THE SCOPE OIN 2009 A

PHASING OU

An IEA, OE

Prepared for the

1 This Joint Report does not express theconstrued as interpreting or modifying aas expressing any legal opinions or havin

The World

OF FOSSIL-FUEL SUBSAND A ROADMAP FOUT FOSSIL-FUEL SUBS

ECD and World Bank Joint Report1

e G-20 Summit, Seoul (Republic of Kore

11-12 November 2010

positions of the G20 Member Countries. Nothing in this Joiany legal obligations under the WTO Agreements, treaty, lawng probative legal value in any proceedings.

Bank

IDIES OR

IDIES

ea)

nt Report shall be w or other texts, or

Page | 2

Table of Contents

Executive Summary ................ 1.The scope of fossil-fuel subs

1.1 Introduction ..................1.2 Measuring fossil-fuel s1.3 Estimate of global foss1.4 Implications of phasing1.5 Recent action taken to

2.A roadmap for phasing out f

2.1. Analytical Framework .2.2 Lessons from recent exp

References ..............................

The World

.........................................................................

idies in 2009 .............................................................................................................................ubsidies ...........................................................il fuel consumption subsidies ..........................

g out fossil-fuel consumption subsidies........... phase out subsidies ........................................

fossil fuel subsidies ...................................................................................................................

perience of energy subsidy reforms ................

.........................................................................

Bank

................... 3

................... 5

.................... 5

.................. 14

.................. 17

.................. 19

.................. 22

................. 27

.................. 27

.................. 37

................. 47

Executive SummaThis report, prepared for the G-2Economic Co-operation and Devfossil-fuel subsidies in 2009 and p

The IEA estimates that direct slowering end-user prices for fossof mechanisms can be identifiedfuel production or consumptioresources under government covia concessional loans or guarantcompared with direct consumer s

Phasing-out fossil-fuel subsidiessecurity, reduce emissions of grhighlighted by estimates from thphased-out by 2020, it would amounts to the current consumoil demand, the savings amountwould also represent an integralin carbon-dioxide emissions wou

Furthermore, OECD and IEA analeconomic gains as in many unsupportable fiscal burden on example, the IEA estimates thatlikely to reach almost $600 billiocountries emerge from the ecoinefficient fossil fuel subsidies, opoverty alleviation, health and ed

Since the commitment taken at toutside the G-20 are moving aheextent of the potential gains will the reforms they are pursuing.

The World Bank’s contributionremoval, revisited through the posome quick diagnostics of the key

• Who has been benefittinterms and for wasteful cremoving the subsidy onSubsidy removal will, housed to satisfy basic nee

2 An OECD expert workshop on 18-1types of fossil-fuel subsidies.

The Worl

ry 20 by the International Energy Agency (IEA), the Ovelopment (OECD) and the World Bank, estimateprovides a roadmap for phasing-out fossil-fuel sub

subsidies that encourage wasteful consumptionsil fuels amounted to $312 billion in 2009. In additd, also in advanced economies, which effectively n, such as tax expenditures, under-priced accntrol (e.g. land) and the transfer of risks to govetees). These subsidies are more difficult to identifysubsidies.2

s represents a triple-win solution. It would enreenhouse gases and bring immediate economiche IEA that indicate that if fossil-fuel subsidies we

cut expected growth in global energy demandption of Japan, Korea and New Zealand combine

t to 4.7 mb/d, or around one-quarter of current l building block for tackling climate change as exld be cut by 2 gigatonnes.

lyses suggest that subsidy reform would bring abocases they are creating market distortions, government budgets and are weakening trade

t, in the absence of reform, spending on fossil-fuon in 2015, or 0.6 percent of global gross domestonomic crisis, the revenues that can be saved for redirected to more directly tackle pressing priducation, will be important.

the Pittsburgh Summit in 2009, many countries bead with reforms. While this is a very encouraging

only be realised if more countries raise the level

provides a road map for implementing fossioverty lens. Such a roadmap may help policy maky problems and the required policy response:

ng from an existing subsidy? If it is primarily the ronsumption, as is often the case, then there is a sn equity grounds as well as for improved econo

owever, have a negative impact on the poor, if theds, rather than to encourage wasteful consump

19 November 2010 will examine methods for estimati

orld Bank

Page | 3 rganisation for

es the scope of bsides.

n by artificially tion, a number support fossil-

cess to scarce ernments (e.g. y and estimate

nhance energy c gains. This is ere completely d by 5%. This ed. In terms of US demand. It pected growth

out immediate imposing an

balances. For uel subsidies is tic product. As

from removing orities such as

oth within and g start, the full of ambition in

il fuel subsidy kers in deriving

ich in absolute strong case for mic efficiency.

he subsidy was ption. Schemes

ng the different

Page | 4

assisting households withome heating (such as thcountries) are described.providing incentives for d

• Assuming that there is athose? The answer will dIf it was to make the support programs or (seenergy access available connection costs) with fu

• After the implementatioenvironmentally welfare eventual adverse social im

Lessons drawn from recent expfuel subsidies that largely benefit

• Well designed rural elecpoor

• Better-targeted compenaimed at protecting the m

• Moves towards automatfor fossil fuels.

The World

th only the portion of residential energy costs he US LIHEAP, successfully replicated in some Eas. Alternative schemes supporting new gas connecdemand side management are discussed.

an impact on the poor, what are the options fodepend in part on what the intended effect of thexisting use of energy more affordable, then

econd best) lifeline tariffs can be considered. If ito new households, then switching the subsidy

ull payment of incremental consumption costs is r

on of subsidy phasing out, the report discussenhancing ways to reallocate the savings to mitig

mpacts.

erience also suggest more effective alternativest higher income households, including the followin

ctrification subsidies to make energy services aff

nsation packages for poorest households or bromost vulnerable.

tic price adjustments mechanisms and fully libe

Bank

that goes for stern European ction as well as

or ameliorating e subsidy was. income-based t was to make to access (e.g.

recommended.

ses socially or gate and offset

s to regressive ng:

ordable to the

oader reforms

ralized system

1. The scope of fo

1.1 Introduction

1.1.1 Background to the

In September 2009, G-20 leadersummit in Pittsburgh, United Stover the medium term inefficienThis move was closely mirroreNovember 2009. These commisubsidies distort markets, impeddeal with climate change.

During the Pittsburgh Summit, tsubsidies and suggestions for thewas presented to the G-20 Toimplementation strategies and ti

The IEA, OECD, and World Bank wthe November 2010 G-20 summextends the analysis presentedquantitative findings to include dfossil fuel subsidies. The report c

• motivations for introduci

• the case for reforming en

• estimates of energy subs

• modelling-based analysis

• recent action taken to ph

• a road map for phasing o

• lessons drawn from rece

1.1.2 Defining energy su

Finding a commonly agreed defcontext and countries have decidfor the purpose of this report, lowers the cost of energy produprice paid by energy consumers. reasons much narrower definitiobe quantified and for which data

The Worl

ossil-fuel subsidies in 2009

e report

rs took a key step towards reforming energy subates. Together, they committed to “rationalize ant fossil fuel subsidies that encourage wasteful ced by Asia-Pacific Economic Cooperation (APEtments were made in recognition that inefficie investment in clean energy sources and underm

he G-20 also requested a Joint Report on the sce implementation of their phase-out initiative. Thoronto Summit in June 2010, during which cometables were tabled.

were subsequently requested to prepare this secmit meeting to be held in Seoul, Republic of Kod in Toronto in June 2010, in particular by

data for the year 2009 and providing a road map fovers:

ing energy subsidies;

nergy subsidies;

sidies;

s of the implications of phasing-out energy subsid

hase out subsidies;

out subsidies, revisited through the poverty lens; a

nt experience of subsidy reforms.

ubsidies

finition of subsidies has proven a major challengded to adopt their own definition of energy subsidan energy subsidy is defined as any governme

uction, raises the revenues of energy producers Many energy subsidies are difficult to measure, s

ons are often adopted that include only those suba are readily available. The broad definition used i

orld Bank

Page | 5

bsidies at their and phase out consumption”. EC) leaders in ient fossil-fuel mine efforts to

cope of energy he Joint Report ountry-specific

cond report for rea. This work updating the

for phasing out

ies;

and

ge in the G-20 dies. However,

ent action that or lowers the

so for practical bsidies that can

n this report is

Page | 6

designed to capture all of the commonly exist.

Energy subsidies are frequently producers or consumers, or whenergy. Fossil-fuel consumption have been phased out by most many emerging and developing ethat seek to maintain or to exsubsidisation, particularly in advaalso been phased out, with thepolicies and liberalisation of interencouraging excessive productiresources and market distortionfossil fuel subsidies that lead to w

1.1.3 Mechanisms of go

Subsides can be further distingadministered; these include bu(Table 1). They can be groupedeployment of fledgling energy end-use prices.

Subsidies to energy consumptionintended to regulate the cost designed to provide consumers Government interventions suppdirect subsidization of domestconsumption through direct budfor the fuels or electricity thus stransfer is a fuel voucher, whichprice. In advanced market econofor low-income households, andthe cost of fuel purchases when p

Similarly, a wide array of tax expof excise-tax concessions on fuesurvey of practices in OECD couterms. For example, OECD estimyear to the agricultural sector fisheries sector. Finally, tax regimencourage the provision by emplof company-paid fuel for those v

Governments provide support tmarkets in such a way as to affecassuming part of their risk, by seand by undercharging for the usone transfer mechanism is inv

The World

diverse and non-transparent types of energy

differentiated according to whether they confeether they support traditional fossil fuels or clesubsidies lower prices to end-users. These typeeconomies in the developed world, but are sti

economies. Production subsidies, by contrast, invxpand domestic supply. They remain an impoanced economies, though many subsidies in this e shift towards more market-oriented economrnational trade. Both production and consumptioon or consumption, can lead to an inefficient

ns. Within this report a particular focus is givenwasteful consumption.

overnment support to energy

guished according to the channels through wudgetary payments, regulations, taxes and trad

d as either direct transfers, such as grants totechnologies, or indirect transfers, such as the

n are provided through several common channels:of energy to consumers, direct financial transwith rebates on purchases of energy products

porting energy consumption often involve the tic prices. However, many economies also sudgetary transfers that do not alter the observablesupported. In developing countries, a common f allows low-income recipients to purchase fuel a

omies, direct budgetary transfers include heatingd subsidies to help particular sectors, such as agrprices rise unexpectedly.

penditures often target consumers. These mostlyel designed to benefit particular users or areas. untries suggests that these could be quite impo

mates that fuel tax concessions are worth some in OECD countries, and at least $1.4 billion pe

mes in a number of advanced market economiesloyers of company-owned or leased vehicles for eehicles.

to energy production in a variety of ways: by ct costs or prices, by transferring funds to recipieelectively reducing the taxes they would otherwisse of government-supplied goods or assets. Ofteolved. For example, a government may fund

Bank

subsidies that

er a benefit to eaner forms of es of subsidies ill prevalent in

volve measures ortant form of

category have ic and energy n subsidies, by t allocation of n to inefficient

which they are e instruments

o expedite the e regulation of

: price controls sfers, schemes and tax relief. regulation or

upport energy e market price form of such a t a discounted -energy grants

riculture, meet

take the form A preliminary

ortant in value $8 billion per

er year to the s inadvertently mployees, and

intervening in nts directly, by se have to pay, en, more than research at a

national laboratory on how to coguarantees to companies investinproduction of such fuels, and exestate-owned lands. The nationalthe fuel than it could have paid fo

Table 1: Common types of energ

Direct budgetary transfers are although the complexity of the budget documents. In the case have traditionally benefited the c

Tax expenditures relating to thestem from favourable tax treatmhigher levels of production thancapital, special rules that allow bat which equipment becomes esubsidies. Immediate deduction

Description

Trade instruments Quotas. Technical reTariffs.

Regulations

Price controDemand gudeploymentMarket-accPreferentiaPreferentia

Tax breaks

Rebates or producer leTax credits allowances.Rebates, refenergy dutienergy in ge

Credit Low-interesloans to pro

Direct financial transfer Grants to pr

Risk transfer Limitation o

Energy-related services provided by government at less than full cost

Direct invesinfrastructuPublic resea

The Worl

onvert coal into a liquid transport fuel, provide gng in synthetic fuels from coal, provide a tax credempt such producers from paying royalties on co government may, in turn, pay the producer a hor an imported, petroleum-derived fuel.

y subsidies

the most straight-forward types of subsidiestask depends on how well they are reported iof European countries, the bulk of direct budge

coal industry.

e production of energy in industrialised countriment for capital or intermediate inputs. These c

n would otherwise be demanded by the market. businesses to deduct depreciation faster than theeconomically obsolete can in some cases imply (expensing) of exploration and development cost

Examples

estrictions. Tariffs on imports of crude petroleum products, makinfuel production more lucra

ols. arantees and mandated t rates. ess restrictions. l planning consent. l resource access.

Gasoline prices regulated ainternational market levelsRegulations that prioritise ucoal for power generation.

exemption on royalties, evies and income taxes. and accelerated depreciation . funds or exemptions on es and CO2 taxes or for eneral consumption taxes.

Favourable tax deductions investments in oil and gas fdeposits. Excise exemptions for fuel international air, rail, or wa

st or preferential rates on oducers.

Loan guarantees to financeinfrastructure.

roducers or consumers. Social payment programmeor earmarked for heat and consumption.

of financial liability. Insurance or indemnificatiofossil fuel producers at belo

stment in energy ure. arch and development.

Provision of seismic data foexploration. Government finance of actenvironmental health and smines.

orld Bank

Page | 7

rants and loan it linked to the

oal mined from igher price for

s to measure, n government etary transfers

es most often can encourage In the case of

e actual speed large indirect

ts is also a case

oil and ng domestic fossil tive.

at below s. use of domestic

for depletion or fields and coal

used in ater transport.

e energy

es conditioned on electricity

on provided to ow-market levels.

or oil and gas

ivities relating to safety in coal

Page | 8

in point although the issue is ctargeted at natural-resource rethere is no deduction for intere(expensing) of investment outlatherefore not necessarily imply sif all the existing expensing and atoo far. Meanwhile, other inputsindustries may be allowed to dedintermediate goods, such as raw

Governments also forego revenuresources) under their control. Tthan otherwise. The most direprovide access to domestic resouexploits for their own use or for a particular type of coal or a giveprovide access to intermediate access to government land for, e

Transfers of risk to governmentsthe case of industrialised countrloans but also security guaranteimportant are the transfers of enwhich often result in governmenamounts disbursed by governmewith abandoned coal mines.

Another area of government invdevelopment (R&D). In 2008, Irelated to fossil fuels amounteexpenditure is R&D related to eproduction; refining, transport acoal and gas conversion.

1.1.4 Motivations for in

The rationale for the introductipolitical, economic, social and emarkets operate. In practice, homeans of achieving their statedenergy subsidies include:

• Alleviating energy poveliving conditions of the accessible (Box 1). For ebiomass.

