REPUBLIC OF MADAGASCAR

SOCIAL SPENDING AND OUTCOMES IN

MADAGASCAR 1

Education, health, and social assistance spending in Madagascar is among the lowest worldwide and
social outcomes such as education quality, malnutrition, basic immunization coverage, and poverty
have deteriorated over the last decade. In a context of social fragility and vulnerability to exogenous
shocks, Madagascar faces significant constraints to execute and deliver social spending and services.
Looking ahead, sustained efforts are needed to implement the authorities’ development agenda in the
“Plan Emergence Madagascar” and support the needed investments in human capital. This includes
creating fiscal space for higher social spending combined with institutional reforms to ensure a more
efficient use of resources.

A. Introduction

1. Madagascar experienced some improvement in development outcomes over 2000–

2010, but the overall situation remains challenging following two decades of income
stagnation and recent pandemic and climate shocks. The authorities’ Plan Emergence (PEM) aims
to increase GDP per capita to US$4,000 in 2040 (compared to US$522 in 2019) and the human
capital index to 0.60 (0.39 in 2020), while at the same time bringing the poverty rate down from
81 percent to 35 percent by 2040. While ambitious strategies have been developed in social sectors,
a disconnect between announced ambitions and limited financial and human resources have led to
slow reform implementation. Two years of pandemic, a series of climate shocks, and a more
challenging external environment have compounded Madagascar’s existing deep-rooted fragilities.

2. This paper takes stock of developments in education, health, and social assistance and
offers policy options. It examines social spending 2 and outcomes in Madagascar. In particular, it
addresses the following questions: (i) How large has social spending been over time and compared
to peers? (ii) How does Madagascar perform on various socioeconomic outcomes? (iii) What are
policy options and how can outcomes be improved?

B. Education

3. Government spending on education is relatively low compared with other Sub-

Saharan African (SSA) countries and low-income peers. Education spending increased albeit at
a slow pace in recent years (from 2.3 percent of GDP in 2011–15 to 2.6 percent of GDP in 2016–20

1 Prepared by Samah Mazraani. The analysis benefitted from helpful comments from the World Food Program (WFP)

and the European Union (EU).

2 Social spending throughout the paper is defined using the budget functional classification. The definition is

consistent with the Classification of the Functions of Government (COFOG) and the IMF Statistics Department
Government Finance Statistics Manual (GFSM 2014). It differs from the definition of the social spending target in the
ECF program which includes domestically financed spending by four social ministries (education, health, water, and
population/social protection) excluding wages.

4 INTERNATIONAL MONETARY FUND

©International Monetary Fund. Not for Redistribution

 REPUBLIC OF MADAGASCAR

on average) but remains low compared with other low-income and SSA countries. While education
spending as a percent of total government spending is higher than in peers (Figure 2, left), this
reflects lower levels of government spending in Madagascar due to still limited tax revenue.
Moreover, education spending per student significantly lags peers both at the primary and
secondary levels (Figure 2, right). The teacher-student ratio is somewhat similar to peer groups with
Madagascar having about 40 primary and 20 secondary school students per teacher in 2018.

4. The quality of education in Madagascar is falling with low school completion rates,
a high share of untrained teachers, and
declining test scores. While Madagascar Figure 1. Education Spending, Central
performs well in adult literacy rate (77 percent Government
in 2021) and net enrollment rate for primary (Percent of GDP)
school (96 percent in 2018), the net enrollment 4.5
4.0
rate in secondary school is still low at 3.5
29.8 percent in 2018 compared to 44 percent 3.0

in low-income countries (Figure 3). Moreover, 2.5

2.0
school completion rates have been on 1.5

a declining trend over the last decade reaching 1.0

0.5
levels below SSA and low-income countries 0.0
(63 percent for primary school and 35 percent
2004-07
2008-10
2011-15
2016-20

2004-07
2008-10
2011-15
2016-20

2004-07
2008-10
2011-15
2016-20

2004-07
2008-10
2011-15
2016-20
for lower secondary school). In addition, the
strategy to use low-paid community-hired
EMs LICs SSA Madagascar
Education
teachers supported by parents’ associations,
the “Maîtres FRAM” (now representing about Sources: IMF, Statistics Department

Figure 2. Government Education Spending

Government Education Expenditure, Government Education Expenditure

Latest Value Available per Student, PPP$ adjusted, Latest
25
1,200
Madagascar
Madagasca
20 SSA 1,000 r

15 800

10 600

5 400

0 200
% government Primary Secondary
expenditure Teacher student ratio, per 0
100 students Primary Secondary

Source: IMF FAD Expenditure Assessment Tool (EAT), World Bank.

