John Oliver has received a lot of attention over the last few days for his 13 minute rant on Net Neutrality aired on the most recent episode of his new HBO show, Last Week Tonight with John Oliver.
Perhaps Oliver and his writers were attempting to take a page from Stephen Colbert’s book, presenting a sketch based on truthiness. If so, they sure missed the target. For when Stephen Colbert uses truthiness, it’s with a clear sense of irony, as he lampoons the very argument he makes. Oliver’s rant, on the other hand, was truthiness argued with the same sense of hubris that Colbert so routinely, and deftly, mocks.
This hardly makes Oliver unique. In fact, most of the arguments made by other net neutrality advocates (which Oliver was largely parroting) are also nothing but truthiness: arguments based on gut feelings and appeals to emotion, rather than on objective, logical and informed examination of facts and evidence.
The Internet is not Broken
It starts from the very beginning, with Oliver declaring:
The internet, in its current form, is not broken—and the FCC is currently taking steps to fix that!
Oliver’s words are much closer to the truth than he realizes, except of course, that he means them in precisely the opposite way.
The internet, indeed, is not broken. The internet has been operated via commercial entities since the late 1980s, and the internet backbone became fully commercial nearly two decades ago, in 1995. In that time, there has been one solitary, legitimate, documented example of a US ISP maliciously blocking or degrading traffic: the 2005 case of Madison River Communications, a small local telephone company serving approximately 40,000 internet subscribers in North Carolina, blocking competing VoIP phone service, Vonage. Within a month of a complaint being filed with the FCC, Madison River had agreed by consent decree to pay a fine of $15,000 and not to block VoIP services again.
So, in more than three decades of commercial internet service, there has been just a single documented example of a legitimate net neutrality violation, resolved expeditiously, without any formal net neutrality regulations in place!
You see, for almost the entire history of the internet—up until 2010—there were no formal net neutrality regulations, whatsoever.
(Net neutrality advocates often claim that Comcast maliciously blocked BitTorrent traffic in 2007, but as we explained in a prior article, that claim is simply false. Nonetheless, after receiving complaints, the FCC objected, and Comcast immediately modified its network management practices, in response—again, despite the fact that there were no formal net neutrality regulations in place.)
The internet is not broken, and history suggests that formal net neutrality regulations aren’t necessary for the internet to function properly. The calls for net neutrality regulations, the formal adoption of “Open Internet Rules” in 2010, and the renewed calls for even stronger net neutrality regulations, after most of the Open Internet Rules were struck down by a federal court earlier this year, and the current FCC effort to institute new net neutrality regulations (which Oliver criticizes), are all attempts to fix something that’s not broken.
Denouncing a Non-Change in Policy
This, coupled with the fact that the 2010 rules were limited in scope, is why it was mostly a yawn back in January, when those rules were largely struck down. To the degree it wasn’t a yawn, it wasn’t a yawn because it eliminated artificial barriers to doing things that would actually be beneficial to consumers.
For example, the particular policy proposal that John Oliver was so vehemently denouncing, was that new net neutrality regulations might not prohibit paid prioritization of traffic.
Technically, this doesn’t represent a policy change, since there’s no current regulation in place that prohibits paid prioritization. But what neither Oliver, nor most other net neutrality advocates, seem to realize is that prioritization was also allowed under the 2010 rules, it just wasn’t clear whether it was allowed for everybody.
When it adopted the 2010 rules, the FCC made clear that it was not prohibiting ISPs from delivering managed services over IP. So, for example, virtually all cable operators offer VoIP phone service, and that is normally implemented as a managed service, where VoIP traffic is prioritized, to ensure it is not adversely impacted by momentary disturbances.
It’s important to note that prioritization of traffic is really only an issue for a small class of current applications: those which require real-time interactions, such as voice calls, and video conferencing, where the protocol is very sensitive to momentary disturbances. Prioritization is not important to virtually all of the traffic most users consume, including web pages, email and even streaming internet video (which over the last several years, has transitioned almost entirely to http adaptive streaming, a delivery method that uses deep buffers to avoid quality impacts due to momentary disturbances).
So, under the 2010 rules, it was very clear that ISPs could prioritize their own managed services. It was less clear whether those ISPs could offer priority service to third parties. (The rules prohibited “unreasonable discrimination,” leaving open many questions as to what was reasonable.)
But if it was actually prohibited for ISPs to offer priority service to third parties, that would seem antithetical to consumer interests. After all, if cable operators are allowed to prioritize their own VoIP traffic, for example, in order to protect service quality, wouldn’t it be best for consumers if over-the-top VoIP providers, such as Vonage, were able to obtain prioritization, also?
From a consumer standpoint, the most sensible regulation would be to actually require ISPs to offer priority service, on a non-discriminatory basis, so that over-the-top providers could compete with ISP managed services on equal footing.
