Clay Shirky used (coined?) the term mental transaction costs to describe the problem with using micropayments (small payments to download articles or music). I believe that economists tend to over-rate the value of peak-load pricing systems, because they fail to take into account mental transactions costs. As Andrew Odlyzko has pointed out, consumers prefer flat-rate pricing even though it costs them more.

In the electricity market, to avoid severe shortages, such as occurred in California two years ago, power companies must induce consumers to reduce their demand at peak times. Often, the reduction is small. However, having electricity rates that vary according to the pressure on capacity imposes costs on the power companies to develop the metering system as well as mental transaction costs on consumers, who would have to spend time monitoring electricity prices.

An alternative, used where I live, is to allow consumers to sign up for a program that allows rationing at peak times. In exchange for a discount on my electric bill, I allow the electric company to shut off my air conditioner for 15 minutes at a time when demand is high. This saves on mental transaction costs, and it accomplishes what otherwise might be accomplished by peak load pricing.

Stephen Kirchner points to this debate about spectrum pricing. Eli Noam argues for a peak-load pricing scheme for spectrum usage. Lawrence Lessig argues that the Internet has avoided over-congestion without peak-load pricing, and that the wireless spectrum could do the same.

For Discussion. For communication networks, what sort of pricing mechanisms would provide incentives against congestion while minimizing mental transaction costs?