-
DISTRIBUTIONAL IMPACTS OF ENERGY POLICIES IN INDIA:
IMPLICATIONS FOR EQUITY IN INTERNATIONAL CLIMATE CHANGE
AGREEMENTS
A DISSERTATION SUBMITTED TO
THE EMMETT INTERDISICPLINARY PROGRAM IN
ENVIRONMENT AND RESOURCES
AND THE COMMITTEE ON GRADUATE STUDIES
OF STANFORD UNIVERSITY
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
NARASIMHA DESIRAZU RAO
AUGUST 2011
-
http://creativecommons.org/licenses/by-nc/3.0/us/
This dissertation is online at:
http://purl.stanford.edu/py027yn9445
© 2011 by Narasimha Desirazu Rao. All Rights Reserved.
Re-distributed by Stanford University under license with the
author.
This work is licensed under a Creative Commons
Attribution-Noncommercial 3.0 United States License.
ii
http://creativecommons.org/licenses/by-nc/3.0/us/http://creativecommons.org/licenses/by-nc/3.0/us/http://purl.stanford.edu/py027yn9445
-
I certify that I have read this dissertation and that, in my
opinion, it is fully adequatein scope and quality as a dissertation
for the degree of Doctor of Philosophy.
Lawrence Goulder, Primary Adviser
I certify that I have read this dissertation and that, in my
opinion, it is fully adequatein scope and quality as a dissertation
for the degree of Doctor of Philosophy.
Debra Satz, Co-Adviser
I certify that I have read this dissertation and that, in my
opinion, it is fully adequatein scope and quality as a dissertation
for the degree of Doctor of Philosophy.
Joshua Cohen
I certify that I have read this dissertation and that, in my
opinion, it is fully adequatein scope and quality as a dissertation
for the degree of Doctor of Philosophy.
David Victor
Approved for the Stanford University Committee on Graduate
Studies.
Patricia J. Gumport, Vice Provost Graduate Education
This signature page was generated electronically upon submission
of this dissertation in electronic format. An original signed hard
copy of the signature page is on file inUniversity Archives.
iii
-
iv
-
v
ABSTRACT
Two-thirds to three-fourths of future global greenhouse gas
emissions by 2030 are
likely to come from large developing economies. Their
participation in mitigating
climate change is imperative to achieving the target of
restricting global average
temperature increase to 2 degrees Celsius above preindustrial
levels that was set in the
Copenhagen Accord of December 2009. But a third of the world’s
poor live in one of
these economies, India. While there is much agreement among
scholars that climate
mitigation should not interfere with the humans’ ability to
enjoy a minimal standard of
living, there is little scholarship on how to carve out such an
“exemption” for poor
subpopulations within states in international climate change
mitigation agreements.
Further, there is little analysis in developing countries of how
the impacts of specific
mitigation policies would be distributed across the population.
This dissertation begins
to fill these gaps.
I analyze one class of issues related to the role of state
governments that would receive
such an exemption from mitigation for their poor, but who may
not be accountable in
international agreements for how such an exemption is
implemented. States influence
the poor’s emissions through policy choices and the institutions
that implement
policies. These policies and institutions affect the number of
people that are entitled to
an exemption, and whether they would actually receive its
associated benefits.
This work consists of three studies, the first two of which are
positive studies of the
income distributional impacts of potential climate mitigation
policies in Maharashtra,
India. The first study examines the impact on income
distribution of removing the
kerosene subsidy. The second study evaluates the leverage
electricity regulators have
over the distribution of mitigation burdens in the electricity
sector. The third study is a
normative assessment of the ethical and practical challenges of
implementing an
exemption for the poor in international climate mitigation
agreements.
In the absence of broad-based institutions for redistribution,
governments subsidize
essential consumption, such as food and household energy supply,
but with well
-
vi
known fiscal and environmental costs, including to climate
change. However, how the
benefits of the kerosene subsidy policy are distributed among
the millions of kerosene
users is not understood. This study formally examines these
benefits for different
income groups and the overall efficacy of kerosene subsidies as
a redistributive policy.
The study shows that households’ allocated quotas far exceed
kerosene demand in
rural areas, which encourages suppliers to divert kerosene to
other sectors. Urban
households, on the other hand, some of whose cooking budgets
would double without
the subsidy, supplement subsidized kerosene use with purchases
in the black market.
A better targeted subsidy in urban areas alone would avoid high
costs of the current
policy, yet avoid the impoverishment of urban users from their
complete removal.
The second study assesses the distributional impacts of
investing in low carbon supply
in Maharashtra’s electricity sector to the extent required to
meet the Indian
government’s pledge in climate negotiations to reduce the
economy’s carbon intensity.
I examine the regulator’s leverage in rate-setting over the
distribution of these
incremental costs across households. Using an economic
simulation model of the
electricity sector and household welfare, I assess the impacts
of economy-wide
electricity price scenarios under different political and
institutional constraints. The
analysis reveals that regulators can insulate low-income
households from welfare
losses without trading off aggregate welfare losses as long as
they can raise prices to
industry and high-income households. While feasible, this
pricing approach may be
politically unacceptable. Mitigation may also have a co-benefit
of reducing supply
interruptions to the poor. These results emphasize the
importance of qualifying
mitigation burdens by the internal policies and institutions on
which they depend.
The third study questions the adequacy of burden-sharing
proposals for climate
mitigation that advocate an exemption for the poor without
accounting for states’
agency over the costs and outcomes of such an exemption. How to
allocate the cost of
this exemption, however, can complicate international
agreements. Participating states
face moral hazards over the choice of future baselines of the
poor’s emissions. I show
- using India for illustration - that the financial stakes for
parties in how future growth
is distributed in India can be up to tens of billions of
dollars. I suggest that there is no
-
vii
clear moral basis to make benefiting states accountable for
minimizing the poor’s
emissions. Getting agreement on the terms of exemption may be
easier if benefiting
states adopt comparative benchmarks of accountability for the
poor’s emissions, but
which do not infringe on particular policy choices. Furthermore,
participating states
should share design agreements to ensure that the poor receive
the benefits of an
exemption.
-
viii
ACKNOWLEDGEMENTS
Dedicated to the memory of Stephen H. Schneider (1945-2010)
I am grateful to Steve Schneider, who was my primary advisor and
mentor, for
encouraging me to pursue this topic and imparting the meaning
and importance of
interdisciplinary research. After his passing, I have been
extremely fortunate to have
the close guidance, open-mindedness, patience and rigor offered
by Larry Goulder.
Joshua Cohen was instrumental in developing my ideas in global
justice. Debra Satz
gave me invaluable guidance and opportunities to present my work
to intimidating
audiences. I am grateful for David Victor for his pragmatism and
commitment to my
development and for financing my fieldwork through the Program
in Energy and
Sustainable Development. Paul Baer has been an informal advisor
and collaborator
whose work served as a launching point for mine. I thank Ashok
Gadgil and Kirk
Smith at Berkeley for useful conversations in the formative
stages. My conversations
with Girish Sant of Prayas Energy group kept me assured of the
policy relevance of
my work. I am indebted to Gayatri Gadag, Ravi Deshmukh and Dipak
Patil in Pune,
India for their help with data gathering and fieldwork. I thank
my research assistants,
Chris Bennett, Evan Woods and Allison Fink, for their
contributions to my analysis.
I am grateful to Pam Matson, Danielle Nelson and Helen Doyle for
the opportunities
and support I have had at E-IPER. My dissertation was made
possible through funding
from the Robert G. Kirby and Philip & Jennifer Arnold Satre
Fellowships. Its
successful completion would not have been possible without the
moral support of
Asha Ghosh.
-
ix
TABLE OF CONTENTS
CHAPTER 1 – CLIMATE CHANGE MITIGATION: A HUMAN RIGHTS
PERSPECTIVE
.................................................................................................
1
CHAPTER 2 – KEROSENE SUBSIDIES: WHEN ENERGY POLICY FAILS AS
SOCIAL
POLICY............................................................................................
27
CHAPTER 3 – DISTRIBUTIONAL IMPACTS OF CLIMATE CHANGE
MITIGATION IN INDIAN ELECTRICITY: CASE STUDY OF
MAHARASHTRA...........................................................................................
