Why Monopolies Rule the Internet and How We Can Stop Them

 /  Jan. 4, 2022, 2:33 p.m.

Tech Image

Monopolies rule the internet. Alphabet owns search; Amazon runs e-commerce; Apple has the hardware; Meta controls social networking; and Microsoft dominates business software. The mantra of modern industry is monopoly, and the internet is one of its victims. Since the beginning of the millennium, the technology sector has concentrated rapidly. After decades of optimism about innovation and the internet, concentration has ushered in an age of backlash and skepticism. Technology companies have enclosed what was once an open, de-centralized system within a number of differentiated platforms, silos which entrench their dominance and lessen the vitality of social discourse. The remaining monopolies possess exceptional power to govern vast digital communities with little accountability. 

The communications system has long been essential to American democracy. The dominance of Alphabet, Amazon, Apple, Meta, and Microsoft subvert what can and should be a profoundly democratizing force. Because these monopolies deter competition, digital services are reduced in quality and created less often. Platforms like Uber and the Amazon marketplace get away with abusive working conditions and massive violations of privacy. In addition, geographic inequities become entrenched as few communities share in the wealth of the digital economy. These structures particularly exclude people of color who are denied employment, business ownership, and representation in the media. Giant corporations extract preferential tax breaks, and their lobbying captures regulatory and legislative bodies. This situation represents an incredible subversion of our democracy. Competition must be restored and a more dynamic and decentralized internet created. 

Public policy should effect these changes. The root of concentration is market failure. As a result, the tech monopolies have considerable profit margins and gargantuan market share. This bounty normally draws competitors, yet the lack of new business is evident. Something about digital markets, then, distorts the process of competition. It is widely accepted that tech companies engage in anti-competitive business practices, but more subtly, there exist natural barriers to entry that deter competitors. Where normal market processes fail, government must intervene. Policymakers must use the tools at their disposal, ranging from traditional antitrust to targeted regulation, to create competition. For this to be possible, we need a shared knowledge of why digital markets are uncompetitive and what solutions are on the table.

Barriers to Competition

Unsurprisingly, the economics of digital markets are novel and complicated. In a comprehensive report, UChicago’s Stigler Center at the Booth School of Business concluded that the tech sector “exhibit[s] several economic features that…push these markets towards monopolization by a single company”. The first of these features is anti-competitive business practices. Since the emergence of the mainstream internet, large platforms have ruthlessly snuffed out their competition. When upstarts have risen to threaten their dominance, they have aggressively acquired these competitors, often shelving their products or integrating them as additional features in their suite of services. Such was the fate of Instagram, which threatened Meta’s hold on social media. In the rare cases in which a startup could not be bought, the internet monopolists employed anti-competitive tactics to kill their opponents through wars of attrition. For example, Amazon slew the e-commerce company Quisdi through artificially lowered pricing

Meta and Amazon, along with the rest of big tech, also benefit from extreme economies of scale. This occurs because digital technologies often involve large upfront costs, but after these initial investments, the costs of serving additional customers are close to zero. As a result, there are few natural limits to exponential growth. Digital platforms are able to balloon in size and quickly improve their services using the huge reserves of data they collect from their massive user bases. Smaller companies lack equal access to this data, so they face substantial barriers to competition.

Economies of scope have similar effects because big data doesn’t follow the law of diminishing returns. Instead, the value of data grows as the breadth and depth of the information it contains increases. Digital advertising is worth more when informed by a wide array of knowledge attributes about very large groups. When Alphabet knows your age and income, it doesn’t get less value—as diminishing returns would predict—from knowing that you searched for a wedding service. In fact, it learns more from this additional piece of information, which is an example of why data often gets more valuable as it gets bigger. This tendency creates an incentive for tech companies to expand into as many markets as possible where they can collect and synthesize data about people from many sources. Google and Amazon, for instance, use data from their smart speakers to enhance their services in other markets like targeted advertising, which makes it difficult for smaller firms to compete with them. This is an example of how economies of scope create barriers to entry for smaller firms; they can’t compete in one market because large platforms like Amazon leverage data from many others.   

Network effects are the fourth anti-competitive attribute. This refers to the tendency of consumers to benefit as the number of users grows for a given service. Intuitively, you are more likely to use Facebook instead of a competitor if most of your friends and family are part of the Facebook network. While this added value doesn’t grow indefinitely, it poses a substantial barrier for digital businesses—such as social media companies and online marketplaces like Uber—that rely on aggregating users. Even though you might prefer a smaller ride-sharing platform, the fact that most drivers use Uber as a default makes it extraordinarily difficult for other firms to survive. Similarly, drivers will continue to use Uber despite increasingly poor treatment because all of their customers use its service. When new competitors enter the market, they face a “chicken and egg” problem: to compete with Uber, they need drivers, but to get drivers, they need customers. In this way, network effects lock in existing platforms and are a tremendous barrier to competition in the digital markets that exhibit them.

