The way antitrust is working in recent years is kind of ridiculous. Google in Europe is becoming a rich case study in antitrust meh.
You find that adwords engages in a blatant "monopolistic abuse" in some specific sense. In the Adwords EU case for example, Google was both monopsonising (google search) and monopolising search advertising. But, the actual violations are specific, like exclusivity requirements and other technical infractions buried in their take-it-or-leave-it contracts with "partners.^"
Google get a fine. They deny culpability. They pay a fine and move on. Whether it's 100m or 2bn doesn't really matter. Google are demonstrably willing to pay far more (eg acquisition, browser kickbacks, etc.) to maintain a monopolistic position. Besides minor tweaks, there's no reason for them to change.
What's the point? Antitrust can't be useful as a violation-fine paradigm. These cases/fines don't affect the monopoly or free the market in any useful sense.
IMO, we need some way to declare a company "Big" and apply special rules to them. If those rules are onerous, that might create some organic pressure to split and/or help smaller incumbents compete.
There is no point to antitrust if its totally reactive, handing out fines for moving violations. The point of antitrust is trust-busting, to affect market structure. The point if to have a competitive market.
All the 1990s stuff is playing out. Just like the MSFT case, we see that operating systems, browsers (now app stores) are monopoly prone and that it's a big deal. On the other side, the 90s "portal" concept is fully realized now. In the digital economy, controlling the windows of consumption is everything. None of the costs, all of the profit... literally all in a lot of cases.
^Newspapers didn't have the advantage search partners had in their monopoly claim, because Google doesn't have agreements with them. No contract, no abusive clauses. The take it or leave it offer is "give us your content for free, or go away." Formally, Google isn't even doing business with them.
> IMO, we need some way to declare a company "Big" and apply special rules to them. If those rules are onerous, that might create some organic pressure to split and/or help smaller incumbents compete.
Past a point, they should be regulated like an utility.
IDK if it's "past a point," but in some cases maybe.
I'm torn on that though. EG, what would it actually mean if FB or twitter became "regulated like utilities?" Interoperability? Portability? IMO, these are pretty finicky problems for a regulator. I think it would likely look like the mandated personal data dumps we have today. Some clunky feature that few use, and don't change the structure of the market. Useful transparency, but not really game changing.
Regulated equal access? Court-like proceeding to confirm bans? I'm not feeling it. I suspect it will entrench monopolies further.
A Telecom or EnergyCo is fundamentally different from a FB or Google. Regulators needed to make sure everyone had access, because it's always more profitable to just exclude certain people/areas. Prices are an issue because infrastructure monopoly. Also because government/regulators typically fund a lot of the capex.
There is almost no cost, marginal or fixed, associated with instant messages, shares, likes, links, etc. If Facebook and Twitter folded tomorrow, we would still have plenty of all of these because they are not scarce. Maybe there's a case for cloud computing as a utility, but even here I'm hesitant.
Enough with analogies to 19th century monopolies. Platform monopolies should be regulated as platform monopolies, not utilities. New thinking.
In any case, google is the perfect target for actual trustbusting. Separate search from adwords and android. Ban default buying. Ban mergers.
The old challenge was (eg) trust busting steel without harming the actual production of steel. Today's problem has little to do with production. It does have a lot to do with share prices. Is there really an appetite for turning Alphabet from a $1.5trn company into a $150bn company?
The EU's initial proposed version of the "Digital Services Act", requires that when a service takes down information that they think is "illegal" or incompatible with the terms of service, They are required to provide the "facts and circumstances" used to arrive at that decision, as well as reference to the specific term of service they think is violated, and why they feel the information violates that.
Furthermore, they must provide an appeals mechanism for both information takedowns and any account suspensions/bans associated with them (and besides things like non-payment, I'd bet the majority of account suspensions/bans on online services are due to some posted content the company finds objectionable).
Lastly, if the internal appeals process fails to satisfy the user, the user initiate a what is basically an electronic binding arbitration process with a certified " out-of-court dispute settlement body" of the user's choosing, with this body having the power to overturn the companies decision.
That last part sound an awful lot like "Court-like proceedings to confirm bans".
I suspect the final law may not be as pro-consumer in these areas as the initial draft is (many big companies will not be happy about this part of the law and will try to kill it off).
