Antitrust and Tech companies

The economist Adam Smith (1723-1790) was a strong advocate of free enterprise, and he warned of the dangers to free enterprise posed by monopolies. When monopolies exist, free markets break down. Ever since, there has been some interest in England and America in legislative solutions to the problem of monopolies and anticompetitive business practices.

Antitrust law targets anticompetitive practices, though not necessarily being a monopoly per se. The 1890 Sherman Act, the 1887 Interstate Commerce Act and the 1914 Clayton act forbid price-fixing, collusion, restraint of trade, and anticompetitive mergers. Price-fixing and collusion describe illegal actions between competitors. As for when mergers are "anticompetitive", usually the FTC weighs in, with courts the ultimate arbiters.

The Sherman Act also prohibits "tying"; that is, requiring purchasers of one item to also purchase related items (such as parts, or ink cartridges) from the same company. It is also, generally speaking, illegal for a company to require that repairs be done only by the company's own repair department.

There is also predatory pricing: underpricing your goods in an attempt to drive competitors out of business. This works best when the marginal cost of goods is low; traditional examples are railroad transport (most of the cost is in building the rail network) and oil (most of the cost is in drilling for oil). There is a legal theory that, in general, predatory pricing makes no sense, but note that computers and especially software also tend to have low marginal costs of goods.

The idea behind antitrust laws is to encourage free markets, on the theory that, ultimately, free markets will bring the lowest prices to consumers. One of the earliest uses of US antitrust laws was to break up Standard Oil Company and the American Tobacco Company in 1911.

AT&T

The first "high-tech" antitrust lawsuit was arguably the one filed by the US government against AT&T in 1949; the government was seeking to separate AT&T from its Western Electric manufacturing unit. The case was settled in 1956, with AT&T keeping Western Electric but agreeing to manufacture only telecommunications equipment, and to license its patents on what would now be called FRAND (Fair, Reasonable And Non-Discriminatory) terms. Also in 1956 the Hushaphone case led to allowing third-party devices to be "connected" to AT&T equipment; at that time, all phones were leased by AT&T to customers. The hushaphone was an acoustic privacy muffler, but within ten years there were demands to connect electronically.

A second antitrust case was launched in 1974, with the US government claiming that AT&T was using the profits from Western Electric to subsidize its phone lines. The case was settled in 1982 with the breakup of AT&T. AT&T wanted very much to get into the business of making computer equipment; the settlement removed, among other things, the 1956 constraint on manufacturing non-telecom equipment was dropped. The breakup left the main ATT no longer in the local phone business, and created the seven "regional Bell operating companies": Ameritech, Bell Atlantic, Bell South, Nynex, PacTel, Southwestern Bell, and US West. Southwestern Bell, later renamed SBC, has since acquired Ameritech, BellSouth and PacTel, and the parent ATT itself (and has taken on the ATT name). Bell Atlantic and Nynex ended up in Verizon, and US West was acquired by Qwest in 2000 which in turn was acquired by CenturyLink in 2011.

The breakup of AT&T probably did benefit consumers, but only slightly in terms of lower prices. The real benefit was the rise in availability of additional telecom services: owning your own phone, modem use, advanced call features, etc.

The current antitrust targets are Facebook, Google, Apple and Amazon. Most services for the first two are free, and the traditional indicator of anticompetitive behavior is that it leads to higher prices for consumers. This rule is therefore a little difficult to apply. Still, there are lots of people working on various theories by which Facebook or Google could be considered anticompetitive in a way that harms consumers. One theory is that the data we give up to Facebook and Google helps the companies discover and thwart potential competitors.

IBM

IBM completely dominated the computer business in the 1960's and 1970's; during this period, almost all computers were large multiuser mainfames. For much of this period there were seven smaller mainframe companies often referred to as the "seven dwarves": Burroughs, Univac, NCR, Control Data, Honeywell, General Electric and RCA. The first five on this list were sometimes referred to as the BUNCH. Digital Equipment Corporation (DEC) was founded in 1957 but was never one of the "dwarves", though it later eclipsed all of them. For a while, Loyola owned a Burroughs mainframe for student use.

IBM was sued by the US government for monopoly in 1969, with the government claiming that IBM intentionally created a monopoly, and asking that IBM be broken up. In that year, IBM agreed to unbundle the sales of hardware and software, effectively creating a market for software for the first time. The suit itself, however, dragged on until 1982, at which point the US government dropped it. IBM's share of the mainframe marketplace had by then dropped from 70% to 62%; by this point, DEC had become a major computer vendor.

