By Marshall Auerback, a market analyst and commentator. Produced by Economy for All, a project of the Independent Media Institute
Senator Elizabeth Warren made national headlines with her recently released plans to break up tech giants Amazon, Google, Facebook, and Apple, making it one of her signature proposals as she campaigns for the presidency. The notion should resonate and echo in our political memory—Teddy Roosevelt made his name as the trust-buster, for going after the great monopolies of the early 20th century in the name of the public interest. Twenty-first-century populist economics in America continues to be adorned the century-old piece of political syntax, “break ’em up.”
Warren’s essential rationale is that these tech companies act as monopolies and need to be cut down in size in order to promote more competitive markets, via traditional antitrust instruments such as the Sherman Act. Furthermore, as legal scholar Kelsey Mullane highlights, Warren is also advocating “new legislation that would regulate large tech platforms and the appointment of federal regulators who will enforce our antitrust laws and retroactively terminate anti-competitive mergers.” The premise is that increased competition via vigorous antitrust activity and more regulatory activism will act as a spur to additional innovation, as it levels the playing field between large and small businesses.
Here is an even bigger goal implicit in Warren’s call for an ambitious exercise of antitrust law. She wants to use it to address other social friction points, notably improving the quality of employment, mitigating wealth inequality, better protecting data privacy and enhancing national security. In short, her policies reflect an expansion of antitrust policy away from the narrow focus of the so-called “Chicago School” (which largely uses the metrics of economic efficiency and consumer welfare to determine whether to deploy antitrust remedies), toward broader philosophical concerns over monopoly’s effects on democracy itself. The notion being that the atomization process of a highly concentrated industry sector is a moment where the market can be refashioned with more equitable rules of play to consumers, employees and stockholders. How do these proposals stack up?
Elizabeth Warren is by no means the only presidential candidate to call for a breakup of big tech, but her proposals provide the most specificity in terms of what she aims to do. Her two most notable planks boil down to this:
Companies with an annual global revenue of $25 billion or more and that offer to the public an online marketplace, an exchange, or a platform for connecting third parties would be designated as “platform utilities.”
These companies would be prohibited from owning both the platform utility and any participants on that platform. Platform utilities would be required to meet a standard of fair, reasonable, and nondiscriminatory dealing with users. Platform utilities would not be allowed to transfer or share data with third parties.
In other words, the rationale for the breakup of tech companies in part turns on “structural separation,” as journalist David Dayen explains, which means that companies like Amazon or Google “would not be allowed to both own the platform and also participate as a seller on that platform.” The idea behind this separation is that it prevents the companies from controlling content via control of distribution, which can create long-term competitive distortions, even though in the short term, consumers might benefit from the lower prices offered by a large-scale retailer like Amazon. This logic moves beyond book retailing: It is what gave America separation of ownership of movie studios and movie theaters before online streaming provided by Amazon and Netflix started taking over.
Regulators,argues the legal scholar Lina Khan, “cannot cognize the potential harms to competition posed by Amazon’s dominance if we measure competition primarily through price and output. Specifically, current doctrine [which largely is governed by consumer welfare considerations] underappreciates the risk of predatory pricing and how integration across distinct business lines may prove anticompetitive.”
Fair enough. But it may still be useful to make a distinction between platforms for advertising and selling commercial goods (e.g., Amazon) and platforms for publication/speech/information (e.g., Facebook and Twitter). In the case of the former, the business models of these platforms depend on as many clicks or purchases as possible—on size. The larger the better. Consequently, if a company like Amazon becomes too restrictive via predatory practices, it will begin to reduce the number of clicks and undermine the basis for its pricing model. This need for size means that there are some constraints on these platforms being able to favor their goods or listings over those of others but, as Khan and Warren would argue, probably not enough, which is why competition should be encouraged via antitrust.
The problem with social media platforms that are publication or speech vehicles is somewhat different, as we are seeing with increasing censorship of some voices or viewpoints. It is unclear whether breaking up Facebook, and having multiple competing platforms, actually solves the problem of, say, dissemination of “fake news.” FCC oversight and regulation would seem to be a better way of dealing with this issue.
As far as the size metric itself goes, it’s hard to understand the rationale for a cut-off point of $25 billion in global revenues as a prima facie reason to break up the tech giants. According to Dayen:
Warren’s campaign sees the $25 billion figure as a clean way to assist regulators with pinpointing market dominance. ‘It has the benefit of a clear rule,’ said one senior campaign adviser, who was not authorized to speak on the record. “We should presume if a company with over $25 billion in revenue is operating a marketplace, it has power and leverage.”
But why should regulators make that sort of presumption in a globalized economy full of companies whose revenues vastly exceed that figure? Ford, GM, Toyota and countless other automobile manufacturers have global revenues well in excess of $25 billion. Seldom have we heard calls to break up Detroit’s “Big Three” despite global revenues in the hundreds of billions. Why? Because there is a widespread recognition that these companies are facing significant challenges in a global market dominated by similarly large competitors.
