Subsidy Reform in MENA: Recent Progress and Challenges Ahead

Giorgia Albertin IMF Resident Representative for Tunisia

September 9, 2014

Plan of presentation

I. Subsidies in MENA: Stylized Facts II. Investigating the Determinants of

Successful Subsidy Reform III. Recent Experiences of Subsidy Reform in

MENA IV. Policy Issues: Topics of Special Interest to

MENA

2

3

I. Subsidies in MENA: Stylized Facts

Subsidies in the MENa region

• Generalized price subsidies are a common feature in many developing and emerging markets economy → pervasive in the MENA region

• In the MENA countries subsidies are integral part of a “social contract” and have been used by governments to achieve several objectives:

- support real incomes and protect the poor by maintaing low prices on widely used-products (energy and food) → form of social protection, compensating for limited social security system;

- shielding the population for shocks caused by large increase of commodity prices, particularly in oil importing countries;

- support employment in the private sector by maintaining lower input prices (eg below market price electricity prices)

- wealth re-distribution in oil-exporters countries

4

MENA countries are the biggest subsidizers in the world in pre-tax terms…

5

Total pre-tax subsidies in MENA:

• Close to $240 billion in 2011–half of the total

– Oil exporters:$204 billion

– Oil importers: $33 billion

• 8.6 percent of regional GDP

• About 45 percent of world energy subsidies

• 22 percent of government revenue

• Subsidy costs have been rising since 2009

Advanced Economies$25.4 billion0.1% GDP

Central and Eastern Europe and

Commonwealth of Independent States

$72.1 billion 1.7% GDP

Emerging and Developing Asia

$102.3 billion 0.9% GDP

Latin America and Caribbean $36.2 billion 0.6% GDP

Middle East and North Africa $236.5 billion

8.6% GDP

Sub-Saharan Africa $19.3 billion 1.6% GDP

Total Pre-Tax Energy Subsidies by Region, 2011 1/$492 billion; 0.7% GDP

Source: Clements and others (2013).1/ Includes petroleum, electricity, natural gas, and coal subsidies.

Often, reflecting reluctance to adjust retail prices

-0.5

0.0

0.5

1.0

1.5

2.0

OM

N

SY

R

EG

Y

ALG

SA

U

TU

N

BH

R

IRA

Q

KW

T

LBY

MA

R

QA

T

UA

E

SD

N

YE

M

IRN

MR

T

DJI

JOR

LBN

MENA: Diesel Pass-through to Domestic PriceMar. 2009 - Dec. 2012

MENA Average

EU

US

Source: National authorities; IMF staff estimates.

0.0 0.5 1.0 1.5 2.0

MENA

South Asia

East Asia and Pacific

US

Latin America

Sub-Saharan Africa

Europe and Central Asia

EU

Average Diesel Price, end-December 2011(US$ per liter)

Source: GIZ and IMF.

Subsidies benefit mainly the upper income groups

7

0

10

20

30

40

50

ElectricityLPGDiesel/Gasoline

Share of Energy Subsidies Benefiting the Bottom Forty Percent of the Population 1/(Direct effect)

Sources: ESMAP (2005); IMF and World Bank reports; Salehi-Isfahani and others (2013); and IMF staff calculations.1/ Latest available data.

Even hand-out

0

10

20

30

40

50

60

Co

ok

ing

oil

Su

ga

r

Wh

ea

t

Loca

l ric

e

Wh

ea

t fl

ou

r

Su

ga

r

Ba

lad

i bre

ad

Oil

Ric

e

Su

ga

r

Tea

Wh

ea

t

Ira

n

Jord

an

Leb

an

on

Mauritania Morocco Egypt Bread

Share of Food Subsidies Benefiting the Bottom Forty Percent of the Population 1/(Direct effect)

Even hand-out

Sources: Diverse IMF and World Bank reports; and IMF staff calculations.1/ Latest available data.

