Retail wheeling: Déjà vu all over again

Jonathan Lesser is senior economist with Green Mountain Power Corp. Previously, he worked as an energy

policy specialist with the Washington State Energy Office in Olympia,

Washington. Dr. Lesser holds a B.S. degree in mathematics from the

University of New Mexico, and an M.A. and Ph.D. in economics from

the University of Washington. Malcolm Ainspan is a rate analyst

with Green Mountain Power. Mr. Ainspan holds a B.A. degree in

economics from Columbia University and is a Ph.D. candidate in economics

at State University of New York-Albany.

The views expressed here are those of the authors and do not necessarily

represent those of Green Mountain Power.

Retail Wheeling: D6ja Vu All Over Again?

Retail wheeling seems unlikely to improve economic efficiency, though it would benefit some large customers with market power. But it is not at all clear whether those benefits would outweigh the costs they engender and even if they did whether such a redistribution of wealth would be acceptable.

Jonathan A. Lesser and Malcolm D. Ainspan

A sk an economist what he thinks of competition, and

you are likely to hear that "More competition is better." The man-

tra for retail wheeling has been: more competition to supply elec-

tricity and more competition to sell it. It sounds like an econo-

mist's dream, but it may be a nightmare in disguise.

Retail wheeling reminds us of the externalities debates that have appeared previously on these and other pages. 1 Everyone knows

what externalities are. It's merely a matter of dealing with them in

the right wa~ Unfortunatel~ de- veloping a cogent policy on externalities has proven to be no sim-

ple task. The retail wheeling debate, like externalities, seems to be suffering from a desire to "do something" - - or not do it - - be-

fore first stepping back and in- quiring just what the associated

policy goals are. "Read~ fire, aim" is a dangerous way to ap-

proach any complex subject. Pre- cisely because retail wheeling is a complex subject that now lacks clear regulatory direction, it is difficult to provide either general ap- proval or condemnation of the concept.

The purpose of this article is to

discuss some of the major issues associated with retail wheeling

and their implications, in hopes of

34 The Electricity Journal

providing decision makers a bet-

ter framework with which to evaluate proposed policies.

Evaluation of retail wheeling has been hampered in several ways. First, we lack a clear defini- tion of just what it is. Defining it

by analogy (e.g., people shopping for electricity the way they shop for groceries, etc.) merely trivial- izes the subject. In the abstract, retail wheeling may be thought of as a form of shopping, although we are concerned that, like the classic "bait and switch," the

product purchased may not be the same one that's advertised.

S econd, evaluating the benefits and the costs of retail

wheeling requires answering at least one basic question: Com- pared to what? To attempt a comparison of some sort, we can again return to a framework for the environmental externalities debates. Those debates can be framed in terms of two recurrent, though unfortunately often un- stated paths: economic efficiency and equity2; i.e., how will retail wheeling change overall welfare, and will those changes in welfare be distributed "fairly"?

Using these two broad catego- ries, retail wheeling may be evalu- ated in two different ways. We can define policy goals (e.g., greater economic efficienc~ less cross subsidization, etc.) and then determine whether a retail wheeling framework exists that meets

those goals. Alternatively, we can

analyze specific retail wheeling frameworks and then assess their performance under desired effi-

ciency and equity criteria. Be-

cause of the relatively undefined nature of retail wheeling at this time, we believe a better under-

standing of the potential impacts of retail wheeling can be gained using the first approach.

I. Goals of Retail Wheeling

Without a clear statement of

what goals retail wheeling is to achieve, it is difficult to design a sound retail wheeling polic~ In this section, we address efficiency and equity goals, and what criteria must be satisfied if retail

The retail wheeling debate, like externalities, seems to be suffer- ingfrom a desire to "do something."

wheeling policies are to meet those goals.

A. Economic Efficiency

If retail wheeling achieves

greater economic efficiency, it must do so in the context of ensur- ing both an efficient, "least-cost" supply and an efficient demand for electrici~. In principle, this requires that all external costs and benefits be fully incorporated into

the supply of electricity and its associated attributes, and that the

price of electricity equal its mar-

ginal social cost. All assessments of the efficiency benefits of retail

wheeling should be judged by

this yardstick. To take an example, one fre-

quently discussed benefit of retail wheeling is that, through the competitive marketplace, it will provide lower-priced electricity to customers, thereby increasing the benefits derived from electricity. This certainly seems to be the benefit in the minds of many in-

dustrial customers, who are most likely to reap the financial re- wards retail wheeling may offer. It is, of course, unlikely that all

customers would see lower prices, especially in light of argu-

ments over existing cross subsidies.

A rguments for retail wheeling imply that the existing

electric industry is riddled with inefficiencies that would be elimi-

nated were retail wheeling to be allowed. This seems doubtful for

several reasons. First, there is already competition in the market

for electricity and its services. This competition exists within the electric industry in the form of self- and co-generation opportuni- ties. It also exists in the form of inter-fuel competition and competition between supply-side and demand-side energy service deliv- er3a Second, capturing the competitive benefits of retail wheeling relies on the assumption of low cost entry and exit. Third, an argument can be made that there remain economies of scale and

scope under the current market structure of the electric industry. 3

Fourth, even in the presence of cross subsidies, there are no com- pelling reasons to believe that

April 1994 35

electric rates are uniformly set too high for those allegedly doing the subsidizing. Those customers may. in fact, be receiving subsidized benefits themselves, as well

as not paying the full social costs associated with electricity con- sumption. Lastly, relationships between utilities in the current mar-

ket structure have provided social benefits in the form of pooled re-

search and development efforts, as well as other cooperative planning efforts. These social benefits may not be captured in a more competitive environment.

