U.S. asks Google for detailed search data in antitrust case
bnnbloomberg.ca
bnnbloomberg.ca
If antitrust is entirely delegated to courts, it's hard to see how it ever becomes more than a "compliance" issue. Fines and/or mandatory adjustments to current MOs. That's hardly trustbusting.
The big pieces aren't a secret. Market share, the big competitive dynamics, income streams and overall structure of the search-ads "industry" and its adjacent, also Google dominated markets like android, youtube and chrome.
How is a judge supposed to decide what to do about all this... Seems like the wrong tool for the job.
Beyond the fact that it isn't "pretending" to be associated with Bloomberg, as another commenter points out, I hardly see anything controversial in the reporting that the DOJ has requested information from Google.
In this new broken up set of companies, what's stopping them from remerging like Bell? Also, what's stopping them from giving each other favorable partnerships?
Personally I find this entire exercise a waste of time - rather than break up Google we should just acknowledge that certain industries have certain types of attributes that result in necessarily "anti-competitive" behavior. In Google's case, since it's an internet company I would say the more users it has the harder it is to beat - with that being said they should just create some sort of progressive taxation as a function of users.
Your users in the USA equal 90%+ of the country, you are subject to rules A. Your users in the USA equal to <10%? You are subject to rules B and are subsidized directly by those whose users equal 90%+. If all companies are less than 50% than we use rules c, for example.
The same logic could be used for manufacturing, utilities, internet providers, and more.
Continued antitrust enforcement. Nobody should ever have approved those mergers. We can learn from our mistakes.
> Also, what's stopping them from giving each other favorable partnerships?
Why would they want to do that once they have separate owners?
> just create some sort of progressive taxation as a function of users.
How does this solve the monopoly problem at all? It has nothing to do with taxation.
Long distance calls where crazy expensive. During weekdays in 1980 AT&T charged $2.17 ($6 in today's money) for a five minute call from NYC to LA. Nights and weekends the same call cost $0.87 ($2.40). Today you can get unlimited nationwide for 34$ per month or pay 3$/month for service and get 10c/minute long distance calls anytime. https://www.att.com/home-phone/landline/long-distance.html
Breakups across verticals make a lot more sense. So separating last mile from long distance, and separating Unix from AT&T, those were actually successes.
Once physical connections is public infrastructure, multiple providers can pay for capacity and sell to the end user with great competition resulting.
The tradeoff is basically: If you don't split the infrastructure, you have people experiencing monopoly of ISPs. If you do, you get ISPs experiencing monopoly of Openreach-like organisation.
Neither situation is amazing, but I believe the second option is still better - and it was a way to prevent BT dictating everything. It's also possible for Ofcom (gov watchdog) now to monitor and slap either ISPs or Openreach for not behaving well.
It doesn't really.
Openreach was not broken up from BT, it only created a seperate department within BT that continued to skirt various rules.
> It's also possible for Ofcom (gov watchdog) now to monitor and slap either ISPs or Openreach for not behaving well.
Ofcom broke up BT so that O2 and BT were seperate entities because allowing BT to have a dominant mobile network too was considered strongly wrong for competition.
Ofcom allowed the larger mobile networks t-mobile and orange to merge together to become a mega huge mobile network.
Turned around and said no to Three and O2 merging despite being a significant market disadvantage even after merger due to being smaller and having access to significantly less Spectrum. Unlike landlines, mobile networks are heavily restricted by needing radio spectrum to grow.
Allowed BT to acquire EE. Allowed BT to acquire one of the larger virtual ISPs PlusNET. Ofcom denied alternate infrastructure provider Virgin from setting up further competitive mobile network infrastructure and forces them to exist a MVNO.
Ofcom have ignored the fact that BT have been able to offer FTTH connections through their infrastructure to customers before any other ISP is permitted to use it over OpenReach.
"like the UK did it" isn't exactly great.
The OpenReach method however is not exactly unlike the FCC, they do requiring a certain amount of sharing through infrastructure via the Wireline Competition Bureau (WCB) and Wireless Telecommunications Bureau (WTB), but one significant difference to offer coverage of other services. Networks operating in the same area have a tendency to reuse the same infrastructure and mutually piggybacking on each other. There is one added difficulty that State laws do exist to frustrate the federal requirements in place.