• Boosting domestic enerindigenous fuel producti

The World

complicated by the special nature of tax and ronts. For cash-flow based natural resource tax st expenditure, neutrality would require immed

ays. Provisions for expensing or accelerated depsubsidies, but for many countries it would be releaccelerated depreciation provisions are warranteds can also attract subsidies. For instance, workerduct part of their wage from their personal incommaterials may be acquired free of excise duty by

ue by offering the use of scarce resources (e.g. lanThis can reduce costs and thereby encourage moct cases relate to the conditions under whichurces of fossil fuels that a private company (or indsale. This sometimes takes the form of a royalty

en project of oil and gas extraction. But many govinputs, like water or electricity, at below mark.g. the construction of roads or buildings.

s are much less transparent and, as such, hard to ries. They include measures related to capital likes as in the case of government-funded oil stocknvironmental and health liabilities from producernts acting as insurers of last resort. An example

ents to compensate residents affected by subside

volvement in energy production is investment inEA data suggest that total government expended to almost $1.7 billion. Included under thienhanced oil and gas production; un-conventionnd storage of oil and gas; oil, gas and coal combu

troducing energy subsidies

ion of energy subsidies has often been to advaenvironmental objectives, or to address problemowever, they have rarely proven to be a success goal. The most common justifications for the i

erty: Consumption subsidies have been used topoor by making cleaner, more efficient, fuels a

example, liquefied petroleum gas (LPG) in place

rgy supply: Production subsidies have been uson in a bid to reduce import dependency. They h

Bank

oyalty regimes system where iate deduction reciation does

evant to review d or if some go rs in particular

me-tax base, or refiners.

nd or fossil-fuel ore production h governments dividuals) then exemption for

vernments also ket prices, and

gauge even in e concessional kpiling. Equally rs to the public would be the nce associated

n research and diture on R&D is category of nal oil and gas ustion; and oil,

ance particular ms in the way ful or efficient ntroduction of

o improve the affordable and e of traditional

ed to support have also been

used at times to supportthe overseas activities of

• Supporting industrial deusers are a source of coinvestment in energy-inotherwise not be profitaor trade restrictions, arperiods of economic dow

• Redistributing nationalconsumption subsidies thsharing the value of indencourage economic divof energy-intensive indus

• Protecting the environmhave introduced suppornuclear power and carincentives to move cleahelp to cost-effectively re

1.1.5 The case for reform

In recent years there has been isubsidies as many were seen resources and to be distorting m(Figure 1). Subsidies have been sprice volatility by blurring marundermine the competitivenessimporting countries, subsidies offor producers they quicken the dover the long term.

The Worl

t a country’s foreign and strategic economic policf national energy companies.

evelopment and employment: Energy subsidieompetitive advantage. They are sometimes usedtensive industries, such as aluminium smelting,ble. Further, production subsidies, usually in the e often used to maintain regional employment

wnturn or transition.

l resource wealth: In major energy-producihat artificially lower energy prices are often seen

digenous natural resources. They are also used iersification and employment by improving the costries, such as petrochemicals and aluminium.

ment: Developed countries and several emergirt programmes to aid the development of renebon capture and storage (CCS). In some casener technologies quickly towards market compeeduce greenhouse-gas emissions and pollution.

ming energy subsidies

increasing momentum to phase out certain typeto be resulting in an economically inefficient

markets, while often failing to meet their intendshown to encourage wasteful consumption, exaceket signals, incentivise fuel adulteration and s

s of renewables and more efficient energy tecften impose a significant fiscal burden on state bdepletion of resources and can thereby reduce e

orld Bank

Page | 9

cies by helping

s to industrial d to encourage , which would form of tariffs

t, especially in

ing countries, as a means of in an effort to ompetitiveness

ing economies ewable energy, es, transitional etitiveness can

es of fossil-fuel t allocation of ded objectives erbate energy-muggling, and hnologies. For budgets, while xport earnings

Page | 10

Figure 1: Potential unintended ef

Source: IEA World Energy Outlook 20

Among the key unintended efferecent years are that they:

• Create fiscal burden on unsupportable financial sell it domestically at lorates, spending on oil alevels well above those seized the opportunity pwithout having a major consumers from the upwithout provoking consu

• Encourage wasteful conthereby leading to fasterationalisation and efficisubsidies would provide proper price signals. Formay not have the choicedecide to build new, nopermanent feature of tencourage a motorist towhen her existing vehicle

• Exacerbate energy pricsubsidies exacerbate en

The World

ffects of fossil-fuel subsidies

010 (forthcoming)

ects of subsidies that have proved to be driver

state budgets: In some cases, high energy prices burdens on countries that import energy at wo

ower, regulated prices. As a share of GDP at maand gas imports in many economies spiked in 2

seen during the first and second oil shocks. Soresented by the fall in prices after mid-2008 to reimpact on inflation (since the fall in world pri

ward pressure on prices resulting from subsidyumer wrath.

nsumption: Subsidies can encourage wasteful r depletion of finite energy resources, and can aiency improvements in energy-intensive industrieconsumers with an incentive to conserve energy

r example, a power company burning oil to prode of switching to a less costly alternative overni

on-oil capacity if it expects higher input prices tthe market. Similarly, a rise in the price of go alter her driving habits and/or buy a more fue is traded or scrapped.

ce-volatility: The price controls that give risenergy price-volatility on global markets by damp

Bank

s of reform in

have imposed orld prices and arket exchange 2008, reaching ome countries

educe subsidies ices cushioned removal) and

consumption, also discourage es. Eliminating y by improving duce electricity ight, but could to persist as a gasoline might el-efficient car

to fossil-fuel pening normal

demand responses to csurprised by the robustcrude-oil prices, during tartificially low energy prof 131 countries carried2008 around two-thirdsinternational prices for gcost of diesel (Coady etprices from government and stronger demand government revenues fo

• Distort markets: Subsidi

market distortions by countries still retain substhe subsidy is directed aworkers who had lost theunlikely to alter demanmaintain production thahigh-cost local coal produsubsidies for oil and gacertain areas. Removingeffect of making domeswould, therefore, tend toand production would bprices would rise or fall curve. In practice, the efenergy prices and consuengage in similar policiesother reasons that suppfuel production subsidieleads to greater imports (Steenblik and Coroyanproducers, they can creaand discourage the uptaalso free up budgetary rLastly, lower energy pricDepending on the degrproducers could boost em

• Adverse impact on the eeffects. Subsidies that ebiomass to modern fupollution. Subsidies for lounit production costs tmitigation in the long tcounterproductive in reaprices dampen incentive

The Worl

changes in international prices. Many market ness of global oil demand, despite the dramatthe first half of 2008. This has now been attribuices in many countries, which blunted market sig

d out by the International Monetary Fund (IMF)s of countries failed to fully pass through thegasoline and half failed to pass through the full it al., 2010). Cutting subsidies, by shifting the bbudgets to individual consumers, would lead to response to future changes in energy prices r other urgent needs.

es for fossil-fuel production can hinder competitpropping up less efficient producers. For exasidies for hard coal mining. In some cases, a signiat covering the cost of closing down mines and eir jobs as a result of earlier rationalisation of thend and supply patterns. However, in other caat would otherwise be uneconomic, for examplucers to compete against imports. Similarly, couns production such as through reduced royalties

g production subsidies such as these would typistic production less competitive compared witho lower indigenous production. The extent to whe shifted to other parts of the world, and the exas a result, would depend on the shape of the

ffect of one country no longer subsidising fossil umption is likely to be small. However, when ms, world prices are likely to be higher than otherwort a close review of the efficiency and effectives. For example, if the removal of coal subsidies iof higher-quality coal, it would clearly benefit the

nnakis, 1995). Furthermore, by propping up te barriers to the introduction of cleaner technolo

ake of more efficient production practices. Their esources that could be better used elsewhere ines can result in energy being substituted for laboee of inter-factor substitution, removing energmployment and investment.

environment: Energy subsidies can have varying nable poor communities to switch from the trad

uels can minimise deforestation and reduce ow-carbon technologies can help to accelerate leao decline, and reducing the overall cost of cterm. However, the vast majority of fossil-fueaching local and global environmental goals. Subes for consumers to use energy more efficientl

orld Bank

Page | 11

analysts were ic increases in uted in part to gnals. A survey found that in sharp rise in increase in the

burden of high a much faster and free up

ion and create ample, several ficant share of compensating

e industry, so is ases, subsidies e, by enabling tries also offer s for leases in cally have the h imports and ich investment xtent to which

e global supply fuels on world

many countries wise. There are eness of fossil-in an economy e environment

less efficient ogies and fuels removal could the economy.

our and capital. gy subsidies to

environmental ditional use of household air arning, causing limate change l subsides are

bsidised energy ly, resulting in

Page | 12

higher consumption anFurthermore, fossil-fuel of renewable energy anattractive.

• Encourage fuel adulteratsubstitution of subsidisekerosene intended for has diesel fuel due to wideis created to sell subsiunsubsidised and, therefthe world, particularly isubsidising countries is countries experience lossales in the legitimate madulterate fuels and to sm

• Disproportionally benefioften intended to help rgoes to those who convehicles, electrical appliIndonesia, for example, of energy subsidies, whilbenefits (IEA, 2008a).

• Threaten investment in einvestment resources. Wprice controls, the effeccompanies – is to reducein, maintain and expaelectricity companies arecertain cases, for free) tthem financially weak, haand in maintaining and prevalent within the elesectors.

• Hasten the decline of exto phase out subsidies, ocost of the subsidies buttime, such subsidies marevenue streams, with imoil exporters, including Apetroleum products, pademand and undermineparticularly acute if refin

The World

nd greenhouse-gas emissions than would othsubsidies undermine the development and com

nd other technologies that could become more

tion: Energy subsidies can encourage fuel adulterd fuels for more expensive fuels. In some countrousehold cooking and lighting is diverted for unae price differentials. Smuggling can also arise, sincidised products in neighbouring countries whefore, higher. This has been an issue for years in in Southeast Asia, Africa and the Middle East. a substantial financial transfer to smugglers, w

sses from uncollected taxes and excise duties, dmarket. Removing subsidies would eliminate incemuggle them across borders.

it the middle class and the rich: Although energyredistribute income to the poor, the greatest bensume the most energy, i.e. who can afford tances, etc. The Co-ordinating Ministry of Econoreported that the top 40% of high-income familile the bottom 40% of low-income families reap o

energy infrastructure: Subsidies can have an adveWhere fossil-fuel consumption is subsidised throct – in the absence of offsetting compensatione energy companies’ revenues. This limits their and energy infrastructure. For example, many

e obliged to provide electricity at heavily subsidiseto certain sections of the community. This has marming their capacity to invest in building new geextending the network. Although this problem

ectricity sector, it also exists in the oil, natural

xports: Several energy-rich exporting countries haor expressed interest in doing so, concerned not ot also the resulting low efficiency in domestic enay even threaten to curtail the exports that emplications for global energy security. A numbeAngola, Iran, Kazakhstan and Nigeria, rely on impartly because low regulated prices preserve ae investment in adequate refining capacity. Thers are not reimbursed by governments for their

Bank

herwise occur. mmercialisation e economically

ration, and the ries, subsidised authorised use ce an incentive ere prices are many parts of The effect in

while recipient ue to reduced

entives both to

y subsidies are enefit typically to own motor omic Affairs of es absorb 70% nly 15% of the

erse impact on ugh consumer

n payments to ability to invest y state-owned ed rates (or, in made many of enerating plant

is particularly l gas and coal

ave also moved nly by the high ergy use. Over arn vital state r of significant orts of refined

artificially high his problem is losses.

The gains from phasing out fosenergy taxation reforms. Rising fossil fuel production and in sommay warrant review to assess ifalso found higher excise taxes ondifficulties controlling income trevenues that can help financingfuel subsidies are phased out acontribute to state building and a

Box 1: Subsidies and energy pov

The IEA’s World Energy Outlook (over 20% of the global populatiothe global population) rely on subsidies are one means of aaffordable and accessible for theand often ineffective means of dobut benefits are conditional upodisproportionately to middle and

Poor households may be unablehave no physical access to themnetwork or a connection to aparticularly difficult to target, gimarket. In comparison, the dismonitored and controlled. We esand electricity in countries witcountries with electrification rrepresented just 15% of consummost of these subsidies in any cathe total benefits from petroleu(Coady, et al., 2010).

Nonetheless, the removal of eimplemented, since low-income removal, as they spend a highesubsidies can bring considerable and more efficient fuels or enhasubsidies must be carefully desigincrease poverty. Providing finanessential to smoothing the patenergy-subsidy reform, the retransparent. In undertaking majoto which the economy and socphase-out of fossil-fuel subsidiestructural reforms are underw

The Worl

ssil-fuel subsidies can be enhanced if combinedworld market prices have greatly enhanced th

me cases, the royalty and tax regimes for natural f they are well balanced. Some middle-income cn energy consumption to be an attractive route tax evasion. Wider reforms can therefore raiseg the measures needed to ensure a social balancand, as emphasised by OECD work on Tax and administrative development.

verty

2010 highlights the alarming fact that today 1.4on) lack access to electricity and 2.7 billion peoplethe traditional use of biomass for cooking. Altlleviating energy poverty, by making energy

e poor, studies have repeatedly shown them to beoing so. The cost of these subsidies falls on the enon the purchase of subsidised goods and thus td higher-income groups.

e to afford even subsidised energy or related sem (for example, rural communities lacking a pun electricity grid). In general, subsidies for liqiven the ease with which such fuels can be sold

stribution of electricity and piped natural gas istimate that subsidies in the residential sector to th limited household access to modern energrates of under 90% or modern fuels access

mption subsidies in 2009. There is considerable ase go to richer households. The IMF has estimatem subsidies in 2009 accrued to the richest 40%

ven poorly targeted energy subsidies needs tohouseholds are likely to be disproportionately aff

er percentage of their household income on enebenefits to the poor when they encourage switchance access to electricity. Therefore, any movesgned so as not to restrict access to essential enencial support for economic restructuring or povertth for fossil-fuel subsidy reform. In most succeemaining support has been well-targeted, teor changes, assessments should be made regardiety can absorb the impacts of the reform. Fur

es should be considered as a package, particulaway or being contemplated. Pre-announcing a

orld Bank

Page | 13

d with broader e profits from resource rents countries have in a context of e considerable ce when fossil-Development,

4 billion people e (some 40% of though energy services more e an inefficient ntire economy, tend to accrue

ervices, or may ublic transport quid fuels are d on the black is more easily kerosene, LPG

gy (defined as s under 75%) evidence that

ed that 80% of of households

o be carefully fected by their ergy. Similarly, hing to cleaner s to phase-out rgy services or ty alleviation is essful cases of emporary and ding the extent rthermore, the arly if broader

strategy and

Page | 14

timeframe for phasing in subsidyreforms (UNEP, 2008).

1.2 Measuring fossi

Measuring both energy consumpthe varying definitions of what Although measuring consumptioenough information is published2009. Estimating subsidies to fosingle country there are typicproducer support. Many subsiditax concessions. And the data nequality or not reported.3

Developing a comprehensive asubsidies is hindered by data connational level, few countries havefuel industries, and even fewer Data are often reported only atgiven expenditure to the variocomplex when lacking sufficienexpenditures requires having a praises numerous issues having tcountries.

