Data Troubles

Digital Distribution in the Platform Economy

 

The landscape of digital distribution has presented “data troubles” of all varieties for the media industries. As we collectively develop new paradigms for understanding how distribution works in a global, digital ecosystem of connected viewing, the market for film and television content has become increasingly imbricated with the international circulation and trade of data. And for content providers, even as distinctions between screen industries are rapidly eroding in the current streaming-obsessed universe, the dimensions of navigation required to distribute this digital data continue to expand.

In the process, the digital distribution of content has come to rely most heavily upon online platforms. This is a category with practical, technological, and regulatory complexities that we are only just beginning to recognize. Georgetown Law and Technology Professor Julie Cohen notes that “[…] the platform is not simply a new business model, a new social technology, or a new infrastructural formation (although it is also all of those things). Rather, it is the core organizational form of the emerging informational economy.” [1] This understanding is also emphasized in Nick Srnicek’s Platform Capitalism, which historically contextualizes the monopolistic tendencies of platforms in the capital economy and argues that there are more continuities than radical novelties in this history, even as 21st century capitalism now relies on extracting and exploiting the raw material of data. While the recent discussion about platforms and their role in the digital economy has grown exponentially in film and media studies [2] and well beyond — Cohen and Srnicek are coming from legal studies and political philosophy respectively — it is the platforms’ role as a central component of the media distribution machine that I would like to focus on in this essay.

When discussing who or what entities are involved in distribution, most often the list consists of media conglomerates, festival curators, marketing executives, television programmers, and — in the last 12 years or so — hardware manufacturers, Internet Service Providers, and most certainly platforms. While many still persist in calling platforms “distribution intermediaries,” Ramon Lobato has helped us to understand the role of platforms, content delivery networks, or other entities in what he terms “the messy middle of media industries” [3] that are all actively shaping the current landscape of distribution. Yet as Alisa Perren has written, “Determining the full range of intermediaries involved in distributive processes, and the types of influence they exercise over content individually or collectively, [has] become a central research challenge.” [4] While I agree with these assessments about intermediaries, I also think it is time to revisit the term.

In many ways, streaming platforms were largely intermediaries early on in their development. But at this point, they are squarely in the center of the distribution equation and not merely connecting agents or messy middlemen — they are in many cases key infrastructures and primary agents of power. In addition to distributing content, Netflix, Hulu, and Amazon, for example, are currently spending $24 billion on production and licensing, [5] so their roles must be understood in relation to more than just intermediation but also as producer/distributors. Apple is spending $6 billion on content for its new streaming service Apple TV+. [6] The company’s original content platform, iTunes, began as more of an intermediary and did not become a digital merchant or content provider until the 4th version of the platform in 2003. Moving from its earliest iteration as a revolutionizing agent in music distribution and access, through its successively cluttered menu of audiobooks, movies, television series, podcasts, college courses, and other content, iTunes is another prime example of a platform expanding well beyond the designation of intermediary, not necessarily to its benefit. As one writer put it, “Rather than the Bentley Apple had promised, iTunes became a clown car.” [7] It has become abundantly clear that we need a more complex taxonomy that represents the widely diverse functions that platforms provide: from gatekeeping and content production, to processing and/or facilitating data flows, or to mining our data for profit, among many others.

The power of platforms in digital distribution calculus has grown so much that windowing practices and the levers of power have begun to shift in ways that are favoring streaming platforms over legacy media companies and their traditional viewing spaces. For example, the domestic theatrical window is now functioning largely as a publicity and marketing space for the platform release of streamers’ original content, which is where their films collect the bulk of their money. The head of films for Amazon studios recently said as much when he explained that “Success for our films is not measured by traditional metrics or simple box office reporting […]. The theatrical release is one path for us to market a film before its Amazon Prime Video release.” [8] It necessarily follows that release dates are even being set based on when they will find the largest audience on Amazon’s streaming platform, not when it will find the largest audience in the theaters.

