Wednesday, June 23, 2010

Explaining the futility of data caps

We've heard it all before, data caps are pointless but it's often very hard to wrap our heads around why. After all, it seems perfectly fair that those who use more of something pay more for what they use. That's how the market for most scarce goods work in our economy. The problem with that arguments lies in the fact that data is not scarce, it's infinite and so scarcity economics do not work properly.

In this article I attempt to demonstrate through analogy why capping data does not make sense when data are not scarce resources - capacity (bandwidth) is.

I've been going over analogies for a while trying to find one that clearly conveys why data limits are silly and if you can exercise your imagination the following example does a pretty good job of it.

The Data Superhighway

Imagine an Internet connection as being a highway that's 4 lanes wide and infinitely long. The cars that drive on this highway also have a width. (For the purposes of this analogy, the cars can be represented as lines, not boxes because the length of the car doesn't matter.) The data on the internet is like the pavement of the highway - every stone, every pebble is a byte of data. Cars travelling this highway always travel in a line, side by side, across the road and drive at the same rate (none are ahead of the others, nor behind).

To help visualize, consider the edge of the highway as "||" and a car as "". At any given time, the highway with 4 users might look like this:


Now, these Internet-travelling cars have a unique property of being able to grow and shrink horizontally depending on how much free space there is on the high way and are only limited by the ISP that controls access to the highway. So if there were only two cars on the same highway as before, it might look like this:


The width of the car is the amount of bandwidth a particular user is using. Over the same amount of time (perhaps 10 minutes), users (cars) with more bandwidth (wider) will be able to access more data (drive over more pebbles) than smaller cars. It makes sense then that without widening the highway (increasing infrastructure) the more people that try to get online at once on the same connection, the less bandwidth each user will have and the less data they will be able to access over the same amount of time.

Time Warner, Verizon, and AT&T all want to place data caps on your usage and charge overage fees. If we apply this concept to our highway model you can easily see that it makes no sense. The problem is our highway isn't wide enough to accommodate everyone with a large bandwidth (wide car), but their solution is not to widen the highway, merely limit the distance you can travel on it without paying more.

Consider the two cars sizes I have above. Let's say the narrower one can access 1Mb every minute of "travel" and the wider one can access 5Mb every minute (because it covers more area over the same period of time). If we have a 20Mb data cap, then the small car can stay online for 20 minutes before the cap is reached, but the larger car can only stay online for 4 minutes before the same cap is reached.

While these are low numbers, it demonstrates that with data caps there is a disincentive to having a larger / wider car on the highway of the Internet because all it buys you is less time accessing the Internet without being overcharged. Why upgrade to fiber or 4G if you can't use it? Might as well stay with dial-up and unlimited data since no one seems to have a problem with that (except for the very real problem in the past of people being online all the time and tying up the modems so others can't connect).

A better solution is to sell minimum widths, not maximum data. I buy a car that is guaranteed to be no less than 5Mbps wide when the highway is full, and when it's empty my car can grow as wide as available to take advantage of the excess unused bandwidth.

Expert Opinions

In case you think I'm just talking crazy like some sort of crazy talking guy about the innovation stifling dangers of data caps, here's what a few other more respected technology experts have to say about the matter.