Regulatory Incentives for Investments in Electricity Networks - CRNI ...

THIRD ANNUAL CONFERENCE ON
COMPETITION AND REGULATION IN NETWORK INDUSTRIES
19 NOVEMBER 2010
RESIDENCE PALACE, BRUSSELS, BELGIUM
Regulatory Incentives for Investments in Electricity Networks
Dr. Konstantin Petrov, Dr. Viren Ajodhia, Dr. Daniel Grote ∗, Denis Resnjanskij
KEMA Consulting GmbH
Kurt-Schumacher-Str. 8, 53113 Bonn, Germany
Abstract
The integration of the European electricity markets, increasing shares of renewable electricity
generation and distributed generation all require significant investment in network capacities in
the near future. Incentives for investments in regulated transmission and distribution networks
are however highly dependent on the specific details of the applied regulatory regimes. A failure
to include adequate capital costs in the revenue requirements could result in underinvestment and
risks for network reliability.
In this paper we analyse the explicit and implicit drivers for network investment set by different
regulatory concepts from a normative point of view and from observations of regulatory practice.
The paper addresses general properties of incentive and rate-of-return regulation, the
incorporation of capital expenditures in the allowed revenue, the use of quantity adjustment
factors and quality incentives, explicit investment allowances, efficiency carry-over and slidingscale
schemes.
Keywords
Investment incentives, network industries, capital expenditure, investment assessment, incentive
regulation.
∗ Corresponding author: KEMA Consulting GmbH, Kurt-Schumacher-Str. 8, 53113 Bonn, Germany, Telephone:
+49(0)228-4469049, Fax: +49(0)228-4469099, E-Mail: Daniel.Grote@kema.com
1

1 Introduction
Reliable and efficient electricity networks are essential for the successful performance of
electricity markets. Increasing cross-border trade and the integration of renewable energy place
significant pressures on the existing network infrastructure that can only be dealt with by further
investment in network infrastructure. 1 Furthermore, the replacement of many aging conventional
thermal and nuclear power plants – reaching the end of their life-cycle – as well as the entry of
new power producers might further shift the load flows in the networks and require additional
network investments. A large increase in the regional integration and interconnection of
European electricity markets as well as investments in smart grids have also been identified as
key issues for the achievement of ambitious reductions in carbon emissions in the future. 2
Several regulators and governmental institutions in the European Union are currently considering
further measures to increase efficient investment under their respective regulatory regimes. In
fact, after a focus on efficiency improvements in the first years after liberalisation, adequate
investment incentives have now (besides quality of supply) become a key area for many
European regulators. Increased investments have also been identified as a crucial issue for the
development of competition, security of supply and the internal energy market in European
electricity and gas legislation. For example the European Directive 2009/72/EC and the
Regulation (EC) No. 714/2009 require, among others, each transmission system operator to
“…build sufficient cross-border capacity to integrate European transmission infrastructure…”
and for those structured as independent transmission or system operators to submit a 10-year
network development plan. Furthermore, according to this Regulation, the European network of
transmission system operators for electricity (ENTSO-E) shall also adopt a European wide nonbinding
10 year network development plan every two years. The first of such plans for electricity
was published in March 2010.
Investments in electricity networks are characterised by economies of scale and lumpiness, so
that ‘overbuilding’ is often the least cost option. These characteristics of network infrastructure
seriously limit the extent to which market forces can be introduced in the provision of network
services. In practice, this means that the costs of network operation and of infrastructure
investment should be recouped via regulated network charges. Various regulatory regimes have
been developed to address the shortcomings of the monopolistic nature of electricity networks
and to facilitate competition on up- and downstream markets. In economic literature as well as in
regulatory practice, the different regulatory mechanisms have often been assessed on their ability
to facilitate non-discriminatory network access and to set adequate incentives for efficient
performance. Each of these regulatory regimes, in their general nature, however also sets
1 Significant network investments are required as electricity from renewable energy sources is often produced at
locations far away from the centres of large demand. Renewable power generation also requires investment in the
network infrastructure beyond the connection of the renewable plants since it is largely dependent on climate
conditions and therefore characterised by large load variations. Furthermore, increases in the share of distributed
generation, in particular from small-scale renewable generation, have in several cases reversed the traditional
unidirectional flow of electrical energy from the transmission to the distribution network.
2 See for example the report commissioned by the European Climate Foundation: “Roadmap 2050: A practical
guide to a prosperous, low-carbon Europe” available at: http://www.roadmap2050.eu/downloads
2

particular incentives implicitly for investment in network infrastructure. These incentives are
often further complemented by specific supplementary schemes to the basic regulatory model.
The paper focuses on the regulatory treatment of investments, the incentive arrangements
associated with different regulatory regimes and their underlying properties. It covers the
inclusion of capital costs in the allowed revenue (ex-ante versus ex-post), the integration of
capital costs in the efficiency assessment models and the application of quantity adjustment and
quality incentive factors in the regulatory formulas. Moreover the paper addresses the application
of explicit investment allowances, efficiency carry-over and sliding-scale schemes.
The remainder of the paper is structured as follows. The next section describes the general
relationship between regulation and investments. Section 3 explains the major regulatory models
and their impact on investments. Section 4 describes two main approaches to integrate capital
expenditures in the regulatory process: explicit ex-ante projections with efficiency checks and
ex-post efficiency assessment using benchmarking on total costs. Section 5 focuses on specific
regulatory arrangements aiming to encourage investments such as quantity adjustment factors,
quality incentive schemes and others. The paper concludes with a summary of the key issues.
2 Regulation and Investments
In a competitive market, a company’s investment decisions are based on incomplete information
and therefore require the company to make an assessment of the potential outcomes and the risk
of undertaking the investment. The company will bear the complete risk, once the investment has
been undertaken. If technological change or by-pass occurs, the company bears the cost of
obsolescence or lower returns. If no technological change or bypass eventuates, the company
will obtain a return on the asset for a longer period of time than initially expected. The company
operating in a competitive market will therefore face periods when it under-recovers the
expected returns and times when it over-recovers the expected returns. Hence, mechanisms like
risk-diversification enable companies to compensate or neutralise risk effects and to start projects
under uncertainty.
Network companies operating in regulated environments are affected by technological change or
changes in demand in a similar way. Given the capital intense and irreversible nature of network
investments, correct risk assessment and risk valuation in the regulatory regime are of crucial
importance for the financial viability of network companies. If the regulator is aiming (and
willing!) to mimic decisions made on a competitive market, investment valuation approaches
that consider the probability of various outcomes should be included in the regulatory process.
Furthermore, the choice of the regulatory regime and the behaviour of the regulator both have a
direct impact on risk, capital costs and investment incentives of regulated network companies.
The financial and economic feasibility of an investment project can change over time. Failure to
include adequate capital related costs as part of the revenue requirement of the regulated
company risks a reduction in investment by the industry below an optimal level. This could
ultimately lead to reductions in cost coverage and quality levels, and hence to a reduction in
reliability of supply in the medium and long term. On the other hand, including inefficient
investment costs in the allowed revenue leads to high network charges, distorted price signals
3

