Restructuring and performance in India's electricity sector

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Request accessible transcript

Transcript (if available)

Content RESTRUCTURING AND PERFORMANCE IN INDIA’S ELECTRICITY
SECTOR
Copyright 2002
by
ARUN KUMAR PANDA
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(POLITICAL ECONOMY AND PUBLIC POLICY)
August 2002
Arun Kumar Panda
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UMI Number: 3094364
UMI
UMI Microform 3094364
Copyright 2003 by ProQuest Information and Learning Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
ProQuest Information and Learning Company
300 North Zeeb Road
P.O. Box 1346
Ann Arbor, Ml 48106-1346
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UNIVERSITY OF SOUTHERN CALIFORNIA
The Graduate School
University Park
LOS ANGELES, CALIFORNIA 90089^1695
This dissertation, w ritten b y
A run kuMAK P a NDA
U nder th e direction o f h.i&... D issertation
Com m ittee, an d approved b y a ll its m em bers,
has been p resen ted to an d a ccep ted b y The
G raduate School, in p a rtia l fu lfillm en t o f
requirem ents fo r th e degree o f
DOCTOR OF PHILOSOPHY
y z —
• ar t o f G raduate S tu d ies
D ate
^ C O - z
DISSER TA TION COMMITTEE
Chairperson
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
DEDICATION
To my wife, Rita, and our two sons, Avishyant and Anwesh.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ACKNOWLEDGEMENTS
I am grateful to my dissertation chair, Professor Jeffrey Nugent, and committee
members, Professor Laurie Brand and Professor Charles Cicchetti for their excellent
guidance and unstinted support. Their timely suggestions and invaluable feedback
helped shape this work.
I am also indebted to my friends and colleagues in India, especially Mr. Satya Sahu,
Mr. Amulya Patnaik, Dr MS Panigrahi and Mr. IR Khan. Without their help in
getting data, this study would not have been possible.
I take this opportunity to acknowledge the unflinching help and support, received
from Dr Farideh Motamedi, Associate Director of the Program of Political Economy
and Public Policy in USC.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CONTENTS
Dedication ii
Acknowledgements iii
List of Tables v
List of Figures vi
List of Acronyms vii
Abstract x
Chapter 1: Introduction, Structure and Regulation of Electricity Sector, and
Purpose of the Study 1
Chapter 2: Literature Review, Political Economy of Regulation,
and Empirical Studies on Performance 38
Chapter 3: India’s Electricity Sector, the Orissa Case, and the International
Comparison 70
Chapter 4: Data, Variables, Model for estimation, and Empirical Issues 122
Chapter 5: Analysis of empirical results regarding the effects and possible
determinants of restructuring 149
Chapter 6: Conclusions, Policy Implications, and suggestions for future research 168
References 182
Appendix-A 190
Appendix-B 207
iv
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LIST OF TABLES
3.1: Rate of Return on Capital (without subsidy) in India’s Electricity Boards 75
3.2: GRIDCO’s Initial Loss Reduction Target (in %) 84
3.3: GRIDCO’s Revised Loss Reduction Target (in %) 85
3.4: Profit/Loss of Distribution Companies before Tax (in crores of rupees) 87
3.5: Profit/Loss of various units in crores of rupees 87
3.6: Average Industrial and Residential Electricity Prices in Hungary (’91-’ 00) 116
4.1: Political Variable, POL 124
5.1: Share of India’ Agricultural Consumption in Total Sale (in %)(1996-97) 161
5.2: Per Capita Net Domestic Product of Indian states at Current Prices(96-97) 162
5.3: Unit Cost of Supply of Electricity in Paise/kWh for 1997 164
5.4: Rate of Return on fixed capital (without subsidy) for Indian SEBs for ’97 165
v
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LIST OF FIGURES
1.1: Structure and Ownership Pattern of electricity industry in some developed
countries 18
3.1: Structure of the Electricity Sector in Orissa 83
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LIST OF ACRONYMS
ADB Asian Development Bank
AP Andhra Pradesh
APSEB Andhra Pradesh Electricity Board
APDP Accelerated Power Development Programme
AVT Average Tariff Collection
BJP Bharatiya Janata Party
BLUE Best Linear Unbiased Estimator
CCGT Combined Cycle Gas Turbine
CEA Central Electricity Authority of India
CEGB Central Electricity Generating Board
CESCO Central Electricity Supply Company
CIDA Canadian International Development Agency
CNE Comision Nacional de Energia
CPI Consumer Price Index
CROSS Cross Subsidization
CV Covariance Estimator
DISTCO Distribution Companies
DEA Data Envelopment Analysis
DFID Department for International Development
DMU Decision Making Unit
ECNZ Electricity Corporation of New Zealand
EPS Electric Power Survey
EMKWH Employees per Million kWh electricity Sold
ESI Electricity Supply Industry
ETHCON Employees per Thousand Consumers
FE Fixed Effects
FCOST Fuel Cost
FOROUT Forced Outage
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
GDP Gross Domestic Product
GLS Generalized Least Squares
GW Gigawatt (= 1000 MW)
GRIDCO Grid Corporation of Orissa
HYDEL Share of Hydroelectric power in Total Generation
HT High tension
IPP Independent Power Producers
KV Kilovolt
KW Kilowatt
KWH Kilowatt Hour
LT Low Tension
LVTOT Share of Low Voltage supply in total sale
MEH Hungary Energy Office
MNC Multi National Corporation
MP Madhya Pradesh
MTBF Mean Time Between Failures
MVM Magyar Villamos Muvek Rt
MW Megawatt (=1,000 kW)
NDA National Democratic Alliance
NESCO Northern Electricity Supply Company
NHPC National Hydroelectric Power Corporation
NPP Norwegian Power Pool
NTPC National Thermal Power Corporation
OECD Organization for Economic Cooperation and Development
OERC Orissa Electricity Regulatory Commission
OHPC Orissa Hydro Power Corporation
OLS Ordinary Least Squares
OMC Operation and Maintenance Cost
OPGC Orissa Power Generation Corporation
OSEB Orissa State Electricity Board
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
PAISE (1 Indian Rupee =100 paise)
PHRD Policy and Human Resources Development
PLAVL Plant Availability
PLF Plant Load Factor
POWCOST Power Purchase Cost
PPA Power Purchase Agreement
PPI Private Participation in Infrastructure
PPP Purchasing Power Parity
RE Random Effects
ROR Rate of Return
SAR Staff Appraisal Report
SEB State Electricity Board
SERC State Electricity Regulatory Commission
SOUTHCO Southern Electricity Supply Company
SREV Sales Revenue as a ratio of Cost
TFP Total Factor Productivity
T&D Transmission and Distribution
TPA Third Party Access
UCOST Unit Cost of Supply of Electricity
UP Uttar Pradesh
URRATIO Ratio of Urban Population in Total Population
USAID United States Agency for International Development
WESCO Western Electricity Supply Company
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ABSTRACT
RESTRUCTURING AND PERFORMANCE IN INDIA’S ELECTRCITY SECTOR
Restructuring and privatization, used as major tools in electricity sector reform, are
often viewed as part of the same process and the terms used interchangeably.
Although related, they represent quite different dimensions of change and reform.
Privatization is the result of change in the management/ownership. Restructuring, on
the other hand, refers to changes in structure such as the unbundling of vertically
integrated utilities, and the introduction of competition.
Most studies attempt to assess the impact of privatization of the electric utilities on
their tariff structure, performance and efficiency. They have not tried to estimate the
effect of restructuring on the performance of the unbundled utilities. Using panel
data on the state electricity boards and the thermal power plants, and employing
variance-component fixed effects and random effects models, this study examines
the effects of restructuring and ownership on the performance of India’s electricity
sector. We also study the effects of absolute majority of political parties on
performance. The study also uses a cross-country-comparison-framework to compare
the electricity sector reforms of India with those of Chile, Hungary and Norway.
x
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Results show that restructuring has significantly positive effects on such
performance indicators as plant availability, plant load factor, forced outage, average
tariff collection, and sales revenue as a ratio of cost. With regard to labor efficiency
indicators, we find mixed results. Restructuring also appears to entail reduction in
the extent of cross-subsidization. However, the cost of supply seems to be unaffected
by restructuring. Absolute majority of the party in government shows adverse effects
on costs, sales revenue as a ratio of cost, and labor efficiency. The effects of
ownership are somewhat mixed, with state ownership (as opposed to federal or
private) indicating adverse effects on plant performance. Interestingly, after
controlling for location-specific effects, we do not find significant difference
between privately owned plants and other plants in areas like plant availability, and
plant load factor. In a developing country like India with a long tradition of public
ownership and vertical integration in electricity sector, this has important policy
implications.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 1: INTRODUCTION, STRUCTURE AND REGULATION OF
THE ELECTRICITY SECTOR, AND PURPOSE OF THE STUDY
The electricity sector plays a crucial role in the growth process. Although itself
highly capital-intensive, this sector has been able to generate employment directly
and indirectly through its backward and forward linkages. Moreover, the availability
of reliable supplies of electricity has facilitated the adoption of modem technologies
in most sectors, raising productivity and contributing substantially to long-term
economic growth. After independence, most developing countries have tried to boost
power production through the establishment of a state-owned monopoly that
manages all three subsectors of electricity, namely, generation, transmission and
distribution. These vertically integrated monopolies in the public sector have been
major claimants of state funding and have contributed to the increasing debt burden
in most developing countries.
Recently, however, this pattern has started to change. For several important
reasons, the continuation of electricity production and distribution in this manner
appears no longer sustainable. First, growth in the demand for electrical energy has
been accelerating. Second, most countries have been facing a supply side problem. In
particular, most of these public sector utilities have been performing poorly, with
inefficient, inadequate and unreliable supply, and have sustained huge losses in
distribution. Third, even aside from short-term supply problems, the governments
have been experiencing accelerating problems in financing additions to increasingly
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2
capital-intensive productive capacities. The increasing debt burden and the demand
for higher resource allocations by the social sector (like health and education) are
making it increasingly difficult to fund such projects. Fourth, technological advances
have enhanced the potential for competition in generation by reducing the minimum
efficient scale. In particular, the combined cycle gas turbine (CCGT) technology has
greatly expanded thermal efficiency and reduced minimum efficient plant size. Fifth,
there is the shift in emphasis from an exclusively engineering/technical approach
(about production) to an economic/political approach with an emphasis on the
importance of institutions and the regulatory framework. As a result, these countries
have started to seek different ways to satisfy their power requirements at minimum
cost and to increase production and distribution efficiency in all three subsectors.
Restructuring and privatization offer the promise of being the major tools in this
reform effort. Different kinds of ownership structures and regulatory frameworks
have been advocated and tried in different countries with varying results.
Understandably, there seems to be no consensus on the specific regulatory
framework most likely to achieve the benefits of reforms.
Privatization and restructuring are too often viewed as part of the same process;
and for this reason the terms are often used interchangeably. Although related, they
actually represent quite different dimensions of change and reform. Privatization is
the result of a change in the management/ownership from state ownership to private
ownership, implying the reduction of state/government influence in the operation of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3
firms. Restructuring, on the other hand, refers to changes in structure such as the
unbundling of vertically integrated electric utilities, and the introduction of
competition, though not necessarily by private enterprises. Thus, privatization is
primarily about ownership rather than control, as utilities can face remarkably
similar regulation under public or private ownership. Restructuring, in contrast,
subjects utilities to market forces, and may induce even more dramatic changes in
performance than privatization.
Most existing studies consider restructuring as a component of privatization, even
though the two processes represent fundamentally different dimensions of reform.
These studies attempt to assess the impact of privatization of the electric utilities on
their tariff structure, performance and efficiency. They have not tried to estimate the
effect of restructuring on the performance of the unbundled utilities. Potentially,
however, restructuring and the unbundling of electrical utilities could have just as
profound effects and be easier to accomplish than privatization.
It is worthwhile to discuss the structure of the electricity sector and the need for
its regulation before examining the effect of restructuring on its performance.
THE STRUCTURE OF THE ELECTRICITY SECTOR
The electricity supply industry can be divided into three vertically related
businesses or subsectors — generation, transmission, and distribution. Generation is
the production of electricity. It involves the transformation of another form of energy
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4
into electrical energy. Electricity production may use coal, oil, natural gas, nuclear
power, hydro-power, renewable fuels, wind turbines, or photo-voltaic technologies.
The diversity of generating technologies and cost structures results in a “least-cost
merit order,” in which different kinds of generators are operated according to their
relative costs.
Transmission and distribution comprise the “wires” functions. Transmission is the
high-voltage transport of electricity. It involves not just the transportation of
electricity but also the management of dispersed generators in a grid to maintain a
suitable voltage and frequency and to prevent a system breakdown. Thus, a modem,
synchronously interconnected transmission system requires close coordination. In the
wires, electricity flows along the line of least electrical resistance, not along
contractual paths. This means that the technical problems one generator experiences,
or the transaction it carries out with a consumer, may affect third parties not involved
in the transaction. Thus, there are significant externalities involved in this process.
Transmission is considered a natural monopoly because competition in transmission
would result in duplication of the existing network.
Distribution is the low-voltage transport of electricity, generally from the
transmission system to the end-users. Distribution often shows strong economies of
density in urban conglomerations, but diseconomies in remote areas. Distribution is
usually considered a natural monopoly over given geographic areas.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5
Supply, which is often seen as a part of distribution, is the contracting for, and
selling of, electricity to end-users. It also includes metering and billing, and can
comprise activities such as customer information and advice. It is not a natural
monopoly and hence, is subjected to competition and outsourcing.
To give an idea of relative cost magnitudes, generation in the UK accounts for
about 65 percent of total cost of the supply chain, transmission, 10 percent;
distribution, 20 percent; and supply, 5 percent (Newbery and Green, 1996, p. 59).
The structure of the electric power sector that emerged historically around the
world has been driven largely by the operating and investment complementarities
between generation and transmission (G & T). As Joskow(1998a; pp.27) points out,
The economic rationale for vertical integration between G & T is that it internalizes the
operating and investment interrelationships between generation and transmission inside
public or private organizations where the potential public goods and externality problems
that arise as a result of the physical attributes of electric power networks, as well as the
challenge of coordinating operations in real time to adapt to changing demand and supply
conditions, can be solved with internal operating hierarchies rather than markets.
But as mentioned before, technological advances have augmented the potential for
competition in generation by reducing the minimum efficient scale. The development
of the combined cycle gas turbine (CCGT) greatly expanded thermal efficiency and
reduced minimum efficient plant size from 1000 MW in the early 1980s to between
50 and 350 MW (IEA, 1999). Furthermore, the low variable costs associated with
CCGT technology make it suitable for base load generation. CCGT plants also
require shorter construction and planning time horizons. At the same time, many
countries have focused on renewable sources and on projects that tend to be small
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6
and locally owned. All these developments have challenged the conventional
arguments and wisdom favoring the vertical integration and the internalization of the
operating and investment interrelationships between generation and transmission
within operating hierarchies.
REGULATION
By its very nature, electricity production and distribution may seem to require
government involvement. The projects need public land, and use natural resources
like rivers and minerals. Moreover, they give rise to emissions and effluents, which
are “public bads” that affect living and non-living entities. There are also other
features that merit mention:
• Capital intensity and economies of scale
• Non-storability with fluctuating demand
• Locational specificity generating locational rents
• Low price elasticity of demand (electricity is a necessity, and not a luxury good)
• Direct connections to customers
• Natural monopoly features (at least in transmission and distribution)
Regulation of the electricity supply industry (ESI) is primarily motivated by the
existence of natural monopoly conditions, externalities, and the public good
characteristics mentioned above. These result from a number of unique economic
characteristics such as the non-storability of electricity and instantaneous and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7
fluctuating demand. Externalities occur because the operations, functions and
malfunctions of each generator affect system conditions throughout the entire
interconnected network. Public good characteristics exist in the sense that additions
to a transmission network benefit all producers and consumers. The externality and
the public good aspects of electricity suggest the need for planning and coordination.
Regulatory institutions are usefully thought of as part of the “economic
institutions of capitalism,” to take the title of Williamson’s book (1985). Williamson
argues that transactions, rather than tastes or technology, are the basic unit of
institutional analysis. The key determinant of the evolutionary success of institutions
is their ability to organize transactions to economize on bounded rationality while
guarding against opportunism. The third aspect of transaction cost economics is the
concept of asset specificity. Regulation of network utilities has to deal with asset
specificity on the part of the utility, bounded rationality on the part of the regulator,
and opportunism by both parties (Newbery 1999).
Regulation is often justified as a policy instrument to minimize the effects of
market failures, i.e., the situations in which the firms operating in the market may
fail to maximize overall welfare. Armstrong et al. (1994) list three classes of market
failures: problems of asymmetric information, externalities and market power. Each
of these may justify the regulation of economic activities. Among all these
arguments, the market power argument is the strongest. It is usually believed that
regulation is necessary to prevent the abuse of monopolistic market power.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8
Demsetz (1968) argues against the regulation of monopoly after it has commenced
production. The problem of monopoly rents can be avoided by franchising the
service to the lowest bidder in the first place. The periodic auctioning of franchising
rights will ensure that the monopolist will be under pressure to reduce costs
continually to the optimal level.
The above argument of Demsetz gets into problems in some cases. The feasibility
of franchising in industries where huge fixed costs are involved and the assets have a
long economic life are noted as potential limitations of the franchising options in
case of industries such as electricity, gas and water supply (Domberger; 1986). There
could be other problems like collusive bidding, or simply the lack of bidders. In
these cases, regulation becomes necessary.
In the pre-liberalized, government-managed set up, government controls the
sector, decides on type and magnitude of investment, and sets the tariffs. But after
liberalization/restructuring, an independent commission determines the type, extent
and scope of regulation. The rationale for such independent regulatory commissions
has been to create expert regulatory bodies that follow well-defined public interest
principles and which are insulated from political pressures created by powerful
interest groups. As Joskow (1998a) puts it, complete insulation from political
pressures is, of course, impossible when regulators are appointed by government
officials, depend on government for the funds they need to perform their jobs
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9
effectively, and are ultimately subject to changes in the laws under which they
operate. Nevertheless, the open independent commission system places significant
constraints on special political deals and corruption because they are more difficult
to hide.
In order to work effectively, the independent regulatory mechanisms have to have
at least the following features. First, they should have clearly-specified and publicly-
known objectives. Second, the members should be selected for their knowledge in
economics, law, technology, and possible scope for competition. Third, autonomy
for the regulatory board/commission must be clearly enshrined in the legislation.
Finally, these commissions should evince transparency in their decisions.
An independent regulatory framework is usually recommended in the post
liberalized power sector to correct any possible market failures. But there can be
regulatory failure, if the regulators suffer from excessive informational asymmetry,
or are “captured” by the industry, or try to bring in too much regulation, thereby
defeating the very purpose of restructuring/ privatization. Stigler and Friedland
(1962), and Stigler (1971) studied US utilities and came up with the famous
“Capture Theory.” Stigler tried to show that the regulators were “captured” by the
utilities/industries, which resulted in higher consumer prices. He used cross-section
data to show the effect of regulation on price. He concluded that regulation led to
higher electricity tariffs. The utilities accepted regulation in order to avert
competition.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10
Later research works have indicated varying results in this regard (Jarrell, 1978;
Rnittel, 1999). Jarrell concluded that the industry sought regulation to escape the
competitive environment. Knittel casts doubt on the capture theory, even though his
results do not overwhelmingly support the rival public interest theory.
So the challenge is to develop a credible system of regulation that would both
attract investment, and help deliver better service at lower cost. It involves both
political and economic components. The regulatory framework, the selection of
regulators, the overall mandate and independence of the regulators depend upon the
existing political culture. But their responsibility mainly involves economic decision
making like setting tariffs, deciding the rate-base, delineating the service area, setting
the norms of prudential investment, internalizing the external costs, etc. So long as
the restructured entities/companies supply services as per the terms and conditions of
the licenses, they are allowed freedom in their operational and financial decisions.
Risks and Uncertainties
The electricity market is subject to a variety of risks and uncertainties. Some of
them are mentioned here. First, demand uncertainty arises out of short-term and
long-term demand shifts, and changes in weather. Second, price uncertainty arises
due to fluctuations in the prices of inputs like coal, fuel, natural gas, etc. Third,
regulatory uncertainty may crop up due to changes in regulatory decisions about
stranded costs. These are costs that have already been incurred by the utility, but that
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
11
have not been recognized by the regulatory commission as prudential, and thus not
taken into account while calculating the rate-base and the allowed revenue. The
regulatory uncertainty may also arise due to changes in transmission and distribution
loss allowance, plant load factor or capacity allowance etc. Fourth, political
uncertainty arises from fluctuating political commitment to the regulatory process.
Fifth, the industry may face uncertainty over regulatory rewards for reducing losses.
Sixth, there could be uncertainties due to equipment failures. Finally, the industry,
the regulators and the consumers may face uncertainties due to fluctuations in
interest rates and inflation.
Under a regulatory regime in which utilities’ prices are set so that their revenues
cover costs, at least in the absence of high inflation and high interest rates, much of
the risk is passed from the electric utilities to end-users. For example, where utilities
are permitted to pass through all changes in fuel price to end-users, they avoid
exposure to fuel price risk. Other risks, such as the size of long-term demand, are
passed on to end-users in the form of unreliable or low quality electric power or high
prices to pay for the cost of unused excess capacity. In some cases, like industrial use
in countries like India, the bilateral contract stipulates some minimum level of
consumption/sale. Thus, it remains an important task for all the stakeholders,
especially the regulators, to deal with these risks and uncertainties.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
12
Theories and Types of Regulation
Various theories of regulation have been propounded. We discuss some of them here
before delineating the various types of regulation, followed in different countries.
• Normative and Positive Theories of Regulation: The former prescribes the way in
which regulation ought to be designed to maximize social welfare, while the
latter predicts the way regulation works in practice.
• Interest Group Theory of Regulation: It recognizes that competition among
interest groups is imperfect or oligopolistic and can reallocate returns in socially
undesirable ways. Regulators are seen as utility-maximizing arbiters between
these various interest groups. Different interest groups have different bargaining
power, depending on their costs of organizing and the benefits of manipulating
outcomes, and they will intervene to redistribute the benefits to their groups at
some additional inefficiency cost. Peltzman (1976) argues that the rents will be
distributed approximately in proportion to the net benefits of the interest groups.
Those groups, which perceive erosion of their ‘special’ privileges, may well
oppose liberalization/restructuring.
Theory predicts important similarities and differences between public ownership and
regulated private ownership. Under public ownership, interest groups will compete
in the political market place for benefits. Under private ownership, the regulator will
represent the interest of the non-owning groups (Newbery, 1999; pp. 142). These
groups consist of domestic consumers, workers and other interest groups.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
13
It is important to distinguish two distinct categories of regulation: regulation of
price structures and control of price levels. As far as the former is concerned, we
come across different “menus” like peak-load pricing, multi-part tariffs, and
interruptible service pricing. Peak-load and interruptible-service pricing provide
incentives for users to shift demand away from peak periods. As far as the regulation
of price levels is concerned, we come across two main types of regulation: rate of
return (ROR) regulation and price cap regulation. ROR regulation is the standard
form of utility regulation in the US. It sets a rate of return on capital employed for
the industry, which is assessed primarily on the basis of the capital required to
produce and supply electricity. As Joskow (1998b) points out,
The potential benefit of a credible cost-plus regulatory system is that it is likely to be very
effective in attracting capital investment to the sector because investors are assured that they
will recover their operating and investment costs (pp.212).
However, the ROR has not been without its share of criticism. The Averch-Johnson
(A-J) hypothesis (1962) criticizes the ROR regulation on the grounds that the utilities
tend to over-invest to inflate their rate-base. Some opine that these problems can be
minimized with appropriate modifications such as the introduction of periodic lags in
determination of rates rather than continuous ROR assumed in the A-J model. Some
others claim that contestable markets and the pressure from the capital markets
minimize these problems.
Price-cap regulation was suggested by Littlechild (1983). It sets limits on price
increases linked to the rate of inflation, reduced by the regulator’s assessment of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
14
possible productivity increases (X), and increased again by allowances for fuel cost
increases (Y). This yields the formula for allowed annual price increases RPI - X +
Y. This rule is usually in place for a pre-announced length of time, such as five
years, and then reviewed. It is the main approach used in the UK. Acton and
Vogelsang (1989) list four characteristics of price-cap regulation. First, the regulator
directly sets a ceiling for prices to be charged for the regulated firm. The firm may
choose prices below the ceiling, though it is very unlikely. Second, price ceilings are
defined for the basket of services offered by the regulated firm. They can be
expressed as price indices for these baskets, and different ceilings may apply to
different baskets. Third, these price indices are adjusted periodically by a pre
announced adjustment factor that is exogenous to the regulated firm. Fourth, the
adjustment factors, basket, and weighting schemes for the indices are reviewed at
intervals of several years.
It is claimed that price-cap regulation avoids the risk of over-investment.
Moreover, it provides incentives to efficiency improvement. The X-factor is
supposed to capture potential efficiency gains, which are to be passed on to the
consumers in form of lower prices. The regulatory lag is also claimed to be helpful.
The X-factor is reviewed almost once every five years. So the firm takes the profit
due to enhanced productivity within that period. So there is incentive built into the
regulatory procedure itself. On the other hand, some of the Nordic countries
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
15
emphasize yardstick regulation, which consists of comparing a utility’s rates with
those of several comparable utilities.
Some authors like Schmalensee (1989) argue that neither a pure fixed price system
(like the price-cap regulation) not a pure cost-plus system is likely to be the most
desirable regulatory mechanism. Given the various “shortcomings” or characteristics
of the developing countries, Joskow (1998b) advocates for a “hybrid” approach-
including quality of service and performance norms and prespecified provisions for
evaluation and adjustment once every few years — as the most effective regulatory
approach (pp.215).
Thus, various approaches have been advocated and countries have adopted them.
Some countries like France have continued to opt for a vertically integrated, state-
owned and managed electricity sector. Others have resorted to rate-of-retum or
price-cap regulation. The Scandinavian countries have accepted yardstick regulation.
However, in the majority of cases, we come across not only a change from
government control to independent regulatory control, but also a reduction in the
regulatory control itself. Thus, most of the countries have been witnessing
liberalization in the electricity sector. We discuss that in the following section.
LIBERALIZATION IN THE ELECTRICITY SUPPLY INDUSTRY
There has been a worldwide trend towards liberalization in the electricity supply
industry. Liberalization has taken several different forms. At its least ambitious, it
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
16
may involve the opening up of contracts to private sector bidders and the
introduction of performance-related pay in the state-owned enterprises. More
extensive liberalization may involve allowing independent generators, sometimes
also called independent power producers (IPPs) to enter the market. It may further be
extended to vertical disintegration or unbundling of state-owned monopolies into
generation, transmission and distribution businesses, the creation of a power pool
and the horizontal separation of incumbent generators. The most comprehensive
liberalization usually culminates in the whole or partial sale of state-owned assets to
the private sector, and competition in generation, transmission and distribution.
Chile was the first country to restructure its state-owned electricity companies.
Restructuring began there in 1978 and privatization followed in the latter half of the
1980s. The most radical liberalization continues to take place in the UK where
partial deregulation was initiated as early as 1983.
Pollitt (1997) summarizes the current state of the liberalization process in OECD
countries, transition economies, Latin American and Asian countries. He enumerates
three main reasons for the liberalization trend. First, the developed economies like
the USA and the UK have argued in favor of the benefits of privatization. Second,
the collapse of the Berlin Wall in 1989 and the end of the Soviet Union witnessed a
simultaneous move to restructure the large state-owned sectors, of which the
electricity industry is just one part. Third, following the debt crises of the early
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
17
1980s and the pressure on government finance in developing countries, many took
steps to cut budget deficits by improving the efficiency of public utilities in Asia and
Africa. The IMF and the World Bank actively supported such efforts to restructure
and privatize.
Significant asset sales (privatization) have occurred in Australia, Canada,
Germany, Spain and the UK, while substantial liberalization has occurred in Ireland,
New Zealand, Norway, Sweden and the USA. Among the developed countries, only
a few appear to have done little about restructuring or reform, most notably France,
where the state-owned monopoly, EdF, has long resisted such attempts. Likewise,
among transition economies, only Romania and the Slovak Republic have done
nothing in the way of restructuring. In Latin America, Chile and Argentina
liberalized their electricity sectors, and have already witnessed good results. In Asia,
Malaysia, the Philippines, and Pakistan are leading the reform movement, and some
states of India and China have undertaken restructuring, but almost no privatization.
Most such reforms have been started since 1991.
Hunt and Shuttleworth (1998) have distinguished the structural and ownership
dimensions and changes therein in matrix form. According to them, and as shown in
Figure 1.1, there are only four ways to structure the electricity industry. Model 1 has
no competition at all. Model 2 allows competition in generation. But there is a single
buyer, who chooses from these generators. Model 3 allows distribution companies
(Distcos) to choose their supplier, which brings competition into generation and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
18
wholesale supply. Model 4 allows all customers to choose their supplier, which
implies full retail competition.
Similarly, the ownership dimension can be divided into three levels. Figure 1.1
applies this to several developed countries. In many countries, the electric industry is
a government department, with no separate accounts. The next level is a distinct
government-owned company, or nationalized industry. The third level is a privately
owned industry.
Figure 1.1: (Structure and Ownership Pattern of electricity industry in some
developed countries)
Model 1 Model 2 Model 3 Model 4
(No competition) (competition in (competition ( retail
generation) in gen. & dist) comp.)
Government
Ownership Australia New Zealand
Public France Scotland NZ ► NZ ► NZ
Corporation England an 1 Wales
y
Australia
Northern Ireland
Private Scotland England
Corporation USA ►US (1978) 4>US(1992) & Wales
Source : Hunt and Shuttleworth (1998)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
19
It is quite evident in Figure 1.1 that most of the countries are moving from the
upper left comer to other cells in the matrix. Only France has continued to remain
where it was. Most movements have been downward, some also to the right and
some like England both down and to the right - a reduction in government control
and an increase in competition and choice.
India started the liberalization process in 1991. The electricity sector was mainly
government-owned and operated. Though the sector had witnessed a high rate of
growth since independence in 1947, it ran into a lot of problems. The sector needed
huge investments that the federal government1 or the state governments could not
afford. The rate of return on capital had become negative, and was going from bad to
worse. The thermal power plants’ performance left a lot to be desired. There was a
very high degree of cross-subsidization of tariffs that forced industries to go further
away from the state-owned electricity grid and depend upon their own power plants.
There were many more problems that the state governments were finding
increasingly difficult to handle. There were other factors too, which we will discuss
in greater detail in chapter three. It will perhaps be sufficient to say here that India
decided to opt for restructuring of its power sector in the face of these problems.
1 Instead o f Union Government or Central Government, we use the term Federal Government, as it is more conventional in
literature and comparable to other countries.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
20
PURPOSE OF THIS STUDY:
Most of the existing studies have focused on the effect of ownership on
performance of the electric utilities. These studies mostly relate to the developed
countries like the USA and the UK. They have econometrically estimated the effect
of privatization on costs and productivity. Some have discussed its effect on
productive as well as allocative efficiency. None has analyzed the effect of
restructuring on performance, the objective of the present study.
Second, there has been no analytical work on the effect of restructuring on the
performance of India’s electric utilities, even though there have been very good
general discussions of the Indian electricity sector. Most of these works follow a
comparative-institutional approach and an historical path, explaining the evolution of
the electric power sector. These studies do not use actual data for analysis and
conclusion.
Third, no study has yet made use of panel data to study the effect of restructuring
on plant performance. Although some studies have used time-series data for
particular countries to assess the impact of privatization, most studies have used
cross-section data (mostly of developed countries) in this regard. The use of cross-
section data, however, has shortcomings for making inferences about the dynamics
of change due to unobserved variables, imposing a possible bias in the estimates. For
instance, just using a dummy variable for restructuring in a performance equation
may provide misleading estimates, because these estimates are likely to reflect other
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
21
uncontrolled cross-state or cross-plant differences. The estimated coefficients from a
single cross-section are more likely to reflect inter-individual or inter-firm
differences rather than intra-individual or intra-firm dynamics. Similarly, a single
time-series data set cannot provide precise estimates of dynamic coefficients either.
Aggregate time-series data are not particularly useful for discriminating between
hypotheses that depend on microeconomic attributes. Nor will a single individual
time-series data set provide enough evidence/information on the population.
If panel data are available, we can utilize the individual differences in values of
the explanatory variables to reduce the problem of collinearity. Moreover, panel data
give us a larger number of data points, increasing the degrees of freedom, and
serving as a means of reducing the magnitude of a key econometric problem that
often arises in empirical studies, namely the omitted variable problem. By utilizing
information on both the intertemporal dynamics and the individuality of the entities,
one is better able to control in a more natural way for the effects of missing or
unobserved variables. Panel data, by providing sequential observations for a number
of individuals or units over a period of time, can allow us to distinguish inter
individual differences from intra-individual differences, and to study a before-and-
after effect. Using appropriate specification tests, we can identify the individual-
specific and time-specific effects.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
22
Why India:
The electric power sector in India is in a state of flux. The implications and
significance of these changes for economic, distributive and political outcomes are
potentially profound.
What makes the choice of India especially fortuitous is that its states are at
different stages of restructuring, thus offering an interesting opportunity to study the
impact of restructuring on the sector’s performance. Second, India’s experience in
restructuring will be of importance and relevance to other developing countries with
low per capita income but high growth rate in per capita power consumption. Third,
the choice of regulatory framework and the formulation of public policy will also be
of importance to those states that have chosen to wait and watch.
As already mentioned, there has been no study that uses panel data to estimate the
effect of this unbundling on the performance of the state electricity boards (SEBs)