And remember ofcoms #1 job is keep Murdoch and the other media barrons happy.
Very few organizations are willing to build a third network when an area is already covered by two networks; the cost to build is too high, and the number of customers you'll attract is too low; and the profit margins are too slim, given how quickly incumbent networks can be made competitive when there's a will.
Vanishingly few would overbuild a fourth network.
Often two, sometimes three, and almost never four competitors is not a healthy marketplace. There's no way for a regular person to say 'my ISP stinks, I'm going to go out and do a better job', like you could in the dial-up days by getting a T1 for data, and a T1 for users to call in on. Yes, T1s were expensive, but they weren't as expensive as wiring up to every customer. If we want competition in ISPs, we need local loop unbundling, but for realsies; the companies owning and operating last mile networks should be strictly separate from the companies providing Internet Protocol services on those networks.
Comcast.
Recently a real competitor opened up. Starlink. They're still more expensive but at least now I technically have a choice other than maybe 7M bit down not even 0.75M bit up DSL. (Century Link, formerly lots of other companies.)
In Washington you can sometimes do better in certain rural counties thanks to municipal fiber networks maintained by the local PUD. However, the rules surrounding them appeared complicated last time I checked and last mile infrastructure varies wildly by location.
What would be really nice (in my opinion) is if there was a way to share installation costs without organizing. I've got an installation quote where it's about $30k to build to my pole (and then $20k more to get from my pole to my building, because I have underground utilities from the pole and a long driveway). If I organize a bunch of neighbors along the route, we can share costs on that $30k; but if I just install for me, everyone else along the route can use it at no cost to them. It would be nice if I could pay the $30k now, and get some of that rebated later if other people use the line in some near term future. Although, probably the neighbors would just wait it out; Comcast serves most of them (but not me), and CenturyLink gets me 80/10 mbps on bonded VDSL2, so I imagine most of them can get pretty similar too.
How many is Google getting right now?
Why or how would they end up with separate owners? The current owners of Alphabet would get equal shares in all demerged companies, I would hope. And presumably with the same 3-class stock system, so they would all be controlled by the same people, too.
Because the established business model is proven to be super effective and all the organizations and infrastructures are already optimized for the such model? The real question is, why would they deviate from the existing anti-competitive behavior if reforming the "Google cartel" with business partnership will give them a roughly same level of functionality?
You gotta understand that their incentive functions are already tightly aligned with each other because that's what network effect is precisely supposed to do, so it will naturally head toward anti-competitive regardless of whether they are separate entities or not. The only meaningful action to prevent that would be highly targeted and specific antitrust remedies.
how would you demonstrate consumer harm when youtube was giving away the product for free?...or instagram...or whatsapp..
https://www.npr.org/sections/money/2019/03/20/704426033/anti...
So the cost is the terms under which the product is offered, e.g. the conditions under which you can be excluded at the behest of advertiser sensibilities. If those conditions aren't that nobody is ever banned for any reason then the cost is higher than it could be in a more competitive market.
You also have similar costs as a result of algorithmic recommendation algorithms that promote things that aren't you over things that are you at the behest of the monopolist, and the amount content providers get paid by the platform (it could be higher or you could be not demonetized for nonsense reasons) etc.
"You are not the customer, you're the product. The customers are advertisers, that are buying YOUR data from the google."
That should make it clearer, right?
That’s a nice thought but clearly doesn’t work in practice. I could be wrong though, are there any examples you can cite of monopolies being broken and staying that way in the USA?
> Why would they want to do that once they have separate owners?
They’re already familiar with each other.
> How does this solve the monopoly problem at all? It has nothing to do with taxation.
You don’t think a unavoidable tax that goes into smaller companies won’t hurt monopolies?
It doesn't work if there is regulatory capture. But if the government is captured then you can't have your tax either, so then what are you proposing as a solution?
> They’re already familiar with each other.
Google and Oracle are familiar with each other. Apple and Intel are familiar with each other. Amazon and Netflix are familiar with each other.
> You don’t think a unavoidable tax that goes into smaller companies won’t hurt monopolies?
If they can keep abusing their monopolies to prevent there from being any competitors? There still won't be any competitors.
> If they can keep abusing their monopolies to prevent there from being any competitors? There still won't be any competitors.
