I had an e-mail exchange with a friend over what this Wall Street Journal article (subscription required) means for the future of universal service and rural areas where, to date, competition has not arrived. Accepting the Journal's analysis, my answer is as follows:
I have a pedantic answer and a practical one, neither of which is probably very comforting.
The pedantic answer is borrowed from Richard Posner's classic essay 'Taxation by Regulation.' There, Posner asserts that the purpose of regulation is not constraint of monopoly prices - it manifestly fails at that - but redistribution of benefits among different favored classes. In telecom, this explains universal service in all its implicit and explicit forms. The regulatory system redistributes "rents" from urban to rural and from business to residential and from LD users to non-LD users. Posner explains that the logic of this system presupposes that it is closed. No one can be able to escape the regulated universe because then the ability to effect the redistribution falls apart. Likewise, no one can be permitted to arbitrage the system too much or it again falls apart. Thus, regulation must keep all substitutable services within its domain or it cannot accomplish its purposes, in this case universal service.
Into this closed market, technology comes along and allows for regulatory escape. Law, meanwhile, remains static with categories that were invented for earlier iterations of the technology. So, states only have jurisdiction (depending on the state statute) over wireline voice, but usually not wireless voice. Federal jurisdiction cleaves into two worlds, the relatively unregulated information services and the regulated telecommunications services. VoIP enters the picture and stresses these classifications to the breaking point. Thus, both law and our expectations of what regulation can accomplish must change. Universal service will have to become explicit and it will no longer be possible to use the incumbents as a piggy bank through which to effectuate transfers of benefits. The open secret about markets is that they distribute goods and services efficiently, but not equally when unequal costs are involved. Consequently, higher cost regions will have to pay those costs because they will no longer be blended into average cost rates.
The practical answer, to my mind, is much less encouraging, absent significant legal overhaul. If the concern is rural, non-competitive areas, your ability as a regulator to force cross-subsidy through implicit universal service mechanisms is dwindling, perhaps rapidly. Explicit mechanisms - state USF funds, geographically averaged retail rates and provider of last resort requirements - are probably not much further behind. However, the very reasons those regions are less competitive is likely because of these explicit and implicit USF mechanisms in the first place. No one will enter a market where the retail prices are below cost or where the service territory entry requirements force the obligation to serve high cost customers. Thus, if you want it, universal service will have to be accomplished through a broadened contribution base (not that I necessarily endorse this) and a distribution method that decides first what is to be supported, a single voice line or a broadband connection, and second, how that support is apportioned. On the second question, I am partial to auctions if you can get two bidders. I think in the end then the solution is to deregulate retail, let the rates trend toward cost, and then see where and who you need to subsidize. In the interim, with the technological tumult and the legal and regulatory institutions unable to catch up, there will be dislocation to be sure.
In the short run, I do think the Bell companies face huge challenges. Their advantages - an installed base of customers; an installed network base; and a competent labor force - have a downside - customers will migrate to cheaper or more convenient platforms in an instant; the installed network base in a declining cost industry is a liability, not an asset; that competent labor force is high wage union labor. The BOCs are disadvantaged compared to wireless and cable, particularly when cable comes hard with VoIP, because these other platforms don't have the regulatory burdens or the universal service obligations of the incumbents. Fiber in some flavor - to the premises or to the node - is their muti-billion dollar bet to stay alive and it is the nature of network industries that that is often a winner-take-all or at least close-to-all proposition (think Microsoft Windows, which beat our at least a half-dozen credible platforms in the mid-80s).
Finally, I do think the IP-ization of voice does mean the end to state regulation in large measure. The closed cross-subsidy world in which the states operated is slipping away onto multiple and, in some cases, unregulable platforms. There may be some federal grant back to the states of certain powers, but the traditional roles of ratesetting, tariffing and service area definition are over.