Overview of Infrastructure Regulation
Overview of Infrastructure Regulation
ON INFRASTRUCTURE REGULATION
The authors thank Farid Gasmi and Jos I. Tvara, who co-authored the original document with us in 2004, for their
important research and contributions. Special thanks are owed to Rossana Passaniti and Cynthia Stehouwer for their
organizational and editorial work, and to Janice Hauge, Lynne Holt, Mark Hoekstra, Hamilton Silva, Troy Quast,
Juan-Daniel Oviedo,Rich Gentry, and Abhay Dhingra for their research efforts, and Michael Rodriguez for his web
design expertise. In addition, the authors thank the members of the 2008 review committee Martin Rodriguez
Pardina, Rohan Samarajiva, Jorry Mwenechanya, and Pippo Ranci for their advice and recommendations, as well
as the World Bank staff and attendees at the original review meetings at Eynsham Hall and Arlie House for their
helpful suggestions and comments. The authors take responsibility for all errors and omissions.
A.
Introduction
Section A of the first chapter examines theories of regulation and the rationale for regulation. Section H of this
Overview and Chapter VII that follow specifically examine how regulators can address this political context of
regulation.
2
In this Overview, the authors generally refer to the government when referring to the development of policies,
and to the regulator or agency when referring the implementation of policy. The authors recognize that the
institutional arrangements for developing and performing regulation vary across countries. For example, in some
countries, regulatory agencies take initiative in opening markets to competition, while in other countries all such
work is done within a ministry. However, it is too cumbersome to try to reflect all possible divisions of
responsibilities for regulatory policy in this narrative, so language is simplified here.
level is considered next, followed by issues of rate design. Finally non-price issues, such as
service quality, environmental impacts, and social issues, are reviewed, as is the regulatory
process, including the management of information.
The remainder of this Overview is organized as follows. Section B defines the regulatory
problem from different perspectives and identifies the basic approaches for overcoming the
market power and information issues that tend to underlie many regulatory policies. Section C
describes the first approach, namely the use of competition. Section D summarizes the second
approach, which is the gathering and use of information on markets and operators. Section E
examines the last approach, the use of incentive regulation. The remaining sections examine
related issues. Section F describes issues in tariff design. Section G covers service quality,
environmental, and social issues. Section H examines the regulatory process. Section I provides
concluding observations.
B.
It seems fair to say that governments establish regulation of utilities to improve sector
performance relative to no regulation. What might be meant by improve sector performance,
however, can be subject to considerable debate. Often improve sector performance means that
the government wants to control market power and/or facilitate competition. It may also mean
that the government wants to address commitment issues; that is to say, the government may
create a regulatory agency to protect operators and customers from politically-driven decisions
that would sacrifice long run efficiency for short term political expediency.
Improve sector performance might also mean that the government has chosen to
regulate in order to favor particular types of customers or to protect operators from competition.
In some countries, for example, regulation has been used to subsidize electricity for farmers. In
the 20th century many counties used regulation to protect monopoly telephone companies from
competition.
Except where otherwise noted, this Overview addresses normative issues of regulation,
with the perspective that regulation, and in particular regulation by a regulatory agency, is
intended to improve welfare. 3 In this context, welfare means the aggregate benefit that
infrastructure services provide, including benefits to consumers, 4 benefits to operators, and
externalities. 5 Externalities are benefits or costs from a transaction that are received or born by
3
Focusing primarily on welfare is not meant to imply that distributional issues in regulation are unimportant.
Distributional issues address how different stakeholder groups are affected differently by how infrastructure services
are provided. Situations where governments use regulation to benefit some groups over other groups have been
already mentioned. Portions of Chapters V, VI, and VII are devoted to distributional issues, namely, assisting the
poor. Welfare is the focus because it is the measure of benefit most often used in research and because policies that
emphasize welfare do not preclude also adopting policies that address distributional issues.
