Lynne Kiesling
I have taken advantage of the relative liberty and calm of the past three days to work on an overdue paper with a co-author, to whom I owe my share of the work. The question is this: in a physical network that retains some natural monopoly characteristics while some of the rest of the vertical supply chain in the industry is increasingly competitive [gee, she can’t be talking about electricity, can she?], can ownership structure substitute for traditional “natural monopoly” regulation? Specifically, if competing retailers own shares in the wires network, and those shares give them use rights that they can use, sell or rent, will the beneficial tensions of those retailers exercise the “socially optimal” discipline on the capacity use, pricing, and construction in the upstream wires sector?
The more I think of this, the more I think there’s an untapped telecom application for this idea. Most sensible people agree that the idea of open access to essential facilities leaves a bit to be desired (i.e., is crap) for many reasons, including the difficulty of determining the price that competitors should be charged for the use of the incumbent’s physical network. In Illinois, for example, Ameritech/SBC had to accept a payment from competitors equal to the minimum long-run average cost if the network were being used at its optimal scale. I could draw you a bunch of simple graphs to show why this is a stupid idea, why it’s totally incentive *in*compatible, and why SBC rightly complained (although they egregiously and foully exercised their political muscle in late 2003 to rectify the matter). Plus, that LRAC is estimated; it’s not like SBC’s CFO has it tattooed on the back of the neck! If it’s estimated through a political process, then the whole panoply of rent seeking kabuki players come out onto the stage, and ya gotta wonder whether or not the LRAC the ICC uses is really LRAC*! I’m thinkin’ … not!
The competitive joint venture idea (CJV) would instead involve this: the FCC (or is it the state PUCs who have jurisdiction at that level? It’s such a muddle I can’t tell) holds an auction for the use rights of the wires network and switching capability constructed by the incumbent. Competing retailers bid for those use rights, using some well-thought-out design for the auction. The proceeds go to the incumbent, so there will be no sniveling or whining about so-called “stranded costs”. From that point on, the local wires+switching is a CJV, with periodic (annual?) auctions for the use rights for the following year. Or, you could just do one auction at the start, for the right to be a member of the CJV in which the use rights would be determined by the retail market share of each of the CJV members. No entry/exit barriers, no restrictions on the ability to build new wires (except for the usual construction and local zoning junk).
It sounds like a reasonable alternative to open access to me, and certainly worth discussion, scrutiny, and testing (experimental, of course). I’m going to hold off on more whimsical speculation until I’ve thought further through this paper and have a version ready to go up on SSRN or some such thing.
Lynne, where power distribution diverges from telecom is that in telecom, the cost per bit — the fundamental of the business — is going down because of technological breakthroughs. In power distribution, no such analogous breakthroughs have happened; we’re still reliant as ever on good old-fashioned copper, and so transmission density doesn’t change once you have the rights-of-way secured. Even if you own the ROW and want to add more capacity, that’s more copper and possibly more towers and therefore more land, unlike telecom where tunneling for the conduit is the expensive bit. Of course, all this is subject to things like the creation of room-temperature superconductors, but I don’t see that changing any time soon.
If the industry in question is a natural monopoly, or at least has a component which acts as a natural monopoly, I am skeptical that an auction would be an efficient means of allocating use rights. Since your example didn’t specify the format of the auction I am going to assume that bidders can not only chose their price but also the amount of bandwidth they would like to bid on. Given that winning the entire network would result in the ability to exercise monopoly pricing power, each agent would value the entire network over the sum of its parts. Assuming this relationship is true for all agents, the one with the highest value would win the entire physical network and proceed to maximize profits as a monopoly.
One might argue that diminishing value for each agent would result in an outcome where many agents win a small piece of the physical network but this relies on the fact that value diminishes quickly enough and over a long enough period, a large network, to erode all benefit one may capture from being able to act as a monopoly. But even then the agents themselves are not static over time and the value proposition of monopoly power will incent agents to combine through various means until their value of the physical network results in a monopoly bid. Without the credible threat of regulation, agents will price the economic value of monopoly power into their valuation of the asset.
Still, this may work in telecom where the price of “building in” as it is referred to in another natural monopoly network industry, railroads, is an affordable option. Like the railroads, the largest factor in restraining monopoly power is likely not the threat of competition from another physical network but actual competition from another technology altogether, namely the wireless networks. The prospects for an unregulated competitive joint venture are far worse in electricity markets with the high cost of “building in” and no relevant competing technology to keep prices down.
David,
Ooooh, I’m gonna have to talk to my colleagues who teach IO about why they don’t teach about Demsetz (1968) — franchise bidding for monopoly rights.
A good overview is available in this outline from Steve Hackett at Humboldt State.
I also think that the contestability and “building in” cost problem is not as dire in electricity as is commonly believed, for precisely the reasons you articulate. Technological change in distributed generation can increasingly make distribution contestable. Right now the utilities have no incentive to allow, let alone embrace, DG even though it offers a lot of reliability and risk management benefits. But with CJV ownership they might change their value proposition to include installation and servicing of DG facilities.