ECONOMIC INSTITUTIONS OF ELECTRICITY:
AN APPLICATION TO THE BRAZILIAN CASE
By
Guilherme Signorini

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Agriculture, Food and Resource Economics – Master of Science
2013

ABSTRACT
ECONOMIC INSTITUTIONS OF ELECTRICITY:
AN APPLICATION TO THE BRAZILIAN CASE
By
Guilherme Signorini

Two branches of the New Institutional Economics (NIE) literature, Institutional
Environment and Transaction Cost Economics (TCE), lack several causal explanations. Namely,
the former has not fully explained why and how substantial institutional change occurs; and the
latter fails to provide formal reasons for why firms may adopt “second-best” governance
strategies to support exchange in a certain industry. In this thesis, the Brazilian electricity sector
is used as a case study to analyze how institutions of electricity evolved from a state-dominated
orientation to a competitive market-oriented organization. Sequentially, the adequacy of the
classic TCE framework in explaining how economic agents adopt governance strategies is
evaluated. Based on an extensive literature review and structured personal interviews with sector
experts, results indicate that institutional changes were driven by ‘high levels of uncertainty’,
‘economic recession’, and the ‘unreliability of utility services’. Moreover, results support the
existence of gaps in the classic TCE framework and recognize necessary refinements in two
additional concepts – complementarity and implementability). In light of those concepts, this
thesis concludes that (i) some transactions conducted in the regulated market channel would be
governed differently if parties were authorized by the institutional framework to choose their
strategies; (ii) free market channel players choose governance strategies that are not always
aligned with the TCE predicted governance mechanism because capital availability, existence of
compatible partners, and reputation play crucial roles in determining strategic decisions.

ACKNOWLEDGEMENTS

I have many people to thank for their contribution in this thesis. I would first like to thank
my committee; namely Dr. Christopher Peterson, Dr. Brent Ross, and Dr. Soren Anderson.
Special thanks to Dr. Alvaro Augusto W. de Almeida from Parana Tech Federal University, Dr.
Marco Antonio Conejero from PwC Consulting, Dr. Marcos Fava Neves from FEARP/USP, Dr.
Hamish Gow from Massey University, Dr. Luiz Henrique Alves Pazzini from CCEE, and Dr.
Leontina M.V.G. Pinto from Engenho Consulting for their insights and inspiration.
Thanks to Celso Zanatto Jr. from Volga Energy, Cesar Sá from Odebrecht, Elias Torres
from Bunge Co., Fabio Pelegrini from Guarani Co., Felipe Bazzo from São Francisco Co., José
Roberto Ferreira from Sermatec, Juliana Vilela from Santelisa Co., Luiz Felipe Pellegrini from
ERB, Matheus Biagi Carvalho from Pedra Agroindustrial, Simone Prado from ONS, Suleiman
José Hassuani from CTC, Tulio Soubhia Ribeiro from Cerradinho Group, and Dr. Zilmar José de
Souza from UNICA for the interviews.
I have made many friends thus far at Michigan State and I want to thank them all for their
friendship. Special thanks to Byron Reyes, Jacob Ricker-Gilbert, and Nicole Mason who have
inspired me to keep up with the hard work. Thanks to Rachel Ford for her editorial assistance.
Thank you to all my Brazilian friends who have also supported me psychologically. Special
thanks to my frat brothers at F.M.R. Kangaço.
A big thank you must be extended to my parents and family for always being by my side
in every step of my life and for providing strength and perseverance in pursuing my goals.
Special thanks goes to my fiancé, Marília, for her constant support, and for keeping my life
balanced between work and leisure. I am forever grateful for her support and friendship.

iii

PREFACE

“O saber a gente aprende com os mestres e os livros.
A sabedoria, se aprende é com a vida e com os humildes.”

“Knowledge we learn from mentors and books.
Wisdom must be learnt from life and through humbleness.”

Ana Lins dos Guimarães Peixoto (1889 – 1985)
Pseudonym, Cora Coralina

iv

TABLE OF CONTENTS

LIST OF TABLES ........................................................................................................................ vii
LIST OF FIGURES ..................................................................................................................... viii
INTRODUCTION .......................................................................................................................... 1
1. ANALYTICAL METHODOLOGY AND DATA ..................................................................... 4
2. THEORETICAL FRAMEWORK .............................................................................................. 7
2.1. The Early Work on Institutions.................................................................................... 7
2.2. The New Institutional Economics (NIE) ................................................................... 11
2.2.1. Institutional Environment ........................................................................... 13
2.2.2. Transaction Cost Economics....................................................................... 16
2.3. NIE Applied to Electricity Sectors ............................................................................ 25
3. INSTITUTIONS AND INSTITUTIONAL EVOLUTION OF THE ELECTRICITY SECTOR
IN BRAZIL ................................................................................................................................... 29
3.1. Background ................................................................................................................ 31
3.2. National Privatization Program (1995-2001)............................................................. 34
3.3. Electricity Sector Reform (1995-1998) ..................................................................... 36
3.4. Later Institutional Changes ........................................................................................ 39
3.5. Organization of the Dual-Channel Wholesale Market of Power ............................... 45
3.5.1. Regulated Market Channel ......................................................................... 45
3.5.2. Free Market Channel................................................................................... 48
3.5.3. Imbalance Market ....................................................................................... 49
3.6. Implications of the Institutional Changes .................................................................. 50
3.7. Conclusions ................................................................................................................ 53
4. ATTRIBUTES OF TRANSACTION, GOVERNANCE STRATEGIES, AND MARKET
PERFORMANCE ......................................................................................................................... 55
4.1. Wholesale Transactions in the Regulated Market Channel ....................................... 56
4.2. Wholesale Transactions in the Free Market Channel ................................................ 71
4.3. Mismatch between predicted and observed governance strategies in the regulated
market channel .................................................................................................................. 80
4.4. Mismatch between predicted and observed governance strategies in the free market
channel .............................................................................................................................. 83
4.5. Conclusions ................................................................................................................ 86
5. CONCLUDING REMARKS .................................................................................................... 88

v

REFERENCES ............................................................................................................................. 90

vi

LIST OF TABLES

Table 1: Categories of Economic Agents in Brazil’s Electricity Sector ....................................... 40
Table 2: Laws and Amendments Related to the Formal Institutions of Electricity in Brazil ....... 44
Table 3: Summary of Auctions in the Regulated Market Channel ............................................... 46
Table 4: Summary of Regulated Market Channel Transactions ................................................... 70
Table 5: Summary of Free Market Channel Transactions ............................................................ 79

vii

LIST OF FIGURES

Figure 1: Efficient Governance According to Williamson (1985) ............................................... 22
Figure 2: Government’s Participation in the Electricity Sector .................................................... 34
Figure 3: Summary of Conditions and Driving Factors of Institutional Change .......................... 39
Figure 4: The Dual-Channel Market of Power ............................................................................. 42

viii

INTRODUCTION

The New Institutional Economics (NIE) literature has not yet provided causal
explanations on how long institutions persist or why and how they suffer dramatic changes
(Arrow, 1999; Ménard & Shirley, 2005). Empirical studies have confirmed that institutional
change takes place at the margin of the institutional framework (North, 1990) and often suffices
to explain economic growth and social progress. These results, however, provide little support to
cases of substantial institutional change. According to Shirley (2005), evidence is still
inconclusive to define a theoretical justification on how dramatic institutional change occurs.
Alongside, the classical notion of transaction cost economics (TCE) has shown some
inability to explain why businesses in a certain industry adopt governance strategies that differ
from the most efficient one to support exchange (Delmas & Tokat, 2005). Studies have yet
questioned if TCE is the most indicated theory to predict optimal governance structures
(Mahoney, 1992; Foss, 1996; Jongwook & Mahoney, 2005). In light of these claimed criticisms,
this thesis uses Brazil’s electricity sector as a case study with the aim to expand the body of
empirical research in NIE and produce satisfactory explanations to motivate theory refinement.
The Brazilian electricity sector provides a particularly interesting case for three reasons:
(i) in less than a century, the sector experienced two dramatic institutional changes – the first
toward centralization and the second toward a market economy (Baer, 2008; Leite, 2009); (ii) the
latter reformulation of institutions along with financial support has created incentives for
production and consumption of renewable electricity (Almeida, 2005; Castro et al., 2008; Neves
& Conejero, 2009), which can be analyzed as economic growth and progress in environmental
protection; and (iii) in order to satisfy the new regulatory framework, wholesale market players

1

have adapted their strategies without being authorized to choose across the entire governance
continuum.
Put differently, the sector characterizes a rich setting for reviewing why and how formal
institutions of electricity were fundamentally redefined twice. Under the new set of institutions,
Brazil’s electricity sector makes a particularly appealing case study for analyzing how firms
choose governance strategies to support exchange; and whether the classic TCE theory
(Williamson, 1985; Peterson et al., 2001) suffices to predict those choices.
Plausible explanations for those why and how questions require details that go beyond the
frontiers of economics and enter the space of political science and strategic management.
According to Carroll et al. (1999), no other theory provides a better framework to address
‘interdisciplinary questions’ than the NIE. Furthermore, NIE not only recognizes the existence of
path dependence (North, 1990; Aron, 2000; Jütting, 2003) between institutional evolution and
definition of business strategies but also outlines a method for discussing complex questions
(such as those related to institutional change and economic agent’s strategic choices) based on
qualitative analyses rather than a quantitative approach (Shirley, 2005). As Mary Shirley states,
‘few empirical studies have attempted to grapple with the messy details of real institutional
change’ (Shirley, 2005 p. 631).
This thesis employs NIE theory and a case study research methodology (Yin, 2009) to
analyze the following questions. (i) What historical facts led the government, in the first
moment, to take control over power utilities; and in a second moment, to restructure the
institutional environment with incentive-driven policies? The objectives here are to provide
compelling empirical evidence for why dramatic institutional changes occur and how these
changes influence development. (ii) Do current institutions guide wholesale market players to

2

adopt the most efficient governance strategy? (iii) Does the classical notion of TCE (Williamson,
1985; Peterson et al., 2001) suffice to predict the governance strategy chosen by trading parties?
Our discussion suggests a set of refinements that might enlighten the path towards further
understanding of exchange relationships.
This thesis is organized as follows: Chapter 1 introduces the analytical methodology and
data used in this study. Chapter 2 reviews the NIE theory and relevant empirical literature related
to electricity markets. Chapter 3 tackles the first research question posed above and analyzes the
relationship between institutions and economic development in Brazil’s electricity sector over
time. Chapter 4 utilizes TCE theory to analyze the second research question and to discuss what
reasons cause misalignments between predicted and observed governance strategies within
Brazil’s wholesale electricity market. Chapter 5 offers concluding remarks, identifies
contributions, and proposes direction for future research.

3

1. ANALYTICAL METHODOLOGY AND DATA

This thesis uses case study (Yin, 2009) as research methodology to evaluate the
institutions of electricity and their implications to wholesale market players. Results of this study
are expected to enhance the body of empirical knowledge on institutional evolution (i.e. what
causes institutional reform; why it is necessary; how it is implemented) and whether the classical
notion of TCE suffices to explain adoption of governance strategies.
Case study corresponds to a research methodology that provides guidance for rigorous
data collection, presentation and analysis. As Yin (2009) suggests, case study fits best for
qualitative analysis if: (i) research focuses on open-ended questions, (ii) researcher is interested
in contemporary context, and (iii) investigator has no control over the set of events analyzed. The
author adds that case studies are appropriate for situations in which multiple sources of evidence
and prior theoretical propositions are considered to guide data collection and analysis.
This thesis matches the conditions mentioned above. From an outside perspective, it
reviews why and how Brazil implemented institutional reforms in its electricity sector.
Sequentially, it relies on the classic TCE framework (Williamson, 1985; Peterson et al., 2001) to
examine how economic agents playing in the Brazilian wholesale market of power select
coordination arrangements. Following the inductive approach (Corbin & Strauss, 1990; Strauss
& Corbin, 1994), the empirical analyses indicate that refinements of the TCE theory might
enhance our ability to predict governance choices.
Case studies can be powerful when they are analytical narratives that follow
methodological rigor and also describe all rich nuances of the institutional setting (Bates et al.,
1998). A plausible counterargument against alternative analytical methodologies is that cases are

4

capable of providing causal explanations to entangled real-life occurrences; whereas
experimental strategies (e.g. econometric methods) need to rely on the notion of ceteris paribus
(Wooldridge, 2010 p.3). Put differently, case study is appropriate as this thesis aims to maintain
the holistic and meaningful characteristics of real-life events without controlling related and
interesting variables. Quantitative methods accept that several factors may be correlated with the
variable of interest but in order to establish causal relationships, the investigator conversely
imposes control to keep the effects of related variables constant.
Data collection was based on two sources: (i) extensive literature review; and (ii) key
informant interviews. Two authoritative studies (Baer, 2008; Leite, 2009) provide the
foundations of the historical background. For more recent years, additional sources were
included in the literature review as they provide important descriptions to the sequence of facts
analyzed (Feldman, 1997; 1998; Ferreira, 2000; Pinheiro & Giambiagi, 2000). The current
regulation governing transactions between utilities of generation and utilities of distribution and
between utilities and independent consumers was collected from publications and websites of
related political bodies (i.e. Ministry of Mines and Energy, EPE, ANEEL, CCEE, and ONS).
Primary data was also collected from a series of structured interviews conducted between
th

th

May 25 and July 16 of 2010 with sector experts who represent a range of interests:
government, industry, and academia. In total, we interviewed fifteen people: seven decision
makers at utilities; five members of consulting companies/market facilitators; one project
manager from a design-building company; one member of CCEE (coordinating agency of the
wholesale market of power, in Portuguese); and one member of ONS (National System
Operator). The data collected in interviews are used as supporting arguments throughout the
thesis.
5

Following the classic TCE framework (Williamson, 1985; Peterson et al., 2001), the
underlying attributes of transaction (i.e. uncertainty, complexity, asset-specificity, and
frequency) are quantitatively analyzed for representative transactions between power generating
firms and wholesale market buyers. The attributes are then categorized discretely in order to
permit comparative assessment of governance strategies. Sequentially, the analysis examines
whether governance strategies predicted by TCE truly maps the strategy chosen by economic
agents. Misalignments between predicted and adopted strategies suggest that TCE theory
becomes more compelling when refining studies are taken under consideration. Three
complementary studies are used as building blocks to connect predicted strategies and strategies
observed in the marketplace.
Preliminary results of this thesis were presented and debated at the Graduate Research
rd

Symposium on March 23 , 2012 in East Lansing, MI; and at the International Society for New
st

Institutional Economics (ISNIE) Conference on June 21 , 2013 in Florence, Italy.

6

2. THEORETICAL FRAMEWORK

“Every student of (political) economy repeats in his own mind the historical evolution
of the schools, and a study of the history of economic theory is not an academic curiosity –
it is a recapitulation of the evolution of our own thinking.”
- John Roger Commons (1934, p.260)

“The reader must do his own systematizing of the rich material put before him; but the
result will amply reward the effort.”
- Wesley Clair Mitchell (1935, p.643)

2.1. The Early Work on Institutions

The study of institutional environment, institutional change, and transaction cost
economics integrate the theoretical framework used in the following chapters to assess the
evolution of the Brazilian electricity sector and to analyze transactions in the wholesale market
of power. No different than any other scientific field, these theories were gradually defined with
chronological contributions. Human knowledge evolves as one thinker proposes a theory, or
discusses a point of view, based on the knowledge put before him. Ostrom (1986 p.4) points that
no scientific field can advance far if participants do not share a common understanding of key
terms in their field: ‘we cannot communicate effectively if signs used by one scholar in a field
have different referents than the same sign used by another scholar in the same field’.