Subsidies have high budgetary costs and crowd out priority public spending

DZABHR

DJI

EGYIRN

IRQ

JOR

KWTLBN

LBY

MRT

MAROMN

PAK

QAT

SAU

SYR

TUNARE

0

1

2

3

4

5

6

7

8

9

10

0.0 0.5 1.0 1.5 2.0

Publ

ic h

ealth

exp

endi

ture

(% o

f GDP

)

Diesel price (average, 2006-11, US$ per liter)

Public Health Expenditure and Diesel Price

Subsidies are a source of distortions leading to over-consumption..

DZA

BHR

EGYIRN JOR

LBN

MRT

MAROMN

PAK

QATSAU

TUNYEM

0

10

20

30

40

50

60

70

80

90

100

0 0.5 1 1.5 2 2.5High

ene

rgy-

inte

nsiv

e ex

port

s in

perc

ent o

fm

anuf

actu

red

expo

rts

Gasoline price (average, 2006-2011, US$ per liter)

2011 High Energy Intensity Exports and Gasoline Price1/

Sources: COMTRADE and IMF staff calculations.1/ Excludes petroleum production and refining.

10

II. Investigating the Determinants of Successful Subsidy Reform

Methodology

Combines a narrative-based country-case studies with quantitative analysis

Covers 25 episodes of food and fuel subsidy reforms in 15 low- and middle-income countries including seven MENA countries

For each reform episode, we completed a questionnaire articulated in 23 questions (in seven themes):

(i) time and scope of the reform (incidence study, pace of adjustment, breadth) (ii) ownership, communication and consultation (government commitment,

consensus building, information campaign) (iii) external factors (commodity price cycle, role of partners) (iv) macroeconomic cycle indicators (economic dowturn, inflationary pressures) (v) fiscal pressures ( fiscal consolidation strategy, fiscal pressure) (vi) socio-economic context of reform and mitigating measures (poverty, exiting

social safety net, additional mitigating measures) (vii) political context of reform (type of government)

11

Methodology

The outcome of the reforms episodes were assessed as failed/partially successful/successful based on three criteria:

(i) Qualitative review of country experience (achieved fiscal savings). (ii) Change in the price gap between domestic and international prices (iii) Magnitude of domestic price adjustment in local currency

The episode of subsidy reform was considered: (i) successful: if the initial price adjustment was followed by other price adjustments

leading to a reduction in the price gap thus reducing the fiscal costs ( 7 episodes) (ii) partially successful: prices were not adjusted beyond the initial adjustment or the price

adjustment was not enough to reduce the price gap (13 episodes) (iii) unsuccessful: the reform attempt was short-lived and the initial price adjustments was

reversed (5 episodes)

11

Subsidy reforms outcome

12

Table 3.1. Summary of Subsidy Reform Outcomes for Fuel and Food Products

Country Reform Episode Reform Outcome

Fuel Products

Bolivia 2010 FailedBrazil 1990s Partially successfulGhana 2001 FailedGhana 2003 Partially successfulGhana 2005 Partially successfulIndonesia 1998 FailedIndonesia 2005 Partially successfulIran 2010 Partially successfulJordan 2005 SuccessfulMauritania 2008 FailedMauritania 2011 SuccessfulNigeria 2012 Partially successfulPoland 1996 SuccessfulSenegal 1998 Partially successfulSyria 2008 Partially successfulYemen 1995 Partially successfulYemen 2005 Partially successfulYemen 2010 Partially successful

Food Products

Iran 2010 Partially successfulJordan 1990 Partially successfulMexico 1990 SuccessfulMorocco 1999 SuccessfulTunisia 1983 FailedTunisia 1991 SuccessfulYemen 1996 Successful

Source: IMF staff calculations.

