Archives For

Microsoft wants you to believe that Google’s business practices stifle competition and harm consumers. Again.

The latest volley in its tiresome and ironic campaign to bludgeon Google with the same regulatory club once used against Microsoft itself is the company’s effort to foment an Android-related antitrust case in Europe.

In a recent polemicMicrosoft consultant (and business school professor) Ben Edelman denounces Google for requiring that, if device manufacturers want to pre-install key Google apps on Android devices, they “must install all the apps Google specifies, with the prominence Google requires, including setting these apps as defaults where Google instructs.” Edelman trots out gasp-worthy “secret” licensing agreements that he claims support his allegation (more on this later).

Similarly, a recent Wall Street Journal article, “Android’s ‘Open’ System Has Limits,” cites Edelman’s claim that limits on the licensing of Google’s proprietary apps mean that the Android operating system isn’t truly open source and comes with “strings attached.”

In fact, along with the Microsoft-funded trade organization FairSearch, Edelman has gone so far as to charge that this “tying” constitutes an antitrust violation. It is this claim that Microsoft and a network of proxies brought to the Commission when their efforts to manufacture a search-neutrality-based competition case against Google failed.

But before getting too caught up in the latest round of anti-Google hysteria, it’s worth noting that the Federal Trade Commission has already reviewed these claims. After a thorough, two-year inquiry, the FTC found the antitrust arguments against Google to be without merit. The South Korea Fair Trade Commission conducted its own two year investigation into Google’s Android business practices and dismissed the claims before it as meritless, as well.

Taking on Edelman and FairSearch with an exhaustive scholarly analysis, German law professor Torsten Koerber recently assessed the nature of competition among mobile operating systems and concluded that:

(T)he (EU) Fairsearch complaint ultimately does not aim to protect competition or consumers, as it pretends to. It rather strives to shelter Microsoft from competition by abusing competition law to attack Google’s business model and subvert competition.

It’s time to take a step back and consider the real issues at play.

In order to argue that Google has an iron grip on Android, Edelman’s analysis relies heavily on ”secret” Google licensing agreements — “MADAs” (Mobile Application Distribution Agreements) — trotted out with such fanfare one might think it was the first time two companies ever had a written contract (or tried to keep it confidential).

For Edelman, these agreements “suppress competition” with “no plausible pro-consumer benefits.” He writes, “I see no way to reconcile the MADA restrictions with [Android openness].”

Conveniently, however, Edelman neglects to cite to Section 2.6 of the MADA:

The parties will create an open environment for the Devices by making all Android Products and Android Application Programming Interfaces available and open on the Devices and will take no action to limit or restrict the Android platform.

Professor Korber’s analysis provides a straight-forward explanation of the relationship between Android and its OEM licensees:

Google offers Android to OEMs on a royalty-free basis. The licensees are free to download, distribute and even modify the Android code as they like. OEMs can create mobile devices that run “pure” Android…or they can apply their own user interfaces (IO) and thereby hide most of the underlying Android system (e.g. Samsung’s “TouchWiz” or HTC’s “Sense”). OEMs make ample use of this option.

The truth is that the Android operating system remains, as ever, definitively open source — but Android’s openness isn’t really what the fuss is about. In this case, the confusion (or obfuscation) stems from the casual confounding of Google Apps with the Android Operating System. As we’ll see, they aren’t the same thing.

Consider Amazon, which pre-loads no Google applications at all on its Kindle Fire and Fire Phone. Amazon’s version of Android uses Microsoft’s Bing as the default search engineNokia provides mapping services, and the app store is Amazon’s own.

Still, Microsoft’s apologists continue to claim that Android licensees can’t choose to opt out of Google’s applications suite — even though, according to a new report from ABI Research, 20 percent of smartphones shipped between May and July 2014 were based on a “Google-less” version of the Android OS. And that number is consistently increasing: Analysts predict that by 2015, 30 percent of Android phones won’t access Google Services.

It’s true that equipment manufacturers who choose the Android operating system have the option to include the suite of integrated, proprietary Google apps and services licensed (royalty-free) under the name Google Mobile Services (GMS). GMS includes Google Search, Maps, Calendar, YouTube and other apps that together define the “Google Android experience” that users know and love.

But Google Android is far from the only Android experience.

Even if a manufacturer chooses to license Google’s apps suite, Google’s terms are not exclusive. Handset makers are free to install competing applications, including other search engines, map applications or app stores.

Although Google requires that Google Search be made easily accessible (hardly a bad thing for consumers, as it is Google Search that finances the development and maintenance of all of the other (free) apps from which Google otherwise earns little to no revenue), OEMs and users alike can (and do) easily install and access other search engines in numerous ways. As Professor Korber notes:

The standard MADA does not entail any exclusivity for Google Search nor does it mandate a search default for the web browser.

Regardless, integrating key Google apps (like Google Search and YouTube) with other apps the company offers (like Gmail and Google+) is an antitrust problem only if it significantly forecloses competitors from these apps’ markets compared to a world without integrated Google apps, and without pro-competitive justification. Neither is true, despite the unsubstantiated claims to the contrary from Edelman, FairSearch and others.

Consumers and developers expect and demand consistency across devices so they know what they’re getting and don’t have to re-learn basic functions or program multiple versions of the same application. Indeed, Apple’s devices are popular in part because Apple’s closed iOS provides a predictable, seamless experience for users and developers.

But making Android competitive with its tightly controlled competitors requires special efforts from Google to maintain a uniform and consistent experience for users. Google has tried to achieve this uniformity by increasingly disentangling its apps from the operating system (the opposite of tying) and giving OEMs the option (but not the requirement) of licensing GMS — a “suite” of technically integrated Google applications (integrated with each other, not the OS).  Devices with these proprietary apps thus ensure that both consumers and developers know what they’re getting.

Unlike Android, Apple prohibits modifications of its operating system by downstream partners and users, and completely controls the pre-installation of apps on iOS devices. It deeply integrates applications into iOS, including Apple Maps, iTunes, Siri, Safari, its App Store and others. Microsoft has copied Apple’s model to a large degree, hard-coding its own applications (including Bing, Windows Store, Skype, Internet Explorer, Bing Maps and Office) into the Windows Phone operating system.

In the service of creating and maintaining a competitive platform, each of these closed OS’s bakes into its operating system significant limitations on which third-party apps can be installed and what they can (and can’t) do. For example, neither platform permits installation of a third-party app store, and neither can be significantly customized. Apple’s iOS also prohibits users from changing default applications — although the soon-to-be released iOS 8 appears to be somewhat more flexible than previous versions.

In addition to pre-installing a raft of their own apps and limiting installation of other apps, both Apple and Microsoft enable greater functionality for their own apps than they do the third-party apps they allow.

For example, Apple doesn’t make available for other browsers (like Google’s Chrome) all the JavaScript functionality that it does for Safari, and it requires other browsers to use iOS Webkit instead of their own web engines. As a result there are things that Chrome can’t do on iOS that Safari and only Safari can do, and Chrome itself is hamstrung in implementing its own software on iOS. This approach has led Mozilla to refuse to offer its popular Firefox browser for iOS devices (while it has no such reluctance about offering it on Android).

