Evaluating the Fiscal and Social Cost of Fuel Subsidies Poverty and Social Impact Analysis (PSIA) Group Fiscal Affairs Department March 2006.

Evaluating the Fiscal and Social Cost of Fuel Subsidies

Poverty and Social Impact Analysis (PSIA) GroupFiscal Affairs Department

March 2006

Structure of Presentation

• Background to PSIA on fuel subsidies

• Objective of the PSIA studies

• Methodology and data (five steps)

• Examples of policy responses

• Policy messages from PSIA

Background I: Market Structure

• Most developing countries control the domestic pricing and distribution of petroleum products– Recent FAD survey found that from 48

countries• 16 had fully liberalized systems• 9 had functioning automatic pricing formulae (+8

suspended recently)• 23 had ad hoc pricing

– Liberalized in Kenya, Tanzania, Uganda– Automatic in Ghana, Senegal, South Africa

Background II: Prices and Subsidies

• World prices have increased >100% since 2000• Controlled prices have resulted in rising budget subsidies

in many countries (% 2005 GDP, projd.)– Yemen, 9.2; Jordan, 6.6; Indonesia, 3.2; Bolivia, 0.8– Subsidy rates typically higher for kerosene and diesel as well as in

exporting countries

• Countries often respond by decreasing taxation, so-called tax expenditures (especially kerosene and diesel)– e.g. Bangladesh, India, Sri Lanka, Kenya, Zambia

• Implicit subsidies also often substantial and take form of quasi-fiscal deficit financed by debt (%GDP2005, projd.)– Azerbaijan, 9.9 (2.8ex); Egypt, 4.1; Ecuador, 3.6; Bolivia, 2.3

Crude oil prices,1970-2005 ($/bbl)

$-

$10

$20

$30

$40

$50

$60

$70

$80

1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004

Real 2005 $ (GDP deflator)

Nominal

Background III: Reform Agenda

• Fuel subsidies seen as undesirable because– High fiscal cost with consequences elsewhere in

budget (Indonesia: subsidies exceeded combined health and education budgets)

– Inefficient: leads to over-consumption• Governments still reluctant to increase domestic

prices in line with world prices– Concerns about impact on poor and politically

unpopular– PSIA can inform choice of appropriate policy response

(so far: Bolivia, Ghana, Jordan, Mali, Sri Lanka...expected to increase!...Angola, Bangladesh, Cameroon, Ethiopia, Gabon, Honduras, Madagascar, Nicaragua, Nigeria)

Objective of PSIA

• To evaluate the social and fiscal cost of consumer subsidies

• To evaluate the aggregate and distributional incidence of their withdrawal on household real incomes

• To identify appropriate mitigation measures to offset adverse impact on poorest households

Methodology and Data

• Higher domestic prices affect consumers through two channels– Direct effect from increase in price of fuels

consumed by households– Indirect effect from increase in prices of goods

and services that use fuel as inputs– Indirect effect often substantial since over 50

percent of total consumption of fuel is as intermediate product

Step I: Identify extent and “location”

• This requires a reference price for each product and required price increases– For most countries, border (cif,fob) price

(plus,minus) domestic trade and transport margins

– Often existing or desired tax levels included in reference price to allow for “tax expenditures”

Fiscal and social cost of subsidies

P

Q

Pm

Pc

Pp

Ps

Qc Qs

Demand

A B C

D E

• Domestic refinery that imports product

• Import at P(m), produce at P(c)• Subsidized domestic price is P(s)• Produces Q(c), imports Q(s)-Q(c)• Total consumer subsidy =

(A+B+C)=Q(s)[P(m)-P(s)]• Where shows up depends on price

to producer. If taxes, P(p), P(s) – Explicit import subsidy=(B+C)

– Loss in profits=(A+D)+E

– Tax revenue=(D+E)

– Net fiscal position

• On budget: (D+E)-(B+C)

• Off budget: -(A+D+E)

Example from Ghana

Petrol Kerosene Diesel Fuel Oil LPG World CIF price (F) 2,890 3,761 3,884 1,750 5,355 Total taxes (F) 1,711 1,342 1,811 1,141 1,543 Domestic margins 577 677 577 0 990 Ex-pump price (F) 5,179 5,780 5,940 2,891 7,888 Ex-pump price (A) 4,444 3,889 3,556 1,927 3,800 Required increase in current prices

17 percent 49 percent 67 percent 50 percent 108 percent

Note: Prices are in cedis per litre, except for LPG which is cedis per kg. A and F denote that prices are based on actual and formula levels, respectively. Ex-refinery price (F) is based on an average of Platt prices for October–December 2004, an exchange rate of c9,133.33 per US dollar and import margins at 13.8 percent of import CIF price. Included in taxes is an ad valorem excise set at 15 percent of the ex-refinery price.

