Verizon is selling 5 million access lines to Frontier. I expect the deal will go through — after all, a dominant carrier is getting smaller, there is no place where VZ and Frontier compete, etc., etc. What makes the deal interesting is what it tells us about the problem of relying on ILEC/Cable competition to drive broadband. Briefly, (a) we will be perpetually without fiber in a lot of places if we are going to wait for cable and ILECs to meet our needs; and (b) the real problem for is not just the high cost of deployment, but the need to show high rates of return to keep Wall St. happy. It is this latter that will keep telecom policy a very unhappy and complicated place unless we get out of our usual silos and start thinking about some holistic solutions.
More below . . . .
Art Brodsky covers the first point pretty well, but it bears some repeating and enhancement. We start with the fact that VZ is the most aggressive of the ILECs of doing fiber-to-the-home (FTTH). But their solution to “how do you serve the high cost/lower rate of return regions” is “we don’t.” Not only is VZ shedding rural customers, but they also stated on the investor conference call that they expected to reach 17 million of their remaining 27 million customers by 2010 and then afterward would only deploy fiber where improvements in technology drive down deployment cost. That is to say, their massive investment in fiber ends next year, leaving 10 million VZ customers without for the foreseeable future.
And again, VZ is the leader among the big broadband access players in fiber, and they are calling it quits with 10 million customers stranded on legacy DSL and after dumping millions of other customers.
Don’t hate VZ. They are just obeying the dictates of the market — which brings us to the second point. It is imperative to understand this isn’t just about the one-time investment in fiber, but about the potential rate of return on investment. And this is where ILEC stock takes a beating compared to cable stock. The major cable cos — TW, Comcast, and so on — avoided the high-cost areas/low return areas. They were able to do this because of the way local franchising works. To serve a market like NYC, Verizon needs to serve the entire state. But Cablevision and Time Warner only apply for local franchises. True, they may have some neighborhoods in NYC that they’d rather avoid — but even those are reasonably profitable because video is a very good business and cable’s core business is still video. (There are lots of rural cable cos, that’s why organizations such as the American Cable Association exist. Because they have such a radically different economic structure from the major cable ops, they need their own lobbying org.)
The other problem for ILECs servicing rural areas is that rural populations are generally poorer and less likely to buy the high-profit additional services where companies make their money. I don’t mean plain vanilla broadband connectivity. With all due respect to the good folks at PEW, I agree with the rural providers that plenty of rural folks want high-speed internet access. But, on average, rural customers have less disposable income and therefore will buy fewer premium video packages and fewer fancy bells and whistles on their telecom packages. Again, that means lower rate of return despite having higher cost of deployment. The business is profitable — just not as profitable as in the denser population regions.
The result from a Wall St. perspective is that the ILEC rate of return is diluted as compared to the cable rate of return (for consumer subs). That makes cable stock preferable to ILEC stock. Since management gets a honking huge chunk of its compensation from stock, this drag on their values makes them extremely unhappy. The logical result is that publicly traded ILECs and cable cos will avoid build out of fiber (or fiber equivalent) where doing so would dilute earnings. So Hello! to rich suburbs and Goodbye! to poorer suburbs and truly rural areas.
We can repeat the same process in microcosm for more rural companies like Frontier, Embarq, and Fairpoint, and with rural wireless providers. They serve areas that would dilute value for an ILEC, but still provide reasonable rate of return (particularly after getting rid of union jobs and benefits because, lets face it, if it’s good for real people Wall St. is gonna hate it). But get outside the denser areas to areas with lower population density (and therefore higher deployment and maintenance cost) and poorer populations (who therefore provide a lower rate of return) and you have the same problem. You reach a point where these companies won’t deploy because even if they could make a profit in these areas, they cannot make enough profit to justify operation on an ongoing basis. Trying to operate in these areas dilutes revenue as a whole, and they therefore avoid them.
