2
A Comparison of Alternative Programs for Climate Policies
About KAPSARC
Legal Notice
The King Abdullah Petroleum Studies and Research Center (KAPSARC) is a
non-prot global institution dedicated to independent research into energy economics,
policy, technology and the environment, across all types of energy. KAPSARC’s
mandate is to advance the understanding of energy challenges and opportunities
facing the world today and tomorrow, through unbiased, independent, and high-caliber
research for the benet of society. KAPSARC is located in Riyadh, Saudi Arabia.
© Copyright 2018 King Abdullah Petroleum Studies and Research Center (KAPSARC).
No portion of this document may be reproduced or utilized without the proper attribution
to KAPSARC.
3
A Comparison of Alternative Programs for Climate Policies
This research paper compares the relative welfare impact of different options for allocating the
nancing burden of climate change mitigation policies. Focusing on efcient ways to nance policies
aimed at climate change mitigation, not only at direct carbon reduction, could delink the issue of
carbon taxation from carbon emissions.
A Pigouvian tax is the traditional way of correcting for negative externalities, or the undesirable
consequences for society arising from the actions of a company or industry sector, by levying additional taxes
on that activity. Pigouvian taxation corrects society’s welfare loss, however, from the viewpoint of the private
sector, such taxation imposes a deadweight economic loss with respect to the original private equilibrium.
As an alternative, we evaluate a methodology that could fund investments to reduce carbon dioxide (CO2)
emissions, and we show that the policy we consider to be optimal from a tax standpoint – Ramsey pricing
– can both improve world welfare and be politically more acceptable than other pricing options. Rather
than focus directly on emissions reduction by taxing energy, a Ramsey pricing solution can be designed to
minimize distortions while raising funds for investment in climate change mitigation.
Ramsey pricing is seen as permitting the application of the principle of recognizing common
responsibility for climate change mitigation, but at the same time differentiating the ability of individual
countries to contribute to the common goal. Applied to energy prices, this means that efcient taxation
should be inversely proportional to the consumer (household) energy demand elasticity of the individual
country. That is to say, the more inelastic a country’s consumer energy demand, the higher the efcient
taxation should be in that country.
With the aid of an extensive data set of 106 countries that were responsible for around 90% of total
world energy consumption and carbon emissions in 2014, we estimate a complete demand system for
world household consumption behavior and use the resulting country price elasticity values to compute
an optimal Ramsey price scheme for nancing investment in climate change mitigation policies.
Compared to other allocation approaches – such as a Pigouvian tax, which is proportional among
countries – we found that the overall world benet of the Ramsey approach is higher. This modeling
exercise suggests that there are a number of cost reduction opportunities in using a Ramsey allocation.
Furthermore, we believe that Ramsey pricing leaves room for negotiating compensations, which could
be politically more acceptable than traditional taxation approaches.
Key Points
4
A Comparison of Alternative Programs for Climate Policies
Summary
In the global carbon policy debate, pricing is
considered to be a key instrument to achieving
the desired levels of emissions reductions.
The Pigouvian tax is theoretically the best solution to
tax carbon emissions, in order to achieve emissions
reduction through nancial investment, but it has not
proved to be politically viable. A Pigouvian tax sets out
to correct negative externalities, or consequences for
society – such as the consequences of climate change
– by levying additional taxes. However, from the
viewpoint of the private sector, such taxation imposes
a deadweight loss with respect to the original private
equilibrium. This generates political resistance that
may impede achieving the theoretical optimal solution.
Most international policy meetings since the Kyoto
Protocol agreement have resulted in lukewarm
commitments from developed economies and strong
resistance from emerging economies over the
fair economic allocation of the burden associated
with the various calls for emissions reduction. This
kind of situation suggests the need for alternative
formulations, in the realm of what economists
call ‘second-best options,’ to tackle the issue of
realistically nancing alternative policies.
This paper considers alternative policy formulation
aimed at funding investment for climate policies,
based on the principle of minimizing deadweight
losses associated with taxation and on consumer
preferences. (A deadweight loss is the added
burden placed on consumers and suppliers when
the market equilibrium is altered because of tax, for
example. It results when supply and demand are
out of equilibrium.)
The policy proposal we examine here is a
Ramsey allocation, which aims at designing an
economically optimal taxation scheme for nancing
climate mitigation investments. A Ramsey pricing
policy, applied to energy prices, would mean that
efcient taxation should be inversely proportional
to the consumer (household) energy demand
elasticity of the individual country. In other words,
the more inelastic a country’s consumer energy
demand, the higher the efcient taxation should
be in that country. The overall taxation scheme is
optimal because it minimizes the deadweight loss.
This strategy is not aimed at directly reducing
emissions, and hence energy consumption. It
can, in a more general way, help to assist with
providing efcient funding for a wider range of
policies, such as carbon sequestration, alternative
fuels, energy efciency, and the earth’s albedo
enhancing. In this framework, notice that carbon
sequestration and articially enhancing the earth’s
albedo represent technological solutions aimed
at reducing carbon dioxide (CO2) concentration
and adding sunlight reecting aerosol in the soil
or stratosphere, thereby cooling the climate in
a different way than reducing carbon emissions
(NAS 1992). The strategy makes explicit use of
household preferences, as expressed through
their energy demand behavior, econometrically
estimated at the world level.
A Ramsey allocation can be integrated into the
general principle of mutual cooperation that
motivates climate agreements, as it reects a
common but differentiated burden of all parties.
5
A Comparison of Alternative Programs for Climate Policies
Introduction
In the global climate change policy debate, carbon
pricing is considered by many economists to be
a key instrument for achieving carbon emissions
reductions. The traditional solution of computing a
Pigouvian taxation based on the criterion of adding
marginal social damage to the marginal private cost
has not proved to be viable, despite a long history
of international political dialogue since the signing of
the Kyoto Protocol. Moreover, the general outcome
of these meetings has been a bitter confrontation
between developed and emerging economies as to
a fair economic allocation of the burden associated
with carbon emissions reduction.
This unfortunate state of affairs has two important
consequences for climate policy. First, in many
cases different countries’ preferences as regards
cost allocation oppose each other. Second, the
societal benet of climate change policy constitutes
a worldwide positive externality and thus could
involve a sizable policy-induced market distortion.
For these reasons, the carbon price has to be
different from the private marginal costs, pointing
to the need for a second-best solution. A Ramsey
(1927) price scheme, which minimizes the
deadweight losses – the added burdens placed on
consumers and suppliers when supply and demand
are out of equilibrium – associated with given market
inefciencies, is a possible theoretical solution to the
problem of quantifying the real costs of not tackling
climate change.
Surprisingly, the vast literature on carbon pricing
has not explored this analytical tool as a mechanism
for sharing the economic burden of climate policy.
The approach used so far involves designing the
individual country commitments in proportion to
emissions or gross domestic product (GDP), valued
at a common marginal price. In some economic
circles, this could be considered as a proxy trade
barrier. The main shortcoming of this is that it
creates a burden for newly industrialized countries
that produce goods which are ultimately consumed
by advanced economies. In other words, energy-
intensive manufacturing countries risk being
penalized for the carbon content of the nal goods
consumed by higher-income economies. There are
relevant differences across major economies, as
shown by the energy embodied in the trade between
major world economies (Table 1), computed after the
major global recession, based on input-ouput data in
2009 (Gasim 2015). Many emerging countries show
positive values, while most industrial countries have
decentralized energy-intensive, and hence carbon-
intensive, production sectors.
This paper examines the potential for an
economically optimal taxation policy to nance
investments in carbon emissions reduction, based on
households’ preferences, as expressed through their
energy demand behavior. This is a more complex,
yet more accurate, way to quantify the ‘polluters
pay’ principle. Households are the nal consumers
of goods and services and consumption. Goods
incorporate energy used in the production process,
whether they are produced domestically or imported.
In addition, households are the ultimate owners of
the corporate sector and the nal beneciaries of
government expenditures. Accordingly, allocating
a tax burden based on household consumption is
a more precise way to account for all the energy
incorporated into a society’s economic activity. There
are two caveats, however. First, ideally it would be
optimal to include the indirect use of energy that
is involved in the production of other goods and
services consumed. Second, if local policies distort
energy prices, the estimated price elasticities may
suffer from these distortions. These issues are
outside the scope of the present work.
We assume that heterogeneous consumers value
the marginal damage resulting from greenhouse
gas emissions differently, due to differences in
interest, perception, income and values across
6
A Comparison of Alternative Programs for Climate Policies
Introduction
the world. For example, since carbon emissions
are of greater concern to younger and more
educated citizens, a younger society, such as
some emerging economies, is likely to have a
different perception of the importance of carbon
emissions than an older society. In addition, the
higher incomes in a developed economy, such
as the United States (U.S.) and those in Europe,
determine higher environmental awareness.
However, the level of concern over environmental
issues differs between generations.
We contrast the Ramsey optimized scheme with two
traditional approaches: uniform price taxation, levied
in proportion to the country’s importance in terms of
world GDP and/or carbon emissions, and a specic
scheme, better aligned with the ability to pay, levied
only on the richest subset of countries.
