Platform Monopolies

Peter Thiel declared in 2014 that "competition is for losers" and advised entrepreneurs to build monopolies. The statement was treated as provocative, but it perfectly captured the logic of the digital economy that was already in place. Google has 88% market share in search advertising. Facebook controls over 70% of mobile social media. Amazon has over 70% of the e-book market. Together, Google and Facebook captured 85 cents of every new dollar spent on online advertising. These aren't competitive advantages within markets — they are, as media scholar Jonathan Taplin argues, as close to natural monopolies as the Bell System was in 1956.¹

The Natural Monopoly Argument

Taplin makes a straightforward case. If you told him you wanted to start a company to compete with Google in search, he would tell you not to waste your energy, time, or money. Classic economics predicts that a business with 35% net margins should attract massive new capital to capture some of those profits. None has materialized. Even Microsoft, with essentially unlimited resources, couldn't get Bing past 5% global market share. The venture capital community laughs at the idea of funding a new search engine.¹

The comparison with Bell Telephone is instructive. The telephone system became a natural monopoly because competing, incompatible networks were wasteful — a "tragedy of the commons" in reverse, where consolidation genuinely improved service. Google is a different kind of natural monopoly, one built on data network effects. More users generate more data, which improves the product, which attracts more users, which generates more advertising revenue, which funds further improvement. The cycle is self-reinforcing and essentially unbreakable from outside.

Robert Bork, in a paper commissioned by Google shortly before his death, argued that consumers could switch from Google to Bing at no cost. But as Taplin points out, this ignores the actual ecosystem. A typical Google user has Gmail, Google Calendar, Google Maps, Google Docs — all your contacts, appointments, navigation history, and documents locked in an ecosystem that represents years of accumulated data. The switching cost isn't monetary, it's informational, and it's enormous.¹

Rent-Seeking, Not Competing

What makes these platforms distinctive isn't just their market share but the nature of their economic position. Taplin defines them as "classic rent-seeking enterprises." Facebook controls a specific asset — 1.9 billion people — and extracts rent in the form of advertising premiums for access to that data. Google similarly extracts advertising premiums for access to its data on what people want. Amazon operates as a monopsony, using its control of the online book market to force prices down.¹

Yanis Varoufakis, whose analysis of techno-feudalism is already developed in Cooperative Economics, frames this in starker terms. In classical capitalism, profit comes from producing and selling goods. In the platform economy, rent comes from controlling the marketplace itself. Amazon isn't a retailer competing with other retailers — it is the marketplace, and it extracts a toll from everyone who participates. Facebook's metaverse ambitions are, in Varoufakis's reading, straightforwardly feudal: a digital fiefdom with a techno-lord.

This framing explains something orthodox economics struggles with: why market concentration keeps increasing despite supposedly competitive conditions. The answer is that platforms aren't competing within a market — they are the market. Network effects ensure nobody leaves (all your friends are on Facebook), and user-generated data continuously improves the platform's ability to extract rent.

Regulatory Capture

The most disturbing element of Taplin's analysis concerns how these monopolies protect themselves politically. During the Obama administration, Google effectively staffed key regulatory positions: the Patent Office, the assistant attorney general for antitrust, the FTC. According to the Google Transparency Project, hundreds of positions in the Obama administration were filled by Google-connected people. Democrats were "just as bad as Republicans, maybe worse."¹

The standard libertarian response — that digital monopolies are fragile and could be displaced by a better product — doesn't hold up against the evidence. Nokia and Motorola were displaced in handsets, but handsets compete on hardware innovation, which is fundamentally different from platform network effects. Taplin won't bet that Google will stop being the dominant search engine in five years, or that Facebook will stop being the dominant social network. The economics of network effects make these positions essentially permanent absent regulatory intervention.