T he realization of a fully competitive market structure

and its supposed benefits will be

hampered by two major factors: transmission and distribution

(T&D) systems, and the social costs and benefits associated with

the current industry structure. Ex- isting T&D systems have characteristics of a natural monopol~ 4 Few would likely advocate a return to competing power lines

strung hither and yon because

such a situation was clearly inefficient economicall3a However, if

we accept that T&D is a natural monopoly, then the interests of economic efficiency require regulation of T&D access and prices in

some form. Unfortunately. as we discuss below, it is not clear whether retail wheeling can be compatible with any sort of T&D regulation.

There are also numerous social

costs and benefits associated with the electric industry. As we pre-

viously noted, much has been made of environmental externalities and proper accounting for

them by utility regulators. There are, however, other social costs and benefits arising from the cur-

rent industry structure. For example, social benefits have been derived from research and development (R&D) activities. 5 R&D

has characteristics of a public good. Under the existing industry structure, R&D benefits have been captured due to cooperative utility efforts, most clearly embod- ied in the work of the Electric Power Research Institute. A fully

Unfortunately, it is not clear whether retail wheeling can be compatible with

any sort of T&D regulation.

competitive market may not be able to capture these benefits ade-

quately if there is no clear mecha- nism to do so, owing to an indi-

vidual firm's incentive to underinvest in public goods. An- other major social benefit has been joint efforts (e.g., power pools) among utilities that insure greater generating and transmission system reliability than is possible for individual utilities. 6

There may also be social costs

arising from retail wheeling, in the form of excessive entry into and exit from markets. Too much entry and exit can lead to less sta-

ble markets, greater uncertainty, and reduced economic efficienc3z 7 These costs must also be consid-

ered when evaluations of the net benefits from changes to an indus-

try's structure are contemplated.

l 'f the additional economic effi- .ciency benefits predicted

from retail wheeling are to be real-

ized, they must: (1) counteract real or perceived economic inefficiencies within T&D functions and the service attributes it provides, despite T&D's natural monopoly characteristics; (2) eliminate cross subsidies, which distort relative prices and reduce eco-

nomic efficiency; (3) be greater than the social costs associated

with costly entry into and exit from the utility industry; and (4) be greater than the social costs of reduced research and develop-

ment efforts, and other planning benefits that have been realized under the current industry structure. Ultimately, a retail wheeling system that has as its primary

goal greater economic efficiency can only come about if the net social benefits are positive. Such a

system will require that the benefits to the winners must clearly ex- ceed the costs to the losers. 8

B. Equity

The existing legislative and regulatory system that now gov- erns the generation, pricing, and sale of electricity was clearly not

developed solely on the grounds of maximizing economic effi-

ciency Historically. goals stress-

ing greater equity and "fairness" have played an important part in the provision and pricing of elec-


tricity. That this is so can be seen

from decades of national legislation, from the Rural Electrification

Act and the various statutes establishing "preference" to federal power, 9 to public utility commis-

sion mandates to ensure that rates

set by those commissions are "fair, just, and equitable."

These legislative and regulatory actions brought real benefits to

millions of individuals and many

industries. Whether they con-

tinue to do so, and whether the cost of doing so has been worth-

while, can be debated. What is important is that a retail wheeling

regime will have to also address legitimate concerns about its ef-

fects on equit3a Since it is conceiv- able that retail wheeling will raise

electricity costs for some consumers, such as rural customers with

high service costs, the "value" of their losses must be compared with the value of the benefits an- ticipated for other customers.

U nfortunatel~ equity and effi-

ciency goals are often at log-

gerheads. Providing electric service to Grandma at her little house

in the woods may be quite expen- sive and offer little economic "re-

turn." But, in its wisdom, society

has decided that Grandma is enti-

fled to heat and light at an "afford- able" price, much as she is entitled

to her social security check. Given that decision, society should strive

to operate in a second best mode:

Provide Grandma with electricity at the lowest possible cost.

Second best considerations will

likely result in cross subsidies, which may either be direct, in the

form of higher rates for other elec-

tric customers, or indirect, in the

form of higher taxes so as to provide Grandma with sufficient in-

come to afford her electrici~. In either case, the result will be re-

duced economic efficiency. Retail wheeling may be able to eliminate

those cross subsidies and improve economic efficiency, but in doing so may exacerbate inequity. 1°

Because equity and efficiency goals are often at odds, they must

be balanced. A system that im-

proves economic efficiency but in-

Retail wheeling may eliminate cross subsidies and improve economic efficiency, but in doing so may exacerbate inequity.

creases inequity may be rejected;

so, too, may one that promotes eq-

uity at the expense of economic ef-

ficienc~ (We assume rational deci-

sion makers who will not choose

policies that reduce both effi-

ciency and equity.) Society as a whole, or its agents - - in the form

of legislators, regulators, etc. - - must decide what the balance shall be between efficiency and

equip, recognizing that there is no

uniquely "right" answer. Thus,

just as with environmental exter-

nality debates, balancing efficiency and equity goals is a public

policy issue. As we discuss in the

next section, retail wheeling will

almost surely force decisions on the preferred balance. But with-

out first resolving that issue, it will be impossible to design a sat-

isfactory retail wheeling system that realizes the benefits claimed.

II. Examining the Major Issues in Terms of Efficiency and Equity

Recognizing that retail wheel-

ing will affect efficiency and eq-

uip, we can address many of the issues that have arisen in discus-

sions about designs for a satisfac- tory retail wheeling policy and ex-

amine the effect those issues may have on achieving the two broad

goals. The issues we examine u

are these: • Multiple ownership of trans-

mission systems;

• Quantification of external benefits of transmission systems;

• Allocation of transmission system costs;

• Increasing risk and higher

costs of capital;

• The obligation to serve and

the right to refuse service;

• Self-generation;

• Elimination of integrated re-

source planning;

• Environmental externalities;

• Costs of excessive market entry and exit;

• Loss of R&D benefits; and • Interactions with other regula-

tions.

We can ask how each of these 11

issues will affect economic effi-

ciency and equity. Then, we can

ask whether a retail wheeling policy can be designed to either fully

exploit improvements in effi-

April 1994 37

ciency and equity, or whether all retail wheeling policies will'be in- compatible with those criteria.

