One of the biggest changes in the internet era has been the introduction of companies that can aggregate demand. To take the canonical example, Facebook is the place where most (internet-connected) users in the world go for social interaction. Each additional user makes the platform more valuable because it becomes a better place to chat with friends - I'd like to leave Facebook but my high school friends refuse to leave it, and I value them more than my desire to leave. It costs Facebook basically nothing to add another user, and thus they have a direct relationship with 2 billion people. This enables them to onboard advertisers on their platform on their terms, giving them significant control (again, each incremental advertiser has no marginal cost). This gives them the best advertising platform on the planet.
Those of you familiar with Stratechery are rolling your eyes because I just described what everyone in the Valley has known for 3-5 years. A fuller version of that description is available on Stratechery by the creator of Aggregation Theory, Ben Thompson.
Aggregation Theory is the webbiest of web stuff - you could not become an Aggregator in a pre-Internet world because no product on Earth had zero marginal costs - and even if it did, how would you distribute it widely to make up for fixed costs? Literal wars were fought because the reality of economics included marginal costs, and driving those down REALLY mattered.
Ben Thompson has recently proposed/noted that because Aggregators are new because the web is new, the way you'd regulate is very different. Thinking people who wanted to antitrust regulate Big Tech (or at least Amazon/Facebook/Google) always had a little tingling feeling that something wasn't quite right. Yes, they are huge companies and they control something just as relevant as rail/oil were in the 20th century, but the Google argument that "competition is just a click away" was absolutely true - it takes literally no effort for me to suffer through Bing. The reason why they're still dominant is these companies controlled demand and had a flywheel running through their business where their scale makes them better at serving each customer. Anti-trust law is built for a zero-sum world where you have total control or don't - the ways that companies like Facebook build power is much more subtle. I think that means they're still very very worth of scrutiny - it's just a different type of scrutiny.
Let's go to one of the first companies broken up because of the Sherman Anti-Trust Act (the Act being the most traditional and well-known implementation of the traditional anti-trust philosophy): the Northern Securities Company. This was a combination of three railroad companies that basically controlled all rail traffic to the West (via Chicago). This case is the origin of the famous tale of J.P. Morgan asking Teddy Roosevelt to not file charges and Roosevelt refusing1. The three companies merged after a very public fight which upset the public quite a bit and probably helped encourage the DOJ to file their case.
It's very obvious what the Northern Securities Company controlled: all supply of rail transportation to the Western US. The US Congress (channeling the public will) didn't want one company to control all that (and raise prices, as they were explicitly planning to do). This is despite rail companies' creation of new businesses that would not have been possible pre-rail (hell, the entire city of Chicago exists basically because of rail), and that there were alternatives to rail (wagons!). Their upfront investment and innovation in transportation were not regarded as sufficient justification for them to be given the ability to abuse their control of all supply3. There's a moral dimension here - it's just not fair - but also a national self-interest one. Capitalism should be a profitable endeavor for the individual to incentivize growth, but if the individual begins taking growth from other areas via their dominance in one, capitalism is failing the national self-interest. Assuming you live in a place where capitalism is a just means to the national self-interest, and not an end (I'm not convinced we do), this is a bad thing.
This brings us to Apple. Let's be clear: the creation of the iPhone is perhaps the most significant moment in human history. Their app store has created billions upon billions in value, much of which they don't capture. I work for a company worth billions that would not exist if the iPhone/App Store didn't exist. They are the posterchild of the success of capitalism: they wanted money and created something great in the process, and basically, everyone is better off as a result.
You should read this before you continue.
Buuuut let's zoom in on the App Store here. Ben Thompson, in the Aggregation Theory article linked above, argued in a brief aside that Apple is an Aggregator when it comes to the App Store. Apple controls all demand for apps via its direct relationship with its customers so developers onboard on Apple's terms - which include a bunch of annoying rules that developers mostly put up with it. I think Ben Thompson2 misses the mark here. I think a better way to look at this is that Apple has total control over the supply of code to the iPhone. It's impossible to install apps they don't like, and via regulations on web engines and the ineptness of whatever engineers are assigned to mobile Safari, the web is not an adequate replacement. Telling someone to use a web app on an iPhone (and even iPads) is like telling a Chicago meatpacker in the 20th century to go use a wagon. It's a recipe for an uncompetitive product that will fail. Apple's control is not quite like a monopolized railroad because Apple doesn't exactly stop unauthorized apps at the network transit level, but via modern cryptography and their maniacal control of APIs, they achieve essentially the same thing.
In my view, this makes the anti-trust case against Apple much stronger. As explained above, however we regulate other big tech companies, it probably has to come via new laws and approaches (and no one has a great one yet). Controlling all supply though? We know how to regulate this - or at least we know how to diagnose the problem. Apple has total control of any ability to ship code to the iPhone (one might call it a monopoly), the dominant platform among wealthy (wealthy is defined here on the global scale - basically every American is wealthy on that scale) people, and part un of a duopoly, regardless. They are using that control to do what any Wall-Street-approved monopolist would do: squeeze tighter to grab more and more profits from independent developers like HEY.
The method of relief here don't seem to map between the situations as well as the sins do. There are actual physical bits to change ownership of with railroads, not so for digital goods. There are a host of solutions one could imagine; perhaps you make Apple allow 3rd party app stores? perhaps you set a ceiling for their cut? perhaps you do what Ben Thompson has suggested and allow alternative payment methods so in-app purchases have to compete on convenience? To be clear, I don't think Apple will do anything independently. Governments will have to make them do it - but I think they can do it being very confident in their diagnosis of the problem. This case is a lot simpler than the other policy issues in tech.
I haven't written anything like this before and I think I enjoyed it more than being angry on Twitter. Let me know what you think!
Makes one wonder how our current President would handle such a request.↩
Who, to be clear, is much smarter than me about this stuff.↩
Notably, later in the 20th century, after the invention of semi-trucks and the interstate, those same companies were allowed to merge. They could be as abusive as they wanted then - trucks served as a check on their power to do so. Perhaps that moment will come with Apple but an extended time of wrong in the meantime is not justified.↩