Hacker News Reader

Comment by Osiris | Hacker News Reader

In an unregulated capitalist economy, the inevitable end result is monopoly. The best way to compete in the economy is to not have to compete.

Without external restraints, any sufficiently large company will focus on eliminating competition either through pricing them out of the market or just outright buying them.

The only reason the US has more than one cellular network right now it because the government has prevented multiple attempts at mergers, and even with that effort we're down to three.

edit: if you disagree, I'd love to hear your opinion. The government has entire agencies dedicated to maintaining competition in the economy. If monopoly was not the inevitable end result of capitalism, then such agencies would not be required.

View on HN

> Without external restraints, any sufficiently large company will focus on eliminating competition either through pricing them out of the market or just outright buying them.

The counterargument is that this only happens in markets with high barriers to entry, which in turn are mostly caused by regulatory barriers.

For example, the government makes permanent spectrum auctions, so if all the holders of spectrum merge together, no new competitors can enter. Suppose instead they held a new auction every year or every month for use of the spectrum in a given region until the next auction. Then mergers wouldn't prevent competitors from entering; a new competitor goes to the next auction, places a bid and can get spectrum to launch a competing provider.

It's clear that this theory is sometimes true -- that regulatory barriers can do this. Whether it's always the cause is hard to say and depends mostly on how you define what "regulatory barriers" means. But consider the consequences either way.

If there are regulatory barriers causing monopolies, you have to break them up. If there are some other things causing monopolies, you have to break them up. Monopolies are bad. And breaking them up, if they are caused by regulatory barriers, can hamper the entities lobbying to keep the regulatory barriers.

The question is what comes after. You don't want them to just reform again, right? Antitrust is a necessary evil, but still an evil -- still means you had a monopoly for long enough to have to bust it up. So if it's mostly or entirely regulatory capture which is causing them, we still need to address that. And then we don't need antitrust enforcement as often.

But we still need it whenever a monopoly/oligopoly forms. That's just not a thing that can remain.

>> Without external restraints, any sufficiently large company will focus on eliminating competition either through pricing them out of the market or just outright buying them.

> The counterargument is that this only happens in markets with high barriers to entry, which in turn are mostly caused by regulatory barriers.

Not necessarily. For instance: if the barriers to entry are zero but the monopoly can price things below cost (e.g. due to cash reserves or strength in other markets), it can bury any competitor. The mere existence of that ability could have a deterrent effect that prevents competitors from forming in the first place.

> For instance: if the barriers to entry are zero but the monopoly can price things below cost (e.g. due to cash reserves or strength in other markets), it can bury any competitor. The mere existence of that ability could have a deterrent effect that prevents competitors from forming in the first place.

If barriers to entry are zero, the deterrent doesn't exist, because the deterrent works by forcing the competitor to pay the cost of entry without ever making any profits to recover it.

Now suppose barriers to entry are low. Not even zero -- say it's $10,000. Some individual shows up, pays the $10,000, and the monopoly slashes prices so they can never recover their $10,000. But they got into it because they're a stubborn former customer who doesn't like how they've been treated, so making billions of dollars isn't the point. Their goal is to screw the monopolist. Which means all they have to do is sit on the ability to make product indefinitely and use it as soon as the monopolist attempts to raise prices above cost.

To prevent this the monopolist would have to sell below cost indefinitely. But in that case the potential competition is keeping them in check as much as actual competition would.

And this works even better when disgruntled former customers get together, so that it's not one guy willing to pay $10,000 to spite the monopolist but 10,000 customers each willing to pay $1.

Where this falls down is when the price of entry becomes billions and involves a major risk of failure.

I mean, I’m pretty sure it fails down when the theory relies on people willing to burn capital solely to stick it to someone else. That’s less like innovation and investment and more like warfare.

It also runs counter to all observed behavior. A lot of these regulatory barriers can be overcome by money and we don’t see disgruntled people/groups banding together to provide services or goods at a loss

> Which means all they have to do is sit on the ability to make product indefinitely and use it as soon as the monopolist attempts to raise prices above cost.