7

By 1945 institutionalism was fractured and diverse, and lacking consensus on its
methodological and theoretical foundations (Hodgson, 2004). There have been three historic
meanings to institutionalism. Firstly, it has been a protest movement against both (i) the market
economy as it has been institutionalized in the twentieth century in the United States and (ii) the
dominant school of neoclassic economics. The second historic meaning of institutionalism has
been problem-solving. Institutionalists have typically followed the path of pragmatism seeking to
accommodate parties with conflicting interests and therefore causing controversy. Thirdly,
institutional economics, no less than other schools of economic thought has attempted to create
its body of knowledge (Samuels, 1988).
Much emphasis has been given to Ronald Coase as the founder of the ‘New Institutional
Economics (NIE)’. Coase (1998) admits that the NIE started with his article “The Nature of the
Firm” (1937). Although his contributions enlightened both research and public policy in the
years to come, many theorists had worked on defining economic terminology largely employed
in his work. The NIE was coined by Oliver Williamson to differentiate the subject from the “old
institutional economics”. For Williamson (1998), what distinguishes the NIE from the earlier
work on institutions is that institutions are susceptible to analysis.
The study of institutions appeared in the 1890s with Thorstein Veblen whose work has
been neglected partially because of ‘the latterly acquired uneasiness among social scientists
about importing ideas from biology’ (Hodgson, 2004). As clearly stated in The Evolution of
Institutional Economics, Hodgson attributes the foundation of institutional economics to Veblen.
‘His (Veblen’s) preliminary analysis on institutional evolution in the Leisure Class was a major
achievement, standing significantly above his precursors’ (Hodgson, 2004). ‘What did mark off

8

Veblen’s work from that of his predecessors was concentration upon the evolution of
institutions’ (Mitchell, 1935 p.650).
Veblen used principles of Charles Darwin to suggest explanations on institutions and
institutional evolution. He was the first economist to present institutional evolution in terms of
natural selection, and his conception of human nature derived from Charles Darwin and William
James. ‘Institutions are not only themselves the result of a selective and adaptive process which
shapes the prevailing or dominant types of spiritual attitude and aptitudes; they are at the same
time special methods of life and human relations, and are therefore in their turn efficient factors
of selection. The changing institutions in their turn make for a further selection of individuals
endowed with the fittest temperament, and a further adaptation of individual temperament and
habits to the changing environment through the formation of new institutions’ (Veblen, 1899
cited in Hodgson, 2004).
Diversely of Veblen, John Commons’ work derived from Jeremy Bentham. Commons
(1934) attempted to develop a theory of institutional economics as derived from the decisions of
the Supreme Court of the United States. ‘Institutional economics is the field of the public interest
in private ownership, which shows itself behavioristically in buying and selling, borrowing and
lending, hiring and firing, leasing and renting. (…) Transactions are the alienation and
acquisition, between individuals, of the rights of future ownership of physical things, as
determined by the collective working rules of society’ (Commons, 1934). In the article
Institutional Economics, Commons advanced that the ‘transaction is properly regarded as the
basic unit of economic analysis’ (Williamson, 1985).
A vast discussion on related economical concepts was engaged by Frank Knight (1921)
Herbert Simon (1978), Kenneth Arrow (1963), Elinor Ostrom (1986) and other theorists. Frank

9

Knight first discussed the concept of uncertainty as we employ in transaction cost economics. In
his book Risk, Uncertainty and Profit (1921), Knight remarks the importance of the concept of
uncertainty to economics: ‘we live only by knowing something about the future; while the
problems of life arise from the fact that we know so little. This is as true of business as of other
spheres of activity. (…) If we are to understand the workings of the economic system we must
examine the meaning and significance of uncertainty’ (Knight, 1921).
Herbert Simon worked on defining bounded rationality. ‘We must give an account not
only of substantive rationality – the extent to which appropriate courses of action are chosen –
but also procedural rationality – the effectiveness, in light of human cognitive powers and
limitations, of the procedures used for choosing actions’. (…) The scarce resource is
computational capacity – the mind. The ability of man to solve complex problems, and the
magnitude of the resources that have to be allocated to solving them, depend on the efficiency
with which this resource, mind, is deployed’ (Simon, 1978).
Kenneth Arrow (1963) pioneering research on asymmetric information was crucial to the
development of the NIE; more specifically, to the study of governance structure, enforcement
mechanisms, monitoring methods, and incentives. Elinor Ostrom (1986) defined rules as the
means by which we intervene to change the structure of incentives in situations. ‘The term rules
should not be equated with formal laws. Enforcement is necessary for a law to become a rule. In
this rule-structured situation, individuals select actions from a set of allowable actions in light of
the full set of incentives existing in the situation. Rules affect the structure of a situation in which
actions are selected’.
Taking the past contributions on institutionalism and definitions of related economics
concepts under consideration, North (1990) settled the lack of consensus about institutions:

10

‘Institutions are the rules of the game in a society. Or more formally, institutions are the humanly
devised constraints that shape human interaction. Institutions affect performance of the economy
by their effect on the cost of exchange and production’. Other definitions of institutions were
1

given throughout history .

2.2. The New Institutional Economics (NIE)

With the definition of institutions in place, the NIE as a theoretical framework can be
examined in detail. Williamson (2000) considers four levels of social analysis: (i) social
embeddedness, (ii) institutional environment, (iii) governance, and (iv) resource allocation and
employment. In his point of view, the NIE is concerned principally with the second and third
levels. Transaction cost economics (TCE) represents the branch of NIE that is predominantly
concerned with governance arrangements. This branch has its origins in Ronald Coase’s
treatment of firms and markets in 1937 and received valuable contributions from Oliver
Williamson, Benjamin Klein, Victor Goldberg, and others. The other branch holds studies of
institutional environment, which has Douglass North as the prominent thinker and received
insights from Thráinn Eggertsson, Mary Shirley, and others.
Williamson (1985) recognizes that the governance framework takes the institutional
environment, at least in the first moment, as given. In concordance, Coase (1998) recognizes that
the costs of exchange depend on the institutions of a country: its legal system, its political
system, its social system, its educational system, its culture, and so on. The institutions of

1

The intention here is to focus on the definition that has largely been accepted by scholars rather
than to neglect researchers’ efforts in developing theory.
11

governance operate at the level of individual transactions, whereas the institutional environment
is more concerned with composite levels of activity.
The study of institutions maintains that there is a crucial distinction between institutions
and organizations. Organizations include political bodies (political parties, the Senate, a city
council, a regulatory agency), economic bodies (firms, trade unions, family farms, cooperatives),
social bodies (schools, universities, vocational training centers). In other words, organizations
are groups of individuals bound by some common purpose to achieve objectives. They are
‘created with purposive intent in consequence of the opportunity set resulting from the existing
set of constrains (formal or informal ones)’ (North, 1990).
Coase (1937) mentions that firms and markets are substitute governance mechanisms to
perform the very same functions. ‘Whether transactions are organized within the firm
(hierarchically) or between autonomous firms (across a market) is therefore a decision variable
for managers and policy makers’ (Williamson, 1985 p.4). Decades later Klein et al. (1978) argue
that the distinction between transactions coordinated within a firm or through spot markets goes
beyond a simple substitution of coordination arrangements as stated in Coase’s work.
Nevertheless, institutional environment defines the sphere of study on formal constraints,
informal constrains and third party enforcers. It creates a framework that leads us towards a
better understanding of institutional evolution. In our case of interest, institutional environment
provides theoretical background for assessing the ‘rules of the game’ in the wholesale market of
power in Brazil over time. In a complementary fashion, TCE forms the theoretical core for
analyzing the underlying attributes of transaction of electricity and how players strategically
choose governance arrangements.

12

It is noteworthy to observe that these two spheres of knowledge are path-dependent
branches of the NIE theory, and therefore must be addressed with equal weights when analyzing
institutional reforms and modes of governance. While institutions of a society fundamentally
influence both why organizations come into existence and how organizations strategically
choose governances; organizations determine how the institutional framework evolves at its
margin (North, 1990; Mahoney, 2005). Ultimately, these two spheres integrate the theoretical
framework necessary for conducting the analysis of interest. The next sections review both
institutional environment and transaction cost economics in more detail.

2.2.1. Institutional Environment

Three dimensions integrate the concept of institutional environment: informal constraints,
formal constraints, and the effectiveness of enforcement (North, 1990). Institutional constraints
include both what individuals are prohibited from doing and, sometimes, under what conditions
individuals are permitted to undertake certain activities. They consist of formal written rules as
well as typically unwritten codes of conduct that underlie and supplement formal rules. As rules
and informal codes are sometimes violated, an essential part of the functioning of institutions is
the costliness of ascertaining violations and the severity of punishment.
Informal constraints come from socially transmitted information and are a part of the
heritage that we call culture. Culture can be defined as the “transmission from one generation to
the next, via teaching and imitation, of knowledge, values, and other factors that influence
behavior” (Boyd & Richerson, 1985). In other words, culture defines the way individuals process
and utilize information and hence may affect the way informal constraints are specified.

13

Conventions and norms are culture specific. Honesty, integrity, trust, and living up to a
reputation can be self-enforced in many societies via culturally informal constraints. Informal
constraints embodied in customs, traditions, and codes of conduct are much more impervious
than formal rules in any type of society.
Formal constraints include political and judicial law, economic regulations and contracts.
The hierarchy of such restrictions, from constitutions, to statue and common laws, to specific
bylaws, and finally to specific contracts defines constraints, from general to particular
specifications. Political laws broadly define the hierarchical structure of the polity. Economic
regulations define property rights. Contracts contain the provisions specific to a particular
agreement in exchange. North (1990) adds that looking only at the formal type of constraints
themselves gives us an inadequate and misleading notion about the relationship between formal
constraints and performance of economies. Research on institutional environment also requires a
great deal of knowledge on informal constraints such as norms of conduct and beliefs. With a
sociological perspective, Granovetter (1985) agrees that a detailed analysis of social structure is
essential to understanding how existing institutions arrived at their current state.
The legal system represents the usual enforcement of formal constraints. In specific
markets, third party enforcers are sometimes established to solve contract breaches between
business parties. As transactions become more complex, there are strong incentives to create
private controlling systems (such as federal regulatory agencies) and minimize reliance on
judicial bodies. The argument is that legal systems may face difficulties in distinguishing
promised behavior from opportunistic behavior in complex and specific transactions (Joskow,
1985). As third party enforcers focus on a limited number of transactions, controlling measures
and punishment on deviators tend to be more effective. The classic case studies of Ellickson

14

(1989) and Greif (1993) illustrate the latter argument. The authors show that enforcing bodies
established by market participants may enforce norms of conduct more efficiently than
2

centralized political bodies .
In other situations agreements between parties are self-enforcing and neither the legal
system nor third parties are necessary. It often happens when the parties to exchange hold
personal relationship and are involved in repeated transactions (North, 1990). Less likely,
impersonal transactions may also be self-enforcing (even when information asymmetries are
present) if contracts create incentives for both parties to live up to the terms of exchange.
Nevertheless, impersonal agreements in the modern society often require enforcement,
which in turn is typically imperfect. The main reasons are due to potential uneconomical
monitoring costs, bonding costs, and residual losses (Jensen & Meckling, 1976); and to the fact
that enforcement is undertaken by agents whose own utility functions influence outcomes
(North, 2005).
The literature on institutional environment has lately paid great attention to understanding
how institutions change and how they cause economic growth. Successful institutional changes
are interpreted by theorists as those able to foster sustained growth and social progress in
countries or specific industries within economies. It is claimed that economic development
requires: (i) institutions that support exchange by lowering transaction costs and encouraging
trust, and (ii) institutions able to influence the state to protect private property and persons rather
than expropriate and subjugate them (Shirley, 2005).

2

Robert Ellickson looked at transactions and norms of the whaling industry (especially between
1750 and 1870). Greif evaluated the maghribi traders’ coalition and its influence over
transactions in the 11th century.
15

North (1990) explains that numerous but small changes constantly occur at the margin of
institutions without affecting culture or the set of beliefs intrinsic to an economic sector or
country as a whole. Despite the great acceptance on North’s view about marginal adjustment,
theorists have not yet explained how dramatic changes occur. North (2005) argues that radical
institutional changes (resulting from revolution, invasion or crisis) take place only if norms of
conduct and beliefs of the parties involved also change. For this reason, Shirley (2005) agrees
with North when she affirms that successful institutional adaptations are likely led by insiders
(i.e. those who have great deal of information about how institutions currently function) as they
understand local norms of conduct and know how to navigate social relationships.
Numerous case studies have attempted to provide plausible explanations to what factors
drive institutional reforms that in turn set the arena for economic progress: adoption of the
common law rather than the civil law (La Porta et al. 1997, 1998, 1999), occurrence of political
conflicts and wars (Herbst, 2000; Bates, 2001; Nugent & Robinson, 2002), cultural influence and
radical change of beliefs (Greif, 1994; Jütting, 2003), and adoption of democracy over other
political arrangements (Bardhan, 2000; Rodrik, 2000; Keefer, 2002). However, a general
explanation about what drives successful institutional reforms and consequent economic growth
is still missing. Shirley (2005) adds that more research on institutional evolution in developing
countries is necessary.

2.2.2. Transaction Cost Economics

Transaction costs include the following: procurement costs, costs of negotiating and
writing contracts; costs of safeguarding an agreement; costs of monitoring contractual

16

performance; costs of enforcing contractual promises; adaptations and haggling costs; bonding
costs; and costs associated with maladaptation. Such costs are different from production costs
and ‘are equivalent of friction in physical systems’ (Williamson, 1985 p.19).
Taking the institutional environment as fixed, at least in the first moment, firms decide
among a wide array of governance arrangements to minimize the sum of production and
transaction costs and safeguard relationship-specific investments against expropriation of quasirents (Klein et al., 1978; Williamson, 1996). At one extreme, spot market is the most efficient
governance structure for minimizing production costs. At the other, vertical integration is
superior for minimizing transaction costs. In between, there is a continuum of potential ‘hybrid’
strategies that firms can adopt (i.e. specification contract, relation-based alliance, and equitybased alliance) (Peterson et al. 2001). Carroll et al. (1999 p.73) comments that the study of
governance is the largest contribution of Oliver Williamson to the field of strategic management.
Williamson (1991; 1996) argues that four attributes of transaction should drive adoption
of governance arrangements: uncertainty, complexity, asset-specificity, and frequency. To
mention, uncertainty arises naturally as trading parties cannot foresee all future contingencies
and are limited in predicting what is going to happen as the transaction unfolds. In fact if
uncertainty was considered absent, the existence of firms would be pointless (Knight, 1921;
Coase, 1937). In real-world transactions, uncertainty associated with uneconomical levels of ex
ante transaction costs (procuring, drafting, negotiating and safeguarding contracts) may force
parties to move partnerships forward based on incomplete agreements. TCE recognizes that
contracts are often incomplete (Grossman & Hart, 1986; Hart & Moore, 1990) due to the
existence of uncertainty and bounded rationality.

17

Incomplete contracts in turn set the stage for holdup problems (Goldberg, 1976; Klein et
al., 1978; Klein, 1996), especially when one or both trading partners have incentives to behave
opportunistically if a particular outcome not ex ante specified in the contract is reached. When
the lack of consensus emerges between parties due to misspecifications, the legal system must be
contacted in order to solve the issue. But, as presented above, courts of justice are not always
capable of solving partnership conflicts with efficiency. The latter argument is especially true
when transactions are complex.
Along with uncertainty, the presence of information asymmetry (that is, when one or both
parties are willing to retain strategic information) might lead players to complex transactions. As
complexity increases, exchange partners tend to be more vulnerable to contract breaches and
therefore, are likely to incur additional safeguards in order to protect specific assets. In other
words, Masten (1984) suggests that the greater the complexity of a transaction, the greater the
likelihood of a party being bound to an inappropriate action of the counterparty.
As complexity of a transaction increases, players have strong incentive to establish third
party enforcers. These political bodies substitute the legal system when solving ex post problems
that emerge due to the lack of ex ante contract specification. A private enforcer is often more
efficient than the legal system to solve complex issues because the former focuses on a specific
set of laws and is better equipped to interpret trading parties’ actions. Higher efficiency,
however, comes at a cost. As one would naturally expect, the establishment and managerial
activities of an industry-specific organization generate additional costs to the involved parties.
Nevertheless, the existence of uncertainty and information asymmetry prevents third parties from
fully eliminating probabilities of opportunism.

18

If parties are interested in carrying on transactions even when contracts are incomplete
and third party enforcers are not fully capable of solving disputes, ex post controlling measures
must be launched to strengthen credibility and reduce threats of quasi-rents expropriation
(Maskin, 1999; Klein et al., 1978). Controlling measures vary in nature and intensity conditional
to the specificity of transactions. Peterson et al. (2001) suggests that ex ante control prevails
when more market oriented governances are strategically chosen by parties and ex post control
dominates when governance strategies are more vertically integrated.
The latter study also observes that the intensity of control increases as governance
structures become more vertically integrated. More specifically, spot market requires mild
intensity of control because parties do not engage in long-term relationships and the only control
parties tend to exercise are over price discovery and the decision of whether to enter into a future
transaction. When firms engage in more vertically integrated modes of governance, managers
tend to exert control with more intensity because holdup problems or moral hazard problems are
often more costly. Therefore, firms tend to adopt more vertically integrated coordination
strategies and more intense controlling measures when idiosyncratic assets expose parties to
higher expropriation of quasi-rents (Klein et al., 1978; Walker, 1988)
From a different perspective, the TCE literature suggests that transactions are likely to be
coordinated under more vertically integrated modes of governance when specific assets are in
place and inter-firms controlling activities are not solid enough to minimize threats of
opportunism. Asset specificity is the degree to which an asset can be redeployed to alternative
uses and by alternative users without sacrifice of productive value (Peterson et al., 2001). There
are five types of asset specificity defined in the literature: site specificity (Joskow, 1985),

19

physical asset (Klein et al., 1978), human-capital assets (Becker, 1962), dedicated assets
(Williamson, 1985 p.55), and intangible assets (Caves, 1982).
i.

Site specificity: when the buyer and seller are located side-by-side reflecting ex
ante decisions to minimize inventory and transportation expenses;

ii.

Physical asset specificity: when one or both parties in the transaction make
investments in equipment and machinery that involves design characteristics
specific to the transaction and have lower values in alternative uses;

iii.

Human-capital specificity: arises as a consequence of learning-by-doing,
investment, and transfer of skills (specific human capital) specific to a particular
relationship;

iv.

Dedicated assets: general investments that would not be undertaken but for the
prospect of selling a significant amount of product to a particular customer. If the
contract is prematurely terminated, it would leave the supplier with significant
excess capacity.

v.

Intangible assets: intangible capital such as brand name loyalty can have
relationship specific attributes. The value of these investments can be tied to the
company’s brand name.

Finally, frequency is the last attribute of transaction that affects adoption of governance
strategies. This attribute refers strictly to buyers activities in the market and specifies how often a
given transaction happens. In some cases, recurrent transactions control opportunistic behavior
as players are afraid of damaging reputation, and thereby affecting future transactions.
Conversely, occasional transactions may force parties to exercise intense levels of control if they
intend to engage in vertically integrated governance structures.