Microsoft

It was Microsoft's turn in 1998. The issue started in 1995, when Netscape released a better browser, and then a year later Internet Explorer was bundled in with Windows. Microsoft, in fact, insisted that IE be the only browser installed on new machines, if a vendor wanted a bulk windows license (individual windows licenses were and are prohibitively expensive.

MS also famously insisted that to get a bulk license, a computer manufacturer had to pay for Windows for all the machines sold, even if some of them were to be sold with a non-Windows OS (what would that have been? Pre-gnome linux? Maybe the target was IBM's OS/2?).

Finally, a big part of the case was that Microsoft made both Windows and Windows applications, and the MS application programmers appeared to have inside information on Windows. Or maybe it was just information that outside devs never got around to reading.

During the trial, MicroSoft submitted a video of a computer allegedly underfunctioning because IE had been removed. Alas for MS, the video -- presented as representing a single session -- had been spliced. Such actions are often considered to be perjury.

From wikipedia:

When the judge ordered Microsoft to offer a version of Windows which did not include Internet Explorer, Microsoft responded that the company would offer manufacturers a choice: one version of Windows that was obsolete, or another that did not work properly. The judge asked, "It seemed absolutely clear to you that I entered an order that required that you distribute a product that would not work?" David D. Cole, a Microsoft vice president, replied, "In plain English, yes. We followed that order. It wasn't my place to consider the consequences of that."

MS's browser strategy was universally seen as a frontal assault on Netscape, because MS apparently had the idea that it was important to achieve dominance in the "browser" market.

But if you're giving it away free, one theory back then was that there was no market. Does that still make sense, given that the offerings of Google, Facebook and similar companies are almost all free? Alternatively, if you're giving it away free, then it is quite difficult to argue that consumers are being overcharged. (There is an argument that Facebook should actually be paying consumers for their data, but that has never caught on.)

In any event, the primary impetus for the antitrust lawsuit against Microsoft was Netscape. Not consumers.

Once upon a time, some people at MS might have had some notion that, after Netscape was broke, they could resume charging for IE. That is the sort of behavior that antitrust law is intended to prohibit. But a more likely idea was that, if MS controlled the browser market, they would somehow "control" a crucial part of e-commerce. And, to be sure, controlling the browser would mean that they could introduce new server features and be able to guarantee that the browsers out there would support that feature.

As it turned out, controlling the browser market brought about as much control of e-commerce as controlling the cash-register-tape market would have brought control over traditional brick-and-mortar commerce.

MS famously lost their case, at the District Court level. For several years they had to make it possible to remove IE from windows, either by owners or resellers. This was also more or less the death knell for MS's plan to "integrate" the browser with the desktop, ie, to build IE into the desktop.

And, for a while, there was an order in place to split Microsoft into the "Baby Bills": one for Windows, and one for applications.

Did this make any sense? In some ways, integrating the browser into the desktop is not a bad idea. (In other ways, it's a terrible idea.)

A browser is now seen as the reason people buy computers. It needs to come with the computer, if for no other reason that you can't download anything without one. How would I install Firefox on a new Windows computer, for example, if I couldn't use IE once to download it? Would I order a CD by mail?

By 2001, Microsoft won parts of its case on appeal, and, in particular, the breakup plan was scrapped. Instead, the government asked for more mundane restrictions, such as fairer licensing terms.

View 1: Microsoft won

One view is that the DoJ case collapsed in 2001, and Microsoft was allowed to continue with minimal restrictions. It was business as usual.

View 2: The constraints on MS hit at a bad time, and MS never recovered

This view starts with the idea that, starting in 1998, Microsoft was desperate not to be seen as pushing a monopoly agenda. They therefore decided not to compete with several new online services. They debated internally whether traffic to one such service would be shunted by Internet Explorer to their own service, but, in light of the legal climate, decided against it.

That upstart service was Google. The Microsoft alternative here was MSN Search. See:

Microsoft was so powerful, and Google so new, that the young search engine could have been killed off, some insiders at both companies believe. "But there was a new culture of compliance, and we didn't want to get in trouble again, so nothing happened"

MS has tangled with Google since, but no longer from a position of strength. In 2009 they tried to get the Wall Street Journal to remove their news content from Google, in exchange for payment. This is an attempt to get people to have a reason to use bing, the new MS search engine.

Does anyone use bing?

Here's a couple articles about the 2009 kerfuffle:

More seriously, is this a case of antitrust? Or is this a case of exclusive content licensing?