56
CHAPTER 4 – IMPLEMENTING AN EXEMPTION FOR THE POOR IN
INTERNATIONAL CLIMATE AGREEMENTS
.......................................... 87
CHAPTER 5 – CONCLUSIONS AND FUTURE RESEARCH
........................................... 111
Appendix A - Maharashtra Kerosene Quota Allocation
........................................................ 119
Appendix B – Detailed Results and Data Tables (Chapter 3)
................................................ 120
Appendix C - Long-Term Demand Response Model (Chapter 3)
........................................ 125
Appendix D - Household Survey Design
...............................................................................
129
-
x
LIST OF ILLUSTRATIONS
Figure 1: Basic Entitlements – Interpretations and Metrics in
Literature .................... 14
Figure 2: Per Capita Carbon Dioxide Emissions: Country Averages
and Internal
Distribution (Illustrative)
.............................................................................
19
Figure 3: Variation in Urban Household Kerosene Use - Mumbai,
Maharashtra and
India 2004-05
...............................................................................................
30
Figure 4: Cooking Fuel Shares by Income Decile – Maharashtra,
2004-05 ................ 31
Figure 5: PDS Kerosene Prices by Quantity Sold – Maharashtra
2004-05 .................. 34
Figure 6: Kerosene Use for Cooking/Water Heating – Maharashtra
2004-05 ............. 37
Figure 7: Kerosene vs. LPG Delivered Fuel Cost Comparison
(2004-05 prices) ........ 39
Figure 8: Population Share by Kerosene Subsidy Benefit –
Maharashtra 2004-05 ..... 43
Figure 9: Kerosene Subsidy Progressivity: Urban and Rural
Maharashtra 2004-5 ..... 44
Figure 10: Kerosene Subsidy Progressivity – Nandurbar District,
2004-05 ................ 45
Figure 11: Kerosene Subsidy Quotas and Actual Use (a) Urban
................................ 46
Figure 12: Households with Insufficient Kerosene Quotas –
Maharashtra 2004-05 ... 48
Figure 13: Subsidy Price by Purchased Quantity – Maharashtra
2004-05 ................... 50
Figure 14: PDS Kerosene Prices, Quantities, and Transport
Distances by District ..... 51
Figure 15: Electricity and Welfare Model Simulation Approach
................................ 58
Figure 16: Electricity Block Tariff – Maharashtra State
Electricity Board, 2004-05 .. 66
Figure 17: Residential and Industrial Price Impact Comparison
.................................. 75
Figure 18: Industry Group Electricity Intensities (2003-2004)
.................................... 76
Figure 19: Household Expenditure Electricity Intensity – by
Income Group .............. 76
Figure 20: Average Residential Prices - Low Carbon Pricing
Scenarios ..................... 79
Figure 21: Distribution of Welfare Losses - Low Carbon Pricing
Scenarios ............... 80
Figure 22: Distribution of Welfare Losses - Energy Efficiency
Sensitivity for the
Economic Efficiency Scenario
.....................................................................
83
Figure 23: Distribution of Welfare Losses - Energy Efficiency
Sensitivity for the
Equity Scenario
............................................................................................
84
Figure 24: Country Intranational Income and Emissions
Distribution: 2007 .............. 93
-
xi
LIST OF TABLES
Table 1: Kerosene Consumption by Region and Market: Maharashtra
2004-05 ......... 32
Table 2: Household Kerosene Use by Function, Region and Priority
......................... 36
Table 3: PDS Kerosene Price Discrimination Model Results
...................................... 53
Table 4: Optimal Prices – Baseline and Low Carbon Scenarios
.................................. 77
Table 5: Welfare Metrics – Baseline and Low Carbon Scenarios
................................ 78
Table 6: Optimal Prices – Industrial Elasticity Sensitivity
.......................................... 81
Table 7: Welfare Metrics – Industrial Elasticity Sensitivity
........................................ 82
Table 8: Optimal Prices - Energy Efficiency Sensitivity for the
Economic Efficiency
Scenario
.........................................................................................................
84
Table 9: Optimal Prices - Energy Efficiency Sensitivity for the
Equity Scenario ....... 84
Table 10: Exemption Costs Under Alternate Development Paths in
India .................. 94
Table 11: Informal Economies in Developing Countries
........................................... 103
-
1
CHAPTER 1 – CLIMATE CHANGE MITIGATION: A HUMAN RIGHTS
PERSPECTIVE
1 Introduction
Global climate change has been called the “perfect moral storm”
(Gardiner 2010).
This is for good reason. Given the unprecedented nature of
climate change, humans
must confront a number of ethical challenges at once. Greenhouse
gases (GHG)
emitted from sources across the globe have accumulated in the
atmosphere over
centuries, to a large extent without our knowledge, causing
unintended and potentially
catastrophic impacts on people and biodiversity in the future.
While these emissions
come largely from industrialized societies, their long-term
impacts fall
disproportionately on poor societies, through weather-related
events that are not easily
traced to their causes, and with considerable scientific
uncertainty surrounding their
severity and timing. Furthermore, that GHG arise primarily from
the use of an
essential input – energy - into most human activity implies that
to minimize the effects
of climate change societies would have to undergo shifts in
infrastructure as well as in
lifestyles at an unprecedented scale, scope and pace. How will
we compromise to
distribute this responsibility across countries? No matter what
or how much action
humans take to combat climate change, these actions will have
moral repercussions,
for people in the future or today, and mostly likely for
both.
The climate change problem presents challenges for
decision-making at the individual,
state and international level. The complex causal chain between
human activity and
climate impacts and the separation in time and space between
emitters and victims
makes it hard for individuals to acknowledge the threat of
climate change, let alone to
modify their behavior (Swim 2009). At the state level, modern
societies‟ reliance on
centralized sources of emissions (“collective emissions”) such
as power plants and
industrial facilities and dispersed sources such as automobiles
necessitate the
establishment of economy-wide institutions and policies to
enable emissions
reductions. At the international level, because the atmosphere
serves as a “global
commons” into which all humans‟ GHG emissions accumulate the
worst impacts of
-
2
climate change cannot be mitigated by the actions of any one
country alone.
Combating climate change requires cooperation among countries,
each of which have
different interests in climate change policy, states of
development and political power.
A critical concern is that the poorest populations in the world
contribute the least to,
have the most to lose from, and have the least bargaining power
to influence
negotiations on, climate change (Sagar 2001).
As such, global climate change is arguably the most complex
international policy
challenge facing humanity. After over three decades of
scientific research, almost two
decades of international negotiations, and the publication of
four reports by the
Intergovernmental Panel on Climate Change (“IPCC”), over 120
countries - including
the United States, European Union members, China and India -
signed the
Copenhagen Accord. This non-binding agreement symbolizes the
overwhelming
acceptance by state governments of human interference with
climate.
We agree that deep cuts in global emissions are required
according to
science, and as documented by the IPCC Fourth Assessment
Report
with a view to reduce global emissions so as to hold the
increase in
global temperature below 2 degrees C, and take action to meet
this
objective consistent with science and on the basis of
equity.
Copenhagen Accord, December 2009
This agreement symbolizes an acknowledgement of our
responsibility to prevent
imposing on future generations the harmful effects of climate
change that would be
avoided with a maximum temperature rise of 2 degrees centigrade
(“2C”).1
To achieve this target, the IPCC indicates that cumulative GHG
emissions in the 21st
century would have to reduce from a projected average of 670
gigatons of carbon
(“GtC”) to 490 GtC (IPCC 2007). Based on current trends
(“business as usual”, or
“BAU”), this implies that by 2030 annual global GHG emissions
would have to
reduce to about half the levels that would otherwise occur
(Project Catalyst 2009). The
pledges made by state governments in the Cancun Agreement of
December 2010
1 Stabilizing GHG concentrations in the atmosphere at 450 parts
per million (ppm) would keep the odds
of increasing average global temperature since pre-industrial
levels by 2 degrees Celsius (“2C”) below
50 percent.
-
3
amount to only 60 percent of the mitigation levels required to
achieve this target.2
Indeed, it has been estimated that even in the best case
scenario that all the pledges are
implemented, we are still virtually certain to exceed the 2C
target (Rogelj, Hare et al.
2009).