A. Multiple Ownership of Transmission Systems

Proponents of retail wheeling often point to the United King- dom, where limited retail wheel-

ing is already in place. Unfortu- natel~ the comparison between

the UK system and the U.S. system is strained. In the UK, there is only one transmission grid and one grid owner. Thus, retail wheeling transactions are "paper" transactions. Power flows the way it always has, with changes made in the accounting of those

power flows. Since there is only one transmission provider, pur-

chasers of power fully compen- sate the costs of transmission. Thus, the question of how to allocate transmission System costs becomes moot. The system in the UK makes sense as a natural monopoly with one owner.

This is not the case in the U.S.

Because there are multiple transmission owners here, retail wheeling could involve changed physical power flows. Since electrons obey the laws of physics rather than contracts, wheeling by one

utility can affect other utilities. While there are ways of mitigat-

ing these physical impacts (e.g., phase shifters), mitigation may be inadequate or extremely costl)~ As a result, transmission externalities may arise. 12 Of course, some- times the external impacts will be beneficial, although far greater attention has been focused on nega- tive impacts.

Economically efficient retail wheeling would be greatly as- sisted by a one-owner transmission grid, or at least regional one- owner grids. That would help move us toward the British "paper" system and avoid some of the transmission issues now faced. Clearl}~ such a transmission system would have to be regulated or government-owned

to prevent aggrandizing actions often favored by monopolies.

From an equity standpoint,

moving to a one-owner system is unlikely, although not impossible. It would require creation of one or more entities (publicly or pri-

vately owned) to purchase existing utility transmission systems,

as well as regulation of the new entities by the Federal Energy Regulatory Commission (FERC). It would also require state regula-

tors to reassess the positions of utilities which have looked to

their transmission investments for a significant share of their earn- ings.

B. Quantification of the Social Benefits Produced by Transmission Systems

Transmission systems have attributes of both private and public goods. Awe]l-run, intercon- nected system provides direct benefits to a utility's customers, for which they can theoretically be charged, and to other utilities

and customers not directly served by the utili~. How can these

benefits (and costs) be allocated efficiently? Is such an allocation equitable?

Under the current system, many utilities are organized into power "pools." A few pools may dictate both generating and trans-

mission system operations, for the mutual benefit of all the mem- bers. Costs are allocated in ways that are "fair," which in practice means in a mutually agreeable manner. 13

Because of the public good characteristics of transmission systems, allocation of benefits and

therefore pricing becomes more difficult. If retail wheeling is

adopted, accounting for the individual products supplied by the transmission system becomes more critical. If a customer con-

tracts with a third-party, for example, rather than the local utility,

both the third party supplier and the customer will want an assur-

ance of reliability so that contrac- tual guarantees can be met. The utility wheeling the power may be willing to supply this reliabil-

ity as long as the cost of doing so is less than the price the utility can charge. However, because


transmission is a natural monopoly, an unregulated utility, like any monopoly firm, would pro-

vide too little reliability and charge too high a price for it.

If utilities can no longer ade- quately recover the costs of main- taining their transmission systems, either because they are more constrained in their ability to recover costs or because of heightened uncertainty about whether there will be a need for additional investment in the future, then the full social benefits of the transmission system may be unrealized. This would lead to economic inefficiency Whether the degree of inefficiency would be greater than under the present system is an empirical question. 14

C. Allocation of Transmission System Costs

Retail wheeling with the exist-

ing ownership system will require a way to allocate transmission sys-

tem costs. Again, the issue be-

comes one of whether the primary goal is to allocate those costs efficiently or equitabl3a

Should Grandma, with her dis- tant and variable demand, pay for

only the embedded costs of providing her with service, requiring that other customers pay higher rates? Or, should she be charged

the full marginal costs for access to and use of the transmission system, even if she will be unable to afford the electricity? There is no

uniquely right answer. Strict economic efficiency would require

her to pay full marginal costs. A utility commission, however, might view such a pricing structure as unfair.

Allocating transmission system costs also faces another major problem: The allocations will be arbitrar~ since transmission systems supply multiple products jointly. A typical textbook exam-

ple, the joint production of meat and hides, provides an analog)a

Suppose you raise cattle. The

Customers will battle ferociously over cost allocation.

marginal cost of raising an indi-

vidual steer is $100. The steer yields $200 worth of meat and

$150 worth of hides. What is the proper allocation of the $100 marginal cost to the production of beef and the production of hides? The answer is that there is no cor- rect answer; the costs cannot be uniquely separated.

U nder the Energy Policy Act, is FERC requires that

utilities choose between embed-

ded cost transmission system allocations and marginal cost allocations when they wheel power in the wholesale market. Since FERC controls all transmission pricing, it is reasonable to con- clude that it will also influence transmission pricing for any future retail wheeling. In the recent Florida Power & Light case, 16 FERC

also said that the utility must justify its transmission system costs. The problem is that there is no right way to allocate those costs.

The natural monopoly characteristics of T&D systems, however, require some form of regulation, if their full efficiency benefits

are to be realized. This means that regulators must devise a way

of allocating T&D costs. The joint nature of the products provided

(e.g., power delivery, reliabili~, voltage support, etc.) means that economic efficiency provides little guidance on such allocations.

Thus, regulators will need to rely on other guidelines. Ones based on equity have often been used. Unfortunately, one of the reasons

for the drive towards retail wheeling is those very cost allocations,

and the cross subsidies they may

April 1994 39

impose on different dasses of customers..

While the FERC order in the

Florida Power & Light case applied only to wholesale transactions, it

is but a short step to applying the network service concept to retail transactions. However, in doing so, the cost allocation issue becomes even more complex. How can utilities plan for future net-

work service requirements if they do not know who among their retail customers will demand it? How can prices for, sa)4 network

services be developed, in light of the joint cost allocation issue?

