By Nate Anderson
I've just about managed to stuff all those gloopy spare bits of brain back into my skull, but it took a while. I blame Harold Ford Jr. and John Sununu, the ex-politicians now shilling for the ISP-backed Broadband for America, for blowing my mind. Their most recent op-ed, which takes shaky aim at Netflix, must be one of the dumbest such pieces I have ever read—and I have read a lot of them.
I would not absolutely swear that their
Mercury News piece makes any coherent points, though vestigial traces of argument run through the post, going nowhere. The overriding idea is that Netflix must hand over wads of cash to Internet providers—$83 million a year being a nice round number.
Netflix didn't get where it is today by handing out $83 million checks to anyone who asks, but the argument—such as it is—says that we need a new payment scheme that's "socially responsible and fair." Translation: Netflix is a dirty freeloader that quite possibly reeks of marijuana smoke.
"The reality is that Netflix and similar services want a free ride on the networks built with more than $250 billion in design, engineering, manufacturing, construction and maintenance," says the op-ed, giving its angle away. That sweet Netflix cash needs to flow to the small cartel of broadband operators who
aren't making any money
are
hugely profitable, so that they can keep building out their networks.
"But," I hear you asking, "doesn't Netflix somehow pay for its bandwidth already? And if not, how do I get a piece of that free Internet action?"
The economics of "freeriding"
Netflix does pay. The company's
2010 annual report (PDF) notes that it uses the "services of third-party cloud computing providers, more specifically, Amazon Web Services, as well as content delivery networks such as Level 3 Communications, to help us efficiently stream TV shows and movies." Neither Amazon nor Level 3 works for free.
Trashing the piece is almost too easy; it's become something of a sporting event in wonkish circles this week.
If Netflix received a free ride, then ISPs would have no way to increase the company's costs. As it is, however, "to the extent that network operators implement usage based pricing, including meaningful bandwidth caps, or otherwise try to monetize access to their networks by data providers, we could incur greater operating expenses."
In the Internet ecosystem, money flows in multiple directions, but no one works for free. Your home Internet fee pays your own ISP for its network, but portions of your bill are paid out by your ISP for "transit" onto other networks. In other cases, ISPs connect directly to networks with similar traffic volumes, called "peering," and the sides simply agree to call things even rather than send nearly identical checks back and forth. Newer arrangements like "paid peering" use the same direct connections but charge a fee, usually because one side is sending far more traffic than the other. Netflix also pays from its end, and its fee also propagates through the Internet thanks to
peering and transit agreements.
Internet distribution is certainly cheap—Netflix estimates streaming costs about a
penny per hour—but it's not free.
The Ford/Sununu op-ed then makes a totally different argument, that the Netflix calculation "ignores the hundreds of billions of dollars in sunken network investments needed to create that one-penny marginal cost efficiency at the customer's end." True, but that's because Netflix is only talking about streaming. Ford and Sununu make it sound like the content is being freely acquired and then transmitted at almost no cost.
This is patently absurd. To get streaming content, the Netflix annual report points out that "we typically enter into multi-year, fixed-fee licenses with studios and other distributors. As of December 31, 2010, we had over $1.2 billion in such contractual commitments covering payments due over the next several years. Furthermore, we plan on increasing the level of committed content licensing in anticipation of our service and subscriber base continuing to grow." And, as Netflix grows in popularity, those fees are going up.
So Netflix pays to acquire the content, then again to stream the content, and consumers pay both their ISP and Netflix to view the content. This is so far from a free ride that Netflix spent $1.15 billion, or 53.4 percent of its 2010 revenues, to acquire its video content and distribute it on DVDs and over the Internet. Its streaming bill in particular increased in 2010 "due to higher costs associated with our use of third-party delivery networks resulting from an increase in the total number of hours of streaming content viewed by our subscribers."
Then the Ford/Sununu piece careens again. Suddenly, the issue isn't about what Netflix pays, it's about what consumers pay. "Broadband networks are delivering more than just the latest sitcom episodes and hottest movies," says the piece. "They facilitate telemedicine, education, job training, telecommuting and many other functions. It hardly seems fair to make users of these services pay more in order to subsidize Netflix's costs of delivering their videos online."
The thinking here isn't really explained, but it
seems to be that Netflix streaming uses tons of bandwidth; grandma just uses the Internet for e-mail; ergo, grandma should pay less for her Internet connection than those Internet-hogging hippies next door. To which the only real answer is: ISPs can certainly try this pricing model if they want! (It has
not proven popular in the past.)
Indeed, Ford and Sununu then note that the FCC has no problem with ISPs "asking subscribers who use the network less to pay less, and subscribers who use the network more to pay more." US ISPs actually do this already, especially on smartphones, where extra fees kick in once users exceed a certain monthly data allowance.
"Ham-fisted analogies"
The op-ed ends with the line, "But many increasingly question a service that forces tens of millions of non-customers to pay for something they never use." Netflix doesn't do this, of course; ISPs do, and they are free to alter their pricing schemes in a non-discriminatory manner and see how the market responds. The exact same argument could be deployed against any higher-than-average-bandwidth Internet service; what it has to do with Netflix is unclear—and my overtaxed brain simply exploded.
I wasn't the only one. Comments on the piece are utterly negative. The very first one says, "With all due respect, this has to be one of the most poorly written or thought out articles I've come across in some time."
Trashing the piece is almost
too easy; it's become something of a sporting event in wonkish circles this week. Matt Yglesias, from the Center for American Progress,
concluded, "Indeed, it’s difficult to see what the policy issue here is supposed to be. Do they want regulators to create a cartel? Why?"
Liberal media watchdog
Media Matters said, "It's possible that there is a point buried somewhere in its ham-fisted analogies and confused understanding of how broadband services work, but I'm struggling to see it."
Target acquired
So what was the target of the op-ed? "Data transfer" is one of the top issues on the Broadband for American website, which
describes the "challenge" before us as: "how can we best ensure ongoing investment in robust broadband networks, let networks continue to work out their traffic exchanges by contract, and permit any disputes to be resolved without involving the courts or the regulators?"
This is aimed at Level 3, the huge transit provider that more recently got into the content delivery business and now hosts much of Netflix's data. Last year, Level 3 and Comcast got into a
nasty spat over peering arrangements. Comcast wanted Level 3 to start paying to deliver traffic, while Level 3 retorted that its Netflix traffic in particular had been requested by Comcast users and it was really doing Comcast a favor. The dispute then
migrated to the FCC.
Comcast and its fellow ISPs want to charge more to content providers, and they don't want the government mucking about with these contracts. Separately, many would also like to lower data caps and institute overage charges under out of "fairness," but consumers have rebelled against low caps (higher caps have passed without too much resistance).
The op-ed seems like an effort to tie both of these ISP concerns to the "Red Menace" that is Netflix. Instead, it manages the impressive feat of making both ideas look just about bankrupt.
Ludhiana, November 20