Perhaps the most divisive issue in international climate
politics has been the issue of
how to distribute between developed and developing countries the
burdens of reducing
GHG emissions (“mitigation”) further to meet the 2C target. The
term „common but
differentiated responsibilities‟ (“CBDR”) in the UN Framework
Convention for
Climate Change (UNFCCC) symbolizes the intent of the signatories
to distribute the
burdens of responding to climate change so that industrialized
countries would take
the lead in reducing emissions, while the less developed
countries would give priority
to their development but aim to integrate climate change
concerns in the future.3 In the
climate ethics literature scholars have interpreted CBDR in many
ways, but generally
place greater responsibility for mitigation on industrialized
countries (Gardiner 2004;
Gardiner, Caney et al. 2010). A meta-analysis of different
burden-sharing proposals
suggests that under most views of equitable burden-sharing
developed countries as a
whole need to reduce their emissions by 25-40 percent below 1990
levels, while
developing countries together would have to reduce emissions by
15-30 percent below
their BAU levels (den Elzen and Höhne 2008) by 2030.4 However,
even though such
equity formulations have been proposed in negotiations since the
first Conference of
Parties (“COP”) in 1995 (Ringius, Torvanger et al. 2002),
developed and developing
countries have failed to have a dialogue about these equity
principles, let alone reach
any agreement on how to allocate mitigation responsibility.
Although equity principles may never drive political
negotiations, the need for
agreement on burden-sharing rules that are perceived as fair is
compelling. The
2 Christine Figueres, UN Secretariat, December 20, 2010,
Reuters.
3 Articles 2 and 3.1of the UNFCCC. The Convention has been
ratified by193 countries, including the
United States. 4 Recent proposals from scholars in China and
India calculate mitigation obligations based on a per
capita entitlement to cumulative historical emissions extending
back to the early 20th
century. These
proposals would exempt China and India from any mitigation
obligation until after 2030.
-
4
cooperation of large developing countries is essential to reach
the 2C target, since
almost half the future energy demand growth by 2030 is expected
to come from just
China and India (International Energy Agency 2008). But both
countries‟ governments
are unlikely to agree to mitigation obligations that are
perceived as demonstrably
unfair. Thus, it is also of political interest to assess
fairness as one lens of burden-
sharing, while also giving attention to how to make these
principles practicable and
how they affect the incentives for participation by state
governments.
In the climate ethics literature, with few exceptions, scholars
treat states as moral
agents, by defining mitigation obligations for them based on
aggregate indicators, such
as GDP or emissions, but without accounting for heterogeneous
national
circumstances. This characterization would either exempt or
impose mitigation
obligations on large developing countries like China and India
without regard for the
different segments of society that have vastly different levels
of development but
which in aggregate terms account for comparable emissions. This
has both ethical
problems and risks being politically untenable. Even the few
exceptions where
scholars account for internal income inequality, states are
treated as passive agents.
In this dissertation, I examine the implications of defining and
implementing
mitigation obligations for subpopulations within developing
countries. I explore the
ramifications of adopting one morally compelling principle of
exempting the least
advantaged in society from the burdens of mitigation as a
minimal basis for
distributing mitigation responsibilities. The central argument
explored in this study is
that sovereign state policies influence the distribution of
mitigation burdens related to
such an exemption among states. This raises new questions about
the role of states that
receive such an exemption in climate agreements. The object of
this study is both
positive and normative: on the positive side, I evaluate how
state policies and
institutions influence the distributional impacts of climate
mitigation policies using
India as a case. On the normative side, I evaluate the
accountability of parties to an
agreement to achieve its objectives in light of these sovereign
influences. I also assess
the issues that arise in getting agreement on the terms for such
an exemption in
international agreements.
-
5
The remaining part of this chapter describes in more detail the
motivation for choosing
this principle, the gaps in literature associated with
operationalizing it, and how the
subsequent chapters contribute towards filling these gaps. This
dissertation makes
other independent empirical contributions to policy literature
regarding the impact of
energy policies on income distribution in India. These are also
described below.
2 Scope of this Study
2.1 Focus on Intra-generational Equity
For convenience, scholars make a distinction between the ethical
question of what
level to stabilize global temperature, from the related but
distinct ethical problem of
how to distribute the sacrifices required to achieve this given
level of temperature
stabilization among the current generation of people.5 The first
is an inter-
generational equity issue that is viewed primarily as a question
of how to balance the
interests of future and present generations of people. Because
of the lagged effects on
climate of accumulating GHG in the atmosphere, global
temperature would not
stabilize until after 2050, so only future generations would
benefit from abatement.
But meeting this target would require making sacrifices today to
shift to a low-carbon
infrastructure and prevent the „lock-in‟ of carbon-intensive
sources of GHG, which
typically have lives of 10-50 years. Strictly speaking, future
generations may have to
make sacrifices to maintain a steady-state level of GHG
emissions. However,
maintaining safe levels of emissions in perpetuity may be less
burdensome once
society has invested in low carbon technologies and climbed the
learning curve of
putting in place and adjusting to a low carbon economy.
5 This separation is driven by practical considerations rather
than moral ones. The same moral principles
apply in resolving both inter-generational and
intra-generational issues. The choice of a particular
climate stabilization level may well have a bearing on the
choice of an equitable distribution of
mitigation burdens today, and present moral trade-offs. If, for
example, we chose to stabilize emissions
at 350 ppm, future generations would face less severe burdens
from climate impacts, but the increased
mitigation requirements would, ceteris paribus, impose greater
risks of burdens on the poor today. Both
issues therefore would merit simultaneous consideration purely
on moral grounds.
-
6
For the purposes of this work, I take this inter-generational
equity issue as resolved, at
least in aspiration. There remain questions on the one hand
about whether 2C is
enough,6 and on the other hand whether 2C is too burdensome or
even feasible.
Without knowing the level to which states actually commit, what
matters to this work
is that the target is ambitious enough that large developing
countries‟ participation is
unavoidable for its attainment. The practical implications of
this assumption are
discussed later. I focus strictly on the intra-generational
problem of how to distribute
the costs of meeting such a target.
2.2 Focus on Mitigation
In the climate justice literature, the intragenerational
burden-sharing issue addresses
distributive justice for both mitigation and adaptation, and
often together. The
distribution problem for adaptation is concerned with how to
distribute the costs of
compensating victims of unavoidable climate change impacts.
These impacts are
likely to affect the lives and livelihoods of millions of
vulnerable populations in the
equatorial belt who do not have the means to adapt to these
changes. However,
mitigation is a separate ethical problem from adaptation, even
if common principles
may apply in their resolution (Vanderheiden 2009).7 I focus only
on the mitigation
problem.
2.3 Focus on Energy-related Carbon Dioxide (CO2) Emissions
I focus specifically on the distribution of burdens associated
with mitigating carbon
dioxide (CO2), which is the GHG with the longest life, fastest
growth, and as a
consequence the greatest cumulative contribution to climate
change.8 With the
6 While the 2C target has become the marker for dangerous
interference, this is a subjective judgment
that was ultimately determined by political rather than
scientific or moral considerations. The
governments of small island states in the Asia Pacific advocate
for a 1.5C threshold to avoid the risks of
sea level rise on these islands. 7 For example, a stronger case
can be made for counting historical emissions in determining
peoples‟
liability for compensating victims of climate change than the
case that can be made for counting
historical emissions in determining mitigation burdens, even
though reasonable arguments have been
made in support of historical emissions in both distribution
problems. See Vanderheiden, S. (2009).
Atmospheric Justice: A Political Theory of Climate Change. New
York, Oxford University Press. 8 Carbon dioxide accounts for 60
percent of the total increase in radiative forcing, compared to
20
percent from methane. This is because although CO2 has a lower
global warming potential (1:25), it has
-
7
exception of Brazil, CO2 emissions account for two-thirds or
more of the GHG
emissions of the 20 largest contributors to global GHG
emissions. Over 90 percent of
CO2 emissions come from burning fossil fuels, which makes the
task of reducing
emissions intricately tied to how we use energy.
3 Support for Defining a Moral Minimum
The notion of exempting a certain category of people from
mitigation burdens derives
from a human rights principle of a universal human entitlement
to a minimal set of
„goods‟. This view finds support in the climate ethics
literature, but has its roots in the
broader global distributive justice and has supporting
international institutions. I
discuss these motivations below.
Basic rights are the morality of the depths. They specify the
line
beneath which no one is allowed to sink.
Henry Shue (1999), Basic Rights.