If it does survive it would be much better then all too common: "We are banning you because we think you broke an unspecified part of the terms of service. We won't be more specific because we don't have to be, and are scared that people trying to break the rules could learn how they got caught. We won't even tell you that this was a fully automated process with no humans involved, that gets things wrong x% of the time. There is no appeal, unless you can get Twitter or the news media sufficiently riled up against us."
I suppose we'll see how it works out, assuming it does become law. There is devil in the generalities here, but plenty of devil in the detail too.
Detail-wise.... Compliance, currently, is the art of ticking boxes while not changing anything fundamental. Too often, politicians and advocates have their eyes on the spirit of (proposed) laws while corporations and their advocates put their eye on how "compliance" will actually work procedurally. The latter approach is more effective.
That doesn't mean no goals are ever achieved, but it does dampen a lot. EU transparency laws, right to be forgotten and such are good examples. They might establish moral rights, but they don't affect really achieve big practical goals.
Generality-wise... I think the problem for (eg) free expression is the platform oligopoly itself, not their corporate policies. That's not my main concern here though. My main concern is: These laws are only necessary because monopoly. The danger is that the laws entrench monopoly, by applying them unnecessarily to smaller platforms.
That said, we'll see when/if it happens. Thanks for the info.
Generally good comment, but if search is separated from ads, how does it make money? I'm totally on board with splitting up Google, but losing Google search (or search becoming notably worse) would not be worth it for me. The competition is just so bad, and always has been.
Part of why searching is as painful as it is is because it has been deemed acceptable to pollute the search space. SEO should be neither encouraged or allowed. It is nothing less than adversarial corruption of a shared index.
Boolean searching works. Go back to a library catalog sometime and be amazed at how quickly you can get relevant results when every Tom, Dick, and Methuseleh aren't shoveling extra irrelevant feces into the index.
There is no practical way to disallow SEO. Search results have to be ranked somehow in order to be usable for most people. And site owners will always figure out tricks to game the ranking algorithms.
It's not SEO that's the problem. SEO is a symptom of the perverse incentives.
So Google has created incentives (via ad embedding in user-sites) for entities to generate content and compete with each-other for who can show the most ads. This in turn "pollutes the index" as the other poster phrased it, which makes it more and more difficult to find actual real (unique?) content, which then Google turns around and supposedly "helps" us with.
Let's be honest. Despite it's surface-level gigantic scale, the internet is actually very tiny when it comes to real genuine content. My wild estimate is that 99% of it is spam. Even in the case of user-generated content, for every genuine blog where someone shares their thoughts and unique insights or content, there are a thousand blogs which are just there to get ad-views by generating any content that might get someone to look there. I can't even find the former type of blog/site anymore. The only place I still encounter those on occasion is when it's linked to from HN. And the ones I do encounter that way, I find that even when I try to, I can't find it via any sort of generic search (i.e. not cheating by using the site name / author / unique phrases).
How is searching painful? It's gotten better with almost every year.
I only see this sentiment here on HN, along with many other shitty takes about how pure the internet used to be. Are you talking about that era when opening the wrong search result would bombard your screen with millions of popups and install 20 toolbars on your machine?
Please tell me when this magical time of no spam in search results existed? 1995?
It wasn't a case of no spam; but rather predictable spam.
After a while you'd start to recognize the Web rings. The "just link" pages became extremely obvious after a while, and you could generally reliably get the same results regardless of where you searched from or what you were doing when the time came to search. You had far fewer syndication mills, and you could even find resources reliably.
The net was a reasonably navigable thing even to children doing school research. Nowadays, I have to explain to kids all the myriad pitfalls that exist nowadays, when they look up something, ho back a bit later, and can't find their resource anymore.
And that's just the bloody legit stuff. When you talk malware, we're in a completely different world nowadays.
You do a search, it says millions of results, the first pages all have results without the quoted terms in the page at all. They don't show "we don't have any good results" if they can display a bunch of ads on a bunch of pages with useless results.
Google will continue to make money as it does now, through adwords as a "search partner." This is also how other SEs make money. The biggest antitrust case Google has lost so far has related directly to this.
Would they make less money? Yes, probably. It's hard to say how much, but it would still likely be enough to run Google.
