Note: This post was previously titled “Why Ajit Pai is Right.” I have changed it to reflect my interest in a substantive debate, not flame-throwing. The article is unchanged (beyond normal edits).

Weirdly, this article about the American broadband market must start in Portugal.

Last week Federal Communications Commission (FCC) Chairman Ajit Pai circulated a draft order that would undo the 2015 reclassification of Internet Service Providers (ISPs) from what are known as “Title I information services” to “Title II telecommunication providers”; Title II of the Telecommunications Act, originally developed to regulate the AT&T monopoly, gives the FCC broad ability to regulate “common carriers” as utilities. Title I, on the other hand, hands off regulatory oversight to the Federal Trade Commission (FTC).

The net effect of this reclassification would be the elimination of FCC rules restricting the ability of ISPs to block or throttle sites or apps or offer paid prioritization of any Internet content. That is certainly a worthy goal! Who could possibly be in favor of ISPs picking-and-choosing what sites you can visit based on what you are willing to pay? Do we really want to be like Portugal?

There’s just one problem with the tweet I embedded above: Portugal uses Euros, and the language is Portuguese; the tweet above has dollars and English. The image is completely made-up.

Congressman Ro Khanna, who represents Silicon Valley, at least went to the trouble of getting an actual page from a Portuguese carrier:

There are the Euros and Portuguese you would expect, but in this case perhaps it was the language difference that introduced its own issues: Congressman Khanna seems to have missed the text at the top (under ‘+ Smart Net’) that clearly stated that the packages were for an additional 10GB/month of data; in addition to what, you may ask? Simply scroll down the page:

So to recount: one Portugal story is made up, and the other declared that a 10GB family plan with an extra 10GB for a collection of apps of your choosing for €25/month ($30/month) is a future to be feared; given that AT&T charges $65 for a single “Unlimited” plan that downscales video, bans tethering, and slows speeds after 22GB, one wonders if most Americans share that fear.

That, though, is the magic of the term “net neutrality”, the name — coined by the same Tim Wu whose tweet I embedded above — for those FCC rules that justified the original 2015 reclassification of ISPs to utility-like common carriers. Of course ISPs should be neutral — again, who could be against such a thing? What is missing in the ongoing debate, though, is the recognition that, ever since the demise of AOL, they have been. The FCC’s 2015 approach to net neutrality is solving problems as fake as the image in Wu’s tweet; unfortunately the costs are just as real as those in Congressman Khanna’s tweet, but massively more expensive.

The Cost of Regulation

Allow me to state this point plainly: I am absolutely in favor of net neutrality. Indeed, as I explained in 2014’s Netflix and Net Neutrality, I am willing to make trade-offs (specifically data caps) to achieve it. The question at hand, though, is what is the best way to achieve net neutrality? To believe that Chairman Pai is right is not to be against net neutrality; rather, it is to believe that the FCC’s 2015 approach was mistaken.

Any regulatory decision — indeed, any decision period — is about tradeoffs. To choose one course of action is to gain certain benefits and incur certain costs, and it is to forgo the benefits (and costs!) of alternative courses of action. What makes evaluating regulations so difficult is that the benefits are usually readily apparent — the bad behavior or outcome is, hopefully, eliminated — but the costs are much more difficult to quantify. Short-term implementation costs may be relatively straightforward, but future innovations and market entries that don’t happen by virtue of the regulation being in place are far more difficult to calculate. Equally difficult to measure is the inevitable rent-seeking that accompanies regulation, as incumbents find it easier to lobby regulators to foreclose competition instead of winning customers in an open market.

A classic example of this phenomenon is restaurants: who could possibly be against food safety? Then you read about how San Francisco requires 14 permits that take 9 months to issue (plus a separate alcohol permit) and you wonder why anyone opens a restaurant at all (compounded by the fact that already-permitted restaurants have a vested interest in making the process more onerous over time). Multiply that burden by all of the restaurants that never get created and the cost is very large indeed.

This argument certainly applies to net neutrality in a far more profound way: the Internet has been the single most important driver of not just economic growth but overall consumer welfare for the last two decades. Given that all of that dynamism has been achieved with minimal regulatory oversight, the default position of anyone concerned about future growth should be maintaining a light touch. After all, regulation always has a cost far greater than what we can see at the moment it is enacted, and given the importance of the Internet, those costs are massively more consequential than restaurants or just about anything else.

To put it another way, given the stakes, the benefit from regulation must be massive, which is why the “net neutrality” framing is so powerful: I’ll say it again — who can be against net neutrality? Telling stories about speech being restricted or new companies being unable to pay to access customers tap into both the Internet’s clear impact and the foregone opportunity cost I just described — businesses that are never built.

