We can’t crush big platforms without destroying their big social benefits – but there’s a better solution

Lesen Sie die deutsche Version hier.

In the fable “Jack and the Beanstalk”, Jack trades for some magic beans, which grow into enormous size. Jack climbs the beanstalk, and finds a dangerous giant who threatens him. Jack steals the goose that lays golden eggs, becomes rich, and lives happily ever after.

The fear of giants is not limited to fables; the threat of monopoly, and enormous firms that can convert control of markets into excessive power and profit, has long been a target of regulatory attention. In the U.S., the Sherman Act (1890) defined both excessive market share, price-fixing and other restraints on exchange, as violations that could be punished with fines or breakup of the firms themselves. In Europe, the conception of public duty was much more encompassing and fundamental, arising in large part from the German “Ordoliberalism” of Walter Eucken and Franz Böhm.

The difference in perspective is significant; where U.S. antitrust was a Progressive, center-left initiative, in Europe competition policy was an objective for liberals. Europe, and especially Germany, had taken full cartelization as the goal of 1930s government policy; the turn toward markets after the war was an attempt to reduce the market power of “giants” through competition.

But markets do not self-create competitive order, in the ordo-liberal view. Competition must be actively constituted, and then nurtured, by law. Further, private economic power can be as dangerous as state power, because cartels and monopolies threaten democracy. The state’s role is the creation of rules, and aggressive enforcement.

Still, and just like in the U.S., the core motivation is the belief that the ability of large firms to act badly will be disciplined by competition from other firms, not micromanagement of pricing decisions handed down by the state. The shared template was called “structure, conduct, performance.” The structure of the industry – if competitive – would lead firms who wanted to increase profits to lower, not raise, their prices, and to improve quality. This conduct results in benefits for consumers and workers alike.

The first major European “competition policy” was the 1957 “Acts Against Restraints of Competition” (Gesetz gegen Wettbewerbsbeschränkungen, or GWB). Also in 1957, the Treaty of Rome enabled groups of nations to take cooperative legal action to break up cartels and collusive practices, and abuse of dominant market position. These rules were set up to apply to firms, supranationally, laying the groundwork for the current practice of applying European law to American firms.

What If the Goose is a Giant?

The platform economy calls almost everything about competition policy into question. Digital platforms are not merely firms that happen to be large; they are matching institutions that sell connections, meaning that their social value increases with size. Treating such institutions as if they were traditional manufacturers misdiagnoses both the source of their power and the appropriate means of constraining it. Ultimately, if we keep the mindset that we must kill giants, we may end up losing out on the golden eggs.

Historically, transactions “take place.” That means both that they happen, and also that transactions require a location for people to meet and engage in exchange. Originally, the “place” was physical: the agora, the trade fair, the souk, the modern mall. The first large scale virtual platform was the Sears mail-order catalog. Sears didn’t sell stuff; like the souk or the mall, Sears sold “stalls,” in this case space on a page where sellers could depict, describe, and price their wares. The only important difference, conceptually, between the Sears catalog and Amazon is the medium that gives “place,” and the speed with which customers can search, purchase, and receive their items.

Such “places” have come to be called platforms, because they support and enable cooperation. Platforms sell reductions in transaction costs, along three important dimensions:

1. Triangulation – identifying a willing trading partner.
2. Transfer – executing payment and delivery.
3. Trust – ensuring that neither party exploits the other.

Platforms exist because they reduce all three simultaneously, in a way that is often bundled seamlessly with the good itself. A rider and driver, or a host and guest, could in principle transact without a platform. But Uber helps the rider find the driver, negotiate a price, make the payment, secure the ride, and trust the outcome. AirBnb helps the guest find the host, agree on a price, make a payment and security deposit, and trust the security of the stay.

But for most activities, the costs of finding one another, arranging payment, and enforcing good behavior still render exchanges infeasible. Platforms convert potential, even seemingly hypothetical, gains from trade into realized exchanges.

Network Economies, and Commodifying Excess Capacity

In economics, these cost savings are an example of “network economies.” It is easier for buyers and sellers to find each other (triangulation) if everyone is looking the same “place.” And using reviews for assurance (trust) works better if the buyers are not split up among different platforms.

In simple evolutionary terms, then, platforms want to be large. In fact, users want platforms to be giants, because that’s where real value of platform use comes from. A platform with more sellers, more buyers, and more reviews is superior along all three dimensions of the transaction costs. Entry by rival platforms is not difficult because incumbents behave badly, but because successful incumbents perform their core function – matching – faster, more reliably, and at much lower cost. The very attributes that antitrust seeks to discipline are those that create those golden eggs.

Economists also have a name for the golden eggs, of course: commodification. In the case of platforms, what is being commodified is both (1) things I have but don’t use and want to sell; and (2) things I have but don’t always use, and am willing to share.