I think it's easier to understand the paradigm with Twitter or FB, since they're simpler. Facebook currently makes about $100bn pa from ads. Circa 2015, they were operating profitably on about $10bn ad revenue. Twitter has had a similar (more modest) trajectory. They spend a lot of money running FB & Twitter, because they can, but there's no reason to think FB or Google Search on far less revenue than they bring in currently.
There are versions of what I'm suggesting that likely affect revenue moderately, and versions that could hit harder. IMO, in either case, I don't believe there is any meaningful consumer harm or ill effects outside of FB's shareholders or perhaps employees.
This isn't true of all monopolies/sectors. It's true of FB, Google, maybe twitter.
That said, I think the risks are way smaller than previous era's trustbusting. I like FB as an example, as I said. I'm pretty confident that FB would be pretty much the same on far less revenue because they were pretty much the same on less revenue in recent years.
Comparable social media services run of far less revenue (scaled whichever way you like). You can also look at their accounts, which reveal a lot. The simplest clue is profit margins.
If we're wrong... one social media service declines and another takes its place. Personally, I see very little risk. The actual risk, IMO, is not really about keeping these services viable. There no "shit! I broke auto-manufacturing and now we don't have cars" risk.
It's about share price. Smallish declines in revenue can mean big declines in profit. Any decline in revenue can lead to a collapse of share prices, regardless of profit. The big techs are a big part of the S&P, and high profile. Quite a lot of the S&P's value relates directly to monopoly, so antitrust (IMO) directly conflicts with equity values... and things could get "macroeconomic" from there.
All that said, I'm not advocating anything intentionally destructive. Google is, I suspect, expecting to be broken up eventually. I think that was part of the Alphabet restructuring logic.
The primary objection that comes to my mind, in the vein of your comment,is a hayek-esque information argument. A Hayekian logic though, is totally incompatible with paragraph 5 above though, and I'm quite confident about that... so I don't know where it fits in.
And we need an effective regulator. For example, I am baffled by the level of oligopoly allowed in the EU cellular/telecom industry. It is quite obvious that prices are synchronized between providers if you watch their dynamics over time. They are currently killing prepaid cards as well.
There are now countless eSIM offers (mostly for data) that are way cheaper than the previous prepaid options, so I'm not convinced of this.
(I'm sure the roaming fee limits played a big part in this, because it effectively allows any mobile operator to resell data at the wholesale rate, or even below it because data is usually sold in packages that are rarely used up in full.)
Doesn't that effectively mean that these companies will become further entrenched and nearly impossible to dislodge from their market positions? A prime example is Google Fibre, who tried to dislodge existing cable/internet companies, and failed primarily because of regulatory entrenchment.
Why not allow the market to be more inclusive? This will help create exactly the competitive pressure that we all benefit from.
Utilities aren't regulated to function like "non-monopolies," generally. They're regulated because (ATT, at least) they were seen as inevitable "natural" monopolies.
There really isn't one "theory of monopolies" that is useful in all times and places. Social Media is not Telecom or electricity. Amazon is not a railroad.
Trustbusting is an anti-monopoly strategy, and it doesn't really work in spaces where monopoly is unavoidable.
Utilities are super heavily regulated. I'm not sure for the US, but for example frequently they can't enter new markets, so they can't leverage their monopoly to crush competition in other markets.
And regarding working well, aren't a lot of people in favor of municipal broadband (an utility) vs Verizon & co. (oligopoly, not even a full blown monopoly)?
California PGE which provides electricity and gas to big portion of public is a moving train wreck. It's corrupted and dangerous. Pipeline explosion, electricity transmission line caused fire burn a whole town happened with worrying frequency. Heavy regulations don't prevent big bonus to executives and degrading infrastructure.
Making an electric or gas company a government entity would at least provide a chance to prosecute its corrupt operators for official misconduct and criminal fraud (see the slow moving train wreck called CalPERS, some executive frog-marching would fix that quickly).
> IMO, we need some way to declare a company "Big" and apply special rules to them.
Careful though. A monopoly may abuse its position, which is bad, but a big company may be big because they are good and people use them, which is good.
This is actually the difference between the US and the EU when it comes to monopoly lawsuits. The US has to prove that the behavior of a monopoly was harmful to customers/consumers. The EU has no burden of proof and can just take market size as a way to go after them.
