Newsletter
Suscríbete a nuestro Newsletter y entérate de las últimas novedades.
https://centrocompetencia.com/wp-content/themes/Ceco
volver
Network effects describe the impact that the entry of a new user of a product or service, or a new participant in an interaction, has on the value that other users —or participants— assign to that product, service, or interaction (Belleflamme & Peitz, 2021).
In recent years, network effects have become very important in various sectors of the economy, especially in digital platforms (Berry et al., 2019). Generally speaking, this trend occurs because in many markets participating agents benefit not only from the product or service they acquire but also from the interactions they establish with other consumers or market agents.
The agents that make up a market —”users”— form a “network” that connects them to one another (Belleflamme & Peitz, 2021). The interactions between users generate external effects; that is, the way one agent interacts with others affects the welfare of the rest of the participants. The existence of these effects makes these interactions difficult to manage. This has given rise to what is currently known as a platform. Thus, a platform is defined as an entity that enables interaction between two or more users, generating value from such interactions (Belleflamme & Peitz, 2021).
As network effects are external impacts, it is important to distinguish between the agent generating the effect (the “sender”) and the one affected by it (the “receiver”). When the sender and receiver belong to the same group of agents, the effects are referred to as within-group or direct network effects. In contrast, when the sender and receiver come from different groups, this is called cross-group or indirect network effects.
Network effects can be positive or negative:
Examples
In a broad sense, the effects a network generates on an individual may depend on the specific people with whom they interact or on their identity. For instance, in media, a person benefits from using a network as long as it is also used by the kind of people said person wishes to interact with. Conversely, in the fashion industry, some may benefit from feeling part of a collective style, while others may prefer having tastes distinct from the majority.
The presence of network effects makes user decisions interdependent. Occasionally, the presence of positive network effects can create a type of positive reinforcement known as “attraction loops.” These loops arise when an increase in users in a given network makes participation in that network more attractive, incentivizing more people to join, and so on.
A good example of a within-group network effect where attraction loops occur is the use of the application “Waze.” Waze is a traffic and GPS navigation app created for mobile devices. The app’s goal is to inform users —in real time— about travel times, route details, and all types of personalized information based on the user’s location.
An important distinction of Waze compared to other navigation platforms is the interaction it establishes with the user. Waze relies on the information provided by its users, incentivizing reports of accidents, congestion, police presence, etc. Furthermore, the app gathers anonymous information from them regarding speed and location. Therefore, practically the entire success of the application depends on the number of users it possesses.
Another aspect worth mentioning is that the use of Waze also generates positive externalities for people who do not use the application —relatively minor compared to the effects perceived by users. This is because the use of Waze helps decongest streets that usually have the most traffic during peak hours.
Cross-group network effects refer to the impact that the entry of a new user into a network has on a subset of users who belong to the same network but to a different user group than the new member (Belleflamme & Peitz, 2021). A typical example of this type of network effect is found in platforms for buying and selling goods and services (“marketplaces”), where users are divided into sellers and buyers. The effects generated between both groups can be positive, negative, or non-existent.
An example where cross-group or indirect network effects have a positive impact is found in platforms like eBay or Airbnb, where the presence of both sellers and buyers is mutually beneficial for both groups. This is known as an “attraction spiral.” In this case, a greater number of sellers and buyers increases the probability that the platform will generate successful interactions. This phenomenon is closely related to the concept of a two-sided market (however, the latter focuses on the role played by the market intermediary rather than the behavior of the network users).
There are cases where indirect effects between users of a platform move in opposite directions. This is known as the “attraction/repulsion pendulum” effect, where one group of users benefits from the presence of another group, while the latter is harmed by the presence of the former. An example of this occurs in web browsers that, upon becoming very popular among users searching for information, attract “hackers,” who negatively affect those users searching for information.
Generally, it is difficult to differentiate the various types of users interacting in the same network, and even more difficult to classify them into groups. In fact, most environments simultaneously have both direct and indirect network effects. For example, in the telecommunications industry, a user may interact intensely with a specific group of users —a inner circle— while also interacting less intensely with other types of users outside that circle (Belleflamme & Peitz, 2021).
The importance of network effects lies in their ability to influence people’s preferences, incentivizing them to coordinate and join the same platform. This is significant for two reasons. First, two platforms that are initially perceived as symmetrical in a market can end up in an asymmetric situation due to the existence of network effects. This occurs because of the intensive use of one platform to the detriment of the other.
Second, network effects can make a user’s decision regarding which platform to use to differ from what they would have chosen if their decision did not depend on what everyone else uses. For example, a computer scientist who prefers working with a paid programming language might end up switching to a free one in order to take advantage of the network effects arising from collaboration among its users.
These two reasons characterize a result known as “winner-takes-all.” Indeed, in recent years, there has been a growing interest in studying how network effects can weaken market competition (Bergemann & Bonatti, 2019; Jullien et al., 2021). This, as network effects can promote user loyalty to a platform, allowing its owners to enjoy greater market power in the short term and increasing barriers to entry in that market (Berry et al., 2019).
In this sense, referring to social media, lawyer Eric Posner and economist Glen Weyl warn that network effects have hindered the entry of new competitors, who could only challenge incumbents if they have sufficient financial backing to subsidize their platform for a long time (Posner & Weyl, 2019). Along the same lines, the European Union’s Guidelines on the Assessment of Horizontal Mergers indicate that network effects may contribute to entry not being profitable unless the new entrant can acquire a sufficiently large market share (para. 72).
In 2001, the United States competition authority accused Microsoft of illegally maintaining its monopoly position. In this case, the impact of positive network effects on application entry barriers was a relevant aspect of the court’s decision (253 F.3d 34 (D.C. Cir. 2001)).
The core argument of the lawsuit was that the size of the Windows user network pressured independent software vendors (ISVs) to develop applications primarily for Windows. This meant Windows users benefited from a wider variety of applications which, in turn, reinforced Microsoft’s dominant position, perpetuating the ISVs’ readiness to continue creating applications for the platform.
The court’s analysis concluded that Microsoft’s abuse of dominant position was the combined product of two phenomena:
Consequently, the fact that more applications were made for Windows —relative to other operating systems— not only attracted more consumers but also kept them locked in. This combined effect between the economics of application development and positive network effects ensured Microsoft’s dominant position. This case set a vital precedent regarding the impact of network effects on market concentration and entry barriers (Werden, 2001).
In 2017, the Tribunal de Defensa de la Libre Competencia (TDLC) issued a regulatory recommendation regarding services associated with the use of universally accepted credit and debit cards. It proposed legal and regulatory modifications to foster competition in the card payment industry (Proposition 19/2017).
In this context, the TDLC considered network effects by analyzing the card payment industry as a two-sided industry. According to the Court, when a cardholder uses their card, they produce a positive externality for other cardholders by incentivizing merchants to accept that payment method. Likewise, each affiliated commerce produced a positive externality for other affiliated commerces by incentivizing cardholders to use those cards. Thus, in this industry, indirect network effects generate the “attraction spiral” effect mentioned previously.
In 2018, the TDLC approved, with conditions, the Joint Business Agreements (JBA) between: (i) LATAM and American Airlines and (ii) LATAM with Iberia and British Airways (Resolution 54/2018).
In this case, network effects were especially relevant in determining the relevant market (“Network Industry”). The merger implied an increase in the routes covered by the alliance. From the passenger’s perspective, a broader network is beneficial because it offers more destinations to redeem rewards or points.
The analysis indicated that network effects have a greater impact on the time-sensitive (premium) passenger segment. A significant part of this segment was related of business travel, requiring flexibility and multi-destination coordination. The breadth of the network —more daily flights and connections— provided an advantage that was difficult for airlines with lower density coverage to replicate.
However, a year later, the Supreme Court (CS) overturned the decision, rejecting the JBAs. The CS was of the idea that the risks of the operation were substantially higher than its efficiencies. Crucially, the CS mentioned that due to the presence of network effects, it was not possible to associate individual anti-competitive risks with isolated measures. Because these risks stem from multiple interrelated causes, they can occur dynamically and reinforce each other (CS, Sentence of 23-05-2019).
