We are now well into America’s Second Gilded Age. The First Gilded Age was the era of industrial capitalism that begins in the 1870s and 1880s and continued through the first years of the 20th century, ultimately giving way to the reforms of the Progressive Era. The First Gilded Age produced huge fortunes, political corruption and vast inequalities of wealth, so much so that people became concerned that they would endanger American democracy.
The Second Gilded Age begins, more or less, with the beginning of the digital revolution in the mid-1980s, but it really takes off in the early years of the Internet Age in the mid to late 1990s, and it continues to the present day characterized by the rise of social media, and the development and implementation of algorithms, artificial intelligence, and robotics. For this reason I call our present era the Algorithmic Society.
If the First Gilded Age is the age of industrial capitalism, the Second Gilded Age is the age of digital or informational capitalism. It too has produced great fortunes and led to concerns that increasing concentrations of wealth and economic inequality are endangering American democracy. Like the First Gilded Age, it is also a time of deep political corruption and despair about the future of American democracy. It has not yet produced a second Progressive Era, yet every day I see signs that this is where we are headed.
There is a large literature criticizing the judicial doctrines of the First Amendment, and how they are slanted toward the interests of corporations (and capital generally) in the Second Gilded Age. The most obvious examples are the federal courts’ recent decisions on commercial speech and campaign finance regulation. These are interesting and important topics, but they are not the subject of this blog post.
My focus here is on the political economy of free speech in the digital age. The basic question is this: How does our political and economic system pay for a digital public sphere? It pays for it largely through digital surveillance and through finding ever new ways to make money out of personal data. Digital capitalism in the Second Gilded Age features an implicit bargain: a seemingly unbounded freedom to speak in exchange for the right to surveil, govern, and manipulate end-users.
The recent Facebook/Cambridge Analytica controversy is a characteristic scandal of the Second Gilded Age. That is because it laid bare how social media companies make their money and how they shape the public sphere in the process. The scandal also reveals a basic problem of freedom of speech in the Second Gilded Age: Digital privacy undergirds our freedom of expression, but the way we pay for freedom of expression perpetually threatens our digital privacy. This is the irony of the digital era: An era that promised unbounded opportunities for freedom of expression is also an era of increasing digital control and surveillance. The same technological advances allow both results. The infrastructure of digital free expression is also the infrastructure of digital surveillance.
The First Amendment plays a role in this story, but not the role that you might expect. One of the interesting features of the digital age is that the protection of freedom of expression has begun to detach from the judicial doctrines of the First Amendment, so that, as interpreted by courts, the First Amendment is increasingly irrelevant to the protection of freedom of speech online.
Why is that? Well, much of our practical ability to speak online depends on an infrastructure of digital communication—broadband companies, domain name registrars and registries, webhosting services, caching and security services, search engines, and social media companies. That infrastructure is owned and operated by private parties, not by the state. Thus, in most cases, the businesses that provide the digital infrastructure of free expression are not state actors bound by the First Amendment. If we want to protect the privacy and freedom of speech of digital speakers from overreaching by digital media companies, the First Amendment will not be our primary line of defense. Nor will the Fourth Amendment, or the rest of the Bill of Rights, for that matter.
Where the First Amendment is often most relevant, sadly, is in a different context. The First Amendment—or more correctly, judicially created doctrines of the First Amendment—may be a potential obstacle to laws that would try to regulate the owners of private infrastructure to protect freedom of speech and privacy. One example would be first amendment attacks on network neutrality. A second would be first amendment defenses against privacy regulations on the ground that data is speech and the analysis, sale, and distribution of data is protected by the First Amendment. The United States has not yet passed comprehensive digital privacy regulation, but when it tries to, I expect that companies will make precisely this kind of argument against passage; and they will probably use the First Amendment to challenge any such legislation in the federal courts.
In short, the First Amendment, as currently interpreted by federal courts, may be of relatively little help in securing the practical ability to speak through the privately-owned digital infrastructures of communication. To be sure, the First Amendment may be of limited help when the state tries to employ privately-owned infrastructure as its tool for speech regulation and surveillance—what I refer to below as “new-school” speech regulation. But in some cases the judicially created doctrines of the First Amendment—wrongly interpreted and extended—may actually be a positive hindrance to freedom of speech online.
Re-engineering the First Amendment for the twenty-first century will mean more than deciding whether coverage or protection should be extended or contracted in particular areas, whether state action should be expanded, whether doctrinal tests should be replaced, and so on. It will require a holistic approach to the institutions and practices that protect freedom of speech and expression in the algorithmic society. The structure of judicial doctrine remains essential but it is only one part of the story. We must adapt judicial doctrine to free-speech protecting structures, institutions, and reforms elsewhere in the system of free expression.
We might make an analogy to the protection of economic freedom at the turn of the 20th century. The Progressive Era and New Deal’s response to police power jurisprudence wasn’t simply eliminating constitutional protections for freedom of contract; it also involved recognizing that substantive economic freedom requires the creation of institutions (like labor unions) and government investment in infrastructure (like social insurance programs, educational institutions, and public works) necessary to give people practical economic freedom. Many if not most of these reforms operated outside of constitutional doctrine but were informed by constitutional values. In the same way, one needs legislation, administrative regulation, institution building, public provisioning (e.g., investment in infrastructure), and technological design to protect free expression in the 21st century. Reform of judicial doctrine is surely important —especially to allow several of the kinds of reforms I list below—but we should see it in the context of a larger program of reform, staged in legislation, administrative regulation, government provisioning, reform of business models, and technological design.