The Abundance Question

The discourse around "abundance" — the idea, popularized by Ezra Klein and Derek Thompson, that the core political project should be building more housing, more infrastructure, more energy — has become a contested site for exactly these questions of platform power. As Dave Karpf observed, if DOGE is part of the Abundance movement, and the people DOGE is illegally firing are also part of the Abundance movement, then "Abundance ceases to mean anything at all. The term is already washed."²

The deeper problem Karpf identifies is that "abundance" has become an empty signifier that allows an alliance between "technocratic libs who hate the left slightly more than they hate fascism, and fascists who hate the left almost as much as they love fascism." The billionaire funding behind Abundance 2025 conferences and publications like The Argument suggests that the concept is being instrumentalized to forge a political coalition that crosses the traditional left-right divide — but in a direction that serves the interests of the platform monopolists who fund it.²

This connects back to the Mondragon question: if cooperative ownership works, why doesn't it happen more often? One answer is that the platform lords now own the marketplace in which even cooperatives must compete. Abundance of what, and for whom, is a question that can't be answered without confronting the structural power of platforms that extract rent from every form of economic activity that flows through them.

Corporations as Alien Invaders

Charlie Stross offered a useful provocation in 2010: we are living in the aftermath of an alien invasion, except the aliens are corporations. Corporations are "hive organisms constructed out of teeming workers who join or leave the collective," pursuing three objectives — growth, profitability, and pain avoidance — with a sociopathic lack of empathy and no loyalty to any nation. They have a mean life expectancy of about 30 years but are potentially immortal. They live only in the present, because short-term accounting regulations discourage attention to the deep future.³

The framing is whimsical but the mechanics are sharp. Corporations lobby trade treaties for favorable operating conditions. They bully individual lawmakers through campaign donations and the threat of unfavorable coverage. The resulting trade agreements define the macroeconomic climate, stripping national politicians of control over their domestic economies. "We are now living in a global state that has been structured for the benefit of non-human entities with non-human goals." Individual humans are either co-opted (you can live very nicely as a CEO or politician, as long as you don't bite the feeding hand) or steamrollered if they resist.³

The Stross framing makes the platform monopoly problem legible in a way the antitrust vocabulary doesn't. When Taplin calls Google a natural monopoly and proposes Bell System-style regulation, he's proposing to regulate an alien organism using rules designed for human institutions. The organism will route around the regulation — as Google did by staffing key regulatory positions during the Obama administration — because that's what organisms do when you threaten their survival.

The Geography of Dominance

Paul Krugman's analysis of US tech dominance adds a dimension that's oddly absent from most discussions of platform power. The US-EU productivity gap, when you decompose it by sector, turns out to be almost entirely about digital technology — and specifically about the production of that technology, not its use. Exclude the main ICT sectors and EU productivity has been roughly at par with the US since 2000.⁴

The deeper insight is Marshallian. Six of the ten highest-valued tech companies are based in Silicon Valley; the other two are in Seattle. US tech dominance may be less about American institutions, culture, or risk appetite than about the fact that, for historical reasons, the world's major technology clusters happen to be in the United States. Alfred Marshall explained over a century ago why industrial clusters persist: "The mysteries of the trade become no mysteries; but are as it were in the air." Venture capitalists on Sand Hill Road aren't inherently bolder than European financiers — they're geographically proximate to the opportunities, which makes them better at judging which risks to take.⁴

This has uncomfortable implications for any proposal to break up the platforms. If network effects are reinforced by geographical clustering, then the concentration of power isn't just a property of individual firms — it's a property of place. You could break up Google and the pieces would still be in Mountain View, still drawing from the same talent pool, still benefiting from the same knowledge spillovers. The platform monopoly problem may be even harder to solve than the antitrust framework suggests, because the monopolistic dynamics are embedded in geography, not just corporate structure.

The Bell System comparison suggests a possible path. When AT&T was designated a natural monopoly, it was required to give up all its patents in return for its protected position. Something analogous — forcing Google to open its search algorithm, or requiring Facebook to make its social graph portable — would at least address the lock-in that makes these monopolies permanent. But as Taplin notes, the regulatory capture is so thorough that even proposing such solutions requires overcoming the very power structures they're meant to constrain.