The general problem I see here is how you slice the market to say a given company has a monopoly. It sometimes feels like those going after large companies try to slice the market in such as way to say they are monopolies, when other slice it in another way to show that it's not. Which slicing is "correct" is really hard, especially when markets shift relatively quickly.
Monopolies are inherently harmful to the customer. A monopoly implies lack of competition. If lack of competition is not harmful to the customer, then competition is not beneficial. If competition is not beneficial in a given market, there is no reason for it to be a market to begin with and the monopoly should be expropriated.
> Monopolies are inherently harmful to the customer. A monopoly implies lack of competition.
I disagree. When you look at a company that is dominant in some part of the marketplace at a given point it time, you need to understand how they got there. The monopoly poster child Standard Oil rose-up when there used to be many more oil processing companies, but Standard Oil was able to do things for cheaper because they used more byproducts from the oil to create different products (which let them sell things like lamp oil for less). So sometimes monopolies rise-up because that one company is able to fill a market segment better than the competition.
The question then becomes, once they have a monopoly and others see this, how can potential competitors react. Can others copy (and improve) the same processing system that the monopoly has to produce a similar product for a similar price (maybe even less)? If so, then they can rise up and the original company will no longer be a monopoly. But if that original company takes steps to squelch the competition, you start getting into what the US has defined as "anti competitive" practices. These were developed because the past showed us that this leads to harming consumers.
We also have a lot of government granted monopolies. Copyright and IP law are time-limited monopolies on a given product. We have private business that operate on federal land (like national parks) that may be given a local monopoly within that region. Sometimes we make tradeoffs and grant monopolies.
There's usually some market reason for a monopoly to exist. Often, it's a temporary reason. That doesn't mean it isn't harmful to consumers, unless you think that market dynamics are inevitably beneficial.
Google, has several market reasons for their monopoly.
(1) The biggest reason is like economies of scale and network effects, but not quite. Online ad markets need to be huge to be good. Bigger market means tighter targeting segments, which is everything. This is why FB & Google make >10X more per customer than their no. 2s and 3s. Data aggregation is like this too. Some data on some users is not proportionally valuable to all data on all users.
(2) Software's infinite economies of scale. Economists like to point to this reason (0 marginal costs) as primary, but FWIW, I don't think it's why google is a monopoly. It probably is why software is monopoly-prone. It does explain the OS/Browser centralisation pretty well.
(3)Normal economies of scale and network effects.
(4)Path dependencies.
Does that mean that Google's monopoly benefits consumers? Who are consumers anyway? Formally, it's advertisers. If monopolies aren't assumed harmful, why do we even need to curb anti-competitive practices in the first place?
>> government granted monopolies; Copyright and IP; federal land, etc
Agreed. I'm not talking about "monopoly," the dictionary definition. I'm talking about the phenomenon in 2021, a handful of companies, most with a tell-tale >30% profit margin. I'm not talking about the exclusive cafe inside on IKEA.
> Monopolies are inherently harmful to the customer. A monopoly implies lack of competition.
Depends why there is a lack of competition. Are there barriers of entry to the market? Were they put up by the company? Or do they just have such an awesome product at a great price they no one can offer a better / commutative alternative?
Antitrust law is weird, so I can't comment on this definitively. That said, in the EU-Adwords case, the focus was on "abuses," primarily in the contract with search partners. Adwords is their buyer, rather than seller. Technically, Google is the monopoly, adwords is the monopsony.
In any case, prosecutors needed to prove harm/abuse affected search partners, and the fine was presumably based around that.
There is a problem defining the market sometimes, but that wasn't a problem here. It's a "crack for economists" type of problem, with a lot more theory than needed for practical purposes.
The problem is/was that these courts aren't a tool that can do the job. Courts can keep monopolies from doing certain things, with varying degrees of effectiveness. They can't prevent a monopoly from being a monopoly.
True, that's why I'm against general rules (aka laws) as the main/only tool. A cloud platform is not a telecom, is not a search engine, etc.
I don't agree with "monopoly may* abuse its position*" though. Monopolies exert their impact by existing as monopolies. They may or may not (as in the adwords case) "abuse" this position in specific, known, ways like anticompetitive contracts in the adwords-EU case... but this tends to be pretty marginal.