We calculate each country’s share of the cost of
climate change mitigation, independent of how that
overall cost is calculated. One potential example
of a cost estimate is the announced level of
investment necessary to achieve the International
Energy Agency’s (IEA’s) 450 parts per million (ppm)
Scenario (IEA 2016). The latest IEA scenarios
project that $100 billion of additional investment per
year will be required to support mitigation policy
to stabilize atmospheric CO2 concentration at that
level by 2030.
Negotiators reached an agreement on goals to
mitigate global climate change at the 2015 United
Nations Conference of Parties (COP 21) (United
Nations 2015) in Paris, after the resounding failures
that plagued previous conferences. The agreement
entered into effect once it was ratied by 55
countries, despite the notication of withdrawal by
the U.S. It aims to limit global warming to less than
2 degrees Celsius by the year 2100, compared
with preindustrial levels. The agreement referred
to a commitment to deliver nancial aid of at least
$100 billion to poorer economies by 2020, a sum
which could increase in the future. However, the
agreement did not set out a way forward on the
allocation of donor country contributions, nor specify
which countries should receive nancial aid.
This research is not concerned with the plausibility
of the COP 21 goals, but with the most reasonable
and responsible outcome — a cooperative
solution reached between the wealthiest countries
representing a large proportion of world emissions.
To achieve this, we consider a hypothetical
agreement that includes the top emitting countries
and the richest countries in terms of GDP per
capita, which together represent at least 55 percent
of total emissions.
In the empirical estimation, we also take account of
geographic differences in evaluating the household
price elasticity of energy consumption, which
is directly related to emissions. The elasticity is
a revealed preference measure of households’
willingness to pay for energy consumption and,
indirectly, for their willingness to pay for emissions
reductions. The inverse of the demand elasticity is
used to calculate the Ramsey proportionality factor
to compute the burden sharing of each country’s
climate policy. The deadweight losses associated
with the Ramsey solution are compared with other
traditional burden-sharing mechanisms.
7
A Comparison of Alternative Programs for Climate Policies
Introduction
Table 1. Embodied energy content in nal consumption – values in $U.S., 2009.
Note: A negative value refers to a net import of embodied energy in traded products, while a positive value indicates a net export.
Source: Gasim 2015.
Country Net energy content
Australia -7560
Austria -4620
Belgium -4200
Brazil -1260
Bulgaria 2100
Canada 2940
China 125580
Cyprus -420
Czech 1680
Denmark 2520
Estonia 420
Finland 420
France -21000
Germany -21000
Greece -3360
Hungary -840
India 420
Indonesia 840
Ireland -2100
Italy -21840
Japan -23520
Latvia -420
Lithuania -420
Luxembourg 420
Malta 0
Mexico -3780
Netherland 1260
Poland 840
Portugal -1680
Romania -840
Russia 74760
Slovakia -840
Slovenia -420
South Korea 15120
Spain -10500
Sweden 0
Taiwan 15120
Turkey -3780
United Kingdom -22260
USA -98700
Rest of world 12180
8
A Comparison of Alternative Programs for Climate Policies
Introduction
Deadweight loss and Pigouvian tax
When the market equilibrium of supply and demand is disturbed by tax – or, in general, by a price
distortion – changes to the consumer and producer surplus result. Intuitively, a price increase for the
product reduces the consumer surplus and a reduction in demand reduces the producer surplus.
However, the additional government revenue is devoted to social goals and thus it can be seen to
contribute to an increase in the welfare of society. The problem is that there is a net loss for society,
described by economists as a ‘deadweight loss.’ It is a non-retrievable loss, calculated as the difference
between the welfare of society before the tax (consumer + producer surplus) and after the tax (consumer
+ producer surplus + tax revenue).
Figure 1a illustrates this. The equilibrium before tax is shown at A. The area ADF represents the sum of the
welfare of society (consumer + producer surplus). After tax, the new equilibrium is at B. The new welfare
(consumer + producer surplus + tax revenue) is the area DBCF (consumer + producer get the area DBE;
government revenue is BCEF). The difference is the area ABC, which is the deadweight loss.
Supply after tax
p
q
0
Supply before tax
Demand
D
B
E
F
C
A
Source: KAPSARC.
Figure 1a. Market equilibrium before and after tax and deadweight loss.
9
A Comparison of Alternative Programs for Climate Policies
Introduction
Source: KAPSARC.
Figure 1b. Market equilibrium before and after the Pigouvian tax.
We can also use a similar analysis to identify the optimal Pigouvian tax on a negative externality. In this
case, the initial equilibrium is inefcient because it does not take into account the social damage that is
inicted by the existence of a negative externality. This initial equilibrium is point A in Figure 1b, where
demand and private supply intersect. In point A, the additional social damage inicted by the externality is
not considered. But if we can add to private supply the correct measure of the cost of the externality (the
damage associated with the externality), we can identify the efcient equilibrium, point B. Operationally, the
Pigouvian tax is the tax that allows the market equilibrium to reach point B. The price increases from F to
E and the quantity decreases from C to D. Point B is efcient because the sum of the marginal social and
private cost is equal to the marginal benet (the willingness to pay expressed by the demand curve).
Private supply plus the
social marginal cost of
the externality
p
q
0
Private supply
Demand
D
B
E
F
C
A
10
A Comparison of Alternative Programs for Climate Policies
Introduction
The Ramsey pricing
When a monopolist is facing different groups of consumers with different behavior, it is convenient to
discriminate among them in order to maximize the prot. Intuitively, if a group of consumers is relatively
price inelastic, an increase in price increases the revenue (because the quantity decrease is less
proportional than the price increase). Conversely, if another group of consumers is price elastic, a
decrease in price increases the revenue (because the quantity increase is more proportional than the
price decrease).
Figure 1c illustrates this. Note that, for simplicity, the marginal cost MC is constant. The prot
maximization equates marginal cost to marginal revenue for both groups. Because there are differences
in behavior, the solution yields different prices. Namely, the price PB* is higher than PA* because the
elasticity of demand DB is lower than that of DA.
Demand
QB*
Q Q
PA*<PB*
DA
DB
MRA
MRB
MC
PP
PA*
0 0
PB*
Source: KAPSARC.
Figure 1c. Ramsey pricing equilibrium.
11
A Comparison of Alternative Programs for Climate Policies
Review of the Applications of Ramsey
Pricing
Policymakers have long studied opportunities
for emissions reduction. In 2005, the
European Union (EU) implemented, with
much fanfare, a carbon emissions trading system
(EU ETS) as the basis for its greenhouse gas
emissions reduction strategy. Local industrial
sectors were lukewarm to it and the ensuing 2009
world recession substantially constrained trading
liquidity (Hu et al 2015). China also aims to reduce
carbon dioxide emissions per unit of GDP by 60%
from the 2005 level by 2020, stop emissions growth
by 2030, and cut carbon intensity by 45% compared
with 2005 levels (UNFCCC 2016).
In 2017, China implemented an effective emissions
trading scheme (ETS), the most efcient and
important policy measure for carbon emissions
reduction in the country. The previous pilot market
test from 2013-2016 showed trade of 94 million
tonnes of emissions allowances at an average price
of $3.72/tonne (Zhang 2017). This is in addition to
previous energy productivity achievements from
1995 onward (Atalla and Bean 2015). However,
most of the climate policies already implemented
are local or regional; global agreements have faced
stiff opposition. The agreement at COP 21 marks a
turning point, but there is no accord on the allocation
of the nancial burden.
Because Ramsey pricing minimizes the welfare loss
associated with taxation, applying this mechanism
in the energy sector could offer a solution for
efcient allocation. It has previously been applied
in tariff regulations for public utility sectors, such
as telecommunications, transport, and electricity
and gas distribution (Laffont and Tirole 1996). Its
theoretical justication is the need to generate
sufcient revenue to support a public utility service
when an economically efcient allocation is
unobtainable. In that context, Ramsey schemes
have been applied at the industrial level, such as
in the oil rening sector analysis of Babusiaux and
Pierru (2007) or the optimal California gasoline tax
proposed by Lin and Prince (2009). In addition, the
analysis of airport fare structures by Hakimov and
Mueller (2014) illustrates the cost recovery problems
of airport operation.
Many applications of Ramsey pricing have analyzed
the heterogeneity of demand elasticity behavior
to justify charging differentiated prices to different
groups of customers. Such studies include analysis
of residential electricity customers in the U.S. (Berry
2002), China (Qi et al. 2008, Sun and Li 2013),
Russia (Nahata et al. 2007), Brazil (Santos et al.
2012), Japan (Matsukawa et al 1993) and European
countries (Deeney et al. 2016). Most of these studies
conclude that the actual tariffs are at variance with
the optimal Ramsey pricing scheme. More recently,
other studies focused on Ramsey pricing to make the
optimal allocation of the social cost of externalities,
such as the environmental cost of air trafc (Martín-
Cejas 2010), or electric network congestion and
security management (Bigerna and Bollino 2016).
Van der Ploeg (2016) and Boeters (2014) discuss
various options for optimal carbon taxation.
Two main observations follow from the literature
review. First, policy strategies tend to impose
inefcient pricing schemes. Typical examples are
the different tariff structures for residential and
industrial electricity users, the various tax rates on
gasoline and diesel in the transport sector, and the
multiple tax rates on electricity and natural gas for
residential consumers. These examples raise the
question of why policy actions tend to be inefcient
and why policymakers do not use Ramsey schemes.
