> But do these wield too much market power to be controlled by the market. That is, sure, Bing and DDG make up 5% of the global search, but is that enough.

This is exactly what antitrust law should not punish. Google has a large market share of search because it offers a good product. Antitrust law should preserve competition without punishing success.

People are free to use alternatives, including Bing. If they don't, that doesn't make Google a monopoly (which indeed in search they are not). It makes Google successful! (Though it is a successful firm which does occasionally violate antitrust laws in other areas and should be punished for that.)

> Google has a large market share of search because it offers a good product

No. Google has a large market share of search because it got in early, presented some synergistic products, heavily biased search results towards its own products to reinforce that, heavily gamed advertising bidding in its favor, placed anti-competitive requirements on hardware manufacturers of Android devices, and locked-in consumers' data so that competitors couldn't access it until it was forced by governments to allow consumers to download a copy.

Google, and Alphabet, do not today offer good products. I can go on for hours describing many years-long problems with the product offerings; culminating in my complete non-use of Google as a search engine because its results are so poor, stopped using Google News because of its tracking and curation biases, stopped using Google Mail because of its inability to provide filters that automatically mark emails as spam, stopped using Google Drive because it actively prevents many features from working in Firefox, and ... well the list can go on but I have better things to do.

> No. Google has a large market share of search because it got in early

"got in early" compared to what?

There were dozens of search engines before Google.

The point is that Google used their search dominance to push YouTube. Then their YouTube dominance to say "this site works better in Chrome" to push Chrome. Then their Chrome and Search Dominance to push AMP.

It's not that Google should be punished. It's that they should be stopped from using one product line to prop up another.

I don't know what the formal definitions of any of these things are, but I assume (hope?) that it's important not to equate being successful with being a monopoly by definition. If Ford came out with an amazing new car tomorrow, and suddenly 95% of new cars purchased were Fords, we probably shouldn't initiate trust-busting action against Ford on that basis alone. We want car companies to try to capture market share by making better cars.

I don't know exactly how we contend with this definitional problem, but intuitively I think it should have something to do with prices and behavior. For example, if we were a fly on the wall in a Google exec meeting, we might be curious about whether they're jacking up the price of ads because they know no one can compete with them, or whether they're sensitive to charging more than their smaller competitors do. If they're behaving like their competitors are a serious threat, then intuitively they're probably not a monopoly, and it probably doesn't make sense to trust-bust them.

But anyway I'm not a lawyer, and I don't know what the current thinking is about any of this stuff.

> I don't know what the formal definitions of any of these things are, but I assume (hope?) that it's important not to equate being successful with being a monopoly by definition. If Ford came out with an amazing new car tomorrow, and suddenly 95% of new cars purchased were Fords, we probably shouldn't initiate trust-busting action against Ford on that basis alone. We want car companies to try to capture market share by making better cars.

There's an important difference between "capturing market share" and "capturing too much the market share." If Ford's going to put its competitors out of business, turn them into niche players, or even just get into a position where it can rest on its laurels for a few decades, our primary concern is no longer making sure Ford can be rewarded. Honestly, I don't see the downside of regularly breaking up a #1 player with too much market share into successors that are the #2 and #3 payers (and forcing major shareholders to divest one or the other). I want companies to be competing to get to the top, but once they get there, give them a prize for their effort and send them back down.

From my experience as a customer, Amazon's e-commerce operation hasn't improved much (e.g. their search is still as shitty as ever) and has arguably regressed in many ways (e.g. more shady products and fake reviews) in the last decade. That's pretty strong signal to me that they're not under enough competitive pressure.

Amazon's search is so bad that it's almost certainly intentional in order to get customers to look at more products. This is a great example of them using their monopoly power to increase their revenue to the detriment of consumers.

For an idea of what it could be, try the search on electronic component distributor sites like Digikey[1] and Newark[2]. They are both improving their search in order to attract more customers benefiting all customers in the process.

