"The court rejected the notion that additional data increases the barriers to competitive entry, and indicated that this issue “requires closer examination and a detailed explanation/review and conclusive presentation by the antitrust authorities.”"
Maybe someone who knows more about this topic could clarify but this suggests that the court believes network effects and scale don't affect the barrier to entry? (pending further reviews)
I would be very interested in learning more about how they came to this decision as from what I've seen a number of large tech products (e.g. youtube) only seem to become feasible at extreme scale.
So I strongly disagreed with that finding, but two possible lines of reasoning:
1) Say you have a machine learning algorithm, where there could be two ways to improve the outcomes: get more data, or pick a better model. Google's massive dataset is a benefit/competitive moat of sorts to its search algorithm, but _in theory_ you could build a better algorithm or model. You can see some evidence of this in vertical specific searches (i.e. travel, shopping, etc.) and in DuckDuckGo's success.
2) The other refrain I hear a lot is that this additional data is not "zero-sum" - i.e. multiple apps can collect my location data or email or [insert data point here,] and I can onboard or offboard that data pretty easily. Therefore this data isn't a limited resource that can be hoarded like a true competitive moat.
I would suggest that monopolies in one area are not that harmful in tech because tech changes so rapidly that monopolies are quickly undermined by "the next thing". What's damaging is when tech companies have a monopoly in one area that they leverage to extend their monopoly into "the next thing". Famous example is Windows bundling Internet Explorer. Google is now doing this kind of thing... Example... Google used to allow businesses to add their own SMS capable number to their "My Business" profile to allow customers to message them. They have since removed that option pending the release of their own API for messaging. They are using their monopoly in search to extend their power into communication. This harms the businesses (like mine) that were assisting companies with this messaging and that harm generates damages. The way for the legal system to stop these monopolies indefinite expansion is to find the actual places where they are restricting competition/innovation and generating monetary harm to others in the process... point out these issues and make them change their behavior, pay or both. If you disagree I'm interested in how you think the power of these tech giants can be constrained?
> companies have a monopoly in one area that they leverage to extend their monopoly into "the next thing."
This is also illegal in the States for being anti-competitive. e.g. if you have a dominant market position in one market, and leverage that to enter another market and/or gain market share, this tends to be the source of many Sherman cases.
Ex: Company X is a monopoly in market A but wants to enter market B. If I use extracted rents from A to significantly undercut prices below competitive level in market B to gain share, even if it's unprofitable in the short term - THAT's a problem.
Note that startups don't get as much scrutiny when they do this, partly because it's very rare for a startup to have dominant market share.
It's not even that they are using "extracted rents from A to significantly undercut prices below competitive level in market B to gain share" they are locking us and other companies out of the market altogether. This is a crime and it's happening right now.
Always useful to keep in mind in discussions of monopoly: there is a big philosophical gap regarding why one might care a monopoly exists, and that gap can guide thinking on the topic in subtle ways.
American law tends to center around harm to consumers. By that reasoning, it can be hard to prove a market with a single player is actually anticompetitive if that player isn't muscling out better options for consumers from the space (i.e. if they corner 90% of the market because they really do offer the best value at the lowest price, who cares).
European laws tend to center around the philosophical core that the market should be a level playing field for producers or vendors, so even if all other choices than the dominant player are worse for consumers, the market needs to stay competitive so that business firms don't die and put a bunch of people out of jobs.
It's useful to know where someone is coming from in considering topics like this.
> American law tends to center around harm to consumers. By that reasoning, it can be hard to prove a market with a single player is actually anticompetitive if that player isn't muscling out better options for consumers from the space.
I think censorship changes the equation a lot. Because of the right of free association, a company can choose to favor a side of the political spectrum. This is not problematic in the absence of a monopoly as there will likely be a diverse range of companies, and new companies can enter that cater to one side or the other.
However, if a company is a monopoly with a huge barrier to entry because of network effects and/or data, it can be argued that the monopoly is harmful to consumers on the other side of the political spectrum because they will not get the chance to promote their views.
The thing is, if Google decides to favor a side and Bing doesn't, Bing search is just a click away. And if consumers value a politically-unbiased (or other-biased) opinion, they can migrate to Bing, right?
They need to know this bias exist first. And they also need to know that there is alternatives. And then they also need to know how to change. This is a lot to ask from a lot of people I'm affraid. Are you willing to bet your and your countrys future on people switching to bing when someone manages to game Googles algorithm to favor a side in the next election in whatever country you live in?
That's a large and complicated question. To a first approximation: if people aren't smart enough to figure out if they need to do that and how to do that, then my country's future is already in a tailspin because my country is a representative democracy and a minimal amount of collective intelligence and savvy is necessary to maintain such a system of government (hence, the public schools).
