Netflix Gives FCC its Broadband Marching Orders

This is a third post on the FCC’s 12th Broadband Progress Report (the others are here and here). Netflix has filed a comment in the inquiry, which isn’t notable by itself. After all, Netflix is a very significant stakeholder in the US broadband services, consuming more than a third of US Internet capacity during prime time.

Netflix Echoes Wheeler

While the firm has a right to be heard, the FCC has established a peculiar pattern with respect to Netflix’s wishes. Two years ago, Netflix declared in its comments on the Tenth Broadband Progress Report that “a 25 Mbps connection is fast becoming ‘table stakes’ in 21st century communications”. This remark echoed verbatim a speech delivered by Chairman Wheeler at 1776 Communications on the very same day as the Netflix filing: “A 25 Mbps connection is fast becoming “table stakes” in 21st century communications“.

Wheeler Echoes Netflix Echoing Wheeler

Oddly, the NOI for the 11th Broadband Progress Report attributed Wheeler’s opinion about the 25 Mbps threshold to Netflix, even though Netflix attributed it to Wheeler. Perhaps Wheeler and Netflix are so closely aligned that neither knows where their shared talking points originate. I suspect this is a common characteristic in “captured regulators”, but that’s not really my field.

If the pattern still holds, we may be able to divine the FCC’s intentions from the Netflix comments on the 12th Broadband Progress Report as well. So let’s see what Netflix is asking the FCC to do.

Is the Internet for television?

First, Netflix wants the FCC to “determine whether American households have access to broadband services sufficient to consume Internet television” (page 2). This is a reasonable request on its face, consistent with my wish for the FCC to determine application requirements. But Netflix offers a peculiar method of determining the answer: it wants the FCC to place television “at the center of a consumer’s expectation for what broadband connections can deliver” by taking the requirement for 4K TV and then doubling it for no good reason. But 4K is not yet mainstream and arbitrary doubling is irrational. This is an annual survey, so it only needs to look at current requirements.

Is the Internet for television? Well yes, that’s part of the what we expect from the Internet. But I’m not convinced that TV is “at the center of the Internet”. TV is a one-way experience, and the Internet is fundamentally a two-way medium. So no, TV is not at the center, interactive communication is. So let’s not exclude the truly advanced applications that are always emerging on the Internet. While I want the FCC to focus on applications, I don’t consider the Netflix application very special.

End All Wireline Data Caps!

Netflix has a simple position on data caps: Wired networks shouldn’t have them at all, and data caps for wireless networks should be high enough that every wireless subscriber can watch all the Netflix she wants without worries: “[T]he Commission should hold that data caps on fixed-line networks — and low data caps on mobile networks — may unreasonably limit Internet television viewing and are inconsistent with Section 706.” This appears to go well beyond the 706(b) mandate because it brings business models into an inquiry that’s supposed to be about technology buildout.

But Netflix may have a point due to the clumsy wording of the Telecom Act. The section directs the FCC to investigate “advanced telecommunications capability” without any more specifics. While it’s generally assumed that the word “capability” means “the things the networks are capable of doing,” it’s not completely unreasonable to interpret it from the user perspective.

Bounding the Inquiry

But it really amounts to the same thing. The problem Netflix runs into is that a broadband provider who offers a cap-free service option is making the capability available. Whether users choose to adopt the most capable package tells us a lot about what the definition of broadband service should be, but it can’t be used to punish providers. So the inquiry should not make policy demands on providers, especially when doing so will inevitably raise the price of broadband. But don’t be surprised if Netflix is foreshadowing the order that follows this inquiry.

Wheeler has a history of tailoring national broadband policy to fit the needs of the Netflix business model after all. Even if Netflix is right that “Data caps and UBP raise the cost of using the the connections that consumers have paid for, making it more expensive to watch Internet television” [sic] that’s not really the issue for the 706 report.

Zero-rating Complaints

Netflix is also upset about zero-rating plans that exempt some services from data caps: “a BIAS provider could also employ data caps in a way that explicitly discriminates in favor of one content source or another, further limiting consumer choice.” But this complaint ignores the technical reality behind common zero-rating schemes. If an ISP is willing to exempt streams that originate at providers who have made infrastructure investments that lower the ISPs costs, zero rating is entirely appropriate.

This is because the investments – which could include things like placing servers inside the ISP’s network – are effectively a form of payment to the ISP, so billing the customer for low-cost data at the same rate as the higher-cost data that is not accompanied by investment amounts to double billing. And as Netflix said about data caps in the previous section, raising the cost of using the connections that consumers have paid for makes things more expensive. Obviously, nobody wants to do that.

It should be pretty easy to distinguish rewards for investment from random fees.

Free Interconnection

Quite often we see advocates bringing their pet issues into inquiries that don’t seem to have much of a nexus. Merger reviews used to bring out demands for net neutrality (before the Title II reclassification) as well as various other demands for subsidized service and relaxation of copyright enforcement. Netflix’s main hobby horse is free interconnection because some carriers charge fees for redirecting investment. Left to their own devices, ISPs will expand connectivity at the natural borders of their networks. Similarly, Netflix would like its infrastructure to grow at its natural borders. When these borders don’t align, somebody has to pay for some new bridges and highways, and each player – Netflix and the ISPs – wants the other one to pay at least some of these costs.

So Netflix is trying to hijack the 706 inquiry to get some cheap or free interconnection capacity because “Without interconnection there is no Internet access.” That’s a bit over-simplified.

Finally, Netflix wants wireless broadband to run at 10 Mbps or faster, so don’t be surprised if we see another round of “table stakes” call-and-response.


The non-network Internet services such as Netflix and Google made out like bandits in the FCC’s Open Internet Order, but there is still more that they’d like the FCC to do for them. Not only are the pending rulemakings on privacy and set top boxes opportunities for additional favors, even routine inquiries such as the annual Broadband Progress Report have become politicized. While I’d like the FCC to stick to its knitting, when it’s being pressured by favored constituencies it rarely says “no”.