Who Should Pay for Netflix? | Hacker News Reader

Who Should Pay for Netflix?

attpublicpolicy.com

18 pointsmcdavis6 comments

The article is conflating the ISP's _costs_ with the _prices_ paid by consumers:

"And I think we can all accept the fact that business service costs are ultimately borne by consumers."

The above statement seems designed to cause confusion. The implication is that the total price consumers pay is determined wholly by the costs incurred in delivering the service. This may be almost true in aggregate, but isn't true at an individual level.

ISPs offer an all-you-can-eat internet service, but aren't happy with the customer-level margins for the most heavy users (including Netflix users). They could raise prices for heavy users (e.g. by applying caps in the same way as mobile providers) but have chosen instead to engage in different tactics to limit usage or get additional revenue from other sources. Throttling P2P traffic (I don't know which ISPs in the US do this) and asking CDN providers to pay for ports are two examples.

This seems to be like they're increasing the maximum speed of the local pipes ('unlimited water up to 100 litres per minutes!') without being willing to increase the size of the pipes into (and capacity within) their own distribution network.

Am I missing something, or is it just that the business model of offering unlimited high speed internet only sustainable at a higher price?

Both this blog post and the Netflix blog post it is responding to seem intentionally confusing. The core issue is that the interests of the consumer and provider are misaligned. In a normal market if the consumer wants more or better service, the provider will make more money by providing the service.

In this case the consumer wants more service in the form of more data for streaming video. In a normal market, the additional sales from the extra data would fund the required infrastructure. In reality, however, since the ISP sells a data rate, not data, they do not get extra revenue from heavy usage customers.

Both of these blog posts miss the mark by misinterpreting the probably rational consequences of this misalignment as unreasonable greed or unfairness. I assume that both of the authors know the real issue here and that these posts are meant to manipulate the public regardless of accuracy.

Especially for capped connections we already are paying for it. Say we get 250 gb a month. We should be allowed to use that as we see fit. If that means millions of cat pictures why should that be treated any different then streaming video. The big problem is that in most areas there is little to no competition so we are all stuck.

I find the argument pretty compelling. If Netflix usage requires more internet pipes to be upgraded, then it seems logical that Netflix users should pay rather than internet users in general.

However, it all presumes that we can attribute increased demand to Netflix and Netflix alone. Suppose you have two services, A and B, which each require 10 ZB of ISP bandwidth. The ISP builds 20 ZB of capacity and the world is happy. Now a third service, C, comes along and requires 10 ZB more of bandwidth, requiring the ISP to upgrade its 20 ZB pipes to 30 ZB. Does it make sense for the ISP/society to charge service C for that network upgrade? C's only crime is arriving last. Otherwise, it is identical to A and B in terms of the load placed on the network. It seems you me you could just as plausibly blame sustained demand for B as much as new demand for C.

I'm conflicted.

IMHO, the extra cost should be passed on to the consumer. This way, we don't get 1 ISP having deals with services A, B and C making people not switch due no other ISP having those same deals and no new ISPs come up because they can't get the same deals because nobody wants to switch.

There is already a vocabulary for why you are conflicted. It is called the insurance effect: Even though all the non-heavy users of the Internet are paying, in their flat rate bill, for the heavy users, most of them are OK with that arrangement because they are "insuring" themselves against the day when they, too, might become a heavy user.

What you are articulating is the risk that one day, you might move a few terabytes of photos, or upload your music collection to Google Play Music, or buy Carbonite backup, and you have only a vague idea what each of those services are going to do to your bandwidth use.

In theory it is rational to pay per bit, but that also assumes you know much more about your future use patterns than any consumer would know.

Now here comes the plot twist: Do m2m customers get cheap rates because they are willing to pay per bit? Noooooooo! They take it up the ass on a cost per bit basis. So maybe ISPs are just greedy bastards.

Reply on news.ycombinator.com