The notion of a moral „minimum‟ occupies considerable space in
annals of political
philosophy. Henry Shue‟s statement above characterizes the
spirit of a moral
threshold. What does respect for human dignity demand that all
humans have
regardless of their culture, nationality, location in history or
place? More importantly,
who has duties to uphold these claims? Of interest here and in
the global distributive
justice literature are the types of duties that such claims
raise for states towards people
in other states. Two types of duties of external states are
typically contemplated: duties
to help realize people‟s claims and to protect them from
infringement by their own
state governments (“positive duties”); and duties to respect, or
not infringe on, other
states‟ abilities to fulfill these rights for their own people
(“negative duties”). While
there is considerable debate among philosophers as to whether
universal human rights
exist at all, the existence of negative duties is relatively
less controversial than positive
duties. That is, notwithstanding the debate over what
entitlements deserve immunity,
a longer lifetime (5-200 years: 8-12 years).CO2 emissions are
growing at over 1 percent per annum
globally, and at almost 5 percent in developing countries, while
methane emissions are constant or
declining (World Resources Institute, cait.wri.org).
-
8
there is little justification, except under extenuating
circumstances, for knowingly
infringing on others‟ basic entitlements.9 As discussed above,
the goal of establishing
a moral minimum for mitigation agreements is to apportion
mitigation responsibility
in a manner that does not infringe on the poor‟s basic rights.
Thus, in this context,
where what is at issue is only whether parties to a mitigation
agreement ought to
respect such a moral threshold, a moral minimum has considerable
theoretical appeal.
3.1 Human Rights International Institutions
A number of international institutions exist that provide some
legal recourse for
international human rights violations. The most relevant one for
the purposes of
climate change mitigation is the International Covenant on
Economic, Social and
Cultural Rights (ICESCR). The ICESCR has been signed by all the
major economies
including the United States, China, European Union, India,
Brazil and South Africa.
The ICESCR includes the right of people to be “free from hunger”
(Article 11.2), the
right to the “enjoyment of the highest attainable standard of
physical and mental
health” (Article 12.1), and to the “right of everyone to an
adequate standard of living
for himself and his family, including adequate food, clothing
and housing, and to the
continuous improvement of living conditions” (Article 11.1). The
Covenant obligates
Parties to “at the very least...ensure the satisfaction of
minimum essential levels” of
economic, social and cultural rights.10
The value of the ICESCR is to lend moral and political support
to the justice concerns
posed by mitigation burdens, rather than to provide a legal
mechanism to uphold
claims related to mitigation. Making a legal case for a human
rights violation from
climate change is fraught with controversy and challenges, such
as demonstrating
causation in proving injury, and identifying defendants.
Particularly with climate
mitigation, assessing rights violations requires an evaluation
of development policies,
which international law does provide a clear means of doing
(Humphreys 2010).
9 Such extenuating circumstances may be where certain rights
have to be infringed upon in order to
prevent other, more serious, rights violations. 10
ICESCR, General Comment 3, The Nature of States Parties‟
Obligations, UN Doc. E/1991/23 (Dec.
14,1990), ¶ 10, cited in Bodansky (2010).
-
9
However, if mitigation burdens can be credibly framed as a human
rights issue under
the ICESCR, this can be used to capture public opinion (Bodansky
2010). Politicians
would be hard pressed to publicly oppose a position that
advocated respecting human
rights, particularly one that demands only forbearance from
causing harm. Indeed,
politicians have de facto supported this principle by not
demanding the participation of
the poorest countries in mitigation agreements (Ringius,
Torvanger et al. 2002). In
comparison, the principle of historical responsibility, which is
one of the contentious
equity principles that divides developed and developing
countries, carries sufficient
moral ambiguity on grounds of past ignorance to climate change,
that US negotiators
have publicly dismissed the principle outright.11
3.2 A Unifying Thread in Climate Mitigation Burden-Sharing
Due to the unique nature of global climate change, the
mitigation burden distribution
problem has been viewed through many lenses of fairness. Each of
these views is
founded on different principles and, when operationalized, can
lead to different
burden-sharing outcomes. A moral minimum represents a relatively
uncontroversial
principle that constrains, but is consistent with, the
application of these other
distributive principles.
The mitigation problem has been viewed at once as one of
distributing: scarce
resources (to the atmosphere); responsibility (for pollution);
and burdens (primarily,
but not only, financial). In the resource-sharing view, global
climate change puts limits
on the amount of CO2 emissions that humans can safely emit. The
carbon absorption
capacity of the earth can be thought of then as a “global
commons” with scarce
capacity that has to be distributed, so that people have an
equal right to pollute up to
the total safe level of CO2 emissions (Agarwal 1991; Jamieson
2001; Vanderheiden
2009). States therefore have rights to emit GHG in proportion to
their population share
of the maximum allowable global emissions (the issue of state
vs. individual
entitlements is discussed later).12
States then have a responsibility to mitigate all but
11 “U.S. Negotiator Dismisses Reparations for Climate”, New York
Times, December 9, 2009.
12 In practice, emissions rights could represent a financial
claim to the value of the rights that can be
traded rather than used.
-
10
their entitled emissions. This view mirrors other moral theories
that view global
economic inequality as arising primarily from an arbitrary and
unequal distribution of
natural resources (Beitz 1979). While drawing some support from
such theories, the
per-capita emissions view seems to represent a kind of resource
fetishism, by
assigning value to what is only a means to an end (Caney 2009).
Further, as the per
capita emissions allocation is derived from a scientific limit,
there is no guarantee that
this allocation protects any fundamental human interest (Hayward
2007), particularly
since the benefits that flow from it – cheap energy - vary
widely based on people‟s
fuel endowments. Thus, this view begs the question of why an
international climate
burden-sharing regime should isolate this one resource for
redressing unequal access,
when the benefits that such a right would provide – energy – can
be obtained from
other resources, such as solar energy.
The other two approaches are both principles of
proportion,13
reflecting the objective
of treating “comparable people comparably”. Mitigation costs
should be apportioned
based on either the responsibility for causing the problem, or
the capacity to bear the
burden of the problem. In the responsibility-based approach,
since global warming
increases roughly in linear proportion to cumulative CO2
emissions, those who emit
more are more at fault for causing global warming, and therefore
ought to bear
proportionately more responsibility for mitigation (“polluter
pays”). This view is
backward-looking, since past and present emissions of a
particular GHG that
accumulate in the atmosphere contribute equally to climate
change.14
However, the
responsibility approach is problematic when applied to
individuals rather than states,
because historical emissions are not obviously the
responsibility of individuals who
happen to have been born in the same state (Caney 2010). This
view also penalizes
states for their fuel endowments by putting at risk the basic
needs of people who can
13 As in Aristotle‟s dictum: “what is just is what is
proportional, and what is unjust is what violates that
proportion”. Aristotle, Nicomachean Ethics: Book V: Ch.3. 14
This responsibility may extend only as far back as the point in
time prior to which people can claim
that ignorance about climate change absolves them of any
liability for the harm caused by prior
emissions.
-
11
ill afford to bear mitigation burdens, and who may have no
control over their
emissions.15
The capacity view, on the other hand, is forward-looking. People
should bear
mitigation costs in proportion to some measure of their “ability
to pay”, such as
income or wealth. This is because CO2 emissions derive from the
combustion of fossil
fuels, whose use is ubiquitous in the global economy. Mitigation
therefore imposes
burdens on people by inducing lifestyle changes, shifts to more
expensive low carbon
energy sources, or emissions reductions from existing energy
sources. Since people
ultimately care about their welfare and not emissions, the
problem should be viewed
as one of distributing the burdens on human welfare that arise
from mitigation.
However, the capacity view penalizes those who may have low
emissions due to
efficient energy use or less carbon-intensive lifestyles.16
And more importantly here,
the capacity approach gives priority to the poor, but doesn‟t
proscribe imposing
mitigation burdens that may cause harm.
Both the principles of proportion thus have their respective
merits and problems
(Caney 2010). These tradeoffs are somewhat irreconcilable, and
subject ultimately to
political, rather than ethical, resolutions.17
But all three approaches to different degrees
risk imposing mitigation responsibility on the poor. This gives
cause to define a
minimum inviolable threshold of human well being upon which
mitigation burdens
should not impinge (Shue 1999; Caney 2009). This threshold thus
serves to
circumscribe the scope of applying burden-sharing principles,
rather than provide an
alternative to them. Note that these distributive principles are
also consistent with a
moral minimum – any of them can be applied to divide mitigation
burdens among the
emissions and people that are not exempt by virtue of this
minimum.