Subsidy reforms: key success factors (I)

• We conducted quantitative analysis: i) Spearman correlations → correlation between each factor and the reform outcome; ii) success rate computation → how often a factor was present when the reform was successful

• The results of our quantitative analysis point to six key factors for success:

(i) Good reform preparation, gradual pace of adjustment, and breadth of reform;

(ii) Strong government leadership, consensus building and communication; (iii) Support from international partners, in particular technical assistance; (iv) Mitigating measures to soften the impact of the reform on the poor; (v) Favorable economic conditions, particularly higher economic growth, and

inclusion of subsidy reform in a broader consolidation effort; and (vi) Presence of a coalition government at the time of the reform.

13

Subsidy reforms: key success factors (II)

- Good reform preparation: accurate diagnostic study provides the basis for better policy discussion and targeting mitigation measures (86% success rate)

- Gradual pace of adjustment: gradualism (price increase of less than 50 percent

at the start of the reform) provides more time to adjust and the mitigating measures to take effect (75% success rate)

- Breadth of the reform: a more comprehensive reform is more likely to be successful triggering a “climate for reform” (76% success rate)

- Strong government leadership: commitment to advance in the reform(100% success rate)

- Consensus building: outreach to beneficiaries and key stakeholders, consultation with civil society (100% success rate)

- Communication strategy: underline the costs of subsidies and the benefits of the reform and costs of lack of reform to the wide public (100% success rate)

13

Subsidy reforms: key success factors (III)

- Role international partners: provide support to the national efforts and technical assistance in the analysis and design of mitigating measures (88% success rate)

- Mitigating measures: reforms more successful when measures to protect the poor and vulnerable groups: cash or in-kind transfers (100% success rate)

- Favorable economic conditions: (i) higher economic growth: public willing to accept the reform (100% success rate) (ii) higher inflation: public less willing to take the impact of the reform (50 % success rate

- Fiscal pressures: (i) fiscal pressures: high fiscal pressures increase the need for the reform (67% success

rate) (ii) fiscal consolidation: subsidy reform is more successful when included broader

strategy to reduce fiscal deficits (67% success rate)

- Presence of a coalition government at the time of the reform: more negotiation may be required but there is broader basis of support (75% success rate)

13

17

III. Recent Experiences of Subsidy Reform in MENA

Recent subsidies reform in MENA countries

• Since 2011, several MENA countries have initiated to reform energy subsidies: Egypt, Jordan, Mauritania, Morocco, Sudan, Tunisia and Yemen

• Renewed impulse to reform subsidies lies on two main reasons:

fiscal and external buffers were eroded by the increase in international oil prices and the increase in higher spending (wages and generalized subsidies) which followed the Arab Spring;

Policy makers and the public more aware of the shortcomings of subsidies

Focus of the reforms: energy (fuel products and electricity tariffs) while less on food subsidies due to much lower fiscal costs and much higher social sensitivity

17

19

Common triggers of recent reforms: large external and fiscal deficits and high debt ratios

-40

-30

-20

-10

0

10

20

30

40

Real GDP growth (percent)

Inflation rate (percent)

Fiscal balance Current account balance

Oil imports

Selected Macroeconomic Variables, 2012 1/(In percent of GDP, unless otherwise specified)

Max Min Median

1/ Include Egypt, Jordan, Mauritania, Morocco, Sudan, Tunisia, and Yemen.Source: IMF World Economic Outlook.

0

20

40

60

80

100

120

Egypt Jordan Mauritania Morocco Sudan Tunisia Yemen

Public Debt(In percent of GDP)

2009 2012

Tunisia: energy subsidy reform (I)

• Objective: Gradually phase out energy subsidies by increasing electricity tariffs and fuel prices and replace them by better targeted social safety nets

• Key measures undertaken: - Ad hoc-increases in fuel prices by 7-8 percent, on average, in September 2012 , in March

2013 and in July 2014;

- Energy subsidies to cement companies were gradually eliminated by increasing in two steps prices of electricity and natural gas for these companies (January and June 2014)

- Electricity tariffs were increased by 7 percent in September 2012 and March 2013; by further 20% in 2014 (two steps in January and May 2014) for medium and low voltage consumers

- An automatic gasoline price formula was adopted.