and their unbundled counterparts in India. We attempt it in this study.
In addition to this, we study the effect of ownership and restructuring at the plant
level. The thermal power plants in India function under various patterns of
ownership. The private sector owned some plants even before India’s independence.
They continue to run those plants. Some more plants have come under private
ownership over time; however, the different states of India (through their SEBs) own
a large majority of the power plants. The federal government started investing in and
constructing thermal power plants in the mid-70s.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
23
Not many states in India have resorted to restructuring their electricity industry.
Although the idea has gained ground since 1991, and expectations of the
stakeholders have risen in various places, it was first implemented in Orissa, a state
in the eastern part of India. Three major factors helped to start the process in Orissa:
(a) very strong political leadership, (b) a very small share of power going to
agriculture (mostly the agriculture is rain-fed) leading to little pressure from
agriculture against changing the highly subsidized tariff in the restructured scenario,
and (c) poor commercial performance in the electricity sector, inadequate tariffs and
large-scale cross subsidies. The plant availability1 for Orissa was extremely low in
the years before the restructuring. The plant load factor was also very low compared
to the all-India average. Similarly, the forced outage3 was 33.50 percent in 1994-95
compared to the all India average of 12.42 percent. The cross-subsidization of tariffs
was also very high. In 1992-93, the ratio of industrial tariff and the tariff for
households was as high as 2.36. The rate of return, exclusive of subsidies, was 21.50
percent in 1995-96, the year in which the restructuring started.
Orissa has not only unbundled generation, transmission, and distribution, it has
also set up an independent regulatory commission. It is also the first state to privatize
its distribution. Other states like Andhra Pradesh, Haryana, Karnataka, Rajasthan and
1 The number of hours in the year the plant was available for generation as a percent of 8760 hours.
2 Plant Load Factor is the actual output as a percent of total potential output.
3 Forced Outage refers to the unscheduled breakdown as a percent of 8760 hours.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
24
Uttar Pradesh are in different stages of restructuring. Yet, some states like Bihar,
Assam, and West Bengal have not yet taken any concrete steps in this regard.
To date, numerous claims and counter-claims regarding the effects of restructuring
on social welfare and efficiency have been made. It is generally claimed that the
situation improves in the long run. In the present context, the question that naturally
crops up is whether the process of restructuring in the given institutional framework
is showing any promise for the future. Providing the necessary analytical arguments
and evidence may prevent a relapse to the previous system.
There are certain advantages in choosing states within a country. They all have the
same political structure, same judicial system, and almost the same type of
bureaucracy. They enjoy/suffer from same type of average price level. There is no
need to take into account PPP (Purchasing Power Parity) exchange rate. They deal
with the same type of exchange rates. They face the same federal fiscal and monetary
policies.
But, these states also differ from each other in the timing of their decisions to
restructure their power sector, their per capita income, their demand, their
consumption patterns, the proportion of hydroelectric power in total power
generation, fuel prices, and the extent of urbanization. They also display variability
in performance in generation, transmission and distribution. This provides us with a
unique opportunity to study the power sector.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
25
Restructuring:
What do we mean by restructuring? Why and how might it affect the performance
of these thermal power plants?
As noted earlier, restructuring in the present context means the unbundling of the
erstwhile vertically integrated SEB into different companies of generation,
transmission, and distribution under separate control. It also implies that the state
does not control the electricity tariff, which is set by an independent regulatory
commission. The restructuring process is supported by reform legislation (like a
State Electricity Reforms Act), and not by an executive decree, which can be altered
by the executive when it wants. The Act, on the other hand, cannot be changed
without the consent of the Parliament or the Legislative Assembly, as the case may
be.
There are three arguments that are cited against restructuring/unbundling and for
vertical integration. First, vertical integration enhances the utility’s ability to control
information to defend its interests. The integration gives the managers more control
over the information, and helps them to coordinate the activities of the various
‘wings’ of the company. Second, the utility can argue that economies of scale and
scope make integration efficient. These integrated companies often cite cost
reduction to support vertical integration. Third, unbundling and competition increase
risks. Competition brings in uncertainties, which are cited as detrimental to the
company’s well being.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
26
Now, the question that naturally arises is that if the vertical integration is so
beneficial, why does anyone need restructuring or unbundling? Before, we discuss
the possible impacts of restructuring, we should recall what Grossman and Hart
(1986) say about vertical integration. According to them, vertical integration is the
purchase of the assets for the purpose of acquiring the residual rights of control. This
“inevitably creates distortions.” Integration “shifts the incentives for opportunistic
and distortionary behavior, but does not remove these incentives.” They further
illustrate the point by saying that if firm i owns firm j, then firm i will use its residual
rights of control to obtain a large share of the ex post surplus, and this will cause
firm i to overinvest and firm j to underinvest. Under non-integration, on the other
hand, the ex post surplus will be divided more evenly. Integration is therefore
optimal when one firm’s investment decision is particularly important relative to the
other firm’s, whereas nonintegration is desirable when both investment decisions are
“somewhat” important (Grossman and Hart, 1986; pp. 717).
In the case of integration, the state (principal) will overinvest in generation,
provide subsidies, keep the tariff low, and undertake social programs like rural
electrification. The state does all this in response to the preferences of the median
voter, who is a household consumer. The manager/employee (agent) will underinvest
in managerial skill and effort. In the case of non-integration, the state loses direct
control over tariff setting. And it no longer has the incentive to invest or subsidize as
much as before.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
27
Now let us discuss how restructuring might affect performance. There could be
various reasons for restructuring to affect performance. First, the unbundling of the
vertically integrated and monolithic structure creates independent action-centers with
accountability. Unlike the integrated case, these restructured units start feeling the
pressure of the market, as the direct control and the protection of the government are
removed. Each of these individual companies is compelled to take care of its own
performance to improve profitability. The accounts of these companies become
separate, and reflect their own performance. The earlier case of possible moral
hazard among the departments gives way to individual accountability.
Second, the government subsidies disappear along with the government control.
The managers and the employees start facing much harder budget constraints.
Seldom are governments in a position to provide any subsidy.
Third, restructuring brings with it a new way of tariff setting. The procedures of
setting tariffs for different consumers in the regime of the vertically integrated utility
are not transparent. The government usually decides the tariffs and the extent of
cross-subsidization. But in the new regime of the independent regulatory
commission, the procedures are made public. The commission hears all concerned
parties, including consumer groups, government, and the generation, transmission
and distribution companies. The orders are made public. Restructuring, thus, puts an
onus on the regulatory commission to explain the tariff structure, including the extent
and rationale for cross-subsidization.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
28
Fourth, restructuring usually brings competition in generation and distribution.
Everywhere in the world, restructuring has created more companies and initiated
more competition among those companies. The earlier monopoly of the state-
controlled vertically integrated monolith gives way to a number of independent
companies, competing with each other. The earlier market power gradually
diminishes. However, it usually takes some time to achieve this competitive
environment that has a downward effect on costs and prices.
Fifth, as the vertically integrated company is replaced by its restructured
counterparts with managerial independence and financial accountability, these
companies try to improve the efficiency and profitability of their respective
companies without looking to the government for help. They decide investment that
is needed to achieve these goals.
Sixth, one of the most important changes that takes place after restructuring is that
income transfers between these companies stop as companies become separate. The
income transfers among various ‘branches’ of the vertically integrated companies
usually conceal inefficiencies in some of those branches. After restructuring, they
function as independent companies and face their budget constraints.
Restructuring may thus be required as part of the process of confining regulation
to the natural monopoly network and introducing competition where possible.
Regulation is inevitably inefficient because of problems of information and
commitment and, more fundamentally, because of inefficient bargaining between
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
29
interest groups over potential utility rents (Newbery, 1999; pp.134). The regulators
may give in to the demands of those groups that have more bargaining power,
thereby deviating from the optimal solution. Competition may help in this regard. It
suggests a rule for structural reform: competition where possible, regulation only
where unavoidable. The reason that competition may be superior to regulation is
twofold: Pricing tends to be more efficient and costs lower. Competition gives
maximum incentive to reduce costs and innovate as the only ways to increase profits.
Competition is more effective than regulation at cutting costs to improve productive
efficiency, and aligning prices with costs to improve allocative efficiency. The
Hayekian discovery mechanisms of competition put considerable pressure on the
companies and tend to improve performance. Therefore, the aim of liberalization and
restructuring is to confine regulation to the core network and thereby minimize the
extent of regulatory inefficiency.
The critical element of vertical separation is to ensure that the link between
generation and transmission is severed so that generators do not own transmission
and the transmission company does not own generation. Separating transmission
from generation increases the transparency of the charges of the generators, and also
enhances the efficient contestability of the market place. Similarly, generation and
distribution should also be separated. “Regulation is simplified if generation and
distribution are also separate, for then the boundary between the natural monopoly
and potentially competitive parts is clearly defined” (Newbery, 1999; pp.202).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
30
As Newbery (1999) further points out aptly, structural reforms to introduce
competition are likely to have far-reaching effects on the distribution of rents, the
distribution of risks, the rate and direction of technical progress, the choice of
investment and the forms of regulation required. Institutional reforms may also have
a profound effect on the path of future regulatory reform by encouraging or
discouraging competition and innovation. Statutory monopolies at one extreme
protect incumbents from the need to experiment to find the least-cost solution to
consumers’ needs, while offering opportunities for building technically attractive
solutions to the company’s desires. A large part of the case for allowing competition
is to liberate the potential for both discovery and market testing. Competitive
entrants reveal potentially embarrassing information to regulators or government
audit committees about the true costs of the industry, and this can lead to a tougher
and less attractive regulatory regime for the incumbent.
Thus, restructuring brings in changes in management, accountability, competitive
pressure, regulatory regime and budget constraints. While some economies of scale
and scope may be sacrificed in the process, other effects may be beneficial. Hence,
the desirability of restructuring becomes an empirical question. Therefore, we
develop hypotheses and use data to test them. We discuss the performance indicators
and the hypotheses in the following section.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
31
Performance Indicators and Hypotheses:
As we have already discussed, restructuring brings with it a basic change in the
structure of the company. It is likely to initiate important changes in management
style and control, regulatory regimes, introduce harder budget constraints (in the
present cases in the electricity sector), and greater accountability. In view of these,
we hypothesize that restructuring will significantly affect performance in both the
short-run and long run.
In this study, we use two different types of performance indicators. The first type
relates to operational efficiency: plant availability, plant load factor, forced outage,
number of employees per thousand consumers, and number of employees for MKWh
of electricity sold. Plant availability and forced outage indicate the quality of plant
management. The plant load factor is an indicator of capital productivity. The last
two are measures of labor productivity. We discuss these variables at length in
chapter four.
The second type of performance indicator is financial in nature. These indicators
include the average unit cost of supply of electricity, average tariff collection, sales
revenue as a ratio of cost, and the extent of cross-subsidization (measured as a ratio
of industrial tariff to domestic tariff). Reliability or the quality of power supply
would have been a good indicator of performance of the electric sector, as it affects
all the consumers. In some countries, the mean time between failures, i.e., power
failure, (MTBF), is used to indicate the reliability or the quality of the power supply.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
32
Since the data on this important variable are not available for all states and all years
under study, it could not be taken into account for analysis. Another alternative could
have been the panel data on random samples of consumers’ opinion about the
reliability or the quality of supply. But that is also not available. Therefore, we will
study the effect of restructuring on the operational and financial performance
indicators mentioned above.
We hypothesize that restructuring would have a significantly positive effect on
plant availability and plant load factor, but a significantly negative effect on forced
outage. Since restructuring initiates changes in management style and monitoring,
and brings in harder budget constraints, it may positively affect labor productivity. In
the present study, we take two indicators as proxies for labor productivity: (1)
number of employees per thousand consumers, and (2) number of employees per
million KWh of electricity sold. We hypothesize that restructuring will have negative
effect on these two variables, indicating an increase in labor productivity.
As far as cross-subsidization is concerned, we hypothesize that restructuring will
bring in tariff rationalization, implying thereby a reduction in cross-subsidization.
The new independent regulatory commission will try to reduce the cross-subsidy
over a period of time. The procedure becomes much more transparent. The tariff
order is made public. It becomes difficult for the commission to justify an extremely
high level of cross-subsidy, especially when the industries are banking more and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
33
more on their own captive power generation to avoid the extremely high industrial
tariff.
As far as tariff collection is concerned, we hypothesize that restructuring would
have a positive effect. This is due to the fact that the new management would try to
improve financial performance, because of the harder budget constraints. For similar
reasons, it would try to improve sales revenue vis-a-vis costs. Thus, restructuring
would have a positive effect on sales revenue as a ratio of costs. We also hypothesize
that restructuring would have negative effect on the unit cost of electricity in the long
run. It may not have significant effect in the short run, as it takes considerable time
to renew contracts and renegotiate agreements.
We use three different indicators of restructuring to analyze the panel data on
performance. We measure restructuring both as a dummy variable, and also as a
continuous variable representing the number of years since the restructuring
commenced. The lagged variable of restructuring (with one-year lag) is used to
analyze its effect on the set of financial indicators on the assumption that impact on
financial performance may take longer to realize than operational changes. It takes
some time to enter into new contracts, renegotiate/renew old contracts with suppliers
and buyers. Similarly, it takes time to modify or rationalize the salary structure of
employees, and to make it performance-based. The management may have to
undertake long negotiations/discussions with the trade unions to arrive at a mutually
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
34
acceptable solution. The management could face injunctions from the courts during
this period.
Restructuring is usually initiated by new legislation that subjects the new
unbundled companies to less direct political/government control. For instance, the
government no longer decides the tariff structure, or controls the policies of the new
companies directly. It does, however, continue to play a significant role, indirectly,
suggesting the need to consider political economy issues.
Political Economy of Restructuring
Political forces play an important role in the decision making of the electricity
industry, especially in a democracy like India. Government plays a direct role in
investment and tariff decisions when the company functions as a vertically integrated
state electricity board. It is quite likely that the political party in the government
would influence various decisions of the board. Even after restructuring, government
still plays an important role. The industry needs government land and resources like
water and coal. Government also decides the types of regulation (like rate-of-retum
or price-cap), appoints the regulators and delimits the scope of the regulatory
framework.
It is natural that the above government actions often reflect the ‘wishes’ of the
prevailing political majority. As Laffont (1995) points out, in developing countries, a
government with majority support can pursue its own objectives more easily than it
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
35
can in more developed countries, where various countervailing powers may constrain
government actions in such a way as to make a better account of social welfare.
As pointed out in the Shapiro-Willig model (1990), through voting, politicians
represent majorities, which have particular interests. In the case of public firms,
Laffont (1996) argues that by definition they belong to the majority in power.
Therefore, changes in majorities will affect the incentives of regulated firms. It is not
only the political majority that affects decision-making; it may also be the type of
majority that can give rise to different effects. For example, in a multi-party
democracy like India, sometimes political parties come together (sometimes before
and sometimes after the election) to form a coalition government. Most of the time,
they act as checks and balances on each other. These checks and balances usually
come up from the grass-root level party workers. If a functionary of one of the
parties in government wants to exercise undue influence, the functionary from the
other political party opposes it. However, when a party has an absolute majority in
the legislative assembly, and does not share power with any other political party in
government, it escapes the above checks and balances temporarily till the next
election.
We come across various accounts of political interference in the functioning of the
electricity sector. As Dubash and Raj an (2001) point out,
The main complaint o f SEB managers was that, as government-appointed officials, their
elected superiors were providing them directives that interfered with day-to-day operations.
They were also frequently transferred out of the sector based on the whims of their superiors
or the vagaries of the elections. Such interference frequently included demands for ad hoc
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
36
extensions to the grid to address the needs of particular constituencies, which worsened the
performance of the distribution network and increased the potential for theft. The
combination of structural inefficiencies introduced through political interference and forced
subsidies without government compensation led to a situation where most SEBs were often
in serious financial trouble.
Thus, it is worthwhile to examine the political economy aspects while analyzing
the effect of restructuring on the performance of the electricity sector. In this study,
we attempt to identify the effect of a single-party majority on performance
indicators. Hence, we construct a dummy variable, POL, to indicate absolute
majority of the political party in the government.
Organization of the chapters
Chapter 2 reviews the literature relating to restructuring, privatization and regulation.
It discusses the various empirical studies conducted to assess the impact of
privatization on performance of electric utilities. It also analyzes the conditions that
are necessary for successful privatization. Moreover, the chapter tries to distinguish
the ownership issues from the structural aspects.
Chapter 3 discusses the evolution, characteristics and political economy issues of
India’s electricity sector. No Indian Act specifically explains the meaning of “public
interest”. This has encouraged different political parties to interpret its meaning in
various ways. We not only look at the election manifestos of major political parties,
but also examine the constitutional provisions and legal interpretations in this
chapter. We also discuss the political economy of subsidies in the electricity sector.
The chapter also delineates the structural and functional dimensions of the electricity
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
37
sector in Orissa, the state that pioneered the restructuring of the state electricity
boards in India in 1995. Further, it compares India’s power sector with those of
Chile, Hungary and Norway. Chile, a developing country like India, was the first to
liberalize its power sector in 1978. Its privatization of the power sector is complete.
It has witnessed commendable achievements. Hungary, a transitional economy,
started restructuring and privatization in 1990s. Its transmission company is still
state-owned, though most of the generating plants and the distribution companies
have been privatized. Its power sector resembles that of India to a considerable
extent. Norway, a developed country, started liberalization in 1990s. Interestingly, it
has restructured its state-owned power sector without privatizing it. It has
encouraged competition with privately owned utilities, and has achieved very good
results.
Chapter 4 discusses the two panel data sets, the variables, the model and the
equations used to estimate the effect of restructuring and ownership on performance
of the state electricity boards and the thermal power plants. It also discusses and
examines the econometric issues involved in the analysis.
Chapter 5 discusses the results of the econometric estimations and the testing of
hypotheses. Chapter 6 delineates our conclusions, and policy implications, including
some suggestions for future research.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
38
CHAPTER 2: LITERATURE REVIEW, POLITICAL ECONOMY OF
REGULATION, EMPIRICAL STUDIES ON PERFORMANCE
The terms “privatization” and “restructuring” have been used interchangeably in
the literature, even though they represent different dimensions of change. Since it has
been commonly assumed that competition is difficult, if not impossible, to achieve
within the public sector, it has been natural to view privatization as a pre-requisite
for liberalization (Newbery; 1997). Thus, it is not surprising that the literature on
power sector reforms mainly deals with issues pertaining to privatization, which is
often assumed as the endpoint of liberalization. Restructuring, on the other hand, has
received attention only in so far as it would constitute a part of the privatization
process.
In his insightful exposition on the political economy of restructuring and
regulation, Laffont (1996) argues that regulation and privatization are two possible
instruments for promoting development. They may appear at first sight
contradictory, since privatization is designed to a large extent to keep the
government’s hands off enterprises, and regulation means government control.
Actually these terms are not contradictory since, even in a world where all firms are
privatized, we still have to worry about monopolistic behavior, and competitive
policy is required.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
39
Political Economy of Regulation
As noted earlier, regulation involves both political and economic components. The
choice of regulatory framework, the selection of regulators, the overall mandate and
independence of the regulators depend upon the existing political culture and
decision-making of the political party (or parties) in the government. But the
responsibility of the regulators mainly involves economic decision-making like
setting tariffs, allowing subsidies, issuing licenses, deciding the rate-base, delineating
the service area, setting the norms of prudential investment, internalizing the external
costs, etc.
As the economic environment and the legal and political institutions of the
developing countries differ substantially from those of the industrial economies,
Joskow(1998b) argues, “while there are lessons to be drawn from the recent reform
experiences of industrial countries, these lessons cannot be applied mechanically to
developing countries.” He argues that introducing competition and deregulation in
infrastructure sectors dominated by state-owned natural monopolies poses a
challenge to a country’s politico-economic equilibrium. Thus any prescription for
regulatory reform must take into account a country’s economic environment, its legal
and political institutions, and its specific characteristics.
The developing countries evince many specific characteristics, which often work
as impediments in the reform path. Laffont (1996, 1998) enumerates these
‘specificities’ of developing countries. First, he shows that an essential concept in the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
40
theory of regulation and privatization is the notion of marginal price of public funds,
that is to say the social price of raising one unit of money. He underscores the fact
that the cost of public funds in developing countries is quite high. This price includes
the deadweight loss because government revenue is raised with distortive taxes. Of
course, the dead-weight losses of taxes depend on the type of tax used since the tax
systems are not generally optimized. According to Jones, Tandon and Vogelsang
(1990), this cost for developing countries is much higher than that for developed
countries.
Secondly, developing countries often lack sophisticated accounting systems,
proper auditing staffs and administrations, and expertise in enforcement. The poor
functioning of audits is due to at least three phenomena. These countries face
political and social difficulties of paying incentive salaries to auditors that induce
effort and discourage corruption (Gould and Amaro-Reyes 1983). The poor
technology of auditing itself makes it harder to discover cost-padding. The inability
to impose high penalties when wrongdoing is documented contributes to the
problem.
Third, cost of side transfers is relatively smaller and more difficult to identify.
Some norms may value positively some types of side transfers when they take place
within families, villages, or ethnic groups.
Fourth, the developing countries usually lack two features— meaningful
constitutional control of the government and some ability to commit and write long
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
41
term contracts. The lack of checks and balances provided by well-functioning
democracies, supreme courts, government auditing bodies and other counter powers
allows the government to be more easily captured by interest groups and makes it
more likely to favor inordinately the voters supporting it.
Fifth, the lack of political democracy and well-functioning political institutions
increases the uncertainty regarding future regulations and makes it difficult for the
government and regulatory institutions to make credible commitments to long-run
polices. Laffont (1998) observes in this regard,
The lack o f credibility associated with weak government is a major problem for developing
countries, one that makes the emergence of a competitive environment difficult and affects
all areas of economic activity. Agents cannot write long-term contracts because the court
system cannot be trusted to protect them. And agents do not find it worth investing because
they fear that their gains will be stolen or expropriated either in their market activities or in
their contractual relationships with the government. This lack of protection for agents who
are no longer residual claimants for the results of their efforts induces the Williamson
underinvestment effect, particularly in non-contractible investments.
He further observes, “Consequently, the economic policies of developing countries
are even more sensitive to ratchet effects and renegotiation constraints”(Laffont
1996; pp. 166).
Sixth, poorly developed capital markets as well as the sheer lack of wealth make
limited liability constraints more binding. Seventh, beyond institutional weaknesses,
competition is weak in developing countries because transactions are localized as a
result of poor communication systems and inefficient trading organizations.
Finally, governments of developing countries are captured by powerful interest
groups— whether local elite or foreign investors. Laffont (1998) maintains,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
42
Little can be done to reform institutions without political support.... An oligopolistic
structure can be maintained under the pretext of regulation, generating rents that can be
shared by those in power Implementing effective competitive policies is the best way to
destroy the sources o f such rents.
He further argues,
Changes such as the information effect o f a more competitive environment, the greater
sustainability o f competitors’ goods, and the lower costs of those competitors can be viewed
as exogenous competitive pressures whose impact must be studied.
These specific characteristics affect the choice of the regulatory regime and then
the decision-making process. Firms can choose from a menu that may include either
a fixed-payment and no reimbursement of cost or a small payment and full
reimbursement of cost. The most efficient firm chooses the fixed price mechanism
(which includes a high level of effort), and less efficient firms choose cost-sharing
mechanisms that induce less effort. The former is sometimes referred to as a high-
powered incentive scheme, whereas the latter is called a low-powered incentive
scheme. Laffont (1996) points out that the high cost of public funds in developing
countries call for higher prices of the commodities produced by the natural
monopoly and for lower-powered incentive schemes with high shares of cost
reimbursement (rate-of-retum regulation).
Following the pioneering application of incentive theory to regulation by Loeb and
Magat (1979), and Baron and Myerson (1982), the 1980s witnessed the emergence
of a new theory of regulation that emphasized the asymmetries of information
between government, regulatory commissions, firms and various interest groups
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
43
(Laffont, 1994a). The asymmetric information about the firm’s costs implies that an
informational rent must be given up to the firm when it is efficient, since it can
always mimic the inefficient firm and realize the same cost with a lower effort level.
If financial penalties are not possible because of limited liability constraints, only
rewards for good performance can induce appropriate effort levels in the agents, i.e.,
informational rents must be given up.
Monitoring of effort generally enables the regulator to reduce the informational
rents and calls for higher-powered incentive schemes. A less efficient monitoring
technology will call for relatively less powerful incentive schemes. It is natural that
the high cost of public funds makes monitoring costly. And it induces low incentives
among utilities both directly and indirectly through a decrease of the more costly
monitoring.
It follows from this that there is a high probability of cost-padding in rate-of-retum
regulation that allows return according to the rate-base or cost incurred. The analysis
by Laffont and Tirole (1992) shows that the imperfect auditing of cost padding calls
for a shift toward higher-powered incentive schemes (price-cap regulation) that will
make the firms the residual claimants of their costs. But there are other factors that
call for even higher shifts toward cost-plus mechanisms. Laffont (1996) enumerates
them as follows:
■ Lower cost of internal transfers because of less monitoring of illegal activities;
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
44
■ Higher cost of incentive payments to the agency because of the higher cost of
public funds;
■ Political difficulties in creating such strong incentive payments.
The possible proclivity of regulators to renege on their commitments can be
mitigated by their continuing relationship with the utilities and the building of the
regulators’ reputation of not expropriating the rents derived by these utilities from
improvement of efficiency in the future (Croker and Reynolds, 1989). But according
to Laffont (1996), this will be less easy to achieve in developing countries.
Regulatory institutions must be particularly scrutinized in developing countries for
their ability to provide long-run incentives. For example, price-capping has been
pushed in the western world as a way to provide high-powered incentives. However,
price caps can be easily renegotiated, while a commitment to a fair rate of return
might be less prone to costly renegotiations (Greenwald, 1984).
In a world of complete contracting with a benevolent constitution, ownership
does not matter (Coase, 1960; Williamson, 1985; Grossman and Hart, 1986;
Sappington and Stiglitz, 1987). But in the world of incomplete contracts and
transaction costs, the private objectives of decision-makers will affect the outcomes,
because they cannot be perfectly controlled by appropriate contracts and because
they depend on the nature of existing institutions.
Contracts require a fair and efficient judicial system, which may not be available.
All possible scenarios in a contract cannot be forecast at reasonable cost. Flexibility
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
45
requires leaving some discretion in decision-making. This phenomenon explains why
a constitution cannot perfectly control politicians or their private agendas. And that is
why the nature and character of democratic institutions are so important, given the
specificities of the developing countries.
Let us now discuss some models that examine the political economy of regulation.
(a) The Shapiro-Willig Model (1990). Two hypotheses underlie their argument.
First, it is assumed that governments have non-controllable private agendas. The
simplicity of the constitutional-judicial rules makes the control of these agendas
quite imperfect. The political evolution of countries and democratic institutions
tend to improve this control over time. Shapiro and Willig expect that developing
countries with short constitutional histories will be lacking in these
constitutional-judicial rules. Second, political interference affects the public
sector firms. Hence, privatization is sometimes cited as a way to avoid this
interference. But privatization may not be a useful instrument in this regard. The
answer depends crucially on the quality of the democratic institutions.
(b) The Schmidt Model (1990). Schmidt also takes as given the fact that
privatization is a commitment not to learn some information about the firm and
therefore it is a credible commitment to give an informational rent to the firm.
Under public ownership, the government is informed and expropriates the rent.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
46
(c) The Laffont -Tirole Model (1991b). The cost of privatization here is due to the
multi-principal structure, since both the regulator of the private monopoly and
the owners attempt to control the managers who have private information about
the firm’s efficiency. The result of this dual control is that each principal fails to
internalize the effects of contracting on the other principal and provides socially
too few incentives for the agents, the firm’s insiders.
(d) Laffont (1996) opines that the costs of privatization seem higher in developing
countries for two reasons. First, a high level of cost of public funds implies
higher cost of monitoring. This, in turn, implies higher foregone rents in favor of
the private companies. Second, a bad auditing system also leads to high rents.
Laffont (1996) talks about three stages of development concerning regulation. In
the first stage (ex: China), the auditing mechanisms are so bad that powerful
incentive schemes should be advocated. This stage should be used to identify and
train good accountants and to develop a good auditing system. The second stage
may gradually evolve less powerful incentive systems. As development
continues, the country may slowly go to stage three when the more powerful
incentive systems could be used. The quality of regulation in each of these stages
depends critically on the ability of the governments to commit credibly to the
implementation of the schemes. Laffont (1998) holds that only a strong state can
implement competition, and that little progress in this regard can be expected in
countries where political will is lacking.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
47
Regulatory Changes
The above models mainly discuss various issues involved in privatization,
especially the political economy aspects, and the importance of institutions. They do
not specifically examine the changes that deregulation and competition might
facilitate. In the monopoly situation, the “captive customers” have no choice and are
forced to pay the price. The uncertainty is relatively small and the commitment of
resources, e.g., financial resources, is unlikely to have an adverse impact on the
firm/utility as prices could be increased by the monopoly if bad decisions are made.
The lack of incentives to become economically efficient often leads these companies
to pursue goals other than financial efficiency. Instead, these might be changed by
political objectives set by the current government, setting goals such as growth,
services, employment, or any other of a large set of possible objectives. But this
scenario changes when restructuring and competition are initiated. As Dyner and
Larsen (2001) point out, as more utility markets are deregulated and competition is
introduced, there is an increasing need to understand how the planning methods used
under monopoly have to change to take the new deregulated environment into
account.
Deregulation and competition create uncertainty. Information is no longer
complete; prices fluctuate; demand varies and becomes difficult to forecast. The
demand from a single company might have little connection with the growth or
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
48
decline in overall demand. The demand that each individual company faces will
increasingly depend on both the reliability and service provided and perceived, and
on the prices and general marketing skills with which a company can deliver
electricity.
Two other factors increase uncertainty for the planning process: (1) increasing
consumer power, especially large well-informed industrial users, and (2) the fact that
the regulator may take a harder line with the companies. In England up to 100,000
customers per week changed supplier after the market became fully competitive
(OFGEM, 2000).
One way of viewing the planning and strategy processes within deregulated
utilities is by considering the time horizon of the decisions that are involved. In the
case of a monopoly, the main planning method makes use of hard optimization type
models at all levels, which might be complemented with sensitivity analysis,
financial as well as technical. In the case of a deregulated market, this will change
significantly. The companies have to use strategic simulation and “gaming” in
addition to the optimization techniques (Dyner and Larsen, 2001; Pp. 1149).
Under monopoly there was little or no outsourcing. Deregulation, however, forces
companies to take another look at what they can justify to keep doing themselves and
what they should outsource, e.g., billing, meter reading, information technology,
research and development etc. In view of this outsourcing, new specialized
companies are likely to be established.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
49
There has been a gradual change from a planning culture to a strategy culture in
the western developed economies. Larsen and Bunn (1999) observe,
When their markets were dominated by stable monopolies, there was little pressure on the
price of the product; in many cases, a government was the sole or main shareholder; there
was never a need to embrace strategy tools. Under the changing circumstances, the planning
process has to change significantly to obtain the flexibility required by competition, customer
choice, brand management, and shareholder value. There will be a shift in the view of