Google bought Doubleclick in 2007. After this, they owned both the biggest search engine and the biggest ad marketplace... their competitors' only revenue source. Throw in Android, Chrome & a $20bn contract to be Apple's default.
At this point it's all about market structure and control over choke points, nothing to do with consumer preferences. I'm not saying shut down google, or prevent people from using it. I'm saying that a market structure with only one choice is a monopoly, and that's bad for everyone but the monopolist.
>but a big company may be big because they are good and people use them, which is good.
I agree in part, but I disagree that 'good' can be a state that applies universally. Personally, I think of it as a tag that you can apply to individual actions on a case-by-case basis. Also, is there any company that has gotten big solely because 'they are good and people use them'?
I'd argue that the latter eventually turns into the former with a probability approaching 1. The incentives are just too screwed up for it to happen any other way.
The fix isn't the fines, it's the requirement to stop given behaviors. Google won't care about the fine it just got, sure, but it also can't block competing apps in this way again, at least in this jurisdiction.
It's a small victory, but it's also a hole in the armor. One of many, as various antitrust cases handle various abuses. Unfortunately, progress is slow but we're making it.
Ive long held that corps should be limited to a single trade category (as per trademarks). I think that change would solve a lot of the complexities that megacorps introduce to the application of legislature.
I don't agree, but I think it's an example of rules that could be applied selectively to megacorps. Banning supermarkets from selling hot coffee, so to speak, is a problem.
It's also a problem on the megacorp scale. If Google wasn't allowed to do gmail, android and such, would we have been better off? That said, if we're only talking about a handful of megacorps, there's no need to be abstract or simple with rules. There are lots of ways to keep them in a lane.
Android, Youtube & adwords are great candidates for IPO. Why not force a float? This immediately fixes most of the issues so far demonstrated in court. It compensates AlphaGoog at market rates, preserves an incentive to innovate...
Before that though... ban mergers! Reverse merges. The whole foods acquisition is a worrying one. The "multiple" discrepancies between amazon and regular retail is so high that there's an arbitrage-like force pushing towards these acquisitions. Tesla & other auto manufacturers are a similar scenario.
Consider this... Acquiring the entire publicly traded retail sector would dilute Amazon's shares by circa 25%-50%, most of that going to Walmart and Costco. They could buy the fashion/apparel industry with petty cash. They could acquire UPS. That would make other online retailers compete with amazon like Google competes with its search partners... same logic as amazon marketplace, but more.
...yet, it will absolutely matter if it's $20B - Google's 2020 operating income was $41B.
You're making the claim that since Google wasn't fined enough, fines don't work - which is absurd.
Try making the fines a reasonably large percentage of gross revenue from the previous year - say, 10%. 10% of Google's 2020 gross revenue ($182B) is $18B, which is almost half of their operating income. That will have an effect.
The problem with this is that you can easily bankrupt a company by codifying exceedingly onerous terms.
Google (to take your example) happens to be able to (barely) afford 10% of their revenue taken from their income. Another company with a different cost structure (perhaps in another sector) might not be able to afford that. The upshot is that providing a codified “one size fits all” punishment could be disastrous (if you consider preserving companies to be desirable: for the sake of this argument I do, but entirely legitimately you might not).
In the case of Italy (I’m Italian) and it’s awkward, glacial justice system, that’s a very dangerous assumption to make. This place is already hostile enough towards businesses.
Fines are, largely calibrated one way or another. It's not marked to gross revenue... but it is usually marked to the infraction. In the EU-Adwords case, the $2bn fine was relative to the size of adwords search partners' EU market share. In this Italian case (I assume) the fine is scaled to the size of the monopolised sector... Italian apps or whatnot.
This is why I called it a "violation-fine paradigm." Courts can't arbitrarily fine companies "enough to make it hurt." It has to relate to laws, specific infractions. Laws have to be general. Etc.
My point is that the infraction itself is not the problem. It's a sign of the problem, a symptom... the problem is monopoly. There's no version of adwords partner agreement (including the post-fine version) that isn't still a problem, even if it doesn't contain abuse of power clauses. The problem is that the only meaningful market for search ads is Google's. Similar issue for Amazon marketplace. There's no version of Amazon marketplace policies where amazon retail competes on a level playing field... they own the field.