On Windows Phone, meanwhile, Bing is integrated into the OS and can’t be removed. Only in markets where Bing is not supported (and with Microsoft’s prior approval) can OEMs change the default search app from Bing. While it was once possible to change the default search engine that opens in Internet Explorer (although never from the hardware search button), the Windows 8.1 Hardware Development Notes, updated July 22, 2014, state:

By default, the only search provider included on the phone is Bing. The search provider used in the browser is always the same as the one launched by the hardware search button.

Both Apple iOS and Windows Phone tightly control the ability to use non-default apps to open intents sent from other apps and, in Windows especially, often these linkages can’t be changed.

As a result of these sorts of policies, maintaining the integrity — and thus the brand — of the platform is (relatively) easy for closed systems. While plenty of browsers are perfectly capable of answering an intent to open a web page, Windows Phone can better ensure a consistent and reliable experience by forcing Internet Explorer to handle the operation.

By comparison, Android, with or without Google Mobile Services, is dramatically more open, more flexible and customizable, and more amenable to third-party competition. Even the APIs that it uses to integrate its apps are open to all developers, ensuring that there is nothing that Google apps are able to do that non-Google apps with the same functionality are prevented from doing.

In other words, not just Gmail, but any email app is permitted to handle requests from any other app to send emails; not just Google Calendar but any calendar app is permitted to handle requests from any other app to accept invitations.

In no small part because of this openness and flexibility, current reports indicate that Android OS runs 85 percent of mobile devices worldwide. But it is OEM giant Samsung, not Google, that dominates the market, with a 65 percent share of all Android devices. Competition is rife, however, especially in emerging markets. In fact, according to one report, “Chinese and Indian vendors accounted for the majority of smartphone shipments for the first time with a 51% share” in 2Q 2014.

As he has not been in the past, Edelman is at least nominally circumspect in his unsubstantiated legal conclusions about Android’s anticompetitive effect:

Applicable antitrust law can be complicated: Some ties yield useful efficiencies, and not all ties reduce welfare.

Given Edelman’s connections to Microsoft and the realities of the market he is discussing, it could hardly be otherwise. If every integration were an antitrust violation, every element of every operating system — including Apple’s iOS as well as every variant of Microsoft’s Windows — should arguably be the subject of a government investigation.

In truth, Google has done nothing more than ensure that its own suite of apps functions on top of Android to maintain what Google sees as seamless interconnectivity, a high-quality experience for users, and consistency for application developers — while still allowing handset manufacturers room to innovate in a way that is impossible on other platforms. This is the very definition of pro-competitive, and ultimately this is what allows the platform as a whole to compete against its far more vertically integrated alternatives.

Which brings us back to Microsoft. On the conclusion of the FTC investigation in January 2013, a GigaOm exposé on the case had this to say:

Critics who say Google is too powerful have nagged the government for years to regulate the company’s search listings. But today the critics came up dry….

The biggest loser is Microsoft, which funded a long-running cloak-and-dagger lobbying campaign to convince the public and government that its arch-enemy had to be regulated….

The FTC is also a loser because it ran a high profile two-year investigation but came up dry.

EU regulators, take note.

The Federal Trade Commission’s recent enforcement actions against Amazon and Apple raise important questions about the FTC’s consumer protection practices, especially its use of economics. How does the Commission weigh the costs and benefits of its enforcement decisions? How does the agency employ economic analysis in digital consumer protection cases generally?

Join the International Center for Law and Economics and TechFreedom on Thursday, July 31 at the Woolly Mammoth Theatre Company for a lunch and panel discussion on these important issues, featuring FTC Commissioner Joshua Wright, Director of the FTC’s Bureau of Economics Martin Gaynor, and several former FTC officials. RSVP here.

Commissioner Wright will present a keynote address discussing his dissent in Apple and his approach to applying economics in consumer protection cases generally.

Geoffrey Manne, Executive Director of ICLE, will briefly discuss his recent paper on the role of economics in the FTC’s consumer protection enforcement. Berin Szoka, TechFreedom President, will moderate a panel discussion featuring:

  • Martin Gaynor, Director, FTC Bureau of Economics
  • David Balto, Fmr. Deputy Assistant Director for Policy & Coordination, FTC Bureau of Competition
  • Howard Beales, Fmr. Director, FTC Bureau of Consumer Protection
  • James Cooper, Fmr. Acting Director & Fmr. Deputy Director, FTC Office of Policy Planning
  • Pauline Ippolito, Fmr. Acting Director & Fmr. Deputy Director, FTC Bureau of Economics

Background

The FTC recently issued a complaint and consent order against Apple, alleging its in-app purchasing design doesn’t meet the Commission’s standards of fairness. The action and resulting settlement drew a forceful dissent from Commissioner Wright, and sparked a discussion among the Commissioners about balancing economic harms and benefits in Section 5 unfairness jurisprudence. More recently, the FTC brought a similar action against Amazon, which is now pending in federal district court because Amazon refused to settle.

Event Info

The “FTC: Technology and Reform” project brings together a unique collection of experts on the law, economics, and technology of competition and consumer protection to consider challenges facing the FTC in general, and especially regarding its regulation of technology. The Project’s initial report, released in December 2013, identified critical questions facing the agency, Congress, and the courts about the FTC’s future, and proposed a framework for addressing them.

The event will be live streamed here beginning at 12:15pm. Join the conversation on Twitter with the #FTCReform hashtag.

When:

Thursday, July 31
11:45 am – 12:15 pm — Lunch and registration
12:15 pm – 2:00 pm — Keynote address, paper presentation & panel discussion

Where:

Woolly Mammoth Theatre Company – Rehearsal Hall
641 D St NW
Washington, DC 20004

Questions? – Email mail@techfreedom.orgRSVP here.

See ICLE’s and TechFreedom’s other work on FTC reform, including:

  • Geoffrey Manne’s Congressional testimony on the the FTC@100
  • Op-ed by Berin Szoka and Geoffrey Manne, “The Second Century of the Federal Trade Commission”
  • Two posts by Geoffrey Manne on the FTC’s Amazon Complaint, here and here.

About The International Center for Law and Economics:

The International Center for Law and Economics is a non-profit, non-partisan research center aimed at fostering rigorous policy analysis and evidence-based regulation.

About TechFreedom:

TechFreedom is a non-profit, non-partisan technology policy think tank. We work to chart a path forward for policymakers towards a bright future where technology enhances freedom, and freedom enhances technology.

The International Center for Law & Economics (ICLE) and TechFreedom filed two joint comments with the FCC today, explaining why the FCC has no sound legal basis for micromanaging the Internet and why “net neutrality” regulation would actually prove counter-productive for consumers.

The Policy Comments are available here, and the Legal Comments are here. See our previous post, Net Neutrality Regulation Is Bad for Consumers and Probably Illegal, for a distillation of many of the key points made in the comments.

New regulation is unnecessary. “An open Internet and the idea that companies can make special deals for faster access are not mutually exclusive,” said Geoffrey Manne, Executive Director of ICLE. “If the Internet really is ‘open,’ shouldn’t all companies be free to experiment with new technologies, business models and partnerships?”