technology, i.e., from a best technical solution to the most cost efficient solution.” The
choice o f the right portfolio of methods is a difficult issue, due to the long transition period
that most electricity industries are faced with when they are deregulated.
We come across divergent views on not only the choice of the right portfolio of
methods but also the pace of reform and regulatory changes. Rodrik (1996) provides
a comprehensive discussion of the various approaches advocated in reforms. Authors
like Sachs (1994) argue for what is known as the “big bang” approach. “In Sachs’
view, it is at best a waste of time to seek a broad coalition for reform because most
people have no understanding of what is required” (Rodrik 1996; pp. 33). In contrast,
authors like Przeworski (1991) “fault the storm tactics favored by Sachs and others
for both weakening democratic political institutions and making errors in economic
policy more likely.” In this connection, Bresser Pereira, Maravall, and Przeworski
(1993) underscore the building of social consensus that enhances the sustainability of
economic reforms. They observe,
We find that subjecting the reform strategy to the competitive interplay of political forces is
superior on three essential grounds. It improves policy, it builds support for the continuation
o f reforms, and it helps consolidate democratic institutions
Joskow (1998b) distinguishes between the two approaches in the context of
electricity sector. He observes,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
50
Under a ‘big bang’ approach privatization, restructuring, and the introduction o f competition
occur at the same time. Electricity restructuring in Argentina and in England and Wales took
this approach with fairly short transition periods. Another approach is to provide for a
relatively long transition period during which the industrial organization and associated
regulatory institutions are allowed to evolve according to a planned transition program.
Electricity restructuring in Columbia took this approach .. .The choice between the big bang
approach and the transition approach largely depends on six factors: the performance of the
existing system, the complexity o f implementing a big bang approach given pricing and other
imperfections that cannot be fixed instantly, the capacity of legal and political institutions to
support competitive markets for infrastructure services, the speed with which reasonably
competitive markets can evolve, the time required to create effective regulatory institutions,
and the government’s ability to credibly commit to a restructuring framework that supports
private investment and competitive entry.
Whatever be the choice of the portfolio of methods and the pace of the transition, we
come across the shift from the planning culture to the strategy culture in the new
dispensation. The proponents of privatization have pointed this as being crucial to
improved performance. Several other theories have also been propounded to explain
the efficiency gains in privatization. We briefly discuss these theories in the
following section. Thereafter, we delineate the conditions that facilitate privatization,
and discuss the empirical studies conducted to analyze the effects of liberalization on
performance.
Theories of liberalization and the effect on performance
Privatization is often assumed to be the endpoint of liberalization, and the major
rationale for privatization is said to be the improvement of economic efficiency of
the commercial enterprise, such as an electric utility. However, governments often
have secondary agendas in seeking to privatize, including reducing the budget
deficit, achieving a greater breadth of ownership, increasing investment, and
reducing the power of public sector unions, etc.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
51
Several theories have been propounded to explain the bases for improved
efficiency following privatization. First, property rights theories (following Alchian,
1965) argue that privatization involves a clearer assignment of property rights
towards those with a comparative advantage in the ownership of particular assets.
This is incentive-compatible and thus, motivates people to perform better. It reduces
the agency problem and monitoring costs.
Second, bureaucracy theories (following Niskanen, 1968) model the behavior of
civil servants and politicians who may be responsible for running the publicly owned
utilities. These individuals may not be interested in minimizing costs, but rather in
maximizing departmental discretionary expenditures. This theory also advocates
privatization of the state-owned utilities as a means of increasing efficiency.
Third, theories of regulation, especially the rate-base one (following Averch and
Johnson, 1962), bring new dimensions into the picture. An unregulated monopoly
firm chooses its output level through its choice of labor and capital so as to maximize
its profit at the monopoly level. A firm whose rate of return is regulated behaves
differently. It may lower its ROR from the monopoly level either by lowering its
profit or by increasing capital (or both). Averch and Johnson argued in their famous
article that rate of return regulation creates incentives to over-invest relative to the
social optimum because this type of regulation allows profits on the basis of the
capital or rate base. They pointed out that a regulated firm has an incentive to
increase its capital relative to the amount of labor it uses (and thereby produce
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
52
inefficiently) in order to maximize its profits. This thesis may explain why studies of
relative efficiency of public and private electric utilities in the USA show that costs
are not significantly lower in private electric utilities (Kwoka, 1996).
Other theories of regulation emphasize the information problems involved in
regulating a utility. Some point out that regulation may introduce uncertainty into the
investment decision-making process. It may also raise the cost of capital as the
companies are deterred from making risky long-term investments, the returns to
which may be appropriated in future regulatory reviews. Moreover, it is also
important to recognize that utilities may be increasingly subject to other sorts of
regulation like tougher environmental regulation because of the rising desire for
sustainable development (Chessire, 1997).
Fourth, privatization alters the relationship between owners, managers and
workers. Redistribution of bargaining power within the firm affects efficiency.
Milgrom and Roberts (1990) refer to the costs arising due to ‘influence activities’ of
various groups to increase their power within the organization. These are deadweight
losses in terms of social welfare. The most obvious example of this occurs in
government departments, which compete with one another for a greater allocation of
funds.
Influence activities arise in organizations when organizational decisions affect the
distribution of wealth or other benefits among members or constituent groups of the
organization. In such a case, in pursuit of their selfish interests, the affected
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
53
individuals or groups attempt to influence the decision to their private benefit. The
costs of these influence activities are influence costs. As Milgrom and Roberts(1992)
point out, “the magnitude of influence costs depends on the existence of a central
authority, the kinds of procedures that govern decision-making, and the degree of
homogeneity or conflict in the interests of organization members.” Thus, influence
activities are possible only when there is a central authority with the ability to affect
the distribution of costs and benefits between individuals or units. In these cases,
some solutions could be decentralization and separation of the units. Thus,
restructuring may reduce, if not eliminate, influence activities. The authors maintain
that another solution could be participatory management through employee stock
ownership plans. Given appropriate data, one can test the hypothesis that employees
as ‘owners’ try to increase efficiency and reduce costs to maximize their benefits.
Fifth, some believe that privatization represents a credible commitment to
reduced future interference by government in the running of a utility. As a result,
Willig (1994) points out that it should encourage investment. Perotti (1995) suggests
that governments may attempt to gain credibility about their commitment through
enforcement of contracts. Boycko et al. (1996) formally model privatization as an
institutional change, which increases the cost of government interference and hence
reduces the demand for it.
The theories above argue for privatization. But there are also arguments against
privatization. One of the apprehensions is the possibility of the rise of monopoly
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
54
power in the private sector that could be especially problematic in the absence of an
adequate regulatory framework. If privatized utilities exploit their monopoly power,
there may be the usual deadweight welfare losses due to high prices, the social waste
of entry deterring activities, and/or excessive entry caused by the high profits of the
incumbent. Another area of concern has been income distribution and the social
impact of privatization. In most developing countries, the governments supply
electricity to the economically disadvantaged groups at a relatively cheaper rate.
Moreover, they undertake such social programs as rural electrification, which is
subsidized by the urban and industrial supply. Privatization, it is feared, may stop
these programs, which are not commercially viable, and also result in labor
displacement from the electricity sector. Yet, another area of fear pertains to foreign
control in the case of any multi-national corporations (MNC) buying the utility. Most
developing countries try to encourage their indigenous industries to develop. Many
political parties in these countries also support some sort of economic nationalism.
The prospects of foreign control and labor displacement through capital-intensive
technologies obviously raise concerns and fear amongst the people in these
countries.
Countries differ not only in their concerns and fear about the prospects of
privatization, but also in the conditions that facilitate or hinder it. We now
distinguish the positive from the normative aspects of privatization, and discuss the
conditions that are usually considered helpful to privatization.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
55
Conditions facilitating privatization
The literature points out the existence of certain conditions that facilitate the
privatization process. Some of them can be enumerated here. First, privatization
needs institutions favoring private property rights. Clearly defined tax laws, trade
policies, bankruptcy laws, prudent accounting standards, and laws allowing profit
repatriation are often cited as necessary to attract private investors. Second, an
independent and smoothly functioning judicial system adds to the confidence of the
investors. Third, a strong but non-interfering government with a progressive
bureaucracy is needed to enforce property rights and contracts, and maintain law and
order, both crucial in attracting private capital. Fourth, a competition policy
framework is necessary to prevent private monopoly power.
Fifth, a pool of domestic savings and skill is also considered very important.
Sometimes, due to lack of adequate domestic savings, the mass sale of shares to
citizens does not become feasible. The highest domestic bidder may sometimes find
it difficult to actually arrange enough money to purchase the asset. Alternatively, the
bids are sometimes too low and collusive, causing the government not to sell the
assets.
Sixth, efforts to privatize will not succeed without political support. Political
parties have often taken stances against liberalization and privatization in their
election manifesto and party platforms. Under Margaret Thatcher, the Conservative
Party’s government in England took the decision to restructure and privatize the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
56
government-owned and controlled organizations like the Central Electricity
Generating Board (CEGB), British Gas, and British Telecom. In India, where
different political parties have been in governments in the states, one can see that the
communist parties and their allies have opposed any attempt to restructure and
privatize. However, after the Congress government began the liberalization process
in 1991, other major national political parties like the Bharatiya Janata Party (BJP)
and the Janata Dal have supported it in their party forums. This political support has
been an important prerequisite for the continuation of the liberalization process. As
Cicchetti and Sepetys (1995) point out, politics and economics will both influence
the final outcome of competition in retail power markets. There needs to be a
political resolution of the utility’s duty to serve, the scope of franchise monopoly, the
‘stranded’ investment by the utilities and the cost responsibility for the social
programs.
Laffont (1996) states that the absence of good financial markets in developing
countries also has an enormous impact on what can be expected from privatization.
First, the lack of competitiveness of the market does not provide enough information
for the correct valuation of the stocks. Second, the groups or individuals with the
best management abilities may not be able to participate easily in the privatization
process due to budget constraints. One way to alleviate those difficulties is to
structure the bidding process in a way that enables bidders to pay in the future.
However, such mechanisms may be difficult to implement because limited liability
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
57
limits ex post the possible penalties, which would induce proper behavior (Bolton
and Rolland 1992; Aghion and Burgess 1992; Maskin 1992). Foreign firms with
technology, information, and financial capabilities may seem tempting. But stability
in regulatory commitment is crucial. The problem of favoritism in the organization
of the auctions/bidding, seen so often, affects the process adversely.
The private sector is increasingly involved in electricity sector in developing
countries, but mainly under power purchase agreements (PPAs) with a single buyer,
except for a few important examples in Latin America. Izaguirre (1998), drawing on
data from the World Bank PPI (Private Participation in Infrastructure) Project
Database, finds that 62 developing countries have involved the private sector in the
electricity supply industry to some extent. He comments that failing to address the
fundamental structural problems leading to underpricing made this private sector
involvement problematic. The poor credit record of the electricity companies and
their inability to set cost-recovering tariffs made investors seek government
guarantees to underwrite the PPAs. As such, the investment was effectively public
sector borrowing.
Thus, the presence or absence of the above conditions may affect the progress of
privatization. Pollitt (1995) argues that the balance of the theoretical arguments
essentially means that it is an empirical issue, whether or not privatization means
lower costs, and improved efficiency and performance. We discuss next different
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
58
types of studies, which have been conducted to analyze the actual or likely
performance affects of electricity liberalization
Empirical Studies on Performance
Pollitt (1997) reports the results of eighteen empirical studies, which can be
classified as follows. First, we come across the long tradition of straight comparisons
of prices and costs in publicly-owned and privately-owned electric utilities. Pollitt
(1995), using an international sample of electricity generation plants, detected a
small difference of 2 to 5 percent in productive efficiency in favor of private plants,
but publicly owned plants had a higher variance in their efficiencies. Kwoka (1996)
found that prices were around 3 percent lower for publicly owned US electric
utilities than privately-owned ones, controlling for other factors. However, the
presence of a competitive supplier lowered prices by 8 percent relative to a
monopoly supplier situation.
Second, simulation studies are also used to measure the likely rather than actual
performance impacts of liberalization in the electricity industry. Pollitt( 1997)
summarizes the results of four simulation studies, one for the UK, one for Norway
and two for the USA. Bunn and Vahlos (1989) examine the likely impact of
privatizing Britain’s CEGB, the former generation and transmission monopoly in
England and Wales. They estimate the price effect of increasing the required rate of
return on existing capital and new investment. This has the effect of significantly
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
59
raising the price of electricity relative to its pre-privatized levels. Bjorvatn and Tjotta
(1993) in their simulation study found that there were modest but positive cost
reductions resulting from both national deregulation and the proposed integration of
the Norwegian, Swedish, Finnish and two Danish power transmission networks.
Kahn, Bailey and Pando (1997) find small cost reductions resulting from the creation
of a Californian and neighboring states power pool in spite of large increases in the
interstate flows of electricity.
Third, there have been studies of the actual impact of liberalization. Yarrow
(1992) compared actual prices in the early years of the restructured electricity supply
industry in England and Wales to those prices that would have been expected given
past trends. His results suggested a sharp rise in prices after restructuring and
privatization. But whether this means social welfare was actually lowered by
privatization is unclear. As Pollitt (1997) points out, it could be the case that
electricity prices were suboptimally low in the public sector, or that prices did not
reflect environmental externalities. Estache and Rodriguez-Pardina (1996) report on
the results of the restructuring and privatization of Argentina’s electricity industry,
and describe the results “as impressive by any standards.” Wholesale prices were
shown to have fallen significantly, and physical performance to have improved
dramatically.
Fourth, a World Bank study (The World Bank, 1995) of the performance of
management contracts using an international sample of firms found that incentives
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
60
were difficult to get right in the absence of political desirability ( the benefits to the
leadership and its constituencies must outweigh costs), political feasibility (the
leadership must be able to enact reform and overcome opposition) and credibility
(promises that the leadership makes to compensate losers and protect investors’
property rights must be believable).
Fifth, several studies have attempted to examine the productivity growth impact of
liberalization. Bishop and Thompson (1992) failed to find any effect on labor or total
factor productivity (TFP) growth rates as a result of the UK’s Energy Act of 1983,
which attempted to introduce limited competition into the electricity generation
industry.
Sixth, studies, using the production frontier approach, involve the construction of
an efficient cost or production frontier against which the performance of individual
firms is measured. Hawdon (1996), using cross-sectional data for developing
countries, found that the privatized firms in the sample had higher productivity
efficiency. Bums and Weyman-Jones (1994) use data envelopment analysis (DEA)
to measure productivity growth in UK electricity distribution utilities. Privatization
might either move companies closer to the efficient frontier (efficiency change) or
shift the frontier out more quickly (technical change). This DEA technique is
essentially comparative in that the overall efficiency measure compares the average
cost of production of a decision-making unit (DMU) to the average costs of the least
cost DMUs in the sample, and not some hypothetical least cost DMU (Pollitt, 1996)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
61
Seventh, there have also been attempts to conduct social cost-benefit analyses of
electricity privatization. It combines the elements of the simulation approach and a
valuation of externalities such as the impact on greenhouse gas emissions. Galal et
al. (1994) apply it to twelve cases of privatization, including two Chilean electricity
firms (CHILGENER, a generator and transmitter of bulk electricity, and ENERSIS,
the distribution utility for Santiago). The Galal study found that the gains in social
welfare were substantial, but that they mostly accrued to the shareholders in the
privatized companies. Applying different social weights to the various groups in the
society affected by the privatization process, one can substantially change the social
value of the unweighted gains.
Eighth, there have also been studies, which have tried to estimate economies of
scale. The results of some of these studies justified the horizontal integration of the
electricity industry to reduce costs. Nerlove (1968) fitted a generalized Cobb-
Douglas cost function to 1955 data on 145 firms in the US ESI and detected the
presence of increasing returns to scale. Christensen and Greene (1976), using data on
124 firms in 1955 and on 114 firms in 1970, examined economies of scale by fitting
a translog cost function to the data in the two years mentioned above. They
concluded that the use of the more general translog functional form resulted in a
flattening of the U curve identified by Nerlove for the 1955 data. Huettner and
Landon (1977), using data on 74 US utilities in 1971 conducted a comprehensive
study of returns to scale in the ESI. They confirmed the non-robustness of the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
62
Nerlove study. They found that out of all the categories of expenses only the sales
expenses exhibited increasing returns to scale over the whole of the output range.
Roberts (1986) stressed the need to take into account the effect of service area
characteristics on scale economies. Hausman and Neufeld (1991), using best-practice
econometric techniques, compared 218 private (unregulated) and 97 municipally-
owned U.S. generating stations. They found that the municipalities had significantly
higher productive efficiency (9 percent higher) than the private undertakings, mostly
because of greater scale economies.
Finally, some studies try to compare the efficiencies of various countries. Yarrow
(1988) compares the time taken to construct power stations in different countries (a
major determinant of the capital cost of electric generation). He finds that Japan,
West Germany, and the United States, where public ownership is less pervasive,
have considerably shorter construction times than Britain and Italy with state-owned
generation (Japan having less than half that of Britain). But France does well despite
state ownership, and private utilities in Belgium and Holland do rather poorly. This
may reflect more on the comparative efficiency of the construction companies in
those countries than their ownership.
As far as studies pertaining to the productive efficiency in electric utilities are
concerned, we come across two different approaches. Fare et al. (1985) used a non-
parametric approach to the testing for the effects of ownership on productive
efficiency. Atkinson and Halvorsen (1986) employed a parametric approach in this
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
63
study. They estimated a translog shadow cost function for the data on an
international sample of electric utilities operating in 1986. Both these studies used
the same data set and arrived at broadly similar results: there is little significant
difference in performance between the two ownership types - public and private.
Restructuring and Ownership
David Newbery (1999) analyzes comprehensively the different facets of
restructuring, ownership and regulation of network utilities. He states that the great
innovation of the post-oil-shock period was not so much privatization as
liberalization and/or restructuring. If regulation could be confined to the core natural
monopoly network, and competition introduced for the services supplied over the
network, then efficiency and innovation could be encouraged. The most important
problem to address is choosing the right structure for the utility that will limit the
need for necessarily inefficient regulation. Vertical separation has the advantage that
given adequate competition, regulation can be confined to the network (pp. 3).
Competition is difficult to sustain in government-owned utilities (though it may be
possible where municipally owned utilities can compete in a national market, as with
Norwegian electricity), and so there is a natural complementarity between
privatization and competition. In that sense, privatization seems to be a necessary but
not sufficient step to achieving the benefits of competition (pp.4).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
64
Privatized monopolies have many of the drawbacks of public monopolies, with the
added disadvantage that the government no longer has the power to order their
reorganization and restructuring. “In Britain the combination of private ownership,
liberalized access and regulation that had a duty to foster competition eventually
induced a degree of post-privatization restructuring, arguably at considerably higher
cost than if it been done before privatization” (Newbery, 1999; pp.5).
Newbery argues that policy toward network utilities must address three questions:
how to ensure that the large amounts of specific sunk capital are financed, and
specifically, how property rights in this capital are to be defined, allocated, and
protected; what the appropriate structure of the utility, both vertically and
horizontally, is; how best to secure the efficient use of the existing networks.
Besides addressing the above questions, the government has to restrain regulatory
opportunism without incurring excessive transaction costs. Spiller (1993, 1994)
argues that these restraints need to operate at two levels: the regulatory agency itself
needs to be restrained and its discretion limited. But the government also needs to be
restrained from changing the regulatory framework. If these restraints are to be
credible, they will require institutions to enforce them. Countries differ in their
institutional endowments, which include the legislative, executive and judicial
institutions, norms of behavior, administrative capability, and the degree of social
consensus within the society. These differences will condition the kind of restraints
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
65
that can be placed on regulatory opportunism, and whether regulated private
ownership is viable (p. 55).
Regulatory flexibility may be desirable to encourage efficiency, by retaining the
right to disallow imprudent costs. But there is a fine line between flexibility and
arbitrary discretion. Regulatory discretion needs to be restrained by dispute
resolution mechanisms that are costly to invoke, deterring actions that undermine
credibility. The government will hold the ultimate right and the responsibility to deal
with any possible regulatory failure. If there is deep-seated dissatisfaction with the
working of utility regulation, the government may initiate an inquiry, issue a White
Paper, and ultimately enact new legislation to remedy the defect (Newbery, 1999).
The trend is toward a more rule-based approach that may constrain regulatory
discretion and penalties that may restrain utility opportunism.
If regulatory institutions are not sufficiently strong to provide adequate
credibility, then private ownership may be infeasible or too costly. The costs may
take the form of a high rate of return required to reward investors for the high
perceived regulatory risks (Newbery, 1999; pp. 73).
Countries do differ in their institutional environments. Therefore, comparisons
between utilities in different countries are particularly difficult. Moreover, input and
factor prices differ as do the quality of managerial talent and the labor force. It is
noticeable that in some countries both public and private enterprises are more
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
66
efficient than both ownership forms elsewhere, perhaps because of the quality of the
institutions and their ability to make intertemporal commitments (Newbery 1992a).
The main lesson to draw is that the quality of regulation is a key determinant of
performance, whether the utility is public or private. Compared to the quality of
regulation and institutions, ownership seems relatively less important, though there
may be more chance of high-quality regulation under private than public ownership.
Ownership changes in both directions often coincide with (and may be provoked by
or induce) structural and managerial changes that can have significant effects on
factor productivity (Newbery, 1999; pp. 127).
The real case for privatization must be that it is easier to sustain efficient pricing
under private ownership in the face of political and populist pressures, and that
privatization will generate additional efficiency gains. The case for privatization is
that it locks in hard-won regulatory reforms that continued state ownership might
erode. The case against hurried privatization is that it foregoes the option of future
restructuring and better regulatory design (Newbery; 1999, p.128).
As Newbery (1999) correctly points out, if the vertically integrated company is to
be restructured/unbundled, then a great deal of additional work is required. The
assets need to be identified and reallocated, as do contracts, liabilities, employment,
pension assets, and the like. The financial structure must be designed and tested for
robustness, pro-forma accounts prepared, and a past history of accounts relating to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
67
the assets of the new firm created to convince financial analysts of the commercial
viability of each proposed firm.
Thus, creating competitive and efficient markets for network services like
electricity paradoxically requires sophisticated regulatory institutions and traditions.
The problems of unreasonable discretion may become less severe as case law
accumulates. The existence of a free press, open hearings and telling orders also help
reduce the potential for regulatory indiscretion.
Restructuring and Competition
Acutt, Elliott and Robinson (2001) maintain rightly that privatization alone is not
necessarily a sufficient condition to ensure full consumer benefits in terms of, for
example, lower prices to both industrial and domestic customers, greater efficiency
and improved quality of service. Adequate competition is also required. The authors
suggest that in the absence of competitive markets, the threat of regulation, as well as
direct regulation of firms, can affect incumbent firms’ pricing and output decisions.
The credible threat of regulation can ensure that market power is not abused by the
firms/utilities. The authors further suggest that if the regulator can establish a
credible reputation for intervention in the face of the abuses of market power, then
constant direct regulation of firms may be avoided.
There are two major ways of introducing competition into network utilities:
unbundling, or separating the competitive activities from the core network, and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
68
liberalizing access to a vertically integrated utility. Different countries have
employed different strategies to introduce competition. Some have preferred to bring
in privatization before bringing in competition, assuming that privatization would
bring in competition. Others have brought in competition before any change of
ownership.
The main goal of utility reform in Britain was to transfer ownership to the private
sector rather than to introduce competition. In contrast, in New Zealand and Norway
the main goal was to improve the efficiency of the network utilities by introducing
competition, without necessarily privatizing them. This raises an obvious question: Is
it possible for publicly-owned utilities to be restructured and regulated to introduce
competition in such a way that the benefits of competition are realized without the
state’s losing its claim on the increased profits from increased efficiency?
If we look around the world, it is clear that there are a number of possible models
of competition involving state-owned utilities. Norway, in common with several
other Scandinavian countries, has municipally-owned generation and distribution
companies competing in a power pool. The innovation in Norway was to restructure
the industry to create direct competition.
Municipal ownership differs from state ownership in several important respects.
Different companies with the same state owner might be encouraged not to compete,
or only to compete in certain allowed areas. However, most obviously different
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
69
utilities in the same industry have different owners under municipal ownership, and
are thus motivated to compete (Newbery, 1999).
Summary
Almost all the studies that try to estimate the productivity gains underscore the effect
of ownership. Many of these studies attempt to identify scale effects between public
and private firms/utilities. They also try to estimate the effect of privatization on the
efficiency and performance of the electric utilities, especially in the areas of
electricity prices and costs. Some studies have also tried to analyze the social welfare
effects of privatization.
As we see in the above review of the literature, these studies do not try to estimate
the effect of restructuring on the performance of the unbundled utilities under the
new regulatory regime. They usually treat restructuring as a part of the privatization
process, though they represent different dimensions of reform.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
70
CHAPTER 3: INDIA’S ELECTRICITY SECTOR, THE ORISSA CASE, AND
THE INTERNATIONAL COMPARISON
In this chapter, we discuss the Indian electricity sector, its evolution, its regulatory
framework, and the political economy issues. We also discuss the special case of
Orissa, which pioneered the reforms in the sector. Then we compare the power sector
of India with those of Chile, Hungary and Norway. This enables us to compare and
contrast the structure and functioning of the electricity sector in various countries.
INDIA’S ELECTRICITY SECTOR
India has a democratic and federal structure. Energy as a subject has been in the
Concurrent List in the Constitution of India (implying that it is in the ambit of both
the Federal and the State Governments). One can see that both the federal and the
state governments have played significant roles in the evolution of this sector.
India did not have an independent regulatory framework. After independence, the
country chose to have a “socialist pattern of society.” The Constitution of India
provided for state control of the “commanding heights” of the economy. Critical
infrastructure like power and telecommunication was owned and managed by
government, both at the federal and state levels.
One of the Directive Principles of State Policies enshrined in the Constitution
requires the government to ensure that the ‘operation of the economic system does
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
71
not result in the concentration of wealth and means of production to the common
detriment.’ In keeping with this requirement, specific consumer products and
consumer groups were subsidized. The government tailored the electricity tariff
accordingly, resulting in large-scale cross-subsidization. Energy pricing was used to
achieve multiple goals: increased access to commercial energy forms for lower-
income groups, increased productivity in agriculture by encouraging irrigated
agriculture, and a wider dispersion of industry to include backward and rural areas.
The Regulatory Framework of India’s Electricity Sector
The power sector was operated directly by the government at both the federal and
state levels. There was no room for an independent regulatory authority. So after
1991 the governments had to amend Acts and Rules, and bring about new legislation
to set up independent regulatory commissions, unbundle the SEBs, and bring in IPPs
(independent power producers). Most of these Acts have remained with some
amendments to permit the new changes. The Electricity Bill, 2000 was to supercede
all previous Acts and Rules. But it is yet to be enacted into law. The federal
government and the states are still discussing various aspects of this bill.
The statutes that govern the electricity sector in India are:
• The Indian Electricity Act of 1910
• The Electricity (Supply) Act of 1948
• The Indian Electricity Rules of 1956
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
72
• The Electricity Law (Amendment) Act of 1991
• The Electricity Regulatory Commissions Act of 1998
• State Reforms Acts (e.g., the Orissa Electricity Reform Act of 1995)
The technical aspects of the electricity business are specified in the Indian Electricity
Act of 1910 and the Indian Electricity Rules of 1956. The most important regulations
from the point of view of private power licensees and private power companies, i.e.,
the IPPs (Independent Power Producers) are the Electricity (Supply) Act of 1948 and
the Electricity Law (Amendment) Act of 1991, respectively. The Electricity Law
(Amendment) Act of 1991 specifically relates to setting up the new generating
capacities by the IPPs. The last two of the above Acts refer to the setting up of
independent Regulatory Commissions at the federal and state levels. Most
importantly, however, the Electricity Regulatory Commissions Act of 1998 (ERC
Act) requires the electricity regulatory commissions to promote competition,
efficiency and economy in the electricity industry, to encourage investment and to
safeguard the consumer interest.
The Evolution of the Sector
In a comprehensive survey and analysis of the Indian power sector, Dubash and
Raj an (2001) have referred to four major moments in history. These four phases can
be categorized as: from independence to 1991, 1991-1995, 1995-1998 and post-
1998. We use these “four major moments in history” to discuss the evolution of
India’s power sector.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
73
(1) Pre-1991 Phase: When India got independence in 1947, private companies or
local authorities supplied more than four-fifths of the total generation capacity in the
country, which amounted to slightly less than 1400 MW (World Bank 1993c). Then
came the Electricity (Supply) Act of 1948. This Act led to the creation of the state
electricity boards (SEBs). The Industrial Policy Resolution of 1956 reserved the
generation, transmission and distribution of electricity almost exclusively for the
government. Accordingly, the industry was nationalized, and the SEBs were created
in various states to ensure coordinated and efficient generation, transmission and
supply of electricity within each state. The Central Electricity Authority was
established to develop a national power policy and ensure coordinated development
of the power sector in India.
The SEBs have functioned entirely at the direction of the state governments.
Section 78(A) of the Electricity (Supply) Act entitles state governments to give
policy directions to the SEBs. The Act also stipulates that members of each SEB are
appointed by the government of that state. By 1991, these SEBs controlled over 70
percent of the generation and virtually all distribution. They controlled almost all the
transmission lines as well.
In several important ways, the pre-1991 institutional arrangements were
remarkably effective in accelerating the development of electricity services in India.
Between 1948 and 1991, they were responsible for increasing generation capacity
more than 50 fold, at the rate of 9.2 percent per year. Electricity generation, which
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
74
was only about 4.1 billion kWh in 1947, rose to 448 billion kWh in 1998-99. The
government played a crucial role in this remarkable growth. By 1998-99, the federal
government controlled 31 percent and the state governments controlled 65 percent of
the total generating capacity.
Out of the total generating capacity of 93,249 MW, thermal plants account for
more than 72 percent. However, these plants account for over 79 percent of the total
power generated in the country. The 15th electric power survey (EPS), conducted by
the Ministry of Power in the federal government estimates the increase in demand at
a compounded annual growth rate of 7.12 percent in the present plan period (until
2002). The growth in the demand for electricity has traditionally been higher than the
growth in generating capacity.
These SEBs have remained vertically integrated, combining the functions of
generation, transmission and distribution at the state level. The performance of these
organizations has been far from satisfactory, with increasing financial losses,
inefficient management, and inadequate and unreliable power supply. Maharashtra
SEB, the largest SEB, has, of course, been quite different. It has performed
consistently better than other SEBs. Relatively lower transmission and distribution
losses, existence of the highly industrialized Mumbai-Pune belt, and commercial
functioning have contributed to this performance. However, other SEBs have left a
lot to be desired. As Table 3.1 shows, the rates of return are negative and growing
more so over time. In view of the burden of these increasing financial losses on state
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
75
budget, there has been an understandable enthusiasm to restructure the SEBs by
unbundling generation, transmission, and distribution, and by encouraging
competition to enhance efficiency.
Table 3.1
RATE OF RETURN ON CAPITAL
(Without Subsidy)
(%)
SEB 1992-93 1993-94 1994-95 1995-96 1996-97 1997-98 1998-99
(RE) (AP)
1. Andhra Pradesh -0.20 -0.60 -22.30 -28.80 -21.80 -12.40 -19.30
2. Assam -43.30 -47.30 -25.70 -32.10 -42.60 -44.40 -37.90
3. Bihar -20.00 -12.70 -19.00 -12.90 -27.60 -24.30 -15.60
4. Delhi(DVB) -26.20 NA NA -29.90 -27.80 -28.60 -32.50
5. Gujarat -16.50 -14.60 -14.30 -24.90 -22.90 -27.40 -29.30
6. Haryana -26.10 -31.20 -27.90 -31.80 -39.30 -34.00 -46.00
7. Himachal Pradesh 0.50 -14.90 -2.50 2.60 4.30 -3.40 -6.50
8. Jammu & Kashmir -39.10 -50.10 -51.70 -48.20 -56.70 -55.20 -51.80
9. Karnataka -2.00 -0.20 -11.40 -29.40 -35.90 -15.20 -8.80
10. Kerala -11.40 -11.40 -17.90 -24.10 -53.00 -45.20 -16.40
11. Madhya Pradesh -14.60 -9.60 -13.50 -14.30 -10.30 -17.10 -29.40
12. Maharashtra 3.10 3.10 4.10 -5.30 -1.20 1.50 2.20
13. Meghalaya -7.90 -4.00 -6.90 -9.60 -6.80 -8.60 -55.10
14. Orissa -8.70 -13.50 -10.20 -21.50 -15.80 -15.50 -13.30
15. Punjab -19.90 -20.90 -19.40 -21.10 -20.30 -47.90 -47.70
16. Rajasthan -11.40 -17.90 -19.10 -16.00 -23.60 -23.70 -27.20
17. Tamil Nadu -8.80 -9.70 -0.10 -1.90 -5.80 -7.70 -6.90
18. Uttar Pradesh -16.70 -17.80 -12.20 -9.60 -14.10 -12.40 -12.90
19. West Bengal -35.30 -29.70 -42.20 -56.10 -74.30 -82.40 -72.40
Average : -12.70 -12.30 -13.10 -16.40 -17.90 -18.00 -18.70
(Source: Planning Commission of ndia, 1999)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
76
Privatization of these boards or at least of the generation part, has been proposed in
