Comment by AnthonyMouse

> Good government is a public good, and elementary economics tells us that public goods are underproduced. Or, to put it another way, concentrated special interests have a larger incentive and more concentrated resources to corrupt government, than dispersed individuals have to keep it from being corrupted.

This is true, but is no reason to abandon the attempt to prevent this, because it's still a matter of degree. More to the point, large companies will engage in regulatory capture and try to use regulation to stifle competition regardless of whether individuals try to exert pressure on governments to preserve competition.

> Also, governments aren't quite monopolies as long as people have the right of exit--if you don't like your state's government, you can move to another state, and if you don't like your country's government, you can move to another country. Granted, the transaction costs for such moves are high (particularly for the latter), but many people still do it. That exerts at least some competitive pressure.

By this definition nothing is really a monopoly. If there is only one grocery store within a 999 mile radius, you could say that it isn't a monopoly because you could go to one which is 1000 miles away. It's a monopoly in practice because the cost of patronizing the "competitor" is too large.

> The real problem with a company like Google is that its users are not its customers, so the whole idea of "competitive pressure" in the usual economic sense doesn't even apply.

Even if you accept the framing then in this context the users who aren't the customers are the "product." They're the suppliers. In exchange for eyeballs they get services. The company is still subject to competitive pressure because if someone else provides better services, they lose the eyeballs. In a competitive market, companies have to compete for supply as well.

> Breaking up Google without changing the basic business model would just create huge incentives for its customers to game the system, getting pretty much the same things they are getting now, but just with more maneuvering under the table.

The "spies on you for advertisers" problem isn't really an antitrust problem, because as you point out, a larger number of smaller companies could do the same thing.

But if there were a larger number of smaller companies, they would have to distinguish themselves. One way to do that is to charge for services instead of using ads, which people like you prefer, and then you could patronize that one. Others would provide the service for free with ads, or with ads but ads based only on the search terms without tracking you all over the internet. And then customers get to choose. Right now there are two search engines and they're both of the "spies on you everywhere you go" variety.

> Another suggestion I've seen is for the government to force Google, and Facebook, and every other such platform, to open up the APIs for all their services, so anyone could write their own client.

This is for services with network effects, like Facebook messenger. But services like that aren't the real problem. It's much easier to convince most of your friends to switch to another messenger than it is to get even one of them to switch to a phone platform which is neither Android nor iOS.

And the solution there isn't APIs, it's breaking up vertically integrated companies. If Google Maps, Android and the Play Store were all separate companies then it would be easier to compete with each of them because they would be less tightly coupled with each other. A competitor could replace one of them without having to replace all of them.

> No, that's nowhere near the worst case. The worst case is something like locking down the entire Internet, so, for example, a site like this one can no longer even exist.

Which is the point of constraining antitrust rules to companies with major market share. None of them would apply to a site like this because it's not big enough. And if there was strong competition then they wouldn't apply to anyone because there would be many independent sites, none of which exceeds the threshold.

Destroying independent sites and apps is the risk of doing nothing, because then two corporations get to decide winners and losers for everyone.

It goes without saying that good antitrust enforcement is better than mediocre antitrust enforcement. But mediocre is the more likely outcome, and that's still better than nothing.