20

Nevertheless, all four attributes of transaction are equally important when parties to trade
seek the optimal governance strategy to coordinate transactions. Examining each attribute
individually may indicate whether more integration is necessary, but a serious decision only
obtains when interactions among attributes are also taken into account. Based on these
directions, Williamson (1985 p.72) proposed a methodology to predict the most efficient
governance strategy. Although his model considered only asset-specificity and frequency as
3

relevant attributes to guide the decision making process , it has not only shown that transactions
are indeed susceptible to analysis but it has also taught us how to perform applied research using
the TCE framework.
Accordingly to Williamson’s model there is an efficient governance structure for each
discrete combination of frequency and idiosyncratic assets. Namely, nonspecific investment
(asset specificity dimension) and both occasional and recurrent contracting (frequency
dimensions) lead to spot market governance. Occasional transactions with mixed and specific
investments are efficiently conducted under trilateral governance. Bilateral governance is
devised when recurring transactions are supported by mixed investments. Finally, vertical
integration is the most efficient mode of governance when asset-specific investments are
combined with recurring or occasional transactions. The efficient match of governance structures
with transactions are shown in figure 1.

3

Uncertainty is assumed to be present in sufficient degree to pose adaptation.
21

Figure 1: Efficient Governance According to Williamson (1985)

Occasional

Idiosyncratic

Trilateral governance
Market governance

Recurrent

Frequency

Nonspecific

Investment Characteristics
Mixed

Bilateral governance

Unified
Governance

Source: Williamson (1985, p.79)

In another well-known study, Klein et al. (1978) present five empirical case studies that
support Williamson’s framework. The paper discusses, in light of the TCE, why less integrated
governance structures may sometimes prevail over vertical integration. Years later Shelanski &
Klein (1995) obtain similar conclusions based on an extensive review of case studies. Both
surveys nevertheless conclude that each case study is unique and in-depth analyses are necessary
4

to understand what governance structure tends to be more efficient (or perhaps less inefficient ).
Since the breakthrough studies of Klein et al. (1978) and Williamson (1985), the TCE
literature has received important contributions. Milgrom & Roberts (1990) for instance have
defined influence costs that arise when transactions are organized within firms. Although
neglected in numerous empirical studies of governance strategy, influence costs are present in
any firm and can be associated with bureaucratic inefficiencies that inevitably develop when

4

TCE literature recognizes that Pareto efficiency is seldom obtainable. Hence, any governance
strategy chosen by decision makers is likely to carry some degree of inefficiency in the
traditional neoclassical sense of the word.
22

transactions are internalized (Masten, 1984). In sum, influence costs include not only direct costs
of influence activities but also costs of bad decisions.
In another work, Milgrom & Roberts (1992) define the concept of complementarity,
which along with asset specificity provides causal explanation to the costliness of holdup
problems or breach of partnerships. Complementarity exists when the output of a vertically
integrated firm is considerably larger than the sum of outputs generated by two or more firms
engaged in a transaction. Such concept was borrowed from the classic paper of Alchian &
Demsetz (1972) on agency cost theory. For them, complementarity is the synergy that emerges
from team production.
Peterson et al. (2001) discuss that the existence of a more efficient governance strategy is
not enough to guarantee its implementation. Although the study uses a business school oriented
vocabulary, the authors propose a framework that provides insights to refine our understanding
about coordination strategies. Specifically, the study points out that better governances need to
satisfy four conditions in order to be implementable (i.e. capital availability, existence of
compatible partners, control competence, and institutional acceptability).
Striking criticism to transaction costs theory has lately come from the emerging strategic
management literature. Jongwook & Mahoney (2005) contend that TCE may not be well
equipped to handle issues of shared ownership (e.g. cooperatives, joint ventures). They also
claim that TCE assumes an initial equilibrium state that is not consistent if trading parties move a
partnership forward based on incomplete contracts. The study suggests rather that the modern
property rights theory (Grossman & Hart, 1986; Hart & Moore, 1990) provides a better
framework for examining shared ownership cases.

23

Another study, also from the strategic management literature, indicates that two or more
governance strategies may be similarly efficient to coordinate transactions within the same
industry (Delmas & Tokat, 2005). The authors evaluate relative efficiency in the US electricity
sector using Data Envelopment Analysis (DEA method). The paper concludes that vertical
integration and spot markets are equally superior governance strategies to intermediate
transactions.

In summary, both branches of the new institutional economics (i.e. institutional
environment and transaction cost economics) have received numerous inputs throughout history.
The institutional environment branch has provided consistent explanation to what institutions are
and to how (marginal) institutional change happens. The branch has, however, failed to explain
how and why dramatic changes to institutions occur. Shirley (2005) argues that ‘the specifics of
institutional change fall through a gap in the literature’.
The TCE branch defines an elegant framework for examining transaction relationships.
Williamson (1991; 1996) and Peterson et al. (2001) have created an empirical research agenda in
which uncertainty, complexity, asset-specificity, and frequency are the underlying attributes.
Concerns have lately emerged from business scholars who argue that the classic TCE neglects
measurement problems (Mahoney, 1992), and assume an equilibrium state for choosing
governance strategies (Jongwook & Mahoney, 2005). Empirical evidence has also questioned
whether the classical notion of TCE sufficiently explains why firms may adopt different
governance strategies to coordinate transactions within an industry (Delmas & Tokat, 2005).

24

2.3. NIE Applied to Electricity Sectors

Electricity sectors have been extensively examined under the NIE lenses across the world
(Joskow, 1985; 1996; Souza, 2000; Delmas & Tokat, 2005; Spiller & Tommasi, 2005; Jolink &
Niesten, 2008; Fetz & Filippini, 2010). Regulatory reforms undertaken by countries over the last
few decades stand as the main focus of those analyses. The objectives however are divided
among implications, consistency as policies, and conformity with economic theory.
Joskow (1985) represents a landmark in the empirical application of TCE as a theoretical
framework. The author offers an extensive literature review and examines organizational
arrangements between power generating firms and coal suppliers in the United States. He
concludes that empirical evidence is consistent with the predictions of TCE: vertical integration
is superior for mine-mouth plants due to the high levels of asset-specificity (site specificity,
physical asset specificity, and dedicated assets). Results also point out that when vertical
integration is not observed, very long-term contracts are preferred alternatives to coordinate
transactions. The author, nevertheless, recognizes that his research agenda is not complete as it
provides little explanation for why coal-based generating firms ‘optimize’ the sum of production
and transaction costs and safeguard idiosyncratic assets through two different governance
strategies.
In another study, Joskow (1996) identifies some physical characteristics of electricity
sectors that tend to complicate analyses of organizational arrangements and coordination
strategies. To mention, network reliability requires supply and demand of power to be balanced
subject to complex constraints in real time rather than relying exclusively on tradable property
rights and spot market transactions to allocate resources efficiently (p. 342). Taking that into

25

consideration, the author suggests a few approaches to resolve coordination problems. The
author also reviews how countries have implemented structural and regulatory reforms in order
to promote competition among generating firms and to eliminate costly government-owned
arrangements.
Souza (2000) covers the first case study about institutions of electricity in Brazil after the
institutional reform (National Privatization Program and Electricity Sector Reform). The author
explores marketing mechanisms for renewable electricity in Brazil in a period marked with high
levels of uncertainty. Based on a qualitative analysis, the study concludes that uncertainty and
asset-specificity of power generation stations led trading partners to settle long-term
specification contracts. The study also suggests the definition of a minimum price policy for
renewable electricity in order to enhance market competitiveness.
Delmas & Tokat (2005) analyze the efficiency of governance structures played by
electric utilities during the process of retail deregulation (1998-2001) in the United States. The
authors analyzed quantitatively which organizational arrangement seems more efficient to
govern transactions between power generators and distribution utilities. Empirical evidences
show that either vertically integrated or market oriented firms are more efficient than firms with
intermediate governances. The study also concludes that the deregulation process has a negative
impact on efficiency in the short term due to high levels of uncertainty. In addition, the authors
discuss that transaction cost economics may not sufficiently explain why firms with different
governance structures reach similar levels of efficiency.
Spiller & Tommasi (2005) discuss about the implications of new regulations on public
electricity utilities. The authors argue that three main characteristics of utilities (large and
specific investments, technologies with important economies of scale, and widespread

26

consumption) cause major impacts on the way formal institutions are drafted. Namely, large
idiosyncratic investments in distribution grids may cause under-maintenance or underinvestment by its operator. Economies of scale imply few suppliers in each locality, which in
turn may facilitate exercise of market power. Finally, the massive consumption of electricity
along with its inelastic demand may bias government’s decisions and set the stage for corruption.
Jolink & Niesten (2008) uses the TCE framework to evaluate whether the liberalization
and re-regulation of European electricity industries direct firms to improve performance in the
Netherlands. Based on a qualitative analysis, results point out that the imposed spot market
governance is not the most efficient structure as large degrees of asset-specificity characterize
transactions. The paper concludes that generating firms would be better off under specification
contracts since the adoption of vertical integration is banned.
Fetz & Filippini (2010) employ an econometric model to quantitatively test whether
vertical integration is present in the Swiss electricity sector and whether it brings advantages to
companies. For motivation, the authors observe that most regulatory reforms introduced to
electricity sectors in Europe require legal and functional unbundling of vertically integrated
firms. Empirical results show that the majority of power generating firms and small-sized
distribution utilities are vertically integrated and enjoy some advantages of such governance.
These results support the European Union policy directive that exempts companies with 100,000
customers or less from any barrier in the organizational arrangement chosen.

Although several studies have applied the NIE framework to electricity sectors, research
gaps remain open. Namely, Joskow (1996) and Spiller & Tommasi (2005) address the question
of how regulatory reforms have taken place and how critical re-regulation of electricity

27

industries is. Neither study however takes a country as case study and scrutinizes the reasons
underlying its reform. Delmas & Tokat (2005) and Joskow (1985) have indicated that power
firms may deliberately choose different governance strategies in order to reach similar levels of
transaction efficiency. Delmas & Tokat (2005) however uses a quantitative approach and does
not analyze the underlying attributes of transaction as proposed in the TCE literature. In our view
this missing analysis could have provided causal explanation to firms’ adoption of distinct
governance strategies. Paul Joskow in turn recognizes that his paper lacks fundamental
explanation for why businesses (in his case, coal-burning generation utilities) choose different
coordination arrangements when the classic TCE theory suggests that they should adopt a unique
efficient strategy.
Aiming to overcome these limitations, the next chapters tackle the case study of Brazil’s
electricity sector. Chapter 3 analyzes the evolution of institutions and its effects on economic
growth from the 1930’s - when Brazil’s government started taking over operations of generation,
transmission, and distribution – to the 2000’s – when the sector reform and privatization program
reduced governmental participation in all three operating spheres. Such analysis lets us examine
the leading factors that motivated institutional reform, job not performed in Joskow (1996) or
Spiller & Tommasi (2005). In chapter 4 the classic underlying attributes of transaction (i.e.
uncertainty, complexity, asset-specificity, and frequency) are evaluated in-depth to guide
consistent discussion on whether the classic TCE framework (Williamson, 1985; Peterson et al.,
2001) suffices to explain adoption of coordination strategies.

28

3. INSTITUTIONS AND INSTITUTIONAL EVOLUTION OF THE ELECTRICITY
SECTOR IN BRAZIL

“[The new institutional economics] suggests a whole agenda for microeconomics
empirical work that must be performed (…). Until that work has been carried out, the new
institutional economics and related approaches are acts of faith, or perhaps of piety.”
- Herbert Simon (1991, p.27)

The evolution of Brazil’s electricity sector is marked by a historical lack of consensus
among governments about long-term economic policy. For over five decades (1934-1989), the
government’s involvement in the electricity sector prevailed and then started decreasing as the
Constitution of 1988 was promulgated. With the new Constitution, a range of neoliberal polices
were adopted including privatization of state-owned enterprises and reform of economic sectors.
In the face of severe international debts and high inflation rates, Brazil fell into recession
between 1982 and 1984 (Baer, 2008). The plan of privatizing state-owned companies was seen
as crucial for the raise of a long-term sustainable economy (Pinheiro & Giambiagi, 2000).
Privatization alone, however, would not fulfill the major goal of recovering the Brazilian
economy. In parallel to the privatization program, several economic sectors in which the
government was present as player (e.g. electricity, telecommunications, railroads, highways,
petrochemicals, steel, fertilizer) had institutions restructured in order to effectively meet the
directives of the new Constitution.

29

The reformulation of institutions of electricity in specific was not an easy task for the
5

reasons highlighted by Spiller & Tommasi (2005) . Not to mention, decision makers had the
additional challenge of designing incentives for new enterprises of power generation to enter the
field given that the major player (Brazil’s government) was in process of privatizing its assets. In
a parsimonious fashion, rule markers defined categories of generation firms and power
consumers, organized the marketplace as a dual-channel market, established regulatory agencies,
and provided autonomy for those agencies to formulate enforcement rules. Almeida (2005)
suggests that the ultimate intentions of the Electricity Sector Reform (ESR) were to create a
6

competitive wholesale market of power and to regulate transmission and distribution services .
Almeida’s suggestion in fact fits almost perfectly to the analysis carried out by Joskow (1996).
The following sections of this chapter review in-depth how the government became a
player after the 1929 Great Depression, how it retired from the role of player more than six
decades after, and how the institutional environment evolved to its current structure. The specific
objectives here are to provide compelling empirical evidence for why dramatic institutional
changes occur and how these changes influence development.

5

Three features define utilities (electricity included): (i) technologies are characterized by large
specific, sunk, investments; (ii) technologies are characterized by economies of scale; (iii)
products are massively consumed. The authors argue that these features are always source of
controversy when institutions are to be implemented or modified (Spiller & Tommasi, 2005
p.518).
6
Joskow (1996) refers to transmission and distribution services as ‘wire’ services. The same
terminology is adopted from here on.
30

3.1. Background

The negative effects of the 1929 Great Depression led occidental countries to adopt
initiatives of market protection and domestic industrial support. Brazil was not an exception.
State intervention in the electricity sector began with the Water Code (1934) which empowered
Brazil’s government to set electricity rates (Pompeu, 2006). The reason for this control was the
fact that electricity rates had been partly indexed to international gold prices and partly to the
domestic currency (Baer, 2008). As the Great Depression caused high volatility of gold prices
and undervaluation of the Brazilian currency, power rates could easily go up and consequently
bring electricity consumption down. Low electricity consumption in turn could adversely affect
domestic production, which was one of the government priorities at that time to overcome the
global depression.
Controlled rates of electricity were considered to be of national interest until the 1950s.
For over three decades the government set relatively low rates as a mechanism of subsidy to
industrial development. From one side, the rates policy (along with other policies defined in the
50s) indeed stimulated industrial development and economic growth (Baer, 2008 p.66). From the
other however, it repelled private enterprises of power generation and distribution that were
operating in the country. In other words, the government’s control over electricity prices limited
7

private power generating firms and distribution utilities (mainly foreign ) from reaching
adequate rates of return on investments (ROI) which in turn led them to leave the country. In
face with a growing electricity demand (associated with the supported industrial development)
7

Brazilian Traction Light & Power Co. and American & Foreign Power Co. (Canadian and
American companies, respectively) owned 70% of the capacity of generation in the country at
that time (Leite, 2009).
31

the state had no other alternative but to gradually enter the fields of power generation and
distribution.
Only two decades later, the state dominated the sector. In 1962, private companies
accounted for 64% of the power generating capacity; in 1977 this proportion was reduced to less
than 20%; and in 1982 almost all generation was run by state-dominated enterprises (Leite,
2009). Ferreira (2000) contends that the centralized model was justified not only to be consistent
with government programs of economic growth but also to permit high levels of technical
efficiency in a hydropower-dominated system. It is noteworthy to mention that the three largest
operating dams in the country (Itaipu, Tucuruí, and Ilha Solteira, respectively) were built
between 1967 and 1975 with government’s financial resources.
During the 70s and 80s a sequence of economy shocks associated with poor financial
8

strategies pushed Brazil into recession times with serious inflationary problems. Government
leaders believed that a large scale import substitution program (financed with international loans)
was the best alternative to fight the first oil shock of 1973. The substitution program was set
under the assumption that loans would be paid back as soon as the stimulated production was in
place, declining imports and increasing exports. The reality, however, was not so: the second oil
shock of 1979 and the interest rate shock of 1982 turned down international demand for
Brazilian products and aggravated debts contracted few years earlier.
Storming debts and high inflation rates forced Brazil’s government to venture its last
attempt to overcome the inevitable crisis. Public enterprises, the last stable segment of the
Brazilian economy, were used as tools of macroeconomic policies. Specifically, production of
8

Brazil contracted numerous international loans in the 70s to finance constructions and to
overcome the crisis imposed by the first oil shock of 1973. Loans, however, were contracted
based on flexible interest rates which were sharply increased in 1982 (triggered by the Volcker’s
policy in the US).
32

public enterprises (electricity, telephone services, iron, and steel) was used as instrument to
control inflation. Companies were forced to borrow more on international markets than they
needed in order to provide the government a continuous inflow of foreign exchange needed to
cope with a deteriorating balance of payments (Werneck, 1987). Baer (2008) points out that as
result of those policies the average electricity rate decreased by 40% between 1979 and 1984.
9

Economy shocks along with bad financial strategies left Brazil in a bankrupt state .
Regarding the electricity sector, the forced debt accumulation placed generating firms and
distribution utilities in unsustainable financial situations. Moreover, increasing demand of power
and frozen supply caused severe power outages in the 1990s.
With the liberal directives of the new Constitution, privatization of state-owned
enterprises and reform of several economic segments were given high priority to overcome the
debt crisis and better match supply and demand of power. The positive effects of such initiatives
were, however, only seen after 2001 once the two major programs for institutional change were
complete. Figure 2 attempts to represent government’s participation in the electricity sector over
time and highlights some of the related historical events.