An additional wrinkle is that significant features of digital free expression are no longer controlled by nation states. For example, the European Union’s decisions about hate speech and privacy regulation have—and will continue to have—predictable effects on American free expression. To some extent we may decide that protecting free expression in the United States should piggyback on what other countries (and super-national institutions) choose to do in antitrust, privacy, and technological regulation. And to some extent, protecting freedom of speech must resist these efforts.
Currently Americans are very worried about propaganda and disinformation campaigns from Russia—and, for all we know, from other countries as well. These raise serious issues, but it is unlikely that judicial doctrines will be central features of the remedy. Rather, combatting foreign propaganda and disinformation will require two different lines of attack. The first, which is already happening as a result of public outcry, is a change in business practices by owners of private infrastructure in the United States. The second, which, alas, did not happen sufficiently in 2016, is the federal government’s use of sanctions and reprisals (short of military force) against nation states that we know are interfering with our electoral system. This approach recognizes that we do not simply face a problem of regulation—rather, we face a problem of foreign policy. Had President Obama had moved quickly and decisively to exact a cost for Russian interference, we might have established a stronger deterrent effect going forward. Although we can expect little in the way of reprisals from what is essentially a pro-Russian Trump Administration, future administrations will not be so compromised.
For all of these reasons, it is very important to distinguish the political value of freedom of speech from the judicially created doctrines of the First Amendment. To make the principles of the First Amendment live in our current age, we must look beyond the latest pronouncements of the federal courts to the political economy of digital speech, to the dangers that political economy has created for end users, and to the kinds of reforms would best protect their interests.
I’ve discussed a number of such reforms in other work. Six of these reforms are worth mentioning briefly here. Some of these reforms may require modest tweaks in current judicial doctrine, but most of them will simply require a halting of the direction in which I fear judicial doctrine is headed.
The first approach is structural telecommunications regulation such as network neutrality and open access rules.
A second approach is to limit what I call “new-school speech regulation”—in which governments attempt to regulate infrastructure providers to pressure these companies to do their surveillance and speech regulation for them. The European Union, for example, is presently engaged in turning social media companies into a private bureaucracy that enforces European speech regulations. Because new-school speech regulation involves state action, some of it may violate the First Amendment (if enacted within the United States, that is). Unfortunately, much of it may not violate the First Amendment. For example, although some features of the Section 230 intermediary immunity may be required by the First Amendment, it is likely that the First Amendment does not require an immunity as great as Section 230 currently provides. It’s also important to remember that the European Union’s regulation of social media companies has important ripple effects on freedom of expression around the world, including in the United States. Therefore dealing with the consequences of new-school speech regulation in Europe may be quite important in addition to anything that American politicians do.
A third reform is government investment in basic internet infrastructure—for example, municipal wi-fi, broadband services and basic platform operations. This, in effect, offers a kind of public option for certain basic internet services. This is important because a public option is not bound to the same business models as private infrastructure owners. On the other hand, public provision requires, if anything, even stricter controls to ensure that the state does not try to leverage its control of digital infrastructure for purposes of surveillance, censorship, and control.
A fourth, and very important reform, is a revival of antitrust and competition law in the digital industries. This too, is important because it can shake up existing business models and change incentives for digital businesses.
A fifth reform is a requirement of what I call curatorial due process for social media sites, expanding on the Manilla Principles on Intermediary Liability. The Manilla Principles require, among other things (1) clear and public notice of the content regulation policies companies actually employ; (2) an explanation and an effective right to be heard before content is removed; and (3) when this is impractical, an obligation to provide to post-facto explanation and review of a decision to remove content as soon as practically possible. Although these principles will do only limited good, they remain important backstops to the system of digital free expression.
A sixth reform, and one I’ve emphasized especially in recent work, is treating digital media companies as information fiduciaries toward their clients and end-users. As information fiduciaries, digital companies should have duties of care, confidentiality and loyalty toward the people whose data they collect, store, and use.
The scope of these duties will vary depending on the nature of the digital business; again, I’ve written in detail about these issues elsewhere. But if you want a simple example of what difference the concept of information fiduciaries would make, take a look at the recent Facebook/Cambridge Analytica scandal. It’s important to focus not only on the particular example of Facebook’s negligence in dealing with Alexsandr Kogan and Cambridge Analytica, but also on the ensuing revelations: Facebook’s practices were merely the tip of a far larger iceberg—a series of unwise decisions through which Facebook allowed its business partners to access its end-users’ social graphs. In my view, Facebook probably violated all three duties of care, confidentiality and loyalty. Facebook did not take sufficient care to vet its business partners, it breached its duties of confidentiality toward its end users, and it allowed its end-users to be manipulated by its business partners.
Why are these reforms important? Some of them directly protect free speech against public and private governance. But the indirect effects, I think, are equally important. The central way to protect freedom of expression in the digital age is to alter the political economy of free expression. That means reforming the business models of digital companies—because these methods of making money, in effect, are how we pay for the public sphere. The current cost of digital free speech these days is submission to digital surveillance and private governance. Nation states, realizing this, have piggybacked on the governing capacities of digital infrastructure companies to achieve their regulatory goals.
For this reason, I expect that three of the six reforms mentioned above will prove the most important in the long run. The first is limiting new-school speech regulation, because this restricts nation states’ abilities to leverage the capacities of digital infrastructure companies to do governments’ dirty work for them. The second is antitrust and competition law, because they can chip away at the current advertising-based and monopolization-of-attention models that have caused many of the greatest problems with social media. The third is treating digital companies as information fiduciaries, who have duties of care, confidentiality and loyalty to their end-users. This last reform is especially important because it will limit the ability of digital infrastructure companies to organize their business models around the surveillance, manipulation and abuse of end-users.