Because of these four economic attributes—exclusionary business practices, economies of scale and scope, and network effects—most markets in the tech sector aren’t very competitive. Instead, they are dominated by a single company that extracts high profits and market value. A small number of competitors are able to survive by differentiating their service from the dominant player. For instance, Twitter and TikTok have roughly 78 and 74 million monthly active users, respectively, in the United States (compared with Facebook’s 302 million) by offering a different kind of content. You can use Pinterest to find gifts and recipes even while checking Facebook for updates from your friends. By removing barriers to competition with smart regulation, policymakers can clear the way for more and higher quality digital services. 

Focus on Competition

These days, regulating the tech sector is a hot topic, and many ideas—some better than others—are in the conversation. For instance, recent years have seen many calls to break up big tech into smaller companies, but unlike the slew of 20th-century break ups, there’s no obvious or easy way for platforms to be separated. Standard Oil and AT&T were broken up by geography, but it makes little sense to create a separate mini Facebook for, say, Florida and California. More importantly, breaking up the large tech companies won’t address the fundamental economic factors which underpin their dominance. The mini Facebooks would still be able to employ anti-competitive tactics and exploit the barriers created by large economies of scale and scope as well as network effects. 

Reforming Section 230 of the Communications Decency Act—which shields tech companies from most legal liabilities for content posted by their users—won’t address these issues either. In fact, this would probably make them worse. Bigger companies would find it easier to weather the costs of compliance and legal liabilities than their smaller competitors. While there may be compelling reasons to reduce liability protection for digital companies, reforming Section 230 is inherently anti-competitive. Hoping to take advantage of this fact, Meta recently launched an advertising campaign calling for changes to the act. Break-up and expanded legal liabilities won’t solve the most fundamental problem of big tech: it’s big.  

Competition policy for the tech sector must directly address the economic problems that reduce competition. When markets are competitive, companies work hard to keep customers from switching to competing services. Imagine if you could choose from ten or eleven social media networks instead of three or four. The alternatives to Facebook that currently exist compete by providing different kinds of content (Twitter, Pinterest, and TikTok) and protecting privacy (Snapchat). Even with these differences, mainstream networks are hopelessly generic. They build features for very large groups without much specialization and dedicate insufficient resources to research and development. If Facebook and these smaller players had a greatly reduced share of the market, consumers could choose from a wide variety of social networks. Competition recently gave us the novel music videos of TikTok. As new companies arise, they would similarly develop new forms of community, social interaction, and entertainment to power their networks, and existing social media platforms would be forced to improve and diversify their offerings to compete. The internet would once again become centered around people since more options would mean we could hold companies more accountable for their decisions. Thus, the primary aim of tech regulation should be promoting competition. 

Creating Competitive Markets

Luckily, there are plenty of ideas out there about how to foster competition in digital markets. The first is the most obvious: update American antitrust laws. The last meaningful change to regulation of communications markets was the Telecommunications Act of 1996. More than 25 years later, technology has evolved profoundly, and antitrust has yet to catch up. In this regulatory vacuum, tech companies have systematically bought and snuffed out potential competitors. The regulator primarily responsible for overseeing mergers—the Federal Trade Commission (FTC)—waved through Meta’s acquisition of Instagram in 2012 without an investigation, and it did the same for Amazon’s acquisition of Whole Foods five years later. 

The recent scrutiny of acquisitions by the FTC is a good step forward, but it matters little when judges evaluate mergers using antitrust precedents poorly suited to the information economy. Unlike the industries of old, most tech companies do not charge money for their products, and they compete in markets with fuzzy boundaries. As a result, the standard tests for market monopolization are rapidly becoming toothless. How could the FTC have stopped Amazon, an e-commerce company, from acquiring Whole Foods, a grocer, on the grounds of reducing competition? Traditional antitrust mistakenly views these companies as existing in separate markets. The law needs to catch up with the times, so Congress should empower the FTC or a new regulatory authority with expanded resources and expertise to police the business practices of tech companies. 

To address the remaining barriers to competition—massive economies of scale and scope as well as network effects—Congress has an array of policies to consider. First, lawmakers could require platforms to share proprietary data with competitors and empower users with the ability to control the usage and storage of their personal data. If Alphabet can out-compete other search engines for advertising dollars by augmenting its service with smartspeaker data, then policymakers should extend the same level of access to other search engines. Admittedly, this “data pooling” is a somewhat radical step, but it would make competition in digital markets more about the merits of the product than having “better” data. Why should Alphabet hold an 88% share in online search without necessarily having a better search engine? 

Another policy that would work along the same lines is data portability. When you choose to leave a streaming service like, say, Spotify, you can’t take your data with you. Even though you no longer use its services, Spotify restricts your ability to transfer all the data it has about you: your playlists, liked songs, search history, music preferences, and more. This inflexibility makes it difficult for you to switch to other services because you have to spend a lot of time re-creating it on another streaming platform. The law could require Spotify to make your information “portable” between services and give you the ability to easily delete it, should you want to. Data portability would require Spotify to let you take your playlists with you to other services. You would no longer be locked into using the service because you’ve spent years listening to music there. Making data portable would empower users and make the market more competitive by reducing the cost of switching between services.