9

Brazil required IMF (International Monetary Fund) assistance in December 1982. The austerity
program continued between 1983 and 1984.
33

Figure 2: Government’s Participation in the Electricity Sector
1988 Constitution
Government’s
involvement in the
electricity sector
1st oil
shock
(1973)

IMF
Austerity
Program

Great
Depression
1929

1982

1962

timeline
1995 2001

Source: Designed by the author, 2011. For interpretation of the references to color in this and all
other figures, the reader is referred to the electronic version of this thesis.
3.2. National Privatization Program (1995-2001)

The National Privatization Program (NPP) was one of the government initiatives
launched to modernize the Brazilian economy through a general liberalization process.
‘Privatization was seen as a safety net or bridge to stability, affording the country some leeway
for resolving its two main disequilibria, the current account and fiscal deficits’ (Pinheiro &
Giambiagi, 2000). Public enterprises of electricity, telecommunication, railroads, highways,
petrochemicals, steel, fertilizer, as well as an aircraft manufacturing firm and a computer firm
were all in the government’s privatization list. The privatization process generated $93.4 billion

34

in revenues and the electricity sector alone accounted for 31.8% ($29.7 billion) of the total
(Coelho, 2000).
Privatization was conducted differently depending upon the economic sector. Law 8,031
10

of 1990

established formal procedures for the privatization process of every sector where

public enterprises existed. This law introduced legal structures not only for private acquisition of
existing enterprises but also for the establishment of a variety of new private firms (Feldman,
1997), including investments in power distribution grids and transmission network.
Five years later, law 8,987 provided general rules for the process of contracting private
companies to operate public assets. In the electricity sector, privatization of generation facilities,
transmission grids, and distribution utilities occurred through the settlement of long-term
concession contracts between the government and private firms. This law also specified the
rights and obligations of concessionaries.
Concession contracts of electricity enterprises were publicly auctioned between 1995 and
2000. The NPP privatized a total of 23 state-owned firms. There was however a clear
improvement on specifying contracts over time (Ferreira, 2000). As contracts were individually
drafted for every public enterprise being conceded, clauses were better specified as the
regulatory reform proceeded (Pinheiro & Giambiagi, 2000). In other words, because concession
of state-owned electricity facilities and the restructure of formal institutions were happening
simultaneously, contracts drafted at the end of the five-year period were better specified than
those written at the beginning of the privatization program.
Concurrent policymaking decisions caused reluctance among potential auction buyers.
Leite (2009) suggests that there was a lack of interest in public auctions because during early
10

Feldman (1997) refers to this law as the ‘Concession Law’.
35

stages of the electricity sector reform, high levels of market uncertainty intimidated buyers:
‘auctions began before the complete definition of market regulations’ (p.54). For Feldman (1997)
Brazil was experiencing a paradox of simultaneous growth, represented by the necessary
privatization; and inertia, associated with the lack of interest and market uncertainty.
The close relation between the National Privatization Program and the Electricity Sector
Reform seems to cause confusion even among energy economists and policy makers. While
Ferreira (2000) associates the three formal laws (8,031; 8,937; 9,074) with the Privatization
Program, Feldman (1997) relates the first two laws with the same program and the third with the
Electricity Sector Reform. Either way, it is noteworthy to observe the close relation between
these programs. Regulations of the electricity sector were written taking into consideration
definitions made through the laws associated with the NPP. Likewise, concession contracts were
drafted based on the specific regulations as they were defined through the Electricity Sector
Reform.

3.3. Electricity Sector Reform (1995-1998)

The Electricity Sector Reform (ESR) and post amendments established a market that was,
until then, missing. Until 1995 market competition was not observed and all categories of
electricity users could only accept products and services offered by local distribution utilities, all
government-dominated. Rights and obligations of players, market regulations and mechanisms of
enforcement were crafted from scratch. Incentives not only to generators but also to consumers
of renewable electricity were created to diversify the electricity supply mix and to motivate entry
of new power generating firms.

36

The Public Service law (law 9,074 of 1995) introduced the guidelines for free market
competition in the electricity sector (Feldman, 1997). It formally defined electricity buyers and
sellers, and their rights and obligations. Specifically, this law formalized the entity of
Independent Power Producers (IPPs) just as the Public Utility Regulatory Policies Act did in
1978 in the United States. In addition, that law validated the entity of Independent Consumers
(ICs) who can obtain ‘wire’ service from their local distributor and purchase power supplies
directly from IPPs.
A questionable measure formalized through law 9,074 was that companies of generation,
transmission, and distribution should be functionally separated or completely restructured
through vertical divestiture. The measure partly meets the approach mentioned by Joskow (1996
p.361) to resolve coordination problems associated with abusive pricing strategies that could
otherwise arise. Feldman (1997) suggests that a good reason for this action came from past
experiences of other Latin American countries: ‘a combination of reform and unbundling has
resulted in a third more closing and twice as much megawatt production, as opposed to cases
where the reform has not been accompanied by unbundling’ (p. 5).
Ferreira (2000) in turn relates the measure to the situation of four state-owned companies:
‘CESP, CEMIG, COPEL, and CEEE were not only responsible for 34.7% of the generation
capacity in 1995 but also owned the largest assets of distribution across the country’ (p.207). As
the country did not intend to concede operation rights along with monopoly power to
concessionaries, these four state-owned companies had their assets divided among 15 smaller
companies. Some of them are still controlled by a single holding but organized under different
subsidiaries (Key informant interviews, 2010).

37

Regulations and enforcement mechanisms were indirectly introduced with the
promulgation of law 9,427 in 1996 which formalized the creation of the Electricity Regulatory
Agency (ANEEL, in Portuguese). ANEEL is an autonomous regulatory agency responsible for
overseeing the electricity sector and for enforcing quality protocols over generation firms,
transmission utilities and distribution utilities.
Law 9,427 also defined the entity of special consumers (SCs), a sub-category of ICs who
are authorized to trade renewable electricity (e.g. small hydro, biomass, solar, and wind-based)
directly with IPPs and to obtain ‘wire’ services from the local distribution utilities. This
definition marks the beginning of the renewable electricity segment, the fastest-growing segment
within Brazil’s electricity sector (Castro et al., 2008; Neves & Conejero, 2009). Captive
consumers are also defined under this law as those consumers of power who must accept prices
and conditions offered by local distributors.
The law 9,648 promulgated in 1998 established two other political bodies and introduced
the first incentive to SCs. First, it formally established the Wholesale Electricity Market (MAE,
in Portuguese), an agency responsible for assisting transactions among players (i.e. generating
firms, distribution utilities, ICs, and SCs). Second, the law assigned the task of coordinating
network operations to the National System Operator (ONS, in Portuguese). Finally, it introduced
the first incentive for SCs: a 50% discount in ‘wire’ service rates. Figure 3 summarizes the
conditions before and after the NPP and ESR; and what factors led Brazil to adopt these tools of
institutional change.

38

Figure 3: Summary of Conditions and Driving Factors of Institutional Change

Initial
Conditions:
- Power generation is
controlled by stateowned enterprises;
-Public utilities supply
all categories of power
consumers.

National
Privatization
Program

95

01
98

96

Electricity
Sector
Reform
Driving factors:
- Brazil at stage of bankruptcy;
- Debt crisis;
- High inflation rates;
- Increasing imbalance between supply and
demand of power.

Current
Conditions:
- Generation segment has
private (majority) and
public enterprises;
- Establishment of
agencies to enforce
regulations (ANEEL),
coordinate transactions
(MAE), and coordinate
operations (ONS);
- Establishment of a
competitive wholesale
market of power.

Source: Designed by the author, 2011

3.4. Later Institutional Changes

Out of the ESR scope, a set of major changes were still promulgated to formalize the
specification of players and the new market design. Minor adjustments also took place after the
conclusion of the ESR in order to stimulate new entry, especially in the segment of renewable
electricity.
In 2000, an amendment of law 9,074 decreased the minimum load (capacity installed)
required for independent consumers to trade directly with IPPs. Since then, ICs that have load of
3MW or higher have been authorized to negotiate and buy electricity from IPPs regardless of the

39

input used for generation. Medium-sized firms and hospitals, for instance, are compatible with
these requirements and trade directly with IPPs.
Under the same amendment, the entity of SCs had specifications formalized: SCs are
final consumers with load of 500 kW or more that are allowed to trade directly with IPPs as long
as the input used to generate power is renewable. Shopping malls, galleries, and museums for
instance can easily meet these technical requirements and be categorized as special consumers.
Final consumers with load equal or less than 500kW (e.g. households, small-sized firms) are
considered captive and must be supplied by local utilities. The law also re-states ANEEL
regulation 264 and protects ICs and SCs rights by allowing them to accept prices and conditions
offered by utilities of distribution rather than trading power themselves if they wish. Table 1
summarizes the definition of the economic agents in Brazil’s electricity sector.

Table 1: Categories of Economic Agents in Brazil’s Electricity Sector
Economic Agents
- Utilities of generation
- IPPs
- Utilities of distribution
- Independent consumers
(ICs)
- Special consumers
(SCs)
- Transmission utilities
- Captive consumers

Description
Concessionaries, private-owned, or government-owned firms
responsible for generating power as determined by ONS.
Independent Power Producers. Private or public enterprises.
Concessionaires or government-owned firms responsible for
providing services of distribution to captive consumers within a
specific geographical region.
Final-users with load of 3MW or more authorized by law to trade
directly with IPPs.
Final-users with load of 500kW or more authorized to trade with
IPPs based on renewable sources.
Concessionaries and private companies responsible for transporting
high voltage power from generating plants to utility-managed
substations.
Final-users of electricity with load less than 500kW;
Final-users with higher load who deliberately want to be provided by
distribution utilities.

40

With the definition of economic agents in place, an additional rule was still necessary to
complete the reformulation of institutions governing the electricity sector. The regulation 5,163
of 2004 played this role and formalized the marketplace as it currently is. The regulation has
designed a dual-channel market in which utilities of distribution are enforced to purchase power
11

through auctions, and the resulting mode of governance must be specification contract . The
same regulation has allowed ICs and SCs to coordinate transactions under a free channel using
any governance strategy except spot markets. Also in 2004, CCEE (Chamber of Electrical
Energy Commercialization) replaced MAE

12

with the mission of coordinating the dual-channel

market defined through regulation 5,163.
Figure 4 below summarizes the organization of both regulated channel and free channel.
It highlights transactions (green arrows) between generating firms, utilities of distribution, and
consumers of electricity. Blue-shaded boxes represent wholesale markets whereas the red-shaded
box corresponds to captive markets. Purple arrows represent the flow of electricity.

11
12

Other governance strategies are allowed but limited to small shares of their load.
Promulgated through law 10,848 of 2004.
41

Figure 4: The Dual-Channel Market of Power
 Regulated Market Channel:
Generatin
g firm

CCEE coordinates

Distribution
Utility

Specification contract
Transmission
Transmission
Transmission
Utility
service rates
service rates

ANEEL regulates

Captive
Consumer

Captive market
Distribution
Utility

 Free Market Channel:

Generating
firm

CCEE assists

Independent
Consumer

Any organizational arrangement except spot markets
Distribution
Transmission
Transmission
Utility
Distribution
Utility
Transmission
service rates
service rates
service rates
Source: Designed by the author, 2011

The long history of blackouts and power outages led Brazil’s government to establish
Energy Research Corporation (EPE, in Portuguese), an auxiliary organization whose main role is
to monitor the country’s energy supply and demand from a strategic standpoint and to develop
13

long-term planning . Except for the latter, all agencies are subordinated to the Ministry of
Mines and Energy (MME).
Minor institutional adjustments were also implemented in order to stimulate production
and consumption of renewable electricity. The ANEEL regulation 281 expanded the discount in
‘wire’ service rates to IPPs: utilities of generation based on renewable sources that were
13

Promulgated through law 10,847 of 2004.
42

established before 2004 receive a full exception in the service rate. The law 10,438 of 2002
expanded the stimulus even further and partially released renewables-based IPPs established
after 2004 from paying the whole ‘wire’ service rate. Still, the law 9,991 of 2000 eliminated the
R&D tax that biomass-, wind-, and small scale hydro-based generating firms had to pay.
Between 2002 and 2003 a national scale program took place to promote production of
renewable electricity and solve the historical imbalance of supply and demand of power (that
date from 1990s). The National Program of Incentives for Alternative Electricity Sources
(Proinfa, in Portuguese) supported the construction of 63 small hydropower mills, 52 windmills,
and 27 biomass-based generators with a total of 3,299MW (2.7% of the current capacity of
generation). Financial and marketing advantages were given to Proinfa investors: (i) financial
support of up to 70% of the total investment cost; (ii) reduced interest rates; (iii) acquisition of
production guaranteed; and (iv) minimum price guarantee for the following 20 years. With the
end of Proinfa in 2003, the Brazilian Development Bank (BNDES, in Portuguese) created a new
line of credit that has provided similar benefits, except for the acquisition and minimum prices
guarantee. Table 2 summarizes the laws and amendments related to the redefinition of formal
institutions governing the electricity sector in Brazil.

43

Table 2: Laws and Amendments Related to the Formal Institutions of Electricity in Brazil
Law/Rule
8,031

Year
1990

Details
- National Privatization Plan (NPP).

8,987

1995

- Rights and Obligations of conceded companies.

9,074

1995

- Definition of independent power producer (IPP);
- Definition of independent consumers (ICs);
- Unbundling of generation, transmission and distribution segments.

9,427

1996

- Establishment of ANEEL;
- Definition of special consumers (SCs).

1998

- Establishment of Wholesale Electricity Market (MAE);
- Definition of ONS functions;
- Creation of 50% discount in ‘wire’ rates for special consumers.

ANEEL 264

1998

- Independent consumers and special consumers may deliberately accept
conditions imposed by distribution utilities and stay under captive
markets.

ANEEL 281

1999

- Removal of ‘wire’ rates charged from IPPs based on alternative sources
that become operational before 2004.

Amendment
of law 9,074

2000

- Redefinition of independent consumers.

9,991

2000

- Elimination of R&D tax from biomass, wind, and small hydropower
based firms.

10,438

2002

- Launch of Proinfa;
- Extension of the 50% discount on ‘wire’ rates to IPPs based on
alternative sources that become operational after 2004.

10,847

2004

- Establishment of Electricity Research Company (EPE).

10,848

2004

- Establishment of CCEE.

Reg. 5,163

2004

- Definition of the current market design: creation of the regulated
contracting channel, free contracting channel, and imbalance market.

9,648

Multiple sources: CCEE website (2011), ANEEL website (2011), National Constitution.

44

3.5. Organization of the Dual-Channel Wholesale Market of Power

With the historical review of institutions and institutional change in place, the current
organization of the wholesale market of power can be presented further. Figure 4 above
summarizes the regulated channel and the free channel of power. As mentioned, blue-shaded
boxes represent wholesale markets whereas the red-shaded box corresponds to captive markets.
Regardless of the channel used for trading, all wholesale market players must liquidate positions
in the imbalance market at the end of every month. Transactions conducted through the (i)
regulated market channel, (ii) free market channel, and (iii) imbalance market are described
separately below.

3.5.1. Regulated Market Channel
In the regulated channel, transactions between generating firms and distribution utilities
are conducted through least-price auctions and are formalized as specification contracts. CCEE is
the regulatory agency responsible for coordinating these auctions where contracts are pre-defined
and parties to trade are left to bid prices under a competitive environment. Market experts name
these auctions as ‘auctions of ordinary supply’ because they supply the largest share of power
demanded by utilities of distribution.
Least-price auctions of ordinary supply are subdivided into two groups depending on
whether the generating plant (seller) is an operating facility or new enterprises. Auctions with
operating facilities are held one year before the actual supply. Auctions with new enterprises are
held three or five years in advance with the objective of ‘allowing investors to better plan

45

financial flows and return on investments’ (Key informant interviews, 2010). Nevertheless,
contracts resulting from this category of auction often last three years or more.
There are two other categories of auction: adjustment and backup. Auctions of
adjustment allow utilities that overestimated captive demand to transfer rights of future
consumption to other distribution utilities that underestimated it. Transactions are conducted via
best-price auctions and distribution utilities are both buyers and sellers. These auctions take place
a few months before the transference rights but can happen multiple times while contracts of
ordinary supply are valid. Contracts resulting from auctions of adjustment last less than one year.
Auctions of backup are exclusively used for contracting backup facilities. These facilities
are left at ONS’s disposition, which in turn activates them in the case of imbalance between
immediate supply and demand. No different from the auctions of ordinary supply, specification
contracts are ex ante determined and parties bid prices through least-price auctions. Table 3
summarizes auction types in the regulated market channel.