That, though, is exactly the problem: opportunity costs are a reason to not regulate; clear evidence of harm are the reasons to do so despite the costs. What is so backwards about this entire debate is that those in favor of regulation are adopting the arguments of anti-regulators — postulating about future harms and foregone opportunities — while pursuing a regulatory approach that is only justified in the face of actual harm.

The fact of the matter is there is no evidence that harm exists in the sort of systematic way that justifies heavily regulating ISPs; the evidence that does exist suggests that current regulatory structures handle bad actors perfectly well. The only future to fear is the one we never discover because we gave up on the approach that has already brought us so far.

ISPs Acting Badly

The most famous example of an ISP acting badly was a company called Madison River Communication which, in 2005, blocked ports used for Voice over Internet Protocol (VoIP) services, presumably to prop up their own alternative; it remains the canonical violation of net neutrality. It was also a short-lived one: Vonage quickly complained to the FCC, which quickly obtained a consent decree that included a nominal fine and guarantee from Madison River Communications that they would not block such services again. They did not, and no other ISP has tried to do the same;¹ the reasoning is straightforward: foreclosing a service that competes with an ISP’s own service is a clear antitrust violation. In other words, there are already regulations in place to deal with this behavior, and the limited evidence we have suggests it works.

Another popularly cited case is Comcast’s attempted throttling of BitTorrent in 2007. While the protocol has legitimate uses, by far the most popular application was piracy; notably, pirate networks typically required users to upload as much content as they downloaded, imposing significant burdens on Comcast’s network. The FCC ordered Comcast to stop in 2008, but a federal court ruled that the FCC lacked the statutory authority given that ISPs were Title I providers (not Title II).

What is important to note, though, is that even before the Court ruled, Comcast had already removed its restrictions, not for fear of regulatory oversight, but by making technical changes to its network to better handle BitTorrent traffic, lending credence to Comcast’s arguments that the initial restrictions were about network management, not content discrimination (and, to be clear, Comcast erred in not being transparent). It is worth noting, by the way, that BitTorrent users were a sort of free-loaders, using massively more bandwidth than the vast majority of Comcast’s customers; this is a case where what is best for end users is much murkier than net neutrality advocates would have you think. What is pertinent, though, is that it happened only once.

Perhaps the most misrepresented episode, though, is MetroPCS. Net neutrality advocates claim that the discount carrier (since bought by T-Mobile) “blocked all video sites except for YouTube”; the reality is that in 2011 MetroPCS unveiled a new pricing plan: $40 for unlimited webpages plus YouTube, $50 for several other additional services, and $60 for unrestricted data. In other words, it wasn’t a net neutrality issue at all: it was an early prototype of what is known as “zero-rating.”

T-Mobile, Zero-Rating, and Competition

Zero-rating means that a particular service does not count against a data cap; widely used all over the world, the practice was popularized in the United States by T-Mobile.

Back in 2011 T-Mobile was a distant fourth-place in the U.S. carrier market, with limited spectrum and a shrinking customer base. That year the company tried to sell itself to AT&T, the second-largest carrier, but that deal was (rightly) blocked by the U.S. Department of Justice for competition reasons. That’s when something amazing happened: T-Mobile decided to actually compete.

The company launched its “Un-carrier” campaign, featuring contract-fee pricing with phone financing, data carryover, and, pertinent to this article, zero-rating on a host of music and video services (the video was downsampled). Customers loved it, leading T-Mobile to grow rapidly, soon overtaking Sprint to become the third-largest carrier in the United States. More importantly, T-Mobile forced the other national carriers to respond: now everyone has phone financing instead of lock-in subsidies, monthly plan prices are significantly lower for everyone, and even AT&T and Verizon, the two largest carriers by far, have returned to unlimited data plans.²

Again, zero-rating is not explicitly a net-neutrality issue: T-Mobile treats all data the same, some data just doesn’t cost money. Net neutrality advocates, though, have railed against zero-rating as violating the “spirit” of net neutrality, and shortly after that 2015 reclassification, the FCC launched an investigation into the practice. I’m sympathetic to the argument; I wrote in 2015:

The problem, though, is that regulations, by virtue of being words on a page, always contains loopholes that violate the “spirit” of the rules and more often than not end up favoring the incumbents, and that is precisely what is happening in the broadband war. Fast lanes would likely have only had an effect on the margins, and consumers would have been only affected indirectly; zero-rate data, though, appeals to consumer pocket books directly in a way that massively benefits whatever Internet companies are signed up to play ball with the ISP…

The FCC has signaled a hesitation to do anything about zero rate plans given the fact they benefit the consumer, at least in the short term. After all, who doesn’t like free? The problem, though, is the effect on competition — particularly the Netflix and Spotify competitors who haven’t yet been borne.