You're underestimating the economic tools at the disposal of large companies. Lucrative contracts with suppliers that prevent selling to competitors, vertical integration where they just buy the suppliers outright, attack advertising, poaching employees, etc. A lot of these can happen even without active malice.

Selling below cost is just one of many tools. There are many many ways an economically powerful player can strongarm competitors into submission.

None of these are foolproof (and nothing in politics or economics is), but they all make the barrier to entry higher, sometimes almost impossibly high even in the absence of regulation (that is to say thinking of barrier to entry as something that is independent of a monopoly is not a good model, monopolies are intrinsically a barrier to entry).

Exactly. Like Amazon.

Many barriers to entry are caused by regulation, but not all, and perhaps not even most.

Some problems are just hard, and tend to favor incumbents with economies of scale. Until very recently (post-GDPR), there were no regulatory barriers protecting Search, for example; nevertheless, new entrants found it very difficult to get traction because search is a hard problem. Same goes for putting up cell towers, and burying fiber lines, and maintaining an electric grid.

I'd also argue that most oligopolies aren't formed by technical or government constraints, but rather by branding and consumer behavior. Consumers usually have a desire to stick with familiar, trusted brands, and to try new brands only when they've been recommended by people they trust. This creates barriers to new entrants, because you don't start out with any baseline trust and you have to compete against companies that have been in use for a century. There's nothing technically difficult about bottling carbonated sugar-water; regardless, Coca-Cola enjoys enormous margins because it's ingrained in the habits of hundreds of millions of people.

When there's actual regulatory capture, we should address that, and untangle the regulations so new firms can enter. But don't expect that solving regulatory capture is going to fix all instances of monopoly.

Microsoft, with massive resources, has been trying to break the Google search monopoly for over a decade without making a dent.

Search startup DuckDuckGo uses Bing on the backend rather than building their own.

Spectrum is kind of a weird case though because it kind of has to be monopolistic - if you allow multiple, independent systems to share the same frequency band, none of them will work.

> Suppose instead they held a new auction every year or every month for use of the spectrum in a given region until the next auction.

This would work if the government or some common third party were responsible for putting up towers (and lease them to the telecoms like MVNOs), but as it is the telecoms build and operate the towers themselves, and the hardware loadout changes slightly depending on which frequency bands that telecom is allowed to use.

> Spectrum is kind of a weird case though because it kind of has to be monopolistic - if you allow multiple, independent systems to share the same frequency band, none of them will work.

In theory you could actually do this by using TDM with real time auctions for spectrum access.

But "monopoly" is still the wrong word because you don't have to allocate 100% of the spectrum to the same company. If there are a hundred and they each have a "monopoly" on a specific 1% of it, they're still in competition with each other.

> This would work if the government or some common third party were responsible for putting up towers (and lease them to the telecoms like MVNOs)

So have the third parties do that. "That would work if we do that thing that would work" means it would work.

> The counterargument is that this only happens in markets with high barriers to entry, which in turn are mostly caused by regulatory barriers.

There are hundreds of entire industries that have gigantic barriers to entry no matter what the regulations are. Covering the country in cell towers, building fleets of container ships, designing an operating systems and full suite of applications — these things aren't cheap

That's only because you're defining the market as "has an entire fleet of container ships," which is just assuming the existence of a consolidated industry.

Each container ship could be its own entity and then you just need marketplaces for people to hire them to ship things.

You don't have to build a national cellular network, you can put up one tower and then sell service to a market of MVNOs which buy service from different tower providers in different regions.

Every piece of an operating system doesn't have to be built by the same entity. See Linux. For that matter, see macOS, which is full of BSD code written by many separate entities.

You're grossly underestimating transaction costs. It is vastly preferrable from a company's perspective to be able to make 1 deal with 1 company that has many ships, as opposed to many deals woth many companies that have 1 ship each. Reducing that cost ends up looking like Uber, where you've just moved your monopoly point to the controller of the marketplace.

Each individual deal an entity makes takes time and money.