Step II: Calculate direct effect

• Need household survey with information on different fuel expenditures

• For each household, calculate budget shares as expenditure on fuel divided by total household consumption

• Multiply required price increases by budget share to get approx. real income impact

• Look at distribution of percentage real income effect across income groups (regressive vs. progressive)

Example from Ghana

Bottom 2nd

Quintile 3rd

Quintile 4th

Quintile

Top

All Household Budget Shares

Petrol 0.001 0.001 0.002 0.002 0.021 0.006 Kerosene 0.059 0.041 0.034 0.024 0.016 0.035 LPGas 0 0 0 0.001 0.002 0.001

Real income effect( percentage change in consumption)

Direct effect 0.029 0.020 0.017 0.013 0.014 0.019 Indirect effect 0.062 0.066 0.067 0.069 0.068 0.067

Total effect 0.091 0.087 0.085 0.082 0.082 0.085

Indirect as percent of total 68 percent 77 percent 80 percent 84 percent 83 percent 80 percent

Share of the aggregate loss

Direct effect 0.135 0.160 0.180 0.193 0.332 1.000 Indirect effect 0.077 0.137 0.184 0.256 0.346 1.000 Total effect 0.088 0.142 0.184 0.244 0.343 1.000

Mean consumption (Ratio to bottom quintile)

1.00 1.76 2.55 3.80 7.48 3.31

Step III: Calculate indirect effect

• An input-output table and a simple model can be used to calculate the increase in prices for other goods and services from higher fuel costs

• Aggregate household consumption data to get budget shares for input-output sectors

• Multiply budget shares by percentage price increases to get percentage real income effect

• Aggregate to get total indirect effect and look at distribution across different income groups

• Add to direct effect to get total impact of fuel price increase on household real incomes and distribution

Example from Ghana

Sector Budget Share (BS)

Price Effect (dP)

Impact=BS*dP

Agriculture 0.452 0.066 0.030 Utilities and mining 0.021 0.116 0.002 Manufacturing 0.253 0.052 0.013 Construction 0.000 0.107 0.000 Trade 0.070 0.107 0.007 Transport 0.032 0.267 0.008 Business 0.025 0.025 0.001 Community 0.097 0.048 0.005 Electricity 0.008 0.000 0.000

Example from Ghana

Bottom 2nd

Quintile 3rd

Quintile 4th

Quintile

Top

All Household Budget Shares

Petrol 0.001 0.001 0.002 0.002 0.021 0.006 Kerosene 0.059 0.041 0.034 0.024 0.016 0.035 LPGas 0 0 0 0.001 0.002 0.001

Real income effect( percentage change in consumption)

Direct effect 0.029 0.020 0.017 0.013 0.014 0.019 Indirect effect 0.062 0.066 0.067 0.069 0.068 0.067

Total effect 0.091 0.087 0.085 0.082 0.082 0.085

Indirect as percent of total 68 percent 77 percent 80 percent 84 percent 83 percent 80 percent

Share of the aggregate loss

Direct effect 0.135 0.160 0.180 0.193 0.332 1.000 Indirect effect 0.077 0.137 0.184 0.256 0.346 1.000 Total effect 0.088 0.142 0.184 0.244 0.343 1.000

Mean consumption (Ratio to bottom quintile)

1.00 1.76 2.55 3.80 7.48 3.31

Step IV: Evaluate targeting efficiency

• Calculate the share of the total subsidy (or, equivalently, the burden of subsidy removal) accruing to each income group

• Can do this separately for each product as well as the direct, indirect and total effects

• Individual product shares useful later when comparing alternative approaches to protecting the real incomes of low-income households

Example from Ghana

Bottom 2nd

Quintile 3rd

Quintile 4th

Quintile

Top

All Household Budget Shares

Petrol 0.001 0.001 0.002 0.002 0.021 0.006 Kerosene 0.059 0.041 0.034 0.024 0.016 0.035 LPGas 0 0 0 0.001 0.002 0.001

Real income effect( percentage change in consumption)

Direct effect 0.029 0.020 0.017 0.013 0.014 0.019 Indirect effect 0.062 0.066 0.067 0.069 0.068 0.067

Total effect 0.091 0.087 0.085 0.082 0.082 0.085

Indirect as percent of total 68 percent 77 percent 80 percent 84 percent 83 percent 80 percent

Share of the aggregate loss

Direct effect 0.135 0.160 0.180 0.193 0.332 1.000 Indirect effect 0.077 0.137 0.184 0.256 0.346 1.000 Total effect 0.088 0.142 0.184 0.244 0.343 1.000

Mean consumption (Ratio to bottom quintile)

1.00 1.76 2.55 3.80 7.48 3.31

Bolivia Ghana Jordan Mali Sri Lanka Size of fuel subsidy (% of 2004 GDP)

2.2 percent 2.2 percent 3.6 percent 2 percent

2.1 percent

Price increases (simulated)

Raising prices world prices and LPG by 67 percent

Restoring a formula price that includes taxes

Eliminating subsidies and imposing a standard sales tax.