4
Benefits to consumers is generally measured as net consumer surplus, which is the difference between the gross
value that the customer receives when consuming the service (called willingness to pay) and the amount the
customer pays.
5
This Overview does not address the individual weight that the government may give to each element. These
weights are important, but they are set aside here because each regulator can use her or her governments own
weighting system to determine which tools described herein to apply and how to apply them.
third parties who are not part of the transaction. Air pollution produced by electricity generation
or by transportation vehicles is an example of a negative externality.
At this point, it is important to note that some observers make convincing arguments that
policymakers sometimes have more nefarious motives than maximizing welfare, for example, to
gain short term political advantage or to benefit political supporters. Such motives raise the issue
of how citizens can regulate their government and the regulator. Review of this issue is reserved
to Section H of this Overview and to Chapter VII.
From a normative perspective, regulation of a service provider may be desirable if (1) the
welfare objectives of the government are different from the objectives of the operator, (2) the
operator has an information advantage over the government, and (3) the operator has market
power.
To illustrate why regulation may be appropriate when the government and the operator
have different objectives, consider a situation in which the government and the operator each has
a single objective, namely, the government wants service expansion in rural areas and the
operator wants to maximize profit. An unregulated operator with market power would restrict
output to maximize profit and would invest in areas that give the highest profit. It is unlikely that
either of these outcomes would be consistent with the governments objective, so the government
may want to take steps that would make it in the operators best interest to expand service in
rural areas. In the case of a state owned operator, the management may have an objective of
maximizing its political influence. This could lead the management to use its resources to
employ a lot of people or to target investments to politically powerful areas. Neither would likely
be consistent with the goal of expanding service in rural areas.
Now consider a situation where the government and operator have the same objective,
say to offer service of a particular service quality throughout the country at the lowest possible
cost. In this case, the government could simply give the operator whatever relevant information
the government had and let the operator pursue this objective on its own. Regulation would not
be needed in this situation because the government could not improve results by regulating the
operator, that is to say, regulation, if designed to persuade the operator to do what the
government wants the operator to do, would be redundant with the operators own strategic
objectives.
Of course the world is not so simple. Many developing countries have not privatized their
utilities, but have established regulatory agencies with varying degrees of independence. An
important challenge for these regulators is that the state-owned utilities often do not operate on a
commercial basis. 6 The utilities sometimes are torn between the short-term social stability
interest and the long-term viability of the utility enterprise. Therefore there are many
opportunities for conflicts between the utility and the government even if the interests would
appear to coincide. 7
This experience is not unique to developing countries. Studies have found that government-owned utilities in
developed countries do not operate on a commercial basis.
7
The authors are grateful to Jorry Mwenechanya for this insight.
C.
ways. The first way is that the operator, in its pursuit of profits, has an incentive to provide
service quality levels and price levels that are best for customers, subject to the operators need
to cover its costs. Competition can provide this result because fully informed customers will buy
only from those operators that provide the most beneficial combinations of quality and price. In
other words, each customer seeks to maximize his net consumer surplus.
Even if the operator in a competitive market is state owned, competition presses the
operator to act as a privately owned operator because the state-owned operator must be
responsive to customers in order to finance its operations. Of course a state-owned service
provider might be able to use its relationship with the government to gain an advantage over
rivals, which would at least in part thwart the discipline of the competitive market. For example,
if a state-owned operator were allowed access to taxpayer-provided monies when cash flows are
unable to support investments, the state-owned operator could have an incentive to make
uneconomic investments that further the operators political goals or reduce competitive
pressures.
The second way that competitive pressure addresses the information asymmetry problem
is that competitive market outcomes reveal actual customer demand, the operators innate ability
to be efficient, and how much effort the operator is willing to exert to be efficient. Even if
competition is weak, competing firms can benefit from using all of the information at their
disposal to succeed in the competitive marketplace.