19

Tunisia: energy subsidy reform (II)

Mitigating measures were introduced to protect the most vulnerable from the impact of the energy subsidy reform leading to rising fuel and electricity prices:

• Creation of a new social housing program for needy families • Increase of income tax deduction for the poorest household. • Introduction of an additional lifeline electricity tariff for households consuming less than

100 kwh per month. • Increase in the SMIG and SMAG (covered 900,000 individuals) • Introduction of a target household program (expansion of PNAF from 220,000 to 250,000

at end 2014, increased amount received by each PNAF family by 10 dinars per month, one –off temporary assistance of 80 dinars to PNAF families and 50,000 poor families identified at the regional level, increase in school allowances)

• Develop unique social identifier to achieve a better targeting system for vulnerable households

20

Recent episodes of reforms in MENA countries

• Egypt: 2012–13: prices for gasoline increased by 112 percent, fuel oil for non– energy-intensive industries by 33 percent, and for energy-intensive industries by 50 percent in 2012. In January 2013: electricity prices to households increased by 16 percent on average, natural gas and fuel oil prices or electricity generation rose by one-third.

• Jordan: June 2012: electricity tariffs increased for selected sectors (banks,telecommunications,

hotels, mining) and large domestic corporations and households. November 2012: elimination of fuel subsidies. January 2013: monthly fuel price adjustment mechanism resumed. August 2013 and January 2014: electricity tariffs increased by 7.5–15 percent for selected consumers.

• Mauritania May 2012: new automatic diesel price formula introduced, bringing domestic fuel

prices up to international levels. January 2012: electricity tariffs increased for the service sector.

21

Recent episodes of reforms in the MENA countries

• Morocco: June 2012: diesel prices increased by 14 percent, gasoline by 20 percent, and industrial

fuel by 27 percent. September 2013: diesel prices increased by 8.5 percent, gasoline by 4.8 percent, and fuel by 14.2 percent. January 2014: gasoline and industrial fuel subsidies eliminated, their prices are reviewed twice a month. February 2014: the per-unit subsidy on diesel was reduced, with additional quarterly reductions announced.

• Sudan: June 2012: gasoline, diesel, and liquefied petroleum gas prices increased by 47 percent, 23 percent, and 15 percent, respectively; jet fuel liberalized. September 2013: diesel prices increased by 74.7 percent, gasoline by 68.0 percent, and liquefied petroleum gas by 66.7 percent.

• Yemen: 2011–12: gasoline prices increased by 66 percent and diesel and kerosene prices

doubled. In 2013: diesel price unified across users, including the electricity sector.

22

Progress made but reforms differed across countries

24

Implementation Status of Most Recent Subsidy Reforms in MENA Based on Key Factors for Success

PreparationGradual Pace of Adjustment

Breadth of Reform

Consensus Building and

Communications Strategy

Role of Partners

Mitigating Measures

Egypt Jordan Mauritania Morocco Sudan Tunisia Yemen Source: IMF staff reports for Article IV Consultations.

• Fiscal deterioration brought subsidy reform on the policy agenda

• Iran only major oil exporter to undertake comprehensive reform

• Dubai (electricity tariffs, water), Qatar (gasoline), Bahrain (gas) undertook limited reform

• In oil exporters with smaller reserves—Algeria, Bahrain, and Oman—the urgency to reform is higher

25

Limited progress in subsidy reform in oil exporters

0 50 100 150 200 250

KWT

LBY

QAT

SAU

UAE

OMN

IRQ

BHR

DZA

IRN

YMN

Figure 20. Fiscal Break-Even Oil Prices(In dollars per barrel)

2009 2012

Average petroleum spot oil price

Source: MCD Regional Economic Outlook.