and constrains the effective market performance. It is the task of the regulator to set the right
balance through transparent, consistent and predictable regulatory decisions.
3 Price Control Methods
A range of forms of energy network price regulation are used by regulators, as are various
classifications of these forms by commentators. We group and explain the different forms of
price regulation under two major categories, namely cost-based (rate-of-return) regulation and
incentive regulation. The following diagram presents the typical regimes for regulatory price
controls:
Figure 1: Regulatory Price Control Models
Regulatory price
controls
Cost-based
Regulation
Incentive
Regulation
Cap Regulation
Rate-of-Return
Revenue-Cap Price-Cap Yardstick
It is however important to note that the regulatory regimes applied in practice are seldom applied
in their pure theoretical form, but often elements of different theoretical regimes are applied
simultaneously and complemented by additional mechanisms. In the following section we
discuss the regulatory regimes in detail.
3.1 Rate-of-Return Regulation
Under rate-of-return regulation, the regulator sets prices for the service provider – generally
every year, or sometimes every two years – in such a way that they cover the service provider’s
costs of production and include a rate of return on capital that is sufficient to maintain investors’
willingness to replace or expand the company’s assets. The forecast of operating costs and
capital expenditures can be based on the previous year’s numbers with an adjustment for price
inflation and specific needs to replace and extend assets.
4

This form of regulation can effectively encourage investments if the regulation is designed to
ensure (e.g. through end-of-year adjustments) that the network company receives a guaranteed
rate of return. Under such a scheme of rate-of-return regulation, the regulated firm is largely
immunised against cost changes as they can be passed through to the customers; so it is in fact
the customers who bear the risk. This could result in a lower risk for the regulated firm and likely
lower costs of capital. On the other hand, it has two well-known and significant disadvantages: it
provides little or no incentive to control costs, let alone reduce them, and it may provide an
incentive for the service provider to over-invest in capital. Other disadvantages include the need
for frequent regulatory reviews, with often detailed information needs, and hence high associated
costs for both the regulator and the service provider. 3
3.2 Cap Regulation
Under cap regulation, prices or revenues are set in advance, usually for a period of three to five
years, allowing the company to benefit from any cost savings made during that period. At the
end of the period, the prices or revenues are recalculated in order to generally bring them back
into line with costs, and thus pass the benefits of any efficiency gains through to customers. 4
The “cap” refers to the upper limit that is placed on prices or revenue, hence the term “price-cap”
or “revenue-cap”. 5 This type of regulation is designed to give the service provider a strong
incentive to reduce costs. This is partly done by setting the prices or revenues that a service
provider can earn over a number of years partially or completely decoupled from the costs it
incurs over this time. It is also achieved by allowing the company to keep, for a period of time, at
least a portion of the benefits of any efficiency improvements above the assumed level of
improvements incorporated in the level of the cap.
In order to take account of unpredictable rates of inflation in an economy, a cap regulation
regime typically allows a firm to vary its prices in any year by an amount linked to the overall
level of inflation, as measured by the percentage change in an appropriate price index. This
inflation-adjusted price level is then usually adjusted by a percentage, often referred to as “Xfactor”,
that reflects (among other things) the productivity improvements necessary to achieve a
level of costs that the regulator assumes is reasonable.
A simple formula that could apply to either a price or revenue cap (the cap or upper limit can
apply to either prices or revenue) is set out below 6 :
3 If the frequent regulatory reviews are very intrusive and discretionary, this can in fact increase the risk for the
regulated firm with the possible result of higher costs of capital.
4 The theoretical properties of cap regulation are for example discussed in: Joskow (2008), Crew and Kleindorfer
(2002), Bernstein and Sappington (1999), Armstrong, Cowan, Vickers (1995) or Liston (1993).
5 Both forms of regulation can be further distinguished by the structure in which the caps are set. Revenue-cap
regulation can either cap total revenues or the revenues per unit of output; price-cap regulation can set a maximum
limit on individual prices or average prices (tariff basket).
6 Sometimes another term, usually denoted as “Y” is added to the right hand side of the equation to represent costs
that the regulated service provider is allowed to “pass through” directly to the customer, usually because they are
costs over which the regulated service provider has no control.
5

R t = R t-1 (1 + CPI t - X)
where:
R t
CPI
X
t
is the limit on prices, or on revenue
is the growth rate of a general (consumer) price index
is a factor that reflects the assumed rate of efficiency increases and other relevant
factors
is the year index
This formula shows that the cap or upper limit on either revenue or prices in the current year,
year t is set equal to:
• the upper limit in the previous year;
• adjusted for general price inflation (the “CPI” term) and the assumed increase in efficiency net of
the effect of other factors (the “X” term)
This is the essence of cap regulation, though in practice other variables tend to be added to the
formula. Often these variables represent costs that are passed through in full to the consumer,
usually because they are outside the control of the service provider. They may also represent
costs related to a performance target, e.g. the allowance for network losses where the target is a
specified percentage of network losses. We provide further details on the application of caps in
the next sections.
3.3 Yardstick Regulation
The third general category of price control regulation is yardstick regulation. Under this
regulation prices or revenues are linked to the average industry performance. Yardstick
regulation is not, or at least not completely, based on an assessment of the cost position of
individual service providers but upon a comparison of prices or cost positions and cost
determinants between firms. 7 For example, under a "yardstick" mechanism based on cost
information, service providers are not allowed to charge higher prices than a particular statistical
mean that is calculated over costs of all service providers, unless perhaps different prices were
justified by their "special operating conditions".
Yardstick regulation for electricity distribution is currently applied in Norway and the
Netherlands, though in quite different forms. The Norwegian model partially links the allowed
revenue to a cost norm. The cost norm is set annually using benchmarking analysis and taking
into account any differences in external conditions. The benchmarking for electricity distribution
applies a Data Envelopment Analysis (DEA) using a national data sample.
7 For a discussion of the theoretical effects of yardstick regulation see for example Tangerås (2002), Yatchew
(2001), Weyman-Jones (1995) or Shleifer (1985).
6

