New York Law School USF Program-- Peter Bluhm speech

About Vermont. First, I'd like to tell you a little about Vermont. Consistent with what you may remember about us from the Bob Newhart Show, we really are a rural state in many respects. We have the highest proportion of people living in small towns in the nation. We also have a high percentage of our population living out in the hills on ten or fifteen acre plots. For those of you who know something about telephones, you know this means long average loop lengths.

Slide 1 About Vermont Most rural state in nation. Defined as highest proportion of people in small towns. High costs from long loops.

Average telephone bills. Another fact of life in Vermont is fairly high telephone bills. There are several reasons -- long loops, lots of poles, shallow soils, and small switches. The FCC reports every year on average telephone bills, and NYNEX Vermont ranked 22nd of 121 large telephone companies in average telephone bills.

NYNEX, U.S. West, MCI, and Sprint have undertaken an enormous task to project forward-looking costs of providing telephone service. As I understand their work, the question they asked was something like this:

How much would it cost to provide a new wire-based network, using existing wire centers and optimum technology?

They performed this study using "census blocks" as their basic unit of measurement. A census block is an area that usually has about 400 to 600 houses. The companies have calculated a cost for each census block in many of the states. They then arrayed these costs by frequency.

Pennsylvania slide. My first slide shows the group's analysis for Pennsylvania. As you can see, there is a peak at about $15 per month and a fairly long tail, out to about $50 per month or even farther. Notice, however, that relatively few census blocks are found above about $30 per month.

Vermont slide. Contrast this to Vermont, which has a bimodal distribution. We still have the peak at about $15, but we have a much broader peak in the $30 to $40 range. And some exchanges exceed $100 per month. Clearly Vermont has an interest in national and state policies that will keep rates low in high cost areas.

Vermont telecommunications history. To understand Vermont's recent work on universal service, you should have a little more background. In 1988, Vermont established a Lifeline program. The benefit, to those who qualify, is a waiver of the $3.50 interstate subscriber line charge. We also added a state benefit amounting to a 50 percent reduction of dial tone charges. The state program was funded by a line charge that started out at about six cents per line. Three years later, in 1991, we used the same mechanism to provide relay services for the hearing impaired.

Slide 2 Vermont Telecommunications History 1988: Lifeline Waiver of $3.50 subscriber line charge. Reduction of about one-half of monthly dial tone cost. 1991: Telecommunications Relay Service (TRS) Both funded through a line charge that began at $.06.

The situation in 1994. By 1994, the line charge had grown substantially. Enrollment in Lifeline was up, and we were collecting 59 cents per month per line. People were growing uncomfortable with the size of the charge and with the method of collection.

Also in 1994 came Enhanced-911, a new program looking for funds. They wanted about $6 million, an amount still worth noticing in Vermont. We proposed legislation that passed the 1994 session.

Slide 3 The Situation in 1994 Line charge had grown to $.59 per month per line. Enhanced-911 was looking for funds on the order of $6 million.

The 1994 legislation. The new bill created an E-911 program in Vermont and assigned its administration to a government board. It also created a "Vermont Universal Service Fund" to finance the three programs: Lifeline, TRS, and E-911. It's called a universal service fund because, with only a little stretching, all the programs are funded to keep the network available to all our citizens. The real point of interest, however, was a High Cost Fund.

The USF revenue source is a surcharge on customer bills. The rate, by the way, is now two percent. The charge applies to all intrastate and interstate telecommunications charges sent to a Vermont address. The principle of the new charge is simple: Those who benefit from the network pay to support universal service.

One significant decision was to appoint an independent fund administrator. This was somewhat foreign to our legislators, who like to have what they consider "tax" money held in the state treasury. We explained to them, however, that this money was designed to replace implicit subsidies now already being paid within the telephone system. We also explained to them that, in the years ahead, they didn't want to be in the position of having to raise a "tax" to support through explicit subsidies what is today supported through implicit subsidies. The argument worked. Some months later after a competitive bid, we selected the National Exchange Carrier's Association.

Slide 4 The 1994 Legislation Created E-911 program in Vermont. Consolidated funding for three programs: lifeline, TRS, and E-911. Revenue source is a surcharge on customer bills. Intrastate and interstate telecommunications bill sent to a Vermont address. Those who benefit from the network pay to support universal service. Independent fund administrator Vermont selected NECA Mandated study on need for a state high-cost fund. Study due January 15, 1996.