Table 3: Summary of Auctions in the Regulated Market Channel
Auctions
Ordinary supply

- Operating facilities
- New facilities

Adjustment
Backup

Time before
supply
one year ahead
3 or 5 years ahead
few months
one year ahead

Duration of
contracts
3 to 8 years
15 to 30 years
one year or less
usually 15
years

Regulated transactions unfold as follows: distribution utilities forecast captive power
demand in their designated area and communicate with CCEE; CCEE combines the demand of
multiple distributors and following the merit-of-order approach, solicits generating firms to
subscribe to the auction of ordinary supply; CCEE moves forward with the auction organization
46

when the number of subscriptions is large enough so the aggregate demand falls below the
potential aggregate supply; CCEE drafts specification contracts and makes them available to
auction participants; generators are provided with some time to review the contract and to
modify their bid size (complete withdrawal is possible at this stage); when the given time
expires, the remaining generating firms can only bid at the auction (withdrawal is no longer
possible); the least-price auction occurs; winning bidders are those generators that offer the least
price per megawatt hour (MWh) to partly supply the aggregate demand; specification contracts
are formalized among all winners and utilities; supply starts.
Merit-of-order is the approach used for soliciting participation of generating firms to
subscribe to auctions. The method guides CCEE to invite generating firms based on the average
production costs, from the lowest to the highest, until the potential aggregate supply covers
distributors demand. Following this approach, CCEE ends up soliciting participation to large
scale hydropower plants more often than any other source or technology of generation.
Estimations suggest that large scale hydropower plants supply approximately 56% of the power
distributed to captive consumers. For the same reason, IPPs are more frequently left to trade
power in the free channel. That, however, does not rule out the ability of IPPs to trade through
the regulated market channel but to do so CCEE must solicit their participation.
As specification contracts evolve, distribution utilities are required to periodically report
to ANEEL the actual amount of power delivered to captive consumers. In cases of mismatch
between power acquired and power delivered, the regulatory framework authorizes utilities of
distribution to engage in auctions of adjustment, also organized by CCEE. Although utilities can
trade up to 1% of their captive market loads, auctions of adjustment are interpreted as tool for
minimizing price risk at monthly imbalance market (Key informant interviews, 2011).

47

As contracts evolve ANEEL also compares gross profit margins across utilities.
Specifically, ANEEL estimates margins based on wholesale prices of power at which utilities of
distribution acquire power and the retail prices at which distributors resell to captive consumers.
This study is periodically conducted for all concessionaries and government-owned distributors
whose gross profit margins are cross-compared to control for unilateral market power.

3.5.2. Free Market Channel
The current institutions set ICs, SCs and IPPs partly free to adopt governance strategies
in the free channel, except spot markets. CCEE only requires parties to report the amount of
power transacted in order to guarantee network reliability and to enforce liquidation of positions
at the end of every month through the imbalance market. Price need not be reported and can be
used as strategic information for future transactions.
Other than ruling out spot markets as alternative governance, CCEE does not impose any
additional restriction to wholesale transactions in the free channel. Higher autonomy, however,
comes with the costs of procuring counterparties, negotiating and drafting contracts, establishing
controlling measures, and monitoring performance as transactions unfold. Some free channel
players, nevertheless, receive incentives (i.e. tax exemptions) that partly offset those transaction
costs. The current regulatory framework states that SCs and generating firms based on renewable
inputs are entitled to receive incentives whereas ICs and generating firms based on nonrenewables are not and must incur those costs themselves.
Transactions coordinated through the free channel account for 26% of the annual
electricity consumed in Brazil (EPE, 2011). Very often ICs and SCs hire assistance from

48

consulting companies to intermediate negotiations with IPPs due to their little understanding of
the ‘rules of the game’. These consulting companies often represent free consumers at CCEE.

3.5.3. Imbalance Market
Regardless of the market channel used to trade, both buyers and sellers must liquidate
14

positions at the end of every month . Consumers must match electricity acquired to electricity
consumed. Likewise, sellers must meet electricity sold with electricity generated in the
imbalance market. CCEE coordinates these transactions and imposes penalties in case of nonliquidation. Joskow (1996 p.360) mentions that the liquidation of imbalances is an essential part
of any credible competition model.
Market players liquidate positions by accepting market-clearing prices defined by CCEE.
Holders of short positions are requested to purchase additional shares of electricity at marketclearing prices. Holders of long position are credited with the difference (electricity acquired
minus electricity consumed or, electricity generated minus electricity sold) at the same clearing
prices.
Market-clearing prices are based on the marginal cost of generation (or average
generation cost of backup power) per subsystem. These prices result from two sequential
stochastic models: Newave15 and Decomp. Specifically, ONS runs a stochastic dual dynamic
programming model (Newave) to optimize reservoir storage and to schedule dispatches of
backup generators taking into consideration uncertainties of water inflow and demand of power
14

In the US electricity market, the imbalance liquidation follows the same rule but with different
temporal granularity: if a seller generates less than its day-ahead schedule, then it must purchase
the difference between this day-ahead schedule and its actual generation from the real-time
market at the real-time price. The same rule applies for consumers.
15
For more details, please refer to Pereira & Pinto (1991).
49

(reported by distribution utilities). Sequentially, using the estimates of Newave and a set of
technical variables as inputs, the ONS runs another dynamic optimization model (Decomp) to
match generation and demand and to determine the marginal costs of generation per subsystem.
The marginal costs are then released to wholesale players as market-clearing prices.

3.6. Implications of the Institutional Changes

The NPP, ESR and later adjustments have intensively modified the industry structure as
well as the opportunity set for economic agents. For instance, incentives (i.e. discounts in ‘wire’
service and R&D rates, and Proinfa) have stimulated the entry of numerous generating plants. As
of 2012, there are 387 small hydropower plants, 459 biomass-based plants, and 96 windmills in
operation that were not generating power before the institutional change. These power plants are
mostly classified as IPPs and participate with approximately 13% of the total capacity of
generation in the country.
As result of the second institutional reform – toward market economy; the entry of
renewable electricity generators has not only diversified the electricity mix but also decreased oil
dependence for electricity production in the country (Castro et al., 2008). Previously, large scale
hydropower plants dominated the segment with more than 80% of the generation capacity. As of
2012, there are 178 large scale hydropower plants in operation that account for approximately
67% of the total capacity installed (ANEEL, 2012). Besides, numbers indicate that diesel fuelfired plants represent no more than 3% of the country’s power supply.
Incentives along with the dual-market design also motivated the entry of free channel
players. There are approximately 587 ICs and 967 SCs trading electricity through the free

50

channel with demand of 982 TWh and 134 TWh per year, respectively (CCEE, 2012). It is
noteworthy to remember that this segment of consumers was absent before 1995, and the only
alternative available for final consumers of electricity was to accept prices and services offered
by local utilities of distribution.
Commentators argue, however, that free consumers (ICs and SCs) might be exposed to
additional procurement costs in the free channel as they need to identify counterparties, draft
agreements and safeguards, monitor agreement performance, and so on. On the other hand,
industry experts reply that ‘incentives not only offset procurement cost but also create marginal
benefits for consuming renewable electricity’ (Key informant interviews, 2010).
In this sense the second institutional reform has created opportunities for some consumers
to acquire more affordable electricity. By trading directly with power suppliers, free consumers
are able to alleviate unnecessary expenses that would be intrinsic to retail prices otherwise (i.e.
margin of the utility and related transaction costs). This interpretation seems to be aligned with
the existing NIE theory: economic development happens as institutional reforms lessen
transaction costs.
To a lesser extent, the same is true for captive consumers (i.e. those with capacity of
consumption of 500 kW or less) who can only accept prices and conditions imposed by local
utilities of distribution. Households, a likely type of captive consumer, would be harmed if
utilities exercised unilateral market power in their conceded geographical area. To minimize the
probability of such harm, the institutional reform has established enforcement organizations that
play pivotal roles in regulating retail prices and services. Specifically, CCEE organizes
competitive-oriented auctions through which utilities must acquire the largest share of their
aggregate demand; ANEEL estimates profit margins based on retail prices and auction prices;

51

ANEEL compares those margins across all operating utilities in order to ensure that profits fall
within a feasible range; also, ANEEL imposes operation standards so the service offered to
captive consumers has a certain level of quality and reliability.
Those regulatory activities certainly generate transaction costs to captive consumers that
were absent before the NPP and ESR, tools of the second institutional reform. However, one
might plausibly argue that such costs are consistently inferior to the costs associated with power
outages and blackouts frequently experienced across the country in 1990s. Besides, one would
also argue that these transaction costs are lower than unilateral market prices that would have
been charged to captive consumers if the institutional reform had not established regulatory
agencies. Therefore, the institutional reform has positively affected performance of Brazil’s
electricity sector by lessening transaction cost to final consumers.
Interestingly enough, this latter implication appears to be aligned with structural reforms
implemented in several other countries worldwide (Joskow, 1996). It also corroborates with
Shirley (2005 p.611) when she suggests that economic development occurs only if the
institutional framework protects firms and citizens against expropriation of rents, and fosters
exchange by lowering transaction costs.
Discussions follow an opposite path as one attempts to analyze the results of the first
institutional change – toward market protection. Due to high levels of uncertainty and worldwide
instability after the Great Depression of 1929, government’s decision forced private utilities to
exit Brazil’s electricity sector. Sequentially, a sequence of economic shocks and poor financial
strategies not only threw the country into recession but also resulted in frequent power outages in
the 1990s. The first institutional reform therefore failed to protect firms and citizens against

52

expropriation of rents and also failed to reduce transaction costs; the two necessary conditions
for economic growth (Shirley, 2005).

3.7. Conclusions

The overall purpose of this chapter was to expand the empirical literature of institutional
evolution. Specific objectives were: (i) to review the causes and implications of institutional
reforms in a developing country; and (ii) to provide empirical evidence on how dramatic, rather
than marginal, institutional change occurs.
Different from earlier studies, which have adeptly analyzed cases of marginal
institutional increment, this chapter focused on a case of dramatic institutional change. Brazil’s
electricity sector provided an interesting case to review because it experienced two dramatic
institutional changes in less than a century – the first toward government centralization and the
second toward market economy. A comprehensive review of historical facts indicates causes and
means for institutional change that had not been observed in earlier work. Specifically, the first
dramatic institutional change was motivated by ‘high levels of worldwide uncertainty’; whereas
the second institutional reform was influenced by ‘economic recession’ (Brazil at stage of
bankruptcy, debt crisis, and high inflation rates) and ‘reliability of utility services’. These factors
may be combined and classified as ‘economic performance’ in order to maintain the same degree
of generality that scholars have used elsewhere.
Results also indicate that the first institutional change failed to satisfy two conditions for
economic growth, as mentioned by Shirley (2005). The first dramatic reform not only forced
private companies to exit Brazil – due to their inability to reach expected returns in a government

53

controlled environment – but also exposed citizens to frequent power outages – forcing business
to incur additional transaction costs. Conversely, the second institutional reform re-established
balance between supply and demand of power, created a favorable environment for medium and
large firms to acquire power at low costs, and permitted diversification of the supply mix through
the entry of generating firms based on renewable sources. Thus, the second institutional reform
appears to have met the necessary conditions (Shirley, 2005) for economic development.
There were a few limitations that we could not overcome. Ideally, one would like to
estimate and compare social, economic, and environmental conditions before the institutional
change to conditions thereafter. In that way the effects of the institutional change could be
estimated, or at least, correlated to the social-economic growth and progress in environmental
protection. As this chapter covered a reasonably long time horizon, data has proven to be absent
or of hard access. Specific employment data and GHG emissions data for Brazil’s electricity
sector, for instance, appear not to be available.

54

4. ATTRIBUTES OF TRANSACTION, GOVERNANCE STRATEGIES, AND
MARKET PERFORMANCE

The institutional change conducted through the NPP and ESR has set the stage for a rich
analysis. Activities of defining players, structuring market channels, establishing regulatory
agencies, and creating incentives for renewable power were undertaken with the ultimate goal of
substituting the former state-dominated structure and creating a competitive-oriented wholesale
environment. Under the new circumstances discussed in chapter 3, a reasonable question arises:
do current institutions direct wholesale market players to adopt the most efficient governance
strategy in the sense of minimizing the sum of production and transaction costs?
This chapter attempts to answer the question by examining wholesale transactions – those
represented as green arrows within blue shaded boxes in figure 4 – using the classic TCE
framework. For every representative transaction, the underlying attributes (i.e. uncertainty,
complexity, asset-specificity, and frequency) are analyzed and discretely categorized to allow
comparative assessment across alternative strategies. Results indicate that not only institutions
but other factors (e.g. existence of compatible partners, capital availability (Peterson et al.,
2001), and reputation (Masten, 1996)) also prevent certain transactions from utilizing what
seems to be the most efficient governance strategy. In sections three and four of this chapter we
discuss about the reasons that have caused frequent misalignments between the TCE predicted
strategies – or the most efficient governance strategies – and the strategies observed in Brazil’s
wholesale power market.
The next two sections of this chapter analyze in-depth seven representative transactions
distributed between the regulated market channel and the free market channel.

55

4.1. Wholesale Transactions in the Regulated Market Channel

Wholesale transactions in the regulated market channel require utilities of distribution as
purchasers of power. Sellers can be large scale hydropower plants, biomass-based generating
plants, windmills, small scale hydropower stations, natural gas burning plants, and diesel fuel
burning plants. The underlying attributes of transaction between large scale hydropower plants
and utilities are analyzed below followed by transactions between biomass-based generating
plants and utilities, windmills and utilities, and diesel fuel burning plants and utilities.
Transactions between small scale hydropower plants and utilities and transactions between
natural gas burning plants and utilities are deliberately left out of our analysis as they carry
similar characteristics; and consequently, lead to insights similar to those obtained through
analyzing transactions between windmills and utilities and transactions between diesel fuel
burning plants and utilities, respectively.

a) Transactions between large scale hydropower plants and utilities of distribution
The first and most important set of wholesale transactions in the regulated channel has
large scale hydropower plants and utilities of distribution as protagonists. Transactions between
these parties represent approximately 56% of the electricity flow in the regulated channel
(CCEE, 2012). From one side of the transaction, utilities must ensure stable supply of power to
captive consumers located in their conceded region. From the other, hydropower plants seek to
optimize generating operations without exposing idiosyncratic assets to expropriation of quasirents.

56

Uncertainty is the first attribute that affects decisions made at those firms. Weather
conditions are, for instance, a source of uncertainty affecting operational decisions at large scale
hydropower plants. Weather not only affects water flows but also decisions regarding reservoir
management. On the other end, utilities of distribution have little certainty regarding electricity
consumption in captive markets; mainly because those markets are composed of numerous end
consumers with distinct characteristics (i.e. preferred voltage, seasonality of consumption,
mobility from and to the free channel).
Despite the sources of uncertainty, decision makers at large scale hydropower plants and
utilities of distribution face low exposure to uncertainty. As defined in the institutional
framework, the National System Operator (ONS) coordinates water flow at large scale
hydropower plants and activates backup facilities when there are expected shortages in supply or
sudden rises in demand. As a result, ONS centralizes operations of water flow and exempts
managers at large scale hydropower plants from responsibilities on power outages that poor
reservoir management might cause.
Decision makers at utilities of distribution also face reduced exposure to uncertainty as a
result of activities undertaken by regulatory agencies. Institutions of electricity dictate that
distributors must periodically forecast power demand in their captive markets. Because such
activity may produce outcomes with different time granularity, time horizon, units of
measurement and so forth; ANEEL provides directives so utilities perform standardized
forecasting analysis. The directives help utilities develop sophisticated models (in order to
provide ANEEL and ONS with a certain set of outputs), which lead to accurate estimates. In
addition, situations in which utilities of distribution receive direct assistance from ANEEL to

57

predict future power consumption of captive markets are not unusual (Key informant interviews,
2010).
To decrease uncertainty of demand even further CCEE authorizes utilities to trade future
rights of consumption through ‘auctions of adjustment’. Utilities that overestimated captive
demand and acquired more rights of power consumption than captive consumers truly demand
may trade up to 1% of those rights with utilities that underestimated demand in their service
areas. In that sense, ‘auctions of adjustment’ can be interpreted as another tool for decreasing
uncertainty of demand and price risk at the imbalance market.
Complexity is also reduced as regulatory agencies decrease the level of vulnerability to
which large scale hydropower plants and utilities of distribution are exposed. Specifically,
trading parties accept contract specification pre-determined by CCEE as well as its monitoring
and enforcing activities. Monitoring activities can easily detect opportunism of either party since
CCEE is fully informed about contract specifications, and rights & obligations of traders.
Enforcement takes shape of severe fines if any party deviates from meeting ex ante
specifications. Market participants suggest that even if market conditions change and contracts
fall out of the self-enforcing range (Klein, 1996), CCEE is in a good position – in terms of
information access – to adjust regulated contracts and decrease probability of opportunistic
behavior (Key informant interviews, 2010). Hence, complexity of transactions between large
scale hydropower plants and utilities is low.
The next attribute to be considered is idiosyncrasy of productive assets. Large scale
hydropower plants have high degrees of asset specificity for three reasons. First, a switch from
the first to the second best use

16

16

is likely to cause costly expropriation of quasi-rents. In other

Footnote 4 (page22) applies here.
58

words, the fact of discontinuing generation of electricity for a few days, until a new transaction is
formalized, may represent onerous costs to the generating firm. Not to mention the cost of
turning water turbines back on after a sudden shut down may reach prohibitive levels (Key
informant interviews, 2010).
Second, evidence of dedicated assets can be extracted from the news. Belo Monte for
instance, a large scale dam under construction in the margins of Xingu River had future rights of
supply auctioned on April of 2010 (five years prior to expected operation). Translating that to
applied TCE, Belo Monte has been built to fulfill a specification contract previously settled
through a regulated auction, which allows us to categorize such investment as transactionspecific investment. This observation seems to agree with Carroll et al. (1999 p.79) where the
authors suggest that regulated power generators are unlikely to invest in highly specific assets
without either very high up-front rents or assurances that once invested rents will not be
expropriated. Besides, market participants argue that ‘Belo Monte dam would not have been
built if the sale contract had not been signed beforehand or if the contract had not been enforced
by CCEE (Key informant interviews, 2010).
Third, maturity of financing projects is also a factor influencing specificity of generation
assets. Two plausible assumptions are made here in support of our argument: (i) investments in
large scale hydropower plants tend to require financial support due to the magnitude of the
projects; (ii) a financed generating power plant has no further obligation with the lending
organization after the end of its first commercial agreement.
The former assumption can be justified since investments are expensive and firms are not
likely to undertake them without contracting loans from either BNDES or other private lending
organization. Moreover, evidence in Brazil’s electricity market supports this assumption. Belo