The battle over The Irishman — the $160 million Martin Scorsese film released in November, 2019 — is an excellent representation of these transformations and tensions that have been brewing over the last decade. This Netflix original film which stars Robert Deniro, Al Pacino, and Joe Pesci had a theatrical run of almost 4 weeks before its debut on Netflix in North America. This brief appearance in the cinema has made almost nobody happy, as it is not long enough for the directors who want their films viewed in theaters or for the theater owners who insist on a nearly three month stay in their window, but it is too long for the platform’s executives who want the film streaming as quickly as possible. The reason for the theatrical run is partially for awards consideration — a film has to have a one week run in Los Angeles County to be eligible for an Oscar. The Academy of Motion Pictures President made a recent public statement about continuing this rule, because the Academy “support[s] the theatrical experience as integral to the art of motion pictures.” [9] It is also partly because of the steadfast insistence by the theater owners and their trade organizations, and the leverage they still have in the distribution landscape. The Cineplex theater chain doubled down on this position at the Toronto International Film Festival recently, refusing to show Amazon or Netflix films in its theaters at all during the event, stating “There are hundreds of fantastic films screening as part of this year’s festival, and with all those options, we asked that our screens feature titles from studios who understand and appreciate the importance of the theatrical release model.” [10] That said, this struggle over the debut window for a film produced by a streaming platform is as much about the cachet of the theatrical space and the cultural hierarchies that are still alive and well among concerns of distributors as it is about awards. Ultimately, the sacred space of the cinema still maintains a strong allure for distributors, directors, talent, and many audiences.

There should be a happy medium, as most films dramatically taper off their earnings well short of the 72-day period of exclusivity demanded by major chains, but at this point the biggest national cineplexes in the US, Canada, and Mexico are holding firm on that window in order to combat piracy and best exploit expensive marketing campaigns for a theatrical debut without having to surrender ticket sales to audiences who will just “wait for it to come out on Netflix.” It’s not the three years of a theatrical window enforced by French exhibitors, but it is at this point an inflexible 2.5 months. It was reported that AMC and Cineplex agreed independently to a 60 day window, Netflix signaled that it would not go above 45 days, and that’s where negotiations ended. [11] Consequently, the largest North American chains such as AMC, Regal, Cineplex, and Cinemark did not show the Irishman because it was only going to be in theaters for 26 days instead of 72 days; nor would they allow any other films that are screened in their theaters to stream on sites like Amazon or Netflix within that mandated period. Traditionally, the theatrical window for such films is just 3 weeks to a month, 28 days being the outer limit that is negotiated only by someone like Scorsese. In 2018, for example, Netflix produced Alfonso Cuarón’s Academy Award-winning film Roma which had a 21-day release in independent and small-chain theaters before it started streaming on their platform and eventually screening on roughly 1,100 screens around the world. [12] Thus, it appears that arthouses will be the destination for such films until the theaters and platform companies can come to a different agreement.

And if Netflix gets its way, perhaps those arthouses (or even larger theaters) will soon be part of the platform’s empire. The company announced in early 2019 that it was making an aggressive bid to buy the famous Egyptian Theater on Hollywood Boulevard which is owned by the nonprofit American Cinematheque. The Egyptian Theater was opened by Sid Grauman in 1922 and is designated a historic-cultural monument in LA. It is a treasured piece of Los Angeles history, and the city of Los Angeles and the state of California have both provided great financial resources to restore and preserve the theater. However, because of complex negotiations with the city and state, the Cinematheque has to submit notices of transaction to sell and other fiscal audits to the California attorney general’s office. So far, they have not done so, citing “lost data and a software failure” [13] — another interesting addition to the data troubles wrought by this and other data driven platforms for the contemporary film industry. But it is likely that deal with go through and Netflix will be embedded even deeper into the historical cinematic fabric of Hollywood.