The reliability of the empirical estimations needed
to compute Ramsey pricing is one challenge:
policymakers need plausible and robust knowledge
12
A Comparison of Alternative Programs for Climate Policies
of the demand behavior of different groups and their
related price elasticities to implement Ramsey pricing.
Additionally, the status quo tariff structure reects
historical lobbying by different constituencies. The
strongest constituencies may oppose tariff changes
and it may be politically difcult to change the tariff
structure based on innovative empirical ndings.
A further challenge is that Ramsey pricing
maximizes efciency, but it does not take into
account equity across groups. It is also politically
challenging because it entails price discrimination
among consumers. A typical example is the fact
that poor consumers are less price elastic: they
cannot afford exible behavior as they do not
often have easy access to more efcient capital
stock. This can be seen in their inability to borrow
to nance purchases of new capital equipment.
However, the possibility of achieving a given
target more efciently would permit the use of
the overall gain to bring about welfare transfers
across groups. As such, this should not be a
serious problem.
Review of the Applications of Ramsey Pricing
13
A Comparison of Alternative Programs for Climate Policies
Optimal Pricing and Demand Behavior
To investigate further the option of
applying Ramsey pricing to international
policymaking, we measure the cost
associated with alternative policy strategies,
contrasting the deadweight loss of the Ramsey
scheme to those of other options based on
consumers’ ability to pay. We recall that even if
a Ramsey scheme does not take into account
equity, this omission can be corrected by means
of income transfer related to the marginal utility
of expenditures (Diamond 1975). We take the
viewpoint of a supranational entity that is interested
in achieving global aggregate efciency and must
consider the distributional effects among countries.
The application of Ramsey pricing to climate
change policy is appropriate, because it solves
the problem of setting an optimal price scheme
in cases where the efcient rule of price equal to
marginal cost has failed or, in other words, when
market failures like free riding behavior lead to a
suboptimal solution. Free riders declare a distorted
willingness to pay, counting that others will bear
the cost, and the aggregate consequence is that
insufcient resources are committed to the target.
This is precisely applicable to the worldwide
dilemma of climate change policy. The literature
on the free riding effect (e.g., Bigerna et al. 2016)
and on the issues that lead to a failure in political
negotiations to agree on the correct amount of
resources to be committed to climate policy (e.g.,
Weitzman 2017) is vast.
The cost of climate policy must be superimposed
on the pure market price of energy, and ideally
this revenue should be collected in the least
distortionary way. Ramsey pricing is an ideal
solution for adapting carbon pricing to different
countries, according to their consumers’ willingness
to pay for such a commitment. Nonetheless, to
the best of our knowledge, the existing literature
contains no application of Ramsey pricing to
climate policy.
Ramsey pricing has been conned to the debate
on public utility regulation, highlighting how price
discrimination is used to maximize prots at the
expense of consumers. In this paper, we view
Ramsey pricing as the application of the principle of
recognizing common responsibility for climate change
mitigation, but at the same time differentiating the
ability to contribute to the common goal. It can be
consistent with the framework set at the international
conventions on climate change, aligning with the
idea that there is an efcient way to make the rich
pay more and the poor pay less through adopting
appropriate compensation schemes.
We also recognize that the adoption of Ramsey
pricing for nal users in an international scheme
is challenging because policymakers do not know
precisely the elasticity of demand of the entire
population. We aim to resolve this difculty by using
the models and computations presented in this paper.
In brief, the aggregate demand of the household
sector is modeled by assuming the individual
heterogeneous agent displays cost minimization
behavior. Our hypothesis considers each country as
a representative agent that rationally optimizes the
simultaneous choice of a bundle of goods, based on
the aggregation of heterogeneous agents (Deaton
and Muellbauer 1980) within each country. The
theoretical model is used to create the parameters
for a model that is estimated econometrically, so that
its revealed behavior shows the demand elasticity of
each country for each good. The quantitative model
arrived at from the empirical estimation is the basis
for the allocation scheme. A detailed description of
the model specication and estimation can be found
in the technical appendix.
14
A Comparison of Alternative Programs for Climate Policies
Empirical Results and Discussion on
Alternative Allocation Options
We construct data for the household sector
by considering the nal consumption
expenditure – composed of two goods,
energy consumption and other goods consumption
– for 106 countries for the period 2000 to 2013.
Quantities of energy consumption, in tonnes of oil
equivalent (toe), include both energy for residential
uses and energy for transportation, representing
the actual direct expenditure of households for all
energy uses, in constant 2005 U.S. dollars. This
allows us to capture the households’ preferences
for direct energy use for all household needs.
The average share of energy consumption in
household expenditure is around 18 percent. For
OECD countries, the average is lower, at around 9
to 10 percent. Brazil and Russia show a share of 8
percent, with China at around 21 percent. Energy is
used as a factor of production in the industrial sector
and is incorporated in the value of goods – both
produced domestically and imported – that are sold
to nal consumers, according to traditional input-
output accounts. For this scheme to be valid, we
assume that households are aware of the indirect
energy content of the goods they consume, and do
not discriminate between direct and indirect energy
consumption, knowing that they are the nal bearer
of all forms of energy taxation for environmental
policy. The same reasoning applies to energy
use in the transportation sector. Households are
responsible for a large share of total diesel and
gasoline consumption, according to the input-output
data for the largest economies, as shown in Table 2.
In particular, households’ consumption of gasoline
is around 90 percent of the total for most countries.
We estimate demand functions according to
equation (7) in the Appendix at both stages, using
the seemingly unrelated regression (SUR) method.
We derive unconditional elasticities for each country
over the period 2000-2013. Our estimations are
based on observations for 106 countries and reect
a plausible accuracy. The econometric tests indicate
that the simultaneous estimation method of the
equations is signicant and better than the single
equation approach. Detailed results are shown in
the Appendix.
Table 2. The share of household gasoline and diesel household consumption.
Note: The gures represent the shares of household consumption as part of total country consumption.
Source: KAPSARC.
Country Diesel share Gasoline share
Italy 31 88
Germany 33 84
U.S. 349
China 15 81
India 22 89
France 47 90
U.K. 32 94
Brazil 17 86
15
A Comparison of Alternative Programs for Climate Policies
Empirical Results and Discussion on Alternative Allocation Options
We analyze household energy demand to meet both
residential and transportation needs; the resulting
demand elasticities reect this aggregate behavior.
We estimate the own-price elasticity, which shows
the (percentage) change in consumption of a
particular fuel when the price for that fuel changes.
In other words, it measures the intensity of consumer
responses to price changes. The estimated own-
price elasticities for energy are signicantly different
across regions and countries and represent results
for household behavior worldwide (shown in Figure 2
and Table A1 in the Appendix).
Note that in Figure 2 the elasticity values range
from low to high absolute levels. The world
average is around 0.28 in absolute value. The
OECD average is higher, at around 0.32. The
least developed countries’ average elasticity is
around 0.21, and the elasticity of energy- abundant
countries is around 0.23. The behavior of Brazil,
Russia, India, China and South Africa, the ‘BRIC’
countries is similar to that of OECD countries.
Many Eastern countries and poor African countries
have values around 0.1, and some large European
countries have values around 0.4.
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
France
OECP average*
Germany
Japan
USA
UK
BRICS
China
India
Energy-abundant countries**
Saudi Arabia
Russia
Least developed***
High income***
World
Average Energy Price Elasticity
(*) OECD excludes Israel, Estonia and Iceland.
(**) Net energy exporters, excludes Iran, Iraq and Venezuela.
(***) Following the World Bank denition, includes economies with a gross national income per capita above $12,735 for high income and below
$1,242 for least developed (for 2014).
Source: KAPSARC.
Figure 2. Energy price elasticities by regional averages – 2013.
16
A Comparison of Alternative Programs for Climate Policies
Empirical Results and Discussion on Alternative Allocation Options
Numerous empirical estimations of energy
demand elasticities appear in the existing
literature. We checked our results against
individual country estimations and worldwide
comparative estimations. In particular, we
considered two recent analyses of household
behavior. Dahl (2012) provides elasticity estimates
for transportation fuels only and Atalla and
Hunt (2015) analyze demand elasticity solely
for residential household energy consumption.
Ourresults are generally consistent with both
studies.The heterogeneity in these elasticities
justies our analysis of optimal prices by taking
into account global wellbeing in constructing the
social optimal pricing scheme.
The shares of world GDP for the main regions and
countries are reported in Figure 3. For example,
the OECD region accounts for 77 percent of world
GDP and 20 percent of the world population,
while energy abundant countries account for 2.5
percent of GDP and 7.9 percent of the population.
0
10
20
30
40
50
60
70
80
90
100
OECD average*
France
Germany
Japan
USA
UK
BRICS
China
India
Energy-abundant countries**
Saudi Arabia
Russia
Least developed***
High income***
Countries’ GDP as a share of global GDP
(*) OECD, excludes Israel, Estonia and Iceland.
(**) Net energy exporters, excludes Iran, Iraq and Venezuela.
(***) As per the World Bank denition, includes economies with a gross national income per capita above $12,735 for high income and below
$1,242 for least developed (for 2014).
Source: KAPSARC.
Figure 3. Countries’ GDP as a share of global GDP – year 2013.