59

Monte plant, the same example used above, has relied on multiple funding sources including
BNDES and private lending banks (BNDES, 2012).
The latter assumption is plausible because lending organizations may lend at reasonable
rates only if the borrower guarantees its engagement in productive activities that will generate
enough returns to repay loans and interests for the duration of the project. Consequently, when
the first commercial agreement ends and a given power generating firm seeks a new partnership,
it is likely that the firm has fully paid loans and interests back to the initial lender.
The argument here is that lending organizations often require investors to ensure
repayment of loan shares plus interests at every agreed period of time. The guarantee in turn
takes the shape of unrelated durable goods or partial appropriation of the investment made. If the
commercial agreement is breached and the investor is forced to redeploy his or her assets, the
next best use will sacrifice a large productive value as the lending organization may claim partial
possession over the investment or over the durable goods previously put as guarantee. It is the
same as saying that financed investments face appropriable quasi-rents that represent not only
the potential costs of redeploying assets to the next best alternative but also the latent loss of
capital put as guarantee to the lending organization.
Hence, large scale hydropower plants have high degrees of asset specificity that are
supported by the fact that (i) reallocation of assets from the first to the second best use is costly;
(ii) some large scale hydropower plants are categorized as dedicated assets; and (iii) financed
investments are likely to incur additional expropriation of quasi-rents in case of ex post breach.
Finally, frequency of transaction is the last attribute of Williamson’s framework. It is
plausible to argue that electricity is a good of continuous generation and continuous
consumption, which imply very high frequency of transaction. That is likely to be true regardless

60

of the technology or market channel used in its generation or transaction. Thus, frequency
provides little insight about potential differences of governance strategies across transactions.
For this reason we will give little attention to frequency as we analyze other transactions.
Based on low exposure to uncertainty, low level of complexity, high degrees of assetspecificity, and very high frequency of transactions, the classic TCE theory plausibly predicts
specification contract as the most efficient governance strategy. Although some commentators
may find this result controversial due to the high degree of idiosyncrasy, specification contract
suffices because uncertainty and complexity are issues of little concern to decision makers.
Alternatively, higher levels of involvement (i.e. alliance structures or vertical integration) in
which parties share risks and benefits in order to strengthen controlling activities may not be
necessary as ANEEL and CCEE provide intensive monitoring actions and punishments schemes.
However, it is noteworthy to mention that specification contract is imposed over
transactions between large scale hydropower plants and utilities rather than a choice made by the
parties. In that sense an intriguing question arises: If predicted and observed governance
strategies are aligned, why does the institutional framework bind strategic decisions of firms? In
other words, why do institutions prevent trading parties from freely making strategic decisions if
the most efficient strategy – and most likely to be chosen – is the same as the one desired by
policymakers? We shall return to these questions below.

b) Transactions between biomass-based generating plants and utilities of distribution
Transactions between biomass-based generators and utilities represent approximately 1%
of wholesale electricity sales in the regulated channel (CCEE, 2012). Not different from the
transactions analyzed above, utilities are responsible for ensuring stable power supply to captive

61

consumers and are closely monitored by regulatory agencies. Biomass-based plants, on the other
hand, minimize the sum of production and transaction costs and safeguard specific assets against
opportunistic behavior in order to optimize profits.
While utilities of distribution continue to receive assistance from ANEEL and CCEE,
managers at biomass-based power generators have to cope with uncertainty exposure without
counting on ONS operations. Quantity and quality of sugarcane biomass, for instance, are likely
to vary across growing seasons which may in turn affect electricity generation (Key informant
interviews, 2010). The variation in input quantity and quality is nevertheless less significant here
than it is at windmills or small scale hydropower plants. Therefore uncertainty can be
categorized as moderate for this transaction.
Market experts suggest that ONS centralizes operations at large scale hydropower plants
but let other generating firms, such as biomass-based generators, cope with uncertainty of supply
for two reasons: (i) unexpected defects at large scale hydropower plants might put the entire
national grid at risk if ONS did not centralize operations in order to improve managerial
efficiency; (ii) the network is conversely more resilient to technical flaws at independent
biomass-based generators whose contracts represent smaller shares of supply (Key informant
interviews, 2010). For these reasons, managers at biomass-based generating plants – as well as
windmills and diesel fuel burning plants – are fully responsible for meeting obligations settled
through regulated partnerships.
Complexity levels can be categorized as low because of the same reasons suggested
above. CCEE defines ex ante detailed contracts for transactions between biomass-based
generators and utilities, which increases task programmability and decreases vulnerability to
opportunism. Besides, ANEEL’s monitoring actions and CCEE’s enforcing schemes also

62

decrease probability of opportunistic behavior of counterparties, leading to low levels of
complexity.
Following with asset-specificity, the third attribute of transaction in Williamson’s
framework, biomass-based generators are often categorized as having moderate to high degrees
of idiosyncrasy. The large variety of biomass-based residuals used for power generation imposes
meticulous specification of equipments, which in turn imply physical specificity. Material
density, granularity, moisture content, and temporal availability of inputs are some of the
technical factors affecting specificity (Key informant interviews, 2010).
Investments in biomass-based generators are also site-specific due to the real purpose of
the investments. Sugar & ethanol manufacturing plants, for instance, generate power based on
bagasse – a residue of the sugarcane plant – in order to meet facilities’ electricity demand. The
same reason leads orange juice producers and rice processors to use orange peel and rice shells,
respectively. These generating plants are often located next to the facility where the core
business takes place in order to minimize transportation and inventory costs of residues. Using
Williamson’s terminology, these biomass-based generating plants are in a “cheek-by-jowl”
relation with the core business plants, which implies site specificity.
Redeployment to alternative uses, however, is likely to sacrifice lower productive value
when compared to large scale hydropower plants for two plausible reasons. First, electricity
generation is a complementary activity to the core business; and second, there is a thick market
for biomass-based generating equipment. Although power generating equipments tend to be
designed for a specific plant, experts mention that adjustments might make redeployment
possible (Key informant interviews, 2010). The same is not true for large scale hydropower

63

plants because of the unique landscape and geological configuration of dams. Therefore,
biomass-based generating plants have moderate to high degrees of asset specificity.
Transactions between biomass-based generating firms and utilities of distribution are thus
characterized with moderate exposure to uncertainty, low level of complexity, moderate to high
degree of asset-specificity, and high frequency. Based on these attributes, classic TCE
framework would predict that a strategic alliance would suffice to minimize the sum of
production and transaction costs as well as safeguard specific assets. Such prediction is plausible
because decision makers might be willing to align moderate uncertainty exposure to their risk
aversion profiles without having to intensify control. Besides, decision makers understand that
assets with moderate to high degree of idiosyncrasy are already safeguarded against
expropriation of quasi-rents through activities undertaken by regulatory agencies.
Reality, however, indicates a mismatch between predicted and adopted strategies.
Biomass-based generators and utilities of distribution are bounded by the institutional framework
and forced to adopt specification contracts to govern transactions rather than a more efficient
alliance. A natural question therefore is why the institutional framework seems to block biomassbased plants and distributors from reaching higher levels of transaction efficiency. Further
discussion follows the analyses of the remaining transactions.

c) Transactions between windmills/small scale hydropower plants and utilities of
distribution
Windmills and small scale hydropower plants together represent 3.1% of the electricity
flow in the regulated market channel (CCEE, 2012). Because of numerous similarities between
these sources, our analysis examines transactions between windmills and utilities of distribution

64

only. The results and arguments used here fit well and also represent transactions between small
scale hydropower plants and utilities of distribution.
As briefly mentioned above windmills and small scale hydropower plants are more
exposed to uncertainty of supply than large scale hydropower plants or biomass-based
generators. The reason is primarily associated with the inability of managers at windmills to
predict availability of input for power generation. Besides, ONS does not extend its assistance on
operations management to windmills as it does to large scale hydropower plants, making
managers at those plants responsible for meeting contractual obligations. Utilities of distribution
though, continue to face reduced exposure to uncertainty since ANEEL assists managers
forecasting power demand and CCEE holds auctions of adjustment, which allows utilities to
transfer rights of power consumption. Thus, windmills and utilities of distribution must cope
with moderate to high levels of uncertainty.
Complexity follows the same pattern discussed above: it tends to increase with the level
of vulnerability to opportunism. But windmills and utilities of distribution simply rely on third
party enforcers (i.e. ANEEL and CCEE) to minimize the probability of opportunism to occur at
the cost of having to agree on pre-determined contract terms. In that sense, transactions between
windmills and utilities have low levels of complexity due to monitoring activities and
punishments imposed over deviants.
In regard to asset-specificity, windmills are designed to optimize power generation as a
function of wind-energy resources available in a given region. The aerodynamic design of
blades, gearbox, generator, and tower height for instance must be precisely built to meet specific
characteristics of a given area. Experts suggest that re-dimensioning existing windmills to meet
characteristics of an alternative site is possible; however, real generation would fall far below the

65

installed capacity since certain features of the new location were not considered in the original
technical project (Key informant interviews, 2010). Hence, windmills characterize a type of
specific physical asset.
But the reason pushing windmills to a higher degree of idiosyncrasy when compared to
biomass-based generating plants has yet to be mentioned. Differently from biomass-based
generators, power generation is the primary business at windmills which would lead to higher
production loss in case of redeployment. In other words, redeploying a wind-based plant from
the most suitable to an alternative use would keep its assets idle for a reasonably long time (e.g.
time necessary to draft a new contract, time needed for technical adjustments) which can be
easily interpreted as a period of continuous productive loss. At biomass-based generating plants,
however, the core business (e.g. sugar & ethanol production, orange juice manufacture) would
stay in operation even if power contracts had to be renegotiated, leading to lower productive
losses. For that reason, asset-specificity tends to be higher at windmills than at biomass-based
generating plants.
With moderate to high level of uncertainty, low level of complexity, high degree of assetspecificity, and very high frequency of transactions the classic TCE framework would predict
that a strategic alliance lead to superior efficiency. That can be plausibly argued because an
alliance structure would give autonomy for firms to mutually share exposure to uncertainty based
on risk aversion profiles. Alternatively, more integrated governance structures (i.e. vertical
integration) might lead to inefficient outcomes because activities undertaken by ANEEL and
CCEE are already safeguarding assets with high degree of specificity and reducing probability of
opportunism through strict monitoring schemes and punishments. Toward the left side of the
governance continuum, specification contract might not create enough involvement between

66

parties so they can indentify mutual objectives and consequently, share risks and benefits from
the transaction.
Nevertheless, the institutional framework prevents decision makers from adopting what
seems to be a superior governance strategy and dictates that specification contract must be
adopted instead. Discussion on why the institutional framework prevents firms from adopting a
strategy that optimizes efficiency follows below.

d) Transactions between diesel fuel/natural gas burning plants and utilities of distribution
Transactions between utilities of distribution and diesel fuel burning plants or natural gas
powered plants represent together 40% of the electricity flow in the regulated channel (CCEE,
2012). This analysis focuses on transactions in which diesel fuel burning plants are the sellers.
As mentioned above, the results obtained here can be extrapolated to transactions between
natural gas burning plants and utilities.
Uncertainty, the first underlying attribute in the classic TCE framework, can be classified
as of low importance for two reasons. First, utilities of distribution continue to receive assistance
from ANEEL and CCEE in order to cope with uncertainties of demand. Specifically, ANEEL
provides sophisticated models for power demand forecasting and CCEE organizes auctions for
utilities to trade future rights of power consumption. Second, diesel fuel is widely and
consistently available across the country, which reduces uncertainty of supply. Although
managers at these generating plants are responsible for coping with uncertainty and cannot rely
on ONS to coordinate operational activities – as it is the case for large scale hydropower plants –
stable supply of diesel fuel reduces uncertainty to an issue of little concern.

67

In concordance to the analyses conducted above, transactions between diesel fuel power
plants and utilities of distribution are also of low complexity. Monitoring activities and
punishment mechanisms strictly enforced by ANEEL and CCEE decrease the probability of rent
expropriation due to opportunism of counterparties. As result, trading parties face low levels of
vulnerability, which implies low complexity of transactions.
Regarding asset-specificity, diesel fuel burning plants seem to have two opposite forces
balancing off the degree of specificity. On one hand, redeployment is likely to drive a given plant
out of its core business until a new contract is settled or until technical adjustments are
performed, leading to high productive losses. On the other, diesel fuel powered plants tend to be
inexpensive investments when compared to other technologies, which leads to little sacrifice of
productive value in case of redeployment. In other words, the opportunity cost of these facilities
is relatively low but keeping assets out of production causes expensive losses. Asset specificity
therefore appears to find its equilibrium at a moderate degree.
Frequency of transaction is the final underlying attribute in the classic TCE framework.
Not different from the analysis conducted above, the continuous generation and continuous
consumption of electricity characterize this attribute as of very high frequency.
Based on low levels of uncertainty, low levels of complexity, moderate degrees of asset
specificity, and high frequency of transactions the classic TCE framework would predict
specification contract as the most efficient governance. Such prediction is plausible as firms must
agree on certain technical standards without having to adopt more vertically integrated structures
in order to protect moderately idiosyncratic assets; especially because activities undertaken by
ANEEL and CCEE seem strict enough to minimize the probability of hold-up problems.

68

The perfect alignment between predicted and observed governance strategies brings us
back to the question posed at the end of our analysis of transactions between large scale
hydropower plants and utilities: why are firms imposed to use specification contract if the TCE
theory indicates that such governance would be the natural choice of decision makers? Table 4
summarizes regulated market channel transactions based on the categorized attributes discussed
here. The remainder of this chapter examines transactions conducted in the free market channel
and discusses the questions left open at the end of each subsection.

69

Table 4: Summary of Regulated Market Channel Transactions
Wholesale
sources
LS Hydro
Biomass
Windmill
SS Hydro
Diesel Fuel
Natural Gas

Buyer
Utility of
Distribution
Utility of
Distribution

Governance Structure
AssetUncertainty Complexity Specificity Frequency Predicted Observed
Low

Low

High

Very high

SC

SC

Moderate

Low

Moderate
to high

Very high

Alliance

SC

Utility of
Distribution

Moderate
to high

Low

High

Very high

Alliance

SC

Utility of
Distribution

Low

Low

Moderate

Very high

SC

SC

70

4.2. Wholesale Transactions in the Free Market Channel

Sellers in the free market channel are biomass-based generators, windmills, small scale
hydropower stations, natural gas burning plants, and diesel fuel burning plants. Buyers are
categorized as special consumers (SCs) – if the input used for generating the electricity acquired
is renewable; or independent consumers (ICs) – if the input used is natural gas or diesel fuel.
Such distinction is important due to a set of incentives (i.e. tax exemptions) offered to SCs but
not available to ICs.
Although the institutional framework does not forbid large scale hydropower plants from
trading here, two reasons drive them away from the free channel. The first is the merit of order
approach utilized by CCEE to solicit participation of sellers in auctions. As large scale
hydropower plants have the least expensive generating cost in comparison to other inputs and
technologies, they are quite often invited to trade through the regulated channel. The second
reason seems to be the overall preference of managers for the regulated channel over the free
channel. Market experts mention that under the current institutional framework, firms are in
favor of trading via specification contracts rather than setting up strategies in an environment
with higher levels of uncertainty and higher exposure to opportunism (Key informant interviews,
2010).
Other than the absence of large scale hydropower firms, the key difference between
channels lies primarily in the underlying attributes that receive direct influence from the
institutional framework (i.e. uncertainty and complexity). Asset-specificity and frequency are
very consistent across channels given that they are mainly related to the technology used for
generation and the general characteristic of electricity as a product. For these reasons the

71

following analysis places special attention on uncertainty and complexity and refers to the
analysis conducted above for asset-specificity and frequency. Transactions between small scale
hydropower plants and distributors and transactions between natural gas burning plants and
distributors are left out for the same reason mentioned in section 4.1 above.

a) Transactions between biomass-based generators and special consumers
Despite the fact that little information is publicly available regarding participation of
transactions in the free channel, two factors indicate the importance of biomass-based electricity.
First, estimates suggest that biomass-based electricity has potential to cover 15% of the national
demand by 2015 (Neves & Conejero, 2009). Second, biomass-based power represents less than
1% of transactions conducted in the regulated channel, leaving the reminder to be traded via free
channel.
Uncertainty, the first attribute of interest in the TCE context, is differently categorized
here when compared to the regulated channel. In the demand side, SCs often rely on consulting
companies to assist them to forecast future demand of power. The uncertainty level to which SCs
end up exposed to is conditional on the ability of consulting companies to perform forecasting
studies. It is well accepted, nevertheless, that consultants are not as well equipped as ANEEL to
predict future demand of power (Key informant interviews, 2010), leading to high levels of
uncertainty of demand.
In the other end, uncertainty of supply remains similar to what was discussed above.
Although managers at biomass-based generating facilities deal with weather uncertainty
themselves and are fully responsible for meeting agreement specifications without further
operational assistance; biomass characteristics and the nature of generators offset some exposure