The last policy idea Congress could consider is mandatory interoperability. Computer systems are interoperable when they can easily exchange information and render it useful for end-users. Social media networks are a great example of systems that aren’t interoperable. You can’t see Twitter posts on your Facebook account, and it’s always really annoying when Android users can’t join the group chats you make on iMessage. The only reason social media sites can’t process posts or messages from other platforms is that tech companies have chosen not to let them. Network effects explain why. Meta can’t monopolize the attention of users in its network when you can see Facebook content on TikTok or other smaller competitors. For that reason, they restrict the flow of content, messages, and other information with other platforms. Regulators could require social media and other tech businesses to interoperate, and in doing so they would make it easier for new companies to compete with big tech companies by sharing in the value of their networks. One of the primary reasons Meta can get away with its monopoly is that it has exclusive control over the connections between its billions of users. If some other company could make it possible to see Facebook content while using its service, it wouldn’t matter as much that your friends are all on that particular platform. The network would be shared, and more social media companies could exist as a result. This is the pro-competition rationale for mandatory interoperability. 

The common thread uniting these pro-competition policies is that they lubricate the flow of information between digital services. Pooling would equip new search engines or social networks with the big data necessary to draw advertisers, and portability would strengthen the link between users and their data as they move between services. Mandating interoperability would establish rules for the continuous sharing of information and other useful data between platforms. All of these steps would reduce the ability of technology companies to create competitive barriers by monopolizing highly valuable sources of data like user search history or content posted to the Facebook network. As a result, these policies would ease the entry of new companies into highly concentrated digital markets and therefore lessen the dominance of Alphabet, Amazon, Apple, Meta, and Microsoft. We would see greater diversity and innovation -- a more open internet. 

A More Open Internet

The curse of bigness is that the monopoly power possessed by technology companies shields them from the consequences their decisions have on the rest of us. Many commentators have criticized Meta for its ethos of “move fast and break things,” but these criticisms miss the more fundamental problem that Meta and the rest of big tech can break things without facing any repercussions. Policies that can rewrite the rules of competition, such as updating antitrust laws, requiring data pooling and portability, and mandating interoperability, are a much-needed remedy for monopolistic concentration. Greater competition is not a panacea, but it does strike at the core of the problem. While much good furnished by the internet is plainly non-commercial, the structures which have closed it are profoundly economic. 

While a more open internet would enrich the entertainment, community, and democracy of digital life, pro-competition policy is not without its downsides. A greater flow of information between platforms would necessarily trade off with privacy and security. One possible form of mandatory interoperability would be a requirement that platforms maintain open Application Programming Interfaces (or APIs, for short). Opening APIs would allow a third-party company to, for instance, aggregate all of your direct messages and texts into one interface or mashup social media feeds so you could see tweets and tiktoks on the same app. These trivial examples demonstrate the innovation that interoperability would allow. However, several open APIs were also why Cambridge Analytica was able to collect the personal data of 50 million Facebook users. A more complex information ecosystem simply entails more vectors of danger, so risks to privacy and security like the case of Cambridge Analytca will require a great deal of care during the implementation of any pro-competition policy. 

But these problems are far outweighed by the fruits of competition and decentralization. Can you imagine an internet where most parts of it are not associated with one or two big names? A crucial reason why Facebook violated the privacy of 50 million users is that the platform is so big! When people can easily buy things from six or seven e-commerce platforms, or query eight or nine search engines, they see more diversity and better service. Competition breeds innovation. When companies have to win the patronage of their customers, they invest more in research and development, create novel products and services, and glean better insights from the data they collect. Now, because monopolies dominate most digital markets, disruptive firms struggle to find funding, and overall investment is lower and directed toward refining existing technologies. New forms of entertainment and community have failed to emerge due to the dominance of Alphabet, Amazon, Apple, Meta, and Microsoft. If the federal government were to require more sharing of information, the internet would be a richer and healthier place. 

It is time to move fast and fix things. While risks do exist, the status quo is broken; monopolies rule the internet. Technology isn’t inevitably good for democracy, and the current concentration attests to this fact. The internet is a powerful force, and used for pro-social ends, it would help revitalize American social discourse. A more open internet would relieve numerous social harms, from geographic inequity and violations of privacy to abusive working conditions and capture of government. Policymakers must strive to make the web work for democracy, and the best tools at their disposal involve promoting competition. People should rule the internet, not monopolies, and policy can make that a reality. Democracy would be better for it.

The image featured in this article is licensed for reuse under the Creative Commons Attribution 2.0 Generic license. No changes were made to the original image, which was taken by tec_estromberg and can be found here.

Zander Arnao


<script type="text/javascript" src="//downloads.mailchimp.com/js/signup-forms/popup/embed.js" data-dojo-config="usePlainJson: true, isDebug: false"></script><script type="text/javascript">require(["mojo/signup-forms/Loader"], function(L) { L.start({"baseUrl":"mc.us12.list-manage.com","uuid":"d2157b250902dd292e3543be0","lid":"aa04c73a5b"}) })</script>