17
A Comparison of Alternative Programs for Climate Policies
Empirical Results and Discussion on Alternative Allocation Options
We design alternative taxation pricing options,
taking as a constraint the amount needed worldwide
for climate policy, which is the exogenous amount
G. This amount is determined in equation (6) in
the Appendix. The latest IEA Scenarios (IEA 2016)
forecast $100 billion of additional investment per
year will be required to support mitigation policy
to stabilize atmospheric CO2 concentration at
450 ppm by 2030. In every scenario we calculate
the allocation of this investment among countries
according to alternative taxation options. The
resource constraint is as follows: we consider that
the amount of $100 billion (in real 2005 U.S. dollars)
is on average 0.2% of world GDP, 0.3% of total
household expenditure worldwide, or 2.5% of total
household energy expenditure worldwide.
We construct seven alternative scenarios for energy
price taxation of the household sector in each
country (Table 3).
In each scenario, we introduce a surcharge on
the existing energy price. The taxation revenue
worldwide is the same for all scenarios and is
constrained by the policy target. Scenario 1
designs a taxation burden for each country that is
proportional to each country’s share of world GDP.
In Scenario 2, the allocation burden is proportional
to each country’s total household consumption
expenditure, and in Scenario 3 the allocation
burden is proportional to each country’s household
expenditure on energy. In Scenario 4, the allocation
burden is proportional to each country’s carbon
emissions. Each of these four scenarios imposes
a taxation burden proportional to a measure of the
size of each country as a share of world GDP.
We nd that Scenario 5 is the optimal Ramsey pricing
scheme, based on the estimated price elasticities.
We compute optimal Ramsey prices for every
country using the inverse of the absolute value of that
country’s estimated energy demand elasticity.
Scenarios 6 and 7 consider only the top countries
in term of emissions and GDP per capita, in the
spirit of the Paris COP 21 agreement. In this way,
we identify the top 67 richest countries, for which
we compute, in Scenario 6, the burden share of
participating countries using their GDP shares, and
Scenario 7 displays the burden based on the optimal
Ramsey shares.
Table 3. Description of various scenarios implemented.
Source: KAPSARC.
Scenario Description
1Allocation based on GDP shares
2Allocation based on household consumption expenditure shares
3 Allocation based on household energy expenditure shares
4Allocation based on carbon emissions shares
5 Allocation based on Ramsey optimal pricing
6 Allocation based on the GDP shares of top 67 countries as per COP 21
7 Allocation based on Ramsey optimal pricing of top 67 countries as per COP 21
18
A Comparison of Alternative Programs for Climate Policies
Empirical Results and Discussion on Alternative Allocation Options
For all scenarios, we calculate the deadweight loss
in each country associated with different taxation
schemes. Deadweight loss is represented by the
area under the energy demand function for each
scenario between the pre- and post-taxation price
of energy, after the imposition of the taxation
surcharge. We set out the detailed results for all
scenarios in Tables 4 and 5.
Scenarios 1 through 4 are variants of the
proportional principle, where each country shares
the world climate policy cost in proportion to
the size of its GDP, household consumption
expenditure, household energy expenditures or
carbon emissions, respectively. This implies that
the weighting of each country in terms of total world
wellbeing forms the basis for its contribution to
climate policy costs.
Scenario 5 is based on household behavior, as
reected in energy price elasticity, to minimize
welfare losses resulting from the policy action taken.
In this case, each country’s contribution is based on
its human and economic behavioral decision-making,
as shown by observed utility maximization behavior.
The optimal taxation regime based on this scenario
may lead to inequalities because poorer developing
countries use largely outdated and inefcient
equipment and have less exibility in fuel choice. This
brings lower demand elasticity. As a result, these
countries may end up being taxed a higher portion of
their income. Equation 5 reects this.
An interesting result from Scenario 5 is that
optimal taxation would impose a lower burden on
leading polluters such as the U.S., Japan, Brazil,
United Kingdom, France and Italy and a higher
burden on China, India and Russia, reecting their
lower elasticities.
Table 5 sets out the results of Scenarios 6 and 7,
in which the 67 richest countries, as measured
by GDP per capita, share the burden of climate
policy. This group includes China, which has the
lowest GDP per capita in the group but which is the
largest emitter. In total, this subset of 67 countries
is responsible for around 80 percent of 2014 global
carbon emissions.
The climate policy cost for each country varies
among the scenarios. In Scenario 1, the U.S.
and China have shares of 29.5 percent and 7.8
percent, respectively. These values decrease to 22
percent and 6 percent in Scenario 5, 28 percent
and 8 percent in Scenario 6 and 27 percent and 10
percent in Scenario 7. These results see China’s
tax burden vary considerably; more so when using
the Ramsey scheme in Scenario 7, compared with
Scenario 5. This shows that China must pay a high
price to join the club of the richest countries, and
thus the group of top donors, under the efcient
Ramsey tax allocation.
Table 4 details the deadweight losses associated
with Scenarios 1 and 5 while Table 5 reports
deadweight losses for Scenarios 6 and 7.
Comparing the loss associated with the various
Scenarios, we nd that Ramsey pricing shows
the least loss. This is not surprising as the object
of this methodology is to reduce economic
inefciencies. A proportional tax set at the same
percentage in all countries yields a deadweight
loss around ve times larger than would be seen
with Ramsey pricing. The scheme charging higher
tax shares according to GDP (Scenario 1) yields
an even greater loss. Efciency would require a
higher burden on upper mid-sized economies, such
as Russia, India and Germany, and a lower burden
on the big three economies of the U.S., China and
Japan. This is quite different from the conventional
negotiation strategies that were implemented at
COP 21 and may implicitly explain the resistance of
some advanced economies that see themselves as
paying unjustly.
19
A Comparison of Alternative Programs for Climate Policies
Source: KAPSARC.
Empirical Results and Discussion on Alternative Allocation Options
Figure 4. Deadweight loss savings in Scenario 5 vs. Scenario 1 – Major countries (million) – 2013.
We can appreciate this point by comparing the
deadweight losses between Scenarios 1 and 5 for
major countries, as shown in Table 6 and Figure 4
(similar considerations can be made for Scenarios
6 and 7). In Scenario 1, the U.S. shows the highest
loss compared with other regions of the world.
Alternatively, in Scenario 5 the 28 EU member
countries show a deadweight loss higher than the
U.S. Note that the reduction in inefciency due
to Ramsey pricing of the rst 10 countries – with
a saving in deadweight loss as against Scenario
1 – is sufcient to compensate for the aggregate
deadweight loss suffered by the last 66 countries.
In addition, note that the monetary benet for the
winners far outweighs that for the losers (Table 6,
bottom rows).
In other words, the burden imposed on the less
efcient countries could be compensated for by the
efciency gains obtained by the richer and heavier-
polluting countries. These results may help to
rationalize the negative U.S. position, i.e., the U.S.
withdrawal from the Paris agreement. The U.S. would
have a signicant proportionately higher burden
under Scenario 1, the conventional burden sharing
mechanism, taking up more than half of the world's
deadweight loss, while its burden would be much lower
in absolute and relative terms, under Scenario 5.
0
500
1000
1500
2000
2500
3000
3500
4000
4500
U.S. Japan China Germany France Italy
U.K. Brazil Spain India
3894.57
787.17
526.75
259.19 132.54 130.23 120.36 63.7 53.39 49.42
20
A Comparison of Alternative Programs for Climate Policies
Table 4. Alternative taxation options – allocation of shares by countries worldwide – GDP and population shares
(average 2008-2012). Numbers in parentheses correspond to scenarios.
Country (1)
GDP
(2)
Total
expend.
(3)
Energy
expend.