One issue is that Google's use of the WSJ is often considered to be fair use. But Google makes a heck of a lot of money by indexing this content, from advertising. The estimate in the articles above is that it's in the range of $10-15 million/year. This is sort of like the youtube lawsuits, where the media companies really want a piece of the advertising market that youtube gets for displaying "their" videos.

Google

Google's primary business is advertising. This in turn is driven by search, including map search. All Google's other projects are attempts (usually not very successful, by business measures) at diversification.

Google owns the lion's share of the online advertising infrastructure. Google owns the ad-display infrastructure, the ad-information-sharing/user-tracking infrastructure and most of the ad-bidding infrastructure. Google's dominance in user tracking stems in large measure from the input it receives via Google Search.

Publishers have recently introduced "header bidding", in an attempt to wrest some control of advertiser bidding from Google.

Google's control of advertising clearly hurts competing providers of advertising services. But does it hurt consumers?

Google and search

Shivaun and Adam Raff created a price-comparison service, foundem.com, in 2006. Such services are often called "vertical" search. According to nytimes.com/2018/02/20/magazine/the-case-against-google.html, the service was better than Google's own, then known as froogle.com. Users were supposed to find foundem.com through a Google search for a product.

For the first two days, all went well. Then, links to Foundem started to drop in the Google-search pages, quickly reaching the wasteland of page 10 and beyond. For a long time, the Raffs tried to contact people at Google about what they saw as a "technical" problem.

But Google internal emails suggest a different cause. "What is the real threat if we don’t execute on verticals?... Loss of traffic from Google.com because folks search elsewhere for some queries.... If one of our big competitors builds a constellation of high-quality verticals, we are hurt badly...." In other words, maybe Google tweaked its algorithms to make the Raffs disappear.

The Raffs worked extensively with the US FTC, to no avail. They also filed a complaint with the EU. It was not until the appointment of Margrethe Vestager, however, as head of the EU antitrust agency that the case made progress. The investigation began in earnest in 2015, and in 2017 Vestager announced a fine of $2.7 billion against Google. By that point, Google's price-comparison service was open only to paid advertisers, and appeared at the top of the page, though other companies' individual-product listings would also show up, in the regular search results. Other price-comparison sites might not appear at all.

Google has argued that price-comparison sites are obsolete, and that users are most interested in the individual-product listings, their own price-comparison service notwithstanding.This is sort of true, but partly because of Google's past aggressiveness here. Still, "vertical search" isn't the strategy it once seemed, even to Google.

Google, Yelp and TripAdvisor

Yelp.com makes reviews available; TripAdvisor.com focuses on reviews of places you might visit traveling. Google Reviews offers reviews too. Guess which reviews end up near the top of Google search? Well, this isn't as clear as it might seem. If we search for "steakhouse downtown chicago", the first two real links are to tripadvisor.com and yelp.com. But Google has their own box+map at the very top.

Google's position is that people don't want google search to turn up sites where users would have to search again; they just want an answer. And Yelp has been accused of loading up the bad reviews for businesses which refuse to buy Yelp ads.

For more, see nytimes.com/2017/07/01/technology/yelp-google-european-union-antitrust.html.

Chrome v Firefox

In a series of twitter posts, Johnathan Nightingale describes how Google put the muscle on Firefox (twitter.com/johnath/status/1116871231792455686?lang=en). When the Firefox team started working with Google, there was no Chrome. But then:

But Google as a whole is very different than individual googlers. Google Chrome ads started appearing next to Firefox search terms. gmail & gdocs started to experience selective performance issues and bugs on Firefox. Demo sites would falsely block Firefox as “incompatible.”

All of this is stuff you’re allowed to do to compete, of course. But we were still a search partner, so we’d say “hey what gives?” And every time, they’d say, “oops. That was accidental. We’ll fix it in the next push in 2 weeks.”

Over and over. Oops. Another accident. We’ll fix it soon. We want the same things. We’re on the same team. There were dozens of oopses. Hundreds maybe? I’m all for “don’t attribute to malice what can be explained by incompetence” but I don’t believe google is that incompetent.

I think they were running out the clock. We lost users during every oops. And we spent effort and frustration every clock tick on that instead of improving our product. We got outfoxed for a while and by the time we started calling it what it was, a lot of damage had been done.

Facebook

Facebook is probably the best example here of a company that might qualify as a "natural" monopoly: its business works best for its users when everyone is in. If some of your friends were on MySpace, some were on Friendster, some were on LinkedIn and some were on Google+, you'd have a lot of work to do to keep up with everyone. In this, Facebook is a little like a utility.