The Dutch model of yardstick regulation is based on the use of a general efficiency factor for all
regulated distribution networks. The general efficiency factor is set equal to the average change
in productivity of all network operators measured ex-post.
The application of yardstick regulation is planned in Sweden and was also discussed in Germany
and a number of other countries as a long-term option, after the application of several regulatory
periods of revenue-cap regulation.
3.4 Rate-of-Return versus Cap Regulation
The key difference between cap and rate-of-return regulation is that with the latter, the upper
limit on prices or revenue is normally determined directly from actual costs in the previous year.
While in the case of cap regulation, the limit is normally determined from the limit in the
previous year. 8 Hence under cap regulation the allowed prices/revenue are determined
independently from last year’s actual costs. The service provider is able to keep completely or
partially the cost savings it makes in the period between two regulatory price reviews above the
assumed efficiency requirements.
Compared to rate-of-return regulation, cap regulation provides stronger incentives to reduce
costs; it can be argued that the longer the time between reviews, the stronger the cost reduction
incentive. Through weakening the relationships between actual costs and regulated prices, cap
regulation minimises some of the deficiencies of rate of return regulation. It avoids the need to
review prices frequently and can arguably provide greater price stability. However, cost
reductions should not be achieved by prohibitive regulatory arrangements that would not allow
investors to earn an adequate rate of return. In setting the caps, the regulator will need to ensure
that their level is sufficient to cover not only the efficient operation and maintenance costs, but
also the capital costs of efficient investments.
In economic theory rate-of-return regulation is generally associated with overinvestment and cap
regulation with underinvestment. 9 Furthermore, it is also reported in economic literature that cap
regulation can result in lower levels of quality of supply. 10 The empirical evidence on the impact
of the regulatory regime (rate-of-return versus cap regulation) on investment is however much
more mixed. While some empirical papers provide evidence that investments are indeed higher
under rate-of-return regulation, other papers provide evidence that investment levels are not
affected or are actually higher under models of incentive regulation. 11 What might actually be
the case is that regulatory regimes named as rate-of-return or incentive regulation by the
respective regulators and regarded as such in these empirical studies, do in fact represent quite
different regulatory models in different countries and sectors. Unless the specifics of a regulatory
8 Depending on the types of caps, the limits can be set on the basis of cost projection (linked caps) or automatic
formula adjustments (unlinked caps, see further below). It both cases prices / revenues during the regulatory period
are decoupled to the actual cost development.
9 See for example Joskow (2008), Armstrong, Cowan, Vickers (1995) and Cabral and Riordan (1989).
10 See for example Kwoka (2009), Elliott (2006) or Liston (1993).
11 See for example Cambini and Rondi (2010), Vogelsang (2002 and 2010), Égert (2009), Ai and Sappington (2002)
or Greenstein, McMaster and Spiller (1995).
7

egime are studied in detail, it is hardly possible to estimate the incentives of a particular
regulatory regime for investments (and on other factors).
3.5 Linked versus Unlinked Caps
In practice regulators often name regulatory regimes with quite different properties as “cap
regulation”. For explanatory purposes we divide the cap regimes into two major groups: linked
caps – the allowed revenues for each year of the regulatory period are determined by the
regulator through a projection of costs at the beginning of a regulatory period – and unlinked
caps – the allowed revenues for each year of the regulatory period are determined by the
regulator through an automatic adjustment formula starting from an initial cost level in a prespecified
year.
Linked caps base the revenues of a regulated company during a regulatory period on an ex-ante
assessment of the efficient levels of operating expenditure (Opex) and capital expenditure
(Capex) by the regulator (often referred to as building blocks approach). Under this approach
both types of costs (Opex and Capex) are usually assessed separately. For investments (Capex),
regulators typically set the allowed level on an assessment of the company’s own investment
projections. For operating expenditure cost projections can be based on a benchmarking of the
historical Opex of different network operators with additional adjustments incorporating major
changes of relevant cost drivers. Linked caps appear attractive because they link revenues to the
projected costs. At the same time they allow the consideration of efficiency increases (the
projected costs incorporated in the allowed revenue are checked for efficiency) and the
immediate allocation anticipated efficiency increases to the customers. Where there are strong
investment and maintenance needs in the near future, these types of caps may be suitable as they
enable a specific allowance to be made for the higher investment and maintenance. Linked caps
using a building blocks approach have been traditionally applied in the UK and Australia.
Unlinked caps do not link revenues to costs during the regulatory period and typically do not
require cost projections. Instead they apply a regulatory formula that annually adjusts the
allowed revenue whereas the starting point is based on the company’s actual cost in a prespecified
year. The regulatory formulas may include several components such as efficiency
improvement factors, an inflation index and quantity and quality terms. Unlinked caps have been
applied in Germany, Austria and Norway (before the start of yardstick regulation).
4 Regulatory Treatment of Investments
Investments may be conducted to meet service standards linked to technical and/or regulatory
requirements, for example the obligation to connect or to fulfil specific reliability standards.
Such investments may be related to network extension or to network replacement. 12 Network
extension investments are investments needed to meet the change in load and production patterns
12 Investments may also be driven by economic reasons. For example, a new transmission project can reduce
congestion and enhance market competitiveness by increasing both the total supply that can be delivered to
consumers and the number of suppliers that are available to serve load.
8