The transaction costs problem is when you have 100,000 ships and 100,000 customers and it becomes ten billion individual relationships.

You do solve this with things like Uber, but such things don't themselves have to be a monopoly, because you don't have to make it 100,000:1, you can make it 100,000:20 and have twenty competing marketplaces but still not an explosively large number of separate relationships.

This works even better with federated marketplaces. There are a lot of problems with the international banking system, but the general idea that someone in New York can make a wire transfer to someone in London even though they each use a different bank is based on protocols rather than a single centralized world monopoly bank, and in that respect is better than the more centralized alternative.

> which is just assuming the existence of a consolidated industry.

Well, consolidated industries currently exist, and they don't take too kindly to competition.

What I mean is, it assumes the industry has to operate as it does when consolidated, i.e. that the market can only exist as conglomerates with fleets of container ships and not individual container ships shipping freight as separate entities.

It's like saying the trucking market has high barriers to entry because you have to buy a fleet of trucks, which only billionaires can afford, and individual owner operators are impossible. Obviously not.

>Monopolies are bad.

Not entirely... there's a tradeoff. People forget that monopolies can be a good thing in that it can force a standard. For example, if English were said to have a monopoly on language, it would become very easy to speak and communicate around the world, as it would be the only language spoken everywhere. Contrast, if we have 200 versions of spoken languages, we'd have communication barriers to overcome when we try to communicate with someone foreign.

> For example, if English were said to have a monopoly on language, it would become very easy to speak and communicate around the world, as it would be the only language spoken everywhere. Contrast, if we have 200 versions of spoken languages, we'd have communication barriers to overcome when we try to communicate with someone foreign.

You don't need a monopoly to do this because people can be bilingual. English is the de facto language of international websites on the internet, so people in Spain who speak Spanish and English and people in Russia who speak Russian and English go to Twitter or Reddit and talk to each other in English.

The same is true of internet standards. Everybody uses TCP/IP. Everybody. But nobody has a monopoly on it and nobody is prohibiting anyone from using a competitor out of anything other than network effect. If IPv8 turns out to be super amazing and has many advantages that overcome the inertia of IPv4, everyone could switch. Which is the thing you don't get with a corporate monopoly.

>You don't need a monopoly to do this because people can be bilingual

True, but it takes more work/effort to learn two languages than one. Whereas if we had one language (as in a monopoly), it'd be even simpler/easier.

>The same is true of internet standards. Everybody uses TCP/IP.

Yes, if protocols were businesses, they then could be thought of as having a monopoly on their domains.

There are not enough similarities between protocols and businesses for that analogy to work or have any explanatory value.

Except that there is:

>a dominant protocol of a certain domain where there's no other widely used alternative: TCP/IP

>a dominant business of a certain industry where there's no other widely used alternative: monopoly

I think you might be missing the point I'm making. If you carefully re-read what I wrote, I never said monopolies came without bad properties or were immune to being bad. I also never said protocols were in fact the same things as businesses. I merely pointed out a good tradeoff people aren't acknowledging: Monopolies can provide a common, standard of something which can make certain things easier (just like TCP/IP does). And that's not the same as saying: "TCP/IP comes without problems..."; "monopolies come without problems..."; "TCP/IP is a monopoly..."

> The only reason the US has more than one cellular network right now it because the government has prevented multiple attempts at mergers, and even with that effort we're down to three.

There actually was a monopoly in the telephone system[0] in the US that had to be broken up via antitrust.

[0] - https://en.wikipedia.org/wiki/Bell_System

And over the last 40 years has slowly consolidated back again.

Thanks, Reagan.

The AT&T breakup actually happened under Reagan, and was initiated in 1974 (which was Nixon/Ford). The reconsolidation happened under both parties throughout the 90s and 2000s.

...and this whole saga, lest we forget (and relevant to the HN crowd), was instrumental in giving rise to the unix wars of the 90s, due to the AT&T copyright dispute with University of California Berkeley. (The relevant point here being not the copyright dispute itself, but the fact that AT&T was not allowed to sell the UNIX system for profit, this being one of the conditions of being allowed to operate as a telephone company.)