Arguments about potential political bias are built on some pretty shaky and dangerous hypotheses. At this point, the number of current and former employees who have seen Google’s ranking algorithm and not whistle blown about some political signal makes its existence highly improbable.
Definitely. To be clear: my previous lines of reasoning here were assuming a political bias from Google and noting that even if it exists, the populace has alternatives. I'm personally of the school of thought "Google seems to have liberal bias in the US because relative to US politics, reality has a liberal bias," and not of the conspiracy-to-delude-the-public school of thought. ;)
That's why I said in "American vernacular" because it's not the dictionary definition. Typically businesses/regulators distinguish between business-to-business (B2B) and business-to-consumer (B2C) transactions, the former is what you're describing, the latter is what OP is describing, and relates to any transaction between a business and an individual. it's also why we have markets like "consumer electronics" which refers to products sold to individuals, ie "consumers", and not industrial/business sectors.
I don't want to argue semantics though, since it's very clear to any native English speaker what "consumer" means in the context of regulation - it means individuals, not businesses. I get that the connotation of the word may be lost in translation, and that's what I was getting at.
In addition, "payment" isn't strictly hard currency, it's the exchange of value. Consumers are purchasing Google services with their data. They aren't the literal product that Google sells to advertisers, the data is.
> They aren't the literal product that Google sells to advertisers, the data is.
Google does not sell data. They sell attention, like all advertising companies. They make a lot of money because 1. they have a lot of attention to sell and 2. they do a better job than most advertisers at finding the particular bits of attention that will be most valuable to the buyers. The data is part of what allows them to do that, or so the story goes.
Also, although I agree with your conclusion, I think you've been slightly misled as to what matters in the antitrust investigation. It is not the colloquial understanding of the word "consumer" that matters, it's the legal one. I'm under the impression that, as it happens, the legal one here comports with the colloquial one. I.e., what matters is the impact of Google's policies and practices on individual private citizens and residents. The impact on other businesses does not matter except inasmuch as such impact transitively harms private consumers.
Then it's a good thing there are a lot of filings to the DOJ from Google competitors (in various niche markets) that say Google has been trying to strongarm them out of the market.
The only reason Google wasn't broken apart in 2012 when the last FTC antitrust case happened, is because Eric Schmidt was best buddies with Obama (who then got Schmidt a job at the Pentagon - while still working at Google - and Schmidt then got Google the Project Maven contract with the Pentagon, and so on, and so forth, if you want to go down that rabbit hole).
I think there is a more fundamental philosophical gap is on why we even have a market economy. If you primarily care about private property, the American policy makes sense as the state only interferes when it's undeniably necessary. If what you care about is the common good and you think letting companies compete is the best way to run an economy, only the European policy makes sense: You probably consider it impossible that lack of competition does not harm the customer. Otherwise it would be hard to defend not nationalizing the sector in question.
“[The court] found that consumers were not harmed by Facebook, which had informed them about data use practices in a lengthy terms of service agreement when they registered as users. “There is no evidence that Facebook obtains the consent of users through coercion, pressure, exploitation of lack of willpower or otherwise unfair means.” “
The court seems to have ignored the cases in which consent is never obtained. Shadow profiling is still happening isn’t it?
The German court review of the investigation declared consumers are not harmed because
> “ There is no evidence that Facebook obtains the consent of users through coercion, pressure, exploitation of lack of willpower or otherwise unfair means.”
> The court further found that “whether the users act out of indifference or because they do not want to spend the necessary time and effort … does not matter. Their decision to sign on is ultimately free, uninfluenced and autonomous.”
- so it’s okay to obfuscate consumers with lengthy legal docs, that’s not considered misleading (gutless but unsurprising ruling)
- no mention of the competitive threat of 3 of 4 giant companies tracking all our online behavior, giant companies with massive VC war chests or close ties to the same, who can pick winners and losers to suit their interests and the interests of their few board-seated investors
As far as I see, this is the death of free-market capitalism. We are entering an era of gated-market capitalism. If will erode incentives and ultimately productivity.
22 comments
[ 670 ms ] story [ 147 ms ] threadMaybe someone who knows more about this topic could clarify but this suggests that the court believes network effects and scale don't affect the barrier to entry? (pending further reviews)
I would be very interested in learning more about how they came to this decision as from what I've seen a number of large tech products (e.g. youtube) only seem to become feasible at extreme scale.