15 See Section 3.1 - Brazil and South Africa have almost
identical average GDP and income inequality,
but Brazil has a quarter of South Africa‟s carbon dioxide
intensity (and half its GHG intensity) due to
their reliance on hydro and coal respectively. Imposing
responsibility-based mitigation burdens on both
countries imposes greater risks to the interests of South
Africa‟s poor population. 16
For example, the United States and Hong Kong have comparable
average GDP, but Hong Kong has
almost 40 percent lower energy and carbon intensity (which
therefore is not due to fuel endowments). 17
One burden-sharing approach in fact combines both in proportion
that can be chosen by decision-
makers (Baer et al 2008).
-
12
Thus, a moral threshold can be thought of as a unifying and
morally compelling thread
that runs across these burden-sharing approaches. Below I expand
on how this
principle serves as a minimal principle of distribution, before
I turn to the content of
such a threshold.
3.2.1 What does ‘minimal’ mean?
There are many senses in which a moral threshold of exemption is
minimal in a
mitigation burden-sharing framework. The most important sense is
that this principle
has lexical priority over all other distributive principles that
are invoked to determine a
fair allocation of mitigation burdens (Caney 2009). As discussed
above, this threshold
is inviolable, and therefore absolutist, 18
in the sense that under no circumstances
should people bear any mitigation burdens that compromise these
rights.
The second sense in which a moral threshold is minimal is that
it alone is not
sufficient to specify a complete allocation scheme for
mitigation burden-sharing. This
principle only defines an exemption from mitigation, but says
nothing about how
burdens should be distributed among those who are not exempt.
One might think of a
„default‟ approach that eschews any principles of „fair‟
distribution, and treats all
present and future emissions equally. In such a case, states
and/or individuals would
reduce their own emissions based on causation, or
proportionately. If the exemption
were incorporated as the only equity consideration, the
resulting allocation scheme can
be thought of as the least demanding claims that can be made by
developing countries,
in principle. As noted earlier, the content of this claim may
still be financially
demanding, depending on the definition of such an exemption.
Lastly, because of its theoretical appeal and importance,
respect for a moral threshold
is the least of justice considerations that a global climate
mitigation regime ought to
incorporate, if it has to eschew all other considerations. When
viewed with this
perspective, a moral threshold for exemption from mitigation may
serve as a common
18 To be clear, it is not absolutist in the context of all human
rights. That is, basic needs do not
necessarily have lexical priority over the right to life and
security.
-
13
ground for agreement between major economies, if cooperation
between developed
and developing countries depends on the incorporation of justice
concerns at all.
I now turn to the content of an exemption threshold.
3.3 Moral Threshold – Equality of What, and How Much?
An inviolable moral threshold “in principle” risks being facile
without some claim
about its content. Even the practical significance of an
exemption threshold in a
climate agreement rests on how expansive a guaranteed minimum
ought to be, rather
than on whether one ought to exist at all. At the same time, the
„Equality of What‟
debate – what features of the human condition should count in
basic entitlements - is
larger than the domain of climate change. This study does not
engage in that debate.
However, it is possible to provide at least reasonable bounds
for a moral threshold for
which there is sufficient support in literature, keeping in mind
that this support need
only apply to the negative duty of non-infringement.
The literature on what constitutes fundamental human
entitlements is extensive.
Definitions vary in how they have been conceptualized and in the
range of their
constitutive elements (Figure 1). Among philosophers, in their
„thinnest‟ form it has
been argued that a minimal set of universal rights comprises
rights to only physical
security. Henry Shue defines subsistence rights as additionally
comprising economic
security, on which human‟s survival also depends (Shue 1980).
Shue argues that these
basis subsistence rights have lexical priority over all other
types of rights, because
other rights cannot be enjoyed by people unless they can
subsist. In the climate ethics
literature, Shue indirectly supports a subsistence-based moral
threshold by
differentiating „subsistence emissions‟ from luxury emissions
(Shue 1993).19
A number of scholars view human rights as instrumental to
ensuring a quality of life
beyond mere subsistence, but do not dwell on the meaning of
„decency‟(Buchanan
19 Simon Caney (2010) points out that Henry Shue does not use
the language of rights despite being an
advocate of basic rights. This may be to avoid alienating
opponents of human rights formulations in
general. However, the content of his formulation of subsistence
emissions derives unambiguously from
subsistence needs, even if he frames them in terms of
emissions.
-
14
2004; Hayward 2007; Caney 2009). Simon Caney, more recently,
casts human rights
as claims to a decent standard of living, or to the emissions
that are required to enjoy
decent living conditions. Nussbaum and Sen conceive of basic
entitlements as
constituting capabilities or opportunities for human function
rather than a set of goods
or services.
Figure 1: Basic Entitlements – Interpretations and Metrics in
Literature
Note: Marks indicate the metrics used for different definitions
of entitlements in literature.
There is also a viewpoint that frames basic entitlements as
development rights, which
include economic, social and political entitlements to
individuals. This approach is
distinguished by its emphasis on a collective entitlement to a
process of change that
leads to the progressive realization of the individual rights
(Andreassen 2006).
Although development rights are a relatively undeveloped and
unexplored concept, the
exploration of the importance of collective processes for
realizing individual rights is
unique and important. The problem is that there is little
theoretical guidance as to what
such a collective entitlement should entail, and how much energy
(and by extension
emissions) that entitlement should include.
Among this range of definitions, it is plausible to infer that
subsistence needs are a
common and minimal set of basic rights. No other functions or
liberties can be
enjoyed without subsistence. Then, subsistence can be thought of
the most minimal
candidate for a moral minimum. However, what additional
entitlements beyond
subsistence should count is an open question. Broader
entitlements related to political
-
15
rights or capabilities are relatively abstract. However, there
is wide support for
including some measure of decent living standards, even if their
specification is
subject to some bargaining.
A few burden-sharing proposals incorporate an exemption for
emissions within
countries. They indeed demonstrate the wide range of possible
interpretations of a
moral minimum. Muller et al and Chakravarty et al estimate that
a 1 ton per capita
threshold provides for subsistence needs. On the more inclusive
side, Baer et al in
Greenhouse Development Rights base their exemption threshold on
an entitlement to
development rights, which they quantify at $20/day per capita.
This represents the
type of „upper bound‟ mentioned above that has some normative
appeal but is hard to
justify as a specific number (Baer 2008).
It is therefore important, but difficult, to objectively define
a more expansive upper
bound for a reasonable threshold than mere subsistence. Though
seemingly obvious,
this observation has serious implications for how an exemption
threshold might be
implemented in practice. Without a normative basis, such a
threshold may have to be
determined through political negotiations. Under this
assumption, I set aside the matter
of the level of the threshold in this study.
I turn next to how to measure such a threshold.
3.3.1 Metric for a Moral Minimum
Income is a frequently used and a practical choice for a metric
for an international
exemption threshold, despite its known limitations as an
indicator of human well-
being (Figure 1). There is a well-known tension between
inclusiveness and their
measurability in defining an indicator of human well being and
its constituents.
Indicators of subsistence needs, such as calorie intake, are
reasonably accurate. The
Human Development Index (HDI) comprises income, life expectancy
and literacy, but
the latter two are crude proxies for health and education. Among
the more inclusive
measures is the Multi-dimensional Poverty Index (MPI), which
contains 10 indicators
of health, education and living standards, including access to
water, sanitation and
-
16
electricity.20
However, the added dimensions complicate measurability and
inter-
personal comparisons. For instance, the electricity access
indicator does not capture
actual energy supply, which is a more accurate indicator of the
quality of life
conferred by electric service. Further, a composite index
implicitly allows trade-offs
between potentially incommensurable indicators, such as life
expectancy and
electricity access. Such indices, while more inclusive, present
a potentially misleading
standard of objectivity in comparing living situations in
different cultural and
geographic contexts.
While the development of these indicators has been motivated to
a large extent by the
limitations of income as an indicator of living standards, they
have yet to match the
benefits of income as a commensurable indicator across
countries. Income provides a
universal measure by which to assess the value of disparate
goods and services that
reflect people‟s living standards, by measuring the costs of
these services in market
transactions. Where market transactions fail to accurately
reflect people‟s living
conditions, techniques have been developed to impute prices for
non-market services
and externalities, and to adjust price indices to reflect
different purchasing powers
(Stiglitz 2008).
As a practical matter, income also correlates with these
multidimensional indicators, at
least at low levels of development, and with emissions. Income
exhibits a relatively
predictable correlation to carbon dioxide emissions at an
aggregate level, both across
and within countries. A National Academy of Science paper shows
that emissions
elasticities of income across and within countries range from
0.7-0.9. With multi-
dimensional indicators, the body of research on these metrics is
too limited to provide
either an understanding or actual data of their relationship to
emissions in different
countries.