Of course, studios have owned theaters in the US since the golden age of cinema and even after they were forced to sell them off after the Paramount decree in 1948, they bought them all back by the 1980s as antitrust enforcement was abandoned in favor of neoliberal market logics. Nevertheless, this is the first purchase of a major theater by a streaming platform. Whether this merger will prove to be as indicative of a major trend as the 2011 Comcast purchase of NBC Universal (the first time a cable company bought a film studio in the US) has yet to be determined. The trend ever since the Comcast-NBC Universal merger has been to fold the operations of film studios into more television- or tech-oriented conglomerates, and then begin to eliminate “middlemen” like Netflix or Amazon by attempting to go direct to consumer with content on proprietary platforms. Recently, for example, we have seen Disney’s purchase of 21st Century Fox’s film and TV properties, and announcing its streaming platform Disney +; NBCUniversal’s plans for their upcoming “Peacock” streaming service; and WB’s merger with AT&T that brings its own promises for an in-house streaming platform for WB content. These developments have aligned Hollywood studios even more tightly with the tech companies and big data than with the theater owners and audiences as they once were.

In addition to understanding platforms for their increasingly complicated role in the distribution ecosystem, determining how these entities should or could be regulated in a global context is another research challenge of primary importance. Indeed, platforms are fast becoming arbiters of data policy in the digital age while remaining unpoliced themselves. Jose Van Dijk, David Nieborg, and Thomas Poell have recently addressed this problem with their recent work proposing new ways of reconceptualizing measurements for platform power and its accumulation and/or abuse, given their importance to social and democratic wellbeing. [14] Often, platforms evade traditional frameworks of control and exist on the outskirts of regulatory regimes. Philip Napoli and Robyn Caplan have written about how platform companies regularly insist they are not media companies, but instead merely “tech” companies, encouraging the disconnect between their image and practice in order to evade regulation and various public interest related obligations that media companies are subject to, as well as to appeal to Silicon Valley investors. [15]

The five most valuable companies in the world right now — Microsoft, Amazon, Apple, Alphabet (Google), and Facebook — are tech and platform companies. And these companies have been among the most egregious violators of our privacy and data security, even as they provide us with endless hours of content, entertainment, and engagement. In 2019, Facebook was ordered by the U.S. Federal Trade Commission to create new layers of oversight for how it collects and handles data and fined $5 billion for deceiving users about their ability to control the privacy of their personal data (and yet the company still got a pass from the FTC for its past transgressions and illegal behaviors and Facebook’s officers and directors were given immunity). However, at the same time the U.S. Securities and Exchange Commission imposed a penalty of $100 million against Facebook for making misleading disclosures to investors about the risks of misuse of user data related to the Cambridge Analytica scandal. Now a group of states’ attorneys general are going after Facebook and Google on antitrust violations. In March 2019, the European Commission fined Google’s parent company Alphabet Inc. 1.5 billion Euro for antitrust violations in the online advertising market — the third fine in three years. [16] In July 2018, European regulators fined Google 4.3 billion Euro for installing their apps on cell phones without permission. This is but a small sampling of the bad behaviors of platforms, but ultimately, these betrayals of privacy, data security, and personal liberties become most instructive when they are brought to bear on the current regulatory crisis involving our data and information stored in the cloud — the same data that constitutes the film and television shows we are streaming, the emails we send, and the information we seek out online. It is then that the vast emptiness of policies safeguarding our digital futures is truly revealed.

Since the cloud is really just another way of saying “someone else’s computer,” and all of that data and computers have to be stored somewhere, I want to first address some regulatory dilemmas related to the operations and functioning of data centers — the heart of “the cloud” and much of its physical infrastructure. Such issues illuminate how policies around data and digital freedoms are evolving in ways that are not always immediately obvious or beneficial to private citizens, even as we imagine ourselves to be simply enjoying entertainment online or surfing for information.