And guess what: that’s not so far from what FCC Chair Tom Wheeler was apparently contemplating. While it didn’t go so far as to require ISPs to offer priority services to third parties, it did contemplate explicitly approving paid prioritization, subject to the requirement that such prioritization be made available to all comers, on commercially reasonable terms. That is, ISPs would not be allowed to play favorites, and offer prioritization to some, but not to others.
Focused on Politics, not Policy
It’s hard to emphasize enough how irrelevant prioritization is to most current services. In a well-managed network, it’s relevant only to a limited set of current services (and potentially some as yet undeployed future services), that are very sensitive to momentary disturbances.
Net neutrality advocates assert, however, that ISPs might intentionally degrade their own service, in order to force content services to pay for prioritization.
In a very poorly managed network—one that is not provisioned with sufficient capacity—it’s certainly true that a broader set of services might benefit from prioritization, as protection from sustained congestion. But it’s important to understand that a network cannot get to that point of such congestion without severely impacting almost all traffic. This is not a minor situation that could go unnoticed. An ISP with a network in this condition would be bombarded with customer complaints, demands for refunds, and would face a massive loss of customers. FCC monitoring would show them completely failing to deliver promised speeds. (Speed tests, including the FCC monitoring tests, effectively measure the unused capacity of the network. In the case being contemplated here, there would be long periods with no unused capacity at all. In such case, speed test monitoring would show performance at a tiny fraction of published speeds—quite possibly so bad as a low, single-digit percentage.)
The notion that any ISP might so completely crater its service in order to strong-arm paid prioritization is, quite frankly, ridiculous. And to intentionally degrade any service for such anti-competitive reasons would go to the heart of principles that have spurred decisive FCC enforcement action, even in the absence of formal net neutrality regulations.
But remember, this is truthiness. It is a game of politics, not serious policy analysis. The net neutrality advocates have already made their minds up that they don’t like ISPs, they don’t trust ISPs, and they believe ISPs have evil intentions, so more regulation is better, and any argument they can exploit to achieve such is fair game. It doesn’t have to make rational sense, and it doesn’t matter that they don’t understand the actual business dynamics, market dynamics, or in most cases, even the technology issues. It just has to be a bogeyman that easily resonates (with themselves, as well as the public). The ends justify the means.
Alleged Assault on the Little Guy
Oliver next asserts (or rather, repeats assertions), that paid prioritization would hurt the little guys: startups attempting to compete with established internet behemoths, but who couldn’t afford to pay for priority.
Nothing could be further from the truth.
First, since very few services would benefit from such prioritization, very few startups would even have interest in such.
Those that actually had a need for such—say, because they’re launching some novel, new, high quality, real-time service—would be looking for the opportunity to obtain prioritization, not running away from it.
The same basic allegation was made when AT&T announced plans for sponsored data service, earlier this year: that startups wouldn’t be able to compete with behemoths.
We made the same point then that the key issue in terms of protecting competition is ensuring that the service is available on commercially reasonable terms to all comers. So long as that’s the case, historically, it is new entrants that have used such services to differentiate themselves and challenge the incumbents. (For example, toll-free 800 numbers and free shipping were used by internet retailers, and before them, mail order retailers, to compete with entrenched brick-and-mortar incumbents.)
But despite all evidence to the contrary, Oliver insists that cable operators would actually degrade non-prioritized traffic:
If we let cable companies offer two speeds of service, they won’t be Usain Bolt and Usain Bolt on a motorbike. They’ll be Usain Bolt and Usain…bolted to an anchor!
All Hail Netflix
As evidence, Oliver cites the recent peering standoff between Netflix and Comcast, and a chart showing Netflix streaming performance on Comcast declining for three months, before increasing significantly after the two struck a deal, suggesting it is obvious evidence of intentional degradation by Comcast to force Netflix’s hand in a deal.
Users love Netflix, and mostly hate their ISPs, so it’s very easy to believe that the ISPs must be bad guys, assailing the seemingly consumer-friendly Netflix.
Nevermind that Oliver’s inference couldn’t be farther from the truth, and even Netflix didn’t suggest intentional degradation by Comcast. (In fact, during the dispute, Netflix CEO Reed Hastings actually acknowledged that wasn’t what was happening. Unfortunately, I’ve been unable to locate the article that reported such, in order to provide a link here.)
First, let’s note that this wasn’t just an issue with Comcast. As can be seen from the chart, AT&T and Verizon speeds declined in an eerily similar pattern, while Cox, Cablevision and Google Fiber speeds increased.
What distinguishes these providers? Cox, Cablevision and Google Fiber all joined Netflix’s OpenConnect program, while Comcast, AT&T and Verizon all declined to do such.
With its OpenConnect program, Netflix has sought to reduce its costs to deliver content over the internet (called “transit”) to nearly nothing, and they have not been shy about using strong-arm tactics to promote that goal. Among their public efforts to apply pressure, they initially blocked users of ISPs who didn’t join OpenConnect from accessing their best quality “Super HD” streams and 3D content, and publicly trumpeted this fact, before bowing to criticism that they were improperly discriminating, themselves.