To summarize, exempting the emissions associated with human‟s
most basic living
conditions from people‟s mitigation obligations is a morally
compelling justice
20 The Oxford Poverty and Human Development Initiative
(http://www.ophi.org.uk/policy/multidimensional-poverty-index/)
http://www.ophi.org.uk/policy/multidimensional-poverty-index/
-
17
principle that ought to have priority over other principles that
might govern how
mitigation costs should be distributed. For a threshold to have
moral and practical
significance in an international climate change agreement but
not escape credulity as a
minimum, such a threshold must be defined to include more than
mere subsistence but
not extend beyond standards for decent living conditions.
However, justifying a
particular basket of goods or a threshold level as a moral
minimum is not important for
the purposes of this study. For the purposes of policy, the
metric of income provides a
reasonable compromise in quantifying a basic minimum between the
needs of
inclusivity and flexibility on the one hand, and the need to
translate such a threshold
into an emissions exemption. I now turn to the challenge
associated with
operationalizing a moral threshold in a climate mitigation
regime.
4 Limitations of a Statist View of a Moral Minimum
The principle of a moral minimum challenges presumptions in
climate ethics and
policy about the beneficiaries and agents of burden-sharing
arrangements. With few
exceptions, the discourse on climate equity assumes states to be
agents of a burden-
sharing arrangement, not just as the parties who would implement
the terms of an
agreement, but the agents whose interests the agreement would be
designed to protect.
This assumption has also been made among scholars who think that
burden-sharing
arrangements should allow for poor peoples‟ development needs
(for instance, by
exempting poor countries from a climate agreement).
As such, scholars have predominantly proposed exemption
thresholds that apply to
states. These entail the notion of „graduation‟ – that countries
abstain from taking
steps to reduce emissions until they reach some threshold
indicator, such as average
GDP or average per capita emissions, or a combination of the two
(Höhne 2006;
Frankel 2007; Michaelowa 2007).
This statist approach to climate equity is compelling from a
practical standpoint. First,
it may be a reasonable simplification to account for disparities
in wealth across
countries if the inequalities within countries are of a second
order. Second, it may be
-
18
prohibitively burdensome to account for and verify internal
socioeconomic indicators
in structuring and implementing agreements.
These considerations are important, if true. However, they are
as yet unexplored. How
much would a statist approach distort the distribution of
burdens, and what is
financially at stake? For large developing economies that have
comparable total
emissions from both wealthy and poor populations, the case for a
statist approach
seems weakest. Consider, for example, the emissions of the
populations earning below
and above the average GDP in China, India, Brazil and South
Africa (“BASIC
countries”), who together contributed 60 percent of non-Annex 1
countries‟ GHG
emissions in 2005. In aggregate terms, the CO2 emissions from
those who earn more
than the countries‟ average GDP are comparable, with those
earning less contributing
over 40 percent of total emissions in all these countries.21
But in per capita terms, the
higher income group comprises a third or more of the population
but have average per
capita emissions that are more than double that of the
population who earn less than
average GDP. This does not even represent the full spread of
emissions (and income)
inequality within these countries. In India, for example, in
2003-04 the top 3 percent
of the population had CO2 emissions of at least 4 tons per
capita, which was four times
the country average.22
If, hypothetically (Figure 2),23
a burden-sharing regime were established where
countries that had average per capita emissions below the world
average (~4.5 tons)
were exempt from mitigation, Brazil and India would be exempt,
but approximately
50 million people in Brazil and 35 million people in India who
emit more than the
21 This figure has been calculated on the basis of an
empirically validated observation that income
distribution best resembles a log-normal distribution, which is
uniquely specified by the average GDP
and Gini coefficient. See Chapter 4 for details on this. Data
sources include: for income and population,
International Monetary Fund (2008); CO2 emissions, US Energy
Information Administration (2008);
Gini coefficient, World Bank (2005) 22
According to Parikh et al (2009), ten percent of the urban
population in India (~360 million) has
average carbon dioxide emissions of about 4 tons/capita.
However, this is an underestimate, because
these data are based on household consumption expenditure, which
the upper income groups are known
to understate in surveys. 23
Variations of this rule have been proposed in the literature.
See Höhne, N., Michel den Elzen, and
Martin Weiss (2006). "Common but differentiated convergence
(CDC): a new conceptual approach to
long-term climate policy." Climate Policy 6: 181-199.
-
19
world average would also be exempt. On the other hand, since
China‟s emissions
exceed the world average, even with such a general exemption
level 800 million
people in China would risk being exposed to mitigation burdens
that the country
would have to adopt if mitigation efforts were not well
targeted.
Figure 2: Per Capita Carbon Dioxide Emissions: Country Averages
and Internal Distribution (Illustrative)
Note: Country emissions distributions show population ordered in
increasing per capita CO2
Data Sources – See Footnote 21.
Thus, it seems that for large developing economies in
particular,24
it would be
unreasonable to either exempt the entire country from mitigation
obligations, or to
impose mitigation burdens without adjusting for an exemption for
the poor. For these
countries, a statist approach would either arbitrarily impose
mitigation burdens on
some people because of their location or ignore important risks
imposed by higher
energy prices and other mitigation burdens on poor
populations.
24 This issue is not unique to developing countries.
Approximately 10 percent of the US population is
below the official poverty line, which implies that they have
low or even negative disposable income.
Their living conditions may be better than the poor in
developing countries, but they may not be in
position to bear any mitigation burdens.
Per
Ca
pit
a C
O2
(to
ns)
35 million
800 million
China
USA
EU (27)
Russia
India
Japan
Brazil
Middle East
Africa0
5
10
15
20
25
30
World Average
-
20
5 Implementing a Mitigation Exemption within States –
Contributions of this Study
Imposing mitigation obligations on states that are intended for
only subpopulations
raises new positive and normative questions related to the
implementation of these
obligations. On the positive side, since mitigation obligations
are intended for only
those above a poverty threshold, to what extent would mitigation
impacts fall on the
poor under existing institutional conditions within developing
countries? This
dissertation examines specifically the influence of state
policies over the distribution
of these mitigation impacts using two cases in India. These
studies are the focus of
Chapters 2 and 3, and are discussed in more detail below
(Section 5.1). The fact that
state institutions are not accountable under international
agreements but can influence
the outcomes of these agreements raises normative questions
about what obligations
states that receive an exemption should have in implementing
such agreements. This
second question is the subject of Chapter 4, and is discussed in
more detail below in
Section 5.2.
5.1 Distributional Impacts of Energy Policies in India (Chapters
2
and 3)
There is still a lack of empirical evidence on the magnitude and
direction of the
interdependence and interaction of sustainable development and
climate change, and
[their] equity implications…
New research is required that studies the linkages between
climate change and
national and local policies [emphasis added]
Technical Summary, Working Group III, IPCC, 2007.
Climate mitigation in the energy sectors affects households
predominantly through the
increased costs of energy services. Prices may increase as a
result of shifts towards
more expensive technologies in the supply, delivery or
consumption of energy, or as a
result of policy instruments, such as energy taxes (or the
removal of subsidies) that
aim to discipline energy consumption. It is well known that in
developing countries
energy subsidy policies and the state-owned bureaucracies that
deliver energy services
to households serve many social and political objectives, such
as redistributing income
-
21
and meeting the needs of special interests (Victor 2007). The
distributive impacts on
households of mitigation policies in energy depend, therefore,
on these baseline
service conditions and on how governing institutions implement
mitigation policies in
conjunction with other social objectives. The novel approach of
the studies in Chapter
2 and 3 is to characterize these baseline conditions and
quantify the distributional
impacts of these institutions. Both studies pay particular
attention to the leverage
policymakers have to influence these impacts through pricing
policies. I investigate
the distributional impacts of two energy policies being
considered or implemented in
India, both of which are being pursued for reasons not related
to climate mitigation but
have the effect of reducing India‟s carbon intensity: removing
the kerosene subsidy
and investing in low carbon electric supply.
In Chapter 2, I formally evaluate the impacts of the kerosene
subsidy on income
distribution, its efficacy as an instrument of redistribution,
and the causes of
misallocations of subsidies. This study provides new insights
into the different impacts
of policy design and implementation failures on the
distributional benefits of
subsidies. I also reveal a previously unrecognized phenomenon of
price discrimination
and rent extraction by private licensees who distribute
subsidized kerosene through the
Public Distribution System.