Despite these challenges, numegovernmental organizations havefuel production, resulting in a cle2010). The OECD Secretariat is estimates of support to fossil-fuworkshops on issues relating tsupport.4

Given the incomplete state of ireport only estimates of supportmeasurements are provided. Whthey have a particularly importaenergy security and the environm

3 The Global Subsidies Initiative (GSI)Development (IISD), has estimated that per year (GSI, 2010). 4 An Expert Workshop on Estimating Su19 November 2010.

The World

y reform can help households and businesses to a

l-fuel subsidies

ption and production subsidies is a complex undeconstitutes a subsidy and the availability of a

on subsidies requires an extensive array of energd to enable a reasonable estimate, as is done in ossil-fuel production is particularly challenging. cally several different sources, recipients and es are administered via indirect mechanisms, su

ecessary to estimate producer support are in many

nd internationally comparable set of estimatenstraints and methodological and conceptual issuee produced comprehensive estimates of support have included support provided by sub-nationa

t broad, programmatic levels, requiring analystsous fuels covered by the programmes. This cant details on subsidy recipients. Meanwhile, qproper benchmark against which to assess themto do with definitions and comparability of tax r

erous government agencies, academic researche recently turned their attention to subsidies benearer picture of their nature and scope (see, e.g.,

currently working with OECD Member countriuel production and consumption, and will be orgto the identification, estimation and reporting

nformation on other types of subsidies to fossit to fossil-fuel consumption that are revealed throhile representing only a subset of total subsidies ant impact on global energy trends affecting ecoment.

), a Geneva-based program of the International Instituteworldwide fossil-fuel production subsidies may be of the ord

upport to Fossil Fuels will be held at the OECD’s Headquarter

Bank

adjust to these

ertaking due to dequate data.

gy pricing data, this report for Even within a categories of

ch as complex y cases of poor

s of producer es. Even at the for their fossil-

al jurisdictions. s to allocate a n prove quite

quantifying tax m. This, in turn,

regimes across

hers, and non-nefitting fossil- Koplow et al., ies to compile ganizing expert g of fossil-fuel

il fuels, in this ough price-gap to fossil fuels,

nomic growth,

e for Sustainable der of $100 billion

rs, in Paris, on 18-

1.2.1 The price-gap app

This report provides estimates This approach compares final cofull cost of supply or, where appof transportation and distributioend-users and subsidies to fosapproach may be conceptually, sectors and computing reference

The price-gap approach is the subsidies.5 It is designed to captuthose that would prevail in a comgap approach do not capture allbe understated as a basis for atrade. For example, the methodunder-collection of energy bills (prife. Despite these limitations, thand for undertaking comparativdevelopment (Koplow, 2009).

For countries that import a givapproach are explicit. That is, theof imported energy (purchased acontrast, for countries that exposubsidy estimates are implicit represent the opportunity cost ocould be recovered if consumersconsumption themselves and imprepresent a combination of oppo

1.2.2 Reference prices

For net importing countries, refeprice: the price of a product at tplus the cost of freight and indistribution and marketing and aare not included in the referencpump price in a given country is be no net subsidy if an excise dexporting countries, reference pproduct at the nearest internatfreight and insurance back to thmarketing and any VAT.

For oil products, average distribuin the United States. The assu 5 Kosmo (1987), Larsen and Shah (19approach.

The Worl

roach

of energy-consumption subsidies using a price-onsumer prices with reference prices, which corrpropriate, the international market price, adjusteon. The estimates cover subsidies to fossil fuelsssil-fuel inputs to electric power generation. Scompiling the necessary price data across diffe

e prices are formidable tasks.

most commonly applied method for quantifyure the net effect of all subsidies that reduce fina

mpetitive market. However, estimates produced u types of intervention known to exist. They, ther

assessing the impact of subsidies on economic d does not take account of revenue losses in coparticularly for electricity) is prevalent, or where

he price-gap approach is a valuable tool for estimve analysis of subsidy levels across countries to

en product, subsidy estimates derived through ey represent net expenditures resulting from theat world prices in hard currency), at lower, regul

ort a given product — and therefore do not pay wand usually have no direct budgetary impact.

of pricing domestic energy below market levels, i.e paid world prices. For countries that produce a pport the remainder (such as Iran), the estimates p

ortunity costs and direct government expenditures

erence prices have been calculated based on thethe nearest international hub, adjusted for quali

nsurance to the importing country, plus the coany value-added tax (VAT). Other taxes, includingce price. Therefore, in the case of gasoline, even

set by the government below the reference priceduty large enough to make up the difference is lprices were based on the export parity price: tional hub adjusted for quality differences, minhe exporting country, plus the cost of internal di

ution and marketing costs for all countries were bumed costs for shipping refined products, by

992) and Coady et al., (2010) among others, for example

orld Bank

Page | 15

gap approach. respond to the d for the costs consumed by Simple as the rent fuels and

ying consumer al prices below using the price-refore, tend to efficiency and

ountries where energy theft is

mating subsides support policy

the price-gap e domestic sale lated prices. In world prices — . Rather, they e. the rent that portion of their presented here s.

e import parity ty differences,

ost of internal g excise duties, n if the pre-tax e, there would levied. For net the price of a us the cost of istribution and

based on costs contrast, vary

e, have used this

Page | 16

according to the distance of the ccosts as reported in industry distribution costs have been estbeen carried out using local pricexchange rates.

Reference prices have been adjufuel. For example, for countries timport small volumes of higher below observed import prices.

Unlike oil, gas and coal, electricitreliable international benchmarkannual average-cost pricing for efrom each generating option). Infor the cost of production, transmfor building new capacity, were ifuels and annual average fuel eff$40/MWh was added to accouresidential uses, respectively. To at the levelised cost of a combine

Some experts suggest that the abparticular, some are of the opinishould be based on their cost oapplied within this analysis. Thetheir natural resources in a way tand that this approach more thinternally at a price below theapproach results in an economigrowth in the longer term.

Cross-subsidies between sectorsas to offset lower prices for otheFor example, in many countries cost so as to finance lower pricesituation can also be found in oable to negotiate special low elean average variance in prices, iregion that are often vitally immarkets. Similarly, it does not purchases, such as the discountefuel rebate schemes.

Box 2: Sample calculation – estim

The first step is to calculate the gasoline in 2009 and therefore product at the border. Taking theUnited States, the fob price is

The World

country from its nearest hub and have been takendata. For natural gas and coal, transportation

timated based on available shipping data. All calces and the results have been converted to dol

usted for quality differences, which affect the mathat rely heavily on relatively low-quality domestiquality coal, such as India and China, reference

ty is not extensively traded over national borders,k price. Therefore, electricity reference prices welectricity in each country (weighted according ton other words, electricity reference prices were mission and distribution, but no other costs, suchncluded. They were determined using reference pficiencies for power generation. An allowance of

unt for transmission and distribution costs for avoid over-estimation, electricity reference price

ed-cycle gas turbine (CCGT) plant.

bove method of determining reference prices hasion that the reference price in countries that areof production, rather than prices on internation

e basis for this view typically is that these countthat effectively promotes their general economichan offsets the notional loss of value by sellinge international price. The counter-argument is cally inefficient allocation of resources and redu

, i.e. where some consumers are charged a priceer consumers, have not been taken into account i

commercial and industrial consumers often pays for the agriculture and residential sectors, whil

other countries (for example, where aluminium ctricity rates). Furthermore, as the price-gap mett does not capture the variability in prices by t

mportant in giving new technologies entry pointpick up direct subsidies to consumers that ar

ed fuel coupons used by some developing countr

mating gasoline subsidies in Venezuela in 2009

appropriate reference price. Venezuela was a nwe start with the free-on-board (fob) price, or e average spot price of gasoline in 2009 at the nes calculated by subtracting the average cost o

Bank

n from average n and internal culations have lars at market

rket value of a ic coal but also prices are set

, so there is no were based on o output levels set to account

h as allowances prices for fossil $15/MWh and industrial and

es were capped

s limitations. In e net exporters nal markets as tries are using development,

g the resource that such an

uces economic

e above cost so n this analysis. a price above e the opposite producers are

thod measures time-of-day or ts into energy

re tied to fuel ries or heating-

net exporter of the price of a

earest hub, the of freight and

insurance to transport gasoline b0.89 bolívares fuertes (VEF) ($0.4fob price is VEF 0.87 per litre. Toprice consumers would see at thany VAT. Assuming distribution ($0.08) per litre, the final referenVAT is applied to gasoline sales in

As average end-use prices for gaprice gap then amounts to VEFgasoline, we take the price gap mlitres), arriving at a gasoline subs

1.3 Estimate of glob

The value of fossil-fuel consumptfossil fuels) is estimated to amoInternational Energy Agency andIEA’s finding is based on an extefuel consumption, as identified economies were identified, esticonsumption. Remaining subsidienergy consumption and pricesenergy is identified as being sHowever, production subsidies implication, the figures for contribution of OECD countries to

In absolute terms, the biggest endowments. For a given fuexpenditures by pricing domesticlong as prices are set above thehighest of any economy), with mhighlighting that estimates for cewhen viewed on a per-capita baexpressed as a proportion of theconomy. The $312 billion estimaand to fossil-fuel inputs to electrthe most heavily subsidised fuerespectively. Subsidies to electri2009. At only $6 billion, coal subs

For the economies surveyed he22%, meaning consumers paid rowas the most highly subsidised fnatural gas are comparatively himarkets, even as the global mawere subsidised at an average ra

The Worl

between Venezuela and the United States. Given 41) per litre and a shipping cost of VEF 0.02 ($0.01o complete the calculation of reference prices anheir local pump, retail and distribution cost are ad

and retail costs equal to those in the United Stnce price for gasoline in 2009 was VEF 1.04 ($0.48n Venezuela.

asoline in 2009 were reported as VEF 0.06 ($0.03F 0.98 per litre. To estimate the total value of tmultiplied by total final consumption (estimated idy of approximately VEF 15.6 billion ($7.3 billion)

bal fossil fuel consumption subs

tion subsidies (including subsidies to electricity gount to $312 billion in 2009. These estimates ared do not represent the official positions of G-20 ensive survey to identify those economies that susing the price-gap method outlined above. In

imated to represent over 95% of global subsidised consumption occurs in economies where res are unavailable. The vast majority of the ecoold below a world reference price were outsiare prevalent in both OECD and non-OECD e

consumption subsidies may under-represent o the total of production and consumption subsid

subsidies are in those economies with the laruel, net-exporting economies do not incur c energy products below their value in internatione cost of production. Iran’s subsidies reached $

most of this sum going to oil products and natural ertain economies may appear high in dollar termsasis or as a percentage of GDP. Fossil-fuel subsihe full cost of supply, vary considerably by fuelate comprises subsidies to fossil fuels used in fina

ric power generation. In 2009, oil products and naels, attracting subsidies totalling $126 billion anicity consumption were also significant, reachingsidies were comparatively small.

re, fossil fuels were subsidised at a weighted-aoughly 78% of competitive market reference pricefuel, at an average rate of 51% in 2009. Subsidisigh since many supplies are still priced within lim

arket for liquefied natural gas continues to growte of 19%, electricity at 18% and coal at 7%.

orld Bank

Page | 17

a spot price of 1) per litre, the d arrive at the

dded as well as ates, VEF 0.17 8) per litre. No

3) per litre, the the subsidy to at 15.9 billion

).

idies

enerated from e made by the countries. The ubsidise fossil-total, 37 such

ised fossil-fuel eliable data on nomies where de the OECD.

economies. By the relative

dies.

rgest resource hard-currency

nal markets, as 66 billion (the gas. It is worth s, but less high disation rates, as well as by

al consumption atural gas were nd $85 billion, g $95 billion in

verage rate of es. Natural gas ation rates for

mited domestic w. Oil products

Page | 18

The magnitude of energy subsiddomestic pricing policy, exchangprices typically have by far the gfossil-fuel prices surged in internenergy consumption subsidies wwhen the total was $343 billion. in the value of subsidies betweenbe attributed to deliberate intervgap) in order to reduce the burde

Some economies manage priceproducts. Although the intent mlevels, rising international eneconsumers (an effect picked-up bthe situation can lead to unexpethat the subsidies arising from Mshown that governments often prices are increasing and not toprice falls. During the rapid run-uautomatic price adjustments in ofor being slow to adjust downwa

Figure 2: Economic value of fossi

Source: IEA World Energy Outlook 20

Note: Subsidy estimates are made bposition of G20 countries.

0

50

100

150

200

250

300

2007

Bill

ion

dolla

rs

The World

des fluctuates from year-to-year with changes inge rates and demand. Of these factors, movemgreatest impact on variations in subsidy levels. national markets during the first half of the yeawas estimated at $558 billion, a dramatic increaDeclining world prices were the main reason for tn 2008 and 2009. However, some of the observedventions to raise consumer prices (thereby, shrinen on government finances.

e volatility by regulating domestic prices for cmay not be to hold average prices over a period ergy prices can inadvertently lead to marketby the price-gap approach). Conversely, when wocted revenues. For example, the fall in oil prices

Mexico’s fuel-excise mechanism all but vanished. Efind it hard to increase domestic prices when

o immediately pass through the full extent of anup in world oil prices in early 2008, many economorder to shield consumers, but they subsequently rd after prices fell sharply later in the year.

l-fuel consumption subsidies by type

010 (forthcoming).

by the International Energy Agency and do not repre

2008 2009

Bank

n world prices, ments in world

In 2008, when r, the value of se from 2007, the sharp drop d drop can also king the price-

certain energy below market

t transfers to orld prices fall, in 2009 meant Experience has n international ny subsequent ies abandoned faced criticism

esent the official

Coal

Oil

Gas

Electricity

1.4 Implications ofsubsidies

1.4.1 Method and assum

This section quantifies the enerconsumption subsidies and thebaseline case in which subsidy rain 2007-2009. Because subsidies reasonable basis for estimatinmagnitude of subsidies may risepremise that subsidies to consumto higher levels of consumptioreference, prices are calculated u

To illustrate the magnitude of thegradual phase-out of all subsidi2020.6 A growing number ofimplemented, would eliminate oimportant to emphasise that socneed to be a central consideratio

Box 3: The IEA energy-subsidy on

As highlighted by the G-20, increan essential step in building moto data on fossil-fuel subsidies wencourage informed debate onallocation of resources or whetalternative means. Transparencyand provide a useful baseline fr(Hale, 2008; Laan, 2010).