Competition has additional benefits. It limits a governments ability to use regulation to
favor certain stakeholders or to sacrifice long term efficiency for short term political goals. It
also limits operators abilities to raise prices and creates opportunities for different firms to try
innovative ways to attract customers. 12
Policy makers or regulators subject operators to competitive pressures by liberalizing
markets and facilitating competition. There are three basic approaches. The first approach is to
have multiple operators compete in the market for customers. This is called competition in the
market and examples include having multiple mobile telecommunications service providers and
multiple operators of electricity generation plants. The second method, called competition for the
market, is to have operators compete for the market by having the operators bid for the right to
be a service provider. 13 Franchise bidding to operate a city water system is an example of this
second approach. The third technique is to have operators in different markets compete by
comparing the efficiency and effectiveness of their operations and rewarding those operators that
provide superior performance. Competition in the market is reviewed next, followed by
competition for the market. The third approach, called competition between markets, but also
called benchmarking or yardstick regulation is covered in the section on Incentive Regulation.
12
This is not to imply that competition is without problems. There may be situations where operators compete
aggressively in lowering prices and so do not have adequate cash flows to make needed investments. Competition
has generally been most successful in telecommunications.
13
Operators may also bid for the right to be a service provider in situations of competition in the market. This might
be the case in mobile telephony, for example, where radio spectrum limits the number of possible operators.
1.
There are several approaches to facilitating competition in the market. When all elements
of the utility service can be competitive, then generally a primary job of the regulator is to
remove barriers to entry or competition. Typical steps include removing licensing restrictions or
large licensing fees, 15 reducing switching costs, and requiring access to essential inputs, such as
telephone numbering resources.
When some elements of the utility service have monopoly characteristics, such as gas
distribution lines, and other elements can be competitive, such as gas production, then regulators
also use tools such as structural separation and unbundling to facilitate competition. Structural
separation separates the potentially competitive portions of the utility service from the noncompetitive portions. For example, electricity generation is generally considered to be potentially
competitive, but electricity distribution is not. 16 These non-competitive, yet essential portions of
the service are called essential facilities. 17 Structural separation prohibits a single operator from
providing both the competitive and non-competitive portion of the service.
The intent of separation is to ensure that the provider of the essential facility does use its
control of the essential facility to hinder competition. Structural separation is sometimes called
unbundling, but some forms of facility or service unbundling may be less severe than structural
separation. With simple unbundling, for example, the regulator may allow a single operator to
combine competitive and non-competitive elements to provide bundled service, but also require
the operator to allow rivals access to the essential facility so that the rivals are not disadvantaged
relative to the operators own competitive service. 18 For example, some regulators require
incumbent fixed line telephone operators to allow rivals to lease local telephone lines, but the
regulators also allow the incumbent operators to continue to offer a retail service that bundles the
local telephone line with usage. Regulators that want to facilitate competition generally take
steps to remove barriers to entry, even if structural separation or unbundling is required.
When structural separation or unbundling do not involve separate ownership, regulators
often require accounting separation or ring fencing to ensure that there is no cross-subsidization
from the non-competitive operations to the competitive operations. Accounting separation is
described in more detail below in the subsection on Financial Analysis and in Chapter III.
Access pricing is an important element of regulatory policies designed to facilitate
competition in the market. When a utility service is unbundled, the rivals often pay the operator
14
an access price for use of the non-competitive element of the service. Because this price is a
source of revenue for the incumbent operator the operator that provides both the competitive
and non-competitive portions of the service and a cost for the incumbents rivals, the
incumbent has an incentive to raise this price to a level that limits competition. In cases such as
telecommunications where competitors must interconnect their networks in order to allow
customers of rival networks to communicate, regulators generally require service providers to
negotiate such interconnection arrangements and adopt cost-based prices.
2.