Which monopolies are you referring to have zero competitors?
> Google and Oracle are familiar with each other. Apple and Intel are familiar with each other. Amazon and Netflix are familiar with each other.
I don’t understand your point here, you’re listing pairs of competitors. Are you really not aware of large companies helping each other out?
In order to pass some kind of tax, you need regulators to pass some kind of tax. If the government is captured by the companies, they don't pass any kind of anything the company doesn't want, so you lose. If it isn't captured by the companies, you can do antitrust enforcement.
> Which monopolies are you referring to have zero competitors?
Apple and Google each have a monopoly on app stores for their respective platforms, for example.
Note that "monopoly" for antitrust doesn't mean that there are exactly zero competitors. It only means that one of them dominates the market and has market power.
> I don’t understand your point here, you’re listing pairs of competitors. Are you really not aware of large companies helping each other out?
Google and Oracle hate each other and are constantly embroiled in bitter litigation. Apple just jettisoned Intel's processors in favor of their own. Netflix has used Amazon (AWS) for hosting and it didn't stop Amazon from going into competition with them for content production (Amazon Prime Video).
Companies do what they like. What they like is to make money. One of the best ways to do that is to go make the money that some other company is currently making, or "commiditize your complement."[1] It isn't to altruistically help out competitors for no reason.
The terms of the consent decree or judgement that gets issued. In terms of partnerships, the terms should require that all deals between the babyGs should be public, and a non-discriminatory, non-exclusive basis; along with requirements for some of the babyGs to not own or operate in some areas, or to only own or operate in some areas.
I think that could go a long way towards making things competitive.
If Google was broken up it's entirely plausible that by the time it re-merged search would either be a commodity service or just not that important.
We also need to continue to regulate monopolies, because if we lapse again and let the mergers flow we’ll be right back where we are again. And again.
The other piece of the Bell breakup that people always overlook is that the federal directive did not merely break up the corporation. It also mandated that all Bell System patents be made available for use without fee by possible competitors. The result was an explosion of innovation and technological development (including within Bell companies!) and the consequences were massive.
This holds for tariffs as well. If we are overly reliant on imported facemasks, we should tax them to subsidize a small amount of domestic production in case of catastrophe.
I disagree. If in a certain industry we cannot accept the risk incurred by being reliant on AWS we should legislate against that by requiring insurance.
That's effectively a tax, outsourcing collection to insurance companies. I'm fine with that, too.
AWS is huge in a segment (cloud) of the industry, but that segment represents a very small % of the overall industry.
Mind you, the Baby Bells were allowed to re-merge too quickly. We still lack competition in many areas. Imagine if there was no breakup at all.
doesn't seem like it.
It's a very specific date range to help build the case. The article was very clear about this.
Antitrust cases are rare, so even if they did do inappropriate things, it would only happen every antitrust investigation.
Plus, these data are already given to NSA in total and large chunks to law enforcement.
Privacy is dead. Everyday people are not concerned about privacy. They don't care that someone will potentially see their "How do I cook a hotdog" search.
Anonymization is partly a result of aggregation, but also I can't see how and PII is relevant to the case here.
Now some more bureaucrats and sys admins can take a peek too.
There is no reason to think this isn't still going on.
But it is also true that Google and ATT are very different companies. The Snowden leaks lead Google to encrypt internal links, once they realized NSA were tapping them; ATT built NSA Room 641A.
But this is rather more complicated than that; I'd bet money there are are people employed at Google who wear more than one hat[1], as there have been at tech companies throughout the history of Silicon Valley. And practices at Google makes NSA's life easier - Organizing the World's Advertising Targets is actually a remarkably compatible goal.
also, with Duck Duck Go (and I assume others) I think it's more than debatable that Google is the defacto "pro-user" choice
Plus, one wonders if Google even holds the data the US government is requesting.
Plusplus, I hate it when "granular" is used as if it has a direction, which it does not. Are the feds asking for fine-grained data, or coarse-grained data?
It's an interesting world view where being a random Googler apparently suggests higher priority than a DOJ request.
Moving that much sensitive data is a big task with invokves many interested parties.
lol... I don't wonder this at all... but either ways, the government could potentially get it in real time (like the NSA, today).