As a contribution to the processestablishing an online database tbreakdowns by economy, by fuesystematic analysis of energy suEnergy Outlook series since 19progress being made by economdatabase on an independent basbasis of the IEA’s own survey economies concerned. www.worldenergyoutlook.org/suextensive survey of end-use price 6 Although the analysis assumes tbetween 2011 to 2020, the commitmedium term inefficient fossil fuel s

The Worl

f phasing out fossil-fuel cons

mptions

rgy savings that would result from the phase-oue implications for CO2 emissions. The compariates from 2010 remain unchanged relative to thei

tend to fluctuate as a result of market volatility, g the impact of the subsidy phase-out, evee or fall sharply in a given year. The analysis is

mers lower the end-user prices of energy productson than would occur in their absence. The unusing the price-gap analysis described above.

e gains possible by eliminating subsidies, the anales to fossil-fuel consumption, globally, over the

f economies have already announced plans or reduce their subsidies well before 2020 (see Secial and equity impacts resulting from energy su

on in the design of any phase-out programme (see

nline database

asing the availability and transparency of energy mentum for global fossil-fuel subsidy reform. Imwill raise awareness about their magnitude and n whether the subsidy represents an economther it would be possible to achieve the samey of subsidy data can also encourage consistenrom which progress to phase out subsides can

s of increasing transparency of energy-subsidy dto allow public access to data on fossil-fuel subsil and by year. This new database represents an ex

ubsidies that the IEA has been undertaking throu999. It will be updated annually as a means oies to phase-out fossil fuel subsidies. The IEA is cois, not at the request of the G20. It has been consand the energy-subsidy data has not been aThe database will be ava

ubsidy.asp. The database has been constructede data. A key source of data was the IEA’s quarter

he complete phase out of consumption subsidies inment among G-20 countries is to "rationalize and phaubsidies."

orld Bank

Page | 19

sumption

ut of fossil-fuel son is with a r average level this provides a n though the based on the

s and thus lead nsubsidised, or

ysis assumes a e period 2011-

that, if fully ection 1.5). It is ubsidy removal e Section 2.2).

subsidy data is mproved access

incidence and ically efficient objectives by

t presentation be monitored

data, the IEA is idies, including xtension of the ugh the World

of tracking the onstructing the structed on the agreed by the ailable at d following an rly publication,

n all economies ase out over the

Page | 20

Energy Prices and Taxes. Other sources include official statistics, international and national energy companies, consulting firms and investment banks’ research reports. The IEA’s network of energy and country experts and their local energy contacts have also contributed substantially to the identification and verification of end-user prices. Additional data were extracted from databases, reports and personal communications with various organisations, including the Asian Development Bank, IMF, Latin American Energy Organization and the European Bank for Reconstruction and Development.

1.4.2 Energy demand

Compared with a baseline case in which subsidy rates remain unchanged, the complete phase-out of consumption-related fossil-fuel subsidies between 2011 and 2020 would cut global primary energy demand by 5%, or 738 Mtoe, by 2020 (Figure 3).7 This reduction is equivalent to the current energy consumption of Japan, Korea, and New Zealand combined. Furthermore, reductions in energy demand (relative to the baseline) would continue to be realised after 2020 as consumers continue to change their behaviour over time.

Where consumption is subsidised, eliminating energy subsidies would reduce dependence on imports and lead to an immediate improvement in the fiscal position of many governments. Moreover, exposing consumers to market-driven price signals would strengthen and accelerate the demand response, which in turn would contribute to reducing volatility in global markets. The phase-out of energy subsidies would have several other positive effects on long-term energy security by encouraging diversification of the energy mix and slowing down the depletion of finite fossil-fuel resources.

Figure 3: Impact of fossil-fuel consumption subsidy phase-out on global primary energy demand

Source: IEA World Energy Outlook 2010 (forthcoming)

7 Although the analysis assumes the complete phase out of consumption subsidies in all economies between 2011 to 2020, the commitment among G-20 countries is to "rationalize and phase out over the medium term inefficient fossil fuel subsidies."

8 000

10 000

12 000

14 000

16 000

1990 1995 2000 2005 2010 2015 2020

Mto

e No subsidy removal

Subsidy removal 2011-2020

1.4.3 CO2 emissions

The phase-out of fossil-fuel consrelated CO2 emissions by 5.8% bremain unchanged (Figure 4). Tequivalent to the current combItaly. Reduced demand growth fomatter and other air pollutants.

Our analysis illustrates the impofuel subsidies in addressing climcommitments under the Copenhchance of limiting the global temgases in the atmosphere is limite(ppm CO2-eq). Based on IEA esti45% of the additional yearly inrequired to meet the 2°C goal. phase-out programme would beremoval, such as creating a compolicy objectives, including the re

Figure 4: Impact of fossil-fuel cemissions

Source: IEA World Energy Outlook 20

Box 4: Subsidies for low-carbon

Policy support for low-carbon edrivers underpin this trend: firstand, second, a desire to diversespecially in 2005-2008). Job especially as a contribution to red

0

5

10

15

20

25

30

35

40

1990 2005

Gt

The Worl

umption subsidies over 2011-2020 would reduce by 2020 compared with a baseline case in whichThis amounts to savings of 2 gigatonnes (Gt) ofbined emissions of Germany, France, the Unitedor fossil fuels would also lead to lower emissions

ortance of the G-20 commitment to phase out inemate change and the role it could play in impagen Accord. According to climate experts there imperature increase to 2°C if the concentration ed to around 450 parts per million of carbon-dioximates, fossil-fuel consumption subsidies in 2009nvestment in low-carbon technologies and eneHowever, a portion of the funds liberated thro

e need to be directed towards the costs involvedmprehensive social welfare net, in order to ensueduction of energy poverty, are also achieved.

consumption subsidy phase-out on global energ

010 (forthcoming)

energy sources

energy has increased considerably over the pastt, the effort to constrain growth in greenhousesify the supply mix (prompted particularly by hcreation has been another factor in governmducing unemployment following the economic cri

2008 2015 2020

No sub

Subsidy2011-2

orld Bank

Page | 21

global energy-h subsidy rates f CO2 by 2020, Kingdom and

s of particulate

efficient fossil-lementing the is a reasonable of greenhouse

xide equivalent 9 amounted to ergy efficiency ough a subsidy d with subsidy ure that other

gy-related CO2

t decade. Two -gas emissions

high oil prices, ment support, isis.

bsidy removal

y removal 020

Page | 22

In the context of this growing policy support, some forms of low-carbon generation have grown significantly during the last decade. Renewables-based electricity output increased by nearly a third from 2000 to 2008; wind power expanded seven-fold and photovoltaic generation grew 16-fold during the same period. At the same time, global consumption of biofuels quadrupled. Along with this growth in deployment, renewable technologies have experienced a fall in costs. Cost reductions are essential to large scale development of renewable energy. Most renewable energy technologies are capital-intensive, requiring significant upfront investments, and most cannot currently compete on price with conventional sources.

Government support for emerging low-carbon technologies can lead to design improvements and the widespread deployment that is necessary to make them cost-competitive. The scope for further cost reductions for these emerging technologies is generally greater than for the more mature fossil fuel technologies. By contrast, fossil fuel prices are expected to increase in the future. Subsidies for low-carbon energy can take the form of consumption or production subsidies. A wide variety of mechanisms can be used to deliver the support, including portfolio standards, green certificates, feed-in-tariffs, premiums, and production, consumption and investment tax incentives.

The IEA estimates that worldwide government support to renewable electricity and biofuels amounted to $57 billion in 2009 — up from $44 billion in 2008 and $41 billion in 2007. These estimates do not include subsidies for renewable heat technologies or other emerging low-carbon energy technologies such as CCS.

While subsidies for renewable energy can yield benefits, they can also be ineffective or inefficient if not well-designed. Good policy design for renewable energy subsidies involves paying close attention to non-market barriers, ensuring that support is predictable and transparent in order to attract investors, reflecting improvements in technology over time by reducing subsidies in line with declining costs, matching support to the needs of individual technologies at differing stages of development, and considering the wider effects of new technologies on the energy system as a whole (IEA 2008b).

In addition to providing support as defined above, governments are engaged in the substantial continuing effort in research and development (R&D) to bring down the costs of renewable energy technologies and improve their performance. Some renewable technologies are mature or almost mature and do not require significant additional R&D, while others depend on further supportive R&D policy measures for their widespread diffusion. Total spending on R&D for renewable electricity technologies and biofuels reached $5.6 billion in 2009, with 45% of this amount provided by governments.

See Annex 4 for more discussion on mechanisms to support low-carbon energy technologies.

1.5 Recent action taken and plans to phase out subsidies

Following the commitment made by the G-20 countries to “rationalize and phase out over the medium term inefficient fossil fuel subsidies that encourage wasteful consumption”, each G-20

member has submitted implemeIn addition to the implementatPittsburgh agreement many ecoimplemented or proposed reformthat would prevail in an undisproducers (Table 2).

These efforts contributed to a smconsumption subsidies in 2009 also had a more noticeable impathe moves have varied from econlasting reform will take hold, inovercome.

In October, 2010, Angola madesubsidies as part of its plans to acountry’s oil sector. The first cudiesel prices by 38%. Angola is cproduct demand, but is seeking tself sufficient in refined products

In Canada, the oil and gas seprovisions. Changes at the fedethese preferences. In the 2003 ftax changes to be phased in ovresource allowance with a deducmeasures were intended to imremission order which had allowrespect of the Syncrude oil sanbudget, Canada announced that phased out over the 2011-2015 minerals. Draft regulations to imp

In less than a generation, Chinbecome the world’s fastest-growglobal energy markets. China hacloser to global market levels anhave contributed to the significafor crude oil produced in China agrades of oil sold in internatiogenerally match the internationcoal and began to introduce a mare now largely set by direct negnatural gas market, prices remaiMay 2010, the government annofollowing an increase in gas transmore efficiently, and should adomestic exploration and produ

8 Details of these subsidy re/Annexes_of_Report_to_Leaders_G20_I

The Worl

ntation strategies and timetables to implement thtion strategies planned by G-20 members in renomies both within and outside the G-20 have i

ms to bring their domestic energy prices into line storted market or to rationalise support given

mall but meaningful reduction in the IEA estimatrelative to 2008. Preliminary data also suggest tact on estimated subsidy levels in 2010. The key nomy to economy, as have expectations over the view of the political and social barriers that fir

e the first of a series of planned cuts to gasolattract foreign investment into the downstream suts led to immediate increases in gasoline pricecurrently dependent on imports for around 70%to build new refining capacity to enable the couns.

ctor has traditionally benefited from certain fral level, however, have moved toward the gradederal budget, the government introduced a numver a five-year period, including the replacemection for actual provincial royalties and mining tax

mprove the neutrality of the resource tax systewed deduction of both royalties and the resourcnds project expired at the end of 2003. In the

the accelerated depreciation allowance for oil saperiod, although it still exists for mines extractingplement the phase-out were released on May 3, 2

na, which was a largely self-sufficient energy cwing energy consumer (and importer) and a ms made significant progress in bringing domestic

nd is continuing to push ahead with new reformsant reduction in energy intensity experienced sincare already determined on the basis of the price fonal markets. Prices for many refined oil prodal levels. In 2007, China lifted its remaining pric

market-based pricing system. Coal prices for powgotiations between coal producers and power comn relatively low compared to those on internation

ounced a 25% increase in onshore natural gas bensmission fees. The increase should induce consumaccelerate investment by China’s national oil uction and development of LNG and long-distanc

forms are available at http://www.g20.org/DocumInefficient_Fossil_Fuel_Subsidies.pdf

orld Bank

Page | 23

his phase-out.8 sponse to the n recent years with the levels

n to fossil-fuel

tes for energy-that they have drivers behind likelihood that

rst need to be

ine and diesel segment of the es by 50% and % of its refined ntry to become

favourable tax ual removal of

mber of income nt of the 25% xes paid. These em. A special e allowance in

e 2007 federal ands would be g conventional 2010.

consumer, has major player in c energy prices . These efforts

ce 1980. Prices or comparable ucts also now ce controls for wer generation mpanies. In the nal markets. In nchmark prices mers to use gas

companies in ce pipeline gas

ments2010/expert-

Page | 24

import projects. After launching a programme to eliminate the preferential tariff arrangements for certain energy-intensive industries and increase the electricity prices for non-residential users in 2009, the government has recently released a proposal to introduce a tiered electricity pricing mechanism for residents (under which prices would increase with consumption).

Earnings from energy taxes in India, which go predominantly to the state governments, far outweigh the cost of subsidies, which is borne by the central government. Nonetheless, the country is in the process of energy price and tax reform (Government of India, 2010). In June 2010, the federal government announced that gasoline prices would henceforth be market-driven and the intention to later apply market-driven pricing for diesel. It also announced immediate price increases for diesel, LPG and kerosene. Natural gas pricing reform was also implemented in mid-2010, allowing state-run Oil & Natural Gas Corp. (ONGC) and Oil India Ltd. (OIL) to sell gas from new fields at market rates instead of regulated prices. Furthermore, the price of natural gas more than doubled under the regulated price regime in 2010. Reforms in India’s steam coal industry are expected to slowly bring domestic coal prices in line with import parity levels, with due allowance for quality differences. In June 2010, state-owned Coal India Ltd, which is responsible for almost 90% of the country’s coal production, announced that it would move to price its premium grades on an import parity basis. As more than 80% of India’s electricity is generated from coal, the implementation of the coal pricing reforms can be expected to impact power prices.

Indonesia has a long history of directly subsidising energy as a means of supporting the incomes of poor households. The size of energy subsidies has fluctuated widely over the past decade, following movements in international prices and the exchange rate and adjustments to the subsidy schemes. Previously, subsidies were available for industry and all segments of the population, but coverage has become increasingly targeted and the number of subsidised fuels has declined. In 2010, Indonesia announced plans to eliminate energy subsidies by 2014. The gap between international and domestic prices is to be progressively reduced, in an effort to minimise the impact on the poor. According to Indonesia’s 2011 state budget, 11% of government expenditure in 2011 will be devoted to energy-consumption subsidies, compared with 13% in 2010 and 19% in 2008. Indonesia has an ongoing programme to phase out the use of kerosene in favour of LPG. The energy ministry is considering a new plan to restrict the use of subsidised fuel to motorcycle, public transportation vehicles and cars purchased before 2005. In June 2010, the Indonesian government raised power tariffs by an average of 10%. This will reduce the overall burden of electricity subsidies on the state budget and boost revenues for Indonesia’s state power company.

With vast reserves, Iran is one of the world’s largest oil and natural gas producers. Oil and gas activities play a central role in supporting Iran’s economy, generating about 80% of its export revenues in 2008. Heavily-subsidised energy consumption has left a legacy of inefficient energy use, environmental degradation, inadequate investment and fuel import dependence. In early 2010, a law outlining far-reaching subsidy reform was enacted in Iran. The subsidy reform law calls for gradual implementation (over 2010-2015) of market-based energy pricing and the replacement of subsidies by targeted assistance to lower income groups. Among the key objectives of the law are to increase the prices of oil derivatives to 90% of the Persian Gulf export price, the price of household gas tariffs to 75% of the Persian Gulf export price, and the price of electricity to a level that reflects the full cost of production. To compensate for higher prices and the impact on low-income groups, 50% of the fiscal benefit resulting from increased prices would be redistributed to low-income consumers via direct cash and non-cash payments.