You are right, but it's not the full picture.

The government also has agencies dedicated to maintaining monopoly. It is done through patents, copyright, trade secret laws.

It's important to maintain a balance between completion and monopoly. Monopoly allows collecting huge margins that could be invested into further research that would not be otherwise be feasible financially.

> The government also has agencies dedicated to maintaining monopoly. It is done through patents, copyright, trade secret laws.

Without those laws the big companies would steal and copy every small business idea or IP with no impunity. The laws protect small companies much more than big companies. Of course big companies sits on a lot of patents so they are still powerful, but not as powerful as if they could just abuse espionage etc as much as they wanted.

For example, what do you think would happen with the harry potter books? Others would just print their own copies, write their own books with the same characters and names, and the company with the biggest marketing budget to throw around would win and sell the most. That doesn't seem very good for the little guy to me, rather it would make it basically impossible for the little guy to compete.

(Intellectual property) rights depend on your ability to enforce them. Large companies have many resources to throw at enforcement while small companies have few and individuals even fewer. In the past, patents may have worked for the little guy but no longer. I don't know why exactly.

On the Harry Potter thing, "the biggest marketing budget would win and sell the most" is indeed what happens, but the original creator still has huge power. For example if George Lucas (who sold the Star Wars IP to Disney) decided to publish his "fan fiction" screenplay for Star Wars Episode VII, Disney would have a massive continuity problem because he's the original creator—no matter how much The Mouse spends on marketing. That is what Disney paid him $4B for, to not do that.

Further, isn't "people will write new stories for characters they love" a good thing? We could certainly use fresh ideas rather than this rut of reboots and sequels. In that sense, letting The Market winnow the winners from the chaff would improve consumer offerings. It's happened before: remember that Twilight fan fiction that became a literary sensation and major movie series?

"write their own books with the same characters and names"

Clearly you've never encountered the vibrant and exceptionally large world of Harry Potter fanfiction.

Anyway, IP laws usually benefit incumbents simply because small companies have neither the legal budget nor attention to fight a lawsuit. This is a recurring problem in the legal system, where the party most likely to win is usually the one that can hire the highest-priced lawyers. It's also a very hard problem to fix.

I would definitely disagree that the laws protect small companies much more than big companies, having both founded small companies and worked at big companies. Big companies steal and copy every small business idea or IP with impunity anyway, because small companies do not have the resources to fight their case in court and still survive in business. I've never once heard of a startup successfully suing a gigantic company over IP rights and later replacing that big company off those IP rights; the best they can hope for is usually a negotiated settlement and an acquisition. Startups that actually want to survive based on IP usually do it by keeping it secret and working on problems that are not obviously a big market.

> Clearly you've never encountered the vibrant and exceptionally large world of Harry Potter fanfiction.

That is the point, it is fanfiction and not books written by giant corporations churning out clones as their main business practice. That makes a huge difference.

> Big companies steal and copy every small business idea or IP with impunity anyway

No they don't. Without these laws this would happen with every new rising startup: A really good guy applies and wants to work for you. First day he gets to see your code he sends it all in an email to his old company, quits and gets a huge bonus from them. The next day that company launches a clone using your codebase, with your branding and they then start to put a lot of people to use the "embrace, extend extinguish" practice where their part is interopable with your part but yours isn't with theirs.

Big companies can't do this today, but if you legalised it then they would do it. Do you see the difference here? And worst is, this is super cheap to do, so big companies could do it for every single startup that gets more than a thousand users, at that point their branding isn't strong enough to protect them from a bigger company taking over their branding entirely. When you search Google for your product the top result would be their perfect clone of it, looking and behaving 100% as it should.

> I've never once heard of a startup successfully suing a gigantic company over IP rights and later replacing that big company off those IP rights

That is because every single big company already knows where the line is and just toes it, they wont steal things in a too blatant manner since then there are no lawyers good enough to save them. If you move the line they will start to get more blatant about it, if you remove it entirely then they will do what I talked about above.

