"Who Pays for Universal Service"

by Crandall-Waverman's

Reviewed By Chip Shooshan and Arturo Briceno

We should be clear upfront that, in our case, Crandall and Waverman are preaching to the converted. Their book provides an excellent cost-benefit analysis of the “universal service” system in the United States, and analyzes several myths and realities about what has become, at least in the long run, an unsustainable system.

As they point out, the basic characteristic of universal service is that residential rates have been set below costs in order to increase penetration of the telephone service. Consumers (in most cases the same consumers who are availing themselves of the subsidized local telephone rates) are paying for the shortfall every time they make a long-distance call. Crandall and Waverman argue that this cross-subsidization is undesirable because it is inefficient from an economic standpoint and regressive in terms of income distribution.

The economic losses for the society arising from inefficient pricing have been long documented in the economic literature in general and in the telecommunication literature in particular. It is well known from economics that efficiency losses are minimized by assigning higher mark-ups to those services for which demand is relatively inelastic and lower mark-ups to more price sensitive services. However, the U.S. telecommunication pricing system—as well as the systems in most developed countries around the world—has been constructed by regulators (and, frankly, by the incumbent providers)—in precisely the opposite way.

An important assertion the authors make is that pricing the residential “connection” (what our British friends refer to as the “line rental”) below cost has a limited effect on the number of households that choose to subscribe to the network, given what they believe to be the low elasticity of demand. The probable effect of such a policy is actually to increase the amount of network usage for all the households. This may explain, for instance, why we have higher telephone usage in the U.S. than in European countries. They are undoubtedly correct when one views the effects today, but it is less clear they are correct if one applies a more historical perspective. More on this point later.

The authors also note that the income distribution effects of the universal service policy are such that they are largely income transfers from some poor households to some rich ones and from heavy long-distance users to light long-distance users. The regressive nature of the universal service system in U.S. is based on the following empirical evidence:

The local telephone bill accounts for a small fraction of disposable income, even in low-income households.

The long-distance bill is a large portion of the telephone expenditures.

There is a large variance in the expenditure on long-distance services, varying among and within income groups.

The distribution of long-distance calling is extremely skewed in every category of income. For instance: 80 percent of households in the lowest income group spend less than average national figure of $25 a month, while 50 percent of households in the highest income group spend less than the average national figure of $25 a month.

Second lines are mainly bought by higher income households.

These are largely unassailable facts and have been borne out by other studies over the years. Crandall and Waverman combine these effects to mount their scathing indictment of the current system.

They also make a very interesting comparison with “universal service” in other regulated sectors. They review regulated utility and non-utility services delivered to U.S. households through networks that share some of the characteristics of telephone networks such as oil, gas, and electricity. Their basic message is that, although these services share similar characteristics with telephone services, only the latter have been subject to a distorted price structure brought about by regulatory intervention. They show that universal service for other services has been provided by market forces rather than by regulators. The authors advocate consistent treatment for similar services—in other words, reduce or eliminate the regulatory distortion of prices.

Crandall and Waverman are compelled to answer the question: if what we observe is so manifestly inefficient and unfair, why and how does it endure? Shouldn’t the weight of all the economic evidence they muster be sufficient to bring this “house of cards” crashing down?

The authors contend that the system survives largely intact because of what they call the“median-voter” hypothesis in representative democracies. That is, regulators cannot remain in office if their policies are sufficiently unpopular with a large share of voters; in particular if more than half (or we would suggest, even if a very vocal minority) of voters would suffer from a change in policy. Because a cost-based rebalancing of rates—raising basic exchange rates and decreasing long-distance charges—may benefit fewer than 50 percent of voters, Crandall and Waverman posit that the rebalancing cannot occur.

Another way of looking at it, it occurs to us, is that, having created what amounts to an entitlement program about which a very vocal (and what is perceived to be politically active) segment of the population (the elderly, the poor and the rural) cares intensely, critics find it difficult to rally the countervailing support to dismantle it. This is the thesis developed persuasively by Jonathan Rauch in his must-read book from several years ago entitled, Demosclerosis: The Silent Killer of American Government.

In the face of this largely pessimistic assessment of the possibility of reform, the authors marshal their best evidence. They estimate the welfare gains from implementing a rebalancing program of telecommunication charges by which the residential rates are increased and long-distance rates are reduced. Their basic results are as expected: the losses arising from higher local rates are more than offset by the gains from the reduction of long-distance rates. In absolute values, they estimate the net welfare gains would be in the range of $5.5 to $7.0 billion a year, depending on the incremental cost assumed for the long-distance service. The problem with economic evidence of this kind is that it really is not an effective flag around which to rally the “victims” of the policy the authors have indicted.