Challenges to reform in oil exporters

Low fuel prices are considered: - rights of citizenship

- key basis for legitimacy, substituting for political participation

Large subsidy programs are more difficult to reform

Subsidy reform seen as potential source of unrest

In some countries, political and security volatility makes

it more difficult to remove subsidies

Challenges

26

The way forward

27

Countries that started reform should: Complete scaling up of

targeted social safety nets Set a timeline to raise

domestic prices gradually Implement more rigorously

automatic price setting mechanisms Tackle energy subsidies and

restructure the sector

Countries that are not ready to reform can: Assess the advantages of

containing subsidies Improve transparency on

subsidy costs and beneficiaries Strengthen administrative

capacity to develop and manage SSNs

28

IV. Policy Issues: Topics of Special Interest to MENA

Addressing macroeconomic impact of subsidy reform

Variables Impact Recommended Policy Response

CPI (first-round effect) Increase in inflation Accommodative monetary policy

CPI (second-round effect)

Increase in core inflation May need tighter monetary policy

Growth (short run) Negative Reduction in spending Higher input cost for energy-intensive industries

Redirect budgetary savings to social transfers and investment

Growth (long run) Positive Elimination of distortions Improved competitiveness

29

Mitigating impact on productive sector

• Removal of subsidies is a shock to the productive sector, in particular in energy intensive sectors→ increase in the costs of inputs, which reduces profits if it cannot passed-through to consumer prices

• In MENA countries, due to large energy subsidies, the economy is more energy intense and subsidies reform will have a negative impact on the productive sector

• Mitigating impact on productive sector: • Improve competitiveness:

Business climate Corporate restructuring

•Train workers exiting non-competitive industrial sector

•Develop energy-saving technology –Gradual increase in prices to allow firms to adjust

Strengthening social safety nets

31

Social safety nets (excluding subsidies) •Cash transfers not widely used or underfunded •Small at 0.7% of GDP •SSN fragmented and suffering from leakages Only 23% of benefits accrue to bottom quintile

•Limited impact on poverty and inequality because not-well targeted and poor monitoring systems

Strengthening social safety nets: •Improve targeting to the neediest

Even targeted, subsidies less cost effective than other SSNs

•Cash transfers are more effective

Do not distort prices Provide flexibility to recipients Relatively low administrative cost

Managing the political economy of subsidy reform

• Political economy factors play a key role in subsidy reforms and could derail the reform implementation process due to resistance in the society:

opposition to the reform can build-up due the lag between short-tem costs

(increase in prices affecting negatively consumers and industries) and the long-term growth-enhancing benefits of subsidy reforms

small groups of losers from the reform tend to organize and generate pressure on government to abandon the reform. On the other hand, the beneficiaries of the reform, typically poorer households, tend not be organized.

lack of trust in the government’s capacity to implement compensatory measures, i.e. better-targeted social safety nets, may hinder public support for the reform.

Long duration of the reform effort: time to solidify opposition and exposes to to cyclical developments which weaken support (growth cycle, electoral cycle)

Managing the political economy of subsidy reform

These political economy factors need to be taken into account through an appropriate design of the subsidy reform design:

Build an objective case for reform: provide evidence of the benefits of the reform to sha[e public perception, e.g. impact studies of price subsidies Increase transparency and implementation capacity of the state: increase trnasprency on how public resources are used; earmark subsidy savings Create public concern through a communication campaign: extensive campaign on the costs of subsidies and befits from reform Create coalitions with those who will benefit from reforms Embed subsidy reform in a wider policy reform plan: eg if an health system reform savings from energy subsides could be used to fund them

•

Conclusions: recent subsidy reform in MENA—good progress but challenges ahead

34

Challenges ahead Historically part of social compact Potential source of unrest Beneficiaries resist losing benefit Difficult under political and security instability Inflationary effect and possible negative impact on economic activity

Good progress achieved more recently in seven countries Fiscal and external pressures were the common factor for reform Reform was comprehensive (fuel products and electricity tariffs) Mitigating measures often accompanied reform Few countries adopted automatic price mechanism (Jordan,

Mauritania, Morocco, and Tunisia)

Annex

35

Reform Themes and Spearman Rank Correlations

36

Correlation with the Reform Outcome 1/

Time and scope of the reformPreparation Did the authorities conduct studies of the

incidence of subsidy removal before designing/introducing the reform?