How will the developers of a retail wheeling policy know they are maximizing the efficiency benefits?

U ~nbundling of transmission services, as has been pro-

posed by some authors, 17 may be

difficult to impose and, owing to

FERC requirements, more difficult still to justify. The public

good characteristics of transmission systems described pre- viousl~ and the joint products they suppl)~ would be difficult to

capture in a fully deregulated market, since utilities would not necessarily have the proper incentives to maintain their systems op- timally. With retailwheeling, wheeling utilities and customers

would want to specify all possible transmission externalities that

could affect delivery of the electricity contracted for, so as to reduce future risks. 18 Such specifications,

however, would be extremely costl)4 and would likely reduce efficiency gains.

D. Increasing Risks and Higher Costs of Capital

Capital markets are generally sensitive to risk. One benefit regulated utilities currently enjoy is access to capital markets at lower costs than many of their unregulated counterparts. By being able to plan for long-term commit- ments and having reasonable as- surance of recovery of investments, utilities provide suppliers

of capital a lower-risk alternative.

Unbundling of transmission services

may be difficult to impose and, owing to FERC requirements,

more difficult still tojustify.

Lower costs of capital provide benefits to electricity purchasers, which can then be flowed through to ultimate customers. In a world where consumers and in- vestors are both risk-averse, lower

costs of capital due to reduced risk will result in increases in (ex-

pected) economic welfare. With retail wheeling, cost recov-

ery by a utility or a third party supplier of electricity is no longer assured. Thus, the risk associated with utility investments may be

perceived as being higher relative to a previously regulated environ-

ment. (The ability to diversify

away this risk will also be reduced.) If cost recovery can no longer be assured, the risk to utilities and independent power pro- ducers will increase, as will the

cost of capital. If the risks of additional investments in both generation and transmission are perceived as high, then there may be too little investment. This could lead to insufficient supplies of

electricity, inadequate transmission system reliabili~, and reduced economic efficiency.

E. The Obligation to Serve

Of all the issues associated with

retail wheeling, none has been more discussed than the obligation to serve. As its name implies, the obligation to serve is funda- mentally an issue of equity that has secondary impacts on economic efficienc~ Historically, utilities have operated their sys-

tems with the expectation that they would serve all customers in

their service territory. This still leaves some risks, since customers

may go out of business, move out of the service territory, or self- or co-generate. These risks, however, are still lower in comparison with a fully deregulated system.

There is little, if an~ disagree-

ment that retail wheeling is incom- patible with an obligation to

serve. A utility cannot be obli- gated to meet the energy needs of

potential customers within its service territory without some guarantee of recovering the costs associated with that obligation. To do otherwise would clearly result in economic inefficiencies. It would also lead to further inequi-


ties, since the costs and risks of meeting a standing obligation would be borne either by the utility's remaining customers or its stockholders.

T he flip side of the obligation to serve has been less fre-

quently discussed, but is also rele- vant in a public policy arena.

This is the right to refuse service: Can a utility refuse to serve a customer within its service territory? 19 Lr~ industries without obvi-

ous capacity constraints, this is not a problem. The grocery store is unlikely to turn me away at the door, refusing to sell me bananas.

In capacity-constrained industries, however, the right to refuse service is far more prevalent. Res- taurants with limited seating capacity, for example, may refuse to seat someone who is shabbily dressed, in favor of seating a customer wearing a suit and tie. The

restaurant owner reasons (cor- rectly or not) that the expected re- turns from his limited seating ca-

pacity will be greater with the nattily dressed customer than with the shabbily dressed one.

Thus, refusals to serve can be viewed as economically efficient.

An electric utility subject to (and quite possibly participating in) retail wheeling faces the same issue. If the utility is unable to plan for additional demand, it will be less likely to increase its

capacity to meet additional demand. Should that additional demand materialize, however, the utility may need to ration it. Eco- nomic efficiency leads to ration- ing by price. (A refusal to serve is effectively an infinite price.) De-

mands for additional wheeled power could lead to demands for transmission system upgrades

and new standby generation to maintain voltage stability. It is not clear how the costs of these investments would be allocated since, even though the investments may be required to meet new obligations to wheel, they may also have beneficial effects for the utility's other customers.

Viewing electricity as an input to a production process, in an un-

It is common in rate cases for customers to testify that any rate increases whatsoever wilt lead to economic calamity.

regulated market willingness to pay will be determined by what

economists call the value of the marginal product. That is, at the

margin, the price customers will be willing to pay for a kWh of

electricity will equal the value of the additional production (e.g.,

heat, light, new widgets, etc.) that the marginal k w h provides. With

a rationed suppl~ that price will be bid upward. If a utility had no obligation to serve any customers whatsoever, the existing distribution of electric sales would change. For the typical competitive market, in which, sa~ a gro-

cery store operates, there is noth- ing wrong with this situation. However, given the natural mo-

nopoly characteristics of a utility's T&D system and the long-lived nature of utility investments, it is not clear whether the effects on

equity would be welcomed, much less whether greater economic efficiency would be achieved.

olitics may also affect elimi- P n a t i o n of the obligation to

serve, much as it did in the estab- lishment of such obligations many years ago. It is common in

rate cases for customers to testify that any rate increases whatsoever will lead to economic calamity. The larger the customer, the more likely it is that such arguments will resonate. Thus, imag- ine a situation where a large industrial customer leaves a utility to arrange a third-party long-term source of supply and, after just

one year, this supply is suddenly not delivered. What does this cus-

tomer do? First, it will approach its service territory utility and re-

quest that it again be served, ideally at the same rate it had ar-

ranged with the third-party supplier. The utility may be physically unable to meet the customer's requirements or, since there is no obligation to serve, may wish to exercise its new found monopoly power and charge the customer an economically inefficient high price.