Their frustration with those who “lack eyes to see” mounts as they criticize the universal service “policy” in the 1996 Telecommunications Act. Because the Act does not call for rate rebalancing, it is deemed inadequate. Furthermore, the authors maintain that the FCC has failed in implementing welfare-enhancing pricing policies in the past, that its current universal service policy is still not in place, and worst of all, once in place, its policy is not likely to solve the problems of distorted pricing: the high cost policy only converts an inefficient implicit subsidy program to a similarly inefficient direct subsidy program with the result that new entrants may collect some of the benefits in the rural areas. They suggest that the FCC can start moving in the right direction by adopting a simple rebalancing program as suggested by two FCC economists (Rogerson and Kwerel): (1) increase the subscriber line charge or the per subscriber interexchange carrier charge inversely with the broad measures of population density, and (2) reduce long-distance rates.

Crandall and Waverman join a growing chorus of voices (to be sure again mostly economists) who have raised concerns about the new universal service embedded into the Telecommunications Act which includes Internet access for school, libraries, rural hospitals, etc. They suggest two threshold questions be answered before the definition of universal service is expanded: (1) Whether a subsidy is required for access to these services to encourage efficient exploitation of externalities, and (2) What the costs are of the subsidy and how those costs would be collected. The authors argue that, of services that rely on the Internet, only email displays some of the traditional externalities. If there are to be new, expanded subsidies, they should be generated in more competitively neutral equitable, and efficient ways by placing them directly in the budgets of educational and health services agencies.

Despite our agreement with much of what Crandall and Waverman have to say about universal service, we have tried to consider how the book would be read by those they are arguably trying to persuade: policymakers who are not economists and who are constrained by political realities. This audience will likely find the book far less satisfying.

In the first place, they approach the issue with a certain 20-20 hindsight. We should remind ourselves that as recently as the end of World War II, less than 50 percent of households were connected to the telephone network. It took a variety of subsidies (including low-interest rural loan guarantees that most economists would also abhor) to get us where we are today; that is, to the point where we can even be having a debate about how to preserve universal service and whether to expand it.

On a related issue, the authors seem to have missed something big that has been going on in the U.S. relative to the rest of the world—the explosive growth of the Internet. Internet penetration and usage in U.S. is greater than in other OECD countries. Some have suggested that this is due to the same “inefficient” pricing structure that Crandall and Waverman criticize. For instance, it seems that relatively cheap, unlimited flat rate local calling plans in the U.S. have led to the growth of the Internet economy and e-commerce which far exceeds that in Europe, Japan and Australia for example. It is not clear whether stimulation can be attributed to the below-cost prices or the unlimited calling or both. But it is a topic worthy of consideration and certainly must be factored into any analysis of the welfare gains associated with repricing local and long-distance services.

The effect of rebalancing on the growth of the Internet also deserves attention. Here we note that a recent OECD study indicates that the trend in the pricing of Internet access in countries with metered local telecommunication charges is towards the separation of local telephone service pricing from Internet access pricing, through discount schemes, free Internet access, etc. These steps would undoubtedly strike Crandall and Waverman as “bad economics,” but perhaps that is because the scope of their economic analysis is not sufficiently broad; that is, it does not take into account the “macroeconomic” effects of widely available, heavily used Internet access.

Finally, their policy prescriptions are thin—taking up just over a page at the end of a 172-page volume. It is telling that they preface their recommendations with the phrase “like most economists, we prefer.…” But that is just the point. Most decision-makers are not economists and have more wide-ranging goals than achieving economic efficiency. In a perfect world (as defined by most economists including those here at SPR!), economists would rule; but we live in a decidedly imperfect world. So, what are the “second best” solutions? We agree that rebalancing should occur; indeed, that is the logical result of our grand experiment in the U.S. with replacing end-to-end monopoly with end-to-end competition. But how do we get from here to there?

One last note. Ironically, Crandall and Waverman’s prescription may be far more applicable to developing countries, many of which are just starting to introduce universal service support following privatization and liberalization. Many of these countries are implementing universal service systems with fewer economic distortions than in the U.S. and other developed countries. When penetration level is low, the focus can be on universal access, a narrower concept to be sure than the traditional universal service concept.

An increasing number of countries are relying on market-based mechanisms to achieve universal access. For instance, in Chile, Colombia, Guatemala and Peru, the provision of universal service is accomplished through the following mechanisms: (1) The government collects the funds for the universal service either through the general budget or from specific taxes on telecommunication revenues. (2) The government chooses low-income rural areas without telephone service, which are going to be served by the universal service fund. (3) The universal service provider is committed to install and run at least one public telephone in a small group of towns. (4) There is a single-round, sealed-bid auction, where all the bidders make their bids for the lowest per-line subsidy. This market-based mechanism is superior to the cross-subsidy approach taken in the U.S. since it avoids distorting relative prices. And the basket of services is narrower than in the case of the U.S., even though in the most recent auctions, some developing countries are enlarging the basket by including services such as Internet access. While less ambitious than the U.S. model (can you imagine FCC Chairman Kennard defining “success” as a payphone on every Indian reservation?!), this is an approach imbued with the “right”economics

Back to Book Reviews Main Page