The US allows monopolistic returns because the country (politicians) benefit from it. If antitrust successfully kills monopolistic returns the beneficiaries of the increase in consumer surplus are the customers, many who are foreign.

Peter Thiel: "Competition is for Losers"

https://web.archive.org/web/20141215224757/https://www.wsj.c...

Thiel's argument, to the extent it's valid at all, doesn't prove that monopolies are good for consumers (or workers, for that matter). It only proves that monopolies are good for monopolists.

http://blog.peterdonis.com/opinions/monopoly-money.html

Yeah, power always consolidates unless people use violence (or threat of violence) to stop it. Companies merges to bigger companies, company owners loves this, that is bad and gives people less freedom. Countries merges to bigger countries, politicians loves this, that is equally bad. The more the power is spread around the better, but everyone who sits on power tries to work against that.

Note that giving all the power currently held by capitalists to politicians is also consolidation of power, that is bad. Or vice versa. So giving the government too much power over companies will hurt on the other end, balance is needed and it is hard to say where the line should be drawn.

Exactly. The phrase "power vacuum" is a good way to visualize this idea. When there is an opportunity for someone to gain power (of which there are many kinds), someone will exert the effort to take that power for themselves.

In sectors where products can have infinite variety (like fashion), monopolies don't seem to happen because the small creative producers can create new market niches faster than the monopolies can enter them.

Exactly. AT&T was a monopoly because it made sense to be one given the capital investment requirements and technology. Universal access was something that lifted up the whole country and had and has enduring value. Think of the bullshit and lack of interop between Facebook, Apple and Google messaging - without AT&T, you would have had that problem with legacy phones.

They were tightly regulated, which worked well until technology moved ahead of the operations.

Tight regulation works well when change rates are slow and you need to make capital investments. Short term decision making driven by final optimization always constrain capital investment. That’s why your power grid is less robust than it was 40 years ago.

That's not clear to me. I don't know much about things like clothes fashion, but e.g. eyeglass fashion is completely dominated by Luxottica (resulting in eye-wateringly high markups) and fashion-adjacent fields such as music are still dominated by just a few labels (this one you can say is basically a product of legal lobbying for protections).

The pattern there is that small producers for the most part fail and fizzle out. The few successful ones get snapped up by the big players.

Eh, maybe 15-20 years ago, but fast fashion producers have made it much harder. You get 6-12 weeks before you get copied, at best. It's still possible to carve out a small business, but a lot of good business operators are being held back by those types of companies.

This should happen with pizza, yet we see dominos successfully applying a monopoly strategy.

https://www.forbes.com/sites/dalebuss/2019/04/30/fortress-ma...

I wouldn’t use cellular as a success story. Even with billions of dollars in investment directly from the Feds, AT&T seems to be wallowing.

Actually there are lots of factors impacting horizontal and vertical integration in the capitalist economy.

In general you are right that "external" restraints are needed to prevent a firm from competing so aggressively or moving to integrate horizontally. However, the external restraints are far and away most often free market forces.

Economies of scale are nothing to laugh at, especially in ventures with very high fixed costs. However, integration is also high risk because it comes with significant downsides.

I don't care to analyze the US cellular market, but I would be curious to what extent the way the government has regulated this market has pushed it toward mergers. And perhaps that is a good thing to have a small number of national cell carriers. People who grew up in the early telephone era would probably agree its better than having to navigate the byzantine maze of regional networks and long distance charges.

There are other mostly unregulated wireless markets that might serve as interesting case studies though. Wireless ISPs have been a booming business for decades and is still extremely dispersed in ownership. They're facing some disruption now from Starlink but that is likely years away from posing a serious industry challenge.

> I don't care to analyze the US cellular market, but I would be curious to what extent the way the government has regulated this market has pushed it toward mergers.

I believe that auctioning spectrum is a huge factor in creating monopolies because only the biggest companies get spectrum. No small guy can afford to compete in the auction. I don't think that auction (to the highest bidder) is the best way to allocate spectrum if the goal is to preserve competition.

Reply on news.ycombinator.com