“The media frenzy around this issue assumes that no one, apart from broadband companies, could possibly question the need for more regulation,” said Berin Szoka, President of TechFreedom. “In fact, increased regulation of the Internet will incite endless litigation, which will slow both investment and innovation, thus harming consumers and edge providers.”

Title II would be a disaster. The FCC has proposed re-interpreting the Communications Act to classify broadband ISPs under Title II as common carriers. But reinterpretation might unintentionally ensnare edge providers, weighing them down with onerous regulations. “So-called reclassification risks catching other Internet services in the crossfire,” explained Szoka. “The FCC can’t easily forbear from Title II’s most onerous rules because the agency has set a high bar for justifying forbearance. Rationalizing a changed approach would be legally and politically difficult. The FCC would have to simultaneously find the broadband market competitive enough to forbear, yet fragile enough to require net neutrality rules. It would take years to sort out this mess — essentially hitting the pause button on better broadband.”

Section 706 is not a viable option. In 2010, the FCC claimed Section 706 as an independent grant of authority to regulate any form of “communications” not directly barred by the Act, provided only that the Commission assert that regulation would somehow promote broadband. “This is an absurd interpretation,” said Szoka. “This could allow the FCC to essentially invent a new Communications Act as it goes, regulating not just broadband, but edge companies like Google and Facebook, too, and not just neutrality but copyright, cybersecurity and more. The courts will eventually strike down this theory.”

A better approach. “The best policy would be to maintain the ‘Hands off the Net’ approach that has otherwise prevailed for 20 years,” said Manne. “That means a general presumption that innovative business models and other forms of ‘prioritization’ are legal. Innovation could thrive, and regulators could still keep a watchful eye, intervening only where there is clear evidence of actual harm, not just abstract fears.” “If the FCC thinks it can justify regulating the Internet, it should ask Congress to grant such authority through legislation,” added Szoka. “A new communications act is long overdue anyway. The FCC could also convene a multistakeholder process to produce a code enforceable by the Federal Trade Commission,” he continued, noting that the White House has endorsed such processes for setting Internet policy in general.

Manne concluded: “The FCC should focus on doing what Section 706 actually commands: clearing barriers to broadband deployment. Unleashing more investment and competition, not writing more regulation, is the best way to keep the Internet open, innovative and free.”

For some of our other work on net neutrality, see:

“Understanding Net(flix) Neutrality,” an op-ed by Geoffrey Manne in the Detroit News on Netflix’s strategy to confuse interconnection costs with neutrality issues.

“The Feds Lost on Net Neutrality, But Won Control of the Internet,” an op-ed by Berin Szoka and Geoffrey Manne in Wired.com.

“That startup investors’ letter on net neutrality is a revealing look at what the debate is really about,” a post by Geoffrey Manne in Truth on the Market.

Bipartisan Consensus: Rewrite of ‘96 Telecom Act is Long Overdue,” a post on TF’s blog highlighting the key points from TechFreedom and ICLE’s joint comments on updating the Communications Act.

The Net Neutrality Comments are available here:

ICLE/TF Net Neutrality Policy Comments

TF/ICLE Net Neutrality Legal Comments

With Berin Szoka.

TechFreedom and the International Center for Law & Economics will shortly file two joint comments with the FCC, explaining why the FCC has no sound legal basis for micromanaging the Internet—now called “net neutrality regulation”—and why such regulation would be counter-productive as a policy matter. The following summarizes some of the key points from both sets of comments.

No one’s against an open Internet. The notion that anyone can put up a virtual shingle—and that the good ideas will rise to the top—is a bedrock principle with broad support; it has made the Internet essential to modern life. Key to Internet openness is the freedom to innovate. An open Internet and the idea that companies can make special deals for faster access are not mutually exclusive. If the Internet really is “open,” shouldn’t all companies be free to experiment with new technologies, business models and partnerships? Shouldn’t the FCC allow companies to experiment in building the unknown—and unknowable—Internet of the future?

The best approach would be to maintain the “Hands off the Net” approach that has otherwise prevailed for 20 years. That means a general presumption that innovative business models and other forms of “prioritization” are legal. Innovation could thrive, and regulators could still keep a watchful eye, intervening only where there is clear evidence of actual harm, not just abstract fears. And they should start with existing legal tools—like antitrust and consumer protection laws—before imposing prior restraints on innovation.

But net neutrality regulation hurts more than it helps. Counterintuitively, a blanket rule that ISPs treat data equally could actually harm consumers. Consider the innovative business models ISPs are introducing. T-Mobile’s unRadio lets users listen to all the on-demand music and radio they want without taking a hit against their monthly data plan. Yet so-called consumer advocates insist that’s a bad thing because it favors some content providers over others. In fact, “prioritizing” one service when there is congestion frees up data for subscribers to consume even more content—from whatever source. You know regulation may be out of control when a company is demonized for offering its users a freebie.

Treating each bit of data neutrally ignores the reality of how the Internet is designed, and how consumers use it.  Net neutrality proponents insist that all Internet content must be available to consumers neutrally, whether those consumers (or content providers) want it or not. They also argue against usage-based pricing. Together, these restrictions force all users to bear the costs of access for other users’ requests, regardless of who actually consumes the content, as the FCC itself has recognized:

[P]rohibiting tiered or usage-based pricing and requiring all subscribers to pay the same amount for broadband service, regardless of the performance or usage of the service, would force lighter end users of the network to subsidize heavier end users. It would also foreclose practices that may appropriately align incentives to encourage efficient use of networks.

The rules that net neutrality advocates want would hurt startups as well as consumers. Imagine a new entrant, clamoring for market share. Without the budget for a major advertising blitz, the archetypical “next Netflix” might never get the exposure it needs to thrive. But for a relatively small fee, the startup could sign up to participate in a sponsored data program, with its content featured and its customers’ data usage exempted from their data plans. This common business strategy could mean the difference between success and failure for a startup. Yet it would be prohibited by net neutrality rules banning paid prioritization.

The FCC lacks sound legal authority. The FCC is essentially proposing to do what can only properly be done by Congress: invent a new legal regime for broadband. Each of the options the FCC proposes to justify this—Section 706 of the Telecommunications Act and common carrier classification—is deeply problematic.

First, Section 706 isn’t sustainable. Until 2010, the FCC understood Section 706 as a directive to use its other grants of authority to promote broadband deployment. But in its zeal to regulate net neutrality, the FCC reversed itself in 2010, claiming Section 706 as an independent grant of authority. This would allow the FCC to regulate any form of “communications” in any way not directly barred by the Act — not just broadband but “edge” companies like Google and Facebook. This might mean going beyond neutrality to regulate copyright, cybersecurity and more. The FCC need only assert that regulation would somehow promote broadband.

If Section 706 is a grant of authority, it’s almost certainly a power to deregulate. But even if its power is as broad as the FCC claims, the FCC still hasn’t made the case that, on balance, its proposed regulations would actually do what it asserts: promote broadband. The FCC has stubbornly refused to conduct serious economic analysis on the net effects of its neutrality rules.

