A New Antitrust Doctrine | Hacker News Reader

A New Antitrust Doctrine

platformonomics.com

32 pointstimsneath13 comments

Dude seems to have a weird personal vendetta against Lina Khan.

I detest most private equity and am somewhat curious to read his thoughts on it, but I tuned out after the like fifth instance of high-school-girl gossip mongering about the FTC under Khan.

Also, it's sort of weird to make efficacy judgments about cases brought by the DOJ/FTC in the current revived antitrust era when they are forced to climb such a mountain of adverse precedent generated by the Law & Economics and Chicago School movements over the last 40-50 years. Hayek's intellectual children have been digging a moat inside the judiciary and private practice at least since Bork wrote The Antitrust Paradox.

Put simply, if your goal is to upend the entire paradigm for evaluating competition harms, your arguments probably aren't going to be able to be regarded as very strong under the existing frameworks.

The line about the Meta/Google ad duopoly fading is false.

The graph in the link just depicts the growth of TikTok, a one time event.

Not even the same market, since they offer largely non-substitutable vertical video ads.

Whole piece is rather casually misleading.

tl;dr: the author wants antitrust action to focus on private equity rather than tech (and oh look, he runs a tech consultancy!). the evidence presented isn't entirely wrong (at least the parts i skimmed), but it's certainly highly selective, in service of reaching a desired conclusion. the digression into different schools of thought on antitrust is a waste of time, to put it mildly (and it is two posts mashed together, contrary to the author's explicit disclamation).

antitrust is really simple in my view: companies should principally be in one line of business[0] as that's the most efficient configuration for the economy as a whole. some of those businesses will be larger than others as a natural consequence of the given industry (e.g., capital-intensive businesses like aerospace), but no business should be larger than it needs to be (in the long-term; short-term dynamism is "inefficient" but promotes longer-term optimality).

that's because companies that are too big inevitably develop internal inefficiencies (e.g., bureaucratizing and politicizing) and suffer from diseconomies of scale, like the coordination problem (the topic of mythical man month). they also develop external inefficiencies, like looking for regulatory leverage via lobbying, arbitraging labor across jurisdictions, or settling into rent-seeking. right-sized companies, those that must focus first and foremost on market competition, don't have time for that sort of bullshit.

[0]: the hard part being the delineation, but it basically boils down to any business that can remain a going concern without positive externalities helping them along (which is not consolidated businesses in the US, including all the major telecoms, defense, education, biotech, oil & gas, etc.). private equity plays in all of these industries but it's silly to think that it's PE that is causing poor economic outcomes, so much so that we solely target PE for antitrust action.

Your framework is interesting, but it would seem to preclude much of the historical antitrust action of the past, including the pioneering actions against Ma Bell and, even earlier, ALCOA and Standard Oil.

Unless you mean that the very existence of a vertical monopoly is a positive externality that keeps the business going. In which case, yes, which is why those firms had to be broken up.

strictly speaking, a monopoly by itself is probably not a positive externality, since it isn't external, other than the regulatory permissiveness that allows it. to me a vertical monopoly isn't necessarily a single business, even if the stack as a whole principally delivers a single type of good or service.

if a layer in the stack can thrive on its own as an independent ecosystem, then that's a "separate business", and trust busters should look askew at conglomerations across such boundaries. independent companies will make the whole industry more efficient in the long run, because of competition. in such a configuration, companies can make deals with each other across stack boundaries and even invest in each other as long as a controlling interest hasn't been reached.

ma bell for instance vertically integrated telegraph, telephone, transmission, real estate, consumer phones, answering services, and who knows how many other whole businesses together to form its monopoly. many, if not all, of those could have been thriving markets on their own.

