Nostrums for rostrums

IN 1949 Frank McNamara, an executive at a struggling finance company, had the idea of a charge card to settle the tab at high-class eateries. First, he had to solve a tricky problem. Restaurants would not accept a charge card as payment unless customers wanted to use one; and diners would not carry a card unless restaurants accepted it. His solution was to give away his card to a few hundred well-heeled New Yorkers: once the elite of Manhattan’s gourmands were signed up, he could persuade a few upscale restaurants to accept his new charge card and also to pay him a commission. Within a year, the Diners Club card was accepted in hundreds of places and carried by over 40,000 people.

The Diners Club may not seem to have much in common with digital giants like Facebook, Google, Uber and Amazon. But such businesses are all examples of “platforms”: they act as matchmakers between various entities and they typically charge different prices to different actors in the market. Google connects websites, consumers and advertisers, who foot the bill. Facebook does something similar for its members. Uber matches passengers and drivers, who pay the ride-hailing app a slice of the fare. Amazon brings together shoppers with retailers, who pay a fee.

The growing clout of online platforms is a boon to society but a headache for trustbusters. Platforms benefit from the power of networks: the more potential matches there are on one side of a platform, the greater the number that flock to the other side. The consequence may be a monopoly. That is normally a red flag for trustbusters, who are scrambling to keep pace with the rise of platforms (see article). But they should tread carefully. The nature of platforms means established rules of regulation often do not apply.

Think different

In a conventional, “one-sided” market, prices are related to the cost of supplying goods and services. If a business can charge a big mark-up over its marginal cost of production, a wise regulator would strive to ensure there are enough firms vying for business or, where that is not possible, to set prices in line with the monopolist’s costs. Such precepts are little use in regulating platforms. Their prices are set with an eye to the widest participation. Often consumers pay nothing for platform services—or are even charged a negative price (think of the rewards systems run by some payment cards). Pushing down prices on one side of the platform may cause charges on the other side to rise, a bit like a waterbed. That in turn may drive some consumers away from the platform, leaving everyone worse off. Such uncertainties mean that regulators must not act precipitously.

But they are right to be thinking about the unique economics of platforms. Tech giants like to claim there is no need for special regulation. The winner-takes-all aspect of networks may mean there is less competition inside the market, but there is still fierce rivalry for the market, because countless startups are vying to be the next Google or Facebook. Unfortunately, incumbents may be able to subvert this rivalry.

One of their strategies is to use mergers. “Shoot-out” acquisitions is the name given to purchases of startups with the aim of eliminating a potential rival. Many claim that Facebook’s acquisition of WhatsApp was in this category. A recent parliamentary report in Britain noted that Google had made 187 purchases of other tech firms. Trustbusters tend to ignore mergers of businesses in unrelated markets and big firms hoovering up small fry. Buyers of firms with an EU-wide turnover of less than €100m do not have to notify the European Commission. Rules that take into account how markets may develop over longer periods will be fiendish to craft. But they are needed.

A second concern is talent. Tech firms are jealous of their secrets. When their best people leave, they take ideas with them. Yet clauses in job contracts that restrict what types of work employees can do once they leave a company are also a means of thwarting the emergence of rivals. California has shown the way by clamping down on such practices.

A third issue is the power of personal data. Google is such an effective search engine in part because its algorithms draw on vast logs of past queries. Amazon can use customers’ trading history to guide its marketing with greater precision. These data troves raise barriers to entry to the next Google or Amazon. There are no easy fixes, however. Even defining who owns information is complex; making data portable is tricky.

As Frank McNamara and his heirs have found, a successful platform company finds ways of balancing the interests of the parties it brings together. Regulators of online platforms face a similar balancing act—between the incentives for new firms to emerge and the benefits to consumers of large incumbents. That will require new ways of thinking and careful judgment. In the meantime, however, the priority for trustbusters must be to ensure they do no harm.