in the future. 13 Network replacement investments are investments related to the replacement of
(technically or economically) aged equipment. 14
Investment needs may also solely result from changes in legal obligations, for example, if new
labour safety rules require safety measures in substations or high voltage pylons, this may create
a need for investments. These investments neither lead to more capacity nor replace aged
components.
Network investments can be either assessed and included ex-ante or ex-post in the regulatory
asset base relying either on cost projections or actual investment levels. Applying a linked cap
for example requires an ex-ante assessment of the projected Capex level by the regulator,
whereas under an unlinked cap actual Capex is assessed ex-post, e.g. through benchmarking with
other regulated companies. The benchmarking model is applied to total costs and includes the
capital costs and operating costs. Moreover the benchmarking does not distinguish between
existing assets and investments but rather provides an integral assessment of the company’s
efficiency position. Both methods are discussed in detail further below.
4.1 Explicit Projection of Investment Costs
4.1.1 Ex-ante Review and Inclusion of Investments
Under this approach the regulator agrees ex-ante on the capital expenditures allowed to be
included in the regulatory asset base (RAB). It is usually used with linked caps based on building
blocks 15 . As previously explained, at the start of the regulatory period the company is asked to
provide the regulator with an overview of its intended investments during the next regulatory
period. Due to the information asymmetry the regulator does not accurately know the appropriate
amount of capital expenditure required by the regulated companies, and should rely on their
submission of this information. The regulated companies have incentives to inflate the reported
capital expenditures relative to their true cost by not quoting the best price offers or by simply
suggesting high work volumes. They may also try to strategically allocate operational
expenditure under Capex if the regulatory scrutiny for the latter cost category is less strict.
Furthermore, by capitalising Opex, the company can further inflate its RAB and consequently
earn higher returns.
13 Distribution network investments are mainly made for a specific customer or a group of customers, while
investments in transmission networks tend to result from an increasing load in a specific network area. The
transmission investment plans rely more on forecasts and less on individual applications for new connections.
Furthermore, the investments in transmission network extensions usually take longer to be completed than those for
distribution networks. Transmission projects are usually rather large which leads to lumpy increases in capacity.
14 Traditionally network equipment has been replaced after a certain period of time without taking into
consideration the conditions of the assets. The other extreme is replacing equipment only if it fails. For electricity
networks, this means in many cases either a direct interruption of supply or a decrease in the reliability in the
system. A third approach, called condition-based maintenance, is used to ensure a proper trade-off between the
maintenance and replacement strategies. According to this approach, both maintenance and replacement decisions of
equipment are no longer based on a pre-determined number of years but rather on the actual and historical
conditions of the equipment.
15 It has also traditionally been used with rate-of-return regulation.
9

The regulator should make a judgement of which investments are efficient and should be
included in the RAB. In order to inform his judgment, the regulator may use business plans,
comparisons against other regulated companies, engineers' reports and audits and cost-benefit
analysis. The problem of evaluating economic efficiency of investments is also complicated by
the possible implications for quality performance. Investments conducted at low costs may not
necessarily be desirable as they may deteriorate quality. Similarly, expensive investments may be
associated with an oversupply of quality.
Regulatory arrangements should include provisions dealing with a divergence between expected
and actual capital expenditure at each review. Depending on the specific design of the price
control, the regulator may decide to review the realised investments at the end of the regulatory
period (see the examples from Australia and UK below). If actual investments turn out to be
lower than the target, then prices are accordingly adjusted downwards. Similarly, no ex-post
allowances would be provided for investments in excess of the target. Alternatively, the regulator
could impose a band of desired investment levels with a minimum and a maximum target;
investments exceeding this band would not or only partially be allowed into the RAB. This
approach however comes at the cost of weaker incentives for efficiency increase on the Capex
front. The regulator would (partially) claw back cost savings, irrespective of whether these are
the result of strategic under-investing and/or deferring or due to genuine productivity
improvements. This may discourage the companies from achieving any productivity
improvements in the area of Capex as there would not be any financial rewards attached to this.
In Australia, as part of the Capex incentive framework, transmission or distribution network
operators are allowed to retain the benefit of lower expenditures (both depreciation and return on
assets) for the remainder of the regulatory control period, should it spend less than the allowance
set by the energy regulator ex-ante. Conversely, should a transmission or distribution network
operator exceed the allowance set by the regulator, it would forgo both return on assets and
depreciation associated with the over expenditure for the remainder of the regulatory control
period. The roll-forward model, in accordance with the regulatory guidelines and the electricity
rules, sets out the methodology allowing adjustments to incorporate differences between
estimated Capex and actual Capex in the previous period to arrive at the closing RAB that
becomes the opening RAB for the next regulatory control period.
In the UK, Ofgem does not approve ex-ante specific capital expenditures for the transmission
and distribution networks but rather identifies their total volume. This volume is used to assess
the capital costs in the allowed revenue for each network company. Once the price control has
been set, the network company is free to invest in whatever projects it considers will most
efficiently deliver the services needed by its customers. At the next price control Ofgem then
reviews the network company’s performance in terms of the efficiency of the investment
decisions. For distribution there is an automatic mechanism (Information Quality Incentive)
which rewards or penalises companies for the level of network expenditure they make relative to
a base level. In the future, companies' expenditures will be assessed ex-post in terms of the
delivery of outputs.
4.1.2 Supplementary Ex-Post Review
As explained, the ex-ante approach may encourage the regulated companies to over-claim
investments ex-ante and then to under-spend the allowed level. Therefore regulators may opt to
10