You fine a company for doing something wrong. Google do things wrong, from time to time... but the main problem is what Google is, not what they do. That's derivative.
Google doesn't really have a choice. Their monetization model rests on providing increasingly complicated dashboards to lure advertisers into believing their money is well spent. As long as they haven't found an alternative model they will just blindly pay up and redo
> IMO, we need some way to declare a company "Big" and apply special rules to them. If those rules are onerous, that might create some organic pressure to split and/or help smaller incumbents compete.
What do we win here? To me it seems that letting big companies get as big as they can allows them to use economies of scale to provide quality services cheaply to all of us. Search is a prime example here, only Google seems to be able to do it as well as they do. Many others have tried and failed.
Isn't the benefit of cheap/free services far outweighed by whatever arbitrary "abuses" that are alleged to have been perpetrated? Shouldn't our attitude be primarily of bewildered amazement at the status quo instead of attempting to micromanage how Google uses the massive, amazingly positive platform it built?
>Isn't the benefit of cheap/free services far outweighed by whatever arbitrary "abuses" that are alleged to have been perpetrated?
No, because pluralism and a robust and diverse ecosystem is more important than cheap shit. Handing effectively entire control over what is visible on the internet to one company is anti-democratic madness. You're basically pulling a "but Mussolini made the trains run on time"
I'm not, I'm suggesting that an inclusive, permissive economy is the best way to fight scarcity. Scarcity, though you may not recall, was a big deal. Its opposite is "cheap shit" which is exactly what I'm advocating for as top priority. In this case, cheap (free) access to high quality internet services for everyone with an unfiltered connection.
Less permissive economic systems continue to fail spectacularly. No one is "handing" any control of anything, control is precariously maintained by staying ahead of competition.
Comparing Google to a dictator is ridiculous - who is Google murdering by deprioritizing someone's service on its platform?
between regular people and companies. I don't think it has to come at the expense of those things, a company doesn't have to be the size and scope of google to be efficient.
Companies are made up of regular people, there is no clear distinction between the two. Efficiency is a spectrum, the more efficient a company, the more the poorest of us benefit from cheaper services. Capping efficiency punishes everyone.
armies are also made up of regular people, that doesn't mean they're incapable of doing harm on a scale beyond the individual. you're also falsely equating size and efficiency
And once these regular people decide in numbers that they are tired, miss their wives, and want to go home; it goes disastrously for the commanders who have planned otherwise. For a particularly famous example see Napoleon's retreat from Moscow.
I'm not equating size and efficiency. I'm saying a combination of the two represents the optimal outcome for everyone. In a permissive economy, large, inefficient companies are vulnerable to competition, and tend towards failure.
You find that adwords engages in a blatant "monopolistic abuse" in some specific sense. In the Adwords EU case for example, Google was both monopsonising (google search) and monopolising search advertising. But, the actual violations are specific, like exclusivity requirements and other technical infractions buried in their take-it-or-leave-it contracts with "partners.^"
Google get a fine. They deny culpability. They pay a fine and move on. Whether it's 100m or 2bn doesn't really matter. Google are demonstrably willing to pay far more (eg acquisition, browser kickbacks, etc.) to maintain a monopolistic position. Besides minor tweaks, there's no reason for them to change.
What's the point? Antitrust can't be useful as a violation-fine paradigm. These cases/fines don't affect the monopoly or free the market in any useful sense.
IMO, we need some way to declare a company "Big" and apply special rules to them. If those rules are onerous, that might create some organic pressure to split and/or help smaller incumbents compete.
There is no point to antitrust if its totally reactive, handing out fines for moving violations. The point of antitrust is trust-busting, to affect market structure. The point if to have a competitive market.
All the 1990s stuff is playing out. Just like the MSFT case, we see that operating systems, browsers (now app stores) are monopoly prone and that it's a big deal. On the other side, the 90s "portal" concept is fully realized now. In the digital economy, controlling the windows of consumption is everything. None of the costs, all of the profit... literally all in a lot of cases.
^Newspapers didn't have the advantage search partners had in their monopoly claim, because Google doesn't have agreements with them. No contract, no abusive clauses. The take it or leave it offer is "give us your content for free, or go away." Formally, Google isn't even doing business with them.