Despite Mexico being the worldhave represented a serious ecoincreasing reliance on refined pliquefied petroleum gas were eqperiod 2005 to 2009 (Nationaarrangements for refined produestimated that gasoline and diesby late 2012. As part of the prgovernment – have been increaspoor customers with better targecash transfer connected to Oporcover their energy needs. This is and avoids creating incentives fo

Substantial progress has been electricity pricing, especially in tterms) have been increased cons2007, the government adopted tand export markets by 2011. Tfollowing the surge in oil prices dmarket restructuring began in 20scheduled for 2011. A process ofsector is due to commence in 201

In South Africa, subsidised electinability of utilities to enforce prelectricity capacity. Rolling blackand plans to further raise tariffSouth Africa (NERSA) granted Eapproximately 25% per year oveBasic Electricity programme, whiamount of free electricity for ess

The United Arab Emirates comm2010 which are aimed at bringinthe losses state-run fuel retailersby 26%. There have been reportsneighbouring Oman (where pricecross the border specifically to pin the United Arab Emirates international market levels.

The Worl

d’s seventh-largest crude-oil producer, subsidisedonomic strain on the government budget and cproduct imports. Subsidies for electricity, gasoliquivalent to more than one and a half per cent ofl Energy Strategy). Mexico is currently reformucts and, contingent on international market coel subsidies could be eliminated by late 2010 and

rocess, retail prices for gasoline, diesel and LPGsing on a monthly basis since December 2009 weted subsidies. Starting in 2008 the government irtunidades that is intended to help very low-incompreferable to price subsidies, as it is better targetr environmentally harmful increases in energy con

made in Russia to introduce more market-bthe industrial sector. Gas tariffs for Russian indusistently since 2000, by approximately 15% to 25%the goal of achieving equal profitability from saleThe target date for full parity was extended

during 2008 and the subsequent economic downt006, and full liberalisation of the wholesale markf scaling back retail electricity price subsidies for 11.

tricity pricing, coupled with non-payment by custoperty rights, has led to a lack of investment and

kouts have provided strong impetus for recent pfs in coming years. In 2010, the National EnergyEskom, the state utility, permission to raise aveer 2010-2013. Through cross-subsidies, it will maich provides targeted subsidies to the poor throuential services.

menced a series of planned increases to gasoline ng prices in line with international market levels s had been incurring on gasoline sales. To date, prs, however, that the moves have pushed up saleses are now much lower) by United Arab Emirates purchase fuel. Unlike gasoline, prices for most of have already been deregulated and fluctuate

orld Bank

Page | 25

d energy prices contributed to ne, diesel and f GDP over the

ming its excise onditions, it is d those for LPG G – set by the hile protecting mplemented a

me households ted at the poor nsumption.

based gas and stry (in rouble

% each year. In es to domestic to 2014-2015,

turn. Electricity ket is currently the residential

tomers and an d a shortage of price increases y Regulator of erage rates by aintain its Free gh a minimum

prices in April and stemming ices have risen s of gasoline in nationals who the diesel sold

e in line with

Page | 26

Table 2: Selected plans for subsidy phase-out

Country Description of announced plans

Angola Cut gasoline and diesel subsidies in September, 2010, leading to a price increase of 50% and 38% for gasoline and diesel respectively.

Argentina Proposes to reduce household subsidy for propane gas as natural gas access is expanded.

China Oil product prices were indexed to a weighted basket of international crude prices in 2008. Natural gas prices increased by 25% in May 2010. China has already removed preferential power tariffs for energy-intensive industries.

Egypt Plans to eliminate energy subsidies to all industries by the end of 2011.

India Abolished gasoline price regulation in June 2010 and plans to do the same for diesel. The price of natural gas paid to producers under the regulated price regime was increased by 230% in May 2010. State-owned Coal India Ltd. announced that it would benchmark its premium grade coal to world prices.

Indonesia Plans to reduce spending on energy subsidies by 40% by 2013 and fully eliminate fuel subsidies by 2014. Electricity tariffs were raised by 10% in July 2010. Has an ongoing programme to phase out the use of kerosene in favour of LPG.

Iran

Plans to replace subsidised energy pricing with targeted assistance to low-income groups over the period 2010-2015. Reforms call for the prices of oil products, natural gas and electricity to rise to market-based levels.

Malaysia In July 2010, announced reductions in subsidies for petrol, diesel and LPG as the first step in a gradual subsidy-reform programme.

Mexico Subsidies to gasoline and diesel are expected to disappear by the end of 2010, and the gap of LPG prices is expected to close in 2012.

Nigeria Plans to remove subsidies on petroleum products by December 2010, or latest end of 2011.

Pakistan Plans to phase out electricity subsidies and has implemented a tariff increase of around 20% in 2010.

Russia Natural gas prices for industrial users are to continue increasing toward international levels through 2014 based on the balancing of revenues from domestic and export sales. Pricing in the wholesale electricity market is scheduled to be fully liberalised in 2011.

South Africa Plans to increase electricity tariffs by approximately 25% per year over 2010-2013.

UAE Commenced reducing gasoline subsidies in April 2010 and plans to bring them in line with international market levels. Diesel prices are already largely deregulated.

Ukraine Raised gas price for households and electricity generation plants by 50% in August 2010 and plans to raise them by another 50% from April 2011.

Sources: IEA World Energy Outlook 2010 (forthcoming)

2. A roadmap for The World Bank’s contribution follows. A simple analytical framproposed in the first Joint Repoderives the lessons drawn from savings from the reduction of projects and cash transfers. Othprice adjustments and oil funds asubsidy removal on the poor (Sec

2.1. Analytical Frame

A necessary first step in implemsubsidies that should be phasconsumption. Identifying which wasteful consumption” from amindividual countries requires undthe impact of the subsidy on con

A simple decision tree could bselected subsidies over the medwe will tailor the decision treeredesign or remove an energy suthe following tests can be usedsubsidy at each phase of the deci

The first two phases consider thidentify those inefficient subsiefficiency and equity issues.

The third phase assesses the cossectoral instruments. It should passed all the tests above furthe

Finally, it is necessary to considea broader perspective it is imporeallocated to other more socialsome extent beyond the scope oto an economy wide perspectivephase (referred to in Figure 5)mechanisms for addressing the n

The Worl

phasing out fossil fuel substo the Joint Report with the IEA and OECD is

ework is discussed, through a step by step decisiort here revisited through the poverty lens (Sectiorecent experience of energy reforms, focusing onsubsidies in reaching the poor through rural

er elements of reforms, including the introductioare also considered as ways to mitigate the impaction 2.2).

ework

menting the reform of fossil-fuel subsidies is idesed out because they are inefficient and lead

specific fossil-fuel subsidies are “inefficient” anmong the universe of fossil fuel subsidies that arderstanding the circumstances of each country, asumption.

be used by individual countries to rationalize adium term, focusing on the impact on the poor. e to allow individual countries to assess whetbsidy, focusing on the impact on the poor. Using

d to assess whether to retain, redesign or remoision tree (Figure 5).

he impact of existing subsidies in order to help idies that lead to wasteful consumption, con

st effectiveness of the subsidy tools compared walso be highlighted that even in the case of

er scrutiny, periodic review and monitoring is nee

er the subsidies in the context of broader policyortant to consider whether the same amount of lly or environmental desirable activities. Even if f energy subsidy reforms, and move the analysis fe for policy makers, we made some reference t) covers both the cost effectiveness of other eneeds of the poor (e.g. cash transfers).

orld Bank

Page | 27

sidies s organized as on tree already on 2.1). It also n recycling the electrification

n of automatic ct of fossil fuel

entifying those d to wasteful nd “encourage re provided by and analysis of

and phase-out In this section

ther to retain, a poverty lens, ove an energy

policy makers nsidering both

with alternative subsidies that eded.

y objectives. In money can be the issue is to

from a sectoral to it. This final economy wide

Page | 28

Figure 5: Decision tree

Source: Adapted from IEA, OPEC, OECD and World Bank (2010)

1.A Adequacy test

Has the subsidy succeeded in the objective of reaching the poor?

Phase 1

1.B Efficiency test

Are the net benefits of the subsidy positive?

2.A Impact on consumption

Does the subsidy increase consumption?

2.B Wasteful consumption test

Is the subsidy used to satisfy basic (not wasteful) needs?

3. Cost effectiveness of alternative sectoral policies

Is the subsidy the most effective sectoral tool to reach the poor?

4. Cost effectiveness of public funds

Is the subsidy the most welfare enhancing economy wide tool to reach the poor?

Phase 2

Phase 3

Phase 4

Phase out

Redesign or replace

NO

NO

NO

NO

NO

NO

Yes

Yes

Yes

Yes

Yes

Keep in an exception list and monitor over time to ensure fulfillment of all tests above

Yes

Sectoral Decisions

Economy wide

decisions

2.1.1 Phase 1 Questionpoor? If yes, at positive

2.1.1.A What share of th

To measure the performance ohelpful to define the probabilisubsidy. This index is known as th

Figure 6: Poor beneficiary incide

So

Among the overall poor popubeneficiary incidence into threecase of electricity, enables someresponses to be derived (for mor

• The share of householdetermined by the coverinfluenced by the develoreach (within a reachabloften the case in rural ainfrastructure (including

• The take-up rate amonghouseholds with potenticomponent captures botfrom affordability constrpresence of cheaper buproblems.

• The share of householdsthe third factor is determsophisticated schemes, sProgram’s (LIHEAP), the variables (see Annex 2 fo

The Worl

s: Has the subsidy succeeded in reanet benefits?

he poor is reached by the subsidy?

of a subsidy in reaching the poor, policy makeity that the targeted group (i.e. the poor) wihe beneficiary incidence (Figure 6).

ence – what share of the poor is reached by the su

ource: Adapted from Wodon et al. (2009)

ulation, policy makers may find helpful to dee components. Such a decomposition, reported e quick diagnostics of the key problems and the rre details see Annex 1):

ds with potential access to the energy sourcrage of the electricity grid among the population wopment of the infrastructure grid network and itle distance from where households live). If A is

areas, the best policy response is to develop theoff grid and rural electrification solutions) to reacgst households with potential access – meaningial access that actually use the energy source (Uth supply and demand side variables. A low value raints, due to the expensive connection rates, anut dirtier substitutes, which can in turn cause

s that are connected that are eligible for the subsmined by eligibility criteria included in the subsidysuch as the United States’ Low Income Home Enereports specific criteria for eligibility, based on s

or a short description of LIHEAP).

Poor household using the service eligible for the subsidy (T p|u)

Poor household using the service (U p|a )

Poor household with access (Ap )

All poor household (P)

Remov

Remo

Imprtarg

orld Bank

Page | 29

aching the

rs may find it ill receive the

ubsidy?

ecompose the below for the

required policy

ce (A). This is which is in turn ts geographical s low, which is most suitable

ch the poor. g the share of

U). This second of U can result

nd/or from the environmental

idy (T). Finally, y design. More ergy Assistance socioeconomic

ve supply side barriers

ove demand side barriers

rove subsidy eting design

Page | 30

The framework above has been originally designed for electricity but can be extended to all primary energy sources. Specific recommendations to reach the poor in rural areas are considered in section 2.2.1A.

What information is needed to assess this set of indicators?

A unifying framework, including indicators of inequality (including the Gini Index) and of poverty (such as Sen’s index and the members of the Foster-Gree-Thorbecke family) can be used to measure social and poverty impacts.

In the case of electricity (and natural gas) it is usually possible to determine whether a household lives in an area where a grid connection is available and used and from household survey information.9

In the case of petroleum products the only information available from household surveys is often limited to whether a household uses a specific fuel.

2.1.1.B. What share of the benefits of the subsidies goes to the poor vis-à-vis the non poor?

To further refine the indicator above one may want to look at the benefit incidence (I), namely what share of the subsidy is received by the poor (Figure 7).

Figure 7: Benefit incidence of the subsidies

Source: Adapted from Wodon et al. (2009)

The additional components, beyond the beneficiary incidence defined above, to be considered are:

• The rate of subsidization, calculated from the ratio between household consumption valued at cost-recovery prices and the actual payment, among those who benefit from the subsidy. This component can be improved by better targeting of the subsidy design.

• The quantity consumed among those who benefit from the subsidy, which depends mainly on income. Jacobson et al. (2005) show how electricity consumption is far more evenly distributed in developed than in developing countries, suggesting that the distributional pattern of electricity consumption depends heavily on a combination of wealth, income distribution and quality of infrastructure provision. Metrics relating

9 Limitations of household surveys, including lack of information on illegal connections should be kept in mind.

Beneficiary incidence Rate of subsidization Quantity consumed x

Remove demand or supply barriers

Improve subsidytargeting design

energy access to incomepricing reforms in meeti(Jacobson et al., 2005).

What information is needed to a

The rate by which households ascheme.

The quantities consumed can binformation on the tariff or pricof the tariff scheme at which dEvidence from a large sample of third of the electricity that is uscases less than the first block gen

In the case of fossil fuels, universuch as the US LIHEAP allow to remay most need subsidies througrecipients spend on average a mhouseholds and their individual a

Among fossil fuels, the share of kerosene, making it the least rekerosene is most likely to be divdiversion is taken into account, a

Results based on simulations suto be strongly regressive for 20and South and East Asia (Arze dtotal benefits from subsidized fuincome quintile receives about sThe indirect impact through highlighting, and private transport) (Figure 8b).

Summing up, where subsidies dothe richest, there is a strong caenergy consumption can be morenergy access and affordability regressive pattern of distribution

Even if subsidies are progressivbelow, after undertaking a cost-

10 Arze del Granado et al (2010) considhouseholds faced with higher prices fconsidered in Fig. 4a. The indirect imhouseholds as higher fuel costs are refle4b.

The Worl

e provide a quantitative basis to evaluate the efng economic efficiency, social equity, and enviro

assess this set of indicators?

are subsidized depends on the specific design o

be determined from household surveys and cce charged to end-users. For electricity this requidifferent prices may apply to different blocks off countries shows that the lowest quintile consumsed by the highest quintile (Komives et al. 2007nerally covered by the tariff structures.

sal subsidies are most common. More sophisticaeach a subsample of the low income household ph a careful targeting of recipients. As shown in An

more than 10 percent higher percentage of bill thaand group burden (in terms of income) is also high

the subsidy that accrues to the poor is usually tegressive fossil fuel. However, f diesel prices arverted to the transport sector and added to dies

a kerosene subsidy can became regressive.

uggest that subsidies on electricity, gasoline and 0 countries from Sub Saharan Africa, Latin Americdel Granado et al., 2010).10 Figure 8 presents theuel prices captured by each income group. On avix times more in subsidies than the bottom quinther prices for goods and services (other than cooconsumed by households shows similar regres

o not reach the poor and/or its benefits are mosase for phasing out subsidies. Alternatives to sure effective and more efficient for providing bene

to the poor (see Phase 3). Even if subsidies an indicates a clear need for changing targeting.

ve, they can be further distinguished into the tbenefit analysis. If net benefits are positive (e.g.

der the impact of a $0.25 per liter increase in fuel prices. Thefor fuels consumed for cooking, heating, lighting, and primpact through higher prices for other goods and servicected in increased production costs and consumer prices is

orld Bank

Page | 31

ffectiveness of onmental goals

of the subsidy

ombined with res knowledge consumption.

mes less than a 7) and in many

ated programs, population that nnex 2, LIHEAP an low income her.

the highest for re higher, then sel fuel. When

LPG are likely ca, Middle East e shares of the verage, the top tile (Figure 8a). oking, heating, sivity patterns

stly directed to ubsidization of efits related to are retained, a

two categories benefits more

e direct impact on ivate transport is ces consumed by considered in Fig.