And Title II would be a disaster. The FCC has asked whether Title II of the Act, which governs “common carriers” like the old monopoly telephone system, is a workable option. It isn’t.

In the first place, regulations that impose design limitations meant for single-function networks simply aren’t appropriate for the constantly evolving Internet. Moreover, if the FCC re-interprets the Communications Act to classify broadband ISPs as common carriers, it risks catching other Internet services in the cross-fire, inadvertently making them common carriers, too. Surely net neutrality proponents can appreciate the harmful effects of treating Skype as a common carrier.

Forbearance can’t clean up the Title II mess. In theory the FCC could “forbear” from Title II’s most onerous rules, promising not to apply them when it determines there’s enough competition in a market to make the rules unnecessary. But the agency has set a high bar for justifying forbearance.

Most recently, in 2012, the Commission refused to grant Qwest forbearance even in the highly competitive telephony market, disregarding competition from wireless providers, and concluding that a cable-telco “duopoly” is inadequate to protect consumers. It’s unclear how the FCC could justify reaching the opposite conclusion about the broadband market—simultaneously finding it competitive enough to forbear, yet fragile enough to require net neutrality rules. Such contradictions would be difficult to explain, even if the FCC generally gets discretion on changing its approach.

But there is another path forward. If the FCC can really make the case for regulation, it should go to Congress, armed with the kind of independent economic and technical expert studies Commissioner Pai has urged, and ask for new authority. A new Communications Act is long overdue anyway. In the meantime, the FCC could convene the kind of multistakeholder process generally endorsed by the White House to produce a code enforceable by the Federal Trade Commission. A consensus is possible — just not inside the FCC, where the policy questions can’t be separated from the intractable legal questions.

Meanwhile, the FCC should focus on doing what Section 706 actually demands: clearing barriers to broadband deployment and competition. The 2010 National Broadband Plan laid out an ambitious pro-deployment agenda. It’s just too bad the FCC was so obsessed with net neutrality that it didn’t focus on the plan. Unleashing more investment and competition, not writing more regulation, is the best way to keep the Internet open, innovative and free.

[Cross-posted at TechFreedom.]

Today the FTC filed its complaint in federal district court in Washington against Amazon, alleging that the company’s in-app purchasing system permits children to make in-app purchases without parental “informed consent” constituting an “unfair practice” under Section 5 of the FTC Act.

As I noted in my previous post on the case, in bringing this case the Commission is doubling down on the rule it introduced in Apple that effectively converts the balancing of harms and benefits required under Section 5 of the FTC Act to a per se rule that deems certain practices to be unfair regardless of countervailing benefits. Similarly, it is attempting to extend the informed consent standard it created in Apple that essentially maintains that only specific, identified practices (essentially, distinct notification at the time of purchase or opening of purchase window, requiring entry of a password to proceed) are permissible under the Act.

Such a standard is inconsistent with the statute, however. The FTC’s approach forecloses the ability of companies like Amazon to engage in meaningful design decisions and disregards their judgment about which user interface designs will, on balance, benefit consumers. The FTC Act does not empower the Commission to disregard the consumer benefits of practices that simply fail to mimic the FTC’s preconceived design preferences. While that sort of approach might be defensible in the face of manifestly harmful practices like cramming, it is wholly inappropriate in the context of app stores like Amazon’s that spend considerable resources to design every aspect of their interaction with consumers—and that seek to attract, not to defraud, consumers.

Today’s complaint occasions a few more observations:

  1. Amazon has a very strong case. Under Section 5 of the FTC Act, the Commission will have to prevail on all three elements required to prove unfairness under Section 5: that there is substantial injury, that consumers can’t reasonably avoid the injury and that any countervailing benefits don’t outweigh the injury. But, consistent with its complaint and consent order in Apple, the Amazon complaint focuses almost entirely on only the first of these. While that may have been enough to induce Apple to settle out of court, the FTC will actually have to make out a case on reasonable avoidance and countervailing benefits at trial. It’s not at all clear that the agency will be able to do so on the facts alleged here.
  2. On reasonable avoidance, over and above Amazon’s general procedures that limit unwanted in-app purchases, the FTC will have a tough time showing that Amazon’s Kindle Free Time doesn’t provide parents with more than enough ability to avoid injury. In fact, the complaint doesn’t mention Free Time at all.
  3. Among other things, the complaint asserts that Amazon knew about issues with in-app purchasing by December of 2011 and claims that “[n]ot until June 2014 did Amazon change its in-app charge framework to obtain account holders’ informed consent for in-app charges on its newer mobile devices.” But Kindle Free Time was introduced in September of 2012. While four FTC Commissioners may believe that Free Time isn’t a sufficient response to the alleged problem, it is clearly a readily available, free and effective (read: reasonable) mechanism for parents to avoid the alleged harms. It may not be what the design mavens at the FTC would have chosen to do, but it seems certain that avoiding unauthorized in-app purchases by children was part of what motivated Amazon’s decision to create and offer Free Time.
  4. On countervailing benefits, as Commissioner Wright discussed in detail in his dissent from the Apple consent order, the Commission seems to think that it can simply assert that there are no countervailing benefits to Amazon’s design choices around in-app purchases. Here the complaint doesn’t mention 1-Click at all, which is core to Amazon’s user interface design and essential to evaluating the balance of harms and benefits required by the statute.
  5. Even if it can show that Amazon’s in-app purchase practices caused harm, the Commission will still have to demonstrate that Amazon’s conscious efforts to minimize the steps required to make purchases doesn’t benefit consumers on balance. In Apple, the FTC majority essentially (and improperly) valued these sorts of user-interface benefits at zero. It implicitly does so again here, but a court will require more than such an assertion.
  6. Given these lapses, there is even a chance that the complaint will be thrown out on a motion to dismiss. It’s a high bar, but if the court agrees that there are insufficient facts in the complaint to make out a plausible case on all three elements, Amazon could well prevail on a motion to dismiss. The FTC’s approach in the Apple consent order effectively maintains that the agency can disregard reasonable avoidance and countervailing benefits in contravention of the statute. By following the same approach here in actual litigation, the FTC may well meet resistance from the courts, which have not yet so cavalierly dispensed with the statute’s requirements.

The Wall Street Journal reports this morning that Amazon is getting — and fighting — the “Apple treatment” from the FTC for its design of its in-app purchases:

Amazon.com Inc. is bucking a request from the Federal Trade Commission that it tighten its policies for purchases made by children while using mobile applications.

In a letter to the FTC Tuesday, Amazon said it was prepared to “defend our approach in court,” rather than agree to fines and additional record keeping and disclosure requirements over the next 20 years, according to documents reviewed by The Wall Street Journal.

According to the documents, Amazon is facing a potential lawsuit by the FTC, which wants the Seattle retailer to accept terms similar to those that Apple Inc. agreed to earlier this year regarding so-called in-app purchases.

From what I can tell, the Commission has voted to issue a complaint, and Amazon has informed the Commission that it will not accept its proposed settlement.