conduct an ex-post review that would allow them to assess the prudency of the actual
expenditure incurred, and prevent allowed but unused investment costs being rolled into the
RAB. Ex-ante reviews do not provide the regulator with the necessary information to assess
whether the regulated company has used its investment program properly. Assessing the actual
investment needs ex-post provides the regulator with a clear understanding of which investment
costs are really needed by the company to deliver the regulated services.
The existence of ex-post reviews forces companies to be diligent in assessing the efficiency of
their planned investments, and to maintain proper records to show that the investments have been
undertaken with due consideration. Ofgem in the UK introduced menu regulation in the last price
control, which aims to combine incentives for companies to accurately forecast Capex ex-ante,
while spending it efficiently ex-post.
4.1.3 Regulatory Tests
Regulatory tests have traditionally been applied in the USA in an environment of rate-of-return
regulation and in Australia in an incentive regulation environment. These tests provide the
regulator with a set of criteria that can be used to assess whether a proposed investment should
be included in the RAB. The tests can also be used on an ex-post basis aiming to encourage
network operators to carry out such a test themselves so they tend to only undertake investment
that will be later deemed on an ex-post basis to pass the test and be included in the RAB.
The regulatory tests are usually based on a cost-benefit analysis focusing on the incremental
costs and benefits of the investment project. The cost-benefit analysis studies should ideally
consider the impact on welfare of all parties (network companies and other stakeholders)
affected by the project. These parties can be located in one or several regions, such as in the case
of interconnection projects. The major objective of the cost-benefit analysis is to study the
economic efficiency of the project, i.e. the welfare impact. If the total welfare is maximised (i.e.
the project will bring maximal net benefits), then society as a whole will be better off as a result
of the project.
Examples of incremental costs incurred by the regulated network company include the
investment and operating costs of the project itself over its operating life and any associated
costs. The incremental benefits incurred by the regulated network company may include revenue
increases (e.g. congestion rentals) and avoided future investment and operating costs.
Examples for incremental costs to other sector stakeholders are additional investment and
operating costs in generation facilities, price increases for consumers (i.e. a decrease in consumer
surplus), revenue losses for generators due to price reductions and decreased sales (i.e. a
decrease in producer surplus). Examples for incremental benefits to third parties are avoided
investment and operating costs in generation facilities, reduction of outages, 16 price decrease for
consumers (i.e. an increase in consumer surplus), revenue gains for generators due to price
increase and additional sales (i.e. an increase in producer surplus).
16 For example an electricity transmission project could potentially enhance system reliability by reducing loading
on parallel facilities, especially under outage conditions. At the regional area level, the expansion of the major
interconnection may also improve the overall system reliability and reduce the loss-of-load probability.
11

External effects can also be included in the cost-benefit analysis. For example an expansion of
the transmission system can bring substantial environmental benefits by avoiding air emissions
otherwise caused by local generation and by reducing the need to procure local air offsets needed
for generation. 17 If the environmental costs are internalised (i.e. included in the investments costs
and in the operation and maintenance costs) the benefits will be automatically accounted for via
the avoided costs. If it is not the case, they should be additionally quantified provided that
sufficient information is available.
4.2 Hindsight Efficiency Analysis Using Total Costs
Under this approach, the problem of investment assessments is effectively bypassed. The
regulator does not need to form an opinion on whether a given investment proposal should be
allowed or not. Rather, the regulator considers the actual total costs (including the capital costs
of new investments) incurred by the network operator and sets the efficiency increase factor
based on an efficiency assessment (benchmarking) of these costs.
The efficiency assessment can be made by comparing the existing companies and using
traditional benchmarking techniques, such as Data Envelopment Analysis (DEA), Corrected
Ordinary Least-Squire (COLS) or Stochastic Frontier Analysis (SFA). When the data sample is
small or incomplete, such techniques cannot be applied effectively. Engineering network models
provide an alternative approach that can be used to fill the gap that traditional benchmarking
techniques cannot cover. In contrast to the traditional techniques, engineering network models do
not rely on existing alternative options, but rather create these options on the basis of
predetermined economic and engineering criteria.
The efficiency incentives of this approach come from the fact that in each regulatory period, the
efficiency increase requirements are set on the basis of performance achieved in previous years.
If the firm manages to increase productivity, its efficiency score will be higher in future periods
and consequently the required efficiency increase will be lower. An additional advantage of this
approach is that it provides an incentive to the service provider to be indifferent to the mix of
inputs and to deliver its required output at the lowest total cost.
The threat that capital costs of investments may be rejected, or partially disallowed, in the
process of benchmarking provides an incentive to the regulated company to only undertake
efficient investment. Such an incentive is considered necessary because the regulated company is
likely to hold better information than the regulator about the prospective efficiency of a proposed
investment. Therefore, by making the company accept the consequences of its investment
decisions, the probability that inefficient investment will take place is weakened. On the other
hand, the regulatory threat that capital costs of investments can be disallowed due to the ex-post
benchmarking could discourage regulated companies from implementing even good investment
projects. Also, there may be capital expenditure that is planned and conducted in good faith that
eventually proves “imprudent” on an ex-post basis. Moreover, the credibility of the efficiency
analysis depends on the data quality, adequacy of the model specification and the robustness of
17 Such reductions may also have secondary effects. They may assist in allowing new industries with higher
economic value to enter the local area by avoiding negative impacts to the local water and natural gas supplies
otherwise required for local generation. Also transmission upgrade may reduce the construction of additional
infrastructures such as gas pipelines, pumping stations, and water and waste treatment systems.
12

the selected benchmarking techniques. 18 Benchmark studies do not generate the absolute truth,
but rather an indication and ranking of relative efficiency levels.
5 Supplementary Regulatory Arrangements to Encourage Investments
5.1 Quantity Adjustment Factors
New investments involve capital costs such as depreciation and return on invested capital. The
majority of such costs are already taken into account by updating the cost base for the revenuecap
periodically. Cost recovery is, however, delayed in time because updates do not occur
continuously. In addition, new investment may have an impact on operation and maintenance
cost. Such (arguably positive) changes will not result in changed allowed revenues for the firm
until the next regulatory review.
This entails that the net present value of the implied revenues may be lower than what would be
necessary to cover new capital costs incurred today. The main purpose of quantity adjustment
factors in the regulatory formulas is to provide continuity in terms of investment (and Opex)
recovery. These factors link the allowed revenue to demand and possibly one or more other cost
drivers, e.g. customer numbers or length of network. The main purpose of linking the allowed
revenue to these variables is to consider customer driven changes in company’s costs, like the
costs of extension (load-related) investments that may not be reflected in the regulatory formula.
The advantage of this type of adjustments scheme is that total revenue can track total costs more
closely, thus reducing the risk of persistent losses for the regulated firm. In ensures additional
revenues when the quantities increase according to the pre-determined mathematical formula.
The allowed revenue moves more closely in line with costs, assuming the cap formula is
designed to mimic the cost function of the regulated service, and the financial risk borne by the
regulated business is lowered, since profits move more closely in line with volumes.
A typical application of this approach is the current regime in Germany where the allowed
revenue in the regulatory period is decoupled from the actual costs but adjusted for changes of
several pre-specified cost drivers. Also in Australia the increase in energy, load transmitted and
connected customers are drivers for an augmentation of Capex. The effects of the Capex drivers
are considered in the Capex forecast rather than explicitly in the regulatory formula.
In the UK the allowed transmission revenue is based on the assets employed by the transmission
company plus allowances for predicted future Capex and Opex. The future Capex and Opex
consider the energy transported, peak demand and other factors. There is a system of "Revenue
18 One of the biggest challenges in the application of total cost benchmarking is the incorporation of the capital
costs into the efficiency analysis. A number of issues resulting from the long-term nature of Capex should be
addressed in order to ensure the comparability of capital costs of the investigated networks. Notable examples of
challenges to ensure Capex comparability relate to differences in depreciation policy, capitalisation policy and
network asset age of the regulated companies. Investments are typically undertaken at different time intervals and
tend to vary considerably in size. Investment lumpiness might be characterised by substantial fluctuations in cash
spending from year to year, which could lead to misleading results in the benchmarking. Averaging Capex spending
for a number of years can partially smooth out the figures, but will not completely account for differences, in
particular when companies turn out to be at different stages of their investment cycles.
13