Still, Facebook is large, and as such it attracts lots of interest today in antitrust restraints. For the core facebook.com, it's hard to envision a breakup proposal, or even a data-sharing proposal with competitors. But Facebook has been acquiring other social-media sites, like Instagram and Whatsapp; there are regular suggestions that Facebook be required to spin these off. There are also frequent regrets that Facebook was ever allowed to buy them.

Facebook apparently sees Instagram as a major alternative to facebook.com, popular with younger users. To Facebook, having to give up instagram would be an existential threat.


Amazon

Amazon is not a monopoly. Yet. But are they doing anticompetitive things?

"Predatory pricing" is charging less to drive your competitors out of business, and then raising prices. Is that Amazon's strategy? See inthesetimes.com/article/21850/is-amazon-using-predatory-pricing-in-violation-of-antitrust-laws-monopoly. The theory presented there, developed by Shaoul Sussman, is that Amazon is using accounting tricks (like leasing everything, instead of buying) to conceal the fact that it is underpricing the items it sells.

And there are the cases of Diapers.com and Zappos.com.

Amazon may also be trying to convert many of its own wholesalers to becoming third-party Amazon Marketplace sellers. This eliminates Amazon's risk, and effectively means that Amazon's costs become lower. There is an alternative, slightly contradictory, theory of Amazon's competitive advantage: they can put their own products (from their own wholesalers) in the top position in Amazon search. Already, the Amazon brand is often seen as the "easy choice" in an overwhelming sea of options. For discussion of this second strategy, see propublica.org/article/amazons-new-competitive-advantage-putting-its-own-products-first.

The New York Times ran an article, nytimes.com/2019/06/23/technology/amazon-domination-bookstore-books.html, suggesting that Amazon's bookstore -- which already is a de facto monopoly -- can shed some light on Amazon's long-term strategy. Book prices are not going up, but book sellers are under enormous pressure to cut deals with Amazon because Amazon otherwise will do very little to remove counterfeit versions of the book from its online store. Amazon doesn't appear to be the source of the counterfeits, but they will only act if the original publisher complains, putting the onus on that publisher to police Amazon's content.

Amazon Prime introduces its own anticompetitive wrinkle. Who is going to sign up for a Prime-style membership with some small competitor to Amazon? Or even with a big competitor (eg walmart.com) to Amazon? List-price Prime is $10/month ($120/year), which won't cover traditional shipping charges for two orders a month. So Prime would appear to be significantly underpriced, as in predatory pricing.

A similar argument applies to Amazon's strategy of offering to let you provide a key, so they can deliver your stuff to the inside of your home. It is hard to see how smaller competitors can match this, because you're not likely to provide a door key to anyone else. Does that give Amazon an unfair leg up?

Mostly, Amazon is seen as providing relatively lower prices for consumers. Traditional antitrust theory requires proof of harm to consumers, and amazon has always argued they reduce prices to consumers. But in 2019 Amazon was charged with causing higher prices; see bloomberg.com/news/articles/2019-11-08/amazon-merchant-lays-out-antitrust-case-in-letter-to-congress.

In a letter sent to federal lawmakers, an online merchant has accused Amazon.com Inc. of forcing him and other sellers to use the company’s expensive logistics services, which in turn forces them to raise prices for consumers. 

But what value do you put on 2-day delivery?

Apple

To run an app on an iPhone, it pretty much has to be in the Apple App Store. Apple's stated reason for this is security, and they have indeed been quite successful at keeping malware and spyware off of iPhones. But they charge 30% of an app's fees (special rules apply to continuing subscriptions, like Spotify, and no fee is charged to free apps that sell non-app merchandise, like Amazon).

(Google also has some restrictive app-store policies, and they also try to keep malware out, though at the latter they have been much less successful than Apple.)

Despite the price break for continuing subscriptions, Spotify is still mad at Apple: see TimeToPlayFair.com. Siri won't play Spotify content. Apple's rules make it very hard for Spotify to offer its subscription plans directly to users. Finally, Apple has a history of rejecting Spotify's app upgrades.

Apple responded to Spotify's TimeToPlayFair with a press release: apple.com/newsroom/2019/03/addressing-spotifys-claims, in which they pointed out the 30% commission falls to 15% after the first year. But, still, Apple's pricing model seems simply to ignore the needs of Spotify. Apple could allow Spotify users to subscribe via credit card, but they do not.