(4)
Carbon
emission
(5)
Optimal
Ramsey
Deadweight
loss
scenario
(1)
Deadweight
loss
scenario
(5)
Population
share
Albania 0.02 0.03 0.04 0.02 0.02 -0.01 -0.03 0.05
Algeria 0.24 0.13 0.08 0.40 0.05 -0.20 -0.09 0.58
Armenia 0.01 0.02 0.01 0.02 0.00 0.00 -0.01 0.05
Australia 1.66 1.67 1.30 1.21 1.03 -17. 3 9 -1.66 0.36
Austria 0.68 0.56 0.70 0.21 0.46 -4.55 -0.75 0.14
Azerbaijan 0.05 0.05 0.04 0.11 0.01 -0.04 -0.02 0.14
Bahrain 0.04 0.03 0.02 0.08 0.04 0.00 -0.06 0.02
Bangladesh 0.16 0.23 0.08 0.19 0.09 -0.07 - 0.14 2.43
Belarus 0.09 0.20 0.16 0.21 0.35 -0.04 -0.57 0.15
Belgium 0.82 0.66 0.83 0.32 0.61 -5.99 -0.99 0.18
Bolivia 0.02 0.03 0.04 0.05 0.03 -0.01 -0.05 0.16
Bosnia-Herz. 0.03 0.04 0.04 0.08 0.08 0.00 -0.13 0.06
Brazil 2.21 2.80 2.07 1.44 0.94 -65.22 -1.52 3.18
Bulgaria 0.07 0.10 0.14 0.16 0.35 -0.02 -0.56 0.12
Burkina Faso 0.01 0.01 0.10 0.01 0.05 -0.02 -0.08 0.26
Cambodia 0.02 0.02 0.08 0.01 0.06 -0.01 -0.10 0.23
Cameroon 0.04 0.05 0.20 0.02 0.36 -0.03 -0.59 0.32
Canada 2.48 2.33 2.38 1.59 3.26 -2 7. 3 6 -5.28 0.56
Chile 0.31 0.36 0.45 0.26 0.60 -0.66 -0.97 0.28
China 7.87 5.18 9.57 29.57 5.55 -540.46 -13.71 21.81
Colombia 0.38 0.45 0.30 0.24 0.37 -0.58 -0.60 0.75
Congo DR 0.02 0.08 0.26 0.01 0.40 -0.02 -0.65 1.06
Costa Rica 0.05 0.07 0.06 0.03 0.06 -0.02 - 0.10 0.07
Cote d'Ivoire 0.04 0.05 0.18 0.02 0.83 -0.01 -1.34 0.32
Croatia 0.10 0.10 0.14 0.07 0.56 -0.02 -0.90 0.07
Cuba 0.11 0.09 0.02 0.12 0.01 -0.01 -0.02 0.18
Cyprus 0.04 0.04 0.06 0.02 0.03 -0.03 -0.05 0.01
Czech Rep. 0.31 0.22 0.42 0.36 2.03 - 0.17 -3.28 0.17
Denmark 0.53 0.42 0.66 0.13 0.44 -3.30 -0.71 0.09
Dom. Rep. 0.09 0.14 0.14 0.07 0.10 -0.12 -0.16 0.16
Ecuador 0.09 0.14 0.08 0.12 0.04 -0.09 -0.07 0.24
Egypt 0.25 0.42 0.12 0.72 0.09 -0.25 -0.14 1.32
El Salvador 0.00 0.06 0.07 0.02 0.04 0.00 -0.07 0.10
Ethiopia 0.04 0.11 0.30 0.02 0.74 -0.03 -1.20 1.35
Finland 0.42 0.36 0.50 0.18 0.38 -1.76 -0.61 0.09
France 4.54 4.06 3.76 1.11 2.98 -137. 3 6 -4.82 1.06
Gabon 0.02 0.01 0.06 0.01 0.03 -0.01 -0.06 0.02
Empirical Results and Discussion on Alternative Allocation Options
21
A Comparison of Alternative Programs for Climate Policies
Gambia 0.00 0.00 0.01 0.04 0.00 0.00 0.00 0.03
Georgia 0.02 0.03 0.03 0.03 0.01 -0.01 -0.02 0.07
Germany 6.11 5.13 6.52 2.39 6.14 -26 9.11 -9.92 1.33
Ghana 0.04 0.11 0.38 0.03 0.38 -0.09 -0.61 0.41
Greece 0.49 0.54 0.54 0.28 0.41 -2.25 -0.66 0.18
Guatemala 0.07 0.11 0.32 0.04 0.25 - 0.17 -0.40 0.23
Guinea 0.01 0.01 0.06 0.01 0.11 0.00 -0.18 0.16
Honduras 0.02 0.03 0.07 0.03 0.03 -0.02 -0.05 0.12
Hungary 0.23 0.20 0.35 0.16 0.35 -0.51 -0.57 0.16
India 2.51 2.97 3.05 6.80 2.78 -53.91 -4.49 19.41
Indonesia 0.78 1.17 1.08 1.85 1.03 -5.73 -1.67 3.91
Ireland 0.42 0.28 0.36 0.12 0.19 -1.81 -0.31 0.07
Italy 3.62 3.46 3.41 1.30 2.16 -123.86 -3.50 0.99
Japan 9.45 7. 3 0 8.20 3.89 5.08 -795.38 -8.21 2.08
Jordan 0.03 0.06 0.07 0.07 0.04 -0.03 -0.07 0.10
Kazakhstan 0.16 0.24 0.11 0.86 0.05 -0.24 -0.08 0.26
Kenya 0.05 0.08 0.59 0.04 2.50 -0.04 -4.04 0.67
Kyrgyz Rep. 0.01 0.01 0.02 0.02 0.01 0.00 -0.02 0.09
Latvia 0.03 0.05 0.07 0.03 0.07 -0.02 - 0.12 0.04
Lebanon 0.06 0.09 0.06 0.07 0.04 -0.03 -0.06 0.07
Libya 0.11 0.05 0.03 0.13 0.03 -0.02 -0.06 0.10
Lithuania 0.06 0.07 0.10 0.05 0.15 -0.02 -0.24 0.05
Luxembourg 0.08 0.05 0.10 0.04 0.11 -0.05 -0.17 0.01
Malaysia 0.37 0.31 0.38 0.74 0.73 -0.46 -1.17 0.46
Malta 0.01 0.01 0.01 0.01 0.01 0.00 -0.01 0.01
Mauritania 0.01 0.01 0.02 0.01 0.02 0.00 -0.03 0.06
Mexico 1.92 2.40 1.46 1.53 0.83 -31.42 -1.34 1.85
Moldova 0.01 0.02 0.02 0.02 0.01 0.00 -0.02 0.06
Mongolia 0.01 0.02 0.02 0.06 0.01 0.00 -0.02 0.04
Morocco 0.15 0.14 0.19 0.19 0.08 -0.43 - 0.13 0.52
Mozambique 0.02 0.03 0.12 0.01 0.27 -0.01 -0.43 0.38
Netherlands 1.40 1.01 1.16 0.55 1.58 -7.6 3 -2.55 0.27
N. Zealand 0.24 0.24 0.25 0.10 0.26 -0.36 -0.43 0.07
Niger 0.01 0.01 0.05 0.00 0.09 0.00 -0.14 0.25
Nigeria 0.32 0.59 1.67 0.29 5.40 -1.06 -8.72 2.58
Norway 0.66 0.55 0.47 0.15 0.27 -3.45 -0.43 0.08
Oman 0.09 0.06 0.04 0.21 0.02 -0.03 -0.04 0.05
Pakistan 0.27 0.55 0.43 0.54 1.44 -0.22 -2.33 2.83
Panama 0.05 0.05 0.06 0.03 0.04 -0.03 -0.07 0.06
Paraguay 0.02 0.03 0.04 0.02 0.06 0.00 -0.09 0.11
Peru 0.23 0.24 0.24 0.17 0.25 -0.33 -0.41 0.47
Philippines 0.27 0.34 0.53 0.27 0.77 -0.63 -1.25 1.52
Poland 0.79 0.79 1.34 1.04 2.50 -3.57 -4.05 0.63
Empirical Results and Discussion on Alternative Allocation Options
22
A Comparison of Alternative Programs for Climate Policies
Table 5. Scenario as per COP 21 Paris Agreement using optimal Ramsey tax and GDP share tax and deadweight
loss in million (average 2008-2012). Numbers in parentheses correspond to scenarios.
Portugal 0.40 0.40 0.49 0.16 0.32 -1.93 -0.51 0.17
Qatar 0.19 0.06 0.02 0.28 0.03 -0.02 -0.05 0.03
Romania 0.24 0.44 0.33 0.28 0.69 -0.23 -1.12 0.35
Russia 1.90 2.97 1.87 5.93 2.45 -17.18 -3.96 2.32
Rwanda 0.01 0.01 0.04 0.00 0.03 0.00 -0.04 0.17
Saudi Arabia 0.74 0.47 0.13 1.71 0.22 -0.37 -0.36 0.44
Serbia 0.06 0.11 0.12 0.16 0.22 -0.02 -0.36 0.12
Slovakia 0.12 0.13 0.15 0.11 0.12 - 0.15 -0.19 0.09
Slovenia 0.08 0.07 0.13 0.05 0.16 -0.05 -0.27 0.03
South Africa 0.60 0.78 0.76 1.56 0.47 -4.65 -0.76 0.81
South Korea 2.07 1.73 2.01 1.93 1.07 -49.79 -1.73 0.81
Spain 2.43 2.15 2.40 0.89 1.62 -56.01 -2.62 0.75
Sri Lanka 0.07 0.11 0.23 0.05 0.41 -0.05 -0.67 0.34
Sudan 0.08 0.14 0.19 0.05 0.13 -0.14 -0.20 0.65
Sweden 0.82 0.63 0.96 0.17 0.91 -5.33 -1.48 0.15
Switzerland 0.88 0.78 0.55 0.12 0.27 -6.29 -0.44 0.13
Tanzania 0.04 0.05 0.15 0.02 0.15 -0.03 -0.25 0.73
Thailand 0.43 0.39 0.69 0.99 0.37 -3.55 -0.59 1.13
Tunisia 0.08 0.09 0.11 0.08 0.19 -0.03 -0.31 0.17
Turkey 1.17 1.70 1.41 1.05 0.92 -16.02 -1.48 1.19
Ukraine 0.19 0.50 0.33 0.94 2.79 -0.05 -4.52 0.75
UAE 0.44 0.54 0.19 0.59 0.12 -0.87 -0.20 0.12
U.K. 4.86 5.15 3.99 1.47 3.62 -136.08 -5.85 1.02
U.S. 26.86 29.50 22.76 17.4 0 22.46 -3930.87 -36.30 5.04
Uruguay 0.05 0.07 0.07 0.03 0.37 0.00 -0.60 0.05
Vietnam 0.15 0.28 0.40 0.57 0.24 -0.66 -0.38 1.42
Total world 100.00 100.00 100.00 100.00 100.00 -6339.24 -166.39 100.00
Source: KAPSARC.