Examples of selling include Ricardo, Amazon, and Digitec Galaxus; examples of sharing include Airbnb and Blablacar. But in both cases owners are able to turn items that were previously only used occasionally , and had otherwise to be stored, into sources of steady revenue. By turning unused, or excess, capacity into commodities, platforms spur a flurry of economic activity and substantially reduce the prices paid for a wide variety of services.

Voluntary exchanges make both parties better off; exchanges that were not possible before, because of transaction costs, now become not just possible but routine. The benefits are obvious: We have enough stuff, we just have to make better use of it, and share. It is possible to increase availability, sharply reduce costs, and reduce the space that each of us takes up by storing and disposing of an endless series of products.

But there are problems, of two types, and those problems have to be acknowledged, and if possible mitigated, if the platform revolution is to continue.

The first problem is the adjustment effects that attach to the rapid evolution of feasible exchanges. These problems are manifold: Residents – especially in tourist cities – face higher rents and fewer available long-term rental units; incumbent producers face falling prices; workers experience increased competition and income volatility. These groups are small, identifiable, and available for political cooptation; the beneficiaries, by contrast, are diffuse, and may not even recognize the potential for benefits if platforms are regulated preemptively. The result is predictable pressure for regulation that targets symptoms rather than causes – zoning restrictions, platform bans, or price controls – often worsening the underlying inefficiencies.

The second problem is not transitional, but existential: Successful platforms become giants, but giants govern access to markets. They write rules, adjudicate disputes, impose sanctions, and extract rents. In short, they function as private governments.

Public choice theory warns that concentrated power – whether public or private – tends to be exercised in the interest of those who control it. Platforms face weak competitive discipline, and political oversight is hampered by regulatory capture, informational asymmetries, and jurisdictional fragmentation. Antitrust enforcement agencies themselves are subject to incentive problems: bureaucratic expansion, political signaling, and ex post intervention driven by visible size rather than invisible benefits.

Breaking platforms up does not solve these problems. Fragmentation undermines triangulation, transfer, and trust while leaving unresolved the governance vacuum that allowed abuse in the first place.

The Missing Solution: Constitutional Constraints on Platforms

The most promising solution – largely absent from mainstream antitrust debates – is to treat platforms as inevitable giants and subject them to constitutional-style constraints rather than attempting to simulate competition.

This approach shifts the question from how many platforms should exist to how platform power should be exercised. Such constraints can include:

• Procedural due process for users and sellers, including transparency, notice, and appeal. Just knowing exactly where my data are going, and for what purpose, would be an advance. And a process for opting out, while cumbersome, would involve people more in the process.
• Data portability and reputational ownership, allowing participants to exit without forfeiting accumulated trust capital. At present, I can have an excellent reputation – reviews on Airbnb, or on Ebay – and be required to “start over” on each new platform. Being able to port over a reputation would make new entry much cheaper and smoother. Of course, that would also mean that a bad reputation would follow me, but that that was the way that villages and towns worked for millennia.
• Limits on self-preferencing, separating the platform’s role as market-maker from its role as competitor. This is fiendishly difficult, because in many cases the connections with other products or services are part of the advantage of a platform. But even just providing for the recognition that one is leaving one service for another within a platform might make it possible for competition to get a foot in the door.
• Rule stability and predictability, reducing arbitrary or retroactive governance. The rules in the “user agreement” are never read, because they are never readable. The announcements of rule changes are little better; we are all too impatient to read them. But there are two principles that would have to guide the content and enforcement of the rules: New rules apply only to actions taken after an announced effective date; and sanctions must cite the rule in force at the time of the conduct.

These measures do not require regulators to outguess market structure. Instead, they focus on constraining opportunistic behavior – precisely the concern at the heart of public choice theory.

Limiting scope rather than size

Platforms are giants because they solve triangulation, transfer, and trust problems at scale. Their size is not an accident, nor is it primarily the result of exclusionary conduct. Antitrust, designed for a different economic environment, misfires when applied to matching institutions whose efficiency depends on concentration.

A public choice perspective suggests that the central problem is not monopoly pricing but unchecked governance power. The appropriate response is therefore not forced competition, but constitutional constraint: rules that preserve the transaction-cost–reducing benefits of platforms while limiting their capacity to exploit concentrated authority.

In the original “Jack and the Beanstalk”, published in 1807, the problem was obvious, and the solution tidy: “At that instant the Giant was beginning to come down; but Jack with his hatchet cut the bean-stalk close off to the root, which made the Giant fall into the garden – the fall killed him.”

Our current problem is more complex: the platform beanstalk has grown, and it’s going to grow more; it can’t be cut down. The giants themselves are the sources of the golden eggs we all need to reduce the material and energy footprint of the consumption-driven, rather than sharing-driven, economy. By limiting the scope, rather than the size, of the giants we can try to live with them, and harness their great strength.