I am thrilled that Amazon seems to have decided to fight the latest effort by a majority of the FTC to bring every large tech company under 20-year consent decree. I should say: I’m disappointed in the FTC, sorry for Amazon, but thrilled for consumers and the free marketplace that Amazon is choosing to fight rather than acquiesce.

As I wrote earlier this year about the FTC’s case against Apple in testimony before the House Commerce Committee:

What’s particularly notable about the Apple case – and presumably will be in future technology enforcement actions predicated on unfairness – is the unique relevance of the attributes of the conduct at issue to its product. Unlike past, allegedly similar, cases, Apple’s conduct was not aimed at deceiving consumers, nor was it incidental to its product offering. But by challenging the practice, particularly without the balancing of harms required by Section 5, the FTC majority failed to act with restraint and substituted its own judgment, not about some manifestly despicable conduct, but about the very design of Apple’s products. This is the sort of area where regulatory humility is more — not less — important.

In failing to observe common sense limits in Apple, the FTC set a dangerous precedent that, given the agency’s enormous regulatory scope and the nature of technologically advanced products, could cause significant harm to consumers.

Here that failure is even more egregious. Amazon has built its entire business around the “1-click” concept — which consumers love — and implemented a host of notification and security processes hewing as much as possible to that design choice, but nevertheless taking account of the sorts of issues raised by in-app purchases. Moreover — and perhaps most significantly — it has implemented an innovative and comprehensive parental control regime (including the ability to turn off all in-app purchases) — Kindle Free Time — that arguably goes well beyond anything the FTC required in its Apple consent order. I use Kindle Free Time with my kids and have repeatedly claimed to anyone who will listen that it is the greatest thing since sliced bread. Other consumers must feel similarly. Finally, regardless of all of that, Amazon has nevertheless voluntarily implemented additional notification procedures intended to comply with the Apple settlement, even though it didn’t apply to Amazon.

If the FTC asserts, in the face of all of that, that it’s own vision of what “appropriate” in-app purchase protections must look like is the only one that suffices to meet the standard required by Section 5’s Unfairness language, it is either being egregiously disingenuous, horrifically vain, just plain obtuse, or some combination of the three.

As I wrote in my testimony:

The application of Section 5’s “unfair acts and practices” prong (the statute at issue in Apple) is circumscribed by Section 45(n) of the FTC Act, which, among other things, proscribes enforcement where injury is “not outweighed by countervailing benefits to consumers or to competition.”

And as Commissioner Wright noted in his dissent in the Apple case,

[T]he Commission effectively rejects an analysis of tradeoffs between the benefits of additional guidance and potential harm to some consumers or to competition from mandating guidance…. I respectfully disagree. These assumptions adopt too cramped a view of consumer benefits under the Unfairness Statement and, without more rigorous analysis to justify their application, are insufficient to establish the Commission’s burden.

We won’t know until we see the complaint whether the FTC has failed to undertake the balancing it neglected to perform in Apple and that it is required to perform under the statute. But it’s hard to believe that it could mount a case against Amazon in light of the facts if it did perform such a balancing. There’s no question that Amazon has implemented conscious and consumer-welfare-enhancing design choices here. The FTC’s effort to nevertheless mandate a different design (and put Amazon under a 20 year consent decree) based on a claim that Amazon’s choices impose greater harms than benefits on consumers seems manifestly unsupportable.

Such a claim almost certainly represents an abuse of the agency’s discretion, and I expect Amazon to trounce the FTC if this case goes to trial.

UPDATE: I’ve been reliably informed that Vint Cerf coined the term “permissionless innovation,” and, thus, that he did so with the sorts of private impediments discussed below in mind rather than government regulation. So consider the title of this post changed to “Permissionless innovation SHOULD not mean ‘no contracts required,'” and I’ll happily accept that my version is the “bastardized” version of the term. Which just means that the original conception was wrong and thank god for disruptive innovation in policy memes!

Can we dispense with the bastardization of the “permissionless innovation” concept (best developed by Adam Thierer) to mean “no contracts required”? I’ve been seeing this more and more, but it’s been around for a while. Some examples from among the innumerable ones out there:

Vint Cerf on net neutrality in 2009:

We believe that the vast numbers of innovative Internet applications over the last decade are a direct consequence of an open and freely accessible Internet. Many now-successful companies have deployed their services on the Internet without the need to negotiate special arrangements with Internet Service Providers, and it’s crucial that future innovators have the same opportunity. We are advocates for “permissionless innovation” that does not impede entrepreneurial enterprise.

Net neutrality is replete with this sort of idea — that any impediment to edge providers (not networks, of course) doing whatever they want to do at a zero price is a threat to innovation.

Chet Kanojia (Aereo CEO) following the Aereo decision:

It is troubling that the Court states in its decision that, ‘to the extent commercial actors or other interested entities may be concerned with the relationship between the development and use of such technologies and the Copyright Act, they are of course free to seek action from Congress.’ (Majority, page 17)That begs the question: Are we moving towards a permission-based system for technology innovation?

At least he puts it in the context of the Court’s suggestion that Congress pass a law, but what he really wants is to not have to ask “permission” of content providers to use their content.

Mike Masnick on copyright in 2010:

But, of course, the problem with all of this is that it goes back to creating permission culture, rather than a culture where people freely create. You won’t be able to use these popular or useful tools to build on the works of others — which, contrary to the claims of today’s copyright defenders, is a key component in almost all creativity you see out there — without first getting permission.

Fair use is, by definition, supposed to be “permissionless.” But the concept is hardly limited to fair use, is used to justify unlimited expansion of fair use, and is extended by advocates to nearly all of copyright (see, e.g., Mike Masnick again), which otherwise requires those pernicious licenses (i.e., permission) from others.

The point is, when we talk about permissionless innovation for Tesla, Uber, Airbnb, commercial drones, online data and the like, we’re talking (or should be) about ex ante government restrictions on these things — the “permission” at issue is permission from the government, it’s the “permission” required to get around regulatory roadblocks imposed via rent-seeking and baseless paternalism. As Gordon Crovitz writes, quoting Thierer:

“The central fault line in technology policy debates today can be thought of as ‘the permission question,'” Mr. Thierer writes. “Must the creators of new technologies seek the blessing of public officials before they develop and deploy their innovations?”

But it isn’t (or shouldn’t be) about private contracts.

Just about all human (commercial) activity requires interaction with others, and that means contracts and licenses. You don’t see anyone complaining about the “permission” required to rent space from a landlord. But that some form of “permission” may be required to use someone else’s creative works or other property (including broadband networks) is no different. And, in fact, it is these sorts of contracts (and, yes, the revenue that may come with them) that facilitates people engaging with other commercial actors to produce things of value in the first place. The same can’t be said of government permission.

Don’t get me wrong – there may be some net welfare-enhancing regulatory limits that might require forms of government permission. But the real concern is the pervasive abuse of these limits, imposed without anything approaching a rigorous welfare determination. There might even be instances where private permission, imposed, say, by a true monopolist, might be problematic.

