Thinking about Caps

Time-Warner’s bandwidth metering plan continues to attract attention, in part because a couple of prominent tech journalists are taking an activist position against it: Nate Anderson is unabashedly opposed to most revenue-enhancing plans that come from ISPs and carriers, and Stacey Higginbotham imagines she’ll be personally affected since she lives in one of the trial cities, Austin. The latest development is a threat by Rep. Eric Massa of upstate New York to ban usage-based pricing by law:

Massa has wasted no time backing the issue, sending out two statements last week about his displeasure with TWC’s caps. “I am taking a leadership position on this issue because of all the phone calls, emails and faxes I’ve received from my district and all over the country,” he said in one. “While I favor a business’s right to maximize their profit potential, I believe safeguards must be put in place when a business has a monopoly on a specific region.”

TWC’s plan to meter usage, which differs from Comcast’s cap system in several significant respects*, wouldn’t seem odd in most of the world: volume-based service tiers are the norm for commercial Internet services in the US, and for residential services in most of the world. This is largely because the costs for providing Internet service are significantly related to volume, owing to the interconnect costs born by ISPs (it’s not continuously variable, it’s more like a step function that ratchets upward in chunks as new hardware has to be added to keep up with peak load.) These folks are essentially wholesalers who buy an interconnect to the larger Internet through a transit provider or a carrier. If they’re too small to build an extensive private network, they buy transit and if they’re larger they pay for circuits to and from peering centers, which aren’t free even if you build them yourself (they take parts to build, and parts aren’t free.)

It’s not unreasonable to tie pricing to volume in principle, given that some users consume hundreds or thousands of times more bandwidth than others; we certainly charge 18-wheelers more to use the freeways than Priuses. The argument is over what’s a reasonable fee.

And to answer that question, we have to understand the role that Internet service plays in paying for the infrastructure that supports it. There has never been a case in the United States or any other country where Internet service alone generated enough revenue for a carrier to cover the cost of building an advanced fiber optic network extending all the way from the core to the detached single-family residence, even in the muni fiber networks toward which the neutralists are so partial; in places like Burlington, VT, Lafayette, LA, and Morristown, TN, the service the city offers over fiber is triple play (Internet, TV, and voice.) Without TV and voice, the take-up rate of the service is too low to retire the bonds. It’s simple economics.

So what happens when triple-play customers decide to download all their TV programs from the Internet and replace their phone service with a combination of cell and Skype? Revenues plummet, obviously. So the cable company wants to hedge its bets by replacing triple-play revenue with a higher bill for the higher usage of the remaining indispensable service. That doesn’t seem evil to me, as long as there’s some competition in the market, and the infrastructure is continually upgraded. Over time, the infrastructure will be paid for, and the price per byte will decline.

One of problems that we have with broadband policy in the US is lack of connection between infrastructure costs and service prices. TWC seems to be trying to solve that problem, and I’d like them to have some freedom to experiment without every member of congress within striking distance of a camera crew giving them grief.

In the meantime, TWC would help themselves a great deal if they adopted the policy of printing each customer’s monthly usage on the bill. They shouldn’t do anything about it for the time being, just show the amount for the next six months. At the end of that period, if they want to run a trial or two, the consumers will be able to place the service levels in perspective, and there will be a lot less whining. If service levels are adopted, there also needs to be a policy of re-evaluating them every year. If TWC had done these two things, this whole brouhaha could have been avoided. And yes, I’d be glad to sign on as a consultant and keep them out of trouble.

*Comcast has an elastic cap that can’t be increased by paying higher fees. If you exceed it for three months in a row, you’re ejected. It’s elastic because it takes three simultaneous conditions to activate.

, ,

One thought on “Thinking about Caps”

  1. Richard, one thing that should be emphasized is that consumers are being even more greedy than they accuse the cable providers of being. Many of them are trying to engage in price arbitrage by canceling their video service and stream TV programs via the Net 24×7. Trouble is, the bandwidth to deliver those individual streams costs many times more than either broadcasting (terrestrial or satellite) or sending the channels down a cable, because streaming is not a broadcast medium. Each new stream uses up more capacity and represents an incremental expense to the provider.

    Paradoxically, under today’s “flat rate” Internet pricing, this costs the consumer less and the provider more. In fact, the provider loses money — a situation that’s not sustainable. That’s why pricing models need to change. The consumer does deserve (and should shop for!) the most for his money, but the provider also deserves a fair return — and certainly should not be expected to lose money.

Leave a Reply

Your email address will not be published. Required fields are marked *