Saturday, March 01, 2003

Real-Time Pricing

One of the consevative-libertarian (or was it libertarian-conservative?) myths is that a panacea for the retail electricity market would have been and is real time pricing. Real time pricing in its purest form would be mean that minute by minute retail electricity rates are changed to reflect the current "market price." Now, let's make a couple of BIG (and likely false) assumptions to simplify this issue. First, let's assume that the retail market (and wholesale, too, but we'll get to that) was in reality a competitive one. I'll get into why that is likely a silly assumption at another time, but let's just stipulate for now that it could be. The implication of that is that the real time price of electricity would be roughly equal to the marginal cost of production. The benefit of this is that price would be a correct signal of cost, and therefore increases in cost (and price) would be accompanied by decreases in the quantity of electricity demanded.

Now, the marginal cost of producing an additional bit of juice would be determined by both 'longer run' and 'shorter run' factors - long run including the price of important inputs of energy production - say, the price of natural gas - and shorter run factors would include the transition from more efficient to less efficient means of energy production as demand increased. That is, with low demand the lowest cost generation technologies are used and with high demand less efficient ones are brought oline. So, the second important, and false, assumption that we need to make is that price fluctuations never actually result from reaching the limits of production capacity. If we allowed this to happen, then even in a competitive market the price would no longer bear any relation to the marginal cost of production.

Even given these two rather silly assumptions, real-time pricing is just a pipe dream. First of all, very few or no residential customers would ever want real-time pricing. No consumer would want to have to keep an eye on the price before deciding whether or not to turn on his/her air conditioner at any give moment. There would be a cost to obtaining and monitoring that information which would just be more than anyone would be willing to deal with. That is, the cost of price uncertainty would far outweigh any gains from being able to exploit lower prices by time-shifting. So, competitive retail suppliers, if they existed, would offer long-term contracts to their customers at set rates, which I am sure most customers would opt for. Such contracts would reflect expected future average costs, not marginal ones, and we'd be right back where we started. Retail rates would effectively be fixed - by contract, not regulatory decree - and wholesalers would be able to game the system as they did previously if conditions were right. The only way to prevent this would be to outlaw such long run contracts which is hardly a libertarian solution. In fact, part of the problem with California deregulation was that long run contracts at the wholesale level were outlawed (with good intentions, but disastrous results), . It is true that outlawing these allowed wholesalers to game the system, but it also follows logically from that criticism that if such contracts weren't outlawed, the retail companies would enter into long run contracts with wholesalers, which would then detach the purchase cost from the production cost in the short term making real-time pricing rather meaningless.

It is also true that incorporating a rough time-varying pricing structure such as peak period pricing might not be so unpopular and would therefore be feasible. And, certain commercial and industrial users might be less resistant to time-varying rates, but only users who were extremely sensitive to energy prices and who were able to time-shift their activities would really be likely to perceive that the benefits of real-time pricing would outweigh the costs of potentially extreme price uncertainty.