PSB study, due 1/15/96. The hard question for Vermont, however, wasn't solved in 1994. That question is what to do about supporting universal service in high cost areas. The legislation mandated the Public Service Board to report, no later than January 1996, on the need for a state High Cost Fund. I will describe for you some of the preliminary conclusions of that study.

Why A High Cost Fund? One might wonder why a High Cost Fund is needed at all. There have been three assumptions underlying our work. First, we assumed that penetration decreases when dial tone increases. This is a traditional economic assumption and is somewhat supported by research.

Second, we assumed that when competition comes on the scene, charges will move toward costs. This will happen in two ways. 1. Existing providers will restructure their rates to eliminate subsidies. This probably looks good for business customers and not so good for residential customers. 2. We anticipate some degree of geographic de-averaging. As competition becomes more of a factor in low cost high density areas, existing carriers will want to reduce their prices. The natural inclination will be to try to raise prices in areas where competition is not so fierce and where costs are higher. This is bad news for rural customers.

Third, we assumed that federal support for rural services won't be adequate. While the current federal effort for High Cost Funds alone is $750 million, for some reason the FCC seems to want to reduce it.

Slide 5 Why a High-Cost Fund? Three assumptions: Penetration decreases when dial tone cost increases. Competition will cause rates to move toward costs. Rate restructuring. More revenue from residential customers. Geographic de-averaging. More revenue from rural areas. Federal support for rural services won't be adequate. Current federal high-cost fund at $750 million for United States. The FCC seems to want to reduce federal support.

Why a High Cost Fund might not be needed. There still are some reasons to think that a High Cost Fund might not be needed. First, the elasticity of demand may not be great. That is, people may still keep their phones even if prices rise significantly.

Second, existing programs may be adequate. Lifeline and Link Up help many people keep phones in the house. Some carriers, most notably AT&T, would place primary reliance on such mechanisms. There also are other steps that can be taken. Some states are allowing local carriers to impose advance credit limits on customers who have credit difficulties. Pennsylvania, for example, has even prevented local exchange carriers from disconnecting because of overdue toll charges.

Third, there may not be any upward geographic averaging. Some local exchange carriers think that they can keep rural rates constant even as urban rates drop. NYNEX is a good example. New York Telephone has entered a seven-year price caps plan with severe limits on dial tone increases.

Finally, technology may come to the rescue. If a customer lives in an area where wireline technology would cost $100 per month, wireless service or satellite service may be the most cost-effective way to keep that customer connected.

One thing that I've learned as we've been preparing our report for the Vermont legislature is that the best way to run a High Cost Fund depends on the telecommunications environment in which it exists. This shouldn't be a surprise, as the High Cost Fund is a patch on the market system. The kind of patch you need depends on the kind of tire you're patching.

Slide 6 Why a High-Cost Fund Might Not Be Needed Low elasticity. People may keep phones, even at high prices. Adequacy of other programs. Life line and link up. AT&T views this as the primary means of protecting universal service. Credit limits. Toll limits. Debit cards. Limits on disconnection for overdue toll charges. Rate restructuring and de-averaging may not happen. NYNEX New York has entered a seven-year price caps plan. Technology to the rescue Wireless. Satellite.

Characteristics of a High Cost program in a monopolistic environment (the old model) Here is how a program would look in the old world of monopoly.

1. One provider, the LEC.
   All LECs are subject to FCC accounting rules.

2. Rates are averaged across wide areas.
   Allows FCC to define a single "study area" that is large and geographically diverse.

3. Benefits distributed based on reported costs and embedded technology.

4. Benefits paid to LEC.
   Benefits can be invisible to customer.

Characteristics of a High Cost program in a competitive environment (the new model) On the other hand, here is how a program might look in a competitive environment.

1. Many providers, using different technologies.
   Not all will be subject to FCC accounting rules.

2. Rates may vary over fairly short distances.

3. Benefit amount based on:
   • Rates charged to consumers.
     The customer decides whether a phone is worth the money.
   • Theoretical cost, using optimum technology.
     Problem of geographic specificity: Costs must be calculated in a small area with uniform costs, but the burden of calculation must be reasonable.
     Optimum technology may be wireless.