What has happened to the U.S. mobile industry has certainly made me reconsider: if competition and the positive outcomes it has for customers is the goal, then it is difficult to view T-Mobile’s approach as anything but a positive.

The Startups of the Future

Still, what of those companies that can’t afford to pay for zero rating — the future startups for which net neutrality advocates are willing to risk the costs of heavy-handed regulations? In fact, as I noted in that excerpt, zero rating is arguably a bigger threat to would-be startups than fast lanes, yet T-Mobile-style zero rating isn’t even covered by those regulations! This is part of the problem of regulating future harm: sometimes that harm isn’t what you expect, and you have regulated and borne the associated costs in vain.

That aside, the idea that ISPs would be able to successfully block sites and apps that don’t pay for delivery is flawed:

• First, as noted above, there is no evidence of this happening on a wide-scale.
• Second, should an ISP try, an increasing number of customers do have alternatives (not enough — more on this in a moment).
• Third, if the furor over net neutrality has demonstrated anything, it is that the media is ready-and-willing to raise a ruckus if ISPs even attempt to do something untoward; relatedly, a common response to the observation that ISPs have not acted badly to-date because they are fearful of regulation is not an argument for regulation — it is an acknowledgment that ISPs can and will self-regulate.

Most importantly, the idea makes zero economic sense. Remember that ISPs bear massive fixed costs, which means they are motivated to maximize the number of end users. That means not cutting off sites and apps those customers want. Moreover, even in the worst case scenario where ISPs did decide to charge Google and Netflix and whatnot, they could price discriminate and charge the Netflix competitor nothing at all! That would be a far superior financial outcome to a “take-it-or-leave-it” price that would foreclose all future startups (and again, there is zero evidence that this scenario has happened or will happen at all).

What is worth noting, though, is that the current regulations do foreclose startups that rely on low latency levels that might be offered by ISPs at a premium. The most commonly cited example is remote medical care, but the nature of future innovation is that we don’t know what sort of services might be created with something like paid prioritization.

I’d also note that companies like Google and Netflix already have massive advantages along these lines: Netflix places its content on servers within ISPs, and Google has an entire worldwide private network to ensure its results are milliseconds faster than they might be otherwise. The startups that challenge them will do so by being different, which means keeping open the number of possible ways to differentiate is a good thing.

Competition and Neutrality

To recap, given that:

• Regulation incurs significants costs, both in terms of foregone opportunities and regulatory capture
• There is no evidence of systemic abuse by ISPs governed under Title I, which means there are no immediate benefits to regulation, only theoretical ones
• There is evidence that pre-existing regulation and antitrust law, along with media pressure, are effective at policing bad behavior

I believe that Ajit Pai is right to return regulation to the same light touch under which the Internet developed and broadband grew for two decades. I am amenable to Congress passing a law specifically banning ISPs from blocking content, but believe that for everything else, including paid prioritization, we are better off taking a “wait-and-see” approach; after all, we are just as likely to “see” new products and services as we are to see startup foreclosure. And, to be sure, this is an issue than can — and should, if the evidence changes — be visited again.

What is worth far more attention is the state of competition in broadband generally: ISPs have lobbied for limits on public broadband in 25 states, and many local governments make it prohibitively expensive for new ISPs to challenge incumbents (and Title II requirements don’t help either).³ Increasing competition would not only have the same positive outcomes for customers that T-Mobile demonstrated, but would solve the (mostly theoretical) net neutrality issue at the same time: the greatest check on an ISP is the likelihood of an unsatisfied customer leaving.

And, I’d add, if neutrality and foreclosed competition are the issue net neutrality proponents say they are, then Google and Facebook are even bigger concerns than ISPs: both are super-aggregators with unprecedented power and the deepest moats ever seen in technology, and an increasing willingness to not be neutral.

---

It’s worth stating one last time: I believe deeply in neutrality and in fostering innovation. I think neutrality is the only way the Internet can function at scale, and innovation is how we will survive the Internet-driven transformation that is only just beginning. It is that belief, though, that compels me to push back against this specific regulation, no matter how much I agree with its associated catchphrase.

NOTE: I have written a follow-up to this piece that explains why I’m not making a “Market Will Fix It” argument, the trade-offs in terms of bandwidth investment, and why antitrust enforcement is so critical. You can read it here.

---

1. AT&T and Apple did have an agreement to limit VoIP applications on the iPhone; it is is difficult to separate Apple’s motivations from AT&T’s as the former was leveraging the latter to break carrier control — another outcome that was better for customers in the long-run even though it meant foreclosure in the short-run ↩

2. Subject to restrictions similar to those I listed above ↩

3. Another option is local loop unbundling, which I have discussed in a Daily Update here  ↩