projects like Google’s News Showcase and Facebook’s News reflect their preference for voluntary initiatives over public policy solutions. Independent journalism and general news are public goods that the market and the general population have never paid the full cost of providing. As Victor Pickard notes in chapter 2, the current juncture presents a perfect opportunity to reimagine the possibilities for ensuring that the public good nature of journalism and news, and their democratic roles, obtain the public support they need to thrive beyond their problematic reliance on ad revenue.

It is also essential to further the development of public service broadcasters as public service media services. In this vein, a reformist agenda requires that they be adequately supported—financially and politically—to meet their mandates. Elsewhere I have offered a more radical proposal: create the Great Canadian Communication Corporation which would merge agencies like Canada Post, the CBC, the National Film Board and Library and Archives Canada to provide universal, affordable internet access, high quality media, information and cultural product, and a national digital archive and library (GC3) (CMCRP 2020). For inspiration, we can rethink the original goal of the US Post Office for contemporary times, namely to bring “general intelligence to every man’s [sic] doorstep” (John, 2010). In terms of funding, the G3C could operate like a crown corporation. Revenues raised from the planned-for digital services tax, harmonized sales taxes for all AVMS services and income taxes applied to the Internet giants could also be earmarked for such ends.

Conclusion: New Models for a New Era of Internet Regulation or “Poisoned Chalice”?

There are valuable and desirable goals animating what has been, or is, on the Internet policy agenda in Australia, Canada, the EU, France, Germany and the US. These include:

  1. 1.

    Ensuring a country’s digital media and Internet systems are governed not just by laissez-faire global market interests, but by sovereign regulators based on public interest policies and democratic values.

  2. 2.

    Striking at the issue of market concentration wherever it might manifest across the increasingly Internet-centric digital communications and media universe (but seldom do).

  3. 3.

    Reflecting the reality that the big digital platforms now function as shared infrastructures for the digital economy and society and, thus, the need to govern them by a minimum baseline of public obligations and values.

  4. 4.

    Harmonizing regulations for functionally equivalent electronic communications services and media content services (a fuzzier category consisting of broadcasting, online VOD, search, social media and web hosting services).

  5. 5.

    Replacing unaccountable voluntary codes of conduct by global Internet giants with formal regulations based on the rule-of-law, legitimate policy processes and democratic norms.

That said, so far, the case for a new generation of Internet regulation rests on superficial analogies of the digital platforms as media companies, cherry-picked and factually incorrect evidence, and a circumspect view of the role, function and identity of the “big tech” companies. In contrast, this chapter has argued it is more effective to consider them in light of the “Goliath vs Goliath” clashes between AT&T/Western Electric and the Western Union in the late-1870s, and between AT&T/Western Electric (“the Telephone Group”) and the “Radio and Electrical Group” later in the 1920s and 1930s that shaped the development of the modern communications and media industries. These industrial battles and the dynamics they put into motion were at least as significant as the contemporary impact of internet platformisation. This is not to say that this earlier period trumps the latter in terms of significance, however, but rather that we must keep both firmly in view and use the lessons of each to better understand the other.

To this end, this chapter has attempted to rebuild the platform/Internet policy agenda from the ground up based on the principles of structural separation, firewalls, public obligations and public alternatives taken from communications and antitrust history but repurposed for our times. These principles are intended to address market dominance in all its manifestation, to pry open the inscrutable technical systems and business models that gird the Internet giants’ influence, and to refortify public values emaciated by decades of neglect. They are also respectful of people’s cognitive and communicative abilities, free speech and free press rights and the urgent need to dismantle the unlimited data harvesting practices and weak privacy and data protection rules that have given rise to ‘the broken Internet’.

Ultimately, while the possibility of designing a new era of sovereign Internet policy and regulation that serves the public interest and democracy have thus far fallen short, no democratic government anywhere should bend over backwards for the big Internet companies. Perhaps the ultimate lesson from events so far is not that the task in front of us is impossible or undesirable but rather that the pursuit of Internet regulation for the public interest and democracy must be simultaneously more ambitious in its goals and more circumspect of who has the power to define them.