Raising oil prices by 34 percent

Raise average oil prices by 36.7 percent

Kerosene 49 percent 79.5 percent 34.0 percent 94 percent LPG 67 percent 108 percent 47.5 percent 34.0 percent n.a. Diesel 40 percent 67 percent 77.1 percent 34.0 percent 40 percent Gasoline 40 percent 17 percent No change since tax>16% 34.0 percent 14 percent Other 50 percent (Fuel Oil) 77-84 percent 34.0 percent n.a Total 50 percent? 50 percent 67.7 percent (excl. gasoline) 34.0 percent 36.7 percent Aggregate real income impact

Direct 1.6 percent (neutral) 1.9 percent (very regressive) 2.0 percent (regressive) 0.85 percent (regressive, U) 1.2 percent (regressive) Indirect 3.3 percent (regressive) 6.7 percent (slightly

progressive) 2.4 percent (slightly progressive) 0.82 percent (regressive, U)

(1.06 if electricity incl.)

1.2 percent (neutral)

Total 5.0 percent (regressive) 8.5 percent (regressive) 4.4 percent (regressive) 1.7 percent (regressive, U) (1.94 if electricity incl.)

2.4 percent (regressive)

Share of subsidy ( poorest 40 percent)

Direct 29.5 percent 22.9 percent 22.6 percent 27.2 percent Indirect 21.4 percent 19.8 percent 24.4 percent 23.0 percent Total 15.3 percent 23.0 percent 21.2 percent 23.9 percent 25.1 percent

Step V: Identify mitigating measures

• Can consider a number of alternatives and simulate using household-level data– Gradual withdrawal of specific fuel subsidies

(kerosene, LPG) to minimize revenue-poverty trade-off– Using some of budgetary savings to finance targeted

public expenditures (education, health, roads, transport, electricity)

– Restructure electricity tariff schedules to reduces cost for poor

– Use savings to finance existing/reformed/new social safety net for poorest households

Example from Ghana

Bottom 2nd

Quint 3rd

Quint 4th

Quint

Top

Benefit Shares

Education

Untargeted 0.215 0.225 0.219 0.187 0.154

Targeted 0.204 0.279 0.249 0.17 0.098

Health

Untargeted 0.149 0.193 0.208 0.207 0.244

Targeted 0.148 0.229 0.208 0.226 0.189

Rural electrification 0.329 0.251 0.212 0.135 0.074

Urban transport 0.299 0.128 0.185 0.28 0.376

Proxy-means targeting 0.373 0.277 0.205 0.111 0.347

Kerosene subsidy 0.178 0.211 0.227 0.209 0.174

Note: Quintiles are based on the national distribution of household consumption per adult equivalent

Example from Sri Lanka

• Subsidy Coverage, Level, and Targeting

• Simulated Effect of Price Increases with Compensating Measures

• Real Income Impact of Electricity Tariff Subsidies

• Real Income Effects of Alternative Tariff Structures

Bottom 2nd

Quintile3rd

Quintile4th

QuintileTop All

Kerosene Subsidy

Coverage 0.777 0.689 0.687 0.618 0.449 0.644

Coverage Share 0.211 0.204 0.212 0.201 0.173 1.000

Avg effect 0.017 0.011 0.009 0.007 0.003 0.010

Amount share 0.211 0.199 0.198 0.215 0.177 1.000

Samurdhi Food Stamps

Coverage 0.634 0.512 0.454 0.328 0.125 0.410

Coverage Share 0.276 0.237 0.227 0.179 0.081 1.000

Avg Effect 0.023 0.013 0.010 0.006 0.001 0.011

Amount Share 0.287 0.241 0.233 0.173 0.065 1.000

Proxy Means Food Stamps

Coverage 0.825 0.566 0.395 0.204 0.033 0.405

Coverage Share 0.391 0.281 0.203 0.105 0.020 1.000

Avg Effect 0.033 0.016 0.009 0.004 0.000 0.012

Amount Share 0.391 0.281 0.203 0.105 0.020 1.000

Bottom 2nd

Quintile3rd

Quintile4th

QuintileTop All

No kerosene price increase

Avg effect 0.012 0.014 0.014 0.015 0.018 0.015

Amount share 0.069 0.111 0.145 0.201 0.474 1.000

Samurdhi food stamps

Avg Effect 0.007 0.012 0.013 0.017 0.020 0.014

Amount Share 0.036 0.092 0.129 0.219 0.524 1.000

Proxy means uniform grant

Avg Effect -0.003 0.009 0.014 0.019 0.021 0.012

Amount Share -0.011 0.075 0.142 0.250 0.544 1.000

Bottom 2nd

Quintile3rd

Quintile4th

QuintileTop All

Real income impact (Percent of consumption)

