Fast Forward: Broadband Caps Can Cost You
What's more important when you sign up for your Internet service: getting a lower price upfront or being assured that your bill won't jump later?
That's the crux of the debate over broadband caps, a pricing system some Internet providers have been pushing lately.
The idea behind it, as these companies explain, is to ensure that the most intensive Internet users pay their fair share. Many customers, however, see this as a duplicitous scheme to charge them extra for enjoying the Internet too much.
Two weeks ago, Time Warner Cable backed down from a broadband-cap proposal under which it would have sold access with varying quotas, starting at $15 a month for a plan allowing just one gigabyte of use. Each extra gigabyte would have cost $1 or $2, depending on the plan, with total surcharges capped at $75 a month.
Time Warner's plan had numerous flaws -- such as the steep overage fees and the fact that Time Warner, like other cable companies, offers no such pay-for-what-you-use option with its cable TV service. But other providers have been able to make broadband caps stick.
Last October, Comcast set a 250-gigabyte monthly cap on its residential Internet service. There's no charge for going over that line, which Comcast communications director Charlie Douglas said "fewer than one-tenth of 1 percent" of its customers do. But if you exceed it twice in six months the company can cancel your service.
Most providers in the United States, however, continue to shy away from specific usage limits. Verizon, for example, bans "high volume" use of its FiOS service but doesn't define the term. (Every responsible Internet provider prohibits such egregious abuses as spamming.)
As long as customers have a choice of providers, it shouldn't matter how dumb any one company's policies are -- the market ought to see to that firm's punishment soon enough. But too many users don't have that flexibility.
At the same time, the idea of charging people for the bandwidth they use isn't necessarily flawed. Done right, it could be a fair and understandable way to bring broadband to more people.
It's too bad, then, that the implementations we've seen so far seem to get everything wrong.
First, the idea of a 1-gigabyte limit has to die, right now. That's barely enough for an average month's worth of security patches, much less such monster updates as Microsoft's 345-megabyte Service Pack 1 for Windows Vista or Apple's latest "combo update" for Mac OS X, a 668-MB download.
It's unconscionable to charge people extra to get the fixes they need to keep their computers -- and everybody else's -- safe.
Second, anybody subject to a usage limit needs an easy way to see how close they are to maxing out. Many companies pushing these caps seem to forget that detail. Comcast, for instance, plans to offer a usage meter but for now can only suggest using the McAfee security software it provides -- which Mac and Linux users can't run.
Don't think you can guess your use. A basic bandwidth-meter program counted 1.14 GB of network traffic on my home computer over 12 hours, of which only 475 MB could be blamed on large downloads (a patch for Microsoft Office and a video documentary).
Here's the catch: I was at the computer for only five of those hours. For most of the other seven, I was asleep. Even discounting overhead confined to my wireless network, that's serious traffic.
Third, providers shouldn't treat people's excessive use as isolated money-making opportunities.
Instead of whacking them with a steep fee, why not bump them up to the next tier of service for that month, then let them switch back the next month? The wireless-phone industry offers other examples of how to make usage-based plans palatable: Providers could let customers roll over unused portions of a quota from one month to the next, or offer a prepaid service for people who want to keep their bills as low as possible.
Instead, some providers pushing these caps seem anxious to hand out fines, as if they're some sort of police agency.
Fourth, a capped broadband plan should be substantially cheaper than competing unlimited plans. Time Warner's $15-a-month rate would have barely beaten Verizon's price for basic digital-subscriber-line service. A $10-a-month plan, with a reasonable limit instead of a measly one gig, might pull in more dial-up users who have rejected broadband because of its costs.
Given all those conditions, a usage-based system could work. But companies might find it easier to stick with the sort of tiered service people already accept online, and which requires no complicated billing to account for each user's downloads and uploads -- a slower connection costs less; a faster connection costs more.