I see. I mostly agree, though I definitely think the delineation problem is more difficult with vertical rather than horizontal monopolies, e.g. (to stretch the example), McDonald's franchises currently cook all the food they sell. They could instead buy pre-cooked food from a separate business, but would such a breakup of functions benefit the economy? Where that line is seems like a hard problem for a court to solve.

mcdonald's is actually a good example. it's at least 3 businesses on top of the nominal one of fast food: franchising, supply chain, and real estate. franchisees pay for access to the mcdonald's brand, trademarks, and shared marketing. in addition, mcdonald's corporate provides complete supply chain support, so franchisees buy all of their food and materials through mcdonald's itself or from its approved suppliers. mcdonald's corporate also owns the real estate that many franchises sit atop, so it's a landlord to boot.

it'd be easy to split those businesses into separate ones, as it's plainly obvious that all three businesses can viably stand alone, apart from serving fast food. you might argue that that'd dilute the brand value because the product wouldn't be as uniform, but you just have to look at the international variation to realize that that's not so important to brand value.

You could have a theory of antitrust that allows for both rules - if a company is across multiple fields of business or if a company is too much of a single field, it gets split up. So Standard Oil and Ma Bell get split up for controlling too much of their industries and using it in anti-consumer practices, and Amazon gets split up (AWS vs everything else) because it’s inherently too large and it’s better for the economy for them to be split up.

of course, but i'd argue that too much of a single field is usually a case of a company that's actually in multiple businesses (see the mcdonald's example in a sibling thread).

my simplified perspective is a melding of the brandeisian (big is bad) and the chicago schools (economic efficiency for the win!), touched on in the article, so i certainly have no problems with synthesis in matters of antitrust. =)

> antitrust is really simple in my view: companies should principally be in one line of business[0] as that's the most efficient configuration for the economy as a whole. some of those businesses will be larger than others as a natural consequence of the given industry (e.g., capital-intensive businesses like aerospace), but no business should be larger than it needs to be (in the long-term; short-term dynamism is "inefficient" but promotes longer-term optimality).

Would this limit vertical integration? E.g., could SpaceX exist under your scheme, making rocket engines, rockets, and satellites? Is that all one line of business, or is it multiple lines of business? Is Apple's hardware and software business one line of business or two (hardware, and separately software)?

It seems to me that your approach is easier to apply when a company strays from its line of business rather than vertically integrates. So a company that makes hardware, then years later gets into software for its hardware, that might be a problem, but then how could that be a problem if Apple-like companies exist concurrently.

I think all antitrust schemes are tricky and slippery. Not that we shouldn't have an antitrust scheme, but that I'm not sure how to construct one that doesn't somehow suck.

spacex, as far as i can tell, is a research company that launches satellites to fuel its space exploration research. the research side couldn't stand alone, as that's not a viable business in itself, so the satellite business is really the core "business" here. the making of rocket engines could possibly be outsourced (or turned into a separate business), but those customized engines are integral to its competitive advantage in more efficiently sending satellites into space than its competitors (and in sending people to colonize mars, its long-term mission). core competencies like that aren't separate businesses; they're the business itself. if spacex decided to write its own accounting software and then sell it to others, then you'd have a breach of the single business focus.

apple, however, is in something like a dozen different businesses that are completely viable as stand-alone entities that don't gain much from being consolidated (brand value, for instance, can be franchised if a unified brand face is desired, like mcdonald's does). incidentally, apple wasn't a hardware company that morphed into software, but rather, it was always a computer marketing business (then more broadly consumer electronics, and now encroaching upon consumer entertainment). nest is a good example of what a spin-out company in an apple breakup might look like.

I like an advertising monopoly: it means higher prices for less advertising (good) and more money for services to the actual population (good) and the high cost means ads I see are more likely to actually tailor to my interests (good). Why would we want a more "efficient" market for this? So we could pay a nano penny less for products we won't be buying anyway and in exchange Gmail, Facebook and every other free site becomes unusable as 98% of the screen is an ad? No Thanks.

Far to dismissive of Brandeis and the ilk.

All the kids who today revere RBG as some kind of demigod with accompanying action figures should know more about Brandeis . . . an actual leftist, populist justice, starkly unlike RBG.

Reply on news.ycombinator.com