Presumabl~ in order to negoti- ate a better deal, the customer would announce that the high

electric prices will force it to close its operations, resulting in layoffs

of hundreds of workers and gen-

April 1994 41

eral economic calamity in the

area. What happens next depends on one of three responses. The customer's request may be denied, upon which the customer may decide to pay anywa~ or it can indeed close its operations. A third outcome - - in our political system the more likely one - - will allow the large customer to negoti-

ate a special lower cost "deal." The costs of serving the large cus-

tomer will then be spread to the utility's remaining customers.

I n a system without retail wheeling, this situation may

be viewed as equitable. Other utility customers may reason that they are willing to pay higher electric prices in order to maintain a large customer's economic vital-

ity and the local economy. It may even be efficient, if the costs arising from the disruption of a large customer's operations would be

greater than the savings that would result if the customer sim-

ply relocated. With retail wheeling, the situ-

ation changes. Now, if the customer secures a great "deal," the utility's other customers will have a greater incentive to seek other suppliers, since their costs will be "too high." These customers, if not already allowed to participate in retail wheeling, will have

greater incentives to do so. The result will be increased instability, greater uncertaint~ and reduced economic efficiency.

E Self-Generation

One argument for retail wheeling is that improved generating technologies are allowing more

customers to self-generate (or co-

generate), effectively providing competition to utilities. If retail wheeling is allowed, customers may be less likely to self-generate

and utilities may be able to capture all or part of the revenues that would have otherwise been lost. (This argument is also used at times to justify "special" rates

to large customers, especially under the general phrase of "economic development.") Utilities

may also be able to charge for spe-

cial transmission services associated with wheeled power more easil~ again recouping some frac- tion of lost revenues.

Self-generation is a form of direct competition, although not

one without risk. An important benefit a utility can provide poten-

tial self-generators, as well as other utility customers, is diversity of supply. Individual generating units can fail without compro- mising service to customers. Self-generation will generally fail

to provide these diversity benefits, unless the utility can be relied on to provide backup.

Eliminating the obligation to

serve eliminates an external benefit that a self-generator may have enjoyed in the past. The more important the value of reliability is, the greater will be the reduction in external benefits. In the ab- sence of an obligation to serve, unless a utility is able to charge the full cost of providing diversity benefits, it will have no incentive

to provide such benefits. How- ever, if that is the case, it is likely that the benefits from self-generation will be reduced, unless we assume that self-generators are in- herently more efficient than utilities in generation. This seems unlikely.

G. Elimination of Integrated Resource Planning

Retail wheeling is thoroughly in-

compatible with integrated resource planning (IRP). For some, this may be considered a benefit, for others a cost. Retail wheeling

may be consistent with some form of T&D system planning,

but even this would be highly un- certain.

IRP assumes the orderly development of resources to meet fore- cast demands for energy services. Because development of re-

sources is complex and time con- suming, and because of the long-

lived nature of most resources, IRP can, in theor~ simplify the regulatory process, such as in making "used and useful" deter- minations in rate case proceed- ings. With retail wheeling, no in-

dividual firm can plan on meeting the needs of a known set of customers unless it first contracts


with those customers and then supplies needed energy services. This, however, is not what is envisioned by retail wheeling proponents, who may envision, rather,

shopping for better deals. IRP can also assist in making

consistent comparisons of resources. Utility commissions, for example, may wish to ensure that resources are developed which reduce ratepayers' exposure to future risks, such as future taxes on

specific generating resources. With retail wheeling, those risks may increase if suppliers become more focused on short-term transactions. Such increases in risk, as discussed above at subsection D, can lead to higher capital costs and reduced economic welfare.

H. Environmental Externalities

To summarize briefly the externalities debate, there are various environmental costs associated with electricity production, much

as there are environmental costs associated with just about every- thing else. Economic efficiency

requires that these costs be inter- nalized, so that social costs of elec-

tricity production and use equal its social benefits.

As has been discussed previously, if economic efficiency is the primary goal, then there may

be legitimate actions that regulators can take to ensure greater economic efficiency. With retail wheeling and the inherent aban- donment of IRP, mechanisms designed to optimize environmental externalities would have to be un- dertaken at the federal level, as the current system of piecemeal state regulations of this area would be both redundant and det- rimental. 2° Of course, there is no

guarantee that federal action would promote greater economic efficiency than do current state systems. In part, this would depend on whether reduction of ex-

It's all in the coordination.

ternalities was primarily driven by economic efficiency or equity goals.

If piecemeal state legislation were maintained, or if out-of-state

impacts could not be accounted for, then the importation of cheap

and "dirty" power may be exacerbated and prudent avoidance of future environmental risks may be reduced. The results will likely

be inefficient and inequitable. Higher cost, "clean" utilities may find themselves unable to compete in the retail market, and their service territories may absorb additional pollutants from "dirty"

utilities.

A s an example, suppose utilities in New England

lose sales to midwestern utilities,

whose resource portfolios are primarily low-cost coal-fired generation. As more power is generated in the Midwest, more sulfur dioxide would be generated in that region, potentially exacerbating

acid rain damage in New Eng- land. (Even with the Clean Air

Act's sulfur dioxide caps, there

can still be greater acid rain damage in specific areas.) The situation could be further exacerbated

by the "clean" utility's stranded investment: If the utility's rates were forced upwards, remaining customers would have greater in-

centives to seek their own lower- cost power.

Tracking environmental impacts, such as establishing that a specific sale of power from a Mid-

west utility to a New England customer was increasing acid rain damage, would be costly and liti- gious, and fail to contribute eco-

April 1994 43

nomic benefits. Furthermore, it would be relatively simple to develop transactions that effectively disguised the true source of the power. In such cases, even if the

individual state utility commissions in New England could ac- count for sales that had direct impacts on the region, the actual environmental costs imposed would remain unaccounted for.

This would reduce both efficiency and equity.