Drivers" that will adjust the revenue restriction depending on whether there are significant
unexpected changes to the level of generation connecting in different parts of the system. This
uses a system of "Unit Cost Allowances" (UCA) that will provide £m (million pounds) for
incremental capacity changes of "Y" MW. The UCA is set with reference to the likely level of
investment needed reflecting the balance of generation and demand in the zone.
5.2 Quality Incentive Schemes
Capital costs generally form a substantial part of the firm’s total costs and investment decisions
have a significant impact on the network’s quality. Assuring that investments are undertaken at
least costs and deliver an adequate quality level can generate significant benefits to society. The
ability to effectively measure the performance of investment proposals for both price and quality
is therefore an important regulatory task.
Regulators have developed different methods to encourage quality of supply. The traditional
method is based on quality performance standards. Standards put a floor on the performance
level of the company measured by selected quality indicators. Violation of the standard can lead
to a fine or tariff rebate. Examples of such quality performance standards refer to customer
minutes lost, percentage of customers with an outage, response to customer requests and
complaints.
Quality incentive schemes can be considered an extension of a quality performance standard.
Alternatively, a standard can be considered as a special case of a quality incentive scheme.
Under quality incentive schemes the company’s performance is compared to some target defined
for a certain quality indicator (e.g. reliability measured by SAIDI or energy not supplied) or a
combination of indicators. Deviations from the target result in either a penalty or a reward.
5.3 Cost of Capital
Regulated service providers compete for finance with companies operating in competitive
markets and thus rely on the same financial market conditions. Equity and debt finance will only
be available to utilities who agree to credit conditions posed to firms that operate in competitive
industries and have a comparable credit ranking. Equity finance will only be available if
profitability can be expected that covers the risk free rate of interest (i.e. yield of long term
credible government bonds) and a risk premium.
Given the capital-intensive nature of electricity networks, the return on asset accounts a
significant share of the allowed revenue. As relatively small changes to the rate of return can
have a significant impact on the total revenue requirement and investment behaviour of the
companies, it is essential that the regulator sets the rate of return at a level that reflects an
adequate commercial return for the regulated companies. The Weighted Average Cost of Capital
(WACC) is a commonly used method for determining a return on an asset base. It is generally set
equal to the sum of the cost of each individual component of the capital structure weighted by its
share. 19
19 The WACC calculation will look different depending on how taxes are treated in the revenue requirements. A
post-tax WACC is defined as the average rate of return needed to provide an appropriate return to investors in the
14

5.4 Construction Work in Progress
Construction work in progress (CWIP) is the amount of money already spent on an asset that has
not yet been commissioned at the relevant time. Due to the large size of investments and long
construction time, in particular for transmission, most regulators do apply specific arrangements
for CWIP. For instance, the regulator may allow capitalisation of debt and equity costs incurred
by the service provider during the construction period. Alternatively the regulator can permit the
inclusion of the cost of capital (allowed return on debt and equity) but not the depreciation in the
allowed revenue during the construction period. Finally, the regulator may decide to partially
include CWIP in the RAB, for instance for projects with a short construction time.
5.5 Explicit Investment Allowances
We explained in the previous sections that the unlinked caps do not consider the capital cost and
Opex arising from new investments unless they occur in a base year when a new regulatory
period starts. In several countries applying unlinked caps (e.g. Germany, Austria) there had been
significant debate on the extent to which this regulation actually hinders new investment, in
particular investment in network extension, as almost all additional costs incurred by the
regulated company are not considered in the allowed revenue. The rationale for the introduction
of explicit investment allowances is therefore to overcome these issues by allowing for the
explicit consideration of the capital cost of investments in grid extensions/expansions in the
allowed revenue during the regulatory period.
A typical example is the application of investment budgets in Germany in the transmission price
control. Once investment costs have been approved by the regulator, the respective capital costs
are considered as permanently non-controllable costs for the duration of the approval period.
They are not subject to the efficiency improvement targets (X-factor). Therefore, during the
approval period of the respective investment budget project, the assets included in the investment
budget asset are rolled forward outside of the general asset base. Once the investment budget
rolls into the general asset base, the efficiency increase, productivity factor and inflation will
apply according to the regulatory regime.
5.6 Efficiency Carry-Over Schemes
Efficiency carry-over means how the gains resulting from efficiency increases (i.e. a saving in
Capex) are transferred between two regulatory periods and how much time is allowed for
company concerned, it assumes that the company’s business tax has already been paid, i.e. has been included in the
revenue requirements. As the tax shield is fully considered through adjustment of debt-component, the taxes
reimbursed through the allowed revenue are calculated without any tax shield. Some regulators apply a “Vanilla
WACC” which does not adjust debt and equity returns for taxes. Similarly to the post-tax, WACC taxes are included
in the revenue requirements, however they are calculated with tax shield (including deductible interests). In contrast,
a pre-tax WACC is the average rate of return needed to provide an appropriate return to investors in the company
concerned and pay the company’s business tax. In order to calculate a pre-tax WACC, the estimate of the post-tax
cost of capital needs to be increased, by dividing by “(1-tax rate)”, so that the tax payments can be met from the pretax
WACC.
15

