On March 4th, Applico CEO Alex Moazed appeared on the BBC World Service to discuss how big tech monopolies should be regulated. Since then, and following recent proposals to break-up modern monopolies, tech media has been buzzing with discussions about platform regulation.
These discussions have quickly become fraught because, over all, most platforms have been good for society and the economy. Amazon’s marketplace, in its infancy, democratized eCommerce for small sellers who couldn’t afford their to create their own eCommerce digital assets, like a web store. Google gave order to the wild, messy, every-growing world wide web and made it simple for website creators to connect with consumers. Apple’s App Store made it easier than ever for software developers to reach a wide base of users, and those coders can come from anywhere around the globe.
In China, where the infrastructure of commerce was less developed than in the West, Alibaba’s marketplaces have single handedly brought millions into the modern economy, as exemplified by the creation of so-called Taobao villages.
When they work well, platforms bring tremendous benefits. So platform regulation must be carefully structured to preserve the good while reining in the bad.
Often, the regulatory discussion shows an unfortunate lack of nuance in terms of how platforms work. For example, there’s no recognition that there are many types of platforms, some of which require different regulatory approaches. There’s a desire to find simple solutions that can apply broadly in every case. But these kinds of approaches are likely to do at least as much harm as good.
These details are important if we’re going to get platform regulation right. To start with, when thinking about platforms, it’s important to account for the fact that they serve two types of customers: producers (the supply) and consumers (the demand). Most current regulation focuses on linear businesses, so they focus just on consumer harm. This is why, for example, most current discussion of the harm of platform monopolies are over-focused on advertising-based business models. These business models create the most tangible potential for harm to consumers.
However, in most cases, platform monopolies typically exercise their market power against their other customer group, producers. Most analysis, especially in the media and by regulators, miss this important fact. However, there’s ample precedent for this conceptual framing of platform monopolies.
Most of the antitrust actions against platforms have in fact focused on the exercise of market power against producers, not consumers. This includes cases like Google vs. Yelp, Microsoft vs. Netscape, and just this week, Apple vs. Spotify. But it also goes as far back as the middle of the 20th century, where sellers have fought back against payment platforms like Visa, Mastercard and American Express for their exercise of market power against merchants.
The most simple example of this anti-producer dynamic is when platforms also act as linear businesses, in effect acting as a producer on their own platform.
As Alex stressed in his BBC interview, the platform’s market power and information asymmetry enables platform monopolies to compete with their third-party supply unfairly. Amazon uses third-party seller data to undercut sellers on its own marketplace with its private label brands. Google favors its own products over competitors in its search results.
Spotify’s recently filed complaint against Apple alleges that Apple has been limiting Spotify’s exposure in the App Store, and that Apple Music has an unfair advantage because it doesn’t have to pay Apple’s 30% take rate on digital purchases, meaning Apple Music can sell its service at a lower price than Spotify.
The typical frame for monopolies focusing on harm to consumers misses all of these potential abuses, because it doesn’t consider producers as customers. However, for a platform, producers are clearly a customer group. For example, Amazon clearly views sellers as a customer. Amazon sells them services – including fulfillment, advertising and more – to third-party sellers. And sellers pay Amazon a take rate in exchange for access to the Amazon marketplace – in other words, Amazon is also selling them access as a service. For another example, look no further than Uber and Lyft who compete fiercely for drivers, often more so than they do for customers.
Through this lens, these platform monopolies are harming customers, just not the customers we typically think of. The supply side of a platform is the platform’s customers too.
Lina Khan’s famous legal article “Amazon’s Antitrust Paradox” argues “that the current framework in antitrust—specifically its pegging competition to “consumer welfare,” defined as short-term price effects—is unequipped to capture the architecture of market power in the modern economy.”
We would argue that this is not true. Current antitrust law that focuses on consumer welfare is sufficient once regulators realize that those who are harmed by platforms (suppliers) are also that platform’s customers. If Amazon had a real competitor, another marketplace of a comparable size, then Amazon and that marketplace would be forced to compete for the best sellers and the best catalog of products.
Current antitrust law that focuses on consumer welfare is sufficient once regulators realize that those who are harmed by platforms (suppliers) are also that platform’s customers.
Much like there are rules in place to protect traditional consumers, there should be rules that govern how platforms must transact and compete with this new, second type of customers: their producer (e.g. Amazon’s vendors, Uber’s drivers, AirBnB’s homeowners, YouTube’s content creators, etc.)
How to implement rules that limit these abuses of market power is the question. One proposal is to completely prevent the platform owner from acting as a producer on its own platform. In other words, force the linear aspects of the platform owner’s business to operate as a separate company from the platform itself. While this solution has simplicity to it that is appealing, in practice it doesn’t make much sense.
For example, this gets tricky when looking at Apple’s iOS and Android. Can Apple not provide any default apps on its phones? Does this include ones like the Camera app that integrate closely with the phone’s hardware? How about Google’s Android, where most of the actual APIs that developers use are built into Google’s apps, in part to overcome lots of fragmentation at the operating system level. Without these Google apps, Android as we know it today would look very different and many apps would not work across most Android phones.
In Amazon’s case, can it not sell any of it’s private-label products on its marketplace? If so, what happens to retailers like Walmart who have been selling private-label products for generations?
Or let’s take another example, Lending Club. Lending Club is a lending marketplace that connects borrowers to lenders. Lending Club also plays an important role on the platform as a lender of last resort. By doing so, it’s a market maker on the fringes. When there’s a lack of producers willing to lend on the platform, Lending Club steps in to maintain liquidity and make sure some borrowers are still able to receive loans.
Doing so makes the marketplace more attractive to borrowers, who are reassured that their needs will be met on Lending Club. By attracting more borrowers, Lending Club also widens the pool of customers for lenders on the platform. Thus, by market making, Lending Club often helps its supply. A rule banning Lending Club from lending on its marketplace would hurt both borrowers and lenders. An argument can also be made that a good amount of Amazon’s linear selling has a similar market-making effect.
The simplicity that the ‘break-them-up’ solution promises falls apart once you start to get into the details. A big part of the challenge comes from understanding how the various platform types operate differently, and how that impacts regulatory solutions.
While big tech platforms will at times try to hide behind complexity to obfuscate and regulatory discussion – as big finance did before it – there is truth to the idea that in many respects, this time is different. Slapping old solutions on top of new business models won’t provide an effective remedy.
While the mental frame is similar to past antitrust regulation – the goal should be to create as level a playing field as possible for third-party producers – the details will likely differ vastly from the days of Standard Oil and the railroads. There are unfortunately no simple solutions.
So how should we approach this conversation? Here’s a good place to start for anyone looking at how to regulate a platform monopoly.
First, understand that the platform’s monopoly power comes from its information advantage over its users combined with the market power generated by its scale that enables the platform to exploit this information asymmetry.
The irony of platform models is that they often reduce asymmetries of information between market participants. However, they simultaneously create a massive asymmetry of information between the platform and everyone else. Regulators should be looking to reduce these asymmetries where possible and should be focused on preventing the platform from unfairly exploiting them to the detriment of both consumers and producers.
Next, ask these three questions:
By answering these questions, regulators can make informed rules that are both flexible and efficient. The current regulatory conversation is too focused on applying old solutions to new problems.
By accounting for the nature of how platforms work, we can adapt these old solutions to new realities, and preserve the good of platform innovation while reducing the rent-seeking harm that monopolies can cause.