Sadly Comcastic!

Mere loathing won't fix bad broadband policy.


Let me make it perfectly clear: I hate Comcast. The years I was sentenced to the company's crappy CATV and broadband service were long, brutal, and demeaning. And that was just the time spent interacting with Comcast customer service.

Still, contrary to much of the recent vilification the company has received, Comcast has a hint of a justification in attempting to "shape traffic" on its network by throttling bandwidth hogs. More importantly, Comcast's misdeeds do not necessitate—as many have assumed—some sort of federal solution and associated broadband policing.

First, Comcast was absolutely wrong not to inform its customers that file-sharing apps like BitTorrent would be disrupted by active Comcast intervention. That is a deceptive business practice, plain and simple. However, there are still sound reasons for the traffic shaping goal—which the AP hysterically labeled "data discrimination"—that require no ill-intent to explain.

Any network—roads, broadband, canals—has peak usage times. The provider of that network wants to maximize the number of customers they can charge to use that network, of course. But you cannot attract customers unless you promise some reliable level of service and try to smooth out usage peaks and valleys.

Applied to the world of Comcast, who are often trying to service many customers through one access point, one or two BitTorrent or Gnutella freaks can eat away at the access promised to other customers. Comcast would have to be suicidal not to take note of this fact and try to address it.

This is not "data discrimination," but common sense. Likewise it is common sense that consumers would not actually benefit by turning broadband access into a quasi-public utility via a declaration that Internet service is a common carrier, with absolutely no differentiation in the traffic carried, and answerable only to regulators on that count.

Think about the kind of consumers we want to be—captive or coveted? In the old telco common carrier model, consumers had no choice in provider, and only scant choice for services. Maybe a choice of Bakelite handset color if they were lucky. Contrast that with today's hypercompetitive cell phone market, where consumers are courted by providers with hundreds, if not thousands, of options to reflect a user's needs.

This brings us to the major confusion with data discrimination/Net neutrality thinking: Bits may be bits, but users are not users. We should not mistake the binary, on-off nature of digital communication for the end result, which is the untold number of different ways that this means of communication can be used.

Old dial-tone voice service did only one thing, so it was fairly easy for regulators to handle. You either were able to make and receive calls, or you were not. Then the government decided that long distance calls should subsidize local service, and that was that. You heard a dial tone when you picked up your receiver, and that was your expectation as a consumer of telephone service.

A broadband consumer expects—what, exactly? Low-latency for gamers. Reliability for businesses and home offices. Upstream vs. downstream—which is most important to you? Downstream for most folks, but not all. Not someone trying to push a 15 megabyte Power Point presentation through 48K. Is voiceover IP a priority for you? IP TV?

A federal "Civil Rights Act" for Internet bits couldn't possibly address all of these competing uses for broadband service.

In fact, the best solution to providing competition and making broadband consumers coveted by providers does not lie in federal legislation; it is found in breaking the monopoly service providers like Comcast have with local governments. If we are really worried that consumer data will be discriminated against, then the best response is to ensure that consumers have multiple choices in local service providers.

Failing that, local governments can and should negotiate levels of service requirements into their franchising agreements with providers like Comcast. Such requirements have long been a part of cable TV franchise agreements, and explain why Comcast or Time Warner will roll a truck for reports of snow on your basic cable tier. But report flaky Net service? If you're lucky, they'll drop off a new modem in a day or two.

Robust and reasonable Internet service requirements would at least start to ensure that providers could not blame shoddy network performance on peer-to-peer users, or go on a witchhunt against them in order to avoid spending money on needed network upgrades. At a minimum, every franchise agreement should have some minimum up-time and advertised speed language.

Would that usher in Net utopia? No, but it would help ensure consumers get what they pay for without engaging the regulatory blunderbuss of Congress. And it would leave time for the truly important things in life.

Like hating Comcast.

Reason contributor Jeff Taylor writes from North Carolina.