Competition for the market may be desirable when competition in the market is infeasible
or impractical. In such cases, the right to be the monopoly provider of the service 20 could be
auctioned off through an efficient auction. An efficient auction is one in which (1) the most
efficient firm wins the auction, and (2) the winning operator gives up most of its monopoly
profits. An efficient auction achieves cost efficiency because the most efficient firm is the firm
that can afford to pay the highest price for the right to be the monopoly. In paying this high price,
the successful bidder gives up at least some portion of its monopoly profits, which can be
distributed to customers. In general, monopoly profits are profits above the operators cost of
equity 21 that result from the operator having market power. Post-auction regulation may still be
necessary if prices need to adjust to unanticipated events, but periodic re-bidding may substitute
for typical price regulation in some situations.
D.
E.
The third approach to dealing with information asymmetries and limiting the effects of
market power is for the regulator to design and implement incentive schemes that reward the
operator for using its private information to achieve the governments objectives. To be most
effective, the reward should (1) provide the operator with additional units of something it wants
for example, profits when the operator gives the government something it wants for
example, lower prices; (2) give the operator performance options that provide higher rewards for
accepting more challenging performance goals; and (3) allow the operator to keep only a
minimal reward for example, accounting profits that are no greater than the operators cost of
capital when the operator chooses the least challenging performance goal. Cost of capital is the
financial return that the operator must give to its investors and creditors to induce them to
provide capital for the firm. 24
1.
Incentive regulation is generally implemented by controlling the overall price level of the
operator. There are four basic schemes to regulating overall price levels. The first approach is
generally called rate of return regulation or cost of service regulation. This regulatory instrument
establishes an overall price level that allows the operator to receive accounting profits that are
just equal to the operators cost of capital at the time the price level is set. Actual profits may
deviate from the cost of capital until the next time the regulator reviews the operators profits.
The second approach is called price cap regulation or RPI-X regulation, which is a
method that establishes the operators overall price level by indexing the price level according to
inflation minus an offset, called an X-factor. 26 The X-factor should reflect the difference between
this operator and the average firm in the economy with respect to their abilities to improve
23
Chapter IV on Regulating the Overall Price Level provides the primary information on incentive regulation,
although Chapter VI on Quality, Social, and Environmental Issues also examines incentives.
24
Section G of Chapter III notes how to estimate the cost of equity and the cost of capital.
25
Section A of Chapter IV covers this topic in more depth.
26
As explained below, some regulators using price cap regulation incorporate elements of rate of return regulation.
efficiency and to changes in input prices. Directly measuring these efficiency and input price
inflation differences to establish an X-factor, while ignoring the operators actual profits, is
called pure price cap regulation.
The third approach to regulating the overall price level is called revenue caps. This is
similar to price caps except that the inflation-minus-X formula applies to revenue rather than
prices.
The fourth approach is called benchmarking or yardstick regulation. This form of
regulation provides competition between markets by comparing operators across markets, in
effect forcing the operator to compete against the performance of comparable operators in other
markets. In practice benchmarking or yardstick regulation is an input used in price cap or
revenue cap regulation, and sometimes in rate of return or cost of service regulation.
Many regulators adopt hybrid incentive schemes, which are approaches that combine
features of the three basic methods of incentive regulation described above. For example, the
U.S. Federal Communications Commission once combined elements of rate of return regulation
and price cap regulation. Under its scheme, operators could choose from a menu of options.
Each option included an X-factor and a formula that determined the proportion of accounting
profits that the operator would be allowed to keep. Options with more aggressive (larger) Xfactors allowed operators to keep larger proportions of their accounting profits. Regulators in the
U.K. often use elements of rate of return regulation and benchmarking analysis to establish Xfactors in price cap regulation. This is described in the subsection Financial Analysis.
Another approach for combining elements of rate of return regulation and price or
revenue cap regulation is to include pass through elements in the scheme. For example a
regulator might use revenue caps to regulate the revenue of an electricity distribution operator,
but allow the operator to pass through changes in fuel costs to the extent that these cost changes
are beyond the operators control.
2.
Financial Analysis 27
It is very rare for incentive regulation to not involve extensive financial analysis. Such
analysis includes determining the operators cost of capital, historical costs, and projected costs.