=1 if an assessment of the incidence of subsidy removal has been conducted before the reform; =0 otherwise

0.4438**

Pace of adjustment Pace of adjustment=2 if the price adjustment at the start of the reform is higher than 50 percent; =0 if lower than 20 percent;=1 otherwise -0.3727*

Breadth of the reform How comprehensive was the reform? =0 if reform was targeted only at a subset of products; =1 otherwise (i.e., comprehensive reform of most/all products)

0.4702**

Government leadership How strong was the ownership and commitment of the government to the reform?

=1 if the government was strongly and publicly committed to reform; =0 otherwise. Possible indications include the president or prime minister speech announcing the reform

0.3683*

External factorsRole of partners Was the reform undertaken with technical

assistance from development partners (IMF or other) and/or in the context of an IMF arrangement?

=1 if technical assistance was provided at the time of reform; = 0 otherwise 0.613**

Macroeconomic cycle indicatorsEconomic downturn Was the reform undertaken in the context of an

economic crisis?=1 if the growth rate in the year preceding the reform implementation is one std deviation below the average growth rate over the previous 5-years; =0 otherwise

-0.4869**

Mitigating measures What mitigating measures did the authorities put in place or expand, if any?

=1 if wage increase was granted; =0 otherwise -0.0761

=1 if cash and in-kind transfers were put in place at the time of the reform; =0 otherwise

0.559**

=1 if subsidies were targeted at a specific group; =0 otherwise 0.4175**=1 if other programs/social safety nets were put in place at the time of the reform; =0 otherwise

0.6866**

Source: IMF staff calculations.1/ *Significant at the 10 percent confidence level; **significant at the 5 percent confidence level.

Ownership, communication and consultation

Socio-economic context of reform and mitigating measures

Success Rates for Selected Questions ( in percent)

37

Yes No

Preparation** Was an incidence of subsidy removal study conducted prior to the reform?

86 20

Pace of adjustment Was the price adjustment higher than 50 percent? 33 75

Breadth of the reform** Was this a comprehensive reform of most/all products? 78 0

Government leadership**

Was there a strong ownership and commitment of the government to the reform?

100 33

Consensus building* Did the government tried to build a consensus to win adoption of the reform?

100 43

Communication strategy**Did the government undertake an information campaign to support the reform?

100 50

External factorsRole of partners** Was the reform undertaken with technical assistance from

development partners?88 0

Economic conditions** Was the reform undertaken in the context of slower growth/economic contraction?

0 78

Mitigating measures Was wage increase granted as a mitigating measure? 50 67

** Were cash and in-kind transfers granted as a mitigating measure? 100 17* Were subsidies targeted at a specific group? 100 38** Were there other programs/social safety nets? 100 0

Did the mitigating measures target/benefit the middle class? 33 67

Government composition Was the government a fragmented coalition? 75 50

Government control of parliament

Was there a single party majority in the parliament? 67 33

Source: IMF staff calculations.

Time and scope of the reform

Macroeconomic cycle indicators

Political context of reform

2/ Higher scores are associated with better outcomes. (*) means that the difference in success rates is significant at the 10 percent confidence level; (**) means that the difference is significant at the 5 percent confidence level.

1/Success rate are calculated using only failed and successful reform episodes.

Ownership, communication and consultation

Socio-economic context of reform and mitigating measures