Incentives Matter: How to Price Internet Service

Lynne Kiesling

Yes, incentives do matter … Brian Micklethwait finds that how he pays for Internet access changes how he uses it. In transition between service providers, he finds himself with a slow, pay-for-time connection, which has changed how he uses the Internet:

Suddenly, I find myself back in the world of “the news”, in the form of a pre-packaged collection of summarised reports and comments, where all that I learn about each item is the little that the journalists have space in their newspaper or time in their TV slot to tell me. I have stopped being the actively curious individual that a permanent internet connection enabled me to be, and have reverted back to being a passive news consumer. I am no longer in continuous charge of my thinking about the world and its latest doings and sayings. I can only choose from among a few packages. After that choice is made, I am at the mercy of the packagers I have chosen.

Brian is experiencing two changes: one of service quality (dialup instead of broadband) and one of pricing (per minute instead of per month). His experience is important as we consider different proposals to price Internet access differently. I have not thought very deeply about the different proposals for volumetric Internet pricing (by time, by activity, etc.), but Brian’s experience makes me think about how pricing would change how we use the Internet. What would the effect be of charging more for video downloads and other high-bandwith uses? Is such pricing enforceable, or are digital bits sufficiently homogeneous that we would be able to circumvent identification that would differentially price different types of streams?

Here’s an interesting news article that discusses pricing in the context of the proposed AT&T/BellSouth merger.


5 thoughts on “Incentives Matter: How to Price Internet Service

  1. For yours, some friends of ours in St. Petersburg, Russia have been paying for metered bandwidth — roubles per MB of transferred data. (I don’t know the delivery technology they use).

    As opposed to paying for bandwidth directly (even an “unlimited” amount), a per-second rate structure incentivizes consumers to maximize bandwith usage every single second. This may or may be cost effective for a network owner, depending on the size of his pipe and the structure of price regulation…

  2. For yours, some friends of ours in St. Petersburg, Russia have been paying for metered bandwidth — roubles per MB of transferred data. (I don’t know the delivery technology they use).

    As opposed to paying for bandwidth directly (even an “unlimited” amount), a per-second rate structure incentivizes consumers to maximize bandwith usage every single second. This may or may be cost effective for a network owner, depending on the size of his pipe and the structure of price regulation…

  3. When I first got an internet connection, through my university, charges were based on kb sent, and then on time connected, and my behaviour was like that of Brian and Kevin’s friends.

    When I paid based on kB I downloaded less. When I paid based on time I found apps that would download lots for me to read offline.

    All-you-can-eat pricing lasts because you don’t have all those mental worries.

  4. The flat rate vs metered pricing issue has implications for real-time pricing of electric power, too.

    I think that most consumers under a fixed rate for power don’t act as if their power consumption is metered in the way Brian describes for his internet use. However, switching to real-time pricing may make the consumer more aware of the constant economic transactions that flow from household uses of electricity. While the consumer may be more “conservation minded” as a result of real time pricing, they may also dislike such pricing if it takes too much of their attention. Consumers tend not to like micro-payments systems, and real time pricing may make use of electric power too much like a constant stream of micro-transactions.

  5. In the context of Brian’s story, imagine how smart people would be if Internet access were practically free? For example, the new municipal networks subsidised by taxes. I doubt his scenario of active vs. passive media consumers extends to lower Internet service price points.

    Perhaps the Internet price effect is on a curve, where it’s affect is primarily seen at higher pricepoints, but is negligible below a certain rate.

    For example, paying $25/month vs. $1/day vs. $.001/minute vs. ‘free’ may minimally impact behaviour since all are rather low cost.

Comments are closed.