etaining these gains. Sharing schemes have been considered in Australia and the UK in the
context of the building blocks approach used in these countries.
The length of the efficiency carry-over reflects the implied benefit sharing between regulated
service provider and network service users. A long term carry-over mechanism provides a
forward looking framework within which the regulated service provider’s actual costs in future
regulatory periods will provide a better indicator of efficient cost levels. On the other hand,
containing the carry-over not only in the following (second) regulatory period but also in the
subsequent (third) one would increase complexity and delay the process of transferring the
benefits to the final consumers.
The carry-over mechanism should provide sufficient incentives for regulated network companies
to pursue efficiency gains. That is, the benefit that regulated network companies retain at the
margin should outweigh the cost of the efficiency improvement. There is no universal formula
for optimal sharing of gains. The optimal relationship between gains retained and efficiencies
achieved depends on the underlying assumptions regarding the responsiveness of the regulated
companies (in terms of cost reduction and innovation) to changes in the share of efficiency gains
they retain.
It could be argued that the greater the share of the benefits regulated network companies are
allowed to retain, the greater their incentive to make efficiency savings and, hence, the greater
the extent of savings to be eventually passed on to consumers. On the other hand, it is evident
that the greater the share that regulated companies are allowed to retain, the longer consumers
will have to wait before the benefits from efficiency savings are passed through to them.
Furthermore, when deciding on the amount of cost reductions to be carried-over, it is difficult for
the regulator to determine which cost reductions are a consequence of management decisions and
which are windfall profits, resulting for example from changes in input prices. 20
20 There are a range of factors that could cause the regulated company to earn a different rate of return from that
forecast at the initial price review. Some of those factors may be within the control of the regulated company, and
others may be external to them. In principle, there are sound reasons to treat unexpected profits differently,
depending on their source. Some profits may result from influences and events external to the regulated company
(i.e. windfall gains) rather than from its direct actions. It follows that they can be shared with customers at the next
review without having any adverse impact on the incentives of the regulated service provider to pursue efficiencies
that are within their control. Conversely, allowing the regulated company to keep the additional profits that arise
from actions under their control (i.e. controllable gains) for a longer period should give them a strong incentive to
continue making such efficiency gains in the future. A typical example of controllable gains is when the regulated
company may have reduced actual expenditure below that forecast via more efficient use of labour, maintenance and
capital expenditure, and/or by introducing more efficient systems for operations.
Additional profits may result from factors that are not within the control of the regulated company. For example,
general economic growth may have been greater than forecast, or market interest rates may have fallen below those
underlying the original forecasts, and this may have increased revenues or reduced financing costs. While the
principle of differentiating between potential sources of increased profitability appears sound, it may be difficult to
distinguish with any precision those different sources in practice. A significant degree of regulatory judgment will
be required in order to distinguish the different reasons for any gap between the returns incorporated at the initial
revenue setting and those now being achieved by the regulated service provider.
16

5.7 Sliding-Scale Schemes
Sliding-scales, or sharing mechanisms as they are also called, set incentives for network
operators to achieve specific regulatory targets by splitting the benefits and costs of over- or
under-achieving these targets between the company and the network users (the customers)
according to a pre-defined rule. Under such schemes the regulator sets a specific target level for
particular cost items such as investment costs, Opex reductions or quality of supply
improvements ex-ante. If the network operator is able to achieve this target level within a predefined
range, benefits or costs of this item are recognised by the regulator to their full extent. If
this range is over- or under-achieved, benefits or costs are shared between the network operator
and its users. Often these sharing levels are complemented by a maximum and minimum level
(cap or floor) above or below which all costs are covered by the network operator or its users
respectively.
Sliding-scale schemes can help to overcome the informational advantage of the network operator
over the regulator, like in the case of linked caps where the regulator relies on the investment
forecasts by the network operators. As sliding-scale schemes would punish the network operator
for deviations of actual from ex-ante predicted cost levels, these schemes would encourage the
network operator to submit realistic investment forecast.
Sliding-scales mechanisms have been widely applied for transmission and distribution price
control in the UK. Electricity distribution network operators for example are required to predict
their operating costs and network investments for each year of the regulatory period and submit
these figures to the regulator, which are then used to determine the company specific revenue
caps for each year. At the end of the regulatory period, distribution network operators are
required to demonstrate that their actual level of network investment in asset replacement and
general reinforcement meets the agreed ex-ante targets. Any deviations of actual cost levels,
observed ex-post by the regulator, that lie outside a given range, have to be shared equally by the
network operator with its customers over the next regulatory period.
Sliding-scale schemes in the UK are also set for a number of other output parameters. Revenues
of the UK electricity transmission system operator National Grid are linked to sliding-scale
schemes for system availability and for system operation costs. Distribution network operators
also receive penalties or rewards for the over- or under-achievement of targets for network
losses, customer satisfaction levels and other measures of service quality.
Sharing mechanisms in the form of a penalty-reward scheme have also been applied for the
NorNed cable, a high voltage DC submarine power cable between Norway and the Netherlands.
Targets set in the regulatory approval of the cable by the Dutch regulator included the overall
investment costs, the date the cable went into operation and the availability of the cable. Sharing
mechanisms have also been popular in the US telecommunication regulation of the 1990s, where
they have also been used in the context of rate-of-return regulation.
17

6 Conclusion
Reliable and efficient electricity networks are essential for the successful performance of
electricity markets. Increasing cross-border trade and the integration of renewable energy place
significant pressures on the existing network infrastructure and require new investments in
network capacity. Investment incentives are however strongly influenced by the regulatory
arrangements. Failure to include adequate capital costs in the allowed revenue risks a reduction
in investments and deterioration in the quality of supply. On the other hand, including inefficient
investment costs in the allowed revenue leads to high network charges, distorted price signals
and hinders the effective market performance. It is a major task of regulators to establish the
right balance through transparent, consistent and predictable regulatory arrangements.
Regulators have been using different mechanisms to encourage investments. Some of them are
assumed in the general properties of the regulatory model or in the policy to set an adequate rate
of return and rules to account for construction work in progress. The theoretical literature and
practical experiences show that rate-of-return regulation can lead to high and inefficient levels of
investments. Incentive regulation, mainly using caps, has been applied in several countries to
encourage efficient investments. Caps can be divided into two major groups: linked and unlinked
caps.
Linked caps (building blocks) base the revenues of a regulated company during a regulatory
period on an ex-ante assessment of the efficient levels of operating and capital expenditure.
Linked caps appear attractive because they link revenues to the projected costs. At the same time
they allow the consideration of efficiency increases (the projected costs are checked for
efficiency) and allocate the anticipated efficiency increases to customers. Where there are strong
investment and maintenance needs in the near future, these types of caps may be suitable as they
enable a specific allowance to be made for the higher investment and maintenance. However,
under linked caps, regulated companies may have strong incentives to inflate the planned capital
expenditures relative to their true cost by not quoting the best price offers or by simply
suggesting high work volumes. The major challenge for the regulator is to provide a judgement
of which investments are efficient and should be included in the regulatory asset base. In order to
inform their judgments, regulators have been using business plans submissions, comparisons
against other regulated companies, engineers' reports and audits and cost-benefit analysis.
In practice, actual investments at the end of the regulatory period can turn out to be different
from the allowed level. Regulatory arrangements should include provisions dealing with a
divergence between expected and actual capital expenditure at each review. Regulators should
not seek for claw backs and should allow network companies to retain efficiency gains during
the regulatory period, or for longer periods of time using efficiency carry-over schemes.
However, regulators should be concerned with understanding that the saving of projected capital
expenditures does not result from deliberate investment deferrals. Such gains are associated with
gaming behaviour and do not represent eligible benefits resulting from management efforts
towards efficiency increase.
18