The cost of capital consists of two elements, the cost of debt and the cost of equity. Regulators
typically obtain debt costs from operators financial reports, where the operators list their longterm debt instruments and the interest rates paid. Estimates of the operators cost of equity can be
obtained using financial models. Regulators combine the operators cost of debt and cost of
equity into a weighted average, called the Weighted Average Cost of Capital (WACC).
Some regulators, such as those in the U.K., use historical and projected operating and
investment costs to set X-factors. (Historical information alone is generally used in rate of return
regulation.) The operators historical operating and investment costs can be obtained from the
operators accounting records. Care must be taken when using historical accounting data in
27
situations where accounting standards were historically weak or inconsistent over time. In the
U.K. approach, projected operating and investment costs, existing net investment in regulatory
assets or rate base, and projected net investment are used in a net present value or equivalent
analysis to establish X-factors. This method involves making demand forecasts, identifying
investment requirements to meet the projected demand, and the forecasting of associated
operating expenses. These projections are analyzed and adjusted by the regulator to determine
how the operators overall price level should be allowed to change relative to inflation.
When using accounting costs, whether they are historical or projected, regulators place
below the line any costs that are not needed to provide the utility service or that are considered
excessive. The effect of this below the line treatment of costs is that prices for regulated services
are not intended to provide revenue to cover the costs. Costs for items needed to provide the
utility service are considered to be used and useful and so are kept above the line, which means
that they can be recovered through prices charged for regulated services. Costs that are
excessive, perhaps because the operator paid too much for an item or made an avoidable mistake
in an investment are considered imprudent and the excess is placed below the line.
3.
4.
28
F.
Tariff Design 30
Once the overall price level has been established for the operator, the work of
establishing the rate (or price) structure still remains. This work is called tariff design or rate
design and refers to relationships among the individual prices (or rate elements) that the operator
charges. In some instances, the regulator may choose not to regulate the price structure.
Examples include (1) situations where the objectives of the operator are in line with, or at least
do not contradict, the objectives of the regulator, at least as they relate to rate design, and (2)
situations where the regulators resources are limited and regulating price structure is a low
priority.
Most economists agree that efficient price structures cover total cost and align prices with
marginal costs. Marginal cost is the additional capital and operating cost that results from
increasing output by a single unit. 31 Marginal cost pricing may be difficult in situations where
there are economies of scale or economies of scope because prices equal to marginal costs would
not cover all of the costs of the operator and thus would not attract investment. In these
situations, regulators and operators generally favor multipart pricing or in some instances
Ramsey pricing. 32 Multipart pricing is an arrangement where the operator charges separate prices
for different elements of the service. For example, a water provider may charge a connection fee
plus a usage fee. With Ramsey pricing, which is also called differentiated pricing or the inverse
elasticity rule, the operator charges higher prices to customers with inelastic demand and lower
prices to customers with elastic demand. 33
than as method of incentive regulation that could be used by itself. It is true that with benchmark or yardstick
regulation prices are generally capped, which arguably makes the system look very much like price cap regulation.
Here it is treated as a separate form of incentive regulation because some regulators treat it that way, but the authors
recognize that benchmark or yardstick analyses are generally used as an element of a hybrid regulation scheme.
30
Chapter V covers tariff design.
31
If the system is capacity constrained, meaning that capacity cannot be increased, marginal cost would also include
the marginal congestion cost.
32
Regulators generally limit the use of Ramsey pricing to services that are not considered basic or essential to
customers the regulator is particularly trying to protect, such as low income or residential customers.
33
Customers have inelastic demand if they do not change the amount they purchase by very much if the operator
changes its prices. Conversely, customers have elastic demand if they respond to changes in prices by making large
changes in the quantities that they purchase. More precisely, inelastic demand means that a one percent change in
G.
possible for the poor, who are generally unable to establish credit for post-paid service, to obtain
service. 36 Competition among entrepreneurs who transport water from wells or streams has also
increased the supply of water to the poor in some instances.