F ederal legislation could also be circumvented in ways

that are inequitable. One current example concerns requirements under the Endangered Species Act to restore salmon runs in the

Pacific Northwest. If the cost of programs instituted to restore these salmon runs is paid for in

the form of higher electric rates, then customers who are able to shop around may be able to find cheaper electricity, even though for years they may have enjoyed the benefits of artificially low-cost electricity (i.e., electricity that did

not reflect the full social cost of its production) without having to be responsible for the environmental damages. In such a situation,

other ratepayers or taxpayers in general would be forced to pay

these costs. Thus, retail wheeling could adversely affect the alloca-

tion of past expenditures to reduce environmental damages, or expenditures designed to reduce potential future financial risks from changes in environmental policies.

The incorporation of externalities and retail wheeling can be

made compatible only if treat-

ment is equitable across jurisdic- tions and only if externalities accounting is expanded from the current focus on planning activities to actual rate setting and dis- patch decisions.

Ultimate134 interactions between externalities and retail wheeling will be decided in the public pol-

icy arena. Concerns over equity will be debated; such debates will pit the "fairness" of providing (potentially) artificially low-cost

power versus the unfairness of not being allowed to compete on

To achieve the full competitive benefits

of retail wheeling, entry and exit should

ideally be costless.

a "level playing field." If economic efficiency is paramount, then competition between alternative electric suppliers must be

made on an equal basis. While this may still leave issues of un-

regulated fuels unresolved, it can at least make decisions about the choice of electric uutility supplier more consistent. If externalities

and retail wheeling are addressed with equity as the primary goal, then it will be critical to define what the key equity is-

sues to be addressed are, what the effects on equity issues not directly addressed will be, and

what the effects on economic efficiency will be.

I. Costs of Excessive Market Entry and Exit

Economic efficiency will also be affected by the social costs associated with excessive entry and exit into the marketplace for electricity and its services. These social costs for electricity and its services have received less attention in the trade than the social costs associated with environmental impacts. Nevertheless, these costs are critical to determining whether retail wheeling is compatible with

greater economic efficiency. Entry and exit costs fall under

the broad mantle of transaction

costs. To achieve the full competitive benefits of retail wheeling claimed by proponents, entry and

exit should ideally be costless - - i.e., such that firms can enter and leave the market without causing any disruptions to that market. This fundamental assumption un- derlies the entire competitive

model, and for many markets (e.g., restaurants, agriculture) is quite reasonable.

Costless entry and exit do not describe the electric industry, however, for a variety of reasons. 21 In-

stead, the industry has high entry and exit barriers. This is most ob- vious in transmission, owing to its natural monopoly charac-

teristics. And, despite the coming of competition, the generation

market also has high entry costs. Firms wishing to develop and

market new generation will en- counter large costs owing to the expense of generating units them-


selves, siting and permitting proc- esses, and the potential difficulties of negotiating contracts for output. There are also high exit costs. A generator leaving the market can impose high costs on the system if there are not easily avail-

able substitutes, and if that exit affects the reliability of the transmission system.

Entry into a market can also en- courage inefficient investment in entry deterrence and reputation building by existing firms. Utili- ties, for example, may commit additional resources towards retain- ing existing customers that are contemplating retail wheeling, di- verting those resources from standard franchise obligations. If the value of the rest of the utility's system is greater than the value of possibly keeping the customer from leaving, then the expendi- ture will be inefficient. 22 It could

also be inequitable.

T here is no guarantee that costly entry and exit will

lead to the provision of efficient

levels of all of the attributes associated with the supply of electrici~. Proponents of retail wheeling often talk of electricity as a stand-

ard commodity, like wheat. Greater competition, they argue, will lead to more efficient pricing of the commodity. However, be-

cause providing electricity also means providing other attributes

(e.g., reliability, voltage stability), achieving greater economic efficiency would require that efficient levels of all of these attributes

were supplied. This may be especially problematic where some of the attributes provided, such as re-

liability, have public good charac-

teristics. Thus, in a retail wheeling environment, the most likely outcome would be price deregulation and reliability regulation. 23

Whether such an outcome would be stable when there was excessive entry and exit is unknown.

If economic efficiency is the primary goal of retail wheeling, then it will have to provide greater wel-

fare benefits than the welfare losses that will accumulate if en-

try and exit are not costless. Un-

Excessive entry and exit may increase pressure to reduce costs associated with serving "'high cost" customers like Grandma.

fortunatel)~ the retail wheeling ex- perience in Great Britain provides

evidence for just such excessive entry and exit, indicating that

transaction costs may signifi- cantly reduce hoped-for benefits. 24

One result of the inefficiencies

of excessive entry and exit is likely to be consolidation of the electric industry. Utilities with low-cost sources of power, even if those costs are artificially low, may drive out providers of more costly power. More utility mergers will follow as the stronger utilities purchase weaker ones or

independent generators purchase

utilities to secure transmission system ownership. If the already-increasing trend of utility mergers continues, the benefits of a "competitive" marketplace may disap- pear. High entry costs may prevent effective competition. Customers who had previously shopped around for low-cost power may then find themselves unable to do so, because the num-

ber of viable sellers will have de- clined precipitousl)a If this situation occurs, then retail wheeling could actually "turn back the clock" in the industry, returning it to the era when utility holding

companies were broken up into smaller exclusive franchises so as

to prevent monopoly abuses. It is hard to conceive of this situation as representing "progress" for the

electric industry. Lastly, excessive entry and exit

may also affect equity. To the extent that firms that are unable to compete, owing to an inability to capture social benefits, leave the industry, there may be disrup-

tions in service, or greater pressures to reduce costs associated

with serving "high cost" customers like Grandma.