72

to uncertainty. Thus, the overall level of uncertainty for transactions between biomass-based
generators and SCs is moderate to high.
Complexity also tends to have higher levels of importance for trading parties in the free
market channel than it does in the regulated channel. In this channel parties must not only
identify counterparties but also write formal agreements, negotiate safeguards, monitor
agreement performance, haggle in case of ex post misalignment of interests; rather than simply
accepting pre-determined contracts drafted by CCEE prior to regulated auctions.
In a first glance, bilateral agreements could decrease the probability to expropriation of
quasi-rents as much as CCEE’s regulated contracts do, but market participants seem not to agree
for two reasons. First, free market players believe that monitoring activities and punishment
mechanisms enforced by CCEE and ANEEL are more effective than any mechanism created to
moderate transactions in the free channel. Disputes would be solved in courts of justice, which in
turn, might have lower capability of judging and penalizing the deviator when compared to
CCEE.
Second, it is widely accepted that CCEE is better informed to adjust contract terms in
case of changes in market conditions than free market players would be to haggle. Thus, free
trading parties understand that the probability of hold-up problems to occur are higher than in the
regulated channel, considering that the contracts are naturally incomplete and related assets are
specific (Key informant interviews, 2010). For those reasons, transactions between biomassbased generators and SCs are relatively more vulnerable, and therefore of moderate complexity.
As pointed out above, asset specificity and frequency are consistent with the discussions
conducted in section 4.1. Biomass-based generators have moderate to high degree of specificity
for three reasons: (i) high degree of physical specificity, (ii) high degree of site specificity, and

73

(iii) low sacrifice of productive value in case of redeployment. Frequency is always high due to
nature of electricity as a product.
Based on moderate to high level of uncertainty, moderate level of complexity, moderate
to high degree of asset specificity, and high frequency, the classic TCE framework would easily
predict that efficient transactions between biomass-based generators and SCs would require
vertical integration as governance strategy. That is a plausible finding because firms must adopt
a reasonably high degree of involvement and mutuality such that uncertainty is shared based on
risk preferences and opportunistic behavior is considerably inhibited.
Market observation intriguingly indicates that transactions are either governed through
vertical integration – thus meeting TCE predictions – or via specification contract. More
interesting is the fact that the institutional framework does not constitute a barrier for adoption of
what seems to be the most efficient governance. There must be, therefore, other reasons for such
misalignment. We return to this point below.

b) Transactions between windmills/small scale hydropower stations and special
consumers
Transactions between windmills or small scale hydropower plants and SCs represent the
second most important set of transactions in the free market channel. These transactions obtain
high levels of uncertainty because generators need to deal with unstable weather patterns, and
consequently unstable supply of inputs; and SCs must cope with uncertain patterns of
consumption using less sophisticated forecasting models (when compared to those used by
ANEEL). Nevertheless, uncertainty of transactions between windmills and SCs tends to reach

74

high levels as opposed to moderate to high levels in transactions between windmills and utilities
of distribution in the regulated channel.
Transactions are moderately complex since parties need to define contract terms (and
incur associated transaction costs) in order to safeguard specific assets and minimize the
probability of opportunism. More specifically, windmills and SCs exert great effort in ex ante
drafting contractual specifications and ex post monitoring counterparties behavior, especially
because enforcing activities of regulatory agencies do not cover free market channel transactions.
It is broadly understood however, that vulnerability to opportunism is low in the
regulated channel and moderate in the free channel, even though firms attempt high efforts in
decreasing exposure to opportunism. As examined above, such difference has to do with higher
monitoring capacity and stricter punishment mechanisms undertaken by ANEEL and CCEE.
Following with asset-specificity, windmills are considered to be highly specialized assets
for the reasons given above. First, windmills are specific physical assets that would reach lower
production capacity in case of redeployment. Second, electricity generation is the main purpose
of a windmill, and therefore high productive value might be lost in case of contract breach.
Frequency, the fourth attribute of transactions, is invariability high for transactions of power
between windmills and SCs.
The classic TCE theory would then plausibly predict that vertical integration is the most
efficient strategy to govern transactions in which uncertainty reaches high level, complexity is
moderate, assets are highly specific, and frequency is also high. Several cases, however, indicate
that specification contract is widely used in the marketplace to moderate transactions between
windmills and SCs. More interesting is the fact that institutions are not blocking windmills and
SCs from adopting vertical integration, the most efficient governance strategy accordingly to the

75

classic TCE theory. Thus, other factors not considered in the TCE framework seem to be
affecting decisions of SCs and managers at windmills. After analyzing the last set of free channel
transactions we look at three additional factors that might explain why parties seem to choose a
governance strategy that is different from the predicted (or efficient) one.

c) Transactions between diesel fuel/natural gas burning plants and independent
consumers
Evidence suggests that agreements between diesel fuel burning plants and ICs or between
natural gas burning stations and ICs represent the third most important set of transactions in the
free channel for two reasons. First, most of the power generated in 1178 plants (11% of the
installed capacity in the country, approximately) is traded through regulated auctions – ordinary
supply and backup – representing over 40% of the power supplied to captive consumers (CCEE,
2012). That evidence alone implies that a reduced number of plants are left to trade in the free
market channel. Second, demand for non-renewable power in the free channel is low due to tax
exemptions given to consumers of renewable energy. Therefore, diesel fuel and natural gas
burning stations tend to suffer substitution effects introduced via incentive for renewables.
For a matter of completeness it is also important to categorize the underlying attributes of
the last representative transaction in Brazil’s wholesale power market and examine its efficiency.
Not different from what was discussed in the regulated channel, low level of uncertainty of
supply categorize this transaction. That is the case because of the consolidated diesel supply
chain, which facilitates input access at inexpensive costs. Uncertainty of demand, on the other
hand, tends to be of moderate concern to ICs. Small business, shopping malls, and museums – all
potential types of ICs – often hire consulting companies to forecast consumption patterns and

76

assist them at buying decisions. Consulting companies, nevertheless, recognize that ANEEL
provides better assistance to utilities of distribution in the regulated channel than they can offer
to ICs in the free channel. Therefore, diesel fuel burning stations and ICs face low to moderate
levels of uncertainty while trading power.
Complexity, the second attribute of transaction, also tends to reach higher level when
compared to that obtained in the regulated channel. Because trading parties cannot rely on either
ANEEL’s monitoring activities or CCEE’s punishments, they must draft agreements in order to
minimize the probability of opportunistic behavior of counterparties. Experts suggest that firms –
often assisted by consultants – formalize moderately effective agreements in the sense creating
long-lasting partnerships. Although those agreements are not as effective as enforcing actions
taken in the regulated channel, market participants believe that their level of vulnerability is
moderate, and therefore complexity reaches moderate levels (Key informant interviews, 2010).
Similarly to the analysis conducted above, a moderate degree of idiosyncrasy categorizes
diesel fuel powered plants. That appears to be the result because redeployment of assets related
to the core business of a given firm might cause irreparable productive losses. However, its
relatively low investment cost, when compared to other technologies, pushes the degree of
specificity down, equilibrating at a moderate degree. Finally, transactions between diesel fuel
powered plants and ICs are very frequent.
With low to moderate level of uncertainty, moderate level of complexity, moderate
degree of idiosyncrasy, and very frequent transactions, the classic TCE framework would predict
that efficiency is likely to be obtained via specification contract. That seems plausible because
parties need to find a well defined set of contract terms that decrease vulnerability and

77

safeguards assets with moderate degree of specificity. Besides, exposure to uncertain events is of
little to moderate concern of firms and hence need not be mutually shared.
In that sense, diesel fuel powered plants and ICs seem to choose the most efficient
strategy to govern transactions. Remarkably this is merely the third set of transactions – out of
seven sets – in which observed and efficient governance strategies are fully aligned. Table 5
summarizes free channel transactions based on the underlying attributes discussed here; and
highlights both predicted and observed governance strategies. Next two sections discuss about
the remarks left unanswered at the end of each subsection above.

78

Table 5: Summary of Free Market Channel Transactions
Wholesale
sources

Buyer

Biomass

Special
consumer

Windmill
SS Hydro
Diesel Fuel
Natural Gas

Governance Structure
AssetUncertainty Complexity Specificity Frequency Predicted Observed
Moderate
Moderate
Moderate
Very high
VI
VI / SC
to high
to high

Special
consumer

High

Moderate

High

Very high

VI

SC

Independent
Consumer

Low to
moderate

Moderate

Moderate

Very high

SC

SC

79

4.3. Mismatch between predicted and observed governance strategies in the regulated
market channel

The analysis above shows that institutions block firms trading through regulated channel
from choosing what the classic TCE predicts as the most efficient governance in half of the
transactions analyzed – two out of four. In this section we look at two reasons that seem
appropriate to explain why the institutional framework prevents trading parties from freely
choosing governance strategies in the regulated section. The first reason complements the study
of Spiller & Tommasi (2005), which we intend to recap here. The second reason has to do with
the source of capital used for developing Brazil’s electricity sector over time, especially those
resources employed to finance investments in idiosyncratic assets of power generation.
Spiller & Tommasi (2005) argue that the nature of the product – electricity – sets utilities
of distribution in a strategic position for exercising unilateral market power. Utilities are
naturally large and specific investments that rely on economies of scale in order to be profitable.
These two characteristics together imply few supplying firms in each location; and as a matter of
fact, there is not a region in Brazil where two or more utilities compete for customers. Adding
that to the fact electricity demand is often inelastic and widely consumed across the country,
utilities have strong incentive to charge higher-than-competitive retail prices to households.
The authors, however, recognize that authorities in modern countries are well informed
about this potential threat; and therefore, regulatory efforts are put into action to protect
households (Spiller & Tommasi, 2005). In Brazil’s case, regulation translates into specification
contracts ex ante defined by a third party enforcer (i.e. CCEE) and into coordinated competitive
auctions through which generators and utilities of distribution settle wholesale prices. Such

80

regulatory structure is ingenious and convenient because it forces utilities to disclose wholesale
prices almost effortlessly, regardless of the source or technology utilized for generation.
What is remarkable, nevertheless, is that no other governance structure would allow
CCEE to observe wholesale prices. If institutions allowed trading parties to adopt an strategic
alliance – as it seems to be efficient for two representative transactions (table 4) – they would
have strong incentives to hide wholesale price information. CCEE would then be incapable of
estimating and comparing gross revenue margins across utilities of distribution; and
consequently incapable of preventing unilateral market prices. In order to do so CCEE would
need to use more intrusive mechanisms (e.g. fixed retail prices) to protect households. Fixing
retail prices would, however, be harder to implement due to the heterogeneity of utilities and
generators in operation in Brazil.
With that in mind commentators question why institutions do not set households free to
purchase electricity directly from generating firms as ICs and SCs are allowed to do in the free
market channel. If that was the case, key informants suggest that severe network problems could
emerge because ONS would face difficulties managing the transmission system and maintaining
reliability of the system as a whole. Utilities would no longer be responsible for predicting future
consumption and consumption patterns in their conceded area, two crucial pieces of information
for ONS and ANEEL. The system operator would consequently be incapable of running
stochastic optimization models (i.e. Newave and Decomp) as it is currently done (Key informant
interviews, 2010).
In that sense specification contract along with regulated wholesale auctions appears to be
a convenient and unobtrusive regulatory mechanism, without which households would certainly
suffer expropriation of rents as discussed by Spiller & Tommasi (2005). Besides, specification

81

contract creates just enough control so ONS can conduct network operations and minimize the
probability of power outage.
The second reason for the lack of leeway in the regulated channel is the source of capital
employed for investing in specific assets of power generation. A close look at the underlying
features of generating firms trading power through the regulated channel reveals that most of
them used financial resources from public accounts. The vast majority was either built in the
1960s and 1970s using direct resources from the country or built with financial support from
BNDES (Brazilian Development Bank) after the NPP. The same is not true for generating firms
trading in the free channel where investments are mainly undertaken using loans from private
lending organizations or using private resources.
There are, therefore, strong evidences that policymakers designed the dual-channel
market and established a third party (i.e. CCEE) to take better enforcing actions over transactions
in which government’s financial resources are at stake, or to guarantee a stable rate of return on
investment (ROI) to concessionaries. CCEE enforcement is reduced to minimal levels in the free
channel because potential holdup problems do not endanger repayment of loans to the
government (through BNDES). Consequently, free channel players need to draft safeguards
without relying on CCEE in order to protect specific assets against opportunistic behavior of
counterparties.

82

4.4. Mismatch between predicted and observed governance strategies in the free market
channel

Although trading parties can freely choose their governance strategy without being
bounded by institutions, the analyses conducted in section 4.2 above indicate that predicted – or
efficient – and observed governance strategies are sometimes misaligned. In agreement with the
classic TCE framework biomass-based generating firms and SCs, for instance, would reach
efficient outcomes if vertical integration was always the chosen strategy; but it turns out that
specification contract is frequently chosen instead. Likewise, TCE predicts that vertical
integration is superior for transactions between windmills and SCs but parties seem to prefer
specification contract at all times. Thus, there must be other factors influencing strategic
decisions that go beyond the scope of the classic TCE literature.
Implementability and reputation, two complementary concepts, seem appropriate to
explain the distinction between adopted and predicted governance strategies for these
transactions. Peterson et al. (2001) argue that the existence of an efficient governance strategy
does not ensure its implementability. In the authors’ view an efficient strategy can only be
implemented if a set of four conditions are jointly met: (i) decision makers have the capital
required to implement the strategy; (ii) decision makers have trading partners with desirable
characteristics to meet the needs of the strategy to be implemented; (iii) decision makers are
capable of exercising the type of control required by the new strategy; and (iv) the strategy is
legal and does not violate the current institutional framework. Reputation, the second concept,
might curb opportunistic behavior in some circumstances and guide firms toward less integrated

83

governance strategies. As Masten (1996 p.15) puts it, ‘reputation may deter opportunism if
counterparties can observe its occurrence and have alternatives for future transactions’.
The concept of implementability becomes functional to explain why transactions between
biomass-based generators and SCs are sometimes governed via specification contract. Market
evidences suggest that vertical integration, the efficient strategy, is inevitably preferred for
transactions between biomass-based generators and co-located manufacturing firms. This
governance appears to hold until the power demanded by “cheek-by-jowl” firms is fully
supplied. From that point on, biomass-based generators find alternative SCs and trade the
remaining power generated. But because alternative SCs seldom present equal advantageous
characteristics, decision makers deliberately incur additional transaction costs and adopt
specification contract, which might plausibly be the second-best governance strategy. Translating
into Peterson et al. (2001) terminology, trading parties do implement the most efficient
governance when compatible partners exist; but implement an alternative strategy – specification
contract – when the second condition binds.
Looking from the alternative SC perspective as opposed to the biomass-based generator
perspective, the concept of implementability would indicate that the binding condition is rather
capital availability. While sugar & ethanol manufacturing firms, for instance, have either
invested in new power generators or renovated existing ones

17

because such a decision creates

competitive advantage; alternative SCs (e.g. museums, shopping malls) have plausibly decided
not to do so because that investment might not give them a competitive edge in their industry.
Thus, alternative SCs might have chosen to draft specification contracts with biomass-based
generators due to the unavailability of capital for investing in power generators. This very same
17

Also called green-field and brown-field projects, respectively.
84

reason might also explain why SCs engaged in transactions with windmills or small scale
hydropower plants also choose specification contract over backward integration.
Interestingly specification contract – the second best governance strategy – implies a
much lower degree of involvement between parties than commentators might expect. Based on
the underlying attributes of transaction analyzed above (summarized in table 5), one might have
expected to observe equity-based alliance (Peterson et al., 2001) – the next strategy moving to
the left along the governance continuum – as opposed to specification contract. The concept of
reputation (Masten, 1996), nevertheless, adds a new set of arguments to explain why
specification contract provides enough involvement between trading parties.
As mentioned above, institutional incentives (i.e. tax exemption) play an important role
in attracting consumers of renewable electricity; and these consumers – referred to as special
consumers (SCs) – tend to honor agreements with renewable power generators (often IPPs) for
two important reasons. First, interviewees suggest that tax exemptions not only offset transaction
costs but also create marginal benefits when compared to alternative purchasing schemes (Key
informant interviews, 2010). Second, there is only a handful of consulting companies assisting
SCs find counterparties, predict consumption patterns, draft and safeguard agreements, and so
on. This latter reason facilitates information flow across trading parties and let renewable power
generators assess credibility and reputation levels of SCs before signing new agreements.
With that in mind it is plausible to argue that SCs are induced to honor agreements in
order to purchase electricity at low prices; especially because renewable power generators can
costlessly observe deviation – or at least access its past occurrence – and because the alternatives
left for a deceiving SC to purchase power would be more expensive. Knowing this self-enforcing
institutional structure, renewable power generators might feel comfortable setting a governance

85

of little involvement – specification contract – rather than an strategic alliance that would
ultimately lead to rent sharing.
In that sense the nonexistence of potential partners with desirable characteristics or the
unavailability of capital drive transacting partners away from the TCE predicted strategy. Parties
appear to find specification contract superior to alliances, as some scholars would have expected
in the first place, because renewable power generators may costlessly observe deviation and SCs
have strong incentives to live up to the agreement.