Policy Take Aways
The critical lesson we learn from Frontier/VZ is that it is not enough to look at cost of deployment as we formulate a national broadband plan. We need to consider ongoing rate of return and how the urban/rural disparity on rate of return will drive deployment in the absence of regulatory intervention. This exercise gives us two policy takeaways.
1) The nature of ILEC v. incumbent cable competition. Again, the difference in regulatory structure confers a significant competitive advantage to cable. If the ILECS had been smart, they would have insisted that everyone get a state franchise and have to serve the entire state. Instead, they went for relief from red lining to cherry pick the areas with better rates of return. But that favors cable, who are already in the highest return areas without obligation to serve the lower rate of return areas outside their franchise areas. Right now, ILECs make up the difference with their lock on the enterprise market and, for certain more rural carriers, inflated special access and high termination rates for calls. Note that this creates the unpleasant dynamic that a positive development in one competitive area (increasing competition in special access) will re-enforce dominance by cable in their existing markets (retail video and broadband access) by weakening their biggest competitors.
That does not mean we need to cave to AT&T and VZ on special access because if we don’t let them exact monopoly rents then cable will eat their lunch. Variations on this have dominated telecom policy since Bill Kennard was FCC Chair and he gave cable everything it wanted so that they could pump up to take on the Bells in local phone competition. Kevin Martin tried it with his special access/USF reform package, which would have ensured AT&T and VZ sufficient rates of return to remain competitive with cable. At its heart, this argument rests on the fallacy that the only way to protect consumers is with duopoly competition, and we maintain duopoly competition in markets we care about (Kennard voice, Martin video) by letting them extract monopoly rents or otherwise benefit from regulatory arbitrage that gives them an advantage. Instead of creating competition, it ends up creating pockets of monopoly control with limited benefits in the most profitable markets and a general screwing of consumers everywhere else.
Rather, this should warn us that as we move into our national broadband plan, we need to totally rethink all this from the beginning. We need to consider how these markets interrelate and where we can rely on market forces and where we are going to need direct regulatory intervention to achieve our (or at least my) public policy goal of ubiquitous affordable high-speed access for everyone. At a minimum, it bears out the wisdom of trying to do special access reform, USF reform, and other painful reforms all at once.
2) We cannot solve our broadband problems with a one-time capital infusion. The problem is not just the cost of build out, but the ongoing rate of return. We either need entities such as co-ops or munies (or, dare I say it, federal providers) willing to run at extremely low rates of return or even at a loss. Or we need to accept that we need a permanent subsidy model. We did this back in the old AT&T days by regulating rates and having more profitable areas subsidize rural areas (and even then we had places the AT&T monopoly refused to serve).
Barring that, we need some kind of clever alternative or we just accept that something like 5% of our population will remain perpetually unserved and another 5% or so get stuck with crap. For example, do we focus on anchor institutions as hubs for wireless? We already pull fiber to schools and libraries via E-Rate. But E-Rate also prohibits schools and libraries from sharing this fiber in any kind of rational way. Do we solve the problem (at least as a first step) by removing this and other artificial restrictions on sharing fiber and other capacity?
Conclusion
Once again, the problem is not a competition problem. The market is functioning perfectly and the decision makers are utterly rational. The result of leaving this to the market, even with rigorous antitrust and things like structural separation and mandatory unbundling to promote competitive entry, we will have a segment of our population that will remain unserved or have only a single provider (what we used to call in regulatory speak “stranded customers”). While these competitive reforms are important, we must abandon the idea that they are enough. If we want to provide meaningful broadband access to all Americans, we cannot rely on competition policy alone. Nor can we rely on a single infusion of capital to bring down deployment costs. We need creative solutions that engage the non-commercial sector and government entities at every level, combined with a realization that we are dealing with a hideously complex market structure where problems are frequently interconnected and that eliminating “regulatory arbitrage” in one area will – if no other action is taken — reenforce market dominance in another. As much as we have treasured the fallacy that private sector competition could do it all, possibly assisted by some competition-enhancing regulation, we need to give up our preconceived frameworks and start asking one question: What will actually work?
Stay tuned . . .