Chapter 3 contributes to the literature on climate mitigation
policy impact analysis in
India. No study has as yet examined the impact of economy-wide
increases in the
prices of goods and services that result from undertaking
mitigation in the electricity
sector. I develop a simulation model of the electricity sector
and household welfare in
the state of Maharashtra, to examine the leverage policy makers
have over the
distribution of future mitigation costs. This study is also
unique in that the simulation
incorporates the entrenched practice of load rationing to manage
supply scarcities and
households‟ responses to outages. It therefore provides the most
realistic simulation of
sectoral conditions among studies of the Indian electricity
sector.
5.1.1 India as a Case Study
India is of critical importance in climate equity by virtue of
the sheer number of
people living in poverty. Among the large developing countries
that are candidates for
-
22
the types of within-country exemptions discussed here, India has
the largest number of
poor people of any country, with a third of the world‟s poor and
a third of the world‟s
malnourished children. China has a comparable population and has
comparable levels
of inequality, but has achieved far more success in providing
basic minimum living
standards. Thus, the prospect of climate mitigation exacerbating
poverty is a less
pressing political concern in China than in India. Brazil and
South Africa would also
serve as good candidates for similar analyses, but the financial
stakes are orders of
magnitude less than in India, making their cases relatively less
important in an
international agreement over burden-sharing. In fact, one may
ask, given that the
overwhelming majority of Indians are poor (Figure 2), why isn‟t
it reasonable to take a
statist view and exempt India altogether? The premise of this
study is that only
countries with substantial rich and poor populations generate
the moral dilemmas
presented here. However, India does have a sizeable and growing
wealthy population.
Households earning more than $2/day (PPP basis) spend more than
40 percent of total
household expenditure in India, even though they comprise 25
percent of the
population. Further, the aggregate emissions from those earning
more than $2/day are
comparable to South Africa and Brazil simply because of their
sheer numbers. India
has a growing consumerist middle class comprising over 200
million people, whose
consumption could pose serious threats for climate change if
measures are not taken
towards low carbon development.
Despite this empirical focus on India, the arguments and
analytical approach used in
this study can be applied to the BASIC countries, and also to
other developing
countries, such as Indonesia, Pakistan and Nigeria. Though these
countries currently
have predominantly poor populations, if they follow similar
development trajectories
as the BASIC countries, they may well develop „top-down‟, in the
sense that they
alleviate poverty while also creating significant wealth among
the non-poor.
-
23
5.2 Implementing a Moral Minimum in International Climate
Agreements (Chapter 4)
In Chapter 4, I discuss the ethical and policy implications of
implementing an
exemption for the poor in climate mitigation agreements. The
issue I grapple with is
that the outcomes of such an agreement depend on the policies
and actions of state
institutions that have no obligations under international
agreements. This study
addresses two specific aspects of this moral hazard, related to
the enforcement and
design of such an exemption for the poor.
5.2.1 Enforcing an Exemption
Beneficiaries of an exemption in a developing country may not
get the benefits of an
exemption granted to a state on their behalf due to
institutional corruption or weak
capacity. Who should be accountable for enforcing the exemption,
if anyone?
Chapters 2 and 3 illustrate the enforcement problem in some
detail. In Chapter 4, I
discuss the ethical obligations of signatories to an
international agreement to ensure
that the poor receive the benefit of such an exemption. I also
provide some guidance
regarding how such an exemption should be structured in climate
policy in light of
these considerations.
5.2.2 Designing the Terms of an Exemption
States can influence the future distribution of income, and
therefore emissions, around
a threshold through their development policies. Sovereign
policies can therefore
change the total level of exempted emissions, which in turn
influences the costs that
everybody else bears to support this exemption. If this is
indeed the case, parties to
such an agreement would have an interest in designing the terms
of such an exemption
to minimize their respective mitigation burdens, which could
undermine the objectives
of the agreement.
These moral hazards raise an important question regarding the
distribution of
mitigation burdens that has not been asked as yet in literature:
how should the costs of
exempting the poor be distributed between benefiting states and
other parties to a
mitigation agreement that incorporates such an exemption? Other
relevant proposals
-
24
implicitly assume that these costs would be born equally by all
the non-exempt
members of parties‟ to an agreement. However, given that states‟
sovereign policies
can influence how these costs are distributed, this assumption
demands further
exploration and justification.
On the practical side, I investigate how these moral hazards
would affect parties‟
interests in getting agreement on the terms of such an
exemption. I quantitatively
assess what is financially at stake for parties to an agreement
in the influence of states‟
policies over the burden of exemption. I suggest preliminary
directions for climate
policy in how to structure and enforce an exemption in a manner
that surmounts these
potential obstacles.
5.3 Precedents in International Agreements
With developing countries‟ increasing participation in
international trade and
environmental treaties, the influence of these international
institutions on poverty has
received much attention (Chen 2005; Clapp 2010). While there are
many parallels
between previous international agreements and a future climate
mitigation agreement
in terms of their impacts on poverty and sovereignty, the
remedies that have been
sought and implemented in other realms have limited
applicability to the case at hand.
In particular, although countries have obtained exemptions from
particular provisions
of trade agreements, there are no precedents for exempting poor
subpopulations in
large developing countries as is contemplated in the climate
ethics literature and in this
study.
Both trade and environmental agreements present numerous
pathways by which
poverty can be exacerbated, some of which are similar in outcome
to some of the
impacts of climate change mitigation – such as to increase the
prices of essential
commodities, such as energy, food and medicines.25
As such, these agreements raise
25 Another important pathway is through livelihood impacts, such
as from reduced global demand for
products whose production support poor livelihoods, or by
through policies that directly discourage
such production in favor of other products and services.
-
25
equally important claims of harm to the poor‟s basic needs as
would a climate
mitigation agreement.
To avoid such impacts, among other motivations, there have been
a few cases where
countries have sought exemptions from certain obligations under
trade agreements. In
these cases, trade restrictions apply in specific sectors, but
they are typically enforced
for countries as a whole. For instance, in the Trade-related
aspects of Intellectual
Property Rights (TRIPS) under the World Trade Organization
(WTO), least developed
countries have obtained exemptions from particular provisions
related to drug patents
that would raises drug prices.26
WTO provisions also allow countries to restrict food
exports to ensure domestic food security, and to restrict
imports that do not comply
with internal health and safety standards.27
While it is likely that the harmful impacts
of higher drug prices, less food supply or unsafe imports may
fall only on certain
populations, and therefore do not necessarily justify absolute
exemptions for states,
political negotiations have not developed to account for such
intranational inequities.
International environmental treaties have also raised complex
issues regarding the
claims of the poor to local environmental resources that have
global benefits and to
freedom from the burdens of restricting the exploitation of
these resources. Here too,
scholars have recognized the disconnect between the design of
environmental
agreements between states and the multiple subnational agents
that affect or are
affected by the environmental resources that these treaties are
designed to protect
(Herring 1999). This is best illustrated in the case of the
Montreal Protocol, which is
not only seen as one of most successful examples of resolving
international conflicts
over the protection of a global environmental resource
(stratospheric ozone) but is also
the most similar international environmental problem to climate
change, since it
involves a “global commons”. While China and India were provided
compensation by
developed countries in exchange for adopting restrictions on
chlorofluorocarbons,
Herring points out that the Montreal process in India was
restricted to a narrow
26 Article 6, Transitional Arrangements, in the TRIPS
Agreement.
27 General Agreement on Tariffs and Trade (GATT), Article XI,
and the Sanitary and Phytosanitary
Agreement under the WTO respectively.
-
26
governing elite, and the obligations and costs within India of
compliance were not
evaluated before or after the treaty.
Thus, the need to examine subnational distributional impacts of
states‟ participation in
international agreements seems important and neglected in
literature. International
agreements in other domains may well benefit from, rather than
inform, the design and
implementation of exemptions for the poor in climate mitigation
agreements. There
may be other lessons that can be learned from the experiences of
previous
international agreements, such as how to design a fair
negotiation process or monitor
compliance. However, drawing these lessons would require a more
comprehensive
comparative analysis of international agreements, which is
beyond the scope of this
study.
-
27
CHAPTER 2 – KEROSENE SUBSIDIES: WHEN ENERGY POLICY FAILS AS
SOCIAL POLICY
1 Introduction
Energy subsidies have attracted renewed attention with the
urgency of climate change.