Situations also arise where services to the poor can be made affordable by offering
services that are of a lower quality than services provided to wealthier customers. For example, a
shared sewage system provides a lower level of service than a system that gives each customer
his or her own connection, but may be more affordable for the poor than the higher quality
system.
Subsidies are also a common feature of policies designed to assist the poor. These
generally take the form of service or infrastructure development obligations for operators.
(Infrastructure development issues are described above.) In these situations, the operator
internalizes the subsidies. In other instances, the subsidies may be explicit. For example, water
customers living in low-income areas of Colombia have received credits on their bills.
Customers in wealthier areas had surcharges on their bills to fund the subsidies to the poorer
customers. Subsidy arrangements should be approached with caution. Research has shown that
traditionally higher income customers benefit more from subsidies than do poorer customers.
H.
Regulatory Process 37
Prepaid service was subsequently adopted even in markets where there was no competition.
Chapter VII covers the regulatory process.
38
This topic is covered in both normative theories of regulation and positive theories of regulation.
37
policy makers by encouraging regulation under the law and independence, transparency,
predictability, legitimacy, and credibility of the regulatory system.
1.
Institutional Arrangements
2.
The review and appeal processes for regulatory decisions includes decision making
processes, choices of regulatory instruments, stakeholder and government roles in regulatory
decision making, mechanisms for appeal of regulatory decisions, and alternative dispute
resolution processes. Regulatory instruments include legislation and licenses, the choice of
which is often determined by the legal traditions of the country and the methods by which these
instruments can be changed. For example, the regulatory process is politicized if a license is the
regulatory instrument and the ministry can change the license at will.
In some countries, regulatory decisions are subject to ministry review, which can also
politicize regulation. To avoid such situations, some countries provide only judicial review of
regulatory decisions or establish administrative tribunals. Some countries allow courts to
overrule the regulator only on legal or procedural grounds and not on the substantive grounds of
the regulatory decision itself. In some situations legal processes can delay regulatory decisions to
such an extent that the decisions cannot be made in a timely fashion, which degrades sector
performance. To avoid such delays, some countries use alternative dispute resolutions
procedures, such as binding arbitration, to speed resolution of conflicts.
3.
Ethical Conduct
4.
Stakeholder Relations
Stakeholder relations affect the independence of the agency and include the use of
advisory boards, communication strategies, grievance procedures, and relationships with the
government, consumers, operators, and investors. Some regulators use advisory boards to
facilitate stakeholder input, especially on issues of long-term planning and on issues that require
ongoing surveillance, such as service quality regulation. Care must be taken when using advisory
boards to ensure that the stakeholders represented do not obtain privileged positions for
influencing the regulator. Regulators generally receive complaints from consumers related to
prices and service quality, and often regulators have special staff designated to handle these
complaints.
Some of the regulators interactions with stakeholders can take the form of negotiations.
Such circumstances make it important for regulators to develop strategic negotiation skills, such
as identifying parties interests and win-win solutions.
Lastly, regulators generally dedicate trained staff to dealing with the press because the
public receives most of its information about regulation through newspapers and other media.
This reliance upon journalists makes it important for regulators to develop good press relations,
provide effective press releases, and learn how to provide timely and accurate information to the
press.
I.
Concluding Observations
and efficient sector performance which is necessary for customers to receive their maximum
benefit from the sector decision-making procedures should be in place to limit information
asymmetries and that provide incentives for operators, government, and regulators to work for
the best interest of customers and the economy. This generally means that (1) effective
competition should be encouraged whenever possible, (2) the regulator should gather
information about the sector and should provide stakeholders with information on the regulator
and her decisions, (3) incentive regulation should reward the operator with the opportunity for
higher profits when he accepts performance goals that make customers better off, (4)
requirements should be established for service quality and access for the poor, and (5) regulatory
processes should align the goals and capabilities of the regulator with the welfare of customers.