Page | 32

than outweigh the costs, includinphased out. However, they needencourage wasteful consumption

Figure 8: Distribution of Subsidy

(a) by fuels (direct impact

S

2.1.2 Phase 2 Questiconsumption?

The increased consumption duewhat extent it is not wasteful. needs (basic heating in cold seathreshold by income or consumcountries' circumstances (for houreference). Subsidies that lead into the two categories, after cothe increase of consumption is kept, subject to continuous monneeds and is wasteful then energ

If the poor does not have conconnection may be needed.

Finally, wasteful consumption m

Different policy responses may b

2.1.2.A. Do fossil fuel subwasteful consumption)?

Some schemes assist householdfor home heating. Funds are avaemergencies (winter or summer

Alternative schemes have been Output Based Aid (GPOBA) and t

ng local environmental damages), subsidies may nd to be subject to the test of phase 2 to ensure thn.

Benefits By Quintile

t only) (b) by overall direct and indirect

Source: Arze del Granado et al. (2010)

ions: Have subsidies resulted in

e to the subsidies must be further analyzed to dFor example, if consumption has been used to

asons) this is surely a non wasteful purpose. Thmption must be defined at the national level, useholds the level of subsistence can represent a to excessive energy consumption can be further

omparing actual consumption related with basic not wasteful in that it satisfies basic needs, suitoring. If the increase in consumption does not

gy subsidies need to be phased out.

nection to the energy sources for basic needs

may depend from lack of demand side manageme

e needed as highlighted in the following section.

bsidies support only or mainly basic n

ds with only that portion of residential energy cailable to eligible households for assistance with crisis aid).

used to support new connections. The Global Pthe World Bank funded a scheme in several coun

0

5

10

15

20

25

30

35

40

45

Direct Indi

Bottom 2 3 4 Top

not need to be hat they do not

t impact

n wasteful

disentangle to o satisfy basic he appropriate depending on useful point of r distinguished need levels. If bsidies can be relate to basic

, subsidies on

ent.

needs (non

costs that goes energy-related

Partnership on ntries including

rect

Armenia providing grants to poheaters and in some cases boiloutputs were met, which provinstallation in a timely and effec35,000 new natural gas connectio

Pricing and rate design can also income countries, low-cost weatalso been used to break the one-sales and generate a price incedeveloped in California in the 19and 13 for electricity (Cavanagh 2

2.1.2.B Do fossil fuel suenergy sources for the p

The links between energy poveswitching from biomass to othefrom biomass collection towardHousehold use of biomass is nocauses the death of some 1.6 meffort to collect the biomass education for children.

In low income countries biomasthe main consumption fuels for – a staggering 95 percent of the use traditional biomass for cookData from other countries in othwith biomass making up the lionis the dominant fuel for this purp

Sometimes subsidies have beenHowever, fossil fuel subsidies m(2004) shows effects of displacecomparable household survey dSouth Africa and Vietnam. The firural areas of Ghana and Nepal aThe result suggests that fossil fufrom biomass, as hoped for.

Natural gas has an increasing pocooking and heating. Natural gasand produces less carbon dioxideelectricity to natural gas providiAnnex 3).

Even if subsidies do not encounecessarily require phasing out can bring substantial benefits to

The Worl

oor households for individual heating solutions lers. The funds were disbursed only after the pvided an incentive for the utility providers to ctive manner. A similar approach was used in Coons were made to poor households.

be used to reduce wasteful consumption. In midtherization projects have also been subsidized. D-for-one linkage between utility profits/revenues entive for energy efficiency (see Annex 4). T

980s, and is now implemented in 22 U.S. states f2009).

ubsidies encourage fuel switching toor?

erty and the use of biomass has been well docuer fuels allow labor (by women and children) to ds more productive purposes and helps reduce t only associated with high levels of indoor air pillion people per year) but also increase substan

restricting time available for other activitie

ss (in the form of charcoal, coal, fuel wood and dhouseholds. The vast majority of Sub Saharan Afrrural population and above 90 percent of the lowing. Similar statistics apply to the case of kerosen

her regions, including India, confirms the rural urbn’s share of rural fuel consumption for cooking. Inpose in urban areas (see Annex 3 for more details)

n justified to bring the poor further up the “enmay not always induce a move away from biomment of biomass use only in urban areas, based data from Brazil, Ghana, Guatemala, India, Nepndings vary substantially in rural areas with no fu

and partial switching in the rural areas of South Afel subsidies may not always encourage the poor

otential among urban households, where it is ms burns more cleanly than other fossil fuels, such e per unit of energy released. Some evidence of sing more affordable services to the poor has be

rage switches to modern sources of energy, t in rural areas, as changes in ways traditional f

o the poor.

orld Bank

Page | 33

based on gas predetermined

complete the olombia where

ddle and higher Decoupling has

and gross unit he policy was or natural gas,

to modern

umented. Fuel be redirected deforestation.

pollution (with tially time and

es—particularly

dung) remains rica population west quintile -- ne for lighting. ban dichotomy n contrast, LPG ).

nergy” ladder. mass. Heltbert on analysis of

pal, Nicaragua, uel switching in frica and Brazil. to move away

mainly used for as oil and coal, switching from

een found (see

his would not fuels are used

Page | 34

To summarize the first two phases, policy makers may want to address the question below: who has been benefitting from an existing subsidy? If it is primarily the rich in absolute terms, as is often the case, then there is that much stronger a case for removing the subsidy on equity grounds as well as for improved economic efficiency. Either way, however, there will be an impact on the poor coming from subsidy removal, unless the subsidy really is only used to satisfy basic needs.

2.1.3 Phase 3 Questions: Can protection of the poor be reached by alternative sectoral policy tools in a more cost-effective way?

Further redesign or fine tuning of energy subsidies may be needed to ensure that they are cost-effective. The subsidy re-design may include, for example, better targeting or the introduction of alternative or complementary sectoral policies.

Assuming there is an impact on the poor, what are the options for ameliorating those? The answer will depend in part on what the intended effect of the subsidy was. If it was to just make energy existing use more affordable, then incomes based support programs or (second best) lifeline tariffs can be considered. If it was to make energy access more viable, then switching the subsidy to access (e.g. connection costs) with full payment of incremental consumption costs would make sense.

2.1.3.A How to improve the targeting performance of existing subsidies?

One approach to improve the targeting performance of electricity subsidies is to move from traditionally used Inverted Block Tariff (IBT) to Volume Differentiated Tariff (VDT) structures, where the lowest price for the lowest block is only available to the poor. This is a feasible option only where metering exist. In countries characterized by high connection rates a move from IBT to VDT and the use of means-tested discounts substantially increases the targeting performance of subsidies. However, for low-income countries, such a change would only have a limited impact on targeting performance.

Where possible, the use of geographical or socio-economic targeting variables substantially improves the targeting performance of subsidies in the case of electricity. Apart from the US LIHEAP, examples include geographically defined subsidies in Colombia, average of provincial means-test subsidies in Argentina and winter heating allowance scheme in Georgia (Tbilisi) are all reaching the poorest quintiles (Figure 9). Geographical targeting may be more problematic for petroleum fuels, as they may be more difficult to be implemented and more vulnerable to smuggling and fuel adulteration.

Figure 9: Distribution of Subsidy

2.1.3.B What alternativepoor by increasing accesAn alternative approach is to Simulations show that connecpopulation living in areas connemost of the cases are also prdistributed in the same way as ththe subsidy would improve but nconnection subsidy will be largerover the consumption subsidy.

2.1.4 Phase 4 Questioreallocated to other mactivities?

For those fossil fuel subsidies thasubsidies in the context of broaenergy subsidies is too high, paanalysis would require comparienergy subsidy compared witinfrastructure. In addition, the nepublic funds.

Reforms of fossil fuel subsidies cross-subsidies among users, thewhether the supplier has run andirectly financing the subsidy,consumption of energy is availab

Once fossil fuel subsidies have bfree funds for addressing the nesubsidy reform. One can consid

The Worl

Benefits By Quintile

Source: Komives et al. (2007)

e sectoral instruments can be used toss?

replace consumption subsidies with connectction subsidies designed to reach a majorityected to the grid are superior to consumption suogressive (Wodon, et al., 2009). If connectionshe existing consumption subsidies the targeting p

not significantly. Nevertheless, the efficiency perfor and that would provide a compelling reason for

ons: Can the same amount of mmore socially and environmentally

at remain, it is then necessary for policymakers tder policy objectives. That is, whether the oppoarticularly in the presence of alternative policy ng at the margin the welfare impact of a unityth an increase in expenditure on health, et benefit of subsidies must be higher than the m

reduce expenditures on financing such policiesen the net effect may be revenue neutral or not, overall deficit. If that is the case, or the governm, then public funds that would have been

ble for other uses.

een removed, there is still need to consider howeeds of the poor, but that issue is beyond the scer for example how much of the overall "econo

orld Bank

Page | 35

o reach the

tion subsidies. y of unserved ubsidies and in s subsidies are performance of ormance of the r considering it

money be y desirable

to consider the rtunity cost of priorities. The

y reduction of education or

marginal cost of

s. If these are depending on

ment has been used for the

w to recycle the cope of energy omic efficiency

Page | 36

dividend" obtained from fossil fuel reform might be applied to such social goals, or should go back to general revenues.

2.1.4.A What alternative economy wide instruments can be used to reach the poor? Policy makers can compare the mean targeting performance of utility subsidies versus other targeting instruments. Komives et al. (2007) show that cash transfers and near-cash transfers (food stamps, etc.) were progressive in the great majority of cases studied. In contrast, consumption subsidy for electricity is regressive, and only one in five of the 37 cases studied was progressive. However, the implementation of targeted transfers can be challenging. Their effectiveness and efficiency depend on the targeting method and administrative capacity.

The impact of the introduction of direct cash transfer is illustrated by the Armenian case. To soften the impact of the tariff increase, a direct cash transfer of 1,450 AMD (approximately $2.70 using 1999 conversion rates) was provided to approximately 30% of households (230,000 households) eligible for the family benefit, plus an additional 9% (70,000) to those expected to have difficulty meeting their electricity payments. A significantly higher percentage of the poor (as compared to non poor) regularly consuming in the first two blocks of the 1998 electricity tariff were receiving the income transfer in 1999.

2.1.4.B How to reallocate the savings from subsidy removal to other more productive activities? A number of empirical studies have used computable general equilibrium models (CGE) to model the welfare impact of removal of energy subsidies. The benchmark dataset needed for a CGE model is generally specified in the form of a “social accounting matrix” or SAM. The construction of an accurate SAM is challenging. The raw materials take the form of the National Accounts, input-output tables, household surveys, and a variety of other data. A number of cautionary remarks need to be made. Concepts and definitions differ between data sources. And even after adjustments have been made to make definitions consistent, the estimates for what are conceptually the same totals coming from different sources will generally differ.

In what follows we summarize the results of “recycling” at least some of the savings coming from reduction of subsidies through alternative policies and explore the likely economic, social and environmental impact, through the most recent CGE literature. Recent studies simulate the distributional impact of subsidies removal and the introduction of alternative policies, including the introduction of carbon cap and dividend policies in the case of California (Kunkel and Kammen, forthcoming), Indonesia (Yusuf, 2008, Yusuf and Ramayandi, 2008), Egypt (Abouleinein et al., 2009), Argentina (Benitez and Chisari, 2010) and China (Lin and Jiang, 2010). Annex 5 reports more details on the results of the models summarized below:

• In most of the cases, fossil fuel subsidy removal has adverse economic and social impacts. Incidence of poverty is significantly lower where the subsidy removal does not

include kerosene, supporting the evidence reported in the previous sections that among petroleum fuels they are the most “progressive” (Yusuf, 2008).

• To mitigate and offset thpercentage of the savinefficiency schemes mitigpoor. Cash transfers can b

poor. Transfers targetheir welfare relativsubsidies.

Energy efficiency inimpact.

• While more difficult to iexternal costs, representdesigned to be highly prthe U. S. Government hrulemaking framework (Ualternative to fossil fuel rthe sectors, and can alsoAnnex 4 reports additiomechanisms.

Having identified the subsidies address the obstacles to reform athe subsidy. The next section wirecent experience of fossil fuel su

2.2 Lessons from reforms

From the analysis of recent expbroad set of characteristics (enefuel and electricity) some interes

2.2.1 Lessons from rural

Impact evaluations confirm the access in rural areas requires a electricity consumption levels of remoteness of the location, as wconsumers to pay fully for the cused and low income of the posystem of cost-effective subsidie

11 The average monthly consumptioconnected customers. This contrast

The Worl

he negative impact on the economy, the re-allocngs either to the poor through cash transfer

gate the economic and social impact of subsidy r

e used to mitigate the social impact of subsidy reted to the poorest quintiles of the income distribve to what it would have been in the presen

nvestment also mitigates the adverse econom

implement, penalizing polluters, e.g. through intt a preferable option. Carbon cap and dividend progressive (Kunkel and Kammen, forthcoming). A

has adopted a social cost of greenhouse-gas emUS Government, 2010). The introduction of a carremoval, penalizes less the output (and employme

o mitigate the impact on the poor (Yusuf and Ramnal mechanism to encourage the development o

that need to be reviewed and formed, governmand identify mechanisms for overcoming resistanll review the lessons learned from case studies reubsidy reforms.

recent experience of energy

perience of energy subsidy reforms by countriesergy endowment, income level, region and subssting lessons emerge as summarized below.

l electrification

welfare enhancing benefits of rural electrificatsystem of subsidies that acknowledges the lowe

f rural households and higher costs of supply resuell as the government’s social objectives.11 The in

cost of supply due to a combination of the off-gpulation in particular highlights the importance

es to ensure the recovery of the costs of an effic

n per customer in rural areas is 30kWh, and around 12ts with the average urban consumption of around 100

orld Bank

Page | 37

cation of given or to energy

emoval on the

emoval on the bution increase nce of energy

mic and social

ernalization of policies can be As an example, issions into its rbon tax, as an ent) of most of

mayandi, 2008). of clean energy

ments need to ce to changing eferring to the

y subsidy

differing by a sidy by type of

ion. Extending er income and ulting from the nability of rural rid technology of designing a

cient operation

2 KWh for newly KWh.

Page | 38

— whatever the amount contributed by the customer — so that the service in rural areas is not neglected.

A study on removing barriers to connection was undertaken for Ethiopia, found that a 10% discount leads to an average 11% increase in the probability to connect, although the effect of subsidy varies greatly in relation with a household’s initial income. Interestingly, the contingent valuation approach led to similar results, support the use of so-called ‘smart subsidies’ designed to promote access to energy for the poor, using explicit and transparent subsidies. A study in Vietnam finds that in addition to affecting income, rural electrification had its strongest impact on school attendance by children in households adopting electricity and this is true for both boys and girls. These impacts obviously have a long-term influence on the welfare of the country as a whole as these children move into the workforce both at higher and more productive levels.