J. Loss of Research and Deve lopment Benefits

Research and development has characteristics of a public good, in that its benefits can be shared by all. Since individual firms may not realize the full social benefits of R&D, they will have a ten- dency to underinvest in it. Thus,

cooperative ventures such as the Electric Power Research Institute

can provide social benefits by dis-

April 1994 45

tributing the costs of R&D invest-

ments and encouraging greater R&D activity than would be fi- nanced by individual utilities. Utilities appear to be increasingly concerned about their R&D expenditures. While information

sharing and joint R&D still exist within the i ndus , , , they appear

to be declining due to competitive pressures. 25 Utilities that are ag-

gressively cutting costs to remain "competitive" are abandoning their memberships in organiza- tions such as EPRI. Given the public good nature of R&D, this is to be expected, although it will

likely result in lost efficiencies.

L arge utilities may still be able to invest in R&D ow-

ing to scale economies. For those utilities, the investment is likely to be recouped. But, in a competitive environment, the benefits of those R&D expenditures will not be shared; indeed there will be an incentive to maintain secrec~ to provide a competitive edge.

R&D is not limited to generation

markets. There may also be useful R&D investments made for im-

provements in T&D efficiency To the extent that small utilities, or

those with higher costs, are unable to invest in R&D, their costs will continue to remain higher and they will be unable to compete. This will lead to more entry and exit of firms from the marketplace, in-

creased transactions costs, and reduced economic efficiency

K. Interactions with Related Legislation

If retail wheeling were intro- duced, its overall efficiency and

equity could be affected by existing federal legislation. PURPA, for example, requires utilities to purchase the output from cogen- erators at a price based on a utility's avoided cost. In an environment where the utility has an obligation to serve, this can be justified, since the energy purchased

will be incorporated into the utility's planning efforts to meet future obligations. If those obligations were removed, however,

there would be no clear method

\: % ',

that would allow the utility to recover the cost of the purchase.

Retail wheeling could also affect federal preference power statutes,

which limit the distribution of power generated at federal sites. Here the issue is not primarily one of efficienc~ but of equity.

With retail wheeling, it would be difficult to prevent "sham" transactions that transferred this low- cost power from its designated re- cipients to those willing to pay

more for it. This might be economically efficient, but it would run counter to the purpose of the legislation.

III. Comparisons with Deregulation in the Telecommunications Industry

Retail wheeling proponents often point to deregulation in the telecommunications industry to

bolster their arguments that retail wheeling will provide greater economic efficiency. And, indeed it is true that the telecommunications industry has undergone radical change, which now allows cus-

tomers a dizzying array of choices and offers the prospect for still more choices in the future. The telecommunications industry is not fully deregulated, however, and differs from the electric utility industry in several important ways. First, technological change continues to remove the natural monopoly characteristics in the "T&D" end of telecommunications. With greater reliance on wireless technologies, transmission costs have fallen greatly This

has also been enhanced by the federal government's not charging

for a scarce resource: the radio spectrum. Second, changes to ca-

pacity can be made far more quickly, thus simplifying the planning aspects. 26

The telecommunications indus-

try is also not immune to some of the problems caused by deregulation, namely stranded investment. What would happen if local tele- phone service were fully deregulated, perhaps due to the development of an extremely low-cost, wireless system that linked to-

gether expanding cellular markets and long-distance services? How

would local phone companies re-


cover their previous investments

in providing subsidized access for all local customers, especially

when many of these investments were legislatively mandated? Un-

doubtedly; regulators will face this question in the near future.

Perhaps the only way around it is to charge fully for the use of the

radio spectrum. Telecommunica- tions are also the subject of numer-

ous equity issues, as well as exter-

nalities. There are social benefits

in providing access to the greatest number of individuals and firms;

such benefits wou ld not be cap-

tured in a fully deregulated sys-

tem. ' f complete deregulation will

.not work given the existing natural monopo ly characteristics inherent in T&D, then a partially

deregulated system is the next op-

tion. This has some appeal, since

generation is largely deregulated

already: Partial regulation, however, is likely to create even more

problems. First, regulators would

have to decide which customers

were eligible for retail wheeling.

This would be done in a conten-

tious atmosphere, with those perceiving the most to gain arguing

most fervently for retail wheeling,

and those perceiving losses argu-

ing against it. Partial regulation would have to determine whether

the obligation to serve still applied to those ineligible for retail

wheeling. It wou ld also have to

address gaming strategies, in

which those ineligible for retail

wheeling "changed themselves"

into entities that were eligible. Partial regulation would also

fail to simplify the planning com-

plexities inherent in the indus~¢.

It could also exacerbate percep- tions of risk and uncertainty, fur-

ther raising capital costs and re-

ducing potential economic gains.

Lastly; it would still be confronted

with stranded investment issues,

and other important equity issues.

IV. C o n c l u s i o n s

Can retail wheeling deliver on

its promise? The answer depends

on what the desired goals are. Is

the sole goal of retail wheeling to

improve economic efficiency? Or,

perhaps, is its purpose to increase

"competition," even though increased competition may prove to

be inefficient? What role will con-

cerns about equity play? The cur-

rent regulatory system places a

great deal of emphasis on equit~

Should this change? Will there continue to be cross subsidization

of certain classes of utility customers, due to society's decision that

providing such customers with

electricity has social benefits?

Just as with the environmental externalities debates, it is impossi-

ble to determine appropriate pol-

icy responses for retail wheeling until policy goals are clearly de- fined. Given some of the con-

cerns raised here, it does not ap-

pear that retail wheeling will

necessarily improve economic efficienc~ There is little doubt that it

would benefit some customers, especially large customers with mar-

ket power. However, it is not at all clear whether those benefits

would outweigh the costs and, even if they did, whether such a

redistribution of wealth would be acceptable from an equity stand-

point. Policy makers must there-

fore determine their goals for re-

tail wheeling and decide which tradeoffs are acceptable, and which a r e n o t . 27 Until that time,

retail wheeling will be unable to

deliver on its promises. •

Endnotes:

1. See, for example, Stephen Wiel, The New Environmental Accounting: A Status Report, ELEC J., Nov. 1991, at 46- 54 ; Paul L. Joskow, Environmental Ex- ternalities: Let's Do It Right, ELEC J., Mar. 1992, at 53-67; A. Freeman, et al., Weighing Environmental Externalities: How to Do It Right, ELEC J., Aug. 1992, at 18-25.

