As of Oct. 1, Comcast customers in a few small markets are now subject to metered data use. Households that use more than 300 gigabytes of data per month will have the choice to pay $10 for an extra 50 gigabytes or $30 per month for unlimited service. The cable giant joins the No. 4 player in the industry, Cox Communications, in offering tiers that vary in price depending on data use.
Metered data has a bad reputation with many Internet users. Nobody wants to calculate how much streaming a movie or downloading a new video game costs in data charges. But in fact, metered data is good for most consumers and for the Internet. Here’s why.
Broadband networks are composed almost entirely of fixed costs—costs that don’t vary very much with usage. Cable companies have to spend many billions of dollars to build and maintain their networks whether or not we use them. One way or another, users of the network have to collectively pay those billions of dollars.
Of course, consumers could divide that cost among themselves in different ways. One way, favored implicitly by everyone who opposes metering or other forms of price discrimination, is for every subscriber household to pay an equal share of the fixed cost. This approach has some appeal. It’s as if we all formed a club and paid equal dues to build the network.
But the “equal dues” approach has a critical flaw: It leaves many households with no Internet access at all. For example, suppose “equal dues” work out to $50 per month, not far from the average cost of broadband today. Some households with limited use for the Internet may only value access—and be able to afford it—at $30 per month. One example is a grandmother on a fixed income who uses email but doesn’t watch online video. Under the equal-dues approach, the grandmother is priced out of the market. She has to go without Internet access altogether.
This isn’t a theoretical point. The Census Bureau’s 2013 American Community Survey Report looks at exactly who is going without Internet access. The 25.6 percent of households that don’t access the Internet at all are overwhelmingly poor, elderly, racial-minority, and low-education. In 2010, the FCC found that 80 percent of those without Internet access subscribe to cable or satellite premium television. The value proposition of current Internet plans just isn’t enough to get them online. If our goal is to get everyone online, we can’t expect everyone to pay the same price.
This means that people have to pay different prices. By charging people varying rates, no household is charged more than its members are willing to pay, and no one is excluded. A self-evident way of doing this, in addition to paying different prices for different speeds, is metering: charging heavy users (who get huge benefits from Internet access) more and light users (who don’t get large benefits) less. Grandma will pay less than $30 per month for her light email usage, and her granddaughter who streams movies all weekend will pay more. Metering, therefore, is a way of expanding broadband Internet access to a wider portion of the population.
We could create a new tax-and-spend program to subsidize access for those who don’t value Internet access highly. But in the long run, there is another reason that relying on metered data instead is good for consumers: It neatly aligns the incentives of cable companies with the desires of consumers.
Think about it: If everyone paid equal prices for unlimited data plans, cable company revenues would be limited by the number of people willing to pay that equal rate. The only way they could increase profit would be to reduce their costs, for example by neglecting maintenance or delaying upgrades. As a result, we would expect the quality of broadband networks to stagnate.
But when users pay for data use, cable companies have an incentive to make it easier than ever to use a lot of data—that is, to invest in speed upgrades. They want you to blow right by your habitual usage amounts, which you will probably do only if you are on a superfast connection. In this way, metered data encourages broadband network upgrades.
Again, this isn’t theoretical. The DOCSIS 3.1 cable modem standard, just now being finalized, will allow downloads over the existing cable network up to 10 Gbps (10 times faster than Google Fiber). Cable companies are now facing a choice as to how fast to roll out support for DOCSIS 3.1. As the theory predicts, Comcast, now experimenting with metering, is planning an aggressive rollout of the new multi-gigabit standard.
To recap, metering allows broadband access to those who otherwise would be priced out of the market, and it aligns the incentives of cable companies with the interests of consumers to invest in broadband networks. So why don’t some people like it?
The answer always seems to come back to “mental accounting costs.” It’s not fun to continually calculate how much you are spending. But we all gladly accept metering for water and electricity with no significant mental accounting costs—why should broadband be so different? Both Comcast and Cox make it easy to track usage. And even if we can’t just get over our mental accounting costs, are they really so significant that we should cite them as an excuse for keeping the poor and elderly offline and letting our broadband networks stagnate?