Antitrust in 60 Seconds: What Are Network Effects?
The 60-Second Read:
A network effect occurs when the value of a good or service increases the more people use it. Think of smartphones: It’s easier to connect with friends when they are already on the service you are using, and it’s easier for developers to release a mobile app on an operating system that already has a lot of users.
When belonging to a network creates this much value, users want to stay, and therefore some digital services seek to create these kinds of valuable networks to entice users to stick around. This means that network effects can impact how competition takes place.
Network effects can be used as evidence of monopoly power by showing that an industry has high barriers to entry. The question is how easily can a new network compete with an existing one. Additionally, a judge may find that more aggressive forms of competition are too effective due to network effects, making a borderline or normally lawful act a punishable anticompetitive act. But modern thinking and a number of counterexamples caution against using network effects to take high barriers of entry as a given without supporting evidence.
What Is a Network Effect?
Network effects began as a cluster of economic theories defined as “the utility that a user derives from consumption of a good increases with the number of other agents consuming the good” or situations in which “one consumer’s value for a good increases when another consumer has a compatible good.” When products become more valuable to consumers as more people use them, that’s a network effect, or what economists call a “network externality.” A well designed smartphone is a great product, it’s an even better product if you can use it to communicate with others who have smartphones, and it is a better product still if it is so popular that it attracts software developers to design new apps for it. The latter two features are network externalities.
Some point to the “stickiness” that network effects can bring as a problem, giving networks of a certain size an advantage over smaller ones. But we know these network effects can be overcome by the right competitors, and the resulting market change can be sudden and dramatic. Consider what happened to previous giants like AOL, Myspace, Atari, GeoCities, and Kodak. But a winner-take-all strategy is not the only way to compete. Increasingly, companies win by convincing customers to use multiple competing services at once, or “multi-home,” so they can live side-by-side with the established competition. For example, many users maintain both LinkedIn and Facebook profiles, and use Twitter and Snapchat. In short, the value of any one network isn’t lost if a user also participates in another network.
What Is Not a Network Effect?
Equally important is what network effects are not. Network effects are not the same as lock-in, although they are frequently confused because both concepts involve switching costs. Lock-in occurs when the costs to switch from one product to another are so substantial that a company can theoretically control aftermarkets for complementary goods or services (although many scholars argue that the reputational effects of exploiting aftermarket customers will cause the company to lose sales in their primary products, discouraging companies from taking advantage of lock-in).
The classic example of lock-in was found in the case Eastman Kodak Co. v. Image Technical Services, Inc., which concerned copy machines. These copy machines were expensive enough, the plaintiff argued, that customers were locked-in once they chose a particular brand. Therefore the brand had market power in the aftermarket for replacement parts because it was unreasonable to expect a customer would buy a new copier if Kodak raised the price of a replacement part.
Unlike the Eastman Kodak case, network effects don’t “lock in” a customer to a network. Instead, customers can lose the benefits of a network if they leave without their fellow customers. This distinction is important because most products with network effects have little to zero internal switching costs, but may have external switching costs that can be mitigated through either the desirability of a new product or a customer successfully persuading other customers to make the switch (or a combination of the two).
Network effects are also not the same thing as a natural monopoly. A natural monopoly is supply-sided, meaning the cost of producing a product or service declines with volume to such an extent that a new entrant cannot afford to price goods at the established competitor’s prices. Network effects are demand-driven — the more consumers demand a product the more attractive that product is for the next potential customer. Consumer demand can shift and the network effects will evaporate. This means consumers have much more control, and can ‘vote with their feet’ on what products receive the network benefits.
Are Network Effects Bad?
Network effects are about the value of relationships; they cannot be bad by their very nature. They are positive externalities that come from interactions with others on the network. These benefits can take several forms. Sometimes it is enough that you can use the product with others, like in the case of the telephone or fax machine. Other times the benefits come from the availability of compatible products.
For example, in the days of film photography, 35mm film was popular in part because you could find it at any drugstore. Therefore 35mm film cameras were popular in part because the film for it was ubiquitous. This does not mean that it was the only film, nor that it was the best film. Many photographers used medium format film standards that provided higher resolution and other attractive qualities. But these photographers had to want these benefits enough to buy film from more specialized retailers and either develop the film themselves or develop through more specialized film processors. Photographers who used medium format film had to trade network benefits for quality benefits.
The concern comes from the fact that network effects are benefits that come from outside the product or service. The fear is that these benefits will be enough that even if a popular product falls behind in quality, price, or other measure, the benefits from its network effects will be enough to prevent customers from switching to a newer and better product. This, if true, would provide a barrier to entry that could become a drag on competition. However, history has shown that the size of these barriers is somewhat blown out of proportion. Network effects do not make a product good on their own; a company must first make a product worthy of attracting enough customers to gain the network effects. And these customers can be wooed away or even convinced to be a part of multiple networks. Just look at the decline of products that benefited from network effects like AOL, Myspace, Yahoo, Netscape, ICQ, AltaVista and Friendster, and, conversely, the rise of companies that started without networks and came to replace them.
How Does Antitrust Law Treat Network Effects?
The theories described as network effects are relatively new, and therefore there is nothing that comes even close to “black letter law” in how judges are to treat them. The theories began developing in the 1970s and economists made great strides in developing network effect theories in the 1990s. The discussion among economists today largely revolves around the degree to which network effects impact competition. Indeed, a forthcoming paper by economists David Evans and Richard Schmalensee uses a number of counterexamples to make the argument that network effects, from platforms to formats to “big data,” do not make established companies immune to competition nor present an insurmountable barrier to new competition.