I have been reflecting on our electricity forum this Monday, which really was top notch in terms of the presenters and discussion. That said, I cannot help but think that the electric world could benefit from a better historical understanding of the telecom world to avoid repeating the same mistakes.

Competition came to telecom more quickly and more easily, in part because Moore's Law eroded the natural monopoly characteristics of telecom more rapidly; in part also because the telecom rate structure was so out-of-whack with costs that it induced inefficient entry by competitors into the business segment. Also, electricity is tougher because of the load-balancing and lack of storage characteristics of the network. PURPA, though, did the same thing for electricity that the rate structure did for telecom: subsidized the creation of new "competitors" in the generation market.

In telecom, the natural monopoly or "essential facilities" problem came in the last-mile from the switch to the loop. In electricity, it comes in the transmission grid. The question then for regulators is how you mandate access to the last-mile in telecom; and the grid in electricity. In telecom, we have had an experience with this dating really from the
Computer II inquiry and the Open Network Architecture (ONA) rules from the 1980s. Of course, the unbundling rules from the 96 Act are the most ambitious efforts of regulators to define the rules and terms of access to the last mile. In electricity, there are the Open Access Tariffs and OASIS systems and, lately, the terms of the RTOs and finally Standard Market Design. These efforts are still a good 20 years behind similar telecom efforts.

The structural versus behavioral question also enters the picture. The answer to this question probably depends on a predictive judgment: whether or not multiple platforms are going to emerge. For telecom, the predictive judgment (now coming true) is that multiple platforms will emerge to make all aspects of the network competitive. For electricity, the predictive judgment more likely supports a structural remedy because there will not likely be multiple platforms. Therefore, the long-term costs of a structural solution to the transmission grid will probably not be that much more than a behavioral remedy.

In the end, electricity can learn three things from telecom. First, dis-integration costs are non-trivial and potentially quite large. Indeed, they may swamp the benefits you gain from competition. (I shudder to think of the costs of the current telecom OSS systems and their relative dis-use.) Second, public choice or rentseeking pressures and perils will be extremely high because all terms are up for grabs in the regulatory forum. Accordingly, there will be a great deal of investment in rentseeking, and this will be pure social loss. Third, error costs are high and reliance interests will quickly metastacize .

In all, I think this cautions heading into devising access rules with great trepidation, and an inclination toward incrementalism.