But this idea that any contractual obligation amounts to a problematic impediment to innovation is absurd, and, in fact, precisely backward. Which is why net neutrality is so misguided. Instead of identifying actual, problematic impediments to innovation, it simply assumes that networks threaten edge innovation, without any corresponding benefit and with such certainty (although no actual evidence) that ex ante common carrier regulations are required.

“Permissionless innovation” is a great phrase and, well developed (as Adam Thierer has done), a useful concept. But its bastardization to justify interference with private contracts is unsupported and pernicious.

Earlier this week the New Jersey Assembly unanimously passed a bill to allow direct sales of Tesla cars in New Jersey. (H/T Marina Lao). The bill

Allows a manufacturer (“franchisor,” as defined in P.L.1985, c.361 (C.56:10-26 et seq.)) to directly buy from or sell to consumers a zero emission vehicle (ZEV) at a maximum of four locations in New Jersey.  In addition, the bill requires a manufacturer to own or operate at least one retail facility in New Jersey for the servicing of its vehicles. The manufacturer’s direct sale locations are not required to also serve as a retail service facility.

The bill amends current law to allow any ZEV manufacturer to directly or indirectly buy from and directly sell, offer to sell, or deal to a consumer a ZEV if the manufacturer was licensed by the New Jersey Motor Vehicle Commission (MVC) on or prior to January 1, 2014.  This bill provides that ZEVs may be directly sold by certain manufacturers, like Tesla Motors, and preempts any rule or regulation that restricts sales exclusively to franchised dealerships.  The provisions of the bill would not prevent a licensed franchisor from operating under an existing license issued by the MVC.

At first cut, it seems good that the legislature is responding to the lunacy of the Christie administration’s previous decision to enforce a rule prohibiting direct sales of automobiles in New Jersey. We have previously discussed that decision at length in previous posts here, here, here and here. And Thom and Mike have taken on a similar rule in their home state of Missouri here and here.

In response to New Jersey’s decision to prohibit direct sales, the International Center for Law & Economics organized an open letter to Governor Christie based in large part on Dan Crane’s writings on the topic here at TOTM and discussing the faulty economics of such a ban. The letter was signed by more than 70 law professors and economists.

But it turns out that the legislative response is nearly as bad as the underlying ban itself.

First, a quick recap.

In our letter we noted that

The Motor Vehicle Commission’s regulation was aimed specifically at stopping one company, Tesla Motors, from directly distributing its electric cars. But the regulation would apply equally to any other innovative manufacturer trying to bring a new automobile to market, as well. There is no justification on any rational economic or public policy grounds for such a restraint of commerce. Rather, the upshot of the regulation is to reduce competition in New Jersey’s automobile market for the benefit of its auto dealers and to the detriment of its consumers. It is protectionism for auto dealers, pure and simple.

While enforcement of the New Jersey ban was clearly aimed directly at Tesla, it has broader effects. And, of course, its underlying logic is simply indefensible, regardless of which particular manufacturer it affects. The letter explains at length the economics of retail distribution and the misguided, anti-consumer logic of the regulation, and concludes by noting that

In sum, we have not heard a single argument for a direct distribution ban that makes any sense. To the contrary, these arguments simply bolster our belief that the regulations in question are motivated by economic protectionism that favors dealers at the expense of consumers and innovative technologies. It is discouraging to see this ban being used to block a company that is bringing dynamic and environmentally friendly products to market. We strongly encourage you to repeal it, by new legislation if necessary.

Thus it seems heartening that the legislature did, indeed, take up our challenge to repeal the ban.

Except that, in doing so, the legislature managed to write a bill that reflects no understanding whatever of the underlying economic issues at stake. Instead, the legislative response appears largely to be the product of rent seeking,pure and simple, offering only a limited response to Tesla’s squeaky wheel (no pun intended) and leaving the core defects of the ban completely undisturbed.

Instead of acknowledging the underlying absurdity of the limit on direct sales, the bill keeps the ban in place and simply offers a limited exception for Tesla (or other zero emission cars). While the innovative and beneficial nature of Tesla’s cars was an additional reason to oppose banning their direct sale, the specific characteristics of the cars is a minor and ancillary reason to oppose the ban. But the New Jersey legislative response is all about the cars’ emissions characteristics, and in no way does it reflect an appreciation for the fundamental economic defects of the underlying rule.

Moreover, the bill permits direct sales at only four locations (why four? No good reason whatever — presumably it was a political compromise, never the stuff of economic reason) and requires Tesla to operate a service center for its cars in the state. In other words, the regulators are still arbitrarily dictating aspects of car manufacturers’ business organization from on high.

Even worse, however, the bill is constructed to be nothing more than a payoff for a specific firm’s lobbying efforts, thus ensuring that the next (non-zero-emission) Tesla to come along will have to undertake the same efforts to pander to the state.

Far from addressing the serious concerns with the direct sales ban, the bill just perpetuates the culture of political rent seeking such regulations create.

Perhaps it’s better than nothing. Certainly it’s better than nothing for Tesla. But overall, I’d say it’s about the worst possible sort of response, short of nothing.

Credit where it’s due — the FTC has closed its investigation of the Men’s Warehouse/Jos. A. Bank merger. I previously wrote about the investigation here, where I said:

I would indeed be shocked if a legitimate economic analysis suggested that Jos. A. Banks and Men’s Warehouse occupied all or most of any relevant market. For the most part — and certainly for the marginal consumer — there is no meaningful difference between a basic, grey worsted wool suit bought at a big department store in the mall and a similar suit bought at a small retailer in the same mall or a “warehouse” store across the street. And the barriers to entry in such a market, if it existed, would be insignificant. Again, what I said of Whole Foods/Wild Oats is surely true here, too:

But because economically-relevant market definition turns on demand elasticity among consumers who are often free to purchase products from multiple distribution channels, a myopic focus on a single channel of distribution to the exclusion of others is dangerous.

Let’s hope the FTC gets it right this time.

The FTC’s blog post on closing the investigation notes that:

Despite limited competition from the Internet, the transaction is not likely to harm consumers because of significant competition from other sources. As in all transactions, FTC staff examined which product markets were likely to be affected and what the competitive landscape looks like in those markets. There were two such markets in this matter: (1) the retail sale of men’s suits and (2) tuxedo rentals. With respect to men’s suits, there are numerous competitors that sell suits across the range of prices of the suits the merging parties offer, including Macy’s, Kohl’s, JC Penney’s, Nordstrom, and Brooks Brothers, among others. The two firms also have different product assortments that reflect their different customer bases. Men’s Wearhouse, which sells branded and private-label suits, has a younger, trendier customer set, while Jos. A. Bank, which sells private-label suits only, has an older, more traditional customer base.

Sounds right — and good to see.

Last week a group of startup investors wrote a letter to protest what they assume FCC Chairman Tom Wheeler’s proposed, revised Open Internet NPRM will say.

Bear in mind that an NPRM is a proposal, not a final rule, and its issuance starts a public comment period. Bear in mind, as well, that the proposal isn’t public yet, presumably none of the signatories to this letter has seen it, and the devil is usually in the details. That said, the letter has been getting a lot of press.

I found the letter seriously wanting, and seriously disappointing. But it’s a perfect example of what’s so wrong with this interminable debate on net neutrality.