Note: Columns 1-5: country share in the scenario. Columns 6-7 deadweight loss $million. Column 8: country population share.
Country (6)
GDP share as per
COP 21
(7)
Optimal
Ramsey pricing as
per COP 21
Deadweight loss
using
GDP weights
Deadweight loss
using
Ramsey weights
Albania 0.02 0.03 -0.02 -0.01
Algeria 0.25 0.07 -0.27 -0.02
Australia 1.77 1.27 -23.63 -0.30
Austria 0.72 0.57 - 6.18 -0.13
Bahrain 0.04 0.04 -0.01 -0.01
Belarus 0.09 0.43 -0.06 - 0.10
Empirical Results and Discussion on Alternative Allocation Options
23
A Comparison of Alternative Programs for Climate Policies
Belgium 0.87 0.75 -8.13 - 0.18
Bosnia-Herzegovina 0.03 0.10 -0.01 -0.02
Brazil 2.35 1.16 -88.62 -0.27
Bulgaria 0.07 0.43 -0.03 - 0.10
Canada 2.63 4.02 - 37.18 -0.94
Chile 0.33 0.74 -0.90 - 0.17
China 8.35 10.45 -734.36 -2.45
Colombia 0.40 0.46 -0.79 - 0.11
Costa Rica 0.05 0.07 -0.02 -0.02
Croatia 0.10 0.69 -0.03 - 0.16
Cuba 0.12 0.02 -0.02 0.00
Cyprus 0.04 0.04 -0.04 -0.01
Czech Rep. 0.32 2.50 -0.23 -0.59
Denmark 0.56 0.54 -4.49 - 0.13
Dominican Rep. 0.10 0.12 - 0.16 -0.03
Ecuador 0.10 0.05 - 0.13 -0.01
Finland 0.45 0.47 -2.39 - 0.11
France 4.82 3.67 -186.64 -0.86
Gabon 0.02 0.04 -0.02 -0.01
Germany 6.49 7. 5 6 -365.66 -1.77
Greece 0.52 0.50 -3.06 -0.12
Hungary 0.24 0.43 -0.69 - 0.10
Ireland 0.45 0.24 -2.47 -0.06
Italy 3.84 2.67 -168.30 -0.63
Japan 10.04 6.25 -1080.75 -1.47
Kazakhstan 0.16 0.06 -0.33 -0.01
Latvia 0.04 0.09 -0.02 -0.02
Lebanon 0.06 0.05 -0.05 -0.01
Libya 0.11 0.04 -0.02 -0.01
Lithuania 0.06 0.19 -0.03 -0.04
Luxembourg 0.09 0.13 -0.07 -0.03
Malaysia 0.39 0.89 -0.63 -0.21
Malta 0.01 0.01 0.00 0.00
Mexico 2.04 1.02 -42.69 -0.24
Netherlands 1.49 1.94 -10.37 -0.46
New Zealand 0.25 0.33 -0.48 -0.08
Norway 0.70 0.33 -4.69 -0.08
Oman 0.09 0.03 -0.04 -0.01
Panama 0.05 0.05 -0.04 -0.01
Peru 0.25 0.31 -0.45 -0.07
Poland 0.83 3.08 -4.86 -0.72
Portugal 0.42 0.39 -2.63 -0.09
Qatar 0.20 0.04 -0.03 -0.01
Romania 0.25 0.85 -0.32 -0.20
Empirical Results and Discussion on Alternative Allocation Options
24
A Comparison of Alternative Programs for Climate Policies
Russia 2.02 3.01 -23.35 -0.71
Saudi Arabia 0.78 0.28 -0.51 -0.06
Serbia 0.06 0.27 -0.03 -0.06
Slovakia 0.13 0.14 -0.21 -0.03
Slovenia 0.09 0.20 -0.07 -0.05
South Africa 0.63 0.58 -6.31 -0.14
South Korea 2.20 1.32 - 67.65 -0.31
Spain 2.58 1.99 -76.10 -0.47
Sweden 0.88 1.12 -7. 2 5 -0.26
Switzerland 0.93 0.34 -8.55 -0.08
Thailand 0.45 0.45 -4.83 - 0.11
Tunisia 0.09 0.23 -0.05 -0.05
Turkey 1.25 1.13 -21.77 -0.27
UAE 0.47 0.15 -1.18 -0.04
U.K. 5.16 4.46 -184.90 -1.05
U.S. 28.52 2 7.66 -5341.16 -6.49
Uruguay 0.05 0.46 -0.01 - 0.11
World total 100.00 100.00 -8526.91 -23.48
Source: KAPSARC.
Note: Columns 1-2: country share in the scenario. Columns 3-4: deadweight loss $million.
Empirical Results and Discussion on Alternative Allocation Options
Note: Column 1: Deadweight loss of scenario 5 minus scenario 1.
Column 2: Deadweight loss of scenario 7 minus scenario 6.
Number of countries in parenthesis.
Source: KAPSARC.
Countries Scenario 5 vs.1 Scenario 7 vs. 6
U.S. 3894.57 5334.67
China 526.75 731.91
Japan 787.17 1079.28
Russia 13.22 22.64
EU_28 718.12 1031.64
Winners 6208.47 (39) 8504.24 (54)
Losers -35.75 (67) -0.75 (13)
Table 6. Deadweight loss savings under the Ramsey scheme ($billion) – year 2013.
25
A Comparison of Alternative Programs for Climate Policies
Conclusions and Policy Implications
We consider that the model we have
developed provides a coherent and
integrated empirical tool for policymakers
to quantify how energy demand responds to policy.
We created it by estimating a complete demand
system for world household energy consumption
behavior and used the resulting country price
elasticity values to compute an optimal Ramsey
price scheme to support investment in climate
change mitigation policies.
Our approach is unique in accurately estimating
demand response to prices by explicitly modeling a
utility-maximizing rational behavior for consumers in
each of the countries studied. The study nds that
households in emerging economies have less price
elasticity than advanced ones.
This leads us to two conclusions:
First, we have identied an efcient worldwide
taxation scheme to fund investments in climate
change mitigation policies. This taxation
strategy depends on the heterogeneity of
household behavior in countries around the
world. Accordingly, policy actions could be
designed around the efciency principle, with
eventual compensation for political reasons,
rather than on a debatable equity principle that
leads to greater economic inefciency.
Second, our empirical estimation shows
signicant differences in the burden allocation
when the allocation involves only the world’s
richest countries. However, this means that
some countries might pay a high ‘access price’
to be part of the group of richest countries.
To be successful, negotiations must deal with
the risk of opportunistic behavior by countries
attempting to avoid this high price.
The issue of raising taxes is a difcult task for
any policymaker because it inevitably involves
distortions. In general, a policymaker is confronted
with funding limitations and the associated efciency-
equity trade-off. Pricing policies based on the ability
to pay have a role in improving the living conditions
of poor households, but usually impose a societal
cost in terms of market inefciency. By contrast,
when implementing policies to promote maximum
economic efciency, the poorest in society often
suffer. Our research has shown that there is room for
compensation without compromising efciency.
In this respect, our approach demonstrates
to policymakers the quantitative range of the
associated efciency-equity trade-off. Our
estimations of demand elasticities should prove
useful in constructing the minimum distortion pricing
policy – the so-called Ramsey pricing – that is one
cornerstone of the efciency-equity trade-off. We
also construct the maximum equitable solution
based on countries’ share of total world GDP and/
or household expenditure and/or carbon emissions.
Policymakers also need to assess the economic
impact associated with equitable intervention, such
as price subsidies for the poor and elderly.
We provide a method of measuring this in terms of
deadweight loss. (A deadweight loss is the added
burden placed on consumers and suppliers when
the market equilibrium is altered because of, for
example, tax. It results when supply and demand
are out of equilibrium.)
Our results suggest that governments could
adopt a more inclusive approach, taking into
consideration the behavior of their populations
rather than purely abstract technical standards.
The estimated elasticities are behaviors that can be
monitored over time: as the population progresses
and becomes richer, tastes and behaviors evolve,
26
A Comparison of Alternative Programs for Climate Policies
Conclusions and Policy Implications
and consequently, the elasticities change and the
Ramsey scheme is updated.
In other words, given the political difculty of
implementing a pure textbook Pigouvian taxation,
our proposal could provide a different route. It
could also prove to be more viable, with higher
chances of political acceptance because it
minimizes the deadweight loss. Consequently,
this proposal is not a mere redistributive policy,
creating winners and losers within the political
constituencies involved, but one that creates a net
welfare improvement, i.e., reduced deadweight
loss. This is a positive item which may be used for
compensation and thus could enhance political
consensus.
In conclusion, this paper advocates a new global
policy stance that takes estimated consumer
demand elasticity as a new basis for discussion
and for differentiating taxation allocation worldwide.
Policymakers will likely be aware that they will
face politically responsible economic agents who
require an efcient proposal that is benecial to the
welfare of society to pay for investments to help
mitigate climate change.