Likewise with some of the global food companies, such as Unilever or Nestle. Or consider the pet food industry, which is largely dominated by three suppliers. Should we be breaking up “Big Cat Food,” or is there something about a $25 billion global revenues metric that is intrinsic to Big Tech? By the same token, is it fair to make a presumption that businesses with much smaller revenues lack potentially abusive power and leverage?
No question, it is fair game for Dayen to highlight the “incestuous” linkages between some of the institutes and think tanks that are broadly supportive of Big Tech and oppose the Warren proposals (such as the American Enterprise Institute). After all, they serve as intellectual advocates for the business models of companies such as Google. But beyond pointing out the potential conflicts of interest, it would be helpful to gain a better understanding of the global revenue metric beyond it simply having the virtue of clarity.
There’s nothing in the economic literature on antitrust that would provide any kind of empirical gauge to assess whether there is something sacrosanct about a $25 billion figure. In that sense, the figure adopted by Warren evokes the arbitrary limitson debts and deficits established for fiscal sustainability rules in the European Monetary Union’s so-called “Stability and Growth Pact” (which also have the “virtue” of clarity, even though the rules themselves are devoid of economic common sense).
In the same piece, Dayen writes:
Andy Kessler at the Wall Street Journal denied that antitrust law has anything to do with bigness.
The idea that John Sherman, author of the Sherman Antitrust Act, was not concerned with bigness would come as news to Sherman, who once said, “If we will not endure a king as a political power, we should not endure a king over the production, transportation, and sale of any of the necessaries of life.”
In truth, however, the Sherman Act was conceived by its eponymous author in large part to avert more radical measures, as was noted by William Letwin in his seminal analysis of the Sherman Act, Law and Economic Policy in America. Much as FDR conceived of the New Deal as rescue for capitalism rather than an embrace of socialism, Sherman introduced antitrust lest Congress pave the way “for the socialist, the communist, and the nihilist.” In the words of antitrust law Professor Daniel Crane: “From the Sherman Act forward… it is certain that antitrust has often been deployed as a foil to more interventionist forms of regulation. The ideological and political implications of that move are complex and not neatly housed in right/left categories.” Sherman and his fellow trust-busters (such as the Roosevelts) may have been unwilling to “endure a king” over production, transportation, etc., but they were certainly willing to tolerate a number of archdukes.
And here we come to the tension inherent in Warren’s antitrust proposals. Is the ultimate goal to break up these companies in order to foster greater market competition? If so, to what end, or is it simply enough to let markets be markets, even if market forces lend themselves precisely to the kinds of pathologies that Warren is seeking to eliminate?
As far as outcomes go, there is a benign assumption in the Warren proposals that increased competition levels the playing field between large and small businesses, which in turn helps to spur greater innovation. However, there is a significant body of research since the days of Joseph Schumpeter (the intellectual godfather of the economics of innovation) indicating that R&D spending and R&D productivity increase with scale. True, smaller firms have less bureaucracy and in theory can adapt their technology quicker to the needs of the marketplace. But a comprehensive examination of the empirical databy Professors Anne Marie Knott and Carl Vieregger at the University of Illinois leads to the following conclusion:
Not only do large firms (using the US Small Business Association definition of greater than 500 employees) conduct 5.75 more R&D in aggregate than small firms, they have 13% higher productivity with that R&D. However this merely captures the private returns to their R&D. A further benefit of large firm R&D is that it generates the spillovers upon which small firm innovation free-rides. (Emphasis added.)
The point of “small firm innovation free-rides” seems particularly salient, since smaller businesses are generally perceived in more benign terms politically than big multinational corporations. True, they do not suffer from “the curse of bigness,” but the myth of the plucky entrepreneur obscures the fact that in many instances smaller businesses are in reality the bad guys when it comes offering decent wages and a generous array of social benefits, such as health care provision or tolerance of unionization (the decimation of which has played a huge role behind the increase of wage inequality in our society).
Many businesses that are relatively small, and have minimal levels of concentration, are also marked by “low productivity levels, productivity growth rates hovering around zero, and low real wages,” according to an INET-supported studyby Professors Lance Taylor and Özlem Ömer. As a result, small businesses are often the first to protest increased regulation or “burdensome” mandates, such as those for additional health care provision or increased unionization. And that hardly makes them ideal candidates for the kinds of things Warren is aiming to achieve with her proposals.
By contrast, one of Warren’s chief targets, Google, has just mandated “that all its temporary and contractor workers based in the United States receive a $15 minimum wage by 2020 and comprehensive healthcare, including eight sick days and 12 weeks of paid parental leave, by 2022,” according to a recent piece by Jillian D’Onfro for Forbes.com. This echoes a recent move by Amazonto raise minimum wages to $15 per hour.
To be clear, there are many disturbing practices worthy of elimination in both Google and Amazon (hereand here). But it is also salient that both companies introduced the increased wages largely in response to mounting political pressure (Amazon explicitly acknowledged this), as opposed to waiting on legislation that would force the change. That provides some implicit support for the view that larger corporations generally have a greater economic capacity to enhance working conditions for their employees than smaller businesses (even if their motivations are to pre-empt even greater change that they strongly oppose).