Below I reproduce the letter in full, in quotes, with my comments interspersed. The key take-away: Neutrality (or non-discrimination) isn’t what’s at stake here. What’s at stake is zero-cost access by content providers to broadband networks. One can surely understand why content providers and those who fund them want their costs of doing business to be lower. But the rhetoric of net neutrality is mismatched with this goal. It’s no wonder they don’t just come out and say it – it’s quite a remarkable claim.

Open Internet Investors Letter

The Honorable Tom Wheeler, Chairman
Federal Communications Commission
445 12th Street, SW
Washington D.C. 20554

May 8, 2014

Dear Chairman Wheeler:

We write to express our support for a free and open Internet.

We invest in entrepreneurs, investing our own funds and those of our investors (who are individuals, pension funds, endowments, and financial institutions).  We often invest at the earliest stages, when companies include just a handful of founders with largely unproven ideas. But, without lawyers, large teams or major revenues, these small startups have had the opportunity to experiment, adapt, and grow, thanks to equal access to the global market.

“Equal” access has nothing to do with it. No startup is inherently benefitted by being “equal” to others. Maybe this is just careless drafting. But frankly, as I’ll discuss, there are good reasons to think (contra the pro-net neutrality narrative) that startups will be helped by inequality (just like contra the (totally wrong) accepted narrative, payola helps new artists). It says more than they would like about what these investors really want that they advocate “equality” despite the harm it may impose on startups (more on this later).

Presumably what “equal” really means here is “zero cost”: “As long as my startup pays nothing for access to ISPs’ subscribers, it’s fine that we’re all on equal footing.” Wheeler has stated his intent that his proposal would require any prioritization to be available to any who want it, on equivalent, commercially reasonable terms. That’s “equal,” too, so what’s to complain about? But it isn’t really inequality that’s gotten everyone so upset.

Of course, access is never really “zero cost;” start-ups wouldn’t need investors if their costs were zero. In that sense, why is equality of ISP access any more important than other forms of potential equality? Why not mandate price controls on rent? Why not mandate equal rent? A cost is a cost. What’s really going on here is that, like Netflix, these investors want to lower their costs and raise their returns as much as possible, and they want the government to do it for them.

As a result, some of the startups we have invested in have managed to become among the most admired, successful, and influential companies in the world.

No startup became successful as a result of “equality” or even zero-cost access to broadband. No doubt some of their business models were predicated on that assumption. But it didn’t cause their success.

We have made our investment decisions based on the certainty of a level playing field and of assurances against discrimination and access fees from Internet access providers.

And they would make investment decisions based on the possibility of an un-level playing field if that were the status quo. More importantly, the businesses vying for investment dollars might be different ones if they built their business models in a different legal/economic environment. So what? This says nothing about the amount of investment, the types of businesses, the quality of businesses that would arise under a different set of rules. It says only that past specific investments might not have been made.

Unless the contention is that businesses would be systematically worse under a different rule, this is irrelevant. I have seen that claim made, and it’s implicit here, of course, but I’ve seen no evidence to actually support it. Businesses thrive in unequal, cost-ladened environments all the time. It costs about $4 million/30 seconds to advertise during the Super Bowl. Budweiser and PepsiCo paid multiple millions this year to do so; many of their competitors didn’t. With inequality like that, it’s a wonder Sierra Nevada and Dr. Pepper haven’t gone bankrupt.

Indeed, our investment decisions in Internet companies are dependent upon the certainty of an equal-opportunity marketplace.

Again, no they’re not. Equal opportunity is a euphemism for zero cost, or else this is simply absurd on its face. Are these investors so lacking in creativity and ability that they can invest only when there is certainty of equal opportunity? Don’t investors thrive – aren’t they most needed – in environments where arbitrage is possible, where a creative entrepreneur can come up with a risky, novel way to take advantage of differential conditions better than his competitors? Moreover, the implicit equating of “equal-opportunity marketplace” with net neutrality rules is far-fetched. Is that really all that matters?

This is a good time to make a point that is so often missed: The loudest voices for net neutrality are the biggest companies – Google, Netflix, Amazon, etc. That fact should give these investors and everyone else serious pause. Their claim rests on the idea that “equality” is needed, so big companies can’t use an Internet “fast lane” to squash them. Google is decidedly a big company. So why do the big boys want this so much?

The battle is often pitched as one of ISPs vs. (small) content providers. But content providers have far less to worry about and face far less competition from broadband providers than from big, incumbent competitors. It is often claimed that “Netflix was able to pay Comcast’s toll, but a small startup won’t have that luxury.” But Comcast won’t even notice or care about a small startup; its traffic demands will be inconsequential. Netflix can afford to pay for Internet access for precisely the same reason it came to Comcast’s attention: It’s hugely successful, and thus creates a huge amount of traffic.

Based on news reports and your own statements, we are worried that your proposed rules will not provide the necessary certainty that we need to make investment decisions and that these rules will stifle innovation in the Internet sector.

Now, there’s little doubt that legal certainty aids investment decisions. But “certainty” is not in danger here. The rules have to change because the court said so – with pretty clear certainty. And a new rule is not inherently any more or less likely to offer certainty than the previous Open Internet Order, which itself was subject to intense litigation (obviously) and would have been subject to interpretation and inconsistent enforcement (and would have allowed all kinds of paid prioritization, too!). Certainty would be good, but Wheeler’s proposed rule won’t likely do anything about the amount of certainty one way or the other.

If established companies are able to pay for better access speeds or lower latency, the Internet will no longer be a level playing field. Start-ups with applications that are advantaged by speed (such as games, video, or payment systems) will be unlikely to overcome that deficit no matter how innovative their service.

Again, it’s notable that some of the strongest advocates for net neutrality are established companies. Another letter sent out last week included signatures from a bunch of startups, but also Google, Microsoft, Facebook and Yahoo!, among others.

In truth it’s hard to see why startup investors would think this helps them. Non-neutrality offers the prospect that a startup might be able to buy priority access to overcome the inherent disadvantage of newness, and to better compete with an established company. Neutrality means that that competitive advantage is impossible, and the baseline relative advantages and disadvantages remain – which helps incumbents, not startups. With a neutral Internet – well, the advantages of the incumbent competitor can’t be dissipated by a startup buying a favorable leg-up in speed and the Netflix’s of the world will be more likely to continue to dominate.

Of course the claim is that incumbents will use their huge resources to gain even more advantage with prioritized access. Implicit in this must be the assumption that the advantage that could be gained by a startup buying priority offers less return for the startup than the cost imposed on it by the inherent disadvantages of reputation, brand awareness, customer base, etc. But that’s not plausible for all or even most startups. And investors exist precisely because they are able to provide funds for which there is a likelihood of a good return – so if paying for priority would help overcome inherent disadvantages, there would be money for it.

Also implicit is the claim that the benefits to incumbents (over and above their natural advantages) from paying for priority, in terms of hamstringing new entrants, will outweigh the cost. This is unlikely generally to be true, as well. They already have advantages. Sure, sometimes they might want to pay for more, but in precisely the cases where it would be worth it to do so, the new entrant would also be most benefited by doing so itself – ensuring, again, that investment funds will be available.