Rural electrification through targeted grid-extension efforts can be most effective in reaching the poor in a relatively short period of time both in low and middle income countries In Vietnam, the establishment of the state-owned utility (EVN) and its targeted rural electrification efforts resulted in a significant increase in electrification levels and higher electrification rates for the poor households. Through an ambitious project supported by the World Bank, the Government of Vietnam has been successful in expanding rural electrification. Under this project, more than 600 communes were connected in initial phase during 2000-2004. The second phase of the project since 2005 was able to connect the remaining communes (see Figure 10).

Figure 10: Vietnam: Percent Electrification Coverage (1990-2008)

Source: World Bank (2004, 2008)

Expansion of the rural electrification system is a better and often more affordable alternative to reach the poor compared to fossil fuel subsidies which are often regressive. Although fossil fuels (including kerosene) subsidies are regressive, kerosene is -- together with biomass--the primary energy source for majority of the unconnected rural population in low income countries. A long-term policy option to reduce kerosene subsidies would be sustainable rural electricity infrastructure development. In the case of Ethiopia, for example, the average connection cost for customers living near the grid -- the so called “last mile” -- is about $ 75, representing about 15 percent of the average household annual income. By and large, kerosene is the fuel of choice in rural areas, even if it is much more expensive than electricity. 96 percent of households, which spend $1.60 per month on average on kerosene, much more expensive than a typical electricity bill that would fare around $ 8–10 per year at prevailing rates. Because

of the “last mile” obstacle, the percent per year. As a result, earural electrification expansion ins

Well designed rural electricity ssubsidy element is included mosThe result is that, in general, foslightly higher than the urban tartimes higher. In the extreme cathe contribution of a low-income

Table 3: Impact of Subsidies on T

VAD secondary grid VAD primary grid Capacity cost Energy cost Total

Note: FOSE is the Electricity SecVAD is the Value Added of Distribu

Innovative subsidy schemes havprogram in attracting private firfinancial and political commitmsector regulator, Comisión Nacioministry, as well as strong regiolaunched in 1994, resulting in thoutput targets. The central govenumber of unelectrified househoof RE projects in the precedingelectrification program and gogovernments in establishing a melectrification projects, as well adequate rate of return (Barnes e

Following the Chilean pioneerOutput Based Aid (OBA) reforimplementation of such schemesand rate of electrification. Annexrural electrification.

The Worl

connection rate has grown very slowly, at a ratasily targeted subsidies for connection can be usestead of regressive fossil fuel subsidies.

subsidies can make service affordable to the pst of the schemes allows making tariffs affordablor off grid solutions rural households pay a taririff, in spite of the fact that the supply cost could ase of a high-cost isolated generation system, sue rural household to 16 percent of the total cost (T

Tariffs for an Isolated Mini-Grid in Peru (US cents/

Full Cost With internal tariff

subsidy With intesubsidy

22.01 9.01 39.58 3.98 13.32 3.32 3

17.80 5.76 252.71 22.07 8

Source: OSINERGMIN ctor Compensation Fund, used to apply a consumption cross-tion, which remunerates services provided by the Distribution

ve proven successful. The success of Chile’s ruralrms to provide the service hinged on a high leve

ment through various national agencies, includinonal de Energía (National Energy Commission) anonal government buy-in. Chile’s rural electrifica

he creation of special mechanism (the fund) linkinernment allocates the subsidy fund to the regionsolds and the progress each region has made in theg year. The government targeted $500 million tovernment agencies provided technical assist

methodology for cost/benefit analysis and prioritias calculating the grant required for projects

et al., 2007).

ring scheme, many other countries are now ms. Box 5 reports the recent experience in ths in several countries characterized by different lex 6 reports in detail the design and effectiveness o

orld Bank

Page | 39

te of about 10 ed to promote

poor. Once the le to the poor. iff that is only be from 2 to 5 bsidies reduce

Table 3).

KWh)

ernal tariff plus FOSE .38 .49 .32 .16 .28

-subsidy; n Companies

electrification el of multiyear ng the energy d the planning tion fund was

ng subsidies to s based on the e development to support the ance to local ization of rural to achieve an

implementing he design and

evels of income of subsidies for

Page | 40

2.2.2 Lessons from social safety nets

Channeling budgetary savings arising from subsidy removal or reduction to finance better-targeted compensation packages for poorest households is a more effective alternative to regressive fuel subsidies that accrue to higher income households.

• Indonesia successfully designed targeted cash transfers that were adopted simultaneously with the fuel price increases in 2005. The Unconditional Cash Transfers program (UCT) is the largest of such programs in the world, covering 19.2 million households, or one third of the Indonesian population. The government budget savings from the cost of fuel subsidies was estimated to be about $10.1 billion in 2005-2006. Before execution of the transfers, each household was given a proxy means test. Recipients were issued smart cards (with instructions printed on the back of the cards), and transfers delivered through the post office system. The program delivered benefits of $ 30 per quarter, significantly more than the increase in energy costs. This served to increase the level of assistance for the poor, and to make fuel price increases acceptable. At the same time by covering the bottom 40% of the population, which is more than the targeted bottom 28%, the program also helped prevent those on the verge from falling into poverty (ESMAP, 2006). Other than transferring cash to lowest income households, the government also used the savings to finance programs in education, rural development, and health. The speed with which the UCT was designed and implemented meant that some leakage, targeting errors, and logistical difficulties were inevitable. However, the government responded quickly to reports of irregularities and in spite of the challenges, the program proved largely successful in reaching the poor—the poorest deciles received 21% of the benefits, while deciles 2, 3 and 4

Box 5 OBA schemes in rural electrification

In 2002 the government of Bangladesh, implemented installation of Solar Home Systems (SHS) component, financed by a Global Environment Facility (GEF) grant of $8.2 million for capital cost buy-down, was implemented as an OBA. Private companies in partnership with microfinance institutions (MFIs) and NGOs supply the SHSs. The project successfully installed its target of 50,000 SHSs by September 2005, three years ahead of schedule and $2 million below the estimated project cost. Indeed, the use of output-based subsidies is being increasingly used for rural electrification. In Argentina for instance, output based subsidies are being used to mobilize private sector expertise and provide off-grid electrification services to rural households. As of November 2008, Argentina had provided 8,000 households and 1,900 schools with access to electricity, primarily through individual solar and wind home systems. In Senegal, the rural electrification program launched in 2003 combined privately operated concessions with output based subsidies to leverage private financing resources and overcome the barrier of high up-front connection costs. The project has launched successful bidding processes, but actual results are yet to be seen (World Bank, 2010). In Ethiopia working with the state-owned utility (EEPCo) GPOBA designed and funded the smart subsidy according to which household would pay only about 20% of the connection fee after they acquired compact fluorescent lamps and metered connection. Similarly, the Government of Mexico has recently launched a new innovative rural electrification initiative based on medium term service contracts with output based subsidies to attract private sector participation and develop a sustainable rural electrification market. Rural households not connected to the grid and located more than 5 kilometers away from the grid are being electrified with stand alone systems. A first pilot is introducing, when they are least cost, decentralized power supply options based on renewable energy technologies. These options may include SHS, wind home systems (WHS), diesel-RET-battery hybrids, small scale biomass projects and micro-hydro plants (micro-grids) (World Bank, 2007).

captured 40%. In the abestimated 5% rise in the

• Mexico’s experience witand has shown commendpoverty government proname in 2002. Oportunrural and urban communeducation, health, and transfers to householdsAnother successful progrtemporary employment communities of up to 15than 5,000 residents. Thlow and opportunities athe environmental preinfrastructure (Gobierno

• In Brazil, Auxilio Gas (AGprogram established in transfers in place of LPGthe subsidy phase-out pr(Grosh et al., 2008). BF issalary. The system relieregistry provides informadministration costs andgeographical poverty mgovernment to concentpoorest areas, resultinghouseholds. The AG caprogram as LPG is commhouseholds earning less tbi-monthly.

Broader reforms aimed at protethe impact on the poor while rem

• In Jordan, the minimumreceiving higher wage inregarded as successful intime bonuses given to ltransfers provided tonongovernmental workegroups (by targeting 13 bdesign and implementat1986 as part of a strong (Kelly, 2009). Both publdelivery. Total public speabout one half spent throabout 8-10% of the popuusing less than 160 kWh

The Worl

bsence of compensation, the price hikes would hpoverty head count index. th government social safety net programs is alsodable progressive results. Oportunidades is Mexi

ogram. The program started as Progresa but theidades has been quite successful in targeting th

nities in order to help them invest in human capitnutrition of children. The distribution mechan linked to regular school attendance and healtram that shows progressive characteristics is theprogram (Empleo Temporal), which operates nat5,000 residents and gives preference to commun

he program is designed to provide work when lare few and far between. Work projects typically eservation of their community and provisFederal 2008).

G) was integrated with Bolsa Familia (BF), an in2004, to compensate low-income families with

G subsidy removal. The AG cash transfer was a virogram as LPG is commonly used as a cooking fus available to those earning less than a quarter of

es on the unified registry, Cadunico, establishedmation on all eligible beneficiaries, and has helpd the duplication of benefits. Data is collected omaps and door-to-door questionnaires. This trate extra social expenditures on households

g in a higher portion of the expenditures reacsh transfers were a vital element of the subs

monly used as a cooking fuel by the poor. Underthan R$90 per month were provided with a paym

ecting the most vulnerable have proved effectivemoving fossil fuel subsidies.

m wage was increased, with low-paid governmencreases than other employees as palliative men dealing with price increases. The package inclulow-income government employees and pensioo other low-income households whose ers or pensioners and tax exemptions aimed abasic foodstuffs). The Jordanian government alsotion of the National Aid Fund (NAF), which was social safety net program, in addition to increasic and private providers are involved in safety

ending on safety nets is estimated at more than 1%ough the NAF. The total number of beneficiaries ulation (World Bank, 2008). An electricity lifeline

per month was kept. A one off compensation fo

orld Bank

Page | 41

have led to an

o considerable co’s main anti-en changed its he neediest in al, as it targets ism uses cash th clinic visits. government’s

tionally in rural nities of fewer bor demand is are related to

sion of basic

ncome transfer targeted cash

ital element of uel by the poor f the minimum d in 2001. The ped to reduce

on the basis of enables the

s living in the hing the poor sidy phase-out r the program,

ment of R$7.50,

e in mitigating

ent employees asures, largely

uded also one-ners and cash heads were

at low-income o improved the

established in sing its funding

net programs % of GDP, with is estimated at tariff for those r the non poor

Page | 42

was also implemented by removing of government sales tax on the non-tourist restaurants, and the temporary removal for retailers with annual turnover below $1.4 million, taxis and public transport. Taxis were permitted to increase their prices, and the cost of public transport also rose. Along with subsidy reform, measures aimed at fuel substitution and energy efficiency were also implemented.

• In the case of Ghana, budget savings from fuel subsidies were directed towards transparent and easily monitorable poverty mitigation actions including the provision of extra funds the Community Health Compound Scheme to enhance primary healthcare in poorer parts of the country and the removal of fees for attending primary and junior secondary schools. In addition, planned investment in the provision of mass urban transport expansion was expedited and the existing rural electrification system was expanded.

2.2.3 Lessons from other reforms

Moving towards cost reflective fossil fuel pricing

Decreasing oil prices in 2009 and 2010 offer a unique opportunity for implementing a move from relatively ad hoc pricing to the introduction of regular reviews, autonomic price adjustments or a fully liberalized system. Liberalizing the pricing system through transparent and automatic price setting mechanisms or adopting a market based scheme can help to support subsidy reforms.

Automatic price adjustments can be a useful transitory step in moving towards a fully liberalizing pricing scheme. They are based on predetermined formulae and at regularly defined periods. The price adjustments are passed on to the market relatively quickly. However, the formulae can be modified to allow price changes only when certain thresholds are exceeded. The prevailing regulations are often monitored by a state institution or a panel of experts. Moving from one system to the other also significantly smoothen the impact of subsidy removal over time, as the trends for Egypt, Indonesia and India show (Figure 11).

Reforms moving towards automatic price adjustments mechanisms and fully liberalized system can be politically challenging. There have been examples of backtracking. In June 2010 Kenya’s parliament passed a bill allowing the country to return to price controls of essential food and fuel goods, after the policy was abandoned in the 1990s in favor of economic liberalization. Since the liberalization of the oil sector in the 1990s, the government has had no price control mechanism in place leading to pump prices being implemented arbitrarily by the dealers based on the international oil prices. In 2009 in Vietnam the government announced the move toward fully liberalized fuel prices, but was forced to suspend the plan to control soaring consumer prices amid record high world crude oil prices. Vietnam, which relies almost entirely on oil product imports as it lacks refineries, slashed retail petrol prices by 5.3 percent in August 2010 (the second time in two weeks). In June 2008, Cote d’Ivoire increased the price of diesel by 44% and petrol was raised by 20%. Following public out roars, the government had to revert from the price hikes two weeks after the announced price adjustments. The budgetary implications of this action were significant in the form of political costs and budgetary effects in terms of transportation costs for civil servants and shortfalls in revenues.

Figure 11: Petroleum price settin

Source: Data upda

Several countries both inside anprice mechanisms and full liberaChina, announced further reformsame. This is the case for Malaysprice is determined by an augovernment is currently propointernational market variations. Tfuel price-stabilization fund that recent average import price leveabove a given price range.

Introducing oil sovereign

Oil SWFs can help producing comain objectives of oil funds areprices, to foster investment in brincome equitably across geneimmediate use into which partiinstrument to manage oil revecontributor to public revenue, integral for the positive developmfrom SWFs can be channeled toAzerbaijan. Table 4 reports the acountries that have adopted SWF

No regular price increases

Egypt, Bangladesh and Cote d’Ivoire

The Worl

ng

ted from Coady et al. (2010), based on national statistics

nd outside the G-20 countries are moving towaalization. G-20 countries, including Indonesia, Mems (as reported in section 1.5). Other countries sia, where prices are now subject to a managed futomatic pricing mechanism, but also Angola.osing a new mechanism to buffer local fueThe old mechanism that expired at the end of Jukept the price of fuel import within a price band

l, effectively subsidizing diesel and gasoline prices

n wealth funds (SWFs)

ountries to protect the poor while phasing out se to shield the domestic economy from the volaranches other than natural-resources exploitationerations. Oil funds are stabilization funds prial state oil revenue is channeled. They can b

enue for countries where natural resources arand as a result the prudent management of t

ment within a stable macroeconomic environmeno implement social safety nets or programs as asset value of the fund for the latest available yeFs.

Automatic price mechanisms

Malaysia

Adopted (but then reversed): Ghana, Morocco, Indonesia

Plans: Chile

Fully liberalize

Most OECD coand Philippines

Adopted (but Kenya

Plans: India, Mexico, Angola

orld Bank

Page | 43

ards automatic exico, India and have done the

flow where the In Chile the l prices from ne 2010 was a

d to match the s if they gained

subsidies. The atility of world n, and to share rotected from

be an effective re a dominant his revenue is t. Partial funds in the case of ar for selected

d pricing

ountries, Uganda s

then reversed):

China, UAE, a

Page | 44

The Norwegian Fund display several features that could serve as a model for other oil funds.