The point is that greed, as opposed to affordability, has driven the big tech companies’ business decisions.
But what is the right legislative response? Is it to tackle the greed via vigorous antitrust enforcement or to introduce labor legislation that facilitates unionization and shifts the power balance back somewhat toward labor? Historically, unionization has proven to be a much more effective route toward greater income equality, as unions proved more successful in mitigating the gap between labor productivity and wages.
Frustratingly, Warren and her fellow Democrats in general evince little interest in tripartite labor-business-government models of the kind that worked so well during much of the post–World War II period (before market fundamentalism assumed intellectual primacy), or structural interventions in the labor market—via unionization or immigration policies, such as those embodied in Barbara Jordan’s Commission. In contrast to Trump’s unthinking (and often racist) restrictionism, Jordan’s immigration proposals sought to balance the 21st-century employment requirements of U.S. businesses with protections for “U.S. workers against unfair competition from foreign workers, with an appropriately higher level of protection for the most vulnerable in our society.” Warren and her fellow Democratic presidential candidates admirably pay heed to the less fortunate via cash transfers from winners to losers with increased public provision. But they fail to consider whether the deregulated labor and goods markets that are hallmarks of today’s antitrust propositions actually exacerbate the problems that create those vulnerabilities in the first place.
Likewise, in regard to the issue of data security, and data privacy, antitrust might not offer the best solution. The EU framework for data protection, or the California Consumer Privacy Act, have the makings of early-stage templates for consumer protections on these issues (although in the case of Facebook, this may be a case of closing the barn door after the horse has bolted, given that American Facebook users are abandoning it by the millions, the company’s “original sharing” is rapidly declining, and its WhatsApp subsidiary, the takeover of which Warren seeks to unwind, is rapidly being superseded in quality by China’s WeChat).
But what about the idea that we should all be thankful “now [that] we have the option of using Google instead of being stuck with Bing,” as Warren herself queriesin support of increasing competition among search engines? Network theory, as I’ve written before, leads to a different kind of assessment, because the value of a search engine expands as an increasing number of other users use it, thereby enhancing its value to the user. The corollary also applies insofar as the benefits of a search engine diminish as the number of users decreases, as more search engines are created (and recall the earlier point about innovation and size, which suggests that more competitive choice does not necessarily mean that a better mousetrap is created). Given that these businesses have reached levels of sophistication and capacity, as well as unique periods of benign growth and talent agglomeration, they can’t be easily replicated, so breaking them up via early 20th-century instruments, such as the Sherman Act, might not solve the problem.
If there is something inherent about network effects whereby online social networks like Facebook or search engines such as Google lend themselves to becoming natural monopolies, then the answer might be to regulate them as utilities, rather than breaking them up, as no less a figure than right-wing populist Steve Bannon has suggested.The advantage of the utility model of regulation is that it removes the determination of wages and prices of necessities from either markets or direct state control and put them in the more neutral realm of state-brokered bargaining (an approach earlier championed by both trust-buster Theodore Roosevelt and New Dealer Franklin D. Roosevelt).
Finally, it has to be asked whether it is wise to formulate antitrust policy on the basis of specific domestic requirements, while ignoring the challenges posed by global state-subsidized Chinese national champions: companies such as WeChat, Huawei, Alibaba or Baidu.
Well-crafted industrial policy is the kind of thing the American economy used to do rather well in fact before the deregulatory zeal of the Chicago School maligned any kind of national development strategies as something akin to rent-seeking crony capitalism. As a general rule, the more one uses antitrust to produce increasing domestic competition and less large companies, the more we may have to close the domestic market to international competition and provide export subsidies in order to compete globally. That won’t fly under the World Trade Organization, but without national protectionism and an assertive geo-economic strategy, the more costly becomes the application of antitrust rules based on national considerations alone.
It is true, as Lina Khan acknowledges, “If markets are leading us in directions that we, as a democratic society, decide are not compatible with our vision of liberty or democracy, it is incumbent upon government to do something.” The question is what that “something” should be. The broader philosophical issue is whether antitrust works as an explicitly democracy- and pluralism-reinforcing institution, or whether these objectives are better achieved via other legislative measures (increased unionization, more progressive forms of taxation, utility-style regulation, campaign finance reform, institutional political reforms that encourage greater participatory democracy and minimize barriers to voting, to list a few examples).
Beyond that, it is worth pondering whether the Big Tech behemoths are a product or a cause of our current democratic dysfunction. The earlier existence of progressive tripartite-type deals amid an industrial backdrop of big companies, big unions and big government suggest that today’s problems with Big Tech are symptomatic of the decay of some of the functioning stools on which America’s liberal democracy rested (declining unionization, increased political corruption), rather than the cause itself.
Unfortunately, in her efforts to act as a cheerleader for capitalismand market competition via aggressive antitrust remedies, Warren ignores other remedies of the kind embodied in the Treaty of Detroit. And her antitrust proposals acknowledge no role for possible utility-style regulation (ironically, that is coming from the populist right).
The problems she defines are real, but Warren’s call to break up Big Tech is by no means the best route to the solutions she seeks to find.