The jurisdiction over data in “the cloud,” i.e., how it is legally understood when it is stored in data centers and accessed remotely by end users and providers, is one of many layers of chaos that characterize how digital distribution infrastructure is regulated, or what I have come to refer to it as “Cloud Policy.” This policy landscape as it currently exists is replete with unresolvable (and often illegible) tensions between the movement of data, the protection of that data, and the determination of the rightful owner of such data. At the outset, the economics of the cloud are currently in conflict with the way we think of data protection law. According to a former FCC official, “data protection law is largely based on an understanding that you know where your data is located within particular borders, whereas the economics of the cloud is dependent on data being able to flow across borders in a fairly seamless way.” [17] The contents of the cloud have been dispersed all over the world but the jurisdiction, along with the sovereignty and governance of data, remains determined by national, regional, and even more localized policy regimes.

Data centers have found some precedent as the legally defined geographic location of data in the cloud. However, whether data is determined to exist in one place (that of one data center) or multiple locations (wherever the data are collected, processed, and stored, usually in fragments at a variety of centers) still remains unresolved and in flux. Data originates in one country, passes through and is stored in others, often simultaneously existing in several international territories and multiple state jurisdictions. As a result, this data passes through just as many national regimes of privacy laws, intellectual property laws, data processing and protection laws, and other regulations affecting the authority and control over that data while it is in the cloud. This multi-jurisdictional nature of cloud computing creates significant problems involving the “choice of law doctrine” (the set of rules used to select which jurisdiction’s laws to apply in a lawsuit) including the question of what courts can even determine or hear disputes about which laws apply — state, federal, even extraterritorial.

Further, the exact location of data in the cloud is a fundamental area of contestation and confusion: where does it exist, exactly? As data is collected, processed, stored, and distributed in an array of remote servers often located in numerous different countries, including the one from where it originated, it actually exists in numerous places and jurisdictions at various times. Moreover, if a data server replicates one’s information for safekeeping, multiple countries may have concurrent jurisdiction over it. [18] Thus, the determination of its actual location for legal and regulatory purposes remains uncertain.

And yet, location matters. Where corporations decide to place their data centers is determined by a variety of factors beyond legislation and rules governing data; they include the sophistication of local infrastructure, local tax codes, and, crucially, proximity to both affordable electricity and other energy resources. Indeed, as Julia Velkova has written, these determinations are heavily affected by “current [international] policy efforts to incentivize new data center projects with the promise of corporate tax reductions, cheap land and electricity cost packages, eased access to high-voltage electricity grids, and low-latency fibre connectivity. Policies like these have already converted the Nordic countries into central nodes in the global cloud infrastructure by hosting the data centers of Microsoft, Amazon, Apple, Google, Facebook, Yandex and global collocation providers like Equinix or Interxion.” [19] The resulting uncoordinated legal maze of transnational data flows have created quite a challenge for regulators and the courts when determining who has control over data in the cloud, at what point, and for how long. The lack of universal legal standards for a global digital ecosystem has created an incredibly hostile legal environment for the data of the future. [20] Solutions for global governance regimes remain largely impossible from a legal standpoint, as cloud infrastructure extends beyond national boundaries but jurisdiction does not…with one major exception: US data.

That is because since 9/11, the PATRIOT Act has significantly expanded the US Government’s ability to compel the handover of data, regardless of where it is located. Developed in response to the September 11, 2001 attacks, the Patriot Act “is one of the only laws that affects the entire cloud computing industry” and requires American companies to release data stored in the cloud, even if those servers are located outside the U.S. [21] Consequently, the laws of the hosting countries are rendered useless. This is but one example of how the global nature of digital data and information exchange has grown to require more expansive vision and regime/s for cloud policy than currently exists.