various states to attract more capital and better technology, and raise the efficiency
of management practices in operations. It is also being advocated on the grounds that
governments at the federal and state levels can now afford to allocate more resources
to the social sectors like health and education, and infrastructure development.
The country started to experience severe power shortage in 1980s. Most of the
SEBs’ operational as well as financial performance left a lot to be desired. The
federal corporations like NTPC did not have enough to invest. The power sector
badly needed some basic changes in policy and its implementation.
(2) The Next Phase (1991):
The year 1991 has become a watershed in the history of India’s macroeconomic
reforms. Many factors contributed to a policy shift. First, the balance of payments
crisis and the overall problems in the macroeconomic front set the ball rolling and
provided the immediate impetus to reform. Second, acute problems within the power
sector (peaking shortages, severe financial crisis in SEBs, very poor quality of power
supply etc.) called for change. Third, with the fall of Soviet Union, liberalization
started looking more attractive. Fourth, politicians/bureaucrats found the prospects of
reforms attractive in a globalizing world because it seemed to offer them special
opportunities to increase the resources under their control (Echeverri-Gent 2000).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
77
Fifth, the World Bank and the IMF also ‘persuaded’ India to reform its
macroeconomic structure. The international donor community was sending explicit
warnings that countries should not rely exclusively on multilateral finance for their
development needs, but would have to increasingly turn to private capital. Sixth,
there was also pressure from the people, fed up with brownouts and blackouts.
Seventh, there was an increasing demand to allocate more government funds to
social sectors like health and education.
Finally, apart from all these factors, the prime minister and especially the finance
minister (a Western-educated economist) wanted to bring about the policy shift from
the erstwhile state-controlled economic framework to a liberalized and more market-
friendly structure.
Reforms in the power sector began in October 1991. The Electricity Laws
(Amendment) Act of 1991 radically revised prevailing legislation by permitting
private entities to establish, operate and maintain generating power plants of virtually
any size and to enter into long-term power purchase agreements with SEBs.
This phase witnessed the formulation and implementation of the policy to
encourage independent power producers (IPPs). The initial government notification
provided generous incentives to these producers (IPPs). The most noteworthy of
these incentives was a guaranteed minimum 16 percent (repatriable) return on equity
for plants that operated at their rated capacity for at least 6000 hours in a year, with
additional bonuses for improved capacity utilization.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
78
Other attractions for potential investors included a five-year tax-holiday, a two-
part tariff (the first part covering fixed cost and the second covering variable costs),
equity requirements that were as low as 20 percent of project costs, and selective
counter-guarantees from the federal government to cover payment default by SEBs.
The rules were clearly intended to attract foreign private capital into the sector. The
response was overwhelming. By mid-1995, there were 189 offers to increase
capacity by over 75 GW involving a total investment of over US $ 100 billion.
But for various reasons, this reform program significantly under-performed. The
potential investors were indignant about delays in obtaining clearances and hurdles
to securing adequate fuel supply, and were generally apprehensive about the
recovery of electricity sale proceeds from SEBs and the country’s overall political
stability. There were also instances in which the legacy of older institutions hindered
IPP development. For instance, the IPPs found it complicated to secure contracts for
Indian coal because the vertically integrated SEBs had traditionally defaulted on
payments to public companies managing coal and railways. These companies still
perceived payment risks and hesitated to accommodate the needs of IPPs. Moreover,
the Central Electricity Authority, which had the authority to make techno-economic
review and approval, and the environment departments stuck to their policy and rule.
The good part of the IPP Policy was that it actually initiated competition in the
system for the first time.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
79
As Dubash and Raj an (2001) point out, these rules to attract investment did not
enjoy the support of the power sector experts. Some of the agreements were
considered to be “outrageously lopsided, with unjustifiable risks having to be borne
by the SEBs and, eventually for projects with counter-guarantee, by the central
government.” Secondly, they argued that the possibility of giving SEBs true financial
and administrative independence from state governments with phased solutions to
the agricultural subsidy problem was never really explored. Third, existing public
sector generating companies like National Thermal Power Corporation (NTPC) were
not being offered a level playing field with the IPPs. Fourth, the IPP policy generated
new opportunities for flagrant violation of environmental norms. While most of the
environmental opposition to IPP projects was organized at the community level, a
few national campaigns were also conducted, one resulting in a Supreme Court
verdict directing a national environmental research agency to review the
environmental clearance.
(3) Mid-1990s Phase: Restructuring in Orissa with the World Bank’s financial and
‘technological’ help started in 1993 and took concrete shape in 1995. This ‘Orissa
Model’, as it is referred to in India, involved the ‘unbundling’ of the vertically
integrated SEB into separate components of generation, transmission and
distribution, the setting up of an independent electricity regulatory commission for
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
80
transparent and independent tariff-setting, and a phased privatization in generation
and distribution.
The World Bank suggested that the problems of the power sector were the logical
outcome of a conflict of interest “between government’s role as owner and its role as
operator of utilities.” The consequences were the “opaque command and control
management, poorly defined objectives, government interference in daily affairs, and
a lack of financial autonomy” (World Bank, 1993b).
Various critics, notably the French utility Electricite de France (EdF), raised
several concerns about the efficacy of privatization (World Bank; 1993a). First,
privatization needed regulatory mechanisms that would be “more complex and more
cumbersome than in an integrated monopoly.” Second, whether these countries
possessed the ‘institutional maturity and balance of power’ to sustain independent
regulation was uncertain. Third, privatization might not attract long-term investors
in developing countries having few resources. Fourth, unbundling, it was feared,
would distort the advantages of economies of scale and scope needed to expand the
sector. Finally, privatization would introduce “contradictions of interests between the
shareholders and the essential public goals” (World Bank 1993a). The Bank’s
contention seemed to be that its framework was flexible enough to address these and
similar concerns.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
81
THE ORISSA CASE
The reform in Orissa, and particularly the passage of the Orissa Electricity Reform
Act of 1995 was a significant landmark in that it marked a departure from the
framework for the power sector enshrined in national laws. It set the stage for a
complete rethinking of the institutional basis for the sector in India. At that time,
some argued that the Orissa Act was unconstitutional. There was also little support
for it from the Federal Ministry of Power, and outright hostility from the Central
Electricity Authority, which presumably feared erosion in their authority and
responsibilities as a result of the law. The law cleared the federal government after
active lobbying by Orissa’s chief minister (Dubash and Rajan, 2001; pp.3378).
Although the Electricity Laws (Amendment) Act of 1991 had paved the way for
private sector participation, especially in generation, it had not provided for any
restructuring of the state electricity boards. As the first homegrown experiment,
Orissa established a precedent for states to deviate from the well-trodden national
path of vertical integration and to opt for a restructured electricity sector.
Some authors like Rajan (2000) opine that the main reasons that Orissa was the
pioneer state were political. First, and perhaps most significant, was strong support
from the chief minister Biju Patnaik for reforms. In other states, notably Haryana,
politicians demonstrated little appetite for a menu of short-term costs associated with
reforms, including electricity price increases and staff lay-off, for the promise of
long-term benefits. Significantly, the succeeding chief minister, J B Patnaik,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
82
maintained this strong level of support. Second, Orissa had no strong farm lobby that
could act as a political brake on reforms. Agricultural consumers accounted for an
insignificant 5.7 percent of total sales, as compared to around 40 percent in many
other states (Rajan, 2000; pp.666). Finally, some have argued that the World Bank
chose a relatively poor state, with a small power sector, low levels of political
mobilization and a minor national profile as an experiment that would fly below the
national radar screen. In sifting through these reasons, the chief minister’s support
appears to be the single most crucial factor in thrusting Orissa to the fore (Dubash
and Rajan, 2001; pp.3377).
Orissa opted for a single-buyer model based on an assessment of the underlying
technical, institutional and commercial base in the state, and its relative inability to
support wholesale or retail market forms. In this model, the vertically-integrated SEB
was ‘unbundled’ into generation, transmission and distribution. The transmission
remained under government control, and was to act as the single buyer of electricity
from the generating units and seller to the distribution companies. Private
participation in generation and distribution was allowed, bringing in more
competition in these sub-sectors. While unbundling the SEB, the socio-political
specificity and institutional context were kept in mind. For example, the political
leadership was aware of the need to minimize retrenchment of workers to avoid
opposition by trade unions and some political parties (specially the left-wing parties).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
83
Next, in order to place the finances of the generating and distributing companies
(being prepared for privatization) on a sound footing, financial liabilities were
effectively transferred to the only segment destined to remain in state hands, the
state-owned transmission company, GRIDCO.
Figure 3.1 depicts the structure of Orissa’s unbundled electricity sector. The
generating units include the National Thermal Power Corporation (NTPC), the
Orissa Hydro Power Corporation (OHPC), the partially privatized Orissa Power
Generation Corporation (OPGC), the private sector plants and the captive power
plants (CPPs), belonging to both private sector and public sector.
STRUCTURE OF THE
ELECTRICITY SECTOR IN ORISSA
Central Sector
(GOI-owned)
GOO-owned
initially,but to be
privatised
G O O -O w ned,
Private Sector
Private /
Public Sector
Independent
Generation
Sources
NTPC s .
\
\
OHPC
\
\ \ / 7 '
/
/
OPGC
IPPs / /
/
/
/ )
/
GR
GOO-owned
Privatised
Distribution
Companies
WESCO
NESCO
SOUTHCO
CESCO
Customers
Customers
Customers
Customers
Orissa is the only State in India to have completed Privatisation o f Distribution
Figure 3.1: Structure of the Electricity Sector in Orissa
Source: An Overview of GRIDCO, 2001
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
84
The four distribution companies (DISTCOs) buy electricity from GRIDCO, and
sell it to consumers. These companies (WESCO, NESCO, SOUTHCO, and CESCO)
have been privatized since 1998.
Orissa’s experience has been an eye-opener for other states, and even the federal
government. Even the World Bank’s assessment went awry, and their appraisal
report had to be redone. During the processing of the World Bank loan in early 1996,
Orissa State Electricity Board (OSEB)’s system losses were estimated at 43% for the
year 1996 and on that basis GRIDCO’s system loss reduction targets were fixed as
follows, aiming for a 19% reduction over a six-year period.
Table 3.2: GRIDCO’s Initial Loss Reduction Target
1997 1998 1999 2000 2001 2002 2003
39.5% 34.8% 29.2% 24.3% 22.7% 21.7% 20.6%
(Source: An Overview of GRIDCO, 2001)
During the project implementation, with the availability of much more detailed
information and data, it turned out that the starting point, the 1996 base figures, were
much higher, on the order of about 52 to 53 percent for OSEB’s last year of
operation in 1996 and 50.5 percent for GRIDCO’s first year of operation.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
85
Accordingly, it was necessary to extend the program period by two years, with the
targets revised as below:
Table 3.3: GRIDCO’s Revised Loss Reduction Target
1997 1998 1999 2000 2001 2002 2003 2004 2005
50.5% 47% 41% 35% 29% 23.7% 22.7% 21.7% 20.6%
(Source: An Overview of GRIDCO, 2001)
Distribution Privatization
The key to ensuring a financially sound and self-sustaining power sector is
efficient management of the distribution business, because this is the source of the
cash flow needed to sustain generation. Talvadkar and Mishra (2000) argue that a
major weakness in the Indian power system has been inefficient management of the
distribution business, which needs to be remedied through sector reform and
subsequently privatization. They maintain that the advantages of distribution
privatization are manifold. It will create real incentives to cut commercial and
technical losses, thereby mitigating the need for tariff increases, or perhaps even
facilitating end-user tariff reduction. Distribution loss reduction could result in
substantial cash infusion into the states’ ailing power sectors. Extensive Latin
American as well as the limited East Asian and Indian experience has shown that an
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
86
aggressive loss reduction program can be implemented in a DISTCO (Distribution
Company) under private ownership and management. In most of the selected
DISTCOs in Latin America that tried a loss reduction program, over 50 percent of
the total loss reduction was achieved in the first three to four years of privatization.
According to the report of the Planning Commission of India, in case of Orissa, the
transmission and distribution losses have fallen from a level of 50 percent in 1996-97
to 43 percent in 1999-2000 (Annual Report, 2001; pp. 166).
In the specific case of Orissa, through a process of international bidding, GRIDCO
offered a 51 percent stake to private sector, keeping 39 percent of the shares with it
and 10 percent share for the Employees Welfare Trust. BSES (a private Indian
company) obtained three of the four distribution zones in the competitive bidding.
Later, AES Transpower, a US-based multinational power company, was given the
distribution rights (through negotiation) in the crucial central zone of the state. It also
owns a substantial share in the thermal power generating company, the Orissa Power
Generation Corporation (OPGC).
At the end of two years of privatization, the distribution companies (DISTCOs)
are showing distinct improvements in billings and collection. However, they
continue to show financial losses, even though the level of losses has dropped (Table
3.4). WESCO made a profit in 1999-2000. Others have reduced their losses. These
losses are mainly because of their inability to meet the transmission & distribution
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
87
loss targets set by the regulatory commission. Even after privatization, GRIDCO is
indirectly funding the losses of the DISTCOs through receivables (Sinha, 2002)
Table 3.4: Profit / Loss of Distribution Companies before Tax (in Crores1 of
rupees)
WESCO NESCO SOUTHCO CESCO
1998-99 -39 -55 -80 -193
1999-00 16 -46 -42 -128
(Source: Sinha, 2001) One Crore = 10 million
Table 3.5: Profit / Loss (without subsidy) of various units in Crores of Rupees
1997-98 1998-99 1999-00 2000-01
OHPC 77.78 55.21 50.38 -2.94
OPGC 66.15 112.80 124.38 109.86
GRIDCO -324.43 -588.50 -111.44 -87.29
DISTCOs -383.21 -324.65
TOTAL -180.50 -420.49 -319.89 -305.02
(Source: OERC, 2001)
Table 3.5 shows the profit/loss of all the restructured parts of the electricity sector
since 1997. One can see a J-curve effect in the overall performance. There is a
sudden dip in 1998-99, and then a recovery.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
88
The distribution companies (DISTCOs) have taken concrete steps for better
collection of revenues. For example, they have stepped up billing and collection of
dues from all consumers including government departments and undertakings.
Second, non-payment of dues has resulted in the disconnection of the power supply.
Third, they have held consumer camps (known as “Bijuli Adalats”) for redress of
complaints and collection of dues. Fourth, they have also used the services of village
committees for rural metering, billing and collecting dues. The results have been
quite good. The collection among domestic and commercial consumers in the 100
villages that were taken up as a pilot project by the distribution company, WESCO,
has increased by 107 percent. This collection was achieved without a single threat of
disconnection on the part of WESCO. Interestingly, in some cases the village
committees themselves recommended disconnection (Sinha, 2001; pp. 58). This
limited experience does suggest that there are potential collateral benefits arising
from decentralized forms of organization in the sector following a loosening of state
control.
Education/training of the committee members and the consumers is being
undertaken in the villages to improve the performance of these committees. Apart
from improving the collection, this process has been able to open the lines of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
89
communication between the rural consumers and the electricity distribution
companies. These committees are to function as Customer Care Centers in the
villages.
Fifth, GRIDCO and the DISTCOs have undertaken a program to convert LT lines
to 11 KV lines through the installation of additional single-phase and three-phase
transformers at appropriate locations. This will not only reduce the technical loss,
but also help minimize theft of power.
Sixth, there has also been a concerted effort to repair defective meters, and to
provide new meters. This is absolutely necessary to monitor the use of electricity,
and reduce the transmission and distribution (T&D) losses.
Finally, GRIDCO and DISTCOs have also tried to regularize some unauthorized
connections, and to bring them into the billing net. This has also helped improve the
tariff collection.
Regulatory Commission
The Orissa Electricity Reform Act of 1995 forms the basis of the new regulatory
regime in Orissa. The state government’s direct control has given way to
independent regulatory control. The Orissa Electricity Regulatory Commission
(OERC), established by the above Act, has a broad mandate to govern the electricity
sector in the state, to issue licenses, regulate the purchase and use of electricity, set
tariffs and ensure quality of service, maintain consumer interest, and promote
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
90
competition. It is the “lynchpin in a new model aimed at independent
operation”(Dubash and Rajan, 2001).
Another big task before the Commission is to convey the message that the price
increase reflects the costs previously suppressed and subsidized by the government.
These institutions (the regulatory commissions) play key roles while providing
assurance to investors that they will be able to recover investment costs, and assuring
consumers that regulatory decisions will be taken with long-term interests of both the
consumers and industry in mind.
The Reform Act provides for various mechanisms to maintain the independence
of the commission. First, funding of all reasonable expenditures of the commission
is not subject to voting in the legislative assembly. The expenditure is funded out of
the consolidated fund of the state. Second, section-4 (6) of the Orissa Electricity
Reform Act of 1995 entitles the state government to appoint one of the two
candidates shortlisted by the selection committee as a member of the Commission.
The selection committee consists of the chairman of the Public Service Commission
as chairman, the Secretary of the state government’s Energy Department as
convenor, and the chairman or any member of the Central Electricity Authority.
Third, section-6 (1) enshrines that members shall hold office for a period of five
years and shall not be eligible for re-appointment.
Although the above provisions have helped maintain the independence of the
commission, some critics have expressed concerns regarding the following aspects
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
91
of the Act. Section- 12 (1) of the Act empowers the state government “to issue
policy directives on matters concerning electricity in the State” to the commission.
In case of any dispute between the government and the commission, the matter will
be referred to the Central Electricity Authority, whose decision will be final and
binding. Webb and Derbyshire (W&D, 2000) opine that the Orissa Act emphasizes
the role of regulator in protecting consumers; duty to protect interests of investors
seems relatively weak. There is little emphasis on the need to promote competition.
W&D favor change in the selection criteria and employment restrictions to increase
private sector expertise among commissioners. They also advocate that the legal
provisions should make it incumbent upon the regulators to ensure financial
viability of efficient companies and to promote competition.
In actual practice, the Orissa Electricity Regulatory Commission (OERC) has set
impressive standards for transparency in India. It has made public procedures
mandatory for approval of PPAs (Power Purchase Agreements). It has held open
hearings in tariff-setting cases. Labor and consumer groups have been able to
represent their cases before the commission in all these hearings.
It has also displayed appreciable independence in the decision-making. Provisions
in section 26(2) of the OER Act of 1995 have guided the Commission in
setting/designing the tariffs. It has passed four tariff orders so far. It has taken a
balanced view of the cross-subsidy in its tariff orders. It has tried to progressively re
balance tariffs. For example, during the period 1996-97 to 2000-01 while domestic
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
92
tariffs increased by 60 percent and 52 percent, industry tariffs increased by a much
smaller amount. The Commission directed the distribution companies to carry out
pilot studies for determination of technical and non-technical losses in their system
for a period of six months from April 2001 to September 2001. But they are yet to be
completed.
The OERC has clarified that the government should finance any subsidy provided
by GRIDCO on government directions (e.g., rural electrification program, subsidy
under industrial policy resolution). The tariff design took this into account. The tariff
order for 1999-2000 said,
The licensee should not take up investment on uneconomic projects which will burden the
consumers. Therefore, without firm commitment of subsidy from the government and
without approval of the Commission for the investment, the capital addition on account of
rural electrification work cannot be allowed to be included in the capital base for earning
return.
Problems in the Orissa Experiment
The Orissa case has been hailed by many as a path-breaking experiment in
India’s electricity sector. It has shown the way to other states. It has generated many
analyses and discussions regarding the right way to reform. It has also been assailed
on various grounds. Several assumptions and the consequent prescriptions have
gone wrong. According to GRIDCO, the following assumptions were wrong. It
considers them primarily responsible for its severe financial problems.
First, the World Bank Staff Appraisal Report (SAR) of April 1996 estimated the
transmission and distribution losses at 39.5 percent which was to be brought down to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
93
24.3 percent by 1998-99. Accordingly, the financial projection was made which
estimated that GRIDCO would break even during the financial year 1997-98. The
real magnitude of the distribution losses, which was about 50 percent, was not
known at that time. Based on the SAR and in the absence of reliable data based on
the detailed studies, OERC in their first tariff order (effective 1 April 1997) directed
that the losses be brought down to 35 percent at the end of the year 1997-98. This
was an impossible target.
Second, the SAR had assumed load growth of 11.4 percent in 1997-98, 16.7
percent in 1998-99 and 9.2 percent in 1999-2000. But in actual practice it turned out
to be substantially less. The big industries resorted to their own captive generation
due to very high industrial tariffs. The anticipated industrial growth also did not take
place. Both these factors led to a lower industrial demand than was projected,
thereby curtailing the income flow to the GRIDCO and the distribution companies.
Third, the SAR had not anticipated that in order to make the distribution
companies attractive to the private investors, GRIDCO would have to retain in its
own books about three times the amount of liabilities passed on to the four
distribution companies. In order to make the distribution business attractive to
private investors, only around Rs.650 crores (nearly $140 million in 2001 exchange
rate) of total liabilities were passed on to the four distribution companies. But
GRIDCO, the transmission company retained Rs.1950 crores (nearly $ 420 million)
of liability in its own books.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
94
Fourth, the SAR had assumed an average increase in tariff of 16 percent in 1996-
97 and 18 percent in 1997-98. This also did not materialize. Inadequate cost recovery
because of lower tariffs than that projected affected the financial performance of
GRIDCO from its inception.
Fifth, the increasing cost of power did not match with the retail tariff realized. The
cost of power purchased from OHPC was much more than the cost incurred earlier.
The Orissa experiment has faced problems of various kinds. Opposition from
various groups, which felt threatened for different reasons, has been quite
substantial. The reform process also witnessed unprecedented “devastation wrought
by the ‘super cyclone’ which hit Orissa in October 1999 and which left the power
infrastructure in tatters” (An Overview of GRIDCO; 2001).
Many authors like Haldea (2001) have opposed the single-buyer model of Orissa.
Haldea argues that contract-driven privatization through state-owned monopolies can
have little change of enduring success. Unlike in the West where competition in
generation and supply is the engine for efficiency gains and tariff reduction, the
Orissa model relies on an interconnected chain of monopolies where competition is
conspicuous in its absence. Orissa represents a ‘single-buyer model.’ There cannot
be a credible market in the absence of multiple buyers. The industry structure thus
continues to be in the command-and-control mode.
He further notes that the long-term power purchase agreements (PPAs) stifle
competition. Since generation typically constitutes over 75 percent of the consumer
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
95
tariff, competition among generation companies selling to multiple buyers would
bring about significant price and efficiency gains. In the restructured industry, the
transmission company must not buy or sell power. It should only transmit power on
payment of regulated wheeling charges with a view to providing open access to its
transmission network. Similarly the distribution companies must provide similar
access to their wires to enable bulk consumers to buy directly from the generation
companies. With these fundamental elements of restructuring in place, market forces
can be relied upon to bring in rapid efficiency gains as well as new investments.
(4) Post-1998 Phase: This period has witnessed the strengthening of the reform
process not only at the federal level, but at the state level as well. More states have
accepted the reform process. For example, Haryana, Andhra Pradesh, Karnataka,
Uttar Pradesh and Rajasthan have taken up restructuring of their power sectors.
Many other states have set up electricity regulatory commissions.
There was an increasing consensus that that the model of restructuring and
subsequent privatization was essentially correct, and that the problem had to be
tackled from the distribution end because poor revenue was the source of the crisis,
not lack of generating capacity as such, which had been the dominant belief in earlier
years (Ahluwalia, 2000). The Orissa model was seen as politically feasible despite its
mixed achievements.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
96
In 1998, the Electricity Regulatory Commission Act was passed, creating an
Electricity Regulatory Commission at the federal level. It allowed states to create
their own commissions either on the basis of the federal Act or through legislation of
their own as Orissa had done before. Thus, the 1998 Act marked, in a sense, the
national coming out of the Orissa model (Dubash and Rajan, 2001).
This phase also saw some key changes, based on the lessons learnt from the
Orissa experience. First, power sector reform effort has been presented within the
broader framework of state-level financial restructuring (e.g., Andhra Pradesh, Uttar
Pradesh). The goal is to avoid leaving the public component of the sector with an
unsustainable debt burden, as happened with GRIDCO in Orissa. Second, all but the
Uttar Pradesh loans were structured as ‘Adaptable Program Loans’(APLs), which are
structured to be released in small amounts over many years, each release depending
on the fulfillment of the conditions. The World Bank is not the only funding agency
in this sector. The UK’s Department for International Development (DFID), the
Canadian International Development Agency (CIDA), the United States Agency for
International Development (USAID) and Japan’s Policy and Human Resources
Development Fund (PHRD) have also provided funding for the reform. The Asian
Development Bank (ADB) is also involved with the reform process in some states
like Gujarat.
During the post-1998 period, the federal government has become increasingly
engaged in state-level reforms. Through regular meetings of state chief ministers, it
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
97
has sought to build a political consensus for the reforms. The series of these meetings
started in 1996, which is repeatedly invoked as a turning point. The chief ministers
agreed to a common minimum action plan which included handing over retail tariff-
setting to state electricity regulatory commissions, and a determination that while
cross subsidies across sectors were to continue, they were to be restricted.
Specifically, no sector would pay less than 50 percent of the average cost of supply,
with the exception of agriculture, where an absolute floor of Rs 0.50 per kWh
(considerably less than 50 percent of average cost) was set, to be raised over three
years to the 50 percent benchmark (Government of India, 2001).
The performance toward these goals has been far less than satisfactory. On the
politically difficult task of removing cross-subsidies for agriculture by 2001, only
nine out of 28 states have achieved the Rs.0.50 floor-price benchmark, and none has
achieved the 50 percent of the average cost goal. Hence these goals were reiterated in
meetings in February 2000 and March 2001. At these subsequent meetings, the chief
ministers agreed to introduce compulsory metering of electricity for all consumers,
and also for energy audit. They have also agreed to focus more on transmission and
distribution network. In the February 2000 meeting, they ended with a statement that
if these goals appeared unattainable within the existing framework, then
“corporatization/cooperatization/privatization of distribution would have to be
undertaken.” (Government of India, 2001). At the most recent meeting in March
2001, the federal government agreed to orchestrate a ‘one-time settlement’ of Rs.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
98
260 billion (nearly $ 5.5 billion) owed by SEBs to federal utilities like NTPC (The
Times of India, March 4, 2001). These meetings illustrate a change in mind-set and a
willingness to address politically tough issues.
Through the passage of legislation, the federal government has tried to walk the
line between asserting control by putting in place an overarching framework, and
allowing states a measure of flexibility. This momentum has survived political
changes. As Dubash and Rajan (2001; pp.3383) point out,
In 1998, when a new BJP coalition government was formed at the centre, many were
surprised by the zeal with which it encouraged private participation, notwithstanding the
ideology o f economic nationalism that many of its members shared The political
acceptance of a reformist discourse was enhanced by the waning political strength of the
centre-left parties, historically the main opponents of large-scale privatisation.
Electricity Bill 2000: In 2000, the power ministry initiated the drafting of a
comprehensive bill to replace all existing legislation in the power sector. The original
bill required states to unbundle their SEBs, establish independent regulatory
commissions, facilitate open access to transmission, develop a spot market for power
and meter all electricity supply (Suri 2000). Notably, although the ministry of power
supports privatization as a way to control losses, the bill does not explicitly require
privatization, but offers the states some flexibility in the form of unbundling. The
draft bill has been widely circulated to gauge public and political opinion.
Unfortunately the power minister died in August 2000 before he could present it in
the parliament.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
99
The bill has again been re-opened for debate. More ambitious market structures
like spot markets for electricity have been expunged, with an emphasis on third-party
access (TPA) and long-term contracts as a way of facilitating power trading. There
has been a growing recognition that the debt overhang that clouds the future of the
SEBs will have to be worked out. The federal and state governments also want to
focus their attention on rural electrification. A ministerial committee has started
looking into this very important aspect of electricity supply. One may see the
introduction of subsidy auctions for those willing to undertake rural electrification,
as inspired by experiences in Chile and Argentina.
The evolution of India’s electricity sector, as we saw, has witnessed various
policy shifts at the federal as well as state level. This is quite evident in
governments’ policies on investment in the electricity sector.
Investment Policy
As far as the investment in the electricity sector is concerned, government policy has
evolved in three major phases after independence in 1947. The first phase, from
1948 to 1975, was characterized by the growth of state-level investments in the
power sector and the decline of private investments and management. The second
phase, from 1975 to 1990, marks the ascension of federal government utilities in
generation and transmission. The supply situation soon changed from surplus to
shortage. The SEBs lacked the resources to invest in additional capacity creation to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
100
keep up with the growing demand. In this changed scenario, the federal government
had to intervene to build additional capacity, and to expand the transmission and
distribution network on a large scale.
The Electricity (Supply) Act of 1948 was amended in 1976 to establish generating
companies cutting across state boundaries. In the mid-1970s, the National Thermal
Power Corporation (NTPC) and the National Hydroelectric Power Corporation
(NHPC) were established to plan and implement power projects at the federal level.
The last phase, from 1991 onwards, marks a turning point in which government
policy has evolved to a more market-friendly approach. It has changed policies to
attract investment by private IPPs (independent power producers). Some states have
started the restructuring process with the amendment of earlier Acts and Rules,
unbundling the generation, transmission and distribution, and setting up independent
regulatory commissions for tariff-setting and other allied functions. Some states have
only unbundled their power sector. Yet, some others have just amended the Acts and
Rules necessary to carry out the restructuring process. There are, however, still some
states that have chosen to wait and watch.
India is yet to attract expected investment from the private sector, both domestic
and foreign. The debacle of Enron-managed Dabhol Power Company, and the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
101
recent withdrawal of some other foreign companies like AES from Orissa’s
distribution company (CESCO) have not augured well for such investment. The
existing labor laws have often been assailed as detrimental to investment.
It is worthwhile to point out that most of the investment has been undertaken in
the area of generation, and not in the areas of transmission and distribution. Two
features in the growth of the transmission system are notable in this regard. First, in
contrast to generation, which grew at about 9.1 percent per year during the 1980s,
the transmission and distribution (T&D) network expanded at the much lower rate
of 6.4 percent per annum.
Second, the need to expand the network to far-flung areas necessitated the
expansion of the 11 KV distribution lines. Thus the ratio of the low-tension (LT) to
high-tension (HT) lines increased from 4.1 in 1965-66 to 7.9 in 1980-81 and further
to 10.8 in 1989-90.
The number of people served per kilometer of transmission line in the Indian
T&D system is greater than the number in most countries. This reflects the fact that
the delivery infrastructure is predominantly low-voltage, which necessarily
contributes to high technical losses and poor quality of service. The large scale
expansion of the rural electrification program without corresponding strengthening
of the transmission and distribution network, contributed to this problem. In
addition to these technical losses, there are significant non-technical losses due to
pilferage and unmetered supplies (Rao, Kalirajan and Chand, 1998; pp.42-43).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
102
The rural electrification program enabled agricultural consumers to use electricity
for irrigation purpose. They were further helped by very low agricultural electricity
tariffs, subsidies and cross-subsidies. We discuss this next.
Political Economy of Subsidies and Restructuring
The Green Revolution in the late 1960s called for the exploitation of ground water
potential among other things. Irrigation had broad appeal because it seemed to be
accomplishing three important political goals— (a) achieving food security, (b)
increasing the profits of the farmers who could thereby be organized into large vote
banks, and (c) helping people towards self-employment. Providing electricity to
exploit the ground water was gradually seen as a potential political tool. During the
1977 elections, the Congress-led southern state of Andhra Pradesh offered flat-rate
tariffs (tariffs based on capacity of the pump rather than on metered consumption) to
farmers as an election promise to help the Congress Party get re-elected. The
neighboring state of Tamil Nadu where a new non-Congress party (the ADMK) that
had just come into power in a fragile four-way election, decided to offer free
electricity to some groups of farmers. Subsequently, political leaders in Maharashtra,
Karnataka and other states began to view the entitlement per se as a remarkably
effective political device. The increasing bargaining power of the large middle-class
farmers also played its role. As a result of this flat - rate/free supply, the meters were
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
103
not monitored any more, or were simply removed and returned to the SEBs. Power
subsidies to farmers have since become routine political instruments.
The tariffs for agricultural consumers have not only been low, but have in fact
declined even in current prices, from a little over 20 paise/kWh in 1975-76 to 15.8
paise/kWh in 1991-92, though thereafter it inched up to 24.5 paise/kWh in 1995-96.
Thus, the difference between average cost and the agricultural tariff per unit of
energy increased from 4 paise/kWh in 1974-75 to about 100 paise/kWh in 1990-91
and further to 146 paise/kWh in 1995-96.
The result of this underpriced or free electricity has been an unprecedented
increase in demand for power in agriculture. It also gave rise to reporting of very
high transmission and distribution (T&D) losses. This situation generated a perverse
benefit to SEBs, because they could hide their losses (theft of power) under the
category of agriculture. This situation quickly became “a race towards
unsustainability because there were virtually no incentives or regulatory checks to
save either energy or groundwater resources” (Dubash and Rajan;2001). Several
studies have shown that the large farmers have actually benefited much more than
the small and marginal farmers from the above system of underpriced or free
electricity. Those large farmers invest in irrigation, use free electricity and grow high
value crops. Sometimes they even sell water to other farmers (Sant and Dixit, 1996).
Even though the constituents who stand to gain most from the supply of free
electricity may be in the minority, they have appeared to be those who maintain just
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
104
enough swing power to capture the attention of most political parties. Some studies
have shown that since the quality of the power actually delivered to farmers has long
been extremely poor (unreliable supply causes damage to the crop), most farmers are
“likely to prefer metered and priced reliable electricity to unmetered, free (or low-
tariff) unreliable electricity” (Reddy, 2000).
Besides the above direct subsidies, the system has thrived on indirect cross
subsidies as well. Curiously, in India the price charged to different categories of
consumers varies inversely with the average cost of supplying power. The average
cost of supplying electricity to LT (Low-Tension supply) consumers-who are
mainly agricultural and domestic consumers — is the highest. But the price charged
to these consumers is the lowest, whereas the price charged to the industrial and
commercial consumers (power supplied through HT (High-Tension) lines, and thus