4. Benefits may follow customer.
   Virtual voucher.

Benefits of a good High Cost program If we do it right, there should be many benefits to a High Cost program. First, consumers will like it because it will allow everyone to have a phone at home, and nearly everyone will be able to reach nearly everyone else.

Second, regulators like it. As they face proposals to increase rates in high cost areas that are now getting implicit subsidies, they can rely on a legislated program to minimize the undesirable social consequences.

Finally, competitors like it. This point is often overlooked. Most visions of local competition see competition occurring only in the most dense areas. Everyone else is just that -- everyone else. But imagine what would happen if an area costs $50 per month to serve, and a virtual voucher of $25 is available to any carrier who provides my telephone. My net cost as a customer will be only $25, and I can continue to afford a phone. And my carrier, whomever it is, gets revenue from my account amounting to $50 per month. In short, a good High Cost program should promote competition in the rural areas that is as vigorous as the competition we all expect in urban areas.

By the way, there also are some risks. Paramount among them is political vulnerability. Unlike existing subsidies within the rate design, a High Cost Fund is an explicit subsidy, and it is therefore subject to political winds of greater force than when the subsidy was hiding inside telephone rates. It's out there for all to see.

This may explain some of the recent proposals and actions of the FCC. The federal High Cost Fund, at $750 million, is less than was estimated when it was created. Yet the FCC seems very interested in reducing its size. It isn't clear how much thought has been given so far as to whether $750 million is too much or too little to maintain universal service in a competitive environment -- a condition well beyond what was imagined when it was created.

Slide 9 Benefits of a Good High-Cost Program Customers like it -- high penetration is maintained. Regulators like it -- rate restructuring need not inspire dread. Competitors like it -- competition can flourish everywhere, not just where there is now cream to skim. Risks of a Good High-Cost Program Subsides are explicit and therefore politically vulnerable. Experience with federal high-cost fund. Process cost is high.

Federal-state cooperation on universal service. It is interesting to imagine how the federal and state governments might work together on maintaining universal service. I see two major requirements. First, both state and federal programs need compatible revenue raising structures. From a state's point of view, this means the overall capacity of the system must be adequate in the face of rate restructuring and geographic de-averaging.

To ensure that USF revenue is fairly collected, Vermont has selected a sales tax type charge proportional to value of services provided over the network. Customers who use the network benefit from the universality of that network, and in Vermont they pay in proportion to their usage. The more they use, the more they benefit, and the more they pay. It is noteworthy that while the current federal program is not financed this way, the FCC has selected similar proportional charges for TRS and for its own fees.

Second, both federal and state programs should have compatible distribution systems. This probably means a significant state role. Whether the old model or the new model for high cost funding is appropriate depends on how far along competition is in each state. State commissions and state legislatures know more about this than the FCC. Also, states may legitimately choose different kinds of solutions. One state may choose a program heavily dependent on income-tested programs. Another state may try to keep all dial tone costs within a set limit. Both approaches are legitimate.

Slide 10 Federal-State Cooperation on Universal Service Compatible revenue-raising structure Overall capacity must be adequate in face of rate restructuring and geographic de-averaging. Vermont has selected a sales-tax type charge, proportional to the value of services provided over the network. Customers get the benefit of universality when they use the network. The more they use, the more they benefit. The FCC has selected similar proportional charges for TRS and for its own fees. Compatible distribution systems. State role in distribution should be significant. Local competition has progressed more in some places than others. States may apply different mixtures of income-targeted approaches and geographic approaches.

Criteria for a Vermont High Cost program. Before I conclude, I want to describe for you the five major criteria we're trying to design into the proposed Vermont High Cost Fund.

a. Complies with existing law
b. Beneficial effect on customers
c. Beneficial effect on industry
d. Simple to administer
e. Accountable for results

Complies with existing law. To comply with the law, our program must be funded through a surcharge on customer bills. It also must distribute funds to support "basic service" as defined in our statute. By the way, this may need to be amended.