Try to find a 34 inch 1080p monitor with both VGA and DisplayPort inputs on Amazon and you'll find yourself reading hundreds of monitor product pages. If Amazon had serious competition you could probably find and buy one in 5 or 6 clicks, or easily determine that such a combination of features doesn't exist.

[1] https://www.digikey.com/ [2] https://www.newark.com/

I’m skeptical that making users spend a long time trying to find the right monitor would actually be good for Amazon’s revenue or that Amazon would intentionally optimize for it. If the typical user really experienced that struggle, many would probably give up without making a purchase.

To hazard a guess, I’d speculate that apparent poor quality of search results is more likely Amazon trying to push customers toward items where Amazon earns more margin. The average customer probably just buys the first/cheapest result and isn’t going to spend hours scouring the product pages. There’s a lot of complexity in Amazon’s relationships with suppliers, fee structure, and warehouses/logistics that could affect the revenue-optimizing search ranking but isn’t obvious to the user.

Amazon deliberately leading users to a suboptimal product because they get more margin seems like exactly the sort of thing that competition should help to resolve (or at least limit). The fact that search is so bad is probably evidence of too little competition.

Amazon has ads in search results.

I have to wonder if it's fundamentally appealing to a different type of customer.

You don't window-shop Digi-key. You're pretty much buying on spec sheets alone. I want to see all 330 ohm 1/4 watt through-hole resistors, now sort by price and availability in the quantity I need, then maybe discard products that don't meet some other criteria that I can't directly filter for. There's very little chance of making a sale through a slicker marketing campaign.

Conversely, people buying a single monitor might be browsing, potentially swayed by the right language and graphics on the landing pages. I'm pretty sure those bloody finches sold more ViewSonic monitors than any other aspect of their product line. For a window-shopping customer, getting the customer to see as many of them as possible improves their chances of a sale.

On the other hand, I will concur that Amazon's search is a disaster for computer products. I suspect it's also terrible in other verticals that have clear, well-defined "faceted search" concepts, but we probably have the most experience with that one here. I suspect it may be a casualty of their broadness of categories-- why spend the labour to provide really killer search if it hasn't proven to be an impediment to sales yet?

> There's an important difference between "capturing market share" and "capturing too much the market share."

No. Not in the economics of antitrust. There is no "too much market share" threshold. Unless consumer welfare is harmed there is no reason to complain that a firm is "too big."

> From my experience as a customer, Amazon's e-commerce operation hasn't improved much (e.g. their search is still as shitty as ever) and has arguably regressed in many ways (e.g. more shady products and fake reviews) in the last decade. That's pretty strong signal to me that they're not under enough competitive pressure.

So go elsewhere! That's how competition should work. Among your other alternatives are Walmart and Target, which have quite robust online operations.

Amazon is not plausibly a monopoly.

The "consumer welfare" standard is a fairly recent (1980s) reinterpretation, and the switch to it happened because of large business lobbying. The original intent of American trust-busting laws was very much to prevent capturing so much market share that competition becomes irrelevant.

https://promarket.org/2019/09/05/how-robert-bork-fathered-th...

And that original interpretation was what worked on the original monopolies (oil etc). The sooner we get back to that, the better.

> No. Not in the economics of antitrust. There is no "too much market share" threshold. Unless consumer welfare is harmed there is no reason to complain that a firm is "too big."

Consumers are harmed by a "too big" firm that's under little to no pressure to improve its products. The retort to that usually "wait a few decades until the market sorts that out," but frankly, I could be dead by then.

Another poster addressed how the "consumer welfare" standard is not so much an economic argument, but rather a legal enforcement framework. I see little mechanism for consumer welfare to be harmed by the creation of multiple successor companies that all share the same successful formula of the "too big" firm but now have to compete with one another.

> suddenly 95% of new cars purchased were Fords, we probably shouldn't initiate trust-busting action against Ford on that basis alone.