the least power loss) is the highest.
This system of subsidies and cross-subsidies is likely to change in the new era of
restructuring. The independent electricity regulatory commissions are likely to work
towards the “rationalization of electricity tariff’, as enshrined in the preamble of the
Electricity Regulatory Commissions Act of 1998. However, the prevailing system of
subsidies to help provide underpriced or free electricity to agricultural consumers has
been hailed by the farm lobby, and by various political parties. These parties have
tried to interpret it as being in the public interest.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
105
Although Indian laws explicitly use the expression “public interest,” no Act has
defined the term. The state is usually considered as the legal representative of the
interests of the citizenry. Therefore, all the actions of the state have been deemed to
be in public interest. After the restructuring and the setting up of the independent
regulatory commissions, the governments become just parties in transactions
involving consumers (public), workers, and industries, especially in tariff hearings.
It is assumed that the regulatory commission will be independent and objective.
The Act also expects the commission to protect the public interest. The obvious
questions that crop up are whether the regulatory changes are in the public interest,
and whether the independent regulators are equipped and empowered to interpret and
act in the public interest.
Dissemination of information is extremely important in order to make the reform
process sustainable. This is more so in the case of developing countries like India,
where considerable portion of the population does not have direct access to mass
media. As Webb and Derbyshire (2000) point out, in the pre-reform period,
information on performance of the sector relative to similar organizations in other
countries/states should be published. The need for tariff increase to bring in future
improvement in service quality should also be explained. The possible benefits of the
reforms also should be disseminated. In the post-reform period, the progress and the
problems should also be published and explained in local languages. Commitment to
employees should be visibly met early on. All these steps will help the reform
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
106
process by projecting it as serving public interest, and thus, will create a constituency
for reforms.
The Supreme Court of India has held that “the concept of public interest must as
far as possible receive its orientation from the Directive Principles, enshrined in the
Constitution of India. If any governmental action is calculated to implement or give
effect to a directive principle, it would ordinarily, subject to any overriding
consideration, be formed in public interest” (Kasturilal Lakshmi Reddy vs. State of
Jammu and Kashmir - 1980SCCI).
Legal provisions have arrogated the responsibility (of interpreting what is in
public interest) to the Government. Section 38(1) of the Electricity Regulatory
Commissions Act of 1998 says, “In the discharge of its functions, the Central
Commission shall be guided by such directions in matters of policy involving public
interest, as the Central Government may give to it in writing.” This provision says
that the government can issue directions to the commission only when policy matters
of public interest are involved. This policy directive has to be given in writing, which
means that the government has to justify its directive explicitly.
Different political parties have viewed liberalization differently. Their election
manifestos and their parties’ official position have tried to interpret public interest in
their own ways. Though there are many political parties in India, we will discuss
three of the major parties here.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
107
(1) Congress: India’s oldest political party started liberalization of the economy in
1991. It has been quite supportive of restructuring and privatization. But the
party does not specifically mention it in its election manifesto. The party leaders
and spokespersons talk about ‘reforms’, but not ‘privatization.’ This is perhaps
due to the fact that privatization is still not accepted as something inherently
good for the society. The long tradition of state ownership and the public sector
as a model employer in a labor-surplus society are yet to be discarded by the
public. This party, nonetheless, considers liberalization to be in public interest.
(2) Bharatiya Janata Party (BJP): It is the largest party in the ruling National
Democratic Alliance (NDA) at the federal level. It has supported liberalization
and restructuring to be in public/national interest. The latest step, taken by the
NDA and supported by BJP is the constitution of a committee under the Prime
Minister to oversee the progress of ‘reforms.’ The federal government has
announced enhanced allocation of resources under Accelerated Power
Development Programme (APDP) to states if they undertake reforms (The
Finance Minister proposed Rsl5 billion for this purpose in his budget speech in
Feb, 2001). The Conference of the Chief Ministers held on March 3, 2001
(presided by the Prime Minister) decided to speed up the reform process. It even
went on to recognize the need to revise some provisions of the Forest
Conservation Act for the expeditious completion of power projects. The
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
108
following excerpts from the election manifestos of BJP and NDA drive home the
point that they support liberalization and restructuring.
■ BJP as a party welcomes foreign direct investment “in a non-predatory role in
joint-ventures rather than in 100 percent subsidiaries.” It also wants to “prune the
public sector”, while taking care of the interest of the workers. It believes in
retraining and redeployment rather than retrenchment.
• It wants a commercially viable power tariff structure and transparent procedures
to facilitate private investment.1
• The NDA in its pre-election announcement indicated some “achievements.” It
referred to the Electricity Regulatory Commissions Act of 1998, the Electricity
Laws (Amendment) Act of 1998 enabling private investment in transmission and
grids, the setting up of the Central Electricity Regulatory Commission, the
provision of special subsidy of 4 percent on interest rates on loan relating to
renovation and modernization, additional 1 percent subsidy for the
underdeveloped North-East region and the states setting up SERCs.
(3) The Communist Party of India (Marxist)3: This party has all along been totally
against liberalization and privatization. The leaders of this party view it as
leading to “the erosion of economic sovereignty.” They maintain that the path of
1 (BJP Election Manifesto 1998, at http://www.bip.org/manifes/chap4.htm)
2 (NDA Election Manifesto, 1999 at http://www.bjp.org/manifes/manife99.htm)
3 A detailed discussion of this party’s viewpoint can be found at (www.cpim.org)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
109
liberalization and privatization has enormously benefited the big bourgeoisie. The
party sees a pro-imperialist trend in the liberalization process, resulting in
joblessness. So naturally this party considers liberalization and privatization as
antithetical to the public interest. Other left parties like CPI (Communist Party of
India), and Forward Bloc share this view.
These interpretations of “public interest” and the court verdicts have often
influenced the pace and scope of liberalization in the electricity sector in India. Now
a comparison with some other countries’ electricity sectors will perhaps be useful in
analyzing their objectives, examining their strategies and assessing their progress.
We undertake that in the next section.
INTERNATIONAL COMPARISON
As Gilbert, Kahn, and Newbery (1996) point out, the history of the electricity
industry in various countries illustrates the variety of solutions that have been found
to the problem of balancing the interests of the consumers and governments while
still allowing for efficient investment. The solutions available are constrained by
politics, history, endowments, technology, and the state of the economy.
In this section, we will compare and contrast the efforts and results of some
countries in the electricity sector reform. We will compare the power sectors of
Chile, Hungary and Norway. Chile, a developing country, was the first country to
undertake reforms in 1978. It has been a success story. Hungary is a transitional
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
110
economy that took up reform in the 1990s. It has also been hailed as a success.
Norway also started the reform process in the 1990s. It opted for restructuring of the
power sector and not privatization: a case unlike most other cases that have
emphasized privatization.
Chile
Chile’s pioneering experiment has been hailed as a success. As Spiller and Martorell
(1996; pp. 83)1 point out,
The drastic transformation of the electricity sector occurred in Chile, where in less than a
decade the sector moved from being government owned to private provision and ownership,
not only of generation but of transmission and distribution facilities as w ell.... it has also
been able to promote large private investments in all areas, with current investment projects
by private electricity firms amounting to close to US $ 2 billion.
Spiller and Martorell (1996) claim that Chile’s success is based on a drastic
transformation of its regulatory structure and institutions, a transformation that
occurred before privatization took place. As the authors further point out, most of
the countries trying to promote power sector privatization have an extremely ad hoc
regulatory system, which not only generates very large inefficiencies but also lacks
the assurances of fair play that private investors naturally would require. As a
consequence, countries promoting Build Operate and Own or Build Operate and
Transfer generation projects find that potential private investors require substantial
government guarantees (e.g., repayment guarantees for external debt, minimum
purchase requirements, exchange rate convertibility guarantees), which seem to
allocate most of the risks to the government.
1 Most of the material presented in this section on Chilean experiment is taken from this paper
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
I l l
Chile started the restructuring of its electricity sector in 1978. Earlier, the tariffs
were based on a rate of return method. Until 1980, electricity companies were
regulated by a tariff commission composed of representatives from government, the
firms, and the consumers. It used to set maximum annual revenues that were to
provide each company with at least 10 percent return on its assets. Today, regulated
tariffs are determined according to long-run marginal-cost principles, with rates for
large customers (and whole rates as well) determined in the open market. Tariffs are
now set by a mechanism that does not allow short-run government interference with
the determination of rates. Customers now have more choices, with differential
pricing such as peak, off-peak, interrupted and non-interrupted pricing.
As far as the ownership is concerned, in 1978 the electric system was based on
two publicly owned integrated companies, ENDESA and CHILECTRA. There were
as many as 11 power generating companies, 21 electricity distribution companies,
and two integrated companies in 1991, many of those being traded in the Chilean
stock exchange (Philippi, 1991). The privatization process has been complete.
Private companies now provide 100 percent of Chile’s electricity.
This restructuring of the sector was achieved by separating generation and
transmission from local electricity distribution. The distribution networks of
ENDESA and CHILECTRA were broken into several distribution companies, each
with coherent geographic, and economic units, and they were subsequently
privatized. These companies are classified as high density, medium density and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
112
low-density distribution companies, depending on the number of customers they
serve. There has been considerable investment by generating and distributing
companies in transmission and distribution network. Chile has good energy
connections with Argentina and Peru. It has also developed a good transmission and
distribution system within the country
Unlike India, Chile has introduced competition and TPA (third party access) in
the transmission grid. Any privately owned company can pay a transmission tariff to
‘wheel’ electricity. However, because of its large share in the sector, ENDESA, the
largest generating and transmission company, literally controls it.
Chile’s electricity system is predominantly hydroelectric (80 percent). Around 20
percent of generation is thermal. After the severe drought and power supply problem
in 1999, there is an ongoing effort to increase thermal and other capacities.
As Spiller and Martorell (1996) point out, the restructuring process has been quite
successful. Electricity prices are closely related to long-run marginal costs, private
investment is taking place in all areas of activity (including hydroelectric projects),
and the electricity companies are widely held. The fact that the major electricity
companies are widely held among small investors and pension plans may have also
contributed to the stability of the regulatory system.
The CNE (Comision Nacional de Energla), the basic regulatory institution in
Chile, functions under the office of the President. It consists of seven ministers and a
secretary. Its two basic functions are tariff setting (which have to be approved by the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
113
office of the Minister of Economics) and coordination of the several independent
generation, transmission, and distribution companies. The system provides firms
with recourse to the courts if the proposed prices seem to be too much below the
long-run marginal costs.
The sum of transmission costs and energy costs are called node prices, because
they are the prices at which transactions between generating and distributing
companies take place. These prices, computed by the CNE, are adjusted every six
months, using the indexing formulas that depend on fuel costs, equipment costs, dam
levels, and exchange rates etc. Chile has witnessed an appreciable drop in these
prices, and thus, in tariffs after the liberalization.
The CNE simulates a competitive situation and derives the tariffs. The rates for
large customers and the wholesale rates are determined in the open market. Thus, the
CNE does not have much discretion. There is specific legislation relating to the price
setting (with dynamic programming models). Some say that it is ‘set-in-stone’
regulation. This also guarantees substantial independence from the political process.
Hungary
Hungary has been the pioneer among economies in transition in unbundling,
deregulating and privatizing the utility industries and taking the first steps towards
market liberalization. It started liberalization in the 1990s. It has had to grapple with
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
114
some of the legacies left behind by the centrally planned economy in the energy
sector. Pesic and Urge-Vorsatz (2001) point out some of them.
First, energy intensities are among the highest in the world, indicating large-scale
energy waste in the economy. In 1996, Hungary required four times as much energy
to produce a dollar of GDP as Germany or Denmark (OECD/IEA, 1999). Second,
the communist regime had dictated cheap, heavily subsidized electricity tariffs for
consumers, which could not be sustained in a market-based system. Lifting subsidies
required a drastic rise in tariffs. Third, like other transitional economies, Hungary
faced further challenges such as the state-owned monopoly’s resistance to
restructuring, the prospects of unemployment of the engineering personnel in the
new dispensation, and the high debts accumulated over time. Lack of sufficient local
capital for privatization also complicated the process.
In this challenging environment, Hungary started liberalization in the 1990s. By
the end of 1997, 55 percent of its utilities were under private ownership (Pesic and
Urge-Yorsatz; 2001). The main objectives of this liberalization were improvement in
efficiency (allocative, technical and x-efficiency) and improvement of public
finances.
The government of Hungary has formulated an official energy policy that seeks to
develop diverse energy supplies and eliminate dependency on imports from the
former Soviet Union, improve environmental protection, increase energy efficiency,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
115
and attract foreign investment. Hungary’s energy policy is oriented toward achieving
accession to the European Union (Pesic and Urge-Vorsatz; 2001).
In 1992, Hungary restructured the state-owned MVM (Magyar Villamos Muvek
Rt) into a two-tier system. The upper-tier, MVM, still owns the transmission grid and
the dispatching center and manages all trade. The government hopes to privatize it
later. There is no third party access (TPA). MVM has monopoly over transmission
rights. The second tier consists of eight independent generating companies, six
regional distribution companies and one basic maintenance company for the
network. Unlike transmission, the distribution companies are mostly owned by
private European multinational companies.
Although private investment in generation has increased manifold in Hungary
since 1992, the state-owned MVM still generates a major bulk of the electricity. Out
of 6 generation companies, 4 are completely privately owned. The state owns just 25
percent of the share in each of the other two companies. Generation is predominantly
thermal (76 percent). These thermal plants use indigenous coal, contributing heavily
to the level of pollution. The state-owned nuclear power plants are major sources of
production, generating as high as 23.4 percent of the total power.
The Hungary Energy Office (MEH) established in 1994 is the main regulatory
institution in Hungary. It still operates under government control, especially in price-
setting issues. It also issues licenses to producers of gas and electricity. It determines
(but does not set) tariffs for third party access to electricity transmission system. It is
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
116
also the lead organization for consumer protection and investigation of any
complaints. It does not have much discretion in tariff-setting. Specific formulas,
using some sort of price-cap regulation, are employed to arrive at the tariffs.
Hungary has been witnessing increasing electricity prices after restructuring and
privatization. Like India, its tariffs had been set at abnormally low levels for political
reasons (socialist policies). As the Table 3.6 shows, the residential rate is now more
than six times what it was a decade ago. The industrial rates have increased almost
three-fold in a decade. The state has been providing limited subsidies since 1992 as
most of the system is being privatized.
Table 3.6: Average Industrial and Residential Electricity Prices in Hungary,
1991-2000 (US $/kWh)
1 1991 i 1002 1993 1994 1995 1996 1997 1998 1999 2000
Indus , 0.01 8 i 0 .0 IS
illilillllii
0.017 0.017 0.021 0.024 0.033 0.041 0.047 0.049
Resid l (i.O O O 1 0.010
1
0.013 0.015 0.022 0.029 0.041 0.046 0.053 0.058
____________ | I : :| I ______________|_____
(Source: Hungarian Energy Office, 2001)
Norway
Norway introduced competition into the bulk electricity market and created Statnett
Marked (as a subsidiary of the state-owned owner of the transmission system
Statnett) to operate the power pool in 1993 without altering the ownership structure
of the industry. The effect has been to induce substantial trade across former
franchise boundaries with a decrease of the dispersion of prices (Moen, 1994).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
117
The Norwegian example should provide an important test of the relative
importance of creating competitive power markets by restructuring the industry,
compared to privatization. We must note that the Norwegian system allows private
generation to compete with state-owned and municipally-owned systems.
The unique feature of the electricity system in the Scandinavian countries has been
the relative success of voluntary cooperative mechanisms, what Hjalmarsson (1996)
calls “club regulation.” This system has maintained coordinated dispatch among
electricity producers of mixed ownership and widely differing scale. The process is
most advanced in Norway. Sweden and Finland have moved much more slowly
along these lines.
In an international comparison, the price level of electricity (net of taxes) is very
low in Scandinavia, in particular in Norway and Sweden. One obvious reason is the
large share of hydropower in all countries except Denmark. Well-functioning
electricity markets constitute another reason.
The electricity industry in Norway is characterized, to a large extent, by publicly-
owned, dominant firm leadership, self-enforced club-regulation, and yardstick
competition. The Scandinavian countries are highly cooperative societies, and the
development of the electric power sector is mainly the result of negotiations,
cooperation, coordination, and self-enforced regulation among the major market
agents (Hjalmarsson, 1996).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
118
The Norwegian electricity market is characterized by a large number of small
generators and retail distributors. Norway is unique in the sense that more than 99
percent of generation is based on hydropower. Concerning ownership structure,
about 75 percent of the total generation and distribution capacity is owned by public
entities (state, county, and municipal). Most of the distribution companies are
municipal or inter-municipal companies. A characteristic feature of the Norwegian
system has been its power pool, the Norwegian Power Pool (NPP), and the spot
market. NPP greatly facilitated the deregulation of the Norwegian electricity market.
Statnett Marked was created in 1993 and has taken over the operation of the power
pool. There has been a rapid entry of traders and brokers into this market.
Deregulation in Norway started with the Energy Law on January 1, 1991. The first
step was to separate the high-voltage transmission system from generation and
distribution. Statnett is the state-owned-enterprise in charge of the national grid.
Second, third party access (TPA) was introduced. In principle, a household can buy
power directly in the spot market for its own consumption. Third, there is no long
term contract obligation for generation and distribution companies. They can choose
their suppliers as and when they want. Fourth, a set of markets has been established
with power pool. The spot market is cleared every hour. Fifth, a new regulatory
regime has been put in place. Sixth, foreign trade has been reorganized. But there are
also some barriers to entry. The market participants sign a contract and pay
transaction charges (fixed and variable). In the new system, there are three organized
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
119
power markets: the spot market, the regulation market and the contract market. All
Norwegian customers, including households, have been free to choose their supplier
since the Energy Act.
The legal framework has been important for the working of the electricity market
in its creation of barriers to entry and obligation to supply in retail distribution. On
the other hand, the non-legal aspects are of larger importance when it comes to the
functioning of the electricity supply industry as a whole in Scandinavia
(Hjalmarsson, 1996; pp.138-39).
Norway is not the only country to have resorted to just restructuring without going
for privatization. New Zealand restructured and corporatized its state-owned
electricity supply industry under the State-Owned Enterprise Act of 1986. Though it
brought in competition, the industry had still not been privatized after 12 years.
New Zealand is a test case in the viability of state-owned companies competing
with each other, despite sharing a common owner. ECNZ (the Electricity
Corporation of New Zealand) nearly doubled its labor productivity (from 4.5 to 8.5
GWh per employee) between 1987 and 1992. ECNZ reduced unit-operating costs by
13 percent in real terms, increased plant availability from 73 to 91 percent for
thermal plants and from 87 to 95 percent for hydroelectric units. It increased profits
from NZ $262 million to over NZ $400 million while reducing wholesale prices by 8
percent in real terms (Culy, Read and Wright, 1996).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
120
Summary
As we have seen, different countries have resorted to different strategies to achieve
their objectives in the electricity sector. These objectives have ranged from increased
private investment to improved and reliable supply. Almost all the countries that
have undertaken reform have tried to improve efficiency through modernization of
supply systems, better management of consumption, and cultivation of good
regulatory regime.
Chile’s objectives have been to promote efficiency and investment. Hungary has
had to reduce energy intensity, diversify supplies to reduce the dependence on the
former Soviet Union, attract foreign capital, and face the challenge of reducing the
subsidies. Norway’s prime objective in power sector reform has been to create and
consolidate competitive electricity markets. Besides striving for enhanced efficiency,
India has tried to obtain private investment in generation to deal with the yawning
deficit in this sector. Besides other tasks, India faces the challenges of improving
efficiency, and reducing the abnormally high transmission and distribution losses.
As we have already discussed, these countries have adopted different strategies to
achieve their objectives. Most reforming countries like Chile and Hungary have
opted for restructuring, and privatization of generation and distribution. Some like
Norway and New Zealand have tried to bring in restructuring and competition, rather
than privatization. India has started the process of restructuring of its electricity
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
121
sector in some states with varying results. It remains to be seen if restructuring has
helped improve performance in this regard. We examine this next.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
122
CHAPTER 4: DATA, VARIABLES, MODEL FOR ESTIMATION, AND
EMPIRICAL ISSUES
The primary objective of this chapter is to discuss the data and their sources, the
variables, the model and the equations, and the econometric issues involved in the
analysis. The analyses will relate not only to the thermal power plants at the federal
and state levels, but also to the State Electricity Boards (SEBs). The thermal power
plants include the plants owned by the federal government, the SEBs, and the private
companies. We use cross-section and time-series variations in panel data to examine
the impact of restructuring on the performance of the SEBs and the thermal power
plants. We also analyze the effect of ownership on these plants.
Data
We analyze two different panel data sets. The first one (let’s call it PLANTS)
relates to the performance of 72 thermal power plants (federal, state and private)
over 11 years from 1990-2000. This data set was constructed with the help of the
Central Electricity Authority, India. This data set is chosen to analyze the impact of
restructuring on the power plants at a micro level. Secondly, we also examine the
effect of ownership on the performance parameters, distinguishing among federal,
state and private ownership, and their effects on these performance parameters.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
123
The second data set (let’s call it SEB) relates to the state electricity boards (SEBs).
Out of the total 19 SEBs, four (Jammu and Kashmir, Kerala, Himachal Pradesh,
Meghalaya) do not have thermal power generation. Therefore, the data sources do
not have information on plant availability, plant load factor and forced outage for
these states. So we had to exclude those state electricity boards from analysis. It is
also pertinent to note that none of these states has opted for restructuring yet. The
hydroelectric projects continue to be under direct control of the state electricity
boards. It is also important to recall that thermal power generation accounts for
nearly eighty percent of the total generation, making it the predominant contributor
to power generation.
The Delhi Vidyut Board attained the status of an electricity board only in 1997.
Moreover, there were constraints in getting the data. So these five SEBs were
excluded from the analysis. Thus, only 14 SEBs, which have some thermal power
generation, were analyzed over the 13-year period (1988-2000).
The second data set, SEB, was constructed from various sources. The main source
was the Planning Commission of India. The major bulk of the data came from the
Annual Reports on the working of the State Electricity Boards and Electricity
Departments. The Power and Energy Division of the Planning Commission of India
publishes these Annual Reports. The population data (total population and the urban
population projections) came from the Office of the Census and Registrar General of
India. The data came from the Report of the Expert Committee on Population
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
124
Projection: Census of India, 1981, Occasional Papers - No.4 of 1988. (Published by
the Office of the Registrar General of India, New Delhi in the year 1988). The
projections are used, as annual data are not available for the variables. During my
personal meetings with the census officials, they informed me that the projections
were quite close to the actual figures, found during the census.
Table 4.1: Political Variable (POL): 1 if absolute majority, 0 otherwise
STS 88 89 90 9 1 92 93 94 95 96 97 98 99 00
AP 294 1 1 1 1 1 1 1 1 1 1
1 *
1 1
AS 126 1 1 1 1 1 1 1 1 0 0 0 0 0
BIH 324 1 1 0 0 0 0 0 1 1 1 1 1 0
GUJ 182 1 1 0 0 0 0 0 1 1 1 1 1 1
HAR 90 1 1 1 1 1 1 1 1 0 0
g*
0 0
KAR 224 1 1 1 1 1 1 1 1 1 1 1
1*
1
M P 320 1 1 1 1 1 1 1 1 1 1 0 0 0
M A 288 1 1 0 0 0 0 0 0 0 0 0 0 0
ORI 147 1 1 1 1 1 1 1 1
I*
1 1 1 0
PUN 117 1 1 1 1 1 1 1 1 1 1 1 1 1
RAJ 200 1 1 0 0 0 0 0 0 0 0 1 1
11
TN 234 1 1 1 1 1 1 1 1 1 1 1 1 1
UP 425 1 1 1 1 1 0 0 0 0 0 0 0
g*
W B 294 1 1 1 1 1 1 1 1 1 1 1 1 1
(STS: Number ol' total seats in Legislative Assem ily, * SEB was
restructured that year)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
125
In order to construct the POL variable, state-specific and annual data about the
majority of the single largest political party (prevailing at that time) were collected
from the Election Commission of India and the book, India Decides (Butler, Lahiri
and Roy, 1995). The details of the seats obtained by each political party in each of
the Assembly elections were collected, and then the POL variable was constructed. It
is a dummy variable with value 1 when the single largest party in the government at
that year has absolute majority and 0 when the same party is in government as a part
of the coalition. Table 4.1 shows the state-wise and year-wise values of this variable.
Data on the projected demand for electricity, the ratio of the low voltage
consumption to total consumption, and the total generation were constructed from
the 14th , 15th and the 16th Electric Power Surveys, conducted and compiled by the
Central Electricity Authority of India (CEA). The CEA conducts these surveys in
fulfillment of its obligation under section 3(l)(i) and 3(l)(iv) of the Indian Electricity
Supply Act of 1948. The low voltage consumption was constructed by taking into
account the domestic/household consumption and the consumption by agricultural
irrigation.
The data on the Consumer Price Index (CPI) for India for the years from 1988-
2000 were obtained from the International Monetary Fund (IMF)’s International
Financial Statistics, 2001. The CPI was required to transform the price and cost data
into real terms.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
126
For the state of Orissa, where the SEB has been restructured and distribution
privatized, data had to be collected from the Orissa Electricity Regulatory
Commission (OERC) and the Grid Corporation of Orissa Ltd. (GRIDCO).
GRIDCO’s publication, “An Overview of GRIDCO (01.04.96 to 31.03.2001)” and
the final report of Prof. Sidharth Sinha of Indian Institute of Management,
Ahmedabad, titled “Study of Orissa’ Power Sector Reforms, Phase II” (June, 2001)
were also helpful.
Variables
In this section, we describe the variables used in the analysis. First, let us discuss the
variables of the SEB data set. The first set of variables, called the performance
indicators, consists of two types: five operational efficiency indicators and four
financial performance indicators.
The operational efficiency indicators are:
(a) Plant Availability: The total number of hours in the year the plant was available
for generation as a proportion of 8760 hours. This indicates the production-
worthiness of the thermal plants. This is an efficiency parameter, as it shows the
extent to which the managers and the technical experts keep the plant ‘available’
for power production.
(b) Plant Load Factor: Actual energy produced by a plant during a given period as a
percentage of the maximum energy that could have been produced had the plant
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
127
generated at full capacity during the same period. Thus, it is the actual output as a
percent of the total capacity output. It reflects the capital productivity.
(c) Forced Outage: The total number of hours the plant was shut down during the
year due to breakdowns as a percentage of 8760 hours. It is different from
planned outage that is resorted to for maintenance. Forced outage takes place
due to problems in plant management. Thus, it proxies for plant performance.
(d) Number of Employees per MKWh sold: a measure of the manpower required to
sell a million kilowatt hour of electricity. This indicates labor
productivity/efficiency.
(e) Number of Employees per Thousand Consumers: This is yet another indicator of
labor productivity.
The financial performance indicators are:
(a) Unit Cost of Supply: This is the average unit cost of supply of electricity. This is
expressed as Paise/kWh.
(b) Average Tariff: This is the average rate of realization of revenue from the sale of
energy, i.e., revenue from sale divided by units sold. It is expressed as
Paise/kWh. It is not the average price of electricity, but rather the average
amount collected from the sale of one kWh of electricity. This reflects the
efficiency of the SEBs in the collection of revenue.
(c) Sales Revenue as a ratio of Total Cost: It indicates the effectiveness/efficiency of
the SEB in its financial management.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
128
(d) Cross-subsidization: The SEBs have resorted to cross-subsidizing the domestic
and agricultural consumers at the cost of industrial and commercial consumers
ever since they started functioning. The ratio of industrial tariff to the domestic
tariff is used as a measure of the extent of cross-subsidization.
The second set of variables in the SEB data set consists of the control variables.
They can be enumerated as follows:
(a) Transmission and Distribution Loss: It is expressed as a percent of the total
actual generation of power.
(b) Hydroelectric share: It is expressed as a percent of the total generation.
(c) Low Voltage Sale as a ratio of total sale: It indicates the share of low voltage
electricity in the total sale.
(d) CPI: Consumer Price Index (base year 1995=100).
(e) Fuel Cost: It is expressed as Paise/kWh of sale of electricity.
(f) Power Purchase Cost: It is expressed as Paise/kWh of sale of electricity.
(g) Operation and Maintenance Cost: It is expressed as Paise/kWh of sale.
(h) Domestic Tariff: It is in Paise/kWh.
(i) Industrial Tariff: It is in Paise/kWh.
(j) Demand: Total demand for electricity in MKWh as measured/predicted by the
Electric Power Surveys.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
129
(k) Total Population: Total population of the concerned state, measured (through
census)/estimated or projected by the Expert Committee of the Office of the
Registrar General and Census Commissioner of India. As annual census
population figures are not available, we take these projections/estimates into
account.
(1) Urban Population: state-specific annual projections/estimates made by the Expert
Committee (mentioned above).
(m)Urban Ratio: Ratio of urban to total population.
(n) POL: As shown earlier in Table 6, this variable is constructed with the help of
data about the number of seats that the single largest party has in the government
during that year. If the ruling party has absolute majority, then POL takes the
value 1, otherwise 0 (in case of coalitions).
The third set of variables relates to restructuring. Three indices are used. First,
RESTR is a dummy variable that takes the value 1 if the state SEB has been
restructured for that year and 0 otherwise. Second, we also construct a variable called
RESTY, that takes the value equal to the number of years after restructuring, and
taking the value of 0 for the years when the SEB is vertically integrated. Third, we
also use the lagged value of RESTR, named as RELAG (with one-year lag).
RESTR, RESTY and RELAG are used as explanatory variables for the equations
that have operational performance indicators as the dependent variables. But we use
RELAG as an explanatory variable in the equations that have financial indicators as
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
130
the dependent variables. This is because of the fact that it takes quite some time (we
assume one year) for the restructuring to affect/change the financial performance
indicators. Existing contracts need to be re-entered/renegotiated, and forward and
backward linkages re-established in the financial arena before significant changes
can be realized. However, in the case of operational performance indicators like
plant availability and plant load factor, the change in management and government
control can improve these much faster. Better monitoring, less interference from the
government, harder budget constraints and better management have the potential to
bring in results more quickly.
Now, let us discuss the other data set, PLANTS. It consists of observations for 72
thermal power plants over 11 years (1990-2000). Among these are plants of various
ownership patterns: federal, state and private. For eight out of these seventy-two
plants, data beginning in 1990 are not available, since these plants did not start
functioning until after 1990. Some plants, which started functioning as late as 1999
or 2000, were not taken into account for the analysis. This panel data set contains
data for only three operational performance indicators, plant availability, plant load
factor, and forced outage, and none for financial indicators, or other useful control
variables such as size or age of the plants.
We construct three dummy variables to capture the effect of ownership. FS is a
dummy variable that takes the value 1 if the plant is a federally owned plant and 0
otherwise. Similarly, SS is a dummy variable that takes the value 1 if the plant is
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
131
owned by the state (SEB), and 0 otherwise. PS is also a dummy variable that takes
the value of 1 if the plant is privately owned and 0 otherwise. RESTR is a dummy
variable that takes the value 1 if the concerned plant (owned by SEB) is in a state
that has restructured its electricity board (SEB). Otherwise, it takes a value of 0.
Model, Alternative Estimators and Empirical Issues:
Panel data offer us some distinct advantages in estimating behavioral equations,
although they have their own limitations. So it is worthwhile to discuss these
advantages and limitations before discussing the various models/techniques, which
can be used to analyze panel data.
The major benefits of using panel data are in at least three areas. First, it helps in
identifying appropriate economic models and discriminating between competing
economic hypotheses. Second, it also helps in eliminating or reducing estimation
bias, arising mainly out of omitted variable bias, bias induced by the dynamic
stmcture of a model, and simultaneity bias. Third, it also reduces problems of data
multicollinearity (Hsiao, 1986; pp.213).
As we have discussed before, panel data, by providing sequential observations for
a number of individuals, often allow us to distinguish interindividual differences
from intraindividual differences and to construct the proper recursive structure for
studying the issue in question through a before-and-after effect.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
132
The use of panel data also has some limitations. Because panel data usually
contain a large number of observations, it might appear that the problem of
efficiency is not as important a consideration as is consistency. But this is not
necessarily the case. Using the least-square dummy variable approach when the
cross-sectional units have a random nature, will yield implausible results as opposed
to the efficient GLS estimates. Thus, one has to check for specification bias by using
proper specification tests. Secondly, most often the panel contains only a few
observations in one dimension (usually the time dimension) and many observations
in the other dimension. It is extremely important to make efficient use of that part of
the relationship that differs substantially from one to the other. Thirdly, it is only by
taking proper account of selectivity and heterogeneity biases in the panel data that
one can put greater confidence in the results obtained.
A simple way to take account of the heterogeneity across individuals and/or
through time is to use the variable-intercept models. The basic assumption of such
models is that, conditional on the observed explanatory variables, the effects of all
omitted (or excluded) variables are driven by three types of variables: individual
time-invariant, period individual-variant, and individual time-varying variables
(Hsiao, 1986; pp. 25). These models assume that the effects of the numerous omitted
individual time-varying variables are each individually unimportant but are
collectively significant and possess the property of a random variable that is
uncorrelated with (or independent of) all other included and excluded variables.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
133
On the other hand, because the effects of remaining omitted variables either stay
constant through time for a given cross-sectional unit or are the same for all cross-
sectional units at a given point in time, or a combination of both, they can be
absorbed into the intercept term of a regression model as a means to explicitly allow
for the individual and/or time heterogeneity contained in the temporal cross-sectional
data (Hsiao, 1986; pp. 26).
A simple model that can be used to account for the cross-sectional and time-series
variation, can be written as
Yit= pXit + uit
i = 1.,....,N (Cross-sectional Units)
t= l, ,T (Time)
X, a vector of variables, believed to influence Y.
The intercept varies over individuals and time. The classic procedure in this regard is
to assume that the effects of omitted variables are independent of X and are
independently and identically distributed. We can introduce the unit or individual-