What is actually reflected in the chart was another aspect of Netflix’s attempt to pressure ISPs. As Netflix streaming naturally grew (it has long been growing at a very fast rate: a 56% CAGR over the last six months), Netflix refused to purchase additional transit capacity for delivering content to the OpenConnect holdouts.
This was a game of chicken, controlled by Netflix. Netflix hoped to create enough pain to force Comcast and the other holdouts to join its OpenConnect program. What actually happened was Netflix felt the pain more acutely, and so made a deal with Comcast (and shortly afterward, Verizon), to resolve the issue.
Netflix actually reduced its costs by striking a transit deal directly with Comcast, rather than the third parties it had been using. So what Netflix has really been complaining about afterward is the fact that it wasn’t able to reduce its transit costs to nearly nothing, as it had hoped.
Politics Makes Strange Bedfellows
As further evidence of the grave threat, Oliver tells us how net neutrality activists have partnered with internet behemoths such as Google, Amazon, Facebook, and yes, Netflix, in opposition to the possibility of paid prioritization, and in favor of “strong” net neutrality regulation, more generally—an alliance he likened to Superman joining forces with Lex Luthor.
The mere fact that these entrenched incumbents are united in opposition to paid prioritization ought to immediately call into question the assertion that such prioritization would somehow favor incumbents over new entrants. Google, Amazon, Facebook and Netflix are not egalitarian enterprises, seeking to protect the ability of new entrants to compete against them. They are sophisticated corporate entities seeking to protect their own interests.
These companies have regularly paid lip service to net neutrality, but conveniently redefined their notion of net neutrality whenever necessary to align with their parochial business interests.
As Brendon Sasso reported, the net has never been neutral, and entrenched incumbents have routinely sought advantages in internet delivery over their smaller competitors. A few years ago, that extended to Google creating its own CDN, and striking co-location agreements with many ISPs in order to improve performance and reduce its transit costs to nearly zero. And upon seeing Google’s success, others, like Netflix, sought to do likewise.
These entities aren’t trying to drastically reduce their own costs for transit because they think it creates a level playing field between them and their smaller competitors. They do it to benefit themselves—and certainly without any reservation over the fact that it gives them a tremendous competitive advantage over their smaller competitors.
Any claim that internet behemoths have adopted a position on net neutrality based on principled, egalitarian concerns over free competition is simply preposterous.
The Consumer’s Interest
Furthermore, the belief that these behemoths’ interests are aligned with consumers is misguided. As with all companies, their interests are sometimes aligned with consumers, and sometimes not. Each case requires an individualized analysis.
For example, when Netflix joins self-proclaimed consumer advocates in opposing all forms of broadband capping or metering, broadband subscribers are inclined to believe that Netflix’s interests are aligned with theirs.
In fact, Netflix’s interests are at best aligned with only the one-third of broadband subscribers who are also Netflix subscribers. The two-thirds of broadband users who don’t subscribe to Netflix, and likely even some light Netflix users, on the other hand, are actually massively subsidizing the cost of Netflix service, and Netflix’s interest is to maintain the status quo of that massive subsidy. Most broadband subscribers would be much better off if broadband service were priced with some form of usage metering or tiering, so that they wouldn’t be forced to subsidize the heaviest users.
Flooding the FCC
Oliver encouraged his viewers to complain to the FCC, and reportedly, that encouragement generated so many submissions that it brought down the FCC’s comment system.
Oliver had joked that internet trolls should redirect their vitriol to the FCC’s net neutrality docket. Again, he probably didn’t appreciate how close to the truth that suggestion was. For Oliver certainly didn’t enlighten his audience with an informed and considered argument. He merely whipped up misguided populist opposition, with nothing but truthiness.
Toward a Sensible Policy
Oliver attacked FCC Chairman Tom Wheeler, on the basis that he once led the cable industry lobby. And while that’s certainly reasonable cause to be wary of Wheeler’s views and motivations, we should be wary of the views and motivations of all politicians.
Historically, Wheeler has expressed strong support for protecting the internet from anti-competitive or unreasonable discrimination. However, he has expressed reluctance to implement such by prescriptive regulations that may have unintended and detrimental consequences.
Rather than create new regulations based on mere predictions of future threats, he suggested it was better to respond to problems as they actually demonstrated a need for intervention.
This would seem a sensible approach, given it proved very effective at quickly remediating the very few instances of real or alleged violations, in the past.
And the reality is that the mere threat of heavy-handed regulatory intervention is a very effective sword of Damocles to ensure ISP compliance with principles prohibiting anti-competitive or unreasonable discrimination.
In other words, all evidence suggests that it’s not necessary to implement heavy-handed regulations that risk harm to future innovation and consumer interests, in order to ensure continued fair treatment of traffic by ISPs. The system has worked for decades without such regulations. It can work for decades more without them, and the threat of heavy-handed regulation, if ISPs don’t comply, can be expected to be very effective at ensuring such.
As John Oliver told us at the outset of his rant, the internet in its current form is not broken. We don’t need for the FCC, or John Oliver, to fix it.