Apple is quite strict with app developers, and their code review is much more rigorous than Google's for Android. The New York Times released an article "Apple Cracks Down on Apps That Fight iPhone Addiction", alleging that Apple was pushing its own anti-online-addiction tools. But the reality is more complex: many of the apps in question were quite invasive in terms of user privacy. Apple is probably right on this one.

Game vendor Epic Games, maker of Fortnite, got itself kicked off the App Store for changing their rules on in-app purchases. Their antitrust case against Apple wrapped up testimony in May 2021, with Apple CEO Tim Cook as a witness.

On the one hand, the Apple App Store (the only way to get apps for an iPhone) does add value, by doing some privacy and security vetting. On the other hand, a 30% flat percentage means big apps do a heck of a lot of subsidizing of the vetting of small apps. There's no real reason to think that the difficulty of vetting an app varies much from app to app, though larger apps might take more work.

Despite occasional rumblings from Microsoft and Apple, it would be inconceivable for general-purpose-computer makers to restrict software that can be run. Why are phones different? Make that "why are iPhones different"?

Note that if Spotify or Epic could offer their content as a web app, they would be home free. Apple couldn't touch them. So Apple's insistence isn't quite the consistent policy it might seem.

Most game-console manufacturers do charge a fee to devs for game publication. But this is a very specialized market: game consoles have to be sold at nearly cost, or even lower, and game devs would have no platform if it weren't for the console maker.

Some references:

And there has been weird collateral fallout: at one point Apple ordered Wordpress, a free app, to create non-free tiers so Apple could get 30% of something. Then Apple backed off. Apple claimed Wordpress had agreed to make changes, but that seems false: www.theverge.com/2020/8/22/21397424/apple-wordpress-apology-iap-free-ios-app.

Also, the CEO of Epic Games had apparently asked weeks before their big in-app-purchase policy change for a break from the 30% rule. Apple said no. See also nytimes.com/2020/08/25/technology/fortnite-creator-tim-sweeney-apple-google.html.

Antitrust law prohibits monopolies due to market coercion. Apple has a monopoly on the iPhone app store. Is that even a "market"? Just what is a market? Antitrust laws also prohibit "tying". If to buy an iPhone you have to buy apps from Apple, is that tying? But it's not really buying apps from Apple; it's that they get a 30% share.

Webkit

Browsers are built on top a browser "engine". Chrome's is blink, and Firefox's is gecko (or maybe quantum). Neither of these are allowed by Apple on iPhones. So Firefox and Chome have to port their browsers to run on top of Apple's engine, webkit. Apple's official reason is security, but Apple's own Safari browser makes it clear that Apple's security leaves something to be desired, both in time to issue patches and in total vulnerabilities identified.

Webkit makes certain app-like features of browsers hard to implement. For example, on webkit-based browsers, notifications are more complicated. One doesn't need notifications for most things, but for a messaging app, or for some games, it is essential. Google would like to ship Chrome for the iPhone based on their blink engine, and Firefox would like to ship their Gecko-based browser. But Apple does not allow this: no app can be allowed into the Apple app store without Apple's agreement.

Why does Apple do this? One theory is that this keeps out smaller browsers; another is that it makes all browsers look as bad as Safari, so what's the point of switching. But maybe the most likely reason is that using webkit puts iOS browsers at a severe disadvantage versus using Apple's app tools, which then locks developers into Apple's 30% app-payment rule. That is, webkit-based browsers force companies to move to apps (not that companies aren't already likely to prefer apps). If a web-based app charges, say for a subscription, Apple can't collect their 30% tax. But apps, absolutely yes. So Apple has a huge financial interest in pushing everything to apps. Examples include newspaper and magazine subscriptions, Spotify, and almost all games. See twitter.com/OpenWebAdvocacy/status/1541318055636369409.

Network Neutrality

One question here is whether ISPs should be allowed to throttle content from content providers that don't pay "fees". Is that antitrust? Or is it all about The Free Market?

Alas, at the heart of Network Neutrality is the fact that it is cable companies that bring most Americans their internet. So when Netflix or Hulu or Google (as owner of YouTube) argue in favor of network neutrality, the opposition is trying to get you to watch TV as cable TV rather than as internet. If an ISP were to reduce customers' bandwidth for YouTube videos because Google did not pay them a fee, that might be a pretty solid example of Network un-Neutrality. But what if Comcast applies their bandwidth cap to Netflix and Hulu, but not to their own TV Anywhere, which is actually a cable product and not, technically, an internet product?