and time-effects variables, by rewriting the residual U jt as follows.
Uit=aj + A ,t + sit!
where
Eoii = Ekx = Eslt = 0, Ea\kt = E a,zlt = EX&n =0,
EajOtj = a 2 a if i = /, and 0 otherwise,
EXtX .s = a \ if t = s, and 0 otherwise,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
134
issitSjs = a 2 E if i=j and t= s, and 0 otherwise,
and
Eon x'it = EXt x'it = Esn x'it = 0.
Thus, a; represents the unit-specific effect, and Xt indicates the time-specific effect.
S it represents the effects of all remaining omitted variables. We assume that Sjt can be
characterized by an independently identically distributed random variable with mean
zero and variance a 2s.
2 2
Given the above assumptions, the variance of Yjt conditional on Xit is, a y = a a +
+ c t 2 e. These are called variance components. Therefore, this kind of model is
sometimes referred to as a variance-components model (Hsiao, 1986; pp.33).
Hooch (1962) used such a procedure to estimate parameters of a Cobb-Douglas
production function based on annual data for 63 Minnesota farms from 1946 to 1951.
His work demonstrated that by introducing the unit- and/or time-specific variables
into the specification for panel data, it is possible to reduce or avoid the omitted
variable bias (Hsiao, 1986; pp. 27).
The two alternative specifications of the model differ in their treatment of the
individual effect (Hausman, 1978). The fixed effects model treats the individual or
cross-section effect as a fixed but unknown constant differing across individuals.
This specification is also referred to as least-square dummy variable approach
(Hsiao, 1986). This is because dummy variables are introduced to account for the
effects of those omitted variables that are specific to individual cross-sectional units
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
135
but stay constant over time, as well as the effects that are specific to each time period
but are the same for all cross-sectional units. Given the assumed properties of elt. the
ordinary-least-squares (OLS) estimator is the best linear unbiased estimator (BLUE).
It is also referred to as covariance estimator (CV) or the within-group estimator,
because the variation within each group is utilized in forming this estimator (Hsiao,
1986; pp.32). This specification allows for correlation between a; and x]t. There is no
need to specify the pattern of their correlation.
The alternative specification for the time series-cross section model is known as
the random effects model. As Hausman (1978) points out, instead of treating the
individual effects as fixed constants, this specification treats them as random
variables drawn from an iid distribution, a-t ~ N (0, a a ), and Cov (a;, X jt) = 0.
Whether the a ,’s are treated as fixed or random, the CV of ( 3 is unbiased and
consistent. However, whereas the CV is the BLUE under the assumption that a,j are
fixed constants, the CV is not the BLUE in finite samples when ctj’s are assumed
random. The BLUE in the latter case is the generalized-least-squares (GLS)
estimator (Hsiao, 1986; pp. 34). The GLS estimator in this case is a weighted
average of the between-group and within-group estimators.
The choice of specification seems to rest on two considerations, one logical and
the other statistical (Hausman, 1978; pp. 1262). The logical consideration is whether
the a; can be considered random and drawn from an iid distribution. If this logical
criterion is satisfied, then the random effects specification seems appropriate.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
136
A statistical consideration is then to compare the bias and efficiency of the two
estimators in estimating (3 , the slope coefficients. Hausman (1978) found that using a
fixed-effects-specification produced significantly different results from a random-
effects-specification when estimating a wage equation using a sample of 629 high
school graduates followed over six years by the Michigan income dynamics study.
Hausman (1978) and many prior studies show that the estimators (CV and GLS)
become identical as T becomes large. Since N is usually large relative to T,
differences between the two estimators are an important problem (Hausman, 1978).
Hausman further points out that under the random effects specification Pols is the
asymptotically efficient estimator while the fixed effects estimator Pfe (or Pcv) is
unbiased and consistent but not efficient. If the assumption E (aj | xit) = 0 is violated,
the random effects estimator is biased and inconsistent while the fixed effects
estimator is not affected by this failure of orthogonality.
Hsiao (1986) opines that if an experiment involves hundreds of individuals who
are considered a random sample from some larger population, random effects are
more appropriate. However, if the situation were one of analyzing just a few
individuals, say five or six, and the sole interest lay in just these individuals, then
individual effects would more appropriately be fixed, not random. The situation to
which a model applies and the inferences based on it are the deciding factors in
determining whether we should treat effects as random or fixed. When inferences are
going to be confined to the effects in the model, the effects are more appropriately
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
137
considered fixed. When inferences will be made about a population of effects from
which those in the data are considered to be a random sample, then the effects should
be considered random (pp. 43).
It follows from the above discussion that we should know if the model has been
properly specified. How do we test for misspecification ? These two estimators have
different properties depending on the correlation between time-invariant effects, o tj
and the regressors.
Specifically,
1. If the effects are uncorrelated with the explanatory variables, the random effects
(RE) estimator is consistent and efficient. The fixed effects (FE) estimator is
consistent but not efficient.
2. If the effects are correlated with the explanatory variables, the FE estimator is
consistent and efficient but the RE estimator is inconsistent. (Johnston, pp. 404)
Thus, the issue is whether or not the conditional distribution of a; given x; is equal
to the unconditional distribution of a;. As Hsiao(1986) points out, when cq is
correlated with X j, we shall call it a fixed-effects model, and when r/j is not correlated
with X j, we shall call it a random-effects model.
Therefore, one way to decide whether to use a fixed-effects or random-effects
model is to test for misspecification. The Hausman specification test (1978) basically
asks if the CV and GLS estimates of p are significantly different. If the random
effects specification is correct the two estimates should be near each other.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
138
Hausman (1978) suggests using the statistic
m = q'Var(qylq,
where
V = Per ~ Po l s » Var(q) = Var{/3cv ) - Var(pG L S )
to test the null hypothesis that E (a, | xit) = 0 against the alternative E (a; | xlt) * 0.
This statistic is distributed asymptotically as x2 with K degrees of freedom under the
null hypothesis that the random effects estimator is correct.
When N is fixed and T tends to infinity,
Pcv - Pg l s
In the above situation, the fixed-effects and random-effects specifications become
indistinguishable for all practical purposes. The more typical case in practice is that
N is large relative to T, so that differences between the two estimators or two
approaches become important issues.
In this study, we will use a reduced form model of the following type for both
estimations.
Yit = a* + A .t + p'Zit + y'Rit + Sjt
Where, Y is each performance measure;
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
139
i is the ith state electricity board (SEB)/plant, i =1,..N;
t is time (year), t = 1,.. .T.
a, the state or plant-specific effect;
% , the time-specific effect;
Z, the vector of all control variables, which are assumed to impact
performance, independently of the restructuring;
R, the index of restructuring;
Z and R are ki x 1 and k2 x 1 vectors;
|3, and y are ki xl, and k2 x 1 vectors of constants respectively;
8jti the error term , is independently identically distributed over i and t,
t 2
with mean zero and variance a E .
We use both fixed-effects and random-effects specifications to estimate various
equations involving different performance indicators. We then use the Hausman
specification test to identify the correct specification.
The individual effects may vary on the basis of what is sometimes called the one
way or the two-way effect. If the specification depends only on the cross-sectional
effect, the model is referred to as a model with one-way effects. A specification that
depends on both cross-section and time-series to which the observation belongs, the
model is referred to as a model with two-way effects.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
140
We will use both the cross-sectional effect and the time-effect in most of the
equations where time may have an effect on the performance indicator. For example,
technical expertise and skill may just increase with time as the managers and
employees improve their ‘learning curve’. Similarly, managerial acumen and work
culture may also undergo change over time. Often we come across changes in
technology and in R&D that cannot be controlled in panel regressions without
specific data. Some of these changes may be happening specific to the cross-
sectional units like the State Electricity Board. It may also be happening in all of the
SEBs at a particular point in time. If we do not control for these possible effects, the
estimates of the explanatory variables like restructuring may be biased. Some of
these individual effects will be specific to the cross-sectional units. Some will be
specific to the time. These different effects can be captured only if we use a two-way
effects model in the equations. But before doing that we will have to reason through
whether one-way or two-way effects seem reasonable for the equation involving the
specific performance indicator.
Estimates could be biased if the correct specification is not used. As we have
already discussed, we deal with this by using alternate specifications, both fixed and
random effects, and also one-way and two-way effects as the particular case merits.
We decide about the correct specification by using Hausman’s specification test.
Panel data helps us in getting the correct estimates of these individual effects.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
141
Cross-section analysis can allow for no individual constant but must assume, as
does random effect analysis, that the right hand side variables are orthogonal to the
residual: if the random effect specification is rejected serious doubt may be cast
therefore on much similar cross-section analysis (Hausman, 1978; pp. 1267). Panel
data analysis with different specifications and tests help us to identify better
estimates.
A frequently observed source of bias in both cross-section and panel data is what
is known as selectivity bias: that the sample may not be randomly drawn from the
population. Sometimes, there is truncation in the data. For example, if the
participants in an experiment are restricted to have earnings less than a particular
level, all individuals with earning more than that will be eliminated from the
experiment. The OLS estimates of the effects of the exogenous variables on earning
will be biased due to truncation in the data. In the present case, however, we will
take into account all the states/SEBs, which have thermal power plants. This is
because most of our performance indicators relate to the thermal power plants’
functioning (like plant availability, plant load factor, forced outage etc.).
Another source of bias is endogeneity of the explanatory variable. As pointed out
by Cornwell and Trumbull (1994), there are two possible sources of endogeneity:
unobserved heterogeneity and conventional simultaneity. They point out in their
study of crime with county-level panel data that cross-section econometric
techniques do not control for unobserved heterogeneity whereas panel data allows to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
142
control for unobservable county-specific characteristics that may be correlated with
the criminal justice variables (the right-hand side variables) in the model. “In
general, failure to condition on these unobservables will result in inconsistent
estimate of the coefficients of these variables” (Cornwell and Trumbull, 1994;
pp.361). With panel data, we can account for unobservable unit characteristics by
conditioning on unit effects in estimation. The individual effects will account for
these otherwise unobserved effects.
A second possible source of endogeneity is the conventional simultaneity. We
assume that the right hand side variables are exogenous and influence the dependent
variables. But the assumed exogenous variables may not be exogenous. They may be
dependent on the left-hand side variables. For example, it is quite natural to doubt
the exogenous nature of the index of restructuring. Bad performance of the states in
the electricity sector might actually be influencing the restructuring. Thus,
restructuring is quite likely to appear as endogenous. But a closer look at this index
will show us that restructuring in this particular context is indeed an exogenous
variable.
An endogenous variable is influenced/determined by the exogenous variables. Its
value would change in accordance with the changes in the exogenous variables.
However, in this case, the status of restructuring does not change once it takes place.
None of the states has reverted back to the earlier state-owned vertically integrated
SEB after undertaking restructuring. Thus, it is not as purely endogenous as the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
143
performance of the thermal power plants, that is determined by various factors, and
keeps changing in accordance with the exogenous variables. That is why the index of
restructuring (RESTR), a dummy variable, never reverts back to 0 after it attains a
value of 1. So one cannot argue that performance is affecting restructuring in a
decisive way. Had that been the case, we would have seen some reversals to the
earlier state.
We have discussed the broad econometric model that we employ to estimate the
parameters. We have also discussed various specifications that we would use. As we
have various performance indicators to analyze, we would use different right hand
side variables to do these analyses. These specific equations deserve discussion.
Equations relating to the plant performance indicators
The operational performance indicators can be divided into two groups. The first
group proxies for plant performance while the second group indicates labor
performance. Let us discuss the equations relating to the first group now.
It is hypothesized that demand (for electricity as a whole) will affect the plant
performance. An increase in demand will put pressure on the plants to increase plant
availability, plant load factor, and reduce forced outage. The management will be
under a lot of pressure (either from the government when it is vertically integrated,
or from the regulatory commission, when it is restructured) to meet the increasing
demand. Therefore, if we do not control for this possible effect of demand on these
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
144
indicators, the estimates may be biased. Thus, we use demand as a right-hand side
control variable in regressions involving plant availability, plant load factor and
forced outage. The operation and maintenance cost may also affect these three
performance indicators. Higher expenditure in operation and maintenance may affect
plant availability and plant load factor positively and forced outage negatively.
Therefore, we use this as a control variable.
Equations relating to labor performance indicators
The second group relates to labor performance. It consists of two indicators:
number of employees per thousand consumers and number of employees per million
kWh electricity sold. These are the dependent variables in panel regressions. We use
the different indices of restructuring and the variable, POL, as right hand side
variables.
We also use some control variables, believed to be influencing the dependent
variables. First, we use the share of low voltage electricity in the total sale. As
already discussed, a large share of the low voltage supply goes to the agricultural
sector in the form of electricity for running irrigation pumpsets. Most of this supply
is either free or billed as a fixed tariff, depending upon the capacities of the
pumpsets. It does not require much manpower to monitor, check the meters and do
the billing. So this factor might be affecting the labor performance. We hypothesize
that this variable will have a negative impact on the dependent variables.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
145
Second, we use the ratio of urban population to the total population as another
control variable. A higher urban ratio will imply more household, commercial and
industrial consumers, which requires more monitoring, billing and collecting effort.
Thus, we expect this variable to be positively related with the dependent variables.
We also control for the hydroelectric share in the total generation. The hydroelectric
projects may affect the dependent variables due to its different labor requirement.
Equations relating to the financial performance indicators
Apart from the indices of restructuring, and POL, which are right-hand side
variables, we use the share of low voltage sales in total sales, and the share of the
urban population in total population as control variables in the equations relating to
three of the four financial performance indicators. These three indicators are average
tariff, sales revenue as a ratio of total cost, and the extent of cross-subsidization.
Restructuring is believed to influence the average tariff collection and the sales
revenue as a ratio of cost positively, and the extent of cross-subsidization negatively
due to reasons discussed in chapter one. The variable, POL, is assumed to have the
opposite effects.
A higher share of the low voltage supply (to agricultural consumers, and
household consumers) is likely to affect the average tariff collection and sales
revenue as a ratio of cost adversely, and the extent of cross-subsidization positively.
It is difficult to predict the effects of the urban share in the total population on tariff
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
146
collection, sales revenue as a ratio of cost, and the extent of cross-subsidization.
They will depend on the composition of industrial and commercial supply vis-a-vis
the domestic supply in the urban area.
The fourth financial performance indicator is the unit cost of supply, i.e., the
average cost of supplying one unit of electricity. In many countries like Chile,
Argentina, and UK, the average cost of supply has come down after restructuring and
privatization. It took quite some time for the supply cost to come down, though the
prices came down relatively sooner. As we have discussed, it takes quite some time
for restructuring to reduce these costs. Therefore, we focus on the lagged index of
restructuring, though we also use other indices. Some critics claim that an absolute
political majority of a single-party government may tend to interfere more in the
management of the electricity boards. This may influence costs. Therefore, we use
POL as a right hand side variable.
We use quite a few control variables, believed to influence costs. The first of these
variables is demand. It is believed to increase costs, especially when the domestic
and agricultural demand has gone up much faster than the industrial and commercial
demand. The second variable is the share of hydroelectric generation. It is well
documented that the hydroelectric share tends to reduce the unit cost of electricity.
This is because, though the initial capital investments in hydroelectric projects are
very high, the recurring expenses are much lower than those in thermal power plants.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
147
The third control variable is the percent of transmission and distribution (T&D)
losses in total generation. As the unit cost of supply increases with increasing T&D
losses, it is obvious to predict a positive relationship between these two variables.
The fourth control variable is the urban population ratio. We use it because higher
urbanization implies higher household and other consumers, and may influence costs
through this consumers-structure. The fifth control variable is the share of low
voltage supply. As we have discussed before, a higher share of the low voltage sale
indicates relatively higher transmission and distribution losses, implying increased
costs of supply. The sixth and seventh control variables are the major components of
electricity supply costs: fuel cost and power purchase costs. Most of the SEBs
actually buy some amount of electricity from the federally-owned power plants and
hydro-electric projects. It is natural that an increase in the purchase cost will increase
the cost of supply. Similarly, an increase in the fuel cost also increases the cost of
supply. Since the SEBs depend upon the federally owned and operated corporations
for the purchase of fuel, they usually have no control over those costs. In order to get
an unbiased estimate, we have to control for all these factors and their possible
effects on cost of supply.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
148
Summary
In this chapter, we discussed the panel data sets, the variables, models and
specifications that we use to estimate the effects of restructuring on operational and
financial performance of India’s electricity sector. We use various performance
indicators, three indices of restructuring and several control variables. We also
construct the variable, POL. It indicates the type of political majority in the
legislative assembly. We use it to identify its possible effects on the performance
indicators.
In all these panel regressions, we use both fixed effects and random effects
specifications. Then we use Hausman test to identify the correct specification. As far
as the cross-sectional effect and the time series effects are concerned, we use the
two-way effect model in all those equations where time is believed to have an effect
on the dependent variable. We examine the results of these analyses in the next
chapter.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
149
CHAPTER 5: ANALYSIS OF EMPIRICAL RESULTS REGARDING THE
EFFECTS AND POSSIBLE DETERMINANTS OF RESTRUCTURING
In this chapter, we discuss the results of our econometric analyses, which involve
two different panel data sets: SEB and PLANTS. We also analyze data on several
other variables like agricultural consumption of electricity, rates of return of the state
electricity boards, per capita net state domestic product, the tariff structures, and their
possible relationships with restructuring.
DESCRIPTIVE STATISTICS
The descriptive statistics of the two data sets are given in Appendix-B. As far as SEB
is concerned, these statistics cover all the electricity boards for all the years between
1988 and 2000. There is wide variation in all the performance indicators. For
example, the minimum plant availability for Andhra Pradesh (AP) and Punjab are
77.4 percent and 77.8 percent respectively, whereas the maximum plant availability
for Bihar and Assam are 59.12 percent and 63.3 percent respectively. The minimum
plant load factor (PLF) for Andhra Pradesh is 62 percent, whereas the maximum
PLF for Assam and Bihar are 28 percent and 37.1 percent, respectively.
Interestingly, the highest maximums in PLF are found in AP, Orissa, Karnataka and
Rajasthan, all states that have already undertaken restructuring. As far as forced
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
150
outage is concerned, restructured SEBs like AP and Karnataka fare much better than
other states.
Now let us look at the labor efficiency indicators. As regards the number of
employees per MKWH electricity sold (EMKWH), we again come across wide
variation. States like AP, Gujarat, and Maharashtra perform much better than other
states like Assam and West Bengal. As regards the number of employees per
thousand consumers (ETHCON), the difference is even more conspicuous. States
like AP and Gujarat perform much better than states like Assam, Bihar, Orissa, and
West Bengal. Interestingly, all these SEBs are in the eastern region of the country.
The SEBs also vary in their financial performance. States like Maharashtra and
Orissa fare better than Assam and Haryana in their financial management, indicated
by sales revenue as a ratio of cost (SREV). The data show that Assam has never been
able to collect more than 65 percent of its cost incurred for supply of electricity.
As far as the real cost of supply per unit of electricity (UCOST) is concerned, we
come across wide variation. The mean real unit cost of supply for AP has been 1.64
rupees, whereas it has been as high as 3.73 rupees for Assam.
The extent of cross-subsidization (CROSS), which is a ratio of industrial tariff and
domestic tariff, differs widely across states as well as time. The minimum has been
0.90, whereas the maximum has been as high as 5.068. Thus, the industrial tariff in
Madhya Pradesh (MP) in a particular year has been as high as five times the
domestic tariff.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
151
The share of low voltage electricity in the total supply (LVTOT) over the years
varies from a minimum of 13 percent to a maximum of 70 percent. This share has
increased over time. As more villages were electrified, more household consumers
and agricultural consumers got power connections, thereby increasing this share. But
the high voltage share has decreased over time. Due to increasingly high industrial
tariffs, industries have increasingly resorted to captive generation.
The hydroelectric share in generation (HYDEL) varies widely. There are four
states, which have no thermal power plants. Some states like Assam have no
hydroelectric generation. This share varies between 0 percent to 78.8 percent in the
14 SEBs. As we have already discussed in chapter 4, four SEBs depend upon 100
percent hydroelectric generation. We also come across a wide variation in the ratio
of urban population (URRATIO) from a minimum of 11 percent to a maximum of
43.5 percent.
RESULTS OF THE T-TESTS OF THE MEANS
We use t-tests to compare the means of plant availability, plant load factor and
forced outage between federal and state-owned plants, federal and privately owned
plants, and state and privately owned plants. We use F-tests to test hypotheses of
equal group variances. The results of the t-tests are in Appendix-B. We find
statistically significant difference in mean plant availability between federal and state
plants (p-value: <0.0001), private and state plants (p-value: <0.0001), and also
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
152
between federal and private plants (p-value: 0.0009). In the case of mean plant load
factor, there is significant difference between federal and state plants (p-value:
<0.0001), and between private and state plants (p-value: <0.0001). But there is no
significant difference in mean plant load factor between private and federal plants (p-
value: 0.3823). As far as mean forced outage is concerned, we find significant
difference between private and state plants (p-value: <0.0001), and private and
federal plants (p-value: <0.0001). The difference between federal and state plants is
marginally significant (p-value: 0.0563).
As expected, the mean value of plant availability in 2000 for all the plants is
higher than that in 1990. The trend is the same in PLF. However, there is not much
difference in the mean values of forced outage for the years 1990 and 2000, though
the relative variability is much more in 2000, compared to 1990. The coefficient of
variation for forced outage for 2000 is 1.10, whereas it is 0.824 for the year 1990.
RESULTS OF PANEL REGRESSIONS
Now let us discuss the results of the panel regressions involving two data sets,
SEB and PLANTS. The t-values, which are significant at the 5 percent level of
significance, will be treated as significant. Those t-values, which are significant even
at the 1 percent level of significance, will be termed highly significant. Similarly,
those t-values, which are significant at 10 percent, will be considered marginally
significant. The results of the regressions are in Appendix-A.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
153
(1) SEB Data set
Let us first discuss the results of the regressions involving the data set, SEB. As
we have already discussed in the previous chapter, we categorize the dependent
variables into two types: operational performance indicators, and financial
performance indicators.
Plant Performance Indicators
We use fixed two-way and random two-way specifications to study the effect of
restructuring on plant performance indicators such as plant availability, plant load
factor and forced outage. The Hausman test justifies the fixed effects specification in
all the regressions. The SEBs show fixed individual effects, as expected.
We use three different indices of restructuring such as RESTR, RESTY and
RELAG. We use POL, demand, and operation & maintenance cost (OMC) as control
variables. The regression outputs show that even after controlling for possible time
effects, restructuring has highly significant and positive effects on plant availability,
plant load factor, and negative effects on forced outage. We find this result in all the
three versions of restructuring. The use of demand conditions as a control variable
does not change the results. POL is insignificant. Demand is positive in equations
relating to plant availability and plant load factor, and negative and insignificant in
the forced outage equation. The real operation and maintenance cost (OMC) is
insignificant in all the regressions.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
154
Labor Efficiency Indicators
The results support fixed two-way effects specification in equations relating to the
two labor efficiency indicators, namely the number of employees per million kWh
electricity sold (EMKWH), and the number of employees per thousand consumers
(ETHCON). RESTR and RESTY are negative and marginally significant in
equations relating to ETHCON, but insignificant in EMKWH.
Interestingly, the POL variable is positive and significant in all the specifications,
vindicating the claim that an absolute majority of a party in the government without
proper democratic institutions might encourage interference in the public sector
organizations, especially in areas of employment.
The share of low-voltage electricity in the total supply (LVTOT) is negative and
significant in all specifications. An increase in this share usually implies an increase
in household and agricultural consumers. The latter are usually not monitored, as the
supply of electricity to them is not metered in most cases. Less monitoring of
agricultural consumption involves less manpower requirement. The second control
variable, the ratio of the urban population (URRATIO), is positive and highly
significant. An increase in the urban ratio is associated with higher power
connections, especially household consumers, which require more monitoring,
billing, and tariff collection effort. This, in turn, requires more manpower. The third
control variable, the share of hydroelectric generation (HYDEL) in the total output,
is not significant. When we drop that variable, the results do not change.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
155
Financial Performance Indicators
We use four financial performance indicators, as discussed earlier. As far as average
tariff collection (AVT) is concerned, all the three indices of restructuring are found
to be positive and highly significant with p-values ranging from < 0.0001 to 0.0003.
The variable, POL, is negative and marginally significant in only one specification,
and insignificant in other specifications. The coefficients of all other control
variables are insignificant. The fixed two-way effects specification is supported by
the Hausman test. The logarithmic transformation of AVT does not change the
results.
The coefficients of RESTY and RELAG are positive and significant in equations
relating to sales revenue as a ratio of cost (SREV). The coefficient of POL is
negative and significant in all specifications. The other control variables, LVTOT
and URRATIO, also evince similar effects.
As far as the extent of cross-subsidization (CROSS) is concerned, the Hausman
test supports random one-way effect specification, with Pr>m ranging from 0.2491 to
0.3004. All the three indices of restructuring are negative and significant, supporting
the claim that the independent regulatory commissions would try to rationalize the
tariffs. The coefficient of variable, POL, is insignificant. The control variables,
LVTOT and URRATIO, are positive and significant. A higher share of low voltage
supply (e.g., supply to agricultural consumers) is associated with and compensated
by higher cross-subsidy. Similarly, a higher urban ratio usually entails a larger
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
156
number of household consumers, implying a higher supply of subsidized electricity
to these consumers. This is compensated by higher industrial tariff, indicating a
positive association with the extent of cross-subsidization.
The unit cost of supply, UCOST, is an important indicator of financial
performance. Obviously, cost of supply is crucial in any discussion of the effect of
reform. The opponents of reform sometimes opine that restructuring would entail an
increase in not only tariffs but also costs. So it is worthwhile to study this using
actual data.
We use the model with two-way effects as time could bring in important changes
in the cost structure. The random effect specification is supported by the Hausman
test as the correct specification in all regressions, the Pr > m ranging from 0.5311 to
0.8313.
The results also show that the index of restructuring, RELAG is negative, but
insignificant. This result remains the same when we use RESTY as the index of
restructuring. In fact, none of the other results changes when we use RESTY.
However, the coefficient of POL is positive and highly significant in all the
specifications, indicating a significantly positive association between the unit cost of
supply and absolute majority of the political party in government (the p-values range
from 0.0086 to 0.0092).
As expected, variables like FCOST and POWCOST have positive and highly
significant effects on unit cost of supply in all the specifications. An increase in fuel
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
157
cost and power purchase cost will obviously influence unit cost of supply positively.
Similarly, the percent of transmission and distribution (T&D) losses, TD, is highly
significant and positive in all the specifications. Higher T&D losses increase the cost
of supply. The hydroelectric share, HYDEL, is insignificant. The share of the low
voltage supply, LVTOT, is insignificant, though one expects that it would be positive
and significant. As expected, the urban ratio, URRATIO, is positive and significant
in the specification that does not control for demand conditions. When demand is
used as a control variable, all the results remain the same except that of URRATIO.
The demand effect of the urban population is perhaps being captured by the demand
variable. The coefficient of the control variable, demand, is positive and significant,
as expected.
The use of the logarithmic transformation of the dependent variable (LNUCOST)
does not change the results except that of HYDEL, whose coefficient now becomes
negative and significant, as expected. Now URRATIO is significant and positive,
even when demand is used.
(2) PLANTS Data set
Now let us discuss the results involving the other panel data set, PLANTS. As we
have already discussed, the situation to which a model applies and the inferences
based on it are the deciding factors in determining whether we should treat the
effects as random or fixed. In the case of the present sample, PLANTS, we would
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
158
expect a random effect specification to be the correct one because, the experiment
involves a fairly large number of plants, which are considered a random sample from
some larger population. The inferences will be made about the said population, and
the effects should be considered random. However, we will also use the Hausman
test to check the correct specification.
Unlike the previous data set, SEB, this data set does not have data on the control
variables. In this case, however, we do have the information on the ownership of the
plants and thereby study the possible effects of ownership on plants’ operational
performance. We once again study the effects of restructuring on three performance
indicators: plant availability, plant load factor, and forced outage. We use two
different indices of restructuring, RESTR and RESTY in this regard. We also control
for the state or location-specific effects. We use two-way effects in all the models, as
it is logical to assume that there could be significant time effects in the operational
performance.
The results show that restructuring is significantly and positively associated with
plant availability and plant load factor. Interestingly, the state ownership, indicated
by SS, is significant and negative in all specifications. The results of restructuring
and state ownership do not change when we control for location (states). The random
effects specification is supported by the Hausman test in all the regressions.
The federal ownership dummy variable, FS, has a positive, but insignificant effect
on plant availability, implying that after controlling for location, the effect of federal
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
159
ownership is not significantly different from the other forms of ownership. The
private ownership dummy variable, PS, also has a positive, but insignificant effect
when we control for location-specific effects. Thus, when we control for location, the
effects of private and federal ownership are not significantly different. But the
negative effects of state ownership and positive effects of restructuring are
significant in all specifications.
As far as the location-specific effect is concerned, we took Bihar as the baseline
state. The results show that in matters of plant availability and plant load factor,
Assam, Haryana, Orissa and Uttar Pradesh are not significantly different from Bihar
at 5 percent level of significance. Other locations/states are significantly and
positively different.
The federal ownership dummy variable, FS, has positive and significant effects on
plant load factor in all the specifications. Even after controlling for location, the
effect of federal ownership is significantly different from the other forms of
ownership. However, after controlling for location, the effect of private ownership is
not significantly different from the state ownership.
As far as forced outage is concerned, we use two indices of restructuring, RESTR
and RESTY. We find both indices to be significantly and negatively associated with
forced outage. However, the state ownership, indicated by SS, has significant and
positive effect on forced outage in all specifications. The effects of restructuring and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
160
state ownership do not change when we control for location (states). The random
effects specification is supported by the Hausman test.
The federal ownership dummy variable, FS, is negative, but insignificant in all the
specifications. It indicates that as far as forced outage is concerned, the federal plants
are not significantly different from the other plants. This does not change when we
control for location effects. As far as private ownership is concerned, PS is negative
and significant when we do not control for locations. However, when we control for
only the locations, PS is negative, but insignificant (p-value: 0.1122).
As far as location-specific effect is concerned, the results show that Assam,
Haryana, Orissa, Uttar Pradesh, and Delhi are not significantly different from Bihar
when we control for state ownership, SS, and restructuring. States like Karnataka,
Rajasthan and West Bengal fare marginally better than Bihar. Other states such as
Andhra Pradesh, Gujurat, Madhya Pradesh, Maharashtra, Punjab, and Tamil Nadu
fare significantly better. Interestingly, indices of restructuring are negative and
significant in all specifications.
POSSIBLE DETERMINANTS OF RESTRUCTURING
The foregoing discussion underscores the effects of restructuring on performance
indicators. As discussed earlier, we treat restructuring as an exogenous variable. In
this section, we analyze data on other variables that could potentially affect
restructuring.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
161
When Orissa decided to restructure its electricity sector, some analysts like Raj an
(2000) argued that the state’s low agricultural consumption of electricity helped
restructuring, and that the absence of a strong farm lobby facilitated the process.
Table 5.1: Share of Agricultural Consumption in Total Sale (%) (1996-97)
Andhra Pradesh 37.3 Karnataka 45.5 Rajasthan 33.7
Assam 2.3 Madhya Pradesh 36.2 Tamil Nadu 26.0
Bihar 18.5 Maharashtra 32.5 Uttar Pradesh 36.2
Gujarat 37.7 Orissa 3.3 West Bengal 14.2
Haryana 45.1 Punjab 35.7 All India 30.8
Source: Annual Report on the State Electricity Boards, Planning Commission, 2001
(Underlined states have restructured their electricity sectors)
One can see from the Table-5.1 that the share of agricultural consumption is very
small in Orissa. However, this share is quite high in most of the states that have
opted for restructuring. For example, Karnataka and Haryana have the largest shares
among all states. Even Andhra Pradesh, Uttar Pradesh and Rajasthan have higher
shares than the national average of 30.8. Thus, higher shares of agricultural
consumption have not prevented these states to opt for restructuring. On the other
hand, Assam, West Bengal and Bihar, states with relatively smaller share, are yet to
undertake restructuring.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
162
Numerous studies1 have shown the high correlation between income and
electricity demand. Higher income is likely to induce higher demand. But could
income be a determinant of restructuring? The cases in India do not seem to support
this.
Table 5.2: Per Capita Net Domestic Product at Current Prices (in Rupees) for
1996-97
Andhra Pradesh 11224 Karnataka 11772 Rajasthan 10171
Assam 7394 Madhya Pradesh 8689 Tamil Nadu 13382
Bihar 4965 Maharashtra 17825 Uttar Pradesh 7743
Gujarat 16287 Orissa 6401 West Bengal 9886
Haryana 16454 Punjab 17447
Source: Economic Survey, 2001-02 by the Ministry of Finance, Government of India
(Underlined states have restructured their electricity sectors)
Table 5.2 shows that per capita income does not seem to have a determining effect
on restructuring. Among the restructured states, Haryana has the highest per capita
income. Interestingly, some states, which have similar or higher income (like Punjab,
Maharashtra, and Gujarat), have not opted for restructuring. Even among the low-
income states, Bihar, Assam and Madhya Pradesh have not yet gone the Orissa way.
1 Our study supports this. The correlation is found to be 0.74 for the available data on income and
