Image via CrunchBase
Apple says users of its iTunes online media service are purchasing over 800,000 TV episodes and 350,000 movies per day. Netflix’s 36 million customers are watching more than a billion hours of movies and TV each month. YouTube is now so large that its parent, Google, accounts for as much as 25% of the entire Internet.
This media maelstrom, however, is changing the economics not only of content but also of Internet infrastructure. And right in the middle of this question is Verizon’s challenge to the Federal Communications Commission’s net neutrality rules. After the recent oral argument, it’s increasingly likely the case hinges on the old concept of “common carriage.” Common carriage said monopoly network industries, like telephones, are public and cannot discriminate – they must take all comers on standard terms and rates. Congress said the FCC may not apply common carriage regulation to the Internet. But the FCC’s net neutrality rules contain a “non-discrimination” provision, and two of the three D.C. Appeals Court judges appear to think it qualifies as common carriage. Legal observers think at least portions of the rule will be overturned.
What does this mean for the Internet? The bottom line is it is probably good for that voracious video viewing breed known as the American consumer.
The media and cloud services from the likes of Facebook, Amazon, Microsoft, and the gaming company Valve, put enormous strain on wired broadband links to homes and businesses and on wireless networks connecting billions of mobile devices. This isn’t necessarily a bad thing -- the cloud services help drive robust demand for broadband and mobile subscriptions. But the economics of the ecosystem is, to use a technical term, out of whack, and it is tilted, perhaps unsustainably, by the net neutrality rules.
That’s because YouTube and Netflix may, between them, consume more than 50% of last mile network capacity during peak hours of the day. These services not only demand lots of raw capacity but also -- because it is maddening to watch glitchy, halting videos -- require low latency and jitter. In other words, they do not tolerate delayed delivery of data packets. This strains the network. That means more network capacity and more cost for consumers.
And this is the rub. The net neutrality rules, as written, shift 100% of the cost of last mile bandwidth to consumers. The rules bar commercial relationships between the broadband service providers and the online content companies, in effect mandating a zero price. For the content companies, it is a regulatory subsidy. For consumers, it’s potentially a bad deal.
What if magazines were banned from selling advertisements and had to push all their costs onto subscribers? Such publications do exist, but a rule requiring this as the sole business arrangement of the publishing sector would dramatically reduce the diversity, efficiency, and creativity of the market. It would raise prices and reduce consumer choice.
Broadband networks, like magazines, are classic multi-sided markets, a point that somehow did not penetrate the mind of FCC counsel. During oral arguments, he asserted, repeatedly, that the online content providers (or “edge providers”) are not the customers of the broadband service providers. The market, counsel said, has the following structure: broadband service providers sell Internet access to consumers, on the one hand, and edge providers buy their own ISP services on the other hand, and the two sides are mediated by “the Internet.”
There are two big problems with this line of reasoning: logic and facts.
Counsel is advocating the legality and desirability of a rule that says the edge provider cannot be Verizon’s customer using the argument that the edge provider is not Verizon’s customer. But the edge company is not the customer of the broadband provider, at least in part, because of the rule. That’s circular reasoning.
The technical and economic substance is an even greater challenge to his case, however. It is true that he describes the basic arrangement of the Internet -- as it existed a dozen years ago. In his book The Dynamic Internet, Christopher Yoo, a professor of law and engineering at Penn, described the early Internet’s typical three-tier structure:
consumer ⇆ dial-up access line (tier 3) ⇆ regional ISP (tier 2) ⇆ public Internet backbone (tier 1) ⇆ regional ISP (tier 2) ⇆ DS3 access line (tier 3) ⇆ Web server.
Today, however, after more than a trillion dollars of investment in networks, data centers, and wireless cells, the Internet looks much different. The broadband service providers peer (or connect) directly with each other. They have paid peering and paid transit relationships with other networks. And content delivery networks (CDNs) like Akamai, which store content closer to network end points and route it efficiently, have exploded -- they now account for more than 50% of network traffic. Google Global Cache appliances, which store most YouTube videos, now reside inside the broadband service provider networks. Netflix is asking to place its OpenConnect appliances inside the broadband networks. In all, as much as 80% of IP traffic -- from companies like Google, Microsoft, Netflix, Facebook, and the broadband providers own backbones -- now bypasses the public Internet.
The distant relationship counsel described is thus largely a thing of the past. Broadband service providers and content providers are ever more closely linked. And the multi-sided market becomes possible, even necessary. At the very least, exploring business and technical arrangements in this highly volatile sector should be an option.
Broadband service providers can, in theory, use pricing strategies to differentiate between high capacity and low capacity users. But there has been lots of pushback against “tiered pricing,” “data metering,” and “pay for priority.” If differential pricing is outlawed or discouraged, broadband providers may have to raise prices for everyone, hurting light users and low-income users the most.
One might argue that if the multi-sided market arises, content companies will just pass along the cost of bandwidth to end consumers, and we’ll be right back where we started. True, they may pass along a portion of the cost. But we’ll still be better off. Such arrangements would allow for far more granular decision making among consumers, broadband network operators, and content providers. The high-end Netflix-watching, online-gaming, YouTube gobbling would still pay more than the email-checker, but the content companies would be paying something for direct access to a trillion-dollar last-mile network, and consumers would pay less than they otherwise would. (Or, as we saw with CBS and Time Warner Cable, perhaps broadband will pay for content. Who knows?) The bottom line is that in a world without net neutrality tipping the scales, all parties would more fully realize a rational return on the value they contribute to this vibrant, and vital, ecosystem.