2. For a discussion, interested readers are referred to D. Dodds and J. Lesser, Appropriate Use of Numeric and Monetary Values for Environmental Impacts of Energy Resource Develop- ment and Use Decisions, State of Washington Interagency Task Force on Environmental Costs, Issue Paper ITF- 3; D. Dodds and J. Lesser, Can Utility Commissions Improve on Environmental Regulations?, LAND ECON. Feb. 1994, at 63-76.

3. See D. Gegax and K. Nowotny, Com- petition and the Electric Utility Industry: An Evaluation, 10 YALE J. ON REG. 63- 77. Economies of scale refers to a declining average cost function.

April 1994 47

Economies of scope refers to a situation where produc t ion of a combina- tion of ou tputs by one firm is cheaper than the produc t ion of those inputs by s ingle-product firms.

4. Regrettably, the defini t ion of "natural monopo ly" is more compl ica ted than the s t anda rd in t roductory economics textbook defini t ion relying on decl ining average costs. Declining average costs are sufficient for a na tura l monopoly, but not necessary. An ex- haust ive discussion on natura l monopoly can be found in WILLIAM BAUMOL, ET AL., CONTESTABLE MARKETS AND THE THEORY OF INDUSTRY STRUCTURE (Har- court Brace Jovanovich, 1982).

5. See P. JOSKOW AND R. SCHMALENSEE, MARKETS FOR POWER: AN ANALYSIS OF ELECTRIC UTILITY DEREGULATION (MIT Press, 1983). Joskow and Schmalensee discuss numerous deregula t ion op- tions. None, however, incorporates retail wheel ing as it is now envisioned. Instead, the authors assess a com- pletely deregula ted sys tem and several al ternat ive indus t ry s tructures with open access wholesa le transactions.

6. Id.

7. Evidence for this s i tuat ion in the deregulated marke t in Great Britain is p rov ided by R. Green and D. New- bery, Competition in the British Electric- ity Spot Market, J. POLIT. ECON. Oct. 1992, at 929-53. The British experi- ence, and its relevance for de te rmin- ing the l ikely benefits from retail wheel ing in the U.S., is d iscussed further below.

8. This is technically known as the Kaldor-Hicks criterion. See, R. ZERBE AND D. DIVELY, BENEFIT-COST ANALYSIS IN THEORY AND PRACTICE (Harper Col- lins, 1994) for a thorough discussion.

9. A discussion of federal preference statutes can be found in J. Lesser, The Economics of Preference Power, 12 RES. IN LAW AND ECON. 131-51 (1989).

10. In an ideal ized wor ld , complete e l iminat ion of cross subs id ies among the different marke ts for electricity would lead to full Ramsey pricing, where the rates charged to customers

depend on the shape of their d e m a n d curves for electricity.

11. There are other issues that may be of concern. However, we believe the issues ment ioned will have the largest potent ial impacts on efficiency and equity.

12. This is noted by Gegax and Nowotn)~ supra note 3. As we discuss below, there are other external costs and benefits that would be affected by retail wheeling.

13. See Joskow and Schmalensee, supra note 5, for further discussion of these benefits.

a ~ , _ _ , . . ~ % a l ~ ' ~ R , ~ - -,, ~ - ~

r I . : J ~. , ; . ~ ~ .

~vL.~ r._~,v~; , ,.~ . . ' . .

14. From a more mundane standpoint, there is also the issue of who is called when service is d i s rupted? Does a retail par ty call the local uti l i ty supp ly ing dis t r ibut ion services or does it call the enti ty generat ing the power?

15. Pub. L. 102-486, 102d Cong., 2d Sess. (1992).

16. Flor ida Munic ipal Power Agency v. Flor ida Power & Light, Docket No. TX93-4-000, Oct. 28, 1993. FERC d id not order any specific pr ic ing mechanisms. Its order did state, however, that if FP&L charged more for network services than point- to-point

t ransmission service, the difference would have to be justified.

17. See R. Frame, Transmission Access and Pricing: What Does a Good "Open Access" System Look Like? (NERA Working Paper #14, Jan. 1992).

18. This is k n o w n as a sys tem of complete cont ingent claims markets . In essence, it means one is wr i t ing insur- ance policies for all possible future outcomes.

19. Under the Energy Policy Act, a ut i l i ty cannot refuse to wheel power, unless a state ut i l i ty commission deter- mines that the wheel ing would not be in the "publ ic interest ."

20. This is the approach preferred by some authors. See e.g., Joskow, supra note 1.

21. For a more complete discussion, see Gegax and Nowotny, supra note 3.

22. For a formal discussion, see P. Mil- grom and J. Roberts, Predation, Reputa- tion, and Entry Deterrence, 27 J. OF ECON. THEORY 280-312 (1982).

23. For further discussion, see T. Bren- nan, Entry and Welfare Loss in Regu- lated Industries, COMPETITION AND THE REGULATION OF UTILITIES (M. Crew, ed., Kluwer Academic Publishers, 1992).

24. See Green and Newbery, supra note 7.

25. A good discuss ion on R&D can be found in J. Reingaum, The Timing of In- novation: Research, Development, and Diffusion, in HANDBOOK OF INDUSTRIAL ORGANIZATION (R. Willing and R. Schmalensee, eds., Elsevier Scientific 1989).

26. Joskow and Schmalensee, supra note 5, reject the analogy be tween de- regulat ion in the electric indus t ry and deregula t ion in the te lecommunica- tions industry.

27. One could conceivably argue that FERC has made these decisions already. However , the impacts of retail wheel ing extend far beyond FERC's regula tory grasp. It is not at all clear whether retail wheel ing has been ex- amined in a b road publ ic pol icy context at this time.


Retail wheeling: Déjà vu all over again

Documents