Regulators may apply incentive arrangements (such as sliding scales) to discourage companies
from inflating their capital expenditures forecast above the efficient level and to underspend this
forecast later on. Based on an ex-ante assessment of investments, regulators may impose a band
of desired investment levels with a minimum and a maximum target; investments exceeding this
band would not or only partially be allowed into the regulatory asset base. Alternatively
regulators may also apply rewards and penalties to companies for gaps between planned and
actual investment levels.
Unlinked caps do not link revenues to costs during the regulatory period and typically do not
require cost projections. Instead they apply a regulatory formula that annually adjusts the
allowed revenue whereby the starting point is based on the company’s actual cost in a prespecified
year. The regulator does not need to form an opinion on whether a given investment
proposal should be allowed or not. Rather, the regulator considers the actual total costs
(including the capital costs of new investments) incurred by the network operator and sets the
efficiency increase factor based on an efficiency assessment (benchmarking) of these costs.
Regulated companies have to take into account at the time of decision making that their
investments will be included in the regulatory benchmarking at a later stage. This provides the
regulated companies with a strong incentive to undertake only efficient investment. On the other
hand, the regulatory threat that capital costs of investments can be disallowed ex-post could
discourage even efficient investment projects and endanger network reliability in the middle und
long term. Moreover, the credibility of the efficiency assessment depends on the data quality,
adequacy of the model specification and the robustness of the selected benchmarking techniques.
Benchmark studies do not generate the absolute truth, but rather an indication and ranking of
relative efficiency levels. For these reasons unlinked caps have been complemented by quantity
adjustment factors and quality incentive schemes. Moreover some regulators, for example in
Germany and Austria, introduced explicit investment allowances for extension investments and
incorporate their costs in the allowed revenue during the regulatory period.
19

Bibliography
AI, C. AND SAPPINGTON, D. (2002), “The Impact of State Incentive Regulation on
the U.S. Telecommunications Industry”, Journal of Regulatory Economics, 22(2), 107-132.
ARMSTRONG, M., COWAN, S., VICKERS, J. (1995) Regulatory reform: Economic analysis
and British experience, Cambridge: MIT Press.
BALDWIN, R. AND CAVE, M. (1999) Understanding regulation: theory, strategy, and
practice. Oxford; New York: Oxford University Press.
BERNSTEIN, J. AND SAPPINGTON, D. (1999) "Setting the X factor in price-cap regulation
plans." Journal of Regulatory Economics, 16(1), pp. 5-26.
CABRAL, L. AND RIORDAN, M. (1989) "Incentives for cost reduction under price cap
regulation." Journal of Regulatory Economics, 1, pp. 93-102.
CAMBINI, C., AND RONDI, L. (2010) “Incentive regulation and investment: evidence from
European energy utilities.” Journal of Regulatory Economics, Vol. 38, issue 1, pages 1-26.
CREW, M. AND KLEINDORFER P. (2002) “Regulatory Economics: Twenty Years of
Progress?”, Journal of Regulatory Economics, 21(1): 5-22.
ÉGERT, B. (2009) “Infrastructure Investment in Network Industries: The Role of Incentive
Regulation and Regulatory Independence.” OECD Economics Department Working Papers 688.
ELLIOTT, D. (2006) "Regulating prices and service quality," in Crew und Parker (eds.),
International handbook on economic regulation, Cheltenham: Edward Elgar Publishing.
GREENSTEIN, S., MCMASTER, S. AND SPILLER, P. (1995) “The Effect of Incentive
Regulation on Infrastructure Modernization: Local Exchange Companies' Deployment of
Digital Technology”, Journal of Economics & Management Strategy, 4(2), 187-236.
JOSKOW, P. (2008) “Incentive Regulation and Its Application to Electricity
Networks,” Review of Network Economics, 7(4): 547-560.
KWOKA, J. (2009) “Investment Adequacy under Incentive Regulation”, mimeo, Northeastern
University.
LISTON, C. (1993) "Price-cap versus rate-of-return regulation." Journal of Regulatory
Economics, 5, pp. 25-48.
SHLEIFER, A. (1985) "A theory of yardstick competition." RAND Journal of Economics, 16(3),
pp. 319-27.
20

TANGERÅS, T. (2002) "Collusion-proof yardstick competition." Journal of Public Economics,
83, pp. 231–54.
YATCHEW, A. (2001) "Incentive regulation of distributing utilities using yardstick
competition." The Electricity Journal, 14(1), pp. 56-60.
VISCUSI, K., HARRINGTON, J. AND VERNON, J. (2005) Economics of regulation and
antitrust, Cambridge, Massachusetts: MIT Press.
VOGELSANG, I. (2002) “Incentive regulation and competition in public utility markets: A 20-
year perspective." Journal of Regulatory Economics, 22(1), pp. 5-27.
VOGELSANG, I. (2010) "Incentive Regulation, Investments and Technological Change,"
CESifo Working Paper Series 2964.
WEYMAN-JONES, T. (1995) "Problems of yardstick regulation in electricity distribution." in
Bishop and Mayer (eds.), The regulatory challenge, Oxford: Oxford University Press, pp. 423-
443.
21