27
A Comparison of Alternative Programs for Climate Policies
Atalla, Tarek N., and Lester C. Hunt. 2015. "Modeling
Residential Electricity Demand in the GCC Countries:
The Importance of Weather and Exogenous Stochastic
Trends." Available at SSRN Electronic Journal,
DOI: 10.2139/ssrn.2652814
Atalla, Tarek N., and Patrick Bean. 2015 "Determinants
of Energy Productivity: A Comparison of 39 Countries."
Available at SSRN Electronic Journal,
DOI: 10.2139/ssrn.2602122
Babusiaux, Denis, and Axel Pierru. 2007. "Modelling and
allocation of CO2 emissions in a multiproduct industry:
the case of oil rening." Applied Energy 84, 7: 828-841.
DOI: 10.1016/j.apenergy.2007.01.013
Berry, Keith, 2002. Generation search costs and Ramsey
pricing in a partially deregulated electric utility industry,
Journal of Economic and Business, 5: 331-343. DOI:
10.1016/s0148-6195(02)00064-4
Bigerna, Simona and Carlo Andrea Bollino. 2015. “A
system of hourly demand in the Italian electricity market.”
Energy Journal 36: 1-19. DOI: 10.5547/01956574.36.4.sbig
Bigerna, Simona and Carlo Andrea Bollino. 2016.
"Ramsey prices in the Italian electricity market." Energy
Policy 88: 603-612. DOI: 10.1016/j.enpol.2015.06.037
Bigerna, Simona, Carlo Andrea Bollino, and Silvia
Micheli. 2016 "Costs assessments of European
environmental policies." Computers & Operations
Research 66: 327-335. DOI: 10.1016/j.cor.2015.03.003
Boeters Stefan. 2014. "Optimally differentiated carbon
prices for unilateral climate policy." Energy Economics
45: 304–312. DOI: 10.1016/j.eneco.2014.07.015
Bollino, Carlo Andrea. 1987. “Gaids: a generalized
version of the almost ideal demand system.” Economic
Letters. 23: 199-202. DOI: 10.1016/0165-1765(87)90039-5
Browning, Martin, and Costas Meghir. 1991. "The
effects of male and female labor supply on commodity
demands." Econometrica: Journal of the Econometric
Society: 925-951. DOI: 10.2307/2938167
Dahl, Carol A. 2012. "Measuring global gasoline and
diesel price and income elasticities." Energy Policy 41:
2-13. DOI: 10.1016/j.enpol.2010.11.055
Diamond, Peter A.. 1975. “A many-person Ramsey tax
rule.” Journal of Public Economics, 4: 335-342.
DOI: 10.1016/0047-2727(75)90009-2
Deaton, Angus and John Muellbauer. 1980. Economics
and consumer behavior. Cambridge: Cambridge
University Press. DOI: 10.1017/cbo9780511805653
Deeney, Peter, et al. 2016. "Inuences from the European
Parliament on EU emissions prices." Energy Policy 88:
561-572. DOI: 10.1016/j.enpol.2015.06.026
Gasim, Anwar A. 2015. "The embodied energy in trade:
What role does specialization play?" Energy Policy 86:
186-197. DOI: 10.1016/j.enpol.2015.06.043
Hakimov, Rustamdjan and John Mueller. 2014. Charges
of uncongested German airports: Do they follow Ramsey
pricing scheme?, Research in Transportation Economics,
45, 57-65. DOI: 10.1016/j.retrec.2014.07.008
Hu, Jing, et al. 2015. "Ex-ante evaluation of EU ETS
during 2013–2030: EU-internal abatement." Energy
Policy 77: 152-163. DOI: 10.1016/j.enpol.2014.11.023
International Energy Agency (IEA). 2016. World energy
outlook 2015. Paris, France. DOI: 10.1787/weo-2016-en
Laffont, Jean-Jacques and Jean Tirole. 1996. "Creating
competition through interconnection: Theory and
practice.” Journal of Regulatory Economics, 10, 3: 227-
256. DOI: 10.1007/bf00157671
References
28
A Comparison of Alternative Programs for Climate Policies
References
Lin, Cinthia and Lea Prince. 2009. “The optimal gas tax
for California.” Energy Policy 37: 5173-5183.
DOI: 10.1016/j.enpol.2009.07.063
Martín-Cejas, Roberto Rendeiro. 2010. “Ramsey pricing
including CO2 emission cost: An application to Spanish
airports.” Journal of Air Transportation Management,
16: 45–47. DOI: 10.1016/j.jairtraman.2009.07.001
Matsukawa, Isamu, Seishi Madono and Takako
Nakashima. 1993. “An empirical analysis of Ramsey
pricing in Japanese electric utilities.” Journal of the
Japanese and International Economies, 7, 3: 256-276.
DOI: 10.10 0 6/jjie.1993.1015
Nahata, Babu, Alexei Izyumov, Vladimir Busygin and Anna
Mishura. 2007. “Application of Ramsey model in transition
economy: A Russian case study.” Energy Economics, 29:
105-125. DOI: 10.1016/j.eneco.2005.09.011
National Academy of Sciences (NAS). 1992. Policy
Implications of Greenhouse Warming: Mitigation,
Adaptation, and the Science Base, Panel on Policy
Implications of Greenhouse Warming, Committee on
Science, Engineering, and Public Policy Washington
D.C.: National Academy Press. DOI: 10.17226/1605
Qi, Fang, Lizi Zhang, Bin Wei and Guanghui Que.
2009. “An application of Ramsey pricing in solving the
cross- subsidies in Chinese electricity tariffs.” Institute of
Electrical and Electronics Engineers, 442–447.
DOI: 10.1109/drpt.2008.4523447
Ramsey, Frank. 1927. “A contribution to the theory of
taxation.” Economic Journal, 37: 47-61.
DOI: 10.2307/2222721
Santos, Paulo, Lima Marangon, Rafael Leme and Tiago
Leite Ferreira. 2012. “Distribution charges for consumers
and microgeneration considering load elasticity
sensitivity.” Energy Economics, 34, 2, pp. 468-475.
DOI: 10.1016/j.eneco.2011.12.003
Sun, Chuanwang, and Boqiang Lin. 2013. "Reforming
residential electricity tariff in China: Block tariffs pricing
approach." Energy Policy 60: 741-752.
DOI: 10.1016/j.enpol.2013.05.023
United Nations. 2015. Conference of the Parties (COP21)
FCCC/CP/2015/L.9/rev1. Accessed April 16, 2016
https://unfccc.int/resource/docs/2015/cop21/eng/l09r01.pdf
United Nations Framework Convention on Climate
Change (UNFCCC). 2016. Communication from the
Chinese delegation, “Enhanced actions on climate
change: China’s intended nationally determined
contributions.” Accessed April 16, 2016
www4.unfccc.int/submissions/INDC/Published%20
Documents/China/1/China's%20INDC%20%20on%20
30%20June%202015.pdf
Van der Ploeg, Frederick. 2016. “Second-best carbon
taxation in the global economy: The Green Paradox
and carbon leakage revisited.” Journal of Environmental
Economics and Management 78:85–105.
DOI: 10.1016/j.jeem.2016.02.006
Weitzman, Martin. L. 2017. “Voting on prices vs. voting
on quantities in a world climate assembly.” Research in
Economics, 71, 2: 199-211.
DOI: 10.1016/j.rie.2016.10.004
Zhang, Junjie, Zhenxuan Wang, and Xinming Du. 2017.
“Lessons Learned from China’s Regional Carbon Market
Pilots.” Economics of Energy & Environmental Policy 6,
2,: 1-20. DOI: 10.5547/2160-5890.6.2.jzha
29
A Comparison of Alternative Programs for Climate Policies
Technical Appendix
The country agent faces the simultaneous
choice between ‘energy’ e and a ‘composite
good’ y, which is a representation of the rest
of the goods and services demanded by the agent.
The optimal allocation is dependent on a set of
exogenous variables not explicitly considered by the
agent in the preference set (Browning and Meghir
1991). These variables include the available capital
stock and country-specic climate conditions.
Both variables inuence the allocation between
energy and other goods, insofar as they capture
the level of available technology and the country’s
natural environment. We are particularly interested
here in accurately modeling energy demand,
recognizing that including capital stock and climate
ensures unbiased empirical estimates (Deaton and
Muellbauer 1980).
Formally, the country agent’s consumption cost
function can be dened as:
C=C(pe,py,K,W, U)= min [(pe e + py y) │ U(e,y,K, W)]
(1)
which depends on the prices of energy and of the
composite good [pe, py], total utility U, the capital
stock K and country-specic climate conditions W.
An indirect utility function can be dened, inverting
equation (1):
U=V(pe, py,K, W, C)
(2)
from which the Marshallian demand function can
be dened using Roy’s identity. The duality theory
allows demand functions to be expressed using
Roy’s identity, which states the demand function is
the negative of the ratio of the partial derivative of
the indirect utility function with respect to price and
the partial derivative with respect to expenditure:
hi= - ∂V/∂pi /∂V/∂C = gi(py, pe, K, W, C) where i= y, e
(3)
Equation (3) denes the demand functions hi for
i=[y, e], where hy is the demand for composite good y
and he is demand for energy. From equation (3) price
demand elasticity for each good can be calculated.
We assume that policymakers have correct
knowledge of country-specic demand functions and
are willing to charge optimal prices to buyers in each
country, taking into account efciency objectives.