Of course if both incumbents and startups decide paying for priority is better, we’re back to a world of “equality,” so what’s to complain about, based on this letter? This puts into stark relief that what these investors really want is government-mandated, subsidized broadband access, not “equality.”

Now, it’s conceivable that that is the optimal state of affairs, but if it is, it isn’t for the reasons given here, nor has anyone actually demonstrated that it is the case.

Entrepreneurs will need to raise money to buy fast lane services before they have proven that consumers want their product. Investors will extract more equity from entrepreneurs to compensate for the risk.

Internet applications will not be able to afford to create a relationship with millions of consumers by making their service freely available and then build a business over time as they better understand the value consumers find in their service (which is what Facebook, Twitter, Tumblr, Pinterest, Reddit, Dropbox and virtually other consumer Internet service did to achieve scale).

In other words: “Subsidize us. We’re worth it.” Maybe. But this is probably more revealing than intended. The Internet cost something to someone to build. (Actually, it cost more than a trillion dollars to broadband providers). This just says “we shouldn’t have to pay them for it now.” Fine, but who, then, and how do you know that forcing someone else to subsidize these startup companies will actually lead to better results? Mightn’t we get less broadband investment such that there is little Internet available for these companies to take advantage of in the first place? If broadband consumers instead of content consumers foot the bill, is that clearly preferable, either from a social welfare perspective, or even the self interest of these investors who, after all, do ultimately rely on consumer spending to earn their return?

Moreover, why is this “build for free, then learn how to monetize over time” business model necessarily better than any other? These startup investors know better than anyone that enshrining existing business models just because they exist is the antithesis of innovation and progress. But that’s exactly what they’re saying – “the successful companies of the past did it this way, so we should get a government guarantee to preserve our ability to do it, too!”

This is the most depressing implication of this letter. These investors and others like them have been responsible for financing enormously valuable innovations. If even they can’t see the hypocrisy of these claims for net neutrality – and worse, choose to propagate it further – then we really have come to a sad place. When innovators argue passionately for stagnation, we’re in trouble.

Instead, creators will have to ask permission of an investor or corporate hierarchy before they can launch. Ideas will be vetted by committees and quirky passion projects will not get a chance. An individual in dorm room or a design studio will not be able to experiment out loud on the Internet. The result will be greater conformity, fewer surprises, and less innovation.

This is just a little too much protest. Creators already have to ask “permission” – or are these investors just opening up their bank accounts to whomever wants their money? The ones that are able to do it on a shoestring, with money saved up from babysitting gigs, may find higher costs, and the need to do more babysitting. But again, there is nothing special about the Internet in this. Let’s mandate zero cost office space and office supplies and developer services and design services and . . . etc. for all – then we’ll have way more “permission-less” startups. If it’s just a handout they want, they should say so, instead of pretending there is a moral or economic welfare basis for their claims.

Further, investors like us will be wary of investing in anything that access providers might consider part of their future product plans for fear they will use the same technical infrastructure to advantage their own services or use network management as an excuse to disadvantage competitive offerings.

This is crazy. For the same reasons I mentioned above, the big access provider (and big incumbent competitor, for that matter) already has huge advantages. If these investors aren’t already wary of investing in anything that Google or Comcast or Apple or… might plan to compete with, they must be terrible at their jobs.

What’s more, Wheeler’s much-reviled proposal (what we know about it, that is), to say nothing of antitrust law, clearly contemplates exactly this sort of foreclosure and addresses it. “Pure” net neutrality doesn’t add much, if anything, to the limits those laws already do or would provide.

Policing this will be almost impossible (even using a standard of “commercial reasonableness”) and access providers do not need to successfully disadvantage their competition; they just need to create a credible threat so that investors like us will be less inclined to back those companies.

You think policing the world of non-neutrality is hard – try policing neutrality. It’s not as easy as proponents make it out to be. It’s simply never been the case that all bits at all times have been treated “neutrally” on the Internet. Any version of an Open Internet Order (just like the last one, for example) will have to recognize this.

Larry Downes compiled a list of the exceptions included in the last Open Internet Order when he testified before the House Judiciary Committee on the rules in 2011. There are 16 categories of exemption, covering a wide range of fundamental components of broadband connectivity, from CDNs to free Wi-Fi at Starbucks. His testimony is a tour de force, and should be required reading for everyone involved in this debate.

But think about how the manifest advantages of these non-neutral aspects of broadband networks would be squared with “real” neutrality. On their face, if these investors are to be taken at their word, these arguments would preclude all of the Open Internet Order’s exemptions, too. And if any sort of inequality is going to be deemed ok, how accurately would regulators distinguish between “illegitimate” inequality and the acceptable kind that lets coffee shops subsidize broadband? How does the simplistic logic of net equality distinguish between, say, Netflix’s colocated servers and a startup like Uber being integrated into Google Maps? The simple answer is that it doesn’t, and the claims and arguments of this letter are woefully inadequate to the task.

We need simple, strong, enforceable rules against discrimination and access fees, not merely against blocking.

No, we don’t. Or, at least, no one has made that case. These investors want a handout; that is the only case this letter makes.

We encourage the Commission to consider all available jurisdictional tools at its disposal in ensuring a free and open Internet that rewards, not disadvantages, investment and entrepreneurship.

… But not investment in broadband, and not entrepreneurship that breaks with the business models of the past. In reality, this letter is simple rent-seeking: “We want to invest in what we know, in what’s been done before, and we don’t want you to do anything to make that any more costly for us. If that entails impairing broadband investment or imposing costs on others, so be it – we’ll still make our outsized returns, and they can write their own letter complaining about ‘inequality.’”

A final point I have to make. Although the investors don’t come right out and say it, many others have, and it’s implicit in the investors’ letter: “Content providers shouldn’t have to pay for broadband. Users already pay for the service, so making content providers pay would just let ISPs double dip.” The claim is deeply problematic.

For starters, it’s another form of the status quo mentality: “Users have always paid and content hasn’t, so we object to any deviation from that.” But it needn’t be that way. And of course models frequently coexist where different parties pay for the same or similar services. Some periodicals are paid for by readers and offer little or no advertising; others charge a subscription and offer paid ads; and still others are offered for free, funded entirely by ads. All of these models work. None is “better” than the other. There is no reason the same isn’t true for broadband and content.

Net neutrality claims that the only proper price to charge on the content side of the market is zero. (Congratulations: You’re in the same club as that cutting-edge, innovative technology, the check, which is cleared at par by government fiat. A subsidy that no doubt explains why checks have managed to last this long). As an economic matter, that’s possible; it could be that zero is the right price. But it most certainly needn’t be, and issues revolving around Netflix’s traffic and the ability of ISPs and Netflix cost-effectively to handle it are evidence that zero may well not be the right price.

The reality is that these sorts of claims are devoid of economic logic — which is presumably why they, like the whole net neutrality “movement” generally, appeal so gratuitously to emotion rather than reason. But it doesn’t seem unreasonable to hope for more from a bunch of savvy financiers.