Historically, because broad-based institutions that enable
direct cash transfers, such as
income tax (Piketty 2009), are lacking in many developing
countries, household fuel
subsidies often serve as instruments of redistribution (Komives
2005). In India, the
high fiscal and environmental costs of these subsidies have been
well documented
(Gangopadyaya 2005; Komives 2005). Economists estimate that
kerosene subsidies,
for example, carry a fiscal burden of $4-6 billion per year, and
efficiency losses of $1-
2 billion per year (International Energy Agency 1999; Morris
2006). Kerosene use
also accounts for over 18 million tons of CO2 emissions
annually,28
or over 1.5 percent
of the country‟s CO2 emissions.
The Indian government is now considering phasing out kerosene
subsidies, and
eventually replacing them with conditional cash transfers
(Parikh 2010).
Notwithstanding the theoretical merits of direct cash payments,
the feasibility, timing
and merits in practice of such an ambitious identification
system are uncertain and
disputed (Dreze 2010).
Though kerosene subsidies are known to be poorly targeted, their
distributional
benefits are not well understood. Over 800 million Indians use
kerosene for lighting,
of which 200 million across income groups also use kerosene also
as a cooking fuel.29
Suppliers and distributors divert 40-60 percent of kerosene
upstream in the supply
chain to other lucrative markets such as transportation,30
forcing some households to
purchase kerosene in the black market (Morris 2006). The
government draws support
28 Assuming emissions of 2.5kg/liter, from ~7.3 billion liters
of annual consumption, based on
household consumption data from National Sample Survey 2004-05
29
Author calculations using National Sample Survey of Consumption
Expenditure 2004-05, discussed
in Section I. 30
In transportation, kerosene is used as a cheap fuel substitute
to diesel. In construction, kerosene is
used in making tar for paving roads.
-
28
for phasing out the kerosene subsidy in part from the finding of
an expert committee
that the income shocks on poor households on average would be
fairly small (Parikh
2010). However, these averages mask the distribution of
benefits.
Even though kerosene subsidies are known to be poorly targeted,
their distributional
benefits are not well understood. Suppliers and distributors
divert 40-60 percent of
kerosene upstream in the supply chain to other lucrative markets
such as
transportation,31
forcing some households to purchase kerosene in the black
market
(Morris 2006). The government draws support for the phase-out of
the kerosene
subsidy in part from the finding of an expert committee that the
income shocks on
poor households on average would be fairly small (Parikh 2010).
However, these
averages mask the distribution of benefits.
The World Bank finds that in general quantity-based utility
subsidies tend to be
regressive because their use increases with income. However,
these do not apply to
kerosene, whose use does not correlate well with income. One
study of fuel taxation in
India finds that the direct benefits from household kerosene use
are progressive (Datta
2010). However, the study offers limited detail on underlying
regional differences and
causes. Further, progressivity alone may not justify public
spending on subsidies.
Other redistributive policies may have higher impacts on poverty
and cost less to
deliver.
In this paper I formally examine the performance of the kerosene
subsidy as an
instrument of redistribution using several measures. I evaluate
the subsidy‟s
materiality (what budget share does the subsidy represent to
households) and its
progressivity (do poor households benefit more than the average
household). For
purposes of comparison with other redistributive policies, I
assess efficacy (what share
of the total subsidy goes to poor households as intended, not
counting indirect
benefits32
). I explore the distinction between implementation failures and
design
31 In transportation, kerosene is used as a cheap fuel
substitute to diesel. In construction, kerosene is
used in making tar for paving roads. 32
Data on the diversion of kerosene to other markets are
unavailable, which makes it difficult to
understand who benefits from these diversions. These diversions
are often controlled by organized
-
29
limitations, and evaluate some of the subsidy benefits under
ideal implementation
conditions. Lastly, I explore the phenomenon of price
discrimination by distributors of
subsidized kerosene, whereby households pay a wide range of
prices for subsidized
kerosene. This price variation is an important, and overlooked,
component of the
benefit incidence for a subset of households.
Overall, this study supports the phase-out of subsidies in rural
areas only. However,
kerosene subsidies are material and progressive in urban areas.
Particularly where
access to wood is lacking, kerosene subsidy benefits can be up
to 5 to 10 percent of
household expenditure, and its removal would more than double
these households‟
cooking budgets. However, the goal of policy should be to phase
out kerosene
demand, rather than the subsidies, by improving access to
LPG.
This study relies on data from the National Sample Survey of
India for Consumption
Expenditure, 2004-05 (NSSO0405) and draws qualitative insights
from a primary
survey conducted of 450 households in urban and peri-urban parts
of Maharashtra in
2009/2010 (See Appendix D).
In Section 2, I describe the various uses for kerosene in
different household in
Maharashtra, and how these functions influence the subsidy‟s
relative benefits. In
Section 3, I discuss the measurement approach and the results of
the subsidy
performance analysis. In Section 4 I discuss the subsidy
performance under ideal
implementation conditions. In Section 5 I discuss price
discrimination by ration shop
owners (RSOs). In Section 6, I discuss policy implications.
2 Kerosene Market and Use Characteristics
Kerosene use is widespread in India and in Maharashtra in
particular. In Maharashtra,
seventy percent of households (68 million people) use kerosene.
Of these, about 50
million users are in rural areas, and 18 million are in urban
areas.
crime syndicates in Maharashtra. See „Maharashtra Cracks Down on
Oil Mafia‟, Hindu Business Line,
January 28, 2011.
-
30
Kerosene use is not homogenous across the country, particularly
between rural and
urban areas. Urban use of kerosene is higher in Maharashtra than
in the rest of the
country, and even more so in the vast slums of metropolitan
Mumbai (Figure 3). The
average kerosene consumption among households earning less than
Rs. 4,000 per
month (or $222 PPP) in urban Mumbai is over 10 liters per month.
As a result, though
26 percent of kerosene users in Maharashtra are in urban areas,
they consume 45
percent of total household kerosene use in the state.
Figure 3: Variation in Urban Household Kerosene Use -
Mumbai, Maharashtra and India 2004-05
Kerosene is also used widely across income groups, but rarely as
a primary fuel
(Figure 4). Rather, kerosene is used more commonly as a backup
fuel, to electricity for
lighting, and to LPG and biomass for cooking. This is because
kerosene is not a
preferred fuel. This feature of kerosene is evidenced by the
fact that households using
multiple fuels rarely, if ever, use kerosene as a primary fuel,
either for lighting or
cooking.33
The poorest in rural areas – who rely on wood for cooking – and
the
wealthiest in urban areas – who use LPG exclusively – hardly use
kerosene. This is in
33 This was found both in the primary survey and NSSO0405, where
households name their primary
lighting and cooking fuels.
0
2
4
6
8
10
12
14
16
1 2 3 4 5 6 7 8 9 10
Ker
ose
ne
(Lit
ers/
Mo
nth
)
Income Deciles
Mumbai Suburbs
Rest of MH
India
-
31
contrast to LPG, which is a preferred cooking fuel, and whose
use correlates well with
income, as shown in Figure 4.
Figure 4: Cooking Fuel Shares by Income Decile – Maharashtra,
2004-05
(a) Rural
(b) Urban
Source: National Sample Survey of Consumption Expenditure,
2004-05
Note: Kerosene use below 4 liters is assumed to be for
lighting.
Two underemphasized characteristics of kerosene use in
households are important in
understanding the benefit incidence of subsidies: actual
kerosene market prices\, and
0
20
40
60
80
100
120
1 2 3 4 5 6 7 8 9 10
Ave
rage
Co
oki
ng
Ener
gy(M
J p
er c
apit
a)
Income Deciles
LPG
Kerosene
Wood
0
20
40
60
80
100
120
1 2 3 4 5 6 7 8 9 10
Ave
rage
Co
oki
ng
Ener
gy(M
J p
er c
apit
a)
Income Deciles
LPG
Kerosene
Wood
-
32
kerosene‟s use as a cooking fuel. These are discussed next,
along with how they affect
subsidy benefits.
2.1 Kerosene Markets
The quantity of subsidized kerosene allocated by government is
intended to meet
household needs. In reality, households in Maharashtra purchase
over 40 percent, and
in urban areas over 50 percent, from secondary (black) markets
(Table 1).
Households are supposed to purchase all their household kerosene
needs through the
Public Distribution System (PDS). As part of the PDS, millions
of retailer