4.5. Conclusions

This chapter contributes with empirical evidence that partly corroborates with Delmas &
Tokat (2005) and Joskow (1985). The classical notion of TCE seems appropriate to predict the
most efficient governance in the sense of minimizing the sum of production and transaction costs
and safeguarding idiosyncratic assets for a given transaction. Our findings, however, suggest that
the TCE framework might gain consistency in explaining why trading parties sometimes adopt
less efficient strategies if two concepts broadly discussed elsewhere were taken under
consideration.
Implementability (Peterson et al., 2001) explains that four conditions must be jointly
satisfied for an efficient strategy to be implemented. ‘Institutional acceptability’ appears to be the
binding condition in the regulated channel of Brazil’s wholesale power market, preventing
parties from adopting the most efficient governance strategy in two transactions analyzed. In the
free market channel – although institutions set parties free to choose governance strategies –
‘existence of compatible partners’ and ‘capital availability’ bind and induce parties to choose

86

less vertically integrated strategies. Reputation, the second complementing concept, is functional
to explain why firms with specific assets and moderate vulnerability to opportunism prefer
specification contract over governance strategies that may provide better safeguards. Institutional
incentives given to consumers of renewable electricity and the relative easiness in observing
deviations make special consumers better off preserving reputation for future transactions.
Finally, the classic TCE theory has proven itself as an elegant framework for predicting
transaction efficiency; but has also shown some inability to explain how firms decide over
governance strategies. Results indicate that including two complementary concepts to the classic
TCE framework might also distinguish TCE for its completeness to explain how trading parties
make decision over governance strategies.

87

5. CONCLUDING REMARKS

This thesis analyzes how Brazil’s electricity sector evolved from a state-dominated to a
market-oriented structure in less than a century. Results indicate that ‘economic performance’ –
high levels of worldwide uncertainty, economic recession, and reliability of utility services – was
the main factor driving dramatic institutional reforms. Only the second institutional reform,
however, was successful in creating economic growth and promoting environmental
sustainability through the introduction of renewable sources of power.
Recognizing the existence of path dependence between institutional evolution and
definition of business strategies, this thesis analyzes further whether the resulting institutional
framework permits economic agents to choose governance strategies. Results indicate that
regulated wholesale market players are bounded because regulations are enforced to protect
households against unilateral market power (as also observed by Spiller & Tommasi, 2005); and
also because public capital was vastly used for financing investments in idiosyncratic assets of
power generation. Results also indicate that free wholesale market players, although free to
choose the desired governance strategy, are seldom capable of implementing the most efficient
governance strategy because capital availability, existence of compatible partners, and future
reputation are issues of concern to decision makers.
For future research the results of this thesis might be aligned with other empirical studies
(such as Joskow, 1985; La Porta et al., 1997; 1998; 1999; Greif, 1994; Bardhan, 2000; Rodrik,
2000; Delmas & Tokat, 2001; Nugent & Robinson, 2002; Keefer & Vlaicu, 2004; and those
covered in Shelanski & Klein, 1995) in order to define a common pattern of institutional
development and strategy formulation. Once a pattern is determined, a tentative hypothesis might

88

guide us towards refining the current NIE theory. Grounded theory (Corbin & Strauss, 1990;
Strauss & Corbin, 1994) might be used as analytical framework to help us identify
complementing concepts not yet considered in theoretical models.
Finally, we would like to add that further advances can only be obtained if we
investigators enhance the body of empirical knowledge and consistently systematize historical
evidence in order to create grounds for refining well-accepted economic theories. We hope this
thesis has succeeded in doing so.

89

REFERENCES

90

REFERENCES

Alchian, Armen A., and Harold Demsetz. "Production, Information Costs, and Economic
Organization." American Economic Review 62, no. 5 (1972): 777-95.
Almeida, A.A. “Política Tarifária e Comercialização de Energia Elétrica no Ambiente de
Contratação Livre.” Anais do Simpósio Brasileiro de Sistemas Elétricos, Curitiba, Brazil,
2005. Retrieved Jun 03, 2011 from the World Wide Web:
http://www.labplan.ufsc.br/congressos/SBSE/anais/089_sbse2006_final.pdf
ANEEL. Agencia Nacional de Energia Elétrica (ANEEL). Brasília, Brazil. Retrieved May 25,
2011 from the World Wide Web: http://www.aneel.gov.br
Aron, Janine. (2000), “Growth and Institutions: A Review of the Evidence”, World Bank
Research Observer, Vol. 15, No. 1, pp. 99-135

Arrow, Kenneth J. "Uncertainty and the Welfare Economics of Medical Care." American
Economic Review 53, no. 5 (1963): 941.
———. “Foreword” In Firms, Markets and Hierarchies – the Transaction Cost Economics
Perspective, Carroll, Glen R., Teece, D.J. (eds). Oxford University Press, 1999
Baer, Werner. “The Brazilian Economy: Growth and Development” Boulder: Lynne Rienner
Publishers, 2008
Banker, R. D., A. Charnes, and W. W. Cooper. "Some Models for Estimating Technical and
Scale Inefficiencies in Data Envelopment Analysis." Management Science 30, no. 9
(1984): 1078-92.
Bardhan, Pranab K. "Understanding Underdevelopment: Challenges for Institutional Economics
from the Point of View of Poor Countries." Journal of Institutional and Theoretical
Economics (JITE) / Zeitschrift für die gesamte Staatswissenschaft 156, no. 1 (2000): 21635.
91

Bates, Robert H. "Proserity and Violence: The Political Economy of Development." New York:
W.W. Norton & Company, 2001
Becker, Gary S. "Investment in Human Capital: A Theoretical Analysis." Journal of Political
Economy 70, no. 5 (1962): 9-49.
Boyd, Robert, and Peter J. Richerson. "Culture and the evolutionary process." Chicago:
University of Chicago Press, 1985
Carroll, Glenn, Pablo Spiller, and David Teece. “Transaction Cost Economics: its Influence on
Organizational Theory, Strategic Management, and Political Economy.” In Firms,
Markets and Hierarchies: The Transaction Cost Economics Perspective, Carroll, G.R.,
Teece, D.J. (eds). Oxford University Press: New York; 60-88, 1999
Castro, Nivalde José de; Guilherme de A. Dantas, André L. da Silva Leite, and Roberto Brandão.
“Bioeletricidade e a Indústria de Álcool e Açúcar: possibilidades e limites.” Rio de
Janeiro, Brazil: Synergia, 2008.
Caves, Richard E. “Multinational enterprise and economic analysis.” Cambridge, UK:
Cambridge University Press, 1982
CCEE. Câmara de Comercialização de Energia Elétrica (CCEE). São Paulo, Brazil. Retrieved
May 15, 2011 from the World Wide Web: http://www.ccee.org.br
Coase, R. H. "The Nature of the Firm." Economica-New Series 4, no. 16 (1937): 386-405.
———. "The New Institutional Economics." American Economic Review 88, no. 2 (1998): 7274.
Commons, John R. "Institutional Economics." New York, US: Macmillan, 1934
Corbin, Juliet, and Anselm Strauss. "Grounded Theory Research: Procedures, Canons, and
Evaluative Criteria." Qualitative Sociology 13, no. 1 (1990): 3.

92

Delmas, Magali, and Yesim Tokat. "Deregulation, Governance Structures, and Efficiency: The
U.S. Electric Utility Sector." Strategic Management Journal 26, no. 5 (2005): 441-60.
Ellickson, Robert C. "A Hypothesis of Wealth-Maximizing Norms: Evidence from the Whaling
Industry." Journal of Law, Economics, & Organization 5, no. 1 (1989): 83-97.
EPE. Empresa de Pesquisa Energética (EPE). Brasília, Brazil. Retrieved May 15, 2011 from the
World Wide Web: http://www.epe.gov.br
Fava Neves, Marcos, and Marco Antonio Conejero. “Estratégias para a Cana no Brasil: um
Negócio Classe Mundial.” São Paulo, SP, Brazil: Editora Atlas, 2010
Feldman, Roger D. "Brazil's Power Privatization: Prelude to New Infrastructure Development
Approaches." Journal of Project Finance 3, no. 3 (1997): 13.
———. "Brazil: Deregulation and the Future of Project Finance." Journal of Project Finance 4,
no. 1 (1998): 41.
Fetz, Aurelio, and Massimo Filippini. "Economies of Vertical Integration in the Swiss Electricity
Sector." Energy Economics 32, no. 6 (2010): 1325-30.
Foss, Nicolai J. "Knowledge-Based Approaches to the Theory of the Firm: Some Critical
Comments." Organization Science 7, no. 5 (1996): 470-76.
Goldberg, Victor P. "Regulation and Administered Contracts." The Bell Journal of Economics 7,
no. 2 (1976): 426-48.
Granovetter, Mark. "Economic Action and Social Structure: The Problem of Embeddedness."
American Journal of Sociology 91, no. 3 (1985): 481-510.
Greif, Avner. "Contract Enforceability and Economic Institutions in Early Trade: The Maghribi
Traders' Coalition." American Economic Review 83, no. 3 (1993): 525.

93

———. "Cultural Beliefs and the Organization of Society: A Historical and Theoretical
Reflection on Collectivist and Individualist Societies." Journal of Political Economy 102,
no. 5 (1994): 912-50.
Grossman, Sanford J., and Oliver D. Hart. "The Costs and Benefits of Ownership: A Theory of
Vertical and Lateral Integration." Journal of Political Economy 94, no. 4 (1986): 691719.
Hart, Oliver, and John Moore. "Property Rights and the Nature of the Firm." Journal of Political
Economy 98, no. 6 (1990): 1119-58.
Herbst, Jeffrey I. "States and Power in Africa." Princeton, NJ: Princeton University Press, 2000

Hodgson, Geoffrey Martin. “The Evolution of Institutional Economics : Agency, Structure, and
Darwinism in American Institutionalism, Economics as Social Theory.” London ; New
York: Routledge, 2004.
Jensen, Michael C., and William H. Meckling. "Theory of the Firm: Managerial Behavior,
Agency Costs and Ownership Structure." Journal of Financial Economics 3, no. 4 (10//
1976): 305-60.
Jolink, Albert, and Eva Niesten. "Governance Transformations through Regulations in the
Electricity Sector: The Dutch Case." International Review of Applied Economics 22, no.
4 (2008): 499-508.
Joskow, Paul L. "Vertical Integration and Long-Term Contracts: The Case of Coal-Burning
Electric Generating Plants." Journal of Law, Economics, & Organization 1, no. 1 (1985):
33-80.
———. “Introducing Competition into Regulated Network Industries: from Hierarchies to
markets in Electricity.” Industrial and Corporate Change 5, no. 2 (1996): 341-382.
———. “Vertical Integration” In Handbook of New Institutional Economics, 319-348:
Dordrecht, UK: Springer, 2005

94

Jütting, Johannes. “Institutions and Development: A Critical Review”. OECD Development
Centre Technical Papers No.210 (2003)

Keefer, Philip and Razvan Vlaicu. “Democracy, Credibility and Clientelism”. World Bank Policy
Research Working Paper No.3472. (2004)

Klein, Benjamin, Robert G. Crawford, and Armen A. Alchian. "Vertical Integration,
Appropriable Rents, and the Competitive Contracting Process." Journal of Law and
Economics 21, no. 2 (1978): 297-326.
Klein, Benjamin. "Why Hold-Ups Occur: The Self-Enforcing Range of Contractual
Relationships." Economic Inquiry 34, no. 3 (1996): 444-63.
Knight, Frank H. “Risk, Uncertainty, and Profit.” Library of Economics and Liberty, 1921.
Retrieved
Jun
11,
2011
from
the
World
Wide
Web:
http://www.econlib.org/library/Knight /knRUP.html
La Porta, R., F. Lopez-de-Silanes, A. Shleifer, and R. Vishny. "The Quality of Government."
Journal of Law, Economics, and Organization 15, no. 1 (1999): 222-79.

———. "Law and Finance." Journal of Political Economy 106, no. 6 (1998): 1113.

———. "Legal Determinants of External Finance." Journal of Finance 52, no. 3 (1997): 113150.

Leite, Antonio D. “Energy in Brazil: Towards a Renewable Energy Dominated System.”
London, UK: Earthscan, 2009
Mahoney, J. T. "The Choice of Organizational Form - Vertical Financial Ownership Versus
Other Methods of Vertical Integration." Strategic Management Journal 13, no. 8 (1992):
559-84.
———. “Economic Foundations of Strategy.” Thousand Oaks, CA: Sage, 2005
95

Maskin, Eric. "Nash Equilibrium and Welfare Optimality." The Review of Economic Studies 66,
no. 1 (1999): 23-38.
Masten, Scott E. "The Organization of Production: Evidence from the Aerospace Industry."
Journal of Law and Economics 27, no. 2 (1984): 403-17
———. “Case Studies in Contracting and Organization” New York, US: Oxford University
Press, 1996
Ménard, Claude, and Mary M. Shirley. "Introduction" In Handbook of New Institutional
Economics, Ménard, C., Shirley, M.M. (eds). Springer Berlin Heidelberg, 2005; 1-18
Milgrom, Paul, and John Roberts. "The Economics of Modern Manufacturing: Technology,
Strategy, and Organization." The American Economic Review 80, no. 3 (1990): 511-28.
———. “Economics, organization, and management.” Englewood Cliffs, NJ: Prentice-Hall,
1992
Mitchell, Wesley C. "Commons on Institutional Economics." American Economic Review 25,
no. 4 (1935): 635-52.
North, Douglass Cecil. “Institutions, Institutional Change, and Economic Performance”, Political
Economy of Institutions and Decisions. Cambridge; New York: Cambridge University
Press, 1990
———. "Understanding the Process of Economic Change." New Jersey, NJ: Pricenton
University Press, 2005
Nugent ,Jeffrey B. and James A. Robinson. “Are Endowments Fate?” Centre for Economic
Policy Research Discussion Paper No.3206 (2002).
ONS. Operador Nacional do Sistema (ONS). Rio de Janeiro, Brazil. Retrieved May 29, 2011
from the World Wide Web: http://www.ons.org.br

96

Ostrom, E. "An Agenda for the Study of Institutions." Public Choice 48, no. 1 (1986): 3-25.
Pereira, M. V. F., and L. M. V. G. Pinto. "Multi-Stage Stochastic Optimization Applied to
Energy Planning." Mathematical Programming 52, no. 1 (1991): 359-75.
Peterson, H. Christopher, Allen Wysocki, and Stephen B. Harsh. "Strategic Choice Along the
Vertical Coordination Continuum." The International Food and Agribusiness
Management Review 4, no. 2 (2001): 149-66.
Pinheiro, Armando C. and Fabio Giambiagi. “Os Antecedentes Macroeconômicos e a Estrutura
Institucional da Privatização do Brasil” In A Privatização do Brazil: O Caso dos Serviços
de Utilidade Publica Pinheiro, A.C., Fukasaku, K. (eds). Brasília, BNDES, 2000;
Available (in Portuguese) at:
http://www.bndes.gov.br/SiteBNDES/bndes/bndes_pt/Institucional/Publicacoes/Consulta
_Expressa/Setor/Privatizacao/200002_12.html
Pompeu, Cid T. “Direito das Águas no Brasil.” São Paulo, Editora Revista dos Tribunais, 2006.
Rodrik, Dani. “Institutions for High-Quality Growth: What They Are and How to Acquire
Them?” National Bureau of Economic Research Working Papers No.7540 (2000).
Samuels, Warren J. “Institutional Economics.” Aldershot, Hants, England: Edward Elgar, 1988.
Shelanski, Howard A., and Peter G. Klein. "Empirical Research in Transaction Cost Economics:
A Review and Assessment." Journal of Law, Economics, & Organization 11, no. 2
(1995): 335-61.
Shirley, Mary M., "Institutions and Development” In Handbook of New Institutional Economics,
Ménard, C., Shirley, M.M. (eds). Springer Berlin Heidelberg, 2005; 611-638
Simon, Herbert A. "Rationality as Process and as Product of Thought." The American Economic
Review 68, no. 2 (1978): 1-16.
———. "Organizations and Markets." Journal of Economic Perspectives 5, no. 2 (1991): 25-44.

97

Souza, Z.J. “Uma Avaliação das Formas de Comercialização da Energia Co-gerada pelo Setor
Sucroalcooleiro.” MS Thesis. Applied Economics Program, Department of Economics,
Administration and Sociology. University of São Paulo, Piracicaba, Brazil, 2000
Spiller, Pablo T., and Mariano Tommasi. "The Institutions of Regulation: An Application to
Public Utilities” In Handbook of New Institutional Economics, Ménard, C., Shirley, M.M.
(eds). Springer Berlin Heidelberg, 2005; 515-543
Strauss, Anselm, and Juliet Corbin. "Grounded Theory Methodology: An Overview." In
Handbook of Qualitative Research., 273-85: Thousand Oaks, CA, US: Sage Publications,
Inc, 1994.
Walker, Gordon. "Strategic Sourcing, Vertical Integration, and Transaction Costs." Interfaces 18,
no. 3 (1988): 62-73.
Werneck, Rogério L.F., “Public Sector Adjustment to External Shocks And Domestic Pressures
In Brazil, 1970-85.” Rio de Janeiro, Departamento De Economia, Pontifícia Universidade
Católica do Rio de Janeiro, 1987
Williamson, Oliver E. “The Economic Institutions of Capitalism: Firms, Markets, Relational
Contracting.” New York, US: The Free Press, 1985
———. "Comparative Economic Organization: The Analysis of Discrete Structural
Alternatives." Administrative Science Quarterly 36, no. 2 (1991): 269-96.
———. “The Mechanisms of Governance.” New York, Oxford University Press, 1996
———. "The Institutions of Governance." The American Economic Review 88, no. 2 (1998): 7579.
———. "The New Institutional Economics: Taking Stock, Looking Ahead." Journal of
Economic Literature 38, no. 3 (2000): 595-613.
Wooldridge, Jeffrey M. “Econometric analysis of cross section and panel data”. Cambridge,
Mass: MIT Press, 2002

98

Yin, R.K. “Case Study Research: Design and Methods” Thousand Oaks, CA: Sage, 2009

99