Beneficial effect on customers. First, it should be geographically neutral. It should actually work to keep phone service affordable in all parts of the state. Implicit in this is that the available funds have to be adequate. Also, it should work beneficially at all income levels. We shouldn't have unacceptable reductions in penetration for any income group. We've thought about using our current penetration statistics, which are fairly high, to define a "non-degradation" standard for the future.

Also, the program probably needs to be of some benefit to nonresidential customers. It won't be a pretty sight if residential customers in high cost areas can get service for $25, but business lines sell for $100 per month.

Finally, we hope the program can be operated in a customer-friendly way. Requiring customers to fill out claim forms should be avoided at all costs.

Slide 12 Beneficial Effect on Customers Geographic neutrality -- no region unable to afford service. Adequate funding implicit. Income neutral -- penetration not unacceptably affected by income levels. Problem for some low-income people who can't afford at any price. What should the standard be? No lower penetration than now? General benefit -- allows both residential and business customers to have service at reasonable rates. Transparent to customer -- no filings or returns.

Beneficial effect on industry. It should be competitively neutral, avoiding any discrimination among classes of competitors. This applies vertically, so that a vertically integrated LEC can compete fairly with a niche reseller. It also should be technologically neutral. As long as they can meet the performance standards, wireless providers should be eligible for benefits.

As I mentioned earlier, another test of success will be whether competition appears everywhere. A program that works will promote competition in the boondocks as well as on Main street.

We also expect the program to maintain good incentives. This applies to incentives for continuing investment. It also avoids creating harmful incentives to manipulate data. The system should use "God-given" data when possible -- like population and topography -- provided the measured data can plausibly predict customer rates for dial tone.

A fair program should also take some account of transition issues. Some companies may be left with stranded investments. Fiscal shocks should be avoided, if possible.

Slide 13 Beneficial Effect on Industry Competitive neutrality -- doesn't favor or disfavor any competitor. Vertical neutrality -- doesn't favor or disfavor vertically integrated competitors. Technological neutrality -- doesn't favor one type of transmission technology over others. Geographic neutrality -- if it's done right, competition will appear everywhere. Good incentives. Creates incentives for investment and productivity. Avoids inputs that can be manipulated. Disregard present location of telco facilities, like switches and cables, if possible. Disregard reported accounting costs, if possible. Rely on "God-given" data when possible (such as population, topography) provided the measured data can plausibly predict customer rates for dial tone. Transition. Careful consideration of compensation for stranded costs. No shocks or windfalls to any participant.

Simple to administer. Data requirements must be reasonable. We might use census blocks, for example, but any smaller areas would present insoluble data collection problems. Larger areas have their own problems also, as they invite competitors to find sub-areas with low cost.

We had an important realization during this study. Vermont has not yet decided how small an area a competitive LEC can serve. A carrier might want to serve, for example, every customer within 500 feet of a fiber trunk. We now understand that our policy about the size of that minimum area depends on how refined we want our High Cost Fund analysis to become. Determining subsidies for small areas requires lots of data and lots of calculation. We disregard this problem only at the risk of pegging the benefit amount based upon a worse case. The result would be wasteful spending.

Slide 14 Simple to Administer Reasonable data requirements. Appropriately sized geographic areas. If too small, the burden of calculation increases. If too large, overlooks cost variations and risks overspending. Recalculation at reasonable frequency. Balance between: Recognizing changed circumstances. Population shifts. Rate design work by LECs.

Accountable for results. We expect the program to work, and we want it to be accountable. This means it should evaluate its own effectiveness. The goals should be to keep penetration high everywhere, to keep it at acceptable levels for low-income people, and to ensure that competition is popping up everywhere. Finally, we want our legislature to retain some role in evaluating the program from year to year and to have the ability to adjust the larger parameters, such as how high the surcharge may go.

Slide 15 Accountable for Results Self-reporting. Designed to evaluate its own effectiveness. Penetration should remain high in all regions. Penetration for low-income folks meets predefined standard. Competition is appearing everywhere, not just in more densely populated areas. Oversight. Allows state legislature opportunity to oversee how well the program is working and to set the larger parameters of program operation.

Thank you for your attention.

Return to the beginning of this paper

www.benton.org/Policy/Uniserv/Conference/bluhm.html
Last updated: 9 December 1996 mkh