• Norges Bank Investment Management (NBIM) manages the fund on behalf of the Ministry of Finance, which owns the fund on behalf of the Norwegian people. The fund is fully integrated with the state budget and that net allocations to the fund reflect the total budget surplus, including petroleum revenue. The ministry determines the fund’s investment strategy, following advice from among others, the NBIM and discussions in Parliament.

• The ministry regularly transfers petroleum revenue to the fund. As of October 2010, the fund’s overall value is $512 billion. The capital is invested abroad, to avoid overheating the Norwegian economy and to shield it from the effects of oil price fluctuations. In 2001 it was established that no more than 4 percent of the fund’s return should over time be spent in financing the non oil budget deficit.

Table 4: Progress in the establishment of SWFs

Country Fund Name Assets ($m) Asset Value

as of Inception Year

Source of Funds

Kuwait Reserve Fund for Future Generations

250,000 2007 1953 Oil

UAE Abu Dhabi Investment Authority 875,000 2007 1976 Oil

Norway Government Pension Fund-Global 512,000 2010 1976 Oil

Canada Alberta Heritage TF 14,400 2010 1976 Oil & Gas

Azerbaijan SOFAZ 18,000 2008 1999 Oil

Iran Oil Stabilization Fund 8,000 2007 1999 Oil

Algeria Fonds de regulation des recettes 58,113 2008 2000 Oil

Kazakhstan NFRK 21,600 2008 2000 Oil

Nigeria Excess Crude Account < 5 2010 2003 Oil

Russia Stabilization Fund 157,000 2008 2004 Oil

Libya Oil Reserve Fund 50,000 2007 2005 Oil

Source: Adapted and updated from IMF

More transparency and better governance has also been achieved by oil funds in the Middle East and Euro Asia

• Not surprisingly, the Abu Dhabi Investment Authority (ADIA) represents the largest oil fund in the world invests the oil surplus of Abu Dhabi, the richest city state within the United Arab Emirates, which also includes Dubai. Since May 2008, ADIA acted alongside the International Monetary Fund, as co-chair of the International Working Group of sovereign wealth funds. The goal was to create an agreed framework of Generally Accepted Principles and Practices that reflected appropriate governance and accountability arrangements, as well as the prudent and sound basis on which SWFs conduct their investments.

• The Azeri State Oil Fund to join the Extractive Indthe fund’s assets reached

• Responding to the receRussia, have tapped oil fpoor from the recent fina

Figure 12: Trends in the Azeri and

Source: IMF Staff Reports (20

• Iran and Nigeria are alsoto support budgetary sh(ECA) and the Nigeria Revenue Account (ECA)million as state govern(Aderinokun, 2010). BothFund in late 2010 to prev

• It is worth noting that mentioned above up to unsuccessful (for details,

02000400060008000

100001200014000

2005 2006 20

miii

ons

of M

anat

SOFAZ

The Worl

(SOFAZ) was established in 1999 and in 2003 wasdustries Transparency Initiative (EITI) in 2003. Asd $18 billion. nt financial crisis, several countries, including Afunds to deploy a large fiscal stimulus program aancial crisis (Figure 12, IMF, 2009 and 2010).

d Kazakh oil funds

009, 2010) and http://www.oenb.at/en/img/feei_2007_1_astrov_tc

o about to set up SWFs to prevent oil proceed to bhortfalls. Experience from the Iranian Excess Rev

fund have been mixed. Since September 200)’s resources have fallen from $20 billion to jus

nors have used Nigeria’s oil revenue to fund h governments announced plans to create a Sovvent from raiding oil proceeds from their funds. another type of oil stabilization funds was useJune 2010) and other countries, but their experi see ESMAP, 2006).

007 2008 2009

Z

0

2000

4000

6000

8000

10000

2007 2008 2009 2bi

iions

of R

uble

s

Russian Oil Fun

NWF

orld Bank

Page | 45

s the first fund s of July 2010,

Azerbaijan and nd protect the

m16-58446.pdf

be used mainly venue Account 07, the Excess st under $ 500

their budgets vereign Wealth

ed in Chile (as ence has been

2010 2011

d

Page | 46

References Abouleinein, S., H. El-Laithy and H. Kheir-El-Din (2009), The Impact of Phasing Out Subsidies of Petroleum Energy Products in Egypt, Egyptian Center for Economic Studies, Working Paper No.145.

Aderinokun, K. (2010), Nigeria: Excess Crude Account to Drop to N528.4 Billion, available at http://allafrica.com/stories/201005270386.html, accessed 24 October 2010.

APPRISE (2005), LIHEAP Energy Burden Evaluation Study. U.S. Department of Health and Human Services. PSC Order No. 03Y00471301D.

Barnes, D. (2007), The Challenge of Rural Electrification, Resources for the Future, ESMAP, Washington, DC.

Benitez, D. and O. Chisari (2010), Recycling Universal Energy Subsidies on Residential Consumers Alternative Fiscal Policies to Help the Poorest. The World Bank.

Boqiang, L. and Z. Jiang (2010a), The Analysis and Design of China’s Residential Electricity Tariff Subsidies. Financial Research Journal 11, pp. 45-58.

– (2010b), Estimates of Energy Subsidies in China and Impact of Energy Subsidy Reform. Energy Economic,s Forthcoming.

Coady, D., et al. (2010), “Petroleum Product Subsidies: Costly, Inequitable, and Rising”, IMF Staff Position Note, SPN/10/05, IMF, Washington, DC.

Del Granado, J., D. Coady and R. Gillingham (2010), The Unequal Benefits of Fuel Subsidies: A Review of Evidence for Developing Countries. IMF Working Paper, WP/10/202.

Ellis, J. (2010), “The Effects of Fossil-fuel Subsidy Reform, a Review of Modelling and Empirical Studies”, Untold Billions: Fossil-fuel Subsidies, Their Impacts and the Path to Reform, Global Subsidies Initiative of the International Institute for Sustainable Development, Geneva, April.

ESMAP (Energy Sector Management Assistance Program) (2006), Coping With Higher Oil Prices, The World Bank, Washington, DC.

Foster, J.E., J. Greer, E. Thorbecke (1984), A Class of Decomposable Poverty Indices, Econometrica 52, pp.761-766.

GSI (Global Subsidies Initiative) (2010), Defining Fossil-Fuel Subsidies for the G-20: Which Approach is Best, Global Subsidies Initiative of the International Institute for Sustainable Development, Geneva.

Government of India (2010), Report of the Expert Group on a Viable and Sustainable System of Pricing of Petroleum Products, Government of India, New Delhi.

Grosh, M., et al. (2008), The Design and Implementation of Effective Safety Nets for Protection and Promotion, World Bank, Washington, DC.

Gomez, M. and S. Silveria (2010), Rural Electrification of the Brazilian Amazon – Achievement and Lessons. Energy policy, Vol. 38, pp. 6251-6260.

Hale, T.N. (2008), “Transparency, Accountability and Global Governance”, in T. Farer and T.D. Sisk (eds.), Global Governance, Vol. 14, No. 1, Lynne Rienner Publishers, Inc., Boulder, CO.

Heltberg, R. (2004), Fuel switchinVolume 26, Issue 5, September 2

IEA (International Energy Agency

– (2008b), Deploying Renewables

– (forthcoming), World Energy O

IEA, OECD (Organisation for Eco(Organization of the Petroleum Subsidies and Suggestions for the20 Meeting of Finance Ministers

IMF (International Monetary FunDC.

– (2010), Russia Country Staff Re

Jacobson, A., A. Milman and D.Lorenz curves of cumulative elecdistribution and equity, Energy Po

Kelly, E. (2009), Pricing for ProspReform, University of London and

Khandker, S., D. Barnes and M. Sfrom Vietnam; Impact EvaluationBank.

Komives, K., et al. (2005), WaterDirections in Development, ISBN-

Komives, K., et al. (2007), UtilTargeting Performance. Developm

Koplow, D. (2009), Measuring ELeave Out?, International Institut

Koplow, D., et al. (2010), “MappCountries”, Untold Billions: FossInitiative of the International Inst

Kosmo, M. (1987), Money to BInstitute, Washington, DC.

Kunkel, C. and D. Kammen (fodividend policies, Energy Policy.

Laan, T. (2010), Untold Billions: Gaining Traction: the ImportancSubsidies, Global Subsidies InitiaGeneva.

Lampietti, J., et al. (2001), Utility

Larsen, B. and A. Shah (1992), “WResearch Working Papers, WPS 1

The Worl

ng: evidence from eight developing countries, Ener2004, pp. 869-887.

y) (2008a), Energy Policy Review of Indonesia, OEC

s: Principles for Effective Policies, OECD/IEA, Paris

utlook 2010, OECD/IEA, Paris.

nomic Co-operation and Development), World BExporting Countries) (2010), Analysis of the Sc

e G-20 Initiative, Joint Report prepared for submiand Central Bank Governors, Busan, Korea, 5 June

nd) (2009), Azerbaijan Country Staff Report, IMF

port, IMF, Washington DC.

. Kammen (2005), Letting the (energy) Gini out ctricity consumption and Gini coefficients as meolicy, 33, pp. 1825-1832.

perity: Consumer Price Subsidies in MENA and thed Fiscal Studies, London, UK.

Samad (2008), Welfare Impacts of Rural Electrifican Series No. 38, Policy Research Working Paper 50

r, Electricity, and the Poor: Who Benefits from Ut-10: 0-8213-6342-5, The World Bank.

lity Subsidies as Social Transfers: An Empirical ment Policy Review 25 (6), pp. 659-679.

Energy Subsidies Using the Price-Gap Approach: te for Sustainable Development, Winnipeg.

ing the Characteristics of Producer Subsidies: A Rsil-Fuel Subsidies, the Impact and the Path to Rtitute for Sustainable Development, Geneva, Augu

Burn? The High Costs of Energy Subsidies, Wo

orthcoming), Design and implementation of ca

Fossil-Fuel Subsidies, Their Impacts and the Patce of Transparency in Accelerating the Reformtive of the International Institute for Sustainable

Pricing and the Poor, WTP497.

World Fossil Fuel Subsidies and Global Carbon Em1002, World Bank, Washington, DC.

orld Bank

Page | 47

rgy Economics,

CD/IEA, Paris.

s.

Bank and OPEC ope of Energy ssion to the G-e.

F, Washington,

of the bottle: trics of energy

e Prospects for

ation: Evidence 057, The World

tility Subsidies?

Evaluation of

What Does it

Review of Pilot Reform, Global ust.

orld Resources

arbon cap and

h to Reform − m of Fossil-fuel

Development,

issions”, Policy

Page | 48

LIHEAP Home Energy Notebook (2008), .S. Department of Health and Human Services.

Maurer, L. and C. Nonay (2009), Output Based Aid in Ethiopia: Dealing with the “last Mile” Paradox in Rural Electrification, OBA Approaches, Note Number 27.

Mourougane, A. (2010), “Phasing Out Energy Subsidies in Indonesia”, OECD Economics Department Working Paper, No. 808, Paris.

Niez, Alexandra; 2010. Comparative Study on Rural Electrification Policies in Emerging Economies: Keys to Successful Policies, Information Paper, IEA.

OECD (2006a), Subsidy Reform and Sustainable Development: Economic, Environmental and Social Aspects, Paris.

– (2006b), Subsidy Reform and Sustainable Development: Political Economy Aspects, Paris.

OECD (2010), “Measuring Support to Energy - Version 1.0", Background paper to the joint report by IEA, OPEC, OECD and World Bank to G20 Leaders on “Analysis of the Scope of Energy Subsidies and Suggestions for the G-20 Initiative” (June 2010). www.oecd.org/g20/fossilfuelsubsidies.

Pantanali, C. and J. Benavides (2006), Subsidios Electricos en America Latina y el Caribe: Analisis Comparativo y Recomendaciones de Politica, Banco Interamericano de Desarrollo.

Perl, L. (2009), The Low Income Home Energy Assistance Program (LIHEAP): Program and Funding. CRS Report for Congress, RL31865.

Shenoy, B.V. (2010), Lessons Learned from Attempts to Reform India’s Kerosene Subsidy, International Institute for Sustainable Development, Geneva.

Steenblik, R. and P. Coroyannakis (1995), “Reform of Coal Policies in Western and Central Europe”, Energy Policy, Vol. 23, No. 6, Elsevier, Amsterdam.

Tanguy, B. and M. Torero (2009), Welfare Impact of Rural Electrification, Short Term Evidence from Ethiopia: A Report for the Ethiopian Electric Power Corporation and the World Bank, International Food Policy Research Institute.

Tanguy, B. (2010), Impact Analysis of Rural Electrification Projects in Sub-Saharan Africa, The World Bank Research Observer.

Terrado, E., A. Cabraal and I. Mukherjee (2008), Designing Sustainable Off-Grid Rural Electrification Projects: Principles and Practices. Operational Guidance for World Bank Group Staff.

UNEP (United Nations Environment Programme) (2008), Reforming Energy Subsidies, UNEP, Nairobi.

UNEP and IEA (2002), Reforming Energy Subsidies: An Explanatory Summary of the Issues and Challenges in Removing or Modifying Subsidies on Energy That Undermine the Pursuit of Sustainable Development, UNEP, Nairobi.

Velculescu, D. (2008), Norway’s Oil Fund Shows the Way for Wealth Funds, IMF Survey Magazine: Policy.

Verma, Sid. (2010), Nigerian lawmakers to discuss oil fund, available at www.emergingmarkets.org/Article/2690779/Nigerian-lawmakers-to-discuss-oil-fund.html, accessed 27 October 2010.

Wagner, A. (2010), Exploit FallMechanisms. GTZ.

Wodon, Q., ed. (2009), ElectricitWorking Paper 12, Africa Infrastr

World Bank (2000), Vietnam Rura

– (2004), Vietnam Second Rural E

– (2006) Peru: Rural Electrificatio

– (2007), Mexico: Integrated Ene

– (2009), Peru Proyecto de Electr

– (2010), “Subsidies in the EnergGroup Energy Strategy, World Ba

Yusuf, A. (2008), The DistributionEnergy Pricing Reform in Indones

Yusuf, A. and A. Ramayandi (2008Instruments from the Economy, Eof Economics, Padjadjaran Unive

The Worl

ling Markets – A Contribution to the Debate o

ty Tariffs and the Poor: Case Studies from sub-Sructure Country Diagnostic, Washington, DC: Wor

al Energy project: Project Appraisal Document

Energy Project: Project Appraisal Document

on Project, Project: Project Appraisal Document

rgy Services Project, Project: Project Appraisal Do

rificacion Rural: Ayuda de Memoria 11 al 22 de ma

gy Sector: An Overview,” Background Paper for thank, Washington DC.

nal Impact of Environmental Policy: The Case of Csia. Research Report No. 2008-RR1.

8), Reducing Fuel Subsidy or Taxing Carbon? CompEnvironment, and Equity Perspective for Indonesirsity. No. 200808.

orld Bank

Page | 49

of Fuel Pricing

Saharan Africa. ld Bank.

ocument

ayo 2009.

he World Bank

Carbon Tax and

paring the Two a. Department