demand.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
163
Reduction of prices and costs has often been one of the primary objectives of
restructuring. High costs of supply and high tariffs are likely to compel countries to
opt for alternate industry structures and ownership. However, as discussed earlier,
India’s electricity prices have been kept unusually low for various reasons. The
tariffs have been kept lower than the average costs of supply. The finance minister of
the government of India in his 2001-02 budget speech observed,
Public resources have been invested in the public sector over the last 50 years for the
provision of infrastructure services in the country. One consequence of this has been that
user charges have inevitably become politically determined. Over time non-merit subsidies
inherent in such low user charges have mounted to over 10 percent of GDP.
He further went on to argue that “theft of electricity must be stopped and economic
tariffs levied.”
Thus, unlike many other countries, one of the objectives of restructuring in India
was not the reduction of tariffs, but the restoration of financial viability of the state
electricity boards through augmented efficiency and collection of “economic tariffs.”
It is, therefore, not surprising that the tariffs in those states have been increasing
since they undertook restructuring of their electricity boards. The data show positive
correlation between restructuring and tariffs. The correlation coefficient of
restructuring and domestic tariff is found to be 0.46893 (p-value: <0.0001).
Similarly, the correlation coefficient of restructuring and industrial tariff is found to
be 0.40371 (p-value: <0.0001).
As far as the costs of supply are concerned, the data do not seem to support the
argument that high cost states are more likely to opt for restructuring than the low
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
164
cost ones. The correlation coefficient of restructuring and unit cost of supply is found
to be 0.099 (p-value: 0.1834).
Table 5.3: Unit Cost of Supply of Electricity in Paise/kWh for 1997
Andhra Pradesh 239.69 Karnataka 179.37 Raiasthan 258.6
Assam 457.17 Madhya Pradesh 231.46 Tamil Nadu 208.12
Bihar 316.04 Maharashtra 215.58 Uttar Pradesh 253.31
Gujarat 247.62 Orissa 351.73 West Bengal 252.48
Haryana 293.4 Punjab 217.18 All India 239.73
Source: Annual Reports on the State ElectricilL y Boards, Planning Commission, 200
(Underlined states have restructured their electricity sectors)
We examine the costs of supply of electricity for the year 1997. Most of the
reforming states (except Orissa, which had already restructured its electricity board)
decided to opt for restructuring around that time. Table-5.3 shows that high cost
states like Assam, Bihar and West Bengal have not yet opted for it, whereas low cost
states like Karnataka and average cost states like Andhra Pradesh have resorted to
restructuring.
Another variable that merits discussion is the rate of return of the electricity board.
It is quite logical to hold that states with highly negative rates of return are likely to
opt for restructuring and/or privatization, as it becomes increasingly difficult for the
governments to continue managing these organizations. The data show a negative
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
165
and significant correlation between restructuring and rates of return (-0.3276 with p-
value < 0.0001).
Table 5.4; Rate of return on fixed capital (without subsidy) for 1997 ( in %)
Andhra Pradesh -33.95 Karnataka -16.5 Raiasthan -22.88
Assam -42.53 Madhya Pradesh -25.9 Tamil Nadu -4.94
Bihar -32.57 Maharashtra -0.14 Uttar Pradesh -26.31
Gujarat -29.79 Orissa -20.28 West Bengal -45.34
Haryana -47.79 Punjab -31.99 All India -22.94
Source: Annual Reports on the State Electricilty Boards, Planning Commission, 200
(Underlined states have restructured their electricity sectors)
The above table shows that in 1997, three (Andhra Pradesh, Haryana and Uttar
Pradesh) of the six states that have restructured so far, had lower rates of return than
the all India average of -22.94. Rajasthan was at par with the all India average,
whereas Karnataka and Orissa fared better. Interestingly, such states as Assam, West
Bengal and Bihar with very low rates of return have not yet opted for reform.
Political factors and support have often been viewed as pivotally important in any
reform program. As discussed earlier, political parties in India have broached the
subject of liberalization differently. The communist parties and their allies have
consistently opposed restructuring, privatization, and participation of multi-national
corporations. It is, therefore, not surprising to find that West Bengal, a state under
communist regime since 1977, has not yet opted to restructure its electricity sector.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
166
As noted earlier, the state has a relatively low share of agricultural consumption, a
considerably high cost of supply and one of the lowest rates of return on the capital
employed in its electricity sector. Significantly, these factors, sometimes believed to
goad governments to opt for reforms, do not seem to have worked in the face of
ideological conviction and/or political opposition.
Many other non-communist political parties also are yet to support liberalization.
We come across this in states like Assam and Bihar, whose electricity boards have
lagged behind other boards in most respects.
Authors like Dubash and Raj an (2001) and Raj an (2000) have argued that strong
political leadership has been primarily responsible for the grounding of reforms. This
holds in at least three of the six cases. Orissa had Biju Patnaik and JB Patnaik.
Haryana had Bansilal. Andhra Pradesh has had Chandrababu Naidu. These political
leaders were able to facilitate the reform process through their strong leadership. In
the other three cases, the political parties in government supported restructuring
when the states opted for reforms in the electricity sector.1 Those political parties had
supported reforms in their national election manifestos as discussed earlier in chapter
three.
It is not just the political leaders or political parties that seem to have influenced the
reform program. The type of majority seems to have had a role to play in this
regard. As noted in chapter four, political parties had absolute majorities in
1 Janata Dal in Karnataka, Congress in Rajasthan and Bharatiya Janata Party in Uttar Pradesh.
2 Absolute majority or Coalition Government.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
167
government in four out of the six cases of restructuring when those states opted for
reform by formally enacting the reform legislation. Only Haryana and Uttar Pradesh
had coalition governments when they approved the electricity reform laws.
Summary
As far as the different performance indicators, control variables, ownership
patterns and the timing of restructuring are concerned, the data show wide variation
among the states. Restructuring started in these states at different times, and took
shape with different nuances and policy subtleties. The results show that
restructuring has had a positive effect on most performance indicators. As far as
plant performance is concerned, this finding holds ground even after controlling for
ownership patterns.
We analyzed some variables, likely to be potential determinants of restructuring.
But most of those variables, as discussed above, do not seem to have a deciding
effect on restructuring, although those factors, in combination or otherwise, might
have had some effects on the decision-making.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
168
CHAPTER 6: CONCLUSION, POLICY IMPLICATIONS, AND
SUGGESTIONS FOR FUTURE RESEARCH
In the final chapter, we use the results to draw conclusions regarding the effect of
restructuring and ownership in the electricity sector of India. We also identify some
issues that could be relevant for future research.
CONCLUSION AND POLICY IMPLICATIONS
In conclusion, we briefly discuss the effects of restructuring, ownership and the
absolute majority of a political party in government on various indicators of plant
performance in India, and their possible policy implications. We cite the conclusions
and findings of other studies that support our findings.
Restructuring of the vertically integrated state electricity boards seems to initiate
beneficial changes in such plant performance indicators as plant availability, plant
load factor and forced outage. It creates independent action-centers with
accountability, and brings competition into erstwhile vertically integrated structures.
The accounts of the restructured companies become separate, and reflect their own
performance. In the face of harder budget constraints arising mainly out of no
government subsidy, the restructured companies tend to work harder to face the
market. As far as plant performance is concerned, restructuring has significant
positive effects, even after controlling for ownership. This obviously has an
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
169
important policy implication, pointing to the need of restructuring of the electricity
boards to improve plant performance.
Second, as far as labor efficiency indicators are concerned, we have found mixed
results. Restructuring is negatively and significantly associated with the number of
employees per thousand consumers served, implying an increase in labor efficiency
in this regard. However, it is not significantly associated with another indicator, the
number of employees per million kWh electricity sold. This may be due to the fact
that the restructured companies promised to give full protection to the jobs of
existing employees in order to avert acute employee resistance to any restructuring
and/or privatization. This controversial feature has perhaps been a practical necessity
in the Indian context.
The Andhra Pradesh Electricity Reforms Act stipulates that the terms and
conditions (for the employees) shall not be less favorable than before
restructuring/privatization. It provides for a tripartite agreement between the
government, the APSEB and the employees. A question arises about the efficacy of
this arrangement that saddles the new companies with a large number of redundant
staff. This can be contrasted with the Sections 84 and 85 of the UK Electricity Act of
1989, which provided for ‘compensation’ to the employees who lost jobs. Besides
the above two approaches, a third method of dealing with the surplus employees has
also been used. Some government departments as well as private companies offer
‘voluntary retirement schemes’ with one-time settlement packages. This scheme
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
170
sometimes suffers from considerable adverse selection problems, with more efficient
employees choosing to leave for alternate employment.
Thus, the SEBs have tried to obtain the co-operation of their trade unions and the
employees mainly by assuring them the same job conditions as existed before
restructuring. Of course, many SEBs have enforced a ‘hiring freeze’ for a number
of years. Interestingly, there is a surplus in unskilled labor, but acute shortage in
skilled categories. There should be some scope for training the surplus staff in new
skills. As Upadhyay (2000) points out, some of these surplus employees, who are
above 50 years of age will retire in the next few years, easing the pressure on the
new companies. The federal government is also trying to bring in reforms in the
labor legislation, which could help create a performance-based remuneration
stmcture for employees.
Third, as our study shows, restructuring appears to entail a reduction in the extent
of cross-subsidization. This has been one of the objectives of the reform legislation.
The preamble to the Central Electricity Regulatory Commissions Act of 1998 clearly
mentions the need to provide for “rationalization of electricity tariff, transparent
policies regarding subsidies.” Section- 29(2)(c) of the same Act states clearly that in
setting tariffs the state commissions shall be guided by the provision that “the tariff
progressively reflects the cost of supply of electricity at an adequate and improving
level of efficiency.” Section-29 (2)(e) of the Act also states that “the consumers pay
for the use of electricity in a reasonable manner based on the average cost of supply
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
171
of energy.” These provisions point towards a gradual rationalization of tariffs. The
setting up of independent electricity regulatory commissions, lessening of direct
government control in tariff setting, the provisions in these newly enacted laws
calling for tariff rationalization, and the transparent procedures seem to be initiating
changes in the pivotal area of tariffs. Besides this, the commissions also have started
to make governments tailor their subsidies for the target groups for better effects,
especially in the area of social programs like rural electrification and the supply of
low-priced electricity to some target groups. This dispensation under independent
regulatory commissions also implies a reduction in government control. In the new
scenario, thus, government’s role will shift from being that of a manager/controller
to a supporter/facilitator.
Fourth, the cost of supply seems to be unaffected by restructuring. This finding is
also significant, especially in the face of some skepticism that restructuring may
result in a significant increase in the cost of supply due to loss of economies of scale
and scope. The outcome of such changes in structure and management is likely to be
revealed after some time. The English, Chilean and other experiences also bolster
this claim. It took considerable time before the costs came down in those countries.
The cost reductions actually took place only after the government and the regulators
introduced competition in generation, transmission and distribution. As noted earlier,
countries like Norway and New Zealand experienced significant decreases in costs
after introducing competition in their electricity sectors. Though, restructuring has
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
172
the potential to bring down costs by initiating competition in all sub-sectors, it is yet
to take place in India. More competition in generation would bring down electricity
prices, as the major portion of the cost relates to generation. Thus, this has significant
public policy implications. Indeed, the central plank of restructuring and reform
should be competition in the supply of electricity. The independent regulatory
commission has to take upon itself the task of promoting competition in generation
and distribution in the first phase and in transmission subsequently. The federal
government and the state governments have to recognize this prime need, and
implement such policies accordingly. Hjalmarsson (1996) delineates the need of
such policies in the context of Scandinavian countries and observes that according to
the Public Procurement Act (of Sweden), the domestic publicly owned firms have to
compete with private firms or international firms. “The importance of this Act for
public-sector efficiency and consumer welfare can hardly be exaggerated.... A main
reason for the very low price levels in telecommunication and electricity markets is
this Act’’(Hjalmarsson, 1996; pp.169).
Fifth, as the results reveal, restructuring has positive effects on some crucially
important indicators of financial performance such as tariff collection, and sales
revenue as a ratio of cost. The state electricity boards (SEBs) in India have fared
poorly, especially in these areas. The financial non-viability of the state electricity
boards has had adverse implications for the financing of private power projects. The
actual risks associated with such projects are increased by the perceived risk of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
173
non-payment by these electricity boards. This has hampered investment by
independent power producers (IPPs) in generation.
Restructuring, with its harder budget constraints and infusion of accountability,
seems to initiate improvement in financial management of these companies. With
restructuring, the income transfers between the ‘branches’ stop and these new
companies, functioning under different managerial control, try to be financially self-
reliant. This augurs well for future private investment in the sector.
Sixth, we also got some pertinent results concerning the possible effect of absolute
majority of a political party in the government. This variable, POL, has no
significant effect on the operational performance indicators like plant availability,
plant load factor and forced outage, and also with financial performance indicators
like tariff collection and the extent of cross-subsidization.
However, it shows a highly significant and adverse effect on indicators of labor
efficiency like the number of employees per thousand consumers and the number of
employees per million kwh electricity sold. This effect is visible even when we
control for restructuring. This seems to vindicate the claim that governments (with
one-party majority) tend to force the public sector undertakings to employ more than
what is required. In the absence of strong democratic institutions and traditions that
work as checks and balances, the political majority seems to be able to pursue its
private agenda. Thus, it is not surprising that the variable, POL, seems to affect the
unit cost of supply and sales revenue as a ratio of cost adversely. Our results support
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
174
the theoretical formulations and findings of such authors as Shapiro-Willig (1990),
and Laffont (1995, 1996, and 1998).
The success of liberalization in developing countries seems to rest crucially on the
country’s ability to establish a credible and stable set of regulatory rules which
avoids future political and other interest group interference and therefore provides
new incentive structures. Consolidating democratic and regulatory institutions would
help liberalization immensely. Similarly, building a good accounting and auditing
system would facilitate this process.
Finally, as far as the effect of ownership on the performance indicators of the
thermal power plants is concerned, we find interesting results. State ownership is
associated with relatively worse plant performance. This is evident in all three areas:
plant availability, plant load factor and forced outage. Federal ownership is found to
be significantly and positively different from other forms of ownership, as far as
plant load factor is concerned. However, it is not so in areas like plant availability
and forced outage. Private ownership is found to be marginally different from state
ownership in areas like forced outage. However, after controlling for location-
specific effects, we do not find significant difference between privately owned plants
and other plants in areas like plant availability, and plant load factor.
Thus, the effects of ownership on plant performance are somewhat mixed.
However, we find conclusive evidence that plants owned and managed by states fare
significantly worse than the other plants in all areas of plant performance. It could be
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
175
due to a variety of factors. First, the plants owned by the states are relatively older
than their federal counterparts. As we have discussed in chapter 3, the federal
government took up the construction of plants only in the 1970s, whereas the SEBs
started it in 1950s. This could be contributing to the performance differential.
Secondly, the federal plants are usually much larger in size/capacity, which might
explain it to some extent. Thirdly, the new technology of the new federal plants
might also be a factor. Fourthly, proximity and direct control by the state
governments make them vulnerable to various types of interference, especially
political and bureaucratic. That might be a factor stifling efficiency. The federal
plants are far from the federal power center. Moreover, the federal government does
not control them as directly as do the state governments. Lastly, perhaps the work
culture of the state plants is also different from that of the federal plants. In fact, the
states differ among themselves in their work culture and management. Not
surprisingly, we come across such significant individual fixed effects among state
electricity boards.
As we just discussed, the ownership effect on plant performance seems to be
mixed. However, restructuring shows significantly positive effects on plant
performance even after controlling for ownership and location. Thus, the results
show that ownership is not as crucial as restructuring in matters of plant
performance. This has significant public policy implication in a country like India
that has had a government-managed, vertically integrated electricity sector.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
176
This assumes more significance as governments’ budget problems are often cited as
a motivation for privatization. But as Laffont (1996) aptly points out, many reasons
militate against a smooth functioning of this privatization scheme. Financial
constraints on potential bidders, lack of competition, but more importantly political
and economic uncertainties and the likely inability to commit not to change the
regulation ex post, will make privatization a very ineffective tool for raising public
funds in developing countries. From the pure point of view of raising funds, it seems
fairly clear that privatization of natural monopolies should be delayed until a credible
regulatory environment is built. “This point is reinforced by the fact that public firms
in developing countries are often the most efficient tools for raising tax revenues.
Too hasty a privatization will produce little revenue and quickly precipitate a sharp
increase in the government deficit” (Laffont, 1996). Sometimes, the country or even
a state may want to wait and see if privatization might be more attractive or less
risky later. It might give more time to the regulatory framework to demonstrate its
effectiveness, and for the assets to be properly valued. The opposition of various
interest groups may be overcome, and the most appropriate industrial structure set in
place. The assets can be properly priced in a market that has increasing confidence in
the commitment of the government to the new structure and ownership patterns.
But the governments’ biggest challenge will come in the area of creating and
consolidating appropriate regulatory institutions that counter political and
bureaucratic interference, strengthen investor confidence, promote competition,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
177
foster accountability and enhance performance. Restructuring provides the initial
platform in this regard.
Thus, unless the utilities are financially viable and the regulatory regime credible,
private investors will hesitate to come forward and participate in the process of
liberalization. Restructuring seems to help in this regard. It brings in regulatory
changes and performance gains, and prepares the companies to go for privatization if
and when they deem it appropriate. It helps to initiate salutary changes in the work
culture of the companies, and also in the mindset of the people and the government.
SUGGESTIONS FOR FUTURE RESEARCH
We studied the effect of restructuring on operational and financial performance of
state electricity boards in India. We also studied the effect of ownership on the
operational performance of the thermal power plants. We used various performance
indicators and indices of restructuring, used panel data, and got some interesting
results. However, there are many issues that still remain to be investigated. We could
not do so mainly due to data constraints.
Future research may address the following questions and issues. First, one of the
important determinants of success of electricity sector reform is the quality or
reliability of power supply. There have been many studies that indicate that some
consumers are willing to pay higher prices to get more reliable electricity without
any “black-outs” or “brown-outs” (very low voltage supply). We could not
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
178
investigate this question due to lack of such data as MTBF (Mean Time Between
Failures) in this regard. Apart from this, the distribution companies may also conduct
studies to assess the consumers’ satisfaction about reliability, prices, etc. The
Regulatory Commissions can instruct the distribution companies to conduct such
studies in their operational areas. These studies will help investigate if restructuring
also tends to improve the reliability or the quality of the power supply.
Second, the Indian power sector suffers from very high levels of transmission and
distribution losses. Interestingly, the actual levels of these losses are being reported
now only in the restructured state electricity boards, where it is no longer possible to
conceal such losses in the guise of agricultural supply. The reduction of such losses
remains perhaps the most difficult challenge for the states. The distribution
companies have already started to take a number of steps in this regard. It will be
interesting to determine whether or not restructuring can significantly bring down
these losses.
Third, India is yet to adopt third party access (TPA), allowing competition in
transmission network. As per the prevailing “single-buyer” model, the state-owned
transmission companies have the monopoly over transmission in India. Many
countries have experienced transmission cost decreases after introducing TPA. Some
states in India may opt for it in future. One could, then, study if TPA results in lower
transmission costs, thereby reducing the cost of supply of electricity.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
179
Fourth, rural electrification is an important political and socio-economic issue in
India. Many fear that restructuring and subsequent privatization would hamper this
effort. The Reform Acts and the Electricity Regulatory Commissions Act clearly
mandate the state governments to make budgetary provisions for such social
programs as rural electrification and supply of subsidized tariff to the economically
disadvantaged groups. It will be pertinent to assess the result after some time.
Fifth, so far no state has relapsed to the earlier state of vertical integration after
taking up restructuring. If it happens in future, it will be worthwhile to investigate
and analyze the factors that could cause such a relapse. This may also help to take
corrective measures to prevent such a relapse.
Sixth, privatization has not yet been accepted as the way out. Indeed, in only one
state, Orissa, the distribution companies (DISTCOs) have been privatized. Similarly,
private companies have started to participate in power generation in a very limited
manner. We have already seen cases where some private companies have left the
projects after signing the memos of understanding, and even after taking up the
management.1 Privatization is still not seen as the obvious alternative for various
reasons. However, the country or some states may opt for it in the electricity sector
in future. In that case, one could study its effect on performance and investment.
1 One such case is the multi-national company, AES, leaving the CESCO in Orissa recently.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
180
Seventh, environmental degradation is an important issue in many countries
including India. The preamble to the Electricity Regulatory Commissions Act of
1998 enshrines the need to promote “environmentally benign policies.” Section-22
(2) (o) of the Act also empowers the state governments to confer upon the State
Commissions the function to “co-ordinate with environmental regulatory agencies
and to evolve policies and procedures for appropriate environmental regulation of the
electricity sector and utilities in the State.” One could study the possible effects of
restructuring on the environment, especially on such indicators as air quality, which
could be measured in terms of the amount of particulate matter in air. We could not
study this due to data constraints. However, this situation could change as reliable
panel data may be made available. That would make such a study feasible.
Lastly, with the advent of competition in generation and distribution, the
companies would have to assess the demand with respect to differential pricing (like
time-of-the-day pricing or interrupted-supply pricing). This will be more relevant in
the agricultural sector. Farmers are already using electricity at off-peak hours for
irrigation in many states. The differential pricing could help them plan their
consumption patterns and time. Moreover, the hikes in tariffs are likely to bring
about further changes in demand. In particular, the tariffs for agricultural
consumption have already started to increase in the reforming states. They are likely
to go up further to reduce cross-subsidization. As discussed earlier, the cross
subsidization is likely to witness a downward trend in the future in keeping with the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
181
new legislation to work towards the “rationalization” of tariffs. This will bring about
considerable changes in the demand for electricity in the agricultural sector.
In view of the increase in tariffs, possible differential pricing and the consequent
changes in demand, the electricity distribution companies would have to prepare
their supply schedules in accordance with their expected shares in the market. They
would perhaps depend on game-theoretic modeling and simulation studies to do
forecasting about demand and supply situations in future. While deciding these
details, they would also have to take into account any change in regulatory
framework, as it might change the possible pay-offs, and therefore, may call for
changes in strategies.
On the basis of evidence to date, restructuring seems to have helped the electricity
sector in India. However, the states will have to tailor restructuring to suit their
requirements, their economic and institutional attributes. The design of the electricity
industry structure and the regulatory regime cannot be settled once and for all. To the
contrary, the task calls for careful policy design and effective implementation, its
continuous monitoring, and innovative thinking about the reform process as such.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
182
REFERENCES:
Acton, J.P., and I. Vogelsang (1989): “Introduction to Symposium on Price Cap
Regulation”, Rand Journal o f Economics, 20(3), pp. 369-372.
Acutt Melinda, Caroline Elliott, Terry Robinson (2001): “Credible regulatory threats”,
Energy Policy 29, pp.911-916
Aghion P., and R. Burgess, (1992): “Financing and Development in Eastern Europe and the
former Soviet Union.” Oxford. Mimeo.
Alchian, A. (1965): “Some Economics of Property Rights”, II Politico, vol. 30, 816-829.
Antonette D’sa., K.V. Narasimha Murthy, and Amulya K. N. Reddy (1999): “India’s Power
Sector Liberalization : An Overview”, Economic and Political Weekly, June 5,
pp.1427-1434.
Armstrong, M, Cowan.S. and Vickers, J (1994): Regulatory Reform: An Economic Analysis
and British Experience, MIT Press.
Averch, H. and Johnson, L.L (1962): “Behavior of the Firm under Regulatory Constraint”,
American Economic Review, vol. 52, pp. 1052-1069.
Baron. D, and R. Myerson, (1982): “Regulating a Monopolist with Unknown Costs.”
Econometrica 50: pp. 911-930
Bishop, M. and Thompson, D. (1992) : “Regulatory Reform and Productivity Growth in the
UK’s Public Utilities”, Applied Economics, vol. 24, no. 11, pp. 1181-90.
Bjorvatn, K. and Tjotta, S (1993): “Deregulation and Common Carriage in the Nordic Power
System”, The Energy Journal, vol. 14, no. 4.,pp. 57-74
Bolton P, and G. Rolland, (1992): “The Economics of Mass Privatization”. Ecole
Poly technique No. 375. Paris
Boycko, M., Shleifer, A.,and Vishny, R.W. (1996): “A Theory of Privatization”, Economic
Journal, 106, pp. 309-319.
Bresser Pereira, Luiz Carlos, Maravall JM and Przeworski Adam (1993): Economic Reforms
in new democracies: A social-democratic approach. Cambridge University Press.
Bunn, D. and Vlahos, K (1989): “Evaluation of the Long-term effects on Electricity Prices
following Privatization”, Fiscal Studies, vol. 10, no. 4, pp.104-116.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
183
Bums, P. and Weyman-Jones, T (1994): Productive Efficiency and the Regulatory Review
of Regional Electricity Companies in the UK, Regulatory Policy Research Center,
Discussion Paper No. 1 .
Chessire, J (1997): “Reform of UK Electricity Regulation”, The Business Economist, vol.28,
no.2 , pp.31-41.
Cicchetti, Charles and Kristina M. Sepetys (1995): Restructuring Electricity Markets: A
World Perspective. Visions Communication.
Christensen L.R., and W.H. Greene (1976): “Economies of Scale in U.S. Electric Power
Generation” in Journal of Political Economy, 84: pp. 655-676.
Cornwell, Christopher, and William N. Tmmbull (1994): “Estimating the Economic Model
of Crime with Panel Data”, The Review of Economics and Statistics, Vol. 76, Issue
2, 360-66
Croker,K.,and K. Reynolds (1989): “Efficient Contract Design in Long-term Relationships :
The Case of Airforce Engine Procurement.” Working Paper 10-89-1. Pennsylvania
State University.
Culy, J. G, G. Read, and B.D. Wright (1996): “The Evolution of New Zealand’s Electric
Supply Structure” in Gilbert and Kahn 1996, pp.312-365.
Demsetz, H (1968): “Why Regulate Utilities ?”, Journal of Law and Economics, (11), pp.
55-65
Domberger, S. (1986): “Economic Regulation through Franchising” in Kay, J., C. Mayer and
D. Thompson (Eds.) (1986), Privatisation and Regulation: The UK Experience.
Oxford University Press.
Dyner, Isaac and Erik R. Larsen (2001): “ From Planning to strategy in the electricity
industry”; Energy Policy 29. pp. 1145-1154.
EC. 1997. Directive 96/92/EC of the European Parliament: Official Journal No. L 027.
30/01/97 P.0020
Electricity Market Reform: An IEA Handbook, 1999. OECD/IEA Publication.
Electric Power Surveys (14th , 15th and 16th ), Central Electricity Authority o f India
Estache, A. and Rodriguez-Pardina, M (1996): “Regulatory Lessons from Argentina’s Power
Concessions”, Public Policy for the Private Sector, September, 1996, pp.5-8, World
Bank.
Galal, A., Jones, L.,Tandon, P., and Vogelsang, I. (1994) : “Welfare Consequences of
Selling Public Enterprises - An Empirical Analysis”. Oxford University Press.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
184
Ghafoor, Abdul, and John Weiss (1999). “Privatization of Electric Power Sector in Pakistan:
Some Important Issues”, The Pakistan Development Review, 38:1 (Spring 1999),
pp.69-84
Gilbert Richard. J and Kahn Edward P. (1996) ed.: International Comparisons of Electricity
Regulation. Cambridge University Press.
Gould. D., and J. Amaro-Reyes, (1983): “The Effects of Corruption on Administrative
Performance.” World Bank Staff Working Paper No. 580. Washington D.C.
Greenwald.B., (1984): “Rate Base Selection and the Structure of Regulation.” Rand Journal
of Regulation, 15: pp. 85-95.
GRIDCO, (2001): “An Overview of Gridco”, Grid Corporation of Orissa Ltd.
Grossman, S., and O. Hart, (1986): “The Costs and Benefits of Ownership: a Theory of
Lateral and Vertical Integration.” Journal of Political Economy, 94: pp. 691-719
Haldea, G (2001): “Whither Electricity Reforms? “, Economic and Political Weekly, April
28, 2001.
Hausman, W.J., and J.L. Neufeld. (1991): “Property rights versus public spirit: Ownership
and efficiency of U.S. electric utilities prior to rate-of-retum regulation”, Review of
Economics and Statistics 73: 414-423
Hausman, J.A (1978): “Specification Tests in Econometrics”, Econometrica, Vol.
46, No. 6. (Nov., 1978), pp. 1251-1271
Hawdon,D.(1996): “Performance of Power Sectors in Developing Countries - A Study of
Efficiency and World Bank Policy using Data Envelopment Analysis”, Surrey
Energy Economics Centre Discussion Paper, No. 88.
Hjalmarsson, L (1996): “From Club-regulation to market competition in the Scandinavian
electricity supply industry”, in Gilbert and Kahn (eds.), International Comparisons
of Electricity Regulation (pp. 138-39)
Huettner, D.A. and J.H. Landon (1977): “Electric Utilities: Scale Economies and
Diseconomies”, Southern Economic Journal, 44:pp. 883-912.
Hsiao, Cheng (1986) : Analysis of Panel Data, Cambridge University Press.
Hunt, S., and Shuttleworth G. (1998): “The Difference between Restructuring and
Privatization”, Viewpoint, Winter 1997-98, pp. 1-7.
Izaguirre, A.K. (1998): “Private participation in the electricity sector - Recent Trends”,
Private Sector, December.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
185
Jones, L., P. Tandon, and I. Vogelsang, (1990): Selling Public Enterprises: A Cost Benefit
Methodology. MIT Press.
Joskow, Paul L( 1998a): “Electricity Sectors in Transition”, The Energy Journal, Vol. 19,
No.2, pp. 25-52.
Joskow, Paul L(1998b): “Regulatory Priorities for Infrastructure Sector Reform in
Developing Countries”, Proceedings of the Annual World Bank Conference on
Development Economics, 1998, pp. 191-223
Kahn, E.,Bailey, S., and Pando, L. (1997): “Simulating Electricity Restructuring in
California: Interactions with Regional markets”, Resource and Energy Economics,
vol.19, pp. 3-28.
Kwoka, J. Jr. (1995): “Public versus private ownership and economic performance:
Evidence from the US electric power industry”, Discussion Paper 1712. Harvard
Institute of Economic Research
Kwoka, J. Jr.(1996): Power Structure: Ownership, Integration and Competition in the US
Electricity Industry. Boston: Kluwer.
Laffont, J.J.(1994a.): “The New Economics of Regulation ten Years After”, Econometrica,
62: pp. 507-537
Laffont, J.J.(1995): “Industrial Policy and Politics.” International Journal of Industrial
Organization
Laffont, JJ.(1996): “Regulation, Privatization, and Incentives in Developing Countries” in
Current Issues in Economic Development, ed. by M.G. Quibria and Malcolm
Dowling. Oxford University Press. 1996.
Laffont, J.J and D. Martimort, (1994): “Separation of Regulators against Collusive
Behavior”. Institut d’Economie Industrielle. Mimeo
Laffont, J.J and J. Tirole, (1991): “Privatization and Incentives”. Journal of Law, Economics
and Organization, 7: pp. 84-105.
Laffont, J.J and J. Tirole (1992): “Cost Padding, Auditing and Collusion.” Annales
d ’ Economie et Statistique, 25-26: pp. 205-226.
Laffont, JJ(1998): “Competition, Information, and Development”, Proceedings of the
Annual World Bank Conference on Development Economics, 1998, pp. 237-257
Larsen, E.R, and Bunn, D.W.,(1999): “Deregulation in electricity: understanding strategic
and regulatory risk”, Journal of the Operational Research Society 50 (4), pp. 337-
344
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
186
Littlechild, S.C. (1983): Regulation of British Telecommunications Profitability.
London:HMSO.
Loeb, M., and W. Magat, (1979): “A Decentralized Method of Utility Regulation.” Journal
of law and Economics, 22: pp. 399-404.
MaCurdy, T.E. (1982): “The Use of Time Series Processes to Model the Error Structure of
Earnings in a Longitudinal Data Analysis”, Journal of Econometrics, 18, pp. 83-114.
Maskin,E., (1992): “Auctions and Privatization.” Harvard University.
Milgrom, P and Roberts, J.(1990): “Bargaining Activity, Influence Costs, and the
Organization of Economic Activity”, in J.E. Alt and K.A. Sheple (eds.), Perspectives
on Positive Political Economy. New York: Cambridge University Press.
Milgrom, P and Roberts, J. (1992): Economics, organization, and management. Englewood
Cliffs, New Jersey: Prentice-Hall, cl992.
Moen, J (1994): “Electricity Utility Regulation, Structure and Competition. Experiences
from the Norwegian Electric Supply Industry.” Oslo: Norwegian Water Resources
and Energy Administration, NVE, 1994.
Navroz K Dubash, Sudhir Chella Rajan (2001): "Power Politics: Process of Power Sector
Reform in India”, Economic and Political Weekly, September 1-7.
Nerlove, M., (1968): “Returns to scale in Electricity Supply” in A. Zellner (eds.), Readings
in Economics Statistics and Econometrics. Boston: Little, Brown and Co. pp.409-
439
Newbery, D (1992a): “The Role of Public Enterprises in the National Economy”, Asian
Economic Review 10: pp. 1-34.
Newbery, D (1997): “Privatization and Liberalization of Network Utilities”, European
Economic Review, 41, pp. 357-383.
Newbery, D (1999): Privatization, Restructuring, and Regulation of Network Utilities. The
MIT Press.
Newbery, D.M.G, and Pollitt, M.G (1997): “The Restructuring and Privatization of Britain’s
CEGB - Was it worth it T', Journal of Industrial Economics, vol. 45, no.3, pp. 269-
303.
Niskanen, W. A. (1968): “The Peculiar Economics of Bureaucracy”, American Economic
Review, Papers and Proceedings, vol. 58, pp.298-305.
OECD/IEA (1999): “Selected Energy Statistics for 1996”, cited in April, 1999.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
187
Pechman, Carl (1993): Regulatory Power: the economics of electricity in the information
age. Boston: Kluwer Academic Publishers.
Peltzman, S.(1976): “Toward a more general theory of regulation”, Journal of Law and
Economics 19: 211-240
Perotti, E.(1995): “Credible Privatization”, American Economic Review, vol. 85, 847-59
Pesic, Radmilo V. and Diana Urge-Vorsatz (2001): “Restructuring of the Hungarian
Electricity Industry”, Post-Communist Economies, Vol. 13, No. 1,2001
Planning Commission of India (http://planningcommission.nic.in)
Pollitt, M.G. (1993): The Relative Performance of Publicly Owned and Privately Owned
Electric Utilities: International Evidence. Ph.D. thesis. Oxford University.
Pollitt, M.G. (1995): Ownership and Performance in Electric Utilities. Oxford University
Press.
Pollitt, M.G.(1997): “The Impact of Liberalization on the Performance of the Electricity
Supply Industry: An International Survey”, The Journal Of Energy Literature, 111(2)
Przeworski, Adam (1991): Democracy and the Market: Political and Economic Reforms in
Eastern Europe and Latin America. Cambridge: Cambridge University Press.
Publications of the Power Ministry, Government of India (http://powermin.nic .in)
Rao, M.G, Kalirajan K.P and R. Chand (1998): The Economics of Electricity Supply in
India. Macmillan India Ltd.
Roberts, M.J (1986): “Economies of Density and Size in the Production and Delivery of
Electric Power”, Land Economics, 62 :pp. 378-387.
Rodrik, Dani (1996): “Understanding Economic Policy Reform”, Journal of Economic
Literature, Vol. XXXIV (March 1996), pp. 9-41.
Sachs.J (1994): The Political Challenge of Economic Transition: The Case of Poland.
Unpublished paper (cited in Rodrik, 1996).
Sappington,D., and J. Stiglitz. (1987): “Privatization, Information and Incentives.” Journal
of Policy Analysis and Management, 6: pp. 567-582.
Schmalensee, Richard(1989): “Good Regulatory Regime”, RAND Journal of Economics,
20(3); pp. 417-36.
Schmidt,K., (1990): “The Costs and Benefits of Privatization.” DP A-330, Univ. of Bonn.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
188
Shapiro, C., and R. Willig. (1990): “Economic Rationales for the Scope of Privatization.”
DP 41. Princeton University.
Shleifer, A. (1985): “A Theory of Yardstick Competition”, Rand Journal of Economics 16 :
319-327.
Shleifer,A., andR. Vishny, (1994): “Politicians and Firms”.
Sinha, S, (2001), “Study On Orissa Power Sector Reforms, Phase II”, Power Finance
Corporation, New Delhi.
Spiller, P. (1993), “Institutions and regulatory commitment in utilities’ privatization”,
Industrial and Corporate Change 2: pp. 317-380
Spiller,P. (1994), Contracts, administrative law or regulatory specificity: workable
regulatory frameworks for developing economies. Washington, D.C: Institute for