Antitrust law isn't about stopping people from accumulating a dominant position. It's about stopping people from using that dominant position to maintain itself or to spread to other areas. Ford produces the "Amazing Car" that everyone wants, costs $1, etc. great! When Ford uses an RFID reader to force it to only fill up at Ford Gas Stations antitrust steps in. The issue isn't that Google is the dominant search engine, but that it uses that to become dominant in ads. It's not that YouTube is dominant in video, but that uses that to encourage people to download Chrome.

Good take, very much agree. That’s why so many of the antitrust work against Google/FB, etc. has focused on acquisitions. They become successful, then use their data and infinite money printers to buy anyone who would compete.

Yes Google was able to catapult YouTube playback pages in all Google search results after acquisition, just one example

The Justice department has some good documentation.

https://www.justice.gov/atr/competition-and-monopoly-single-...

This is interesting because it defines monopoly and market power in terms of prices. How does the traditional monopoly definition fit when we're talking about "free" goods (to the general consumer) like FB or Google search? Is the thought that they exert too much power over ad pricing?

This is a fascinating and unresolved question in antitrust economics, but explains why very few people in the profession think FB is a monopoly. And indeed - the FTCs failure to make a plausible market definition argument which showed why FB was a monopoly in that market is why the FTC got slapped by the judge in January of this year.

~Antitrust Economist.

It doesn't fit. The point is that we need new regulatory language that does fit.

Is there a proposed new language/definition? The article's discussion still seems to be centered on price.

"her paper highlights how the consumer welfare view of antitrust fails to curb the predatory pricing".

I said it elsewhere, but I not against the idea of trust-busting, but the industry analogies typically brought up don't really fit. The closest I can find in general discussion is that the services are a public good, but we tolerate/promote monopolies in those sectors (like utilities).

Agreed. I don't think there are really any mainstream alternative proposals in the policy sphere, at least not that I've seen. I've seen technologists grapple with what might make sense, like Ben Thompson for instance, but they don't seem to have the ear of government actors. Which is why I'm so frustrated by the conflation with monopoly regulation. It is actively a distraction to the policy discussion. Every time they go to Congress, there is all this focus on "breaking them up", but it's all just really misguided and beside the point, and I blame the monopoly confusion for that.

What do you think is a better way to frame the policy discussion? I assume they use the monopoly angle (perhaps in a misguided fashion) because it is a tried-and-true template for regulating corporations with too much power.

Yes, that is why, but they are wrong.

A better framing would be to focus on specific problems: "Facebook owning both facebook.com and Instagram and both Messenger and WhatsApp give them too much market power in social sharing and messaging, so Instagram and WhatsApp should be spun back out", "Amazon should not be able to use sales metrics of other platform sellers to inform their own product development", etc. Identify market issues, figure out regulatory language that speaks to those issues.

But "big companies are monopolies that must be broken up" is unhelpfully reductive.

I agree with you, and I'm just following up for a finer point.

It's my understanding the whole intent of anti-trust is consumer welfare. The connection between Amazon and pricing/competition seems to be a straight-forward connection between hurting consumer welfare and its market power in terms of Amazon's market power. But where does the connection come in between consumer welfare and the social media apps? I.e., how does FB hurt me by having the most/best information regarding my social network?

Or is your stance it shouldn't be framed in terms of consumer welfare at all?

Yeah I don't really buy the consumer welfare framing. Or maybe I do buy it for existing anti trust regulation. But I don't buy that it is the only thing we (as a society) should be looking at when considering regulation of companies. This isn't novel, we regulate other externalities even when the regulations themselves harm consumers in the direct sense. For instance, more environmental regulations result in higher prices. But they are often good for other reasons, like not poisoning drinking water. We just don't use anti trust regulation to do this.

Basically I just don't see anti trust regulation as being the best way to target the diverse set of issues society sees with the big tech companies of the moment.