Why voice isn't free

Until the economics of the last mile change, voice will not be free - unless local carriers are rechristened as local access providers, and we start paying a flat monthly charge for something called access rather than local voice service.

Last month's Federal Communications Commission ruling on universal service is a vivid example of why voice service isn't going to be free any time soon.

By now we've all heard the rationale for free voice: IP-based communications are more efficient than what we've been using - and now that data is coming on strong and can share the same facilities, it soon will cost almost nothing to deliver a voice call. Technologies such as WDM certainly have brought down the cost of long-haul transport and, when it comes to the long-distance business, the arguments made by the "IP will inherit the earth" camp are largely true.

But it's a different story when it comes to local service. The cost savings of IP are based on delivering services to many people over a single pipe. But in most cases, the communication terminates as a single service to a single customer over a single line that was installed several decades ago - and that line can't support data unless further investment is made in it.

The cost of providing that last-mile connection varies tremendously from one part of the country to another - and, rather than requiring people in high-cost areas to pay higher rates than everyone else, our forefathers established the universal service system to provide subsidies to carriers serving high-cost areas. Today, long-distance customers foot most of the bill for these subsidies. But if universal service and access reform move ahead as expected, that will change. The cost of serving high-cost areas instead likely will be covered through higher local service charges.

But wait, you say, even if the price of providing the local loop hasn't come down as much as the cost of long-distance transport, surely it has come down somewhat. Also, wireless may offer a lower cost alternative in some of these high-cost areas.

All that may be true.

Local carriers, however, write off their costs of investing in the local loop over periods of up to 20 years. Although the FCC has adopted a forward-looking cost model aimed at determining what it should cost to deliver service using today's technology, it continues to adopt "hold harmless" clauses to allow carriers to recover their embedded costs. Based on embedded cost data, subsidies for high-cost areas total about $1.7 billion per year - and that doesn't include the hidden subsidies built into long-distance access charges that will start getting collected through local rates.

What about minimizing voice costs by delivering data over the same line? This is problematic because it requires such a costly upgrade that carriers have not considered discounting voice services that share a data connection. The cost of upgrading a local loop to support DSL runs between $350 and $800, according to TeleChoice.

The economics of alternative methods of delivering voice and data over the last mile aren't any better. The cost of just adding data to a cable TV system averages about $400 to $700 per customer, according to TeleChoice - and an LMDS wireless system costs more than $3000 per customer location just to support data.

Until the economics of the last mile change, voice will not be free - unless local carriers are rechristened as local access providers, and we start paying a flat monthly charge for something called access rather than local voice service. We wouldn't be able to make long-distance calls if we didn't have a last-mile connection, but unless something changes, we soon may find ourselves paying less to call across country than across town.