Policy Reform, Working Paper IPR97.
Spiller Pablo T. and Luis V. Martorell (1996): “How should it be done? Electricity
regulation in Argentina, Brazil, Uruguay, and Chile” in Gilbert and Kahn (1996),
82-125
Stigler, George (1971): The Economic Theory of Regulation, Bell Journal of Economics, ’71
Talvadkar, Vivek, and Binoy Mishra (2000), “Distribution Privatization in India: Issues and
Approach”, ProjectFinance, November
Tenenbaum, B., R. Lock, and J. Barker (1992): “Electricity Privatization — Structural,
Competitive and Regulatory Options”, Energy Policy 20:12, 1134-1160
Thatcher, M. (1993): The Downing Street Years. London: HarperCollins.
Williamson, O. (1985): “The Economic Institutions of Capitalism” New York: Free
Press.
Willig, R.D (1994): “Public versus Regulated Private Enterprise”, Proceedings of the World
Bank Conference on Development Economics, World bank.,pp.155-180.
World Bank (1993a): “Power Supply in Developing Countries; Will Reforms Work?”,
Proceedings of a roundtable co-sponsored by World Bank and Electricite de France,
April 27-28, Washington D.C.
— (1993b): The World Bank’ s Role in the Power Sector, Washington D. C.
— (1993c): Conference on Power Sector Reforms in India, Jaipur, October 29-31, ESMAP,
Washington D.C.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
189
World Bank (1995): Bureaucrats in Business. Oxford University Press.
Yarrow, G.K. (1988): “The Price of Nuclear Power”, Economic Policy 6; 81-13
Yarrow, G.K. (1992): British Electricity Prices since Privatization. Oxford: Regulatory
P olicy Institute.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
APPENDIX - A
Panel Regressions Results.
A.I. Dependent Variable: Plant Availability (SEB Data Set)
Variables
(1) (2) (3) (4)
Restr 4.42“ * 4.10*“
— . . . .
Resty
— —
6.08***
. . . .
Relag
— . . . . —
4.13***
POL 1.00 0.97 1.54 1.21
Demand 1.84* 1.84* 2.46* 1.88*
Omc
. . . .
-0.13 -0.17 0.14
Constant
— . . . . —
Fixed Effects (CS & Time)’* (CS & Time)** (CS & Time)** (CS & Time)**
P r> m 0.0384 0.011 0.0196 0.0155
Hausman’s test supports fixed effects specification at 5% level o f significance
DF 153 151 151 151
R-sq. 0.8307 0.8301 0.8483 0.8304
(t-statistics reported)
*** Significant at 1% level of significance
** Significant at 5% level of significance
* Significant at 10% level of significance
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
191
A.2. Dependent Variable: Plant Load Factor (SEB Data Set)
Variables (1)
(2) (3) (4)
Restr 4.96*** 5.02***
. . . . —
Resty —
—
7.54’* *
. . . .
Relag —
. . . . —
5.11’* *
POL 0.01 -0.07 0.57 0.20
Demand 1.13 1.86* 1.25
Constant —
— —
Fixed Effects (CS & Time)** (CS & Time)** (CS & Time)** (CS & Time)**
Pr > m 0.0406 0.0429 0.0298 0.0316
Hausman’s test supports fixed effects specification at 5% level of significance
DF 154 153 153 153
R-sq. 0.8796 0.8806 0.8987 0.8812
(t-statistics reported)
*** Significant at 1% level of significance
** Significant at 5% level of significance
* Significant at 10% level of significance
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
192
A .3. Dependent Variable: Forced Outage (SEB Data Set)
Variables
(1) (2) (3) (4)
Restr -4.59*** -3.85**’
— —
Resty
— —
- 4.98***
—
Relag
— . . . . —
-3.31***
POL -0.52 -0.45 -0.87 -0.61
Demand -0.23 -0.06 -0.46 -0.05
Omc
. . . .
-1.25 -1.39 -1.60
Constant
— . . . . —
Fixed Effects (CS & Time)*’* (CS & Time)*** (CS & Time)*** (CS & Time)**
Pr > m 0.0011 0.0052 0.0044 0.0282
Hausman’s test supports fixed effects specification at 5% level of significance
DF 153 151 151 151
R-sq. 0.7554 0.7574 0.7711 0.7516
(t-statistics reported)
*** Significant at 1% level of significance
** Significant at 5% level of significance
* Significant at 10% level of significance
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
193
A.4. Dep. Var: Number of Employees per Million KWH Electricity Sold (SEB Data Set)
Variables
(1) (2) (3)
Restr 1.55
— —
Resty
. . . .
1.76*
. . . .
Relag
. . . . —
1.49
POL 2.51** 2.67*** 2.59**
Lvtot - 2.09** -2.31** - 2.09**
Urratio 5.43*** 5.54*** 5.87***
Hydel -1.48 -1.54 - 1.74*
Constant
— . . . . . . . .
Fixed Effects (CS & Time)** (CS & Time)** (CS & Time)**
Pr > m 0.0201 0.0203 0.0187
Hausman’s test supports fixed effects specification at 5% level of significance
DF 138 138 138
R-sq. 0.9613 0.9615 0.9612
(t-statistics reported)
*** Significant at 1% level of significance ** Significant at 5% level of significance
* Significant at 10% level of significance
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
194
A .5. Dep. Var: Number of Employees per Thousand Consumers (SEB Data Set)
Variables
(1) (2) (3) (4)
Restr -1.96* -2.03**
— . . . .
Resty
. . . . —
-1.78*
. . . .
Relag
— — —
- 1.62
POL 3.81*" 3.17**’ 2.96***
~ ~ _ * * *
3.05
Lvtot
• ~ _ * * *
-4.86 -4.11*** -3.65***
A 1 ,!* * *
- 4.14
Urratio 5.44*** 5.54*** 5.43*** 5.38***
Hydel
—
-0.92 -0.75 -0.61
Constant
. . . . — —
Fixed Effects (CS & Time)*’* (CS & Time)** (CS & Time)** (CS & Time)**
Pr > m < 0.0001 0.0168 0.0171 0.0172
Hausman’s test supports fixed effects specification at 5% level of significance
DF 152 138 138 138
R-sq. 0.9291 0.9310 0.9305 0.9302
(t-statistics reported)
*** Significant at 1% level of significance
** Significant at 5% level of significance
* Significant at 10% level of significance
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
195
A.6. Dependent Variable: Extent of Cross-subsidization (SEB Data Set)
Variables
(1) (2) (3)
Restr -2.60**
— —
Resty
—
-3.44***
. . . .
Relag
— . . . .
-2.67***
POL -0.17 -0.42 -0.34
Lvtot 4.22
**#
4.62 4.32***
Urratio 2.42 2.48 2.15**
Constant 0.74 0.71 1.13
Pr > m 0.2607 0.2491 0.3004
Hausman’s test supports random effects specification
DF 176 176 176
R-sq. 0.2163 0.2378 0.2163
(t-statistics reported)
*** Significant at 1% level o f significance
** Significant at 5% level of significance
* Significant at 10% level of significance
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
A.I. Dependent Variable: Sales Revenue as a Ratio of Cost (SEB Data Set)
Variables
(1) (2) (3)
Restr 0.37
— —
Resty
. . . .
2.33**
—
Relag
— —
2.06**
POL - 2.28** -2 .1 8 ’* - 2.24**
Lvtot - 2.45** - 3.25 -3.00***
Urratio -2.41** -3.28*** -3.11***
Constant
. . . . — —
Fixed Effects (CS & Time)’* * (CS & Time)*** (CS & Time)***
Pr > m 0.0048 < 0.0001 < 0.0001
Hausman’s test supports fixed effects specification
DF 152 152 152
R-sq. 0.7174 0.7269 0.7249
(t-statistics reported)
*** Significant at 1% level of significance
** Significant at 5% level of significance
* Significant at 10% level of significance
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
197
A.8. Dependent Variable: Average Tariff Collected (SEB Data Set)
Variables
(1) (2) (3)
Restr 4.39"*
. . . . —
Resty 4.19***
—
Relag
. . . . . . . .
3.66***
POL -1.88* -1.46 - 1.55
Lvtot 0.18 -0.42 0.30
Urratio - 1.24 -0.99 -0.56
Constant
— — —
Fixed Effects (CS & Time)*** (CS & Time)*** (CS & Time)*’*
P r> m < 0.0001 < 0.0001 < 0.0001
Hausman’s test supports fixed effects specification
DF 152 152 152
R-sq. 0.6619 0.6584 0.6499
(t-statistics reported)
*** Significant at 1% level o f significance
** Significant at 5% level of significance
* Significant at 10% level of significance
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
A.9. Dependent Variable: Log (Average Tariff Collected) (SEB Data Set)
Variables
(1) (2) (3)
Restr 4.18***
. . . . —
Resty
. . . .
4.42“ *
—
Relag
— —
3.70***
POL - 1.45 -1.05 - 1.15
Lvtot -0.58 - 1.28 -0.51
Urratio -0.26 -0.18 0.34
Constant
— — —
Fixed Effects (CS & Time)*** (CS & Time)**’ (CS & Time)***
P r> m < 0.0001 < 0.0001 < 0.0001
Hausman’s test supports fixed effects specification
DF 152 152 152
R-sq. 0.6717 0.6758 0.6643
(t-statistics reported)
*** Significant at 1% level of significance
** Significant at 5% level of significance
* Significant at 10% level of significance
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
199
A .10. Dependent Variable: Unit Cost of Supply (SEB Data Set)
Variables
(1) (2) (3) (4)
Resty -1.26 -1.15
— —
Relag
. . . . —
-1.43 - 1.48
POL 2.66” *
_ ^ * * *
2.64 2.66*** 2.65*’*
Demand
—
3.32***
. . . .
3.38***
Fcost 4.24*” 4.71*** 4.34*** 4.85**’
Powcost 2.91” * 3.21*** 3.03*** 3.33*’*
Td 7.66*** 8.47*** 7.74 8.60***
Hydel 0.26 0.55 0.35 0.64
Lvtot -1.39 - 1.29 - 1.51 - 1.42
Urratio
_ _ * *
2.39 1.08
— ~ — * *
2.32 0.99
Constant -0.22 -0.00 -0.10 0.09
Pr > m 0.6292 0.5186 0.6538 0.5311
Hausman’s test supports random effects specification
DF 159 158 159 158
R-sq. 0.4624 0.4980 0.4632 0.5001
(t-statistics reported)
*** Significant at 1% level of significance
** Significant at 5% level of significance
* Significant at 10% level of significance
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 0 0
A. 11. Dependent Variable: Log (Unit Cost of Supply) (SEB Data Set)
Variables
(1) (2)
Relag - 1.17 - 1.09
POL 2.40** 2.34**
Demand
—
2.40*’
Lnf 3.13*** 3.15***
Lnpw 6.59*** 6.65***
Td 8.94*** 9.44***
Hydel -2.41**
_ . . * *
-2.44
Lvtot -1.48 - 1.39
Urratio 3.39***
_ _ * * * *
2.71
Constant -0.40 -0.52
P r> m 0.3186 0.2993
Hausman’s test supports random effects specification
DF 159 158
R-sq. 0.5943 0.6108
(t-statistics reported)
*** Significant at 1% level of significance
** Significant at 5% level of significance
* Significant at 10% level of significance
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
201
A. 12. Dependent Variable: Plant Availability (PLANTS Data Set)
Variables
(1) (2) (3) (4) (5) (6)
Restr
—
2.49**
. . . .
2.57’*
SS -3.01*** -3.12***
—
-2.29** -2.09*’ - 2.20**
FS —
. . .
1.97**
. . . . . . . .
PS —
. . . . . .
0.89
. . . .
—
. . . . . . . . . .
3.63*** 3.61***
AS —
. . . . . . —
0.04 0.11
GU —
. . . . . . —
3.05*** 3.13” *
HA —
— . . . —
1.27 1.21
KA
. . . . . . . . . . . . .
_ * *
2.03 2.02**
MP
. . . . . . . . . —
2.84*** 2.93*”
MH
. . . . . . . . . —
_ ,_***
3.67 3.76*”
OS
. . .
—
. . . . . . .
1.41 1.32
PU
. . . . . . . . . —
2.63*** 2.70*”
RA
. . . . . . . . . —
1.84* 1.86*
TN
. . . . . . . . . . . . .
2.65***
_ _ _ * # *
2.73
UP
. . . . . . . . . —
1.36 1.40
WB
. . . . . . . . . —
2.43*’ 2.49”
DE
. . . . . . . . . —
2.14** 2.20**
Constant 28.12 28.31 37.23 23.54 7.52 7.50
P r> m 0.7008 0.5355 0.3756 0.4941 0.4587 0.4109
Hausman’s test supports random effects specification
DF 764 763 764 763 750 749
(t-statistics reported) *** Significant at 1% level of significance
** Significant at 5% level of significance * Significant at 10% level of significance
s® The details of States are on page 206.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 0 2
A .13. Dependent Variable: Plant Availability (Contd.) (PLANTS Data Set)
Variables
(1) (2) (3)
FS 1.43
....
1.58
PS
—
1.22 1.39
y^p*@
3.49” * 3.76 3.47
AS 0.10 0.51 0.06
GU 3.20*” 3.03**’ 2.86*”
HA 1.22 1.31 1.21
KA 1.94* 2.00” 1.94*
MP 2.75” * 3.15*** 2.71*”
MH
_ „_***
3.77 3.63’* * 3.46***
OS 1.42 2.83*” 1.33
PU 2.52** 2.60*** 2.51”
RA 1.76* 1.82* 1.75*
TN 2.57* 3.00*** 2.53”
UP 1.35 1.79* 1.31
WB 2.81***
_ __ * * *
2.72 2.27**
DE 2.08**
_ . _ * *
2.42 2.05**
Constant 7.50 7.34 7.52
Pr > m 0.2245 0.0589 0.2961
Hausman’s test supports random effects specification at 5% level of significance.
DF 750 750 749
(t-statistics reported) *** Significant at 1% level of significance
** Significant at 5% level of significance * Significant at 10% level of significance
s® The details of States are on page 206.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
203
A. 14. Dependent Variable: Plant Load Factor (PLANTS Data Set)
Variables
(1) (2) (3) (4) (5) (6)
Restr
—
3.00*"
— —
3.08*"
SS -3.13"* -3.26***
—
-2.71*** -2.54" -2.66***
FS —
. . .
2.47**
— . . . . . . . .
PS —
. . . . . .
0.34
— . . . .
—
. . .
—
—
. _ . * * *
3.34 3.34"*
AS
. . . . . . . . . —
-0.79 -0.73
GU
. . . . . . . . . —
2.47" 2.57"
HA
. . . . . . . . . —
1.20 1.15
KA
. . . . . .
—
. . . .
2.06** 2.06"
MP
. . . . . . . . . —
2.77***
__ _ * * *
2.87
MH
. . . . . . . . . —
3.10"* 3.20*"
OS
. . . . . . . . . —
0.45 0.40
PU
. . .
—
. . . —
2.18" 2.26**
RA
. . . . . . . . . —
2.30" 2.34"
TN
. . . . . . . . . —
2.79*’* 2.88*"
UP
. . . . . . . . . . . . .
1.51 1.55
WB
. . . . . . . . . —
1.53 1.60
DE
. . . . . . . . . —
1.76* 1.83*
Constant 18.79 19.03 22.27 16.28 4.58 4.56
Pr > m 0.8584 0.1252 0.8879 0.9506 0.6703 0.2000
Hausman’s test supports random effects specification
DF 764 763 764 763 750 749
(t-statistics reported) *** Significant at 1% level of significance
** Significant at 5% level of significance * Significant at 10% level of significance
s-! The details of States are on page 206.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
A. 16. Dependent Variable: Plant Load Factor (Contd.) (PLANTS Data Set)
Variables
(1) (2) (3)
FS 1.95*
—
2.11“
PS
. . . .
1.27 1.53
^p$@
3.27*** 3.56*’*
_ _ ,*#*
3.24
AS -0.72 -0.18 -0.76
GU
_ __***
2.75 2.57** 2.39“
HA 1.18 1.27 1.17
KA 2.02** 2.05** 2.01“
MP 2.72*** 3.19“ * 2.68*“
MH
_ _ ^
3.33 3.17*’* 3.00
OS 0.52 2.52“ 0.40
PU 2.13** 2.20 2.12”
RA 2.25** 2.28** 2.24“
TN 2.73 3.25*“ 2.69*“
UP 1.50 2.06“ 1.45
WB 2.02** 2.00“ 1.47
DE 1.73* 2.14“ 1.70*
Constant 3.93 3.63 3.96
P r> m 0.7833 0.0639 0.3942
Hausman’s test supports random effects specification at 5% level of significance
DF 750 750 749
(t-statistics reported) *** Significant at 1% level o f significance
** Significant at 5% level o f significance * Significant at 10% level of significance
s® The details of States are on page 206.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
205
A. 17. Dependent Variable: Forced Outage (PLANTS Data Set)
Variables
(1) (2) (3) (4) (5) (6)
Restr — - 2.40**
. . . . — —
-2.52**
SS 2.11** 2.23**
. . . . —
1.95* 2.07“
FS —
. . .
-0.76 - 1.09
— . . . .
PS —
. . . . . .
_ ,**
-2.46
— . . . .
Ap$ @
—
. . . . . . —
-3.06*** -3.02***
AS
. . . . . . . . . —
1.14 1.07
GU
. . . . . . . . . —
-2.87** -2.96*“
HA —
. . . . . . —
-0.46 -0.37
KA
. . . . . . . . . —
-1.83* -1.80*
MP
. . . . . . . . . —
-2.39** -2.47**
MH
. . . . . . . . . —
-2.79*** -2.88***
OS
. . . . . . . . . —
-0.88 -0.74
PU
. . . . . . . . . —
-2.21** -2.29**
RA
. . . . . . . . . —
-1.64 -1.65*
TN
. . . . . . . . . —
-2.30** -2.38“
UP
. . . . . . . . . —
-0.64 -0.67
WB
. . . . . . . . . —
-1.60 -1.67*
DE
. . . . . . . . . . . . .
-1.13 -1.20
Constant 5.08 5.08 10.16 10.45 3.81 3.86
Pr > m 0.9295 0.8048 0.4651 0.5974 0.9790 0.9686
Hausman’s test supports random effects specification
DF 764 763 764 763 750 749
(t-statistics reported) *** Significant at 1% level of significance
** Significant at 5% level of significance * Significant at 10% level of significance
$ ® The details o f States are on page 206.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
206
A .18. Dependent Variable: Forced Outage (Contd.) (PLANTS Data Set)
Variables
(1) (2) (3)
FS -1.02
—
-1.20
PS
—
-1.59 -1.72*
AP -3.00*** -3.20 -2.98’“
AS 0.94 0.69 0.99
GU
_ ^ * * *
-3.03 -2.76**
_ , , * * *
-2.64
HA -0.46 -0.51 -0.45
KA -1.76* -1.80* -1.76*
MP -2.39** -2.67 -2.35**
MH -2.93***
— _ „ * * *
-2.70 -2.58**
OS -1.08 -1.86* -0.98
PU -2.14** -2.19 -2.13**
RA -1.58 -1.62 -1.58
TN -2.31 -2.63 -2.27**
UP -0.74 -1.04 -0.69
WB -2.03** -1.76 -1.44
DE -1.19 -1.42 -1.15
Constant 5.50 5.61 5.49
Pr>m 0.6220 0.1456 0.7309
Hausman’s test supports random effects specification at 5% level of significance
DF 750 750 749
(t-statistics reported) *** Significant at 1% level of significance
** Significant at 5% level of significance * Significant at 10% level of significance
AP: Andhra Pradesh, AS: Assam, GU: Gujarat, HA:Haryana, KA: Karnataka, MP:
Madhya Pradesh, MH: Maharashtra, OS: Orissa, PU: Punjab, RA: Rajasthan, TN:
Tamil Nadu, UP: Uttar Pradesh, WB: West Bengal
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
APPENDIX - B
B .1: D e s c rip tiv e S t a t i s t i c s of SEBs over th e y e a rs
V aria b le N Mean Median Std Dev Minimum Maximum
p la v l 182 70.8345055 75.0000000 14.5560274 30.4000000 93.8000000
p lf 182 54.4923077 57.9500000 18.0787701 18.2000000 85.6000000
fo ro u t 182 17.7723077 13.0000000 12.1699120 1.6000000 58.2800000
emkwh 182 5.1222967 4.4000000 2.6981351 1.5500000 16.2000000
ethcon 182 16.6214835 14.0900000 9.4754661 5.1100000 55.2000000
c ro ss 181 2.1952048 2.0887513 0.7214866 0.9027982 5.0682062
srev 182 0.7312359 0.7137408 0.1462014 0.3274783 1.0597290
ucost 182 2.0465806 1.8961366 0.5909136 1.0371387 4.5024715
omc 181 0.0901928 0.0822254 0.0416897 0.0400000 0.3721750
f c o s t 181 0.4542767 0.4154966 0.2160183 0.0136000 1.3584028
powcost 181 0.7081792 0.6064639 0.5160922 0.1702509 5.4839611
avt 182 1.4396968 1.4237936 0.2674832 0.8180608 2.4525000
re s ty 182 0.1868132 0 0.6958772 0 5.0000000
r e s t r 182 0.0879121 0 0.2839482 0 1.0000000
pol 182 0.7362637 1.0000000 0.4418736 0 1.0000000
td 182 23.0759890 21.5000000 6.3642308 14.3100000 49.4000000
demand 182 23070.83 20629.00 14097.44 1679.00 78586.00
u r r a tio 182 0.2872783 0.2961399 0.0827466 0.1138886 0.4357709
to tp o p 182 581483.22 545461.50 327317.02 155131.00 1626948.00
lv to t 182 0.4484341 0.4575000 0.1351801 0.1290000 0.7010000
hydel 169 24.8837278 18.6000000 23.0458184 0 78.8000000
207
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
B .2: D e s c rip tiv e S t a t i s t i c s of A N D HRA PRADESH
V aria b le N Mean Median S td Dev Minimum Maximum
p la v l 13 84.5276923 85.3000000 5.1071031 77.4000000 90.9000000
p it 13 73.1000000 70.2000000 7.7349855 62.1000000 85.2000000
fo ro u t 13 7.4815385 6.4200000 3.2324986 3.0000000 12.7300000
emkwh 13 3.4746154 3.4000000 0.6522859 2.5400000 4.7000000
ethcon 13 8.4638462 8.6000000 1.7145045 5.7200000 11.0000000
c ro ss 13 2.1481318 2.1791861 0.1901848 1.8145315 2.4161220
srev 13 0.7794191 0.7295515 0.1714619 0.5151284 0.9971510
ucost 13 1.6498327 1.4214994 0.4377700 1.2266160 2.4535417
omc 13 0.0706967 0.0696233 0.0084758 0.0581000 0.0857639
f c o s t 13 0.4154547 0.3465371 0.1263924 0.2724014 0.5833815
powcost 13 0.6209792 0.4228433 0.4776218 0.2642586 1 .7779167
avt 13 1.2242692 1.2253289 0.1286540 0.9710000 1.4311927
r e s ty 13 0.4615385 0 0.9674179 0 3.0000000
r e s t r 13 0.2307692 0 0.4385290 0 1.0000000
pol 13 1.0000000 1.0000000 0 1.0000000 1.0000000
td 13 24.2176923 19.9000000 6.4187345 18.9000000 33.0900000
demand 13 29249.54 29182.00 8703.61 16247.00 42710.00
u r r a tio 13 0.2933998 0.2928792 0.0192522 0.2641738 0.3236520
to tp o p 13 671209.92 672820.00 38238.64 610100.00 727293.00
lv to t 13 0.5476154 0.5880000 0.0752103 0.4490000 0.6390000
hydel 12 37.4500000 38.5500000 12.0047339 22.0000000 52.8000000
208
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
B .3: D e s c rip tiv e S t a t i s t i c s of A SSA M
V aria b le N Mean Median Std Dev Minimum Maximum
p la v l 13 48.1415385 49.0000000 10.3218682 31.4000000 63.3000000
p lf 13 23.7230769 24.6000000 3.8886917 18.2000000 28.6000000
fo ro u t 13 37.8038462 35.4000000 13.3410379 10.5400000 58.2800000
emkwh 13 12.4123077 12.8000000 2.6919205 7.4100000 16.2000000
ethcon 13 35.2546154 32.7000000 11.2947249 21.4000000 55.2000000
cro ss 13 1.8660746 1.9017595 0.5020362 1.0464625 2.6865217
srev 13 0.4782564 0.4741008 0.1059146 0.3274783 0.6519957
ucost 13 3.7337200 3.7615278 0.4645112 3.0698663 4.5024715
omc 13 0.1429886 0.1209000 0.0627853 0.0686047 0.2627376
f c o s t 13 0.5006063 0.5220507 0.1722368 0.1512103 0.7168459
powcost 13 0.7379486 0.6614162 0.2756645 0.4417763 1.2482659
avt 13 1.7617549 1.6705323 0.3389442 1.3309249 2.4525000
re s ty 13 0 0 0 0 0
r e s t r 13 0 0 0 0 0
pol 13 0.6153846 1.0000000 0.5063697 0 1.0000000
td 13 26.4138462 24.9000000 6.5737553 20.5000000 38.1300000
demand 13 2960.31 2844.00 1065.79 1679.00 4710.00
u r r a tio 13 0.1229636 0.1230152 0.0057907 0.1138886 0.1317866
to tp o p 13 267232.08 267274.00 21924.28 233632.00 300431.00
lv to t 13 0.2402308 0.2700000 0.0835884 0.1360000 0.3220000
hydel 13 0 0 0 0 0
209
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
B .4: D e s c rip tiv e S t a t i s t i c s of BIHAR
V aria b le N Mean Median Std Dev Minimum Maximum
p la v l 13 42.2515385 41.8000000 8.6271894 30.4000000 59.1200000
p lf 13 25.1846154 24.6000000 5.1977313 19.0000000 37.1000000
fo ro u t 13 34.1592308 33.5400000 8.2501338 21.9000000 48.3000000
emkwh 13 6.0884615 6.2000000 1.6751161 3.3900000 8.5000000
ethcon 13 23.9138462 23.1000000 6.4363687 15.7400000 33.3000000
c ro ss 13 2.2314087 2.2634950 0.2934403 1.6374039 2.5184271
srev 13 0.6414587 0.6376905 0.0618618 0.5243534 0.7368158
uco st 13 2.5062595 2.5240000 0.1533347 2.2628593 2.7791118
omc 13 0.1383079 0.1432564 0.0339257 0.0804861 0.1838816
f c o s t 13 0.3227270 0.2438809 0.1668260 0.1717361 0.7072243
powcost 13 1.1380706 1.0590461 0.2824886 0.6653992 1.5871575
avt 13 1.6041541 1.6821293 0.1480327 1.3898611 1.7910085
re s ty 13 0 0 0 0 0
r e s t r 13 0 0 0 0 0
pol 13 1.0000000 1.0000000 0 1.0000000 1.0000000
td 13 22.6823077 22.8000000 2.0676805 19.0000000 25.9000000
demand 13 13061.15 12807.00 4009.08 8695.00 19894.00
u r r a tio 13 0.1685400 0.1677307 0.0155302 0.1458849 0.1937947
to tp o p 13 923881.46 922664.00 74449.43 811560.00 1039715.00
lv to t 13 0.2876154 0.3220000 0.0545138 0.2140000 0.3340000
hydel 12 7.6250000 6.1000000 3.9374946 4.0000000 15.5000000
210
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
B .5: D e s c rip tiv e S t a t i s t i c s o f GUJARAT
V ariab le N Mean Median Std Dev Minimum Maximum
p la v l 13 78.9515385 78.2000000 3.2697142 74.2700000 85.0000000
p lf 13 61.7538462 61.6000000 3.4779341 56.1000000 66.9000000
fo ro u t 13 9.4584615 7.9000000 2.3865102 6.6000000 12.9000000
emkwh 13 2.1746154 2.2000000 0.4896702 1.5500000 3.1000000
ethcon 13 8.2876923 8.1000000 0.8035873 7.4000000 9.5000000
cro ss 13 1.8914201 1.9320134 0.1560981 1.5936439 2.0971429
srev 13 0.7216203 0.7083825 0.0411657 0.6695703 0.7892023
ucost 13 1.9865122 1.9115607 0.1643673 1 .8110197 2.3304861
omc 13 0.0613022 0.0630650 0.0118608 0.0458092 0.0767341
fc o s t 13 0.8521324 0.8786108 0.1084586 0.7011561 1 .0161290
powcost 13 0.4588977 0.3229769 0.2817311 0.1774194 0.9802778
avt 13 1.4299756 1.4526616 0.0951419 1.2828947 1.5753425
re sty 13 0 0 0 0 0
r e s t r 13 0 0 0 0 0
pol 13 0.6153846 1.0000000 0.5063697 0 1.0000000
td 13 20.8346154 21.3000000 1.2198539 18.3000000 22.1000000
demand 13 31537.92 30609.00 9022.34 18396.00 46049.00
u r r a tio 13 0.3553847 0.3547500 0.0147307 0.3332141 0.3786389
totpop 13 426519.54 427450.00 22178.34 390706.00 459679.00
lv to t 13 0.4663846 0.4860000 0.0357527 0.4080000 0.5000000
hydel 12 4.1250000 4.0000000 1.6761021 1.0000000 6.5000000
V aria b le N Mean
p la v l 13 64.2746154
p lf 13 45.5769231
fo ro u t 13 27.7553846
emkwh 13 5.9400000
ethcon 13 14.8300000
c ro ss 13 1.9554954
srev 13 0.6135059
u cost 13 2.0519314
omc 13 0.0990022
f c o s t 13 0.6400128
powcost 13 0.8036631
avt 13 1.2599030
re s ty 13 0.4615385
r e s t r 13 0.2307692
pol 13 0.6153846
td 13 27.3576923
demand 13 14365.46
u r r a tio 13 0.3051720
to tp o p 13 172211.15
lv to t 13 0.6256154
hydel 12 29.2916667
D e s c rip tiv e S t a t i s t i c s of H A R Y A N A
Median S td Dev Minimum Maximum
63.6000000 4.5593176 56.3000000 70.3000000
45.8000000 4.9798182 34.6000000 53.0000000
25.1000000 9.6553833 16.3900000 55.4000000
5.8000000 0.7282056 4.8700000 7.1000000
15.7000000 3.2469627 8.5800000 18.4000000
1.9464286 0.3085540 1.5895000 2.5448635
0.6259364 0.0518772 0.5036276 0.7095553
2.0097205 0.3357820 1.6076046 2.5119863
0.1019692 0.0164503 0.0727390 0.1180000
0.4771863 0.3234521 0.3921720 1.3584028
0.7342649 0.4884202 0.2509506 1.7400694
1.2216097 0.2372876 0.9366925 1.6041096
0 0.9674179 0 3.0000000
0 0.4385290 0 1 .0000000
1.0000000 0.5063697 0 1.0000000
26.4000000 4.5835997 17.5000000 33.8100000
14572.00 5252.90 6878.00 22644.00
0.3024164 0.0342318 0.2564766 0.3621179
172918.00 10413.88 155131 .00 187092.00
0.6200000 0.0306853 0.5820000 0.6800000
28.1500000 23.9240639 5.0000000 54.0000000
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
B .7: D e s c rip tiv e S t a t i s t i c s of K A R N A T A K A
V ariab le N Mean Median Std Dev Minimum Maximum
p la v l 13 81.6576923 86.2000000 9.5067889 56.6000000 90.7000000
p lf 13 70.6230769 70.2000000 9.6769618 49.4000000 82.3000000
fo ro u t 13 7.7230769 6.4400000 6.1516480 2.8000000 26.5700000
emkwh 13 3.3030769 3.0000000 0.9696854 2.3100000 5.8000000
ethcon 13 7.2207692 7.0000000 2.0297104 5.1100000 11.0000000
cro ss 13 1.9638183 2.0537535 0.4283296 1.3857965 2.5003012
srev 13 0.8582798 0.8485254 0.0788781 0.7506008 0.9848577
ucost 13 1.5563295 1.5234000 0.2778205 1.2502584 2.2267361
omc 13 0.0740742 0.0726098 0.0167705 0.0437500 0.1171965
fc o s t 13 0.2758733 0.2411908 0.1150607 0.1433692 0.5009722
powcost 13 0.9596946 0.8973384 0.2875066 0.6862717 1.6679913
avt 13 1.3214920 1.2976914 0.1650502 1.1440000 1.7561111
re sty 13 0.2307692 0 0.5991447 0 2.0000000
r e s t r 13 0.1538462 0 0.3755338 0 1.0000000
pol 13 1.0000000 1.0000000 0 1.0000000 1.0000000
td 13 21.6661538 19.3000000 4.7773503 18.4100000 30.0000000
demand 13 20437.77 20398.00 5626.62 11983.00 29543.00
u r r a tio 13 0.3678101 0.3662427 0.0282387 0.3264253 0.4132980
to tp o p 13 474367.54 475705.00 27731.97 429698.00 515206.00
lv to t 13 0.5643846 0.6180000 0.1082101 0.4410000 0.7010000
hydel 12 67.4083333 69.4500000 9.4893681 50.0000000 78.8000000
K >
u >
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
B .8: D e s c rip tiv e S t a t i s t i c s of M A D H Y A PRA D ESH
V aria b le N Mean Median Std Dev Minimum Maximum
p la v l 13 74.7169231 76.4000000 3.7216739 68.2000000 79.0000000
p it 13 58.4769231 58.2000000 7.2293561 49.2000000 69.4000000
fo ro u t 13 12.7984615 11.7300000 3.0395719 9.0000000 19.3000000
emkwh 13 4.6369231 4.0000000 1.4764280 2.9500000 7.3000000
ethcon 13 14.0438462 12.7000000 3.1299615 11.0600000 18.5000000
c ro ss 13 4.1007150 3.7777778 0.7073934 3.3903148 5.0682062
srev 13 0.7529330 0.7674521 0.0901476 0.6121595 0.8777893
ucost 13 1.8583508 1.8846098 0.0870577 1.7114068 1.9833028
omc 13 0.0816851 0.0817740 0.0093749 0.0657111 0.0967742
f c o s t 13 0.3498781 0.3484871 0.0425624 0.2768467 0.4211470
powcost 13 0.5454330 0.5919806 0.1460158 0.2652330 0.7179861
avt 13 1.3956307 1.4288864 0.1471148 1.1556358 1.6190826
re s ty 13 0 0 0 0 0
r e s t r 13 0 0 0 0 0
pol 13 0.7692308 1.0000000 0.4385290 0 1.0000000
td 13 20.7607692 20.2000000 1.5044737 19.5000000 24.2000000
demand 13 26866.54 27018.00 8477.86 14131.00 39931.00
u r r a tio 13 0.2748099 0.2733023 0.0261527 0.2367371 0.3175346
to tp o p 13 677228.69 678859.00 44898.59 606255.00 742702.00
lv to t 13 0.4552308 0.4930000 0.1114541 0.2940000 0.5850000
hydel 12 10.2841667 11.1500000 3.0044133 5.5200000 14.6000000
214
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
B .9: D e s c rip tiv e S t a t i s t i c s of M A H A R A SH TR A
V ariab le N Mean Median Std Dev Minimum Maximum
p la v l 13 81 .1807692 82.9000000 4.7239540 72.1900000 86.4000000
p lf 13 63.9384615 64.1000000 5.7842514 53.5000000 72.7000000
fo ro u t 13 10.9700000 10.1900000 2.0473235 7.7900000 14.0700000
emkwh 13 3.0846154 3.0000000 0.7159797 2.1200000 4.3000000
ethcon 13 11.2123077 11.0000000 2.0845829 8.9600000 14.9000000
cro ss 12 2.3431219 2.3517829 0.1429638 2.1088825 2.5571207
srev 13 0.9333462 0.9596054 0.0640259 0.8464101 1.0597290
ucost 13 1.8062078 1.7992775 0.0739593 1.6760750 1.8977064
omc 13 0.1052095 0.1060837 0.0105580 0.0902457 0.1286184
f c o s t 13 0.5658131 0.5797945 0.0546237 0.4519444 0.6379928
powcost 13 0.4368507 0.3845376 0.1134527 0.3174905 0.6476389
avt 13 1 .6841167 1.6900000 0.1084102 1.4988593 1.8286756
re sty 13 0 0 0 0 0
r e s tr 13 0 0 0 0 0
pol 13 0.1538462 0 0.3755338 0 1.0000000
td 13 16.2661538 16.4000000 1.0420136 14.3100000 17.7200000
demand 13 52227.15 49535.00 14935.95 31883.00 78586.00
u r r a tio 13 0.4068982 0.4067988 0.0185186 0.3784401 0.4357709
totpop 13 787202.85 785916.00 44197.74 720907.00 856977.00
lv to t 13 0.4339231 0.4540000 0.0533361 0.3590000 0.4980000
hydel 12 10.6391667 11.0850000 3.8656587 6.0000000 16.7000000
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
B.10: D e s c rip tiv e S t a t i s t i c s of ORISSA
V aria b le N Mean Median Std Dev Minimum Maximum
p la v l 13 71.7130769 63.6000000 15.5205785 53.4000000 93.8000000
p lf 13 51.8923077 35.6000000 22.2099250 29.0000000 85.6000000
fo ro u t 13 17.9246154 19.0000000 12.2016253 1.6000000 33.7700000
emkwh 13 6.3513846 6.1700000 0.9859758 4.9400000 8.0000000
ethcon 13 29.1838462 24.6000000 8.7821766 18.8000000 44.0000000
c ro ss 13 1.7771656 1.9318581 0.3632773 1.2203704 2.3608903
srev 13 0.8761234 0.9070642 0.1024612 0.7125731 1.0181189
ucost 13 1.7852589 1.9836806 0.5484770 1.0371387 2.9596330
omc 12 0.1296813 0.0995163 0.0824090 0.0800380 0.3721750
fc o s t 12 0.1718137 0.1141256 0.1195515 0.0136000 0.4011927
powcost 12 1.1763807 0.7266000 1.4639660 0.1715461 5.4839611
avt 13 1.5501262 1.6515987 0.4527650 0.9407514 2.2178082
re s ty 13 1 .1538462 0 1.7722939 0 5.0000000
r e s t r 13 0.3846154 0 0.5063697 0 1.0000000
pol 13 0.9230769 1.0000000 0.2773501 0 1.0000000
td 13 34.3361538 24.0000000 12.3379405 22.5000000 49.4000000
demand 13 10409.23 9823.00 3968.37 5808.00 17859.00
u r r a tio 13 0.1842712 0.1827282 0.0243729 0.1490427 0.2242502
to tp o p 13 329597.46 330020.00 19635.39 298916.00 358869.00
lv to t 13 0.3375385 0.3890000 0.1768816 0.1290000 0.5550000
hydel 12 66.3475000 69.1850000 8.4352282 53.9000000 77.1000000
216
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
B.11: D e s c rip tiv e S t a t i s t i c s of PUNJAB
V ariab le N Mean Median Std Dev Minimum Maximum
p la v l 13 81.7769231 81.4000000 2.7778630 77.8000000 88.1000000
p lf 13 62.5230769 60.8000000 8.2734670 52.8000000 77.9000000
fo ro u t 13 9.8269231 10.2000000 2.6945698 4.7000000 13.6400000
emkwh 13 4.7238462 4.5000000 0.8188767 3.6600000 6.5000000
ethcon 13 17.3592308 16.9000000 1.0346534 15.7600000 19.5000000
c ro ss 13 1.3775430 1.3852903 0.2287825 0.9027982 1.6279461
srev 13 0.6204278 0.6561250 0.0933379 0.4372967 0.7259390
ucost 13 1.7483648 1.7832073 0.1235559 1.4273121 1.8707224
omc 13 0.0534123 0.0522814 0.0073824 0.0411184 0.0656944
f c o s t 13 0.5091612 0.5197133 0.0735456 0.3572254 0.6063183
powcost 13 0.2754112 0.2977083 0.0784200 0.1702509 0.4267979
avt 13 1.0836173 1.1936460 0.1766534 0.8180608 1.2653253
re s ty 13 0 0 0 0 0
r e s t r 13 0 0 0 0 0
pol 13 1.0000000 1.0000000 0 1.0000000 1.0000000
td 13 18.1492308 18.3000000 0.8312487 16.8100000 19.0000000
demand 13 21212.00 21151.00 5590.26 13127.00 30233.00
u r r a tio 13 0.3504940 0.3488044 0.0280206 0.3099727 0.3962905
to tp o p 13 204626.38 204715.00 8738.72 190869.00 218084.00
lv to t 13 0.5290769 0.5250000 0.0248175 0.4930000 0.5800000
hydel 12 36.3250000 35.7500000 13.6367901 18.0000000 54.8000000
217
V a ria b le N Mean
p la v l
p lf
f o ro u t
emkwh
ethcon
c ro ss
srev
u co st
omc
f c o s t
powcost
avt
re s ty
r e s t r
pol
td
demand
u r r a tio
to tp o p
lv to t
hydel
78.3969231
72.7846154
10.4930769
5.2253846
15.3030769
2.1807705
0.7092345
2.0382750
0.0769277
0.3727433
0.7489015
1.4305236
0.0769231
0.0769231
0.3846154
24.9723077
18423.85
0.2722691
480648.31
0.4530769
30.6333333
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
12
D e s c rip tiv e S t a t i s t i c s of RAJASTHAN
Median S td Dev Minimum Maximum
80.3000000 9.9432283 54.0000000 89.3000000
75.6000000 10.5319234 50.2000000 85.2000000
9.4000000 6.0657555 3.7000000 23.6300000
5.0000000 1.7818839 3.2300000 8.4000000
14.0000000 4.2703247 11.0700000 23.9000000
2.2115807 0.3130740 1.5773196 2.5741787
0.7057344 0.0761872 0.5809672 0.8216397
2.1317000 0.2369979 1.6374277 2.4174855
0.0790000 0.0113844 0.0592361 0.0978899
0.3671835 0.0570508 0.2395437 0.4411806
0.7763062 0.1553735 0.5197133 1.0038295
1.4044798 0.0793602 1.3453757 1.6086473
0 0.2773501 0 1.0000000
0 0.2773501 0 1.0000000
0 0.5063697 0 1.0000000
25.0000000 2.4870403 21.5600000 29.4600000
17927.00 6776.19 9331.00 29655.00
0.2709063 0.0226327 0.2395233 0.3094003
480299.00 44524.05 412948.00 549148.00
0.4600000 0.0291561 0.4030000 0.5000000
35.6500000 13.6549913 4.0000000 44.9000000
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
B.13: D e s c rip tiv e S t a t i s t i c s of TAMIL N A D U
V aria b le N Mean Median Std Dev Minimum Maximum
p la v l 13 77.0830769 77.7000000 5.6495890 65.8000000 84.5000000
p lf 13 67.4538462 68.1000000 5.9185605 55.7000000 76.1000000
fo ro u t 13 10.8930769 10.4000000 3.9232350 4.5000000 18.5100000
emkwh 13 4.4884615 3.9000000 1.2883054 3.0200000 6.6000000
ethcon 13 9.5769231 9.2000000 2.0628192 6.6700000 13.1000000
cro ss 13 2.0072800 2.0541173 0.2171757 1.6314779 2.2631579
srev 13 0.8609649 0.8662507 0.0966688 0.6775977 0.9880279
u co st 13 1.7383716 1.7334342 0.1207692 1.4932081 1.9141667
omc 13 0.0455190 0.0445736 0.0038120 0.0400000 0.0520913
fc o s t 13 0.6614098 0.6533627 0.0709470 0.5636561 0.7715596
powcost 13 0.4379664 0.4220532 0.0863520 0.3500917 0.6242361
avt 13 1.4921566 1.5108382 0.1567778 1.1688213 1.6663527
re s ty 13 0 0 0 0 0
r e s t r 13 0 0 0 0 0
pol 13 1.0000000 1.0000000 0 1.0000000 1.0000000
td 13 17.4215385 17.0000000 0.7530698 16.5000000 18.5000000
demand 13 27576.92 27860.00 6976.10 17268.00 38447.00
u r r a tio 13 0.3659919 0.3657599 0.0114053 0.3483694 0.3836495
to tp o p 13 584230.46 585088.00 26660.62 541775.00 624208.00
lv to t 13 0.4078462 0.3990000 0.0180409 0.3890000 0.4350000
hydel 12 24.7583333 24.1500000 4.8951661 17.7000000 33.2000000
219
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
B .14: D e s c rip tiv e S t a t i s t i c s o f UTTAR PRADESH
V aria b le N Mean Median Std Dev Minimum Maximum
p la v l 13 64.7207692 64.3000000 5.2562415 54.3000000 72.9000000
p lf 13 49.5923077 49.1000000 3.5005860 43.9000000 56.7000000
fo ro u t 13 25.0230769 25.6000000 4.1283681 15.8600000 32.1000000
emkwh 13 4.1407692 3.8000000 1.0825546 2.8900000 6.3000000
ethcon 13 17.7584615 16.6000000 6.0003512 10.1200000 29.1000000
cross 13 2.7087645 2.7822581 0.6875207 1.6030534 3.7475401
srev 13 0.6684756 0.6668898 0.0306614 0.6245104 0.7334097
uco st 13 1.9744077 1.9821705 0.1178258 1 .7270231 2.1687500
omc 13 0.1056439 0.1063356 0.0076748 0.0899669 0.1188213
f c o s t 13 0.3852576 0.3563584 0.0670746 0.3034682 0.5513308
powcost 13 0.5409879 0.5732659 0.1056817 0.2794677 0.6665857
avt 13 1.3190306 1.3151734 0.0868208 1.1517341 1.4688356
re s ty 13 0.2307692 0 0.5991447 0 2.0000000
r e s t r 13 0.1538462 0 0.3755338 0 1.0000000
pol 13 0.2307692 0 0.4385290 0 1.0000000
td 13 25.1330769 25.5000000 1.5858194 22.6000000 27.9600000
demand 13 37371.54 36561.00 11093.01 21828.00 56440.00
u r r a tio 13 0.2592715 0.2569510 0.0308591 0.2155525 0.3105864
to tp o p 13 1452073.62 1450311.00 112173.49 1282409.00 1626948.00
l v to t 13 0.5770769 0.5850000 0.0265501 0.5020000 0.5980000
hydel 12 23.8666667 22.7500000 2.3160834 21 .0000000 28.0000000
220
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
B.15: D e s c rip tiv e S t a t i s t i c s of W EST B EN G A L
V ariab le N Mean Median Std Dev Minimum Maximum
p la v l 13 62.2900000 64.7000000 6.7871889 52.3000000 71.2000000
p lf 13 36.2692308 36.2000000 3.7428564 30.8000000 41.2000000
fo ro u t 13 26.5015385 30.4000000 8.7051899 13.9000000 40.3700000
emkwh 13 5.6676923 5.2000000 1.9112699 3.5200000 8.5000000
ethcon 13 20.2923077 19.6000000 7.3344247 10.7200000 34.0000000
cro ss 13 2.1925357 2.1342806 0.5242345 1.2297064 2.9942775
srev 13 0.7232567 0.7209284 0.0524298 0.6119656 0.7962364
ucost 13 2.2183070 2.2720121 0.2141395 1.8840367 2.5853618
omc 13 0.0812861 0.0749306 0.0166396 0.0587329 0.1093190
f c o s t 13 0.3152633 0.2917313 0.0813783 0.2324000 0.4980989
powcost 13 1.0693387 0.9907514 0.1577630 0.8897338 1.3424306
avt 13 1.5990044 1.6138728 0.1388427 1.3903933 1 .8294677
re s ty 13 0 0 0 0 0
r e s tr 13 0 0 0 0 0
pol 13 1.0000000 1.0000000 0 1.0000000 1.0000000
td 13 22.8523077 22.0000000 2.7218687 20.0000000 28.0000000
demand 13 17292.23 17062.00 5659.63 10421.00 26481.00
u r r a tio 13 0.2946206 0.2938115 0.0111349 0.2785397 0.3128386
to tp o p 13 689735.62 691323.00 40238.08 625419.00 749330.00
lv to t 13 0.3524615 0.4190000 0.1100618 0.2130000 0.4550000
hydel 12 1.6916667 1.9000000 0.7525210 0.6000000 3.0000000
to
fo
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
B .16: D e s c rip tiv e S t a t i s t i c s of For th e Year 1988
V ariab le N Mean Median Std Dev Minimum Maximum
p la v l 14 66.6328571 69.5100000 10.7147823 39.3600000 80.5600000
p lf 14 49.6714286 51 .8500000 13.3540215 27.9000000 69.4000000
fo ro u t 14 21.0428571 17.3850000 14.4383397 3.7100000 58.2800000
emkwh 14 7.2285714 6.8000000 3.0582621 3.1000000 16.2000000
ethcon 14 23.5857143 19.0000000 13.2779665 9.4000000 55.0000000
cro ss 14 1.7827570 1.6364595 0.6386908 0.9027982 3.4166667
srev 14 0.7318023 0.7443853 0.1654650 0.3710256 0.9516429
ucost 14 1.9806898 1.7504753 0.7997492 1.2266160 4.5024715
omc 14 0.1010864 0.0908745 0.0528298 0.0520913 0.2627376
fc o s t 14 0.4733840 0.4876426 0.2225241 0.1121673 0.9144487
powcost 14 0.4524715 0.3697719 0.2421388 0.1730038 0.8973384
avt 14 1.3592884 1.3347909 0.2654908 0.8180608 1.8294677
re s ty 14 0 0 0 0 0
r e s t r 14 0 0 0 0 0
pol 14 1.0000000 1.0000000 0 1.0000000 1.0000000
td 14 20.3335714 20.7500000 2.8072971 14.3100000 26.5000000
demand 14 13420.00 12555.00 7577.95 1679.00 31883.00
u r r a tio 14 0.2568743 0.2603252 0.0803656 0.1138886 0.3784401
totpop 14 522166.07 485736.50 296439.23 155131.00 1282409.00
lv to t 14 0.3696429 0.4085000 0.1444766 0.1400000 0.5860000
hydel 14 25.4892857 20.9000000 24.8791245 0 73.5000000
222
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
B.17: D e sc rip tiv e S t a t i s t i c s of For th e Year 2000
V aria b le N Mean Median Std Dev Minimum Maximum
p la v l 14 74.0142857 80.5500000 18.9896675 34.2000000 93.8000000
p lf 14 62.3357143 69.8000000 23.1920280 18.6000000 85.2000000
fo ro u t 14 16.2357143 10.9500000 13.8326040 3.0000000 42.5000000
emkwh 14 3.5800000 3.1250000 1 .8526156 1.5500000 9.1300000
ethcon 14 11.3035714 10.4200000 5.0394301 5.1100000 21.4000000
c ro ss 13 2.1934301 2.1342806 0.6173915 1.4892196 3.4103331
srev 14 0.6892495 0.6612283 0.1212439 0.5151284 0.9242079
ucost 14 2.2752183 2.2746528 0.4969202 1.7164583 3.7615278
omc 14 0.0814831 0.0818750 0.0231956 0.0400000 0.1138889
f c o s t 14 0.4984425 0.4465625 0.2873499 0.1717361 1.3584028
powcost 14 1.0308929 0.9570139 0.4477640 0.2977083 1.7779167
avt 14 1.5475099 1.5274653 0.3331912 1.1579167 2.4525000
re s ty 14 1.1428571 0 1.6104057 0 5.0000000
r e s t r 14 0.4285714 0 0.5135526 0 1.0000000
pol 14 0.5714286 1.0000000 0.5135526 0 1.0000000
td 14 25.1392857 23.5000000 7.8880255 16.5000000 43.9900000
demand 14 34513.00 29944.00 18228.51 4710.00 78586.00
u r r a tio 14 0.3209721 0.3205933 0.0868110 0.1317866 0.4357709
to tp o p 14 639691.57 586678.00 376738.04 187092.00 1626948.00
lv to t 14 0.5004286 0.4930000 0.1068563 0.3100000 0.6760000
hydel 14 15.6428571 7.0000000 17.4691250 0 55.0000000
223
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
B.18: D e s c rip tiv e S t a t i s t i c s of Thermal Power P la n ts
V aria b le N Mean Median S td Dev Minimum Maximum
p la v l 766 73.9596475 79.7050000 17.6126013 0 100.0000000
p lf 766 56.1890992 59.9450000 20.3167188 0 94.9000000
fo ro u t 766 15.4727807 10.3550000 14.0151365 0 100.0000000
re s ty 767 0.0677966 0 0.3899431 0 5.0000000
r e s t r 767 0.0391134 0 0.1939913 0 1.0000000
B.19: D e s c rip tiv e S t a t i s t i c s f o r th e Year 1990
V aria b le N Mean Median S td Dev Minimum Maximum
p la v l 66 70.2346970 75.5650000 16.6288575 31 , .3500000 100.0000000
p lf 66 50.4183333 54.0700000 17.7134288 0 83.8500000
fo ro u t 66 17.3787879 13.1900000 14.3267100 0 57.7200000
B.20: D e s c rip tiv e S t a t i s t i c s f o r th e Year 2000
V aria b le N Mean Median S td Dev Minimum Maximum
p la v l 72 76.2111111 83.5500000 20.9128342 0 100.0000000
p lf 72 61.6097222 68.6500000 23.5739601 0 94.6000000
fo ro u t 72 16.4763889 9.6000000 18.1694488 0 100.0000000
224
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
B.21: D e sc rip tiv e S t a t i s t i c s of F e d e ra l P la n ts
V aria b le N Mean Median Std Dev Minimum Maximum
p la v l 148 78.1415541 82.8500000 14.0053492 36 .1800000 100.0000000
p lf 148 63.9511486 68.0050000 19.7467073 0 93.7000000
fo ro u t 148 14.5120946 8.7450000 13.8658754 0 63.8200000
re s ty 149 0 0 0 0 0
r e s t r 149 0 0 0 0 0
B.22: D e s c rip tiv e S t a t i s t i c s of SEB P la n ts
V aria b le N Mean Median Std Dev Minimum Maximum
p la v l 546 71.3513370 77.9650000 18.0057615 0 97.0000000
p lf 546 52.7741026 55.8500000 19.9933756 0 94.9000000
fo ro u t 546 17.0240293 11.8300000 14.2563444 0 100.0000000
re s ty 546 0.0952381 0 0.4594524 0 5.0000000
r e s t r 546 0.0549451 0 0.2280819 0 1.0000000
B.23: D e s c rip tiv e S t a t i s t i c s of P riv a te P la n ts
V aria b le N Mean Median Std Dev Minimum Maximum
p la v l 72 85.1431944 87.8050000 15.2309633 0 100.0000000
p lf 72 66.1308333 69.2500000 16.0069681 0 86.7000000
fo ro u t 72 5.6838889 3.6650000 6.3739113 0 30.1900000
re s ty 72 0 0 0 0 0
r e s t r 72 0 0 0 0 0
225
B.24: T -tests for means o f F ed and Seb
226
— . x— y— T—
C O C M
+ ■ > o o O o C O C O
-----
o o o o LO LO
o o o o o o
A Ll_ C O IO o
V V V V o o o o C O
£_ A o h - CD
CL o 0 0 C O
L_
Q_ o o o
C D LO o LO CD
3 CM CD o O CD 0 )
rH C D LO C O C O
cd xfr C O C O
1—
3 C O o o
> i I i—I
cd
•w >
Ll_
Li. CM CM CM LO C M CO
O O i 0 ) CD CO CD C O c n
C O C M C O CM C O C M C D LL h * h »
o a
c 7“ T — T —
< D
O
C
cd
cd
>
• o
o
1-1 1 —1 rH
cd cd cd
rH 3 r-H 3 rH 3
cd c r cd c r cd c r
3 C D 3 C D 3 C D
c r c c r c c r c
L L1 3 LU 3 LU 3
a) C D C D
•H ■ M •H
•H ■ H •H
cd cd cd
5 5 5
3 3 s z
■ H + J + ■ »
*D c_ T3 c_ u t -
a> C D C D C D C D < D
•H ■ M 1 “ 1 ■ H 1 —1 -H
o
4J
o ■ M o ■ H
o cd o cd o cd
CL C O Q _ CO 0_ CO
cd
>
H H 3 3
> > O O
cd C d 4- 4 — £- £-
r l r l r l H O O
Q- CL Q_ Q . 4 - M-
C d
>
O
c r
L U
c
0 )
a
10 LO LO
LO LO LO
u . L L li_
TJ ■a T J ■a
o 0) C D C D
3 T3 ■o *o
■H rH 1 — 1 rH
C D o o o
s LL L i. L i.
C D
rH
3 ■ H
cd rH 3
•H > O
C _ cd 4- 1_
cd r— i r — 1 o
> CL CL 4 -
C D
co
1 0
C M
O O O O O O
o o o o o o
o o o o o o
V V V V V V
C D
T-
CO LO CO CM
3 CM o CO h -
•H
cd
>
CO N. LO CO CO T”
■ M
Li. CO CO CO CO LO
Q T — 1 “ o 1 — 00
CO 0) CO 1 — CO 1 —
o rH rH i—1
c cd cd cd
cd rH 3 rH 3 rH 3
•H cd c r cd c r cd c r
L . 3 C D 3 C D 3 c d
cd c r c c r C c r c
> LU 3 LU 3 LU 3
C D C D C D
■P +- » +- »
•H •H •H
cd cd cd
5 5 5
3 3 3
■H + J ■H
T J * o £_ ■a C- ■a C _
o C D C D C D C D C D 0 )
3 rH +-» rH ■H rH •H
+J O •H O ■M O +- »
C D o cd O cd O cd
2 0_ CO a . CO CL C O
C D
rH
3 ■H ■H
cd rH rH 3 3
■H > > O O
£_ cd cd 4 - 4— c_ L_
cd rH rH rH rH o O
> CL Q . Q . Q . 4 - 4 -
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
E q u a lity of V arian ces
V a ria b le Method Num D F Den D F F V alue Pr > F
p la v l Folded F 545 71 1 .40 0.0803
p lf Folded F 545 71 1 .56 0.0211
fo ro u t Folded F 545 71 5.0 0 <.0001
B.26: T - t e s t s f o r means of P ri and Fed
T -T e sts
V aria b le Method V arian ces D F t V alue Pr > | t |
p la v l Pooled Equal 218 3.3 8 0.0009
p la v l S a tte r th w a ite Unequal 131 3 .2 8 0.0013
p lf Pooled Equal 218 0 .8 2 0.4159
p lf S a tte r th w a ite Unequal 170 0.8 8 0.3823
fo ro u t Pooled Equal 218 -5 .1 4 <.0001
fo ro u t S a tte r th w a ite Unequal 217 -6 .4 7 <.0001
E q u a lity of V arian ces
V aria b le Method Num D F Den D F F V alue Pr > F
p la v l Folded F 71 147 1.18 0.3952
p lf Folded F 147 71 1 .52 0.0487
fo ro u t Folded F 147 71 4 .7 3 <.0001
227

Asset Metadata

Creator Panda, Arun Kumar (author)

Core Title Restructuring and performance in India's electricity sector

Contributor Digitized by ProQuest (provenance)

Degree Doctor of Philosophy

Degree Program Political Economy and Public Policy

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag economics, general,Energy,OAI-PMH Harvest

Language English

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-542781

Unique identifier UC11339878

Identifier 3094364.pdf (filename),usctheses-c16-542781 (legacy record id)

Legacy Identifier 3094364.pdf

Dmrecord 542781

Document Type Dissertation

Rights Panda, Arun Kumar

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA