Network effects, which are any effects that a large user base has on a platform’s market dominance, have been an accomplice to antitrust complaints for some time. Today, large tech firms are subject to claims that their network effects contribute to anticompetitive harm, as was successfully alleged against Microsoft in 2001. These accusations have been utilized to promote breakups and to block mergers and acquisitions. Since then, a lot has changed in terms of technology; how regulators evaluate large online platforms should change as well.

In the 1990s, Microsoft operating systems were the dominant consumer access point to the internet and online applications. Thus, an issue arose when Microsoft began bundling its operating system with its less market-dominant web browser software. The operating systems market was such that users had few alternatives to Microsoft, making bundling an unfair use of market power.

This is often used as an example of anticompetitive network effects, which can help justify antitrust regulatory intervention. Because customers predominantly used the Windows operating system, Microsoft held immense power over the web browser software market. Internet Software Vendors (ISVs) would write software for Windows since it had the largest customer base, and in return, this would make Windows more desirable for customers.

A search engine, by contrast, exists on web browsers, and users can easily change from one search engine to another. These low barriers to entry and exit can diminish the network effects of a popular search engine in a way unlike that of operating software. Because ISVs had to usually just pick one operator to work on (single-home) and spend months or years developing their programs for it, there was a large barrier to exit, amplifying the network effects of Windows. 

However, the digital marketplace has evolved immensely since 2001. Most of these problems are nonexistent today. Users can now move from one platform to another with ease.

A report on network effects by Catherine Tucker provides an overview of the changes that have occurred in the platform space since the Microsoft case. Underpinning the ease by which users migrate from one platform to the next is the fact that platforms are no longer hardware-based but instead exist mostly online. If Facebook decides to remove a vendor from their marketplace, then that vendor is free to move to Amazon instead. This lowers the switching costs integral to maintaining market dominance with network effects. It even extends to platforms that once dominated a certain market, like MySpace, in the early 2000s. Because the switching costs were relatively low to move from MySpace to Facebook, Facebook was able to overtake MySpace and its dominant network.

Furthermore, multi-homing and single-homing play a significant role in how network effects function with competition and are vital for antitrust regulators. Multi-homing means that users will operate on multiple platforms at once. For example, someone may possess a Facebook, TikTok, and LinkedIn account. With single-homing, users can only access one platform like Windows. Multi-homing diminishes the lock-in of networks since users are not tied to any one platform. Because most platforms are accessible through multiple means, such as Androids, iPads, and PCs, all social media users can easily multi-home and usually do so since different social media platforms fulfill different consumer niches.

Not even single-homing is as effective today as it was in the 1990s. This is because the low barriers to entry and exit lead to frequent flocking between platforms. Some platforms like Yik Yak will start, quickly gain a massive following, and then disappear, even if there aren’t alternatives. It’s why platforms dominating a new niche market are not protected from eventual supplanting or irrelevancy.

Nothing is stopping users from simply looking up an alternative to Amazon’s marketplace since there are very few barriers to exit, other than perhaps needing to cancel a subscription. The Federal Trade Commission (FTC) has claimed that Amazon’s user reviews and data from customers give it anti-competitive network effects, but this is not necessarily enough to make that case.

A great example of network effects only theoretically guaranteeing success to a platform is Google Plus. The Google social media platform should have been a success since the network of users was already there, and the data necessary to make it a success already existed. Much like Amazon, there were users and data. Yet, it was a flop, outcompeted by a myriad of other platforms without existing networks.

One possible reason for Google Plus’s failure could be that in today’s digital marketplace, a network does not always guarantee success. A platform’s success often hinges on meeting a niche in the market.

However, growing a network can even hurt its appeal to the next generation. A prime example of this is Facebook, which once marketed itself as social network for college students, but slowly shifted into a platform for older generations. On paper, this development appeared to be a major win for Facebook which succeeded in expanding its user base. However, upon closer inspection, it is clear that this shift made the platform less appealing to younger demographics. Though there is nothing wrong with having a site geared toward older users, this was not the company’s original intent.

Since Microsoft, research on network effects and how they relate to competition has greatly expanded. New understandings of this relationship and how technology affects its positive and negative aspects should shape how antitrust regulators evaluate market dominance. These changes inform us how antitrust regulation ought to be applied to large online platforms like Amazon.

Isaac Schick is with the American Consumer Institute, a nonprofit education and research organization. For more information about the Institute, visit www.TheAmericanConsumer.Org or follow us on X @ConsumerPal.