We also assume that policymakers consider the
observed price without carbon-associated externalities
and want to determine the optimal charge tj to be
levied to each country j that must be added to the
market price to satisfy the constraint that the total tax
revenue equals the agreed-upon world target:
pej*=pe + tj
(4)
In equation (4) the optimal Ramsey price is dened
as the sum of the observed energy market price and
the country's optimal tax.
The optimal Ramsey (1927) price can be computed
as for all countries j as:
[(p*ej – pe)/pe] / [(p*ei - pe)/pe] = (1/|εj|) / (1/|εi|)
(5)
subject to the constraint:
G = ∑pejej = ∑pej*ej
(6)
Equation (5) states that the relative increase in
the observed market price in country j over that in
country i is inversely proportional to the ratio of the
demand elasticities of the two counties. This denes
the efcient price p*ej increase over the market
price, where p*ej and p*ei are optimal prices, pej are
historical country prices, ej are quantities, and εej
are estimated own price elasticities of energy for all
countries i and j.
30
A Comparison of Alternative Programs for Climate Policies
Technical Appendix
The quantity G in equation (6) is the exogenous
target revenue. It can be interpreted as the world
revenue derived by charging tj to each country j
according to its behavior. In the Ramsey scheme,
there is no room for distributive equity considerations
because the aim is to minimize inefciency.
The empirical specication of equation (3) is the
Generalized Almost Ideal demand model proposed
by Bollino (1987), which satises consumer theory
restrictions, i.e., adding up, symmetry, homogeneity
and heterogeneous consumer exact aggregation
constraints. We use this parametrization because
it is suitable for the estimation of exible demand
behavior, especially with large variability across
heterogeneous agents (Bigerna and Bollino 2015).
Formally, the parametric function to be estimated is:
hij= γij + Ej
s/pi [αij + ∑t αtij ln(pt) + βij ln(Ej
s/ps)] Ejs =
Ej–(∑γij pj)
ps = ∑wj ln(pj)
(7)
For the two demand functions, i = e,y and for each
country j. Ej and Ej
s denotes the expenditure and
supernumerary expenditure, respectively; while ps
denotes the Stone price aggregator.
The estimated parameters are γij, γij expressing
the committed quantity parameters; αij, αij, αitj, αifj,
βij, βij are structural coefcients; wi, are average
budget shares. After the econometric estimation of
the structural parameters, we take the derivatives
of equations (7) and (8) with respect to prices to
compute the elasticities εij for goods i and for each
country j, i at the equilibrium prices and quantities.
We set out below the shares of goods in the
household budget and the estimated elasticities
(Table A1).
31
A Comparison of Alternative Programs for Climate Policies
Technical Appendix
Table A1. Shares of goods and energy in nal household consumption and energy demand elasticity – average
2008-2012(*)
(*)W1 = share of other goods; W2 = share of energy; Elasticity = energy demand elasticity
Country W1 W2 Elasticity
Albania 0.822 0.178 -0.53
Algeria 0.926 0.074 -0.40
Armenia 0.910 0.090 -0.75
Australia 0.911 0.089 -0.32
Austria 0.858 0.142 -0.38
Azerbaijan 0.912 0.088 -0.84
Bahrain 0.897 0.10 3 - 0.17
Bangladesh 0.962 0.038 -0.22
Belarus 0.885 0.115 -0.12
Belgium 0.856 0.144 -0.35
Bolivia 0.879 0.121 -0.31
Bosnia-Herz. 0.872 0.128 - 0.15
Brazil 0.913 0.087 -0.56
Bulgaria 0.837 0.16 3 - 0.10
Burkina Faso 0.171 0.829 -0.54
Cambodia 0.625 0.375 -0.32
Cameroon 0.506 0.494 - 0.14
Canada 0.883 0.117 -0.18
Chile 0.856 0.144 - 0.19
China 0.785 0.215 -0.44
Colombia 0.924 0.076 -0.20
Congo DR 0.639 0.361 - 0.16
Costa Rica 0.904 0.096 -0.24
Cote d'Ivoire 0.546 0.454 -0.05
Croatia 0.838 0.162 -0.06
Cuba 0.979 0.021 -0.30
Cyprus 0.851 0.149 -0.50
Czech Rep. 0.785 0.215 -0.05
Denmark 0.823 0.177 -0.38
Dominican Rep. 0.881 0.119 -0.36
Ecuador 0.931 0.069 -0.48
Egypt 0.967 0.033 -0.34
El Salvador 0.860 0.140 -0.41
Ethiopia 0.615 0.385 -0.10
Finland 0.841 0.159 -0.33
France 0.894 0.10 6 -0.32
Gabon 0.564 0.436 - 0.41
Gambia 0.659 0.341 -3.60
32
A Comparison of Alternative Programs for Climate Policies
Georgia 0.850 0.150 -0.64
Germany 0.855 0.145 -0.27
Ghana 0.616 0.384 -0.26
Greece 0.884 0.116 -0.34
Guatemala 0.676 0.324 -0.32
Guinea 0.538 0.462 - 0.14
Honduras 0.763 0.237 -0.51
Hungary 0.800 0.200 -0.25
India 0.877 0.123 -0.28
Indonesia 0.891 0.109 -0.27
Ireland 0.855 0.145 -0.48
Italy 0.888 0.112 -0.40
Japan 0.872 0.128 -0.41
Jordan 0.863 0.137 -0.43
Kazakhstan 0.948 0.052 -0.57
Kenya 0.186 0.814 -0.06
Kyrgyz Rep. 0.809 0.191 -0.47
Latvia 0.819 0.181 -0.25
Lebanon 0.923 0.077 -0.39
Libya 0.938 0.062 -0.20
Lithuania 0.833 0.167 - 0.16
Luxembourg 0.766 0.234 -0.25
Malaysia 0.858 0.142 -0.13
Malta 0.893 0.107 -0.42
Mauritania 0.721 0.279 -0.24
Mexico 0.930 0.070 -0.45
Moldova 0.837 0.163 -0.57
Mongolia 0.796 0.204 -0.57
Morocco 0.851 0.149 -0.60
Mozambique 0.536 0.464 - 0.11
Netherlands 0.869 0.131 - 0.19
New Zealand 0.882 0.118 -0.24
Niger 0.500 0.500 -0.14
Nigeria 0.608 0.392 -0.08
Norway 0.902 0.098 -0.45
Oman 0.935 0.065 -0.39
Pakistan 0.908 0.092 -0.07
Panama 0.855 0.145 -0.36
Paraguay 0.828 0.172 - 0.18
Peru 0.888 0.112 -0.24
Philippines 0.819 0.181 - 0.17
Poland 0.806 0.194 - 0.14
Portugal 0.861 0.139 -0.39
Qatar 0.954 0.046 -0.20
Technical Appendix
33
A Comparison of Alternative Programs for Climate Policies
Source: KAPSARC.
Romania 0.915 0.085 - 0.12
Russia 0.926 0.074 -0.19
Rwanda 0.552 0.448 -0.44
Saudi Arabia 0.967 0.033 - 0.15
Serbia 0.870 0.130 - 0.14
Slovakia 0.870 0.130 -0.32
Slovenia 0.784 0.216 -0.20
South Africa 0.887 0.113 -0.41
South Korea 0.867 0.133 -0.48
Spain 0.872 0.128 -0.38
Sri Lanka 0.771 0.229 -0.14
Sudan 0.822 0.178 -0.38
Sweden 0.825 0.175 -0.27
Switzerland 0.919 0.081 -0.52
Tanzania 0.763 0.237 -0.22
Thailand 0.795 0.205 -0.48
Tunisia 0.854 0.146 - 0.15
Turkey 0.904 0.096 -0.39
Ukraine 0.922 0.078 -0.03
U.A.E. 0.958 0.042 -0.40
U.K. 0.912 0.088 -0.28
U.S. 0.912 0.088 -0.26
Uruguay 0.877 0.123 -0.05
Vietnam 0.827 0.173 -0.43
Technical Appendix
34
A Comparison of Alternative Programs for Climate Policies
Carlo Andrea Bollino
Carlo Andrea is a visiting fellow at KAPSARC and Full Professor of
Economics at the University of Perugia, Italy, specializing in microeconomic
analysis, energy economics and environmental policy. He holds a Ph.D. in
Economics from the University of Pennsylvania.
Simona Bigerna
Tarek Atalla
Simona is Associate Professor of Economics at the University of Perugia,
Italy, specializing in energy economics, environmental policy analysis,
transport and international economics. She holds a Ph.D. in Economics.
Tarek is a former KAPSARC research associate evaluating energy
productivity investments, the economics of energy vulnerability and the
effect of climate on energy consumption patterns. He holds a Ph.D. in
Economics from Université Paris Dauphine.
About the Authors
About the Project
Although there has been extensive analysis of different schemes of allocations of nancial
burdens across countries to implement climate mitigation policies, no attempt has previously
been made to estimate the use of efcient Ramsey allocation schemes.
For the rst time, and with the aid of an extensive data set, demand elasticities have been
estimated for 106 countries, from which a Ramsey allocation scheme has been computed.
This modeling exercise showed a number of cost reduction opportunities in using a Ramsey
allocation scheme.
35
A Comparison of Alternative Programs for Climate Policies
Notes
36
A Comparison of Alternative Programs for Climate Policies
www.kapsarc.org