Into this void have stepped private corporations (many of which are platforms) that have become the tail wagging the dog in this policy landscape as increasingly, cloud policy is being dictated by Terms of Service (TOS), End User Licensing Agreements (EULAs), and agreements between providers and carriers. Of course, these are all written by corporate lawyers instead of being guided by any principled application of regulatory philosophy that embraces a duty to the public interest over private enrichment. In “The End of Ownership,” Aaron Perzanowski and Jason Schultz write that license terms and agreements “are long, inscrutable, and full of bad news. They are the Lars von Trier films of legal documents […]” and they now “function as a form of privately made law […]” that functions at the expense of consumers. [22] These “informal” policies [23] are now dictating the terms for all users and their data in the digital ecosystem, most often at the expense of consumers, and determining the limits of behavior and constellation of rights in the digital space.

The Apple-FBI standoff in 2016 over iPhone encryption is another interesting example of the tensions between private and public, and formal vs. informal policy. This occurred when Apple refused to unlock an iPhone of a dead suspected terrorist to help the FBI with their investigation of the shooter who killed 14 people in San Bernadino. Essentially, Apple argued that the company didn’t work for the government, and hacking their own technology would be bad for the brand, and bad for business. While this was an incident that received a great deal of media attention, it was also just a very dramatic example of how corporate-driven cloud policy operates in a sea of daily power struggles currently being won by Silicon Valley. It is also an example of how tragically outdated the laws regarding digital rights and freedoms are in the U.S. — so outdated that in the FBI’s attempt to compel Apple to create the backdoor into their phone’s operating system, the U.S. government invoked the All Writs Act of 1789 which was signed by George Washington, the first American president, 230 years ago. This Act carries a complex, vague, and extreme power that gives federal judges the ability to issue orders that force people to do things within the limits of the law (it has also been used extensively in the “war on terror”). It is a gap filler, where the legislative apparatus is unclear or incomplete and that description certainly applies to the arena of digital media infrastructure such as platforms, operating systems, and ‘intermediaries.’

I recognize the tremendous challenges and obstacles involved in protecting privacy dependent on globally traded data, including the cultural differences in the definitions of “privacy” and “security;” a lack of coordinated global efforts in Internet governance, and clashing international standards/absence of consensus about technological and even psychological norms for data security. However, I see these inconsistent frameworks as a welcome mat for media and other humanities scholars to think about the stakes of these policy visions, and what they really mean for our cultural definitions of autonomy, personal information, and privacy. They are also an invitation to consider the creep of tyranny in this landscape, and better understand the price of surveillance, the sacrifices and legacies of mobility and industrialization and convenience, as well as the mighty power of access, and the vulnerabilities baked in to our desire to share. This is particularly salient in this current age of what Shoshanna Zuboff has termed “surveillance capitalism,” which she defines as, among other things, “a new economic order that claims human experience as free raw material for hidden commercial practices of extraction, prediction, and sales.” As Zuboff writes, “surveillance capitalism operates through unprecedented asymmetries in knowledge and the power that accrues to knowledge. Surveillance capitalists know everything about us whereas their operations are designed to be unknowable to us. They accumulate vast domains of new knowledge from us, but not for us.” [24] Zuboff’s work helps us to understand this environment that is determining the geopolitics of data and information is the very same landscape of global content distribution.

Still, despite the fact that these issues now control over much of our culture and our national characters, the boundaries of our personal liberties, and the contours of public expression, as an object of study they have not historically been conceptualized as central to humanities scholarship and research. I have been interested in enacting that shift because after all, media and platform regulation and controlling our ‘data troubles’ is not merely about maintaining market competition from an economic, industrial, or legal perspective. It is about the public interest, the valuation of privacy, the profiteering of private corporations on the backs of public utilities and the exploitation of their customers and citizens, and the longue durée of these histories. Policy is a gateway through which all media must pass, and the arena where many of our individual and cultural freedoms are either defended or destroyed. Articulating these histories and what they mean for the future of digital citizenry helps us understand the true stakes of “cloud policy” and shines a bright light on the platforms and data that are at its core.

_How to Cite

Jennifer Holt. “Data Troubles: Digital Distribution in the Platform Economy.” On_Culture: The Open Journal for the Study of Culture 8 (2019).

CC-BY 4.0

_Endnotes