If You Wanted to Build a National 100-Mbps Access Network, Could You?

The Federal Communications Commission says it will cost $350 billion to build a single, nationally available broadband access network operating at 100 Mbps and reaching virtually every American. The FCC also says it is studying whether telcos and cable companies should be forced to offer open access to third parties that want access to their networks.

Assuming one believes that both ubiquitous access and 100 Mbps speeds are a desirable thing, and virtually everyone might agree, in principle, that that is a worthy goal, the issue becomes "how to get there."

At some fundamental level, policymakers will have to decide whether they want maximum deployment and innovation in terms of new physical facilities, or mazimum third party access. 

Some will argue this is a false choice. That is possible. There is no way to predict with certainty what will happen if robust open access policies are instituted. 

That would be especially true if cable operators, for the first time in industry history, also were forced to open up their facilities for open access. 

Many will point to mandatory open access policies existing elsewhere in the world, and argue the same sorts of benefits can accrue in the U.S. setting. Some consumer advocates say open access is one reason why Internet service is cheaper and faster in those countries. it's a complicated question to answer, however. 

In most, if not all countries where robust open access rules apply to telcos, the competitive landscape is quite different from that of the United States. Few other countries have ubiquitous cable broadband and telco broadband. 

That might not seem, at first blush, to be much of an issue. It is, and the reason is as simple as pointing out that competitive markets are distinctly different from monopoly markets. Keep in mind that a single provider of very-high-speed access, operating on an open access model, still is a monopoly. There is one network and all comers can pay to use it. 

The issue is that such a provider, or providers, as would be the case in the United States, would not be able to operate as a monopoly, because there no longer is any such thing in the U.S. broadband communications business. 

In most communities, there already exist two fixed broadband access providers in the cable and telephone company. In addition, there are places where a third fixed operator exists, or one or more fixed wireless providers.

Then there are two national satellite broadband providers, Wildblue and HughesNet.

Beyond that, there are four mobile providers with existing or partially-built mobile broadband networks, as well as Clearwire, also in the process of building its own national broadband network.

So here's the problem. Where open access broadband networks are most successful, there is not a ubiquitous cable competitor fighting head to head for customers. Assume for the sake of argument that cable providers, nationally, have about 48 percent share of the fixed market, all telcos collectively have 38 percent, and other providers have the rest. 

Assume away all the issues of changing the business models of the whole industry so that one provider in each locality is charged with building a 100-Mbps access network, and is then free to provide service to all comers, at government-mandated rates.

Assume away the problem of the actual wholesale rate, which was part of the Telecommunications Act of 1996. That Act imposed just such an open access policy on major U.S. telcos.

To simplify what happened in the aftermath, telcos violently disagreed with the wholesale rates, while competitors argued just as vociferously that the mandated rates were too high. At the same time, investment in faster broadband facilities slowed dramatically, for one simple reason. Telcos saw no advantage to investing in expensive new facilities that provided a financial return unappealing to the entities who would have to lend the money.

All of that changed when new rules were written that exempted new fiber-based facilities from the open access requirements. Keep in mind that cable companies still do not have any open access requirements of any sort, and that any new broadband policies might well require them to provide wholesale access as well, and that they might also object to the mandatory wholesale rates. 

But ignore that for the moment. Here's the investment problem. Companies have to raise $350 billion in private capital to build the network. And when they develop their financial projections, they will have to note that the new revenue from building the $350 billion network is based on the incremental difference between what typical customers now pay for broadband access, and what they will pay for 100 Mbps access.

But there are other services on the network, you might point out. That's true. But here's the problem. The new network only replicates voice and video revenue already earned on the existing networks. No smart lender is going to okay huge sums based on replicating existing revenues. They will want to know what new and additional sources of revenue will exist. 

The providers can argue that where consumers now pay $40 a month for single-digit megabits per second of access, they will pay $100 to $200 a month for 100 Mbps access. Then the providers will have to model what percentage of customers will do so. When the number turns out to be quite small, the money will not be raised.

There just aren't all that many customers willing to pay $100 to $200 a month to get 100 Mbps when they can do nicely with 20 Mbps to 40 Mbps for lots less money. Ask people. They will tell you what they'll do.

You might argue that take rates will be very high if people can buy 100 Mbps for $40 a month. And that's correct. The problem is again that $350 billion cannot be raised if the new network has no ability to pay a return, in a reasonable amount of time, on the investment. And at anything like $40 a month, no lenders are going to cooperate. 

But matters actually are more complicated than that, as if that was not a show stopper. Recall that most people who want broadband access already buy it. Recall that cable providers, with their own networks, serve about 48 percent of the customers. 

Ask any cable executive you can find whether they would be willing to stop using their own network and just buy access from the telco. Go ahead. Ask anybody you can find. Let me know when you find anybody that says they will do so. 

But ignore that. Say the local telco is charged with building the 100-Mbps access network, and that somehow lenders are convinced that large numbers of people will buy the more-expensive 100 Mbps service. How many of its own customers, and customers of other providers, will switch to buying the 100-Mbps service? 

Be generous and say 20 percent of all broadband access customers can be convinced to buy the 100-Mbps service. That means about eight percent of the telco's own retail customers will do so. 

Say 20 percent of cable customers desert. That adds another 10 percent of U.S. broadband customers. Then assume 20 percent of all the other customers likewise make the move. That adds another three percent of current broadband customers.

What that all works out to is that about one in five homes or locations the new 100-Mbps network passes will buy the higher-priced access service. So the issue is whether an adequate financial payback can be built on serving one of five locations passed with a single new service.

You might argue there also is voice and video, but the problem is that the existing networks already provide those services. Additional revenue is not created just because the network changes.  

But assume an investment of $2700 per passing to build the network. Assume the 20 percent take rate and $60 a month incremental revenue per customer ($100 a month). 

Based on those assumptions, the network costs $13,500 per customer, since only one in five homes is a buyer. At an incremental $60 a month in revenue, breakeven (even at zero interest cost) is 225 months, or 18.75 years per customer.

Nobody will lend money for a breakeven of 18.75 years, and that is assuming zero interest on borrowed money.

An open-access 100-Mbps network might be a worthy public policy goal. But it is hard to see how money can be raised to build it.