Current Tariff Structure 0.012 0.018 0.019 0.026 0.032 0.022

Simulation 1 0.009 0.013 0.014 0.020 0.031 0.018

Simulation 2 0.013 0.019 0.018 0.021 0.020 0.019

Simulation 3 0.013 0.018 0.018 0.020 0.020 0.018

Simulation 4 0.013 0.021 0.020 0.023 0.019 0.020

Share of Subsidy

Current Tariff Structure 0.048 0.101 0.128 0.224 0.499 1.000

Simulation 1 0.059 0.120 0.146 0.235 0.440 1.000

Simulation 2 0.070 0.140 0.165 0.249 0.376 1.000

Simulation 3 0.069 0.131 0.160 0.233 0.407 1.000

Simulation 4 0.077 0.153 0.178 0.259 0.333 1.000

Note: Total subsidies under the current tariff structure (see Table 7) are equivalent to 2.4 percent of total household consumption and 2.1 percent of GDP in 2004. The budget subsidies under each of the various simulations is reduced by 28 percent compared to the existing subsidy total, i.e. to approximately 1.5 percent of GDP in 2004.

0.0

1.0

2.0

3.0

4S

ubs

idy

Imp

act,

%

4 6 8 10Log Per Capita Consumption

Current subsidy Simulation 1Simulation 4

Figure 1. Real Income Effects of Tariff Structures

Mitigating Measures: Ghana

• Introduced formula in January 2003 with 90 percent price increase. – But formula abandoned and subsidies of 2.2% GDP2004

• February 2005 introduced new formula and set up National Petroleum Authority (broad stakeholder group) to depoliticize implementation. – Prices increased in March/June/August/October 2005 and initial

moves to liberalizing markets (import tendering)

• Announced range of mitigating expenditures (financed by “mitigating levy” in formula)– Removal of fees for primary and junior secondary school– Increased investments in mass urban transport– Expansion of rural electrification scheme

Mitigating Measures: Jordan

• In 2004 subsidy of 3.2 percent GDP, projected at 8.5 percent for 2005

• A 68 percent increase in prices (including taxes) needed to eliminate subsidies

• Price increases would lead to 4.4 percent reduction in real incomes (5.4% for bottom quintile)

• In July 2005 increased prices by over 25%, reducing subsidies to 3% annual basis

• Introduced range of mitigating measures– Raised minimum wage and increased salaries for low-paid state

employees– Maintained lifeline electricity tariff– Provided one-time bonus to government employees and pensioners

earning less than JD400/month– Will increase when targeting is improved– LPG, diesel, kerosene prices expected to reach import parity by March

2007 and intend to liberalize thereafter

Mitigating Measures: Indonesia

• Ad hoc system froze prices between 2002 and February 2005, when subsidies grew to– gasoline (58%), diesel (60%) and kerosene (88%)

• Prices increased by 29% in March 2005 and planned increase of 30% in October

• Announced 114% increase in October resulting in subsidies– gasoline (20%), diesel (23%), kerosene (67%)– Subsidies projected to be 3.2% GDP2005 and 1.8% in 2006

• Introduced unprecedented cash transfer program delivered through Post Office– Coverage of 15.5 million poor families (60million persons)– Each family to receive Rp.300,000 every 3 months (around

US$30/mth)– Annual cost estimated at RS.20 trillion– Additional “incentive package” also introduced

Policy messages from PSIA

• Fuel subsidies are often substantial fiscal drain and badly targeted• So should be able to identify alternative approaches to social

protection that provide same or better protection at substantially lower fiscal cost

• Access to effective system for targeting expenditures can be a crucial component for promoting efficiency-enhancing structural reforms

• Important to announce reforms as part of a package where budgetary savings will be used to finance an effective safety net as well as expand social and infrastructure expenditures that benefit low- and middle-income households

• Gradual reduction of better targeted fuel subsidies should be seen only as short term measure are developed since revenue-poverty trade off is large and efficiency cost from inter-fuel substitution large

• PSIA provides very useful approach for highlighting shortcomings of fuel subsidies and providing insights into alternative approaches for protecting the poor and reallocating expenditures