1) Say you have a machine learning algorithm, where there could be two ways to improve the outcomes: get more data, or pick a better model. Google's massive dataset is a benefit/competitive moat of sorts to its search algorithm, but _in theory_ you could build a better algorithm or model. You can see some evidence of this in vertical specific searches (i.e. travel, shopping, etc.) and in DuckDuckGo's success.
2) The other refrain I hear a lot is that this additional data is not "zero-sum" - i.e. multiple apps can collect my location data or email or [insert data point here,] and I can onboard or offboard that data pretty easily. Therefore this data isn't a limited resource that can be hoarded like a true competitive moat.
This is also illegal in the States for being anti-competitive. e.g. if you have a dominant market position in one market, and leverage that to enter another market and/or gain market share, this tends to be the source of many Sherman cases.
Ex: Company X is a monopoly in market A but wants to enter market B. If I use extracted rents from A to significantly undercut prices below competitive level in market B to gain share, even if it's unprofitable in the short term - THAT's a problem.
Note that startups don't get as much scrutiny when they do this, partly because it's very rare for a startup to have dominant market share.
American law tends to center around harm to consumers. By that reasoning, it can be hard to prove a market with a single player is actually anticompetitive if that player isn't muscling out better options for consumers from the space (i.e. if they corner 90% of the market because they really do offer the best value at the lowest price, who cares).
European laws tend to center around the philosophical core that the market should be a level playing field for producers or vendors, so even if all other choices than the dominant player are worse for consumers, the market needs to stay competitive so that business firms don't die and put a bunch of people out of jobs.
It's useful to know where someone is coming from in considering topics like this.
I think censorship changes the equation a lot. Because of the right of free association, a company can choose to favor a side of the political spectrum. This is not problematic in the absence of a monopoly as there will likely be a diverse range of companies, and new companies can enter that cater to one side or the other.
However, if a company is a monopoly with a huge barrier to entry because of network effects and/or data, it can be argued that the monopoly is harmful to consumers on the other side of the political spectrum because they will not get the chance to promote their views.
a person who purchases goods and services for personal use.
And definition of “customer”
a person or organization that buys goods or services from a store or business.
In both cases the consumer or customer is the person paying. Who is paying and what is the “product”?
I hate to go back to the tired saying but, the people who search on Google are the “product” that Google is selling to the customer - the advertiser.
Why would anyone expect Google to optimize for anyone else except for the paying customer/consumer?
I don't want to argue semantics though, since it's very clear to any native English speaker what "consumer" means in the context of regulation - it means individuals, not businesses. I get that the connotation of the word may be lost in translation, and that's what I was getting at.
In addition, "payment" isn't strictly hard currency, it's the exchange of value. Consumers are purchasing Google services with their data. They aren't the literal product that Google sells to advertisers, the data is.
Google does not sell data. They sell attention, like all advertising companies. They make a lot of money because 1. they have a lot of attention to sell and 2. they do a better job than most advertisers at finding the particular bits of attention that will be most valuable to the buyers. The data is part of what allows them to do that, or so the story goes.
Also, although I agree with your conclusion, I think you've been slightly misled as to what matters in the antitrust investigation. It is not the colloquial understanding of the word "consumer" that matters, it's the legal one. I'm under the impression that, as it happens, the legal one here comports with the colloquial one. I.e., what matters is the impact of Google's policies and practices on individual private citizens and residents. The impact on other businesses does not matter except inasmuch as such impact transitively harms private consumers.
Have we forgotten how Google killed Skyhook?
https://www.theverge.com/2011/05/12/google-android-skyhook-l...
The only reason Google wasn't broken apart in 2012 when the last FTC antitrust case happened, is because Eric Schmidt was best buddies with Obama (who then got Schmidt a job at the Pentagon - while still working at Google - and Schmidt then got Google the Project Maven contract with the Pentagon, and so on, and so forth, if you want to go down that rabbit hole).
The court seems to have ignored the cases in which consent is never obtained. Shadow profiling is still happening isn’t it?
> “ There is no evidence that Facebook obtains the consent of users through coercion, pressure, exploitation of lack of willpower or otherwise unfair means.”
> The court further found that “whether the users act out of indifference or because they do not want to spend the necessary time and effort … does not matter. Their decision to sign on is ultimately free, uninfluenced and autonomous.”
- so it’s okay to obfuscate consumers with lengthy legal docs, that’s not considered misleading (gutless but unsurprising ruling)
- no mention of the competitive threat of 3 of 4 giant companies tracking all our online behavior, giant companies with massive VC war chests or close ties to the same, who can pick winners and losers to suit their interests and the interests of their few board-seated investors
As far as I see, this is the death of free-market capitalism. We are entering an era of gated-market capitalism. If will erode incentives and ultimately productivity.