Note: Latest data available for Madagascar is 2018 (left panel) and 2012 (right panel).

80 percent of teachers in primary schools) may have had a negative impact on education quality, as
the majority lack formal credentials or teacher training (Figure 4). Harmonized test scores have also

INTERNATIONAL MONETARY FUND 5

©International Monetary Fund. Not for Redistribution

REPUBLIC OF MADAGASCAR

fallen significantly since 2010 3 (Figure 4). As a result, 97 percent of 10-year-old children are not able
to read and understand a simple text in French (World Bank, 2022b) 4.

5. Raising outcomes in education

Figure 3. Education Outcomes
requires a concerted effort to allocate
more resources to the sector, ensure Education Indicators, Latest Value Available
transparent and merit-based teacher 100

recruitment mechanisms, and strengthen Madagascar

80 SSA
teacher training and incentives. The
government’s policy initiated in October 60

2020 of canceling school enrollment fees is 40

welcome and would facilitate universal

20
access to school education. It is important
to prioritize education spending by 0
Adult literacy rate Net En rollment, Net En rollment,
allocating more resources to the sector 5 primary secondary

and undertaking public financial Source: IMF FAD Expenditure Assessment Tool (EAT), World Bank.
management (PFM) reforms to improve
budget execution. The preparation of
annual expenditure commitment plans by social ministries in 2023, in line with sectoral strategies,
together with the streamlined spending commitment process (both commitments under the ECF
program) should help in this regard. Improving public investment prioritization of projects
accompanied with appropriate costing of operational and maintenance costs (e.g., of school
buildings) would contribute to greater resource efficiency in the sector. Additional
recommendations include:

• Given limited fiscal space in Madagascar, integrating all FRAM teachers into the civil service
is not feasible in the short term. An alternative approach is to gradually integrate them
following a phased approach over several years (World Bank 2020, 2022a, 2022b).

• This should be done in a transparent and merit-based way following a competitive

recruitment process to award contracts based on qualifications and competency tests—
conducted for example by an independent agency for all civil service teachers (World Bank
2020, 2022a, 2022b).

• Undertake a biometric census of all civil servants including FRAM teachers and volunteer
health personnel and utilize data results to verify the quality of the AUGURE database—an IT

3Student learning outcomes have multi-sectoral drivers which could include in addition to teacher quality, factors
such as malnutrition, student illness, general learning environment, etc.
4A more recent study by the Ministry of Education and the World Bank suggests an improvement in learning
outcomes between 2016 and 2021 despite the effects of the COVID-19 pandemic. The final results of the study
should be released soon.
5While the government is committed to allocate at least 20 percent of the national budget to the education sector,
actual budget allocations and realizations have been lower, ranging from 12–16 percent during 2021–23.

6 INTERNATIONAL MONETARY FUND

©International Monetary Fund. Not for Redistribution

 REPUBLIC OF MADAGASCAR

tool for civil service personnel management—including to eliminate any “ghost teachers”.
(Public Expenditure Tracking Survey, 2021).

• Ensure the timely release of funds to public school teachers 6 and to community teachers
(World Bank 2020, 2022a, 2022b).

• Increase decentralized management in line with the decentralization policy 7 by allocating

additional resources to schools (through school grants or transfers to “Caisse Ecole”) while
improving transfers’ timeliness (World Bank 2020, 2022a, 2022b).

Figure 4. Education Outcomes

Primary Completion Rate Lower-Secondary Completion

(% of relevant age group) Rate
(% of relevant age group)
80 50

70 45

40
60
35
Percent

Percent

50 30

40 Madagascar 25 Madagascar
Low in com e 20 Low in com e
30 Sub-Saharan Africa Sub-Saharan Africa
15
20 10
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020

2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
Trained teachers, primary Harmonized test scores
(% of total teachers) (300 = Minimum, 625 = Maximum)
80 460

70 440

60 420

50 Madagascar 400
Percent

40 Low in com e 380

SSA
30 360

20 340
Madagascar
10 320 SSA
0 300
2012

2013

2014

2015

2016

2017

2018

2019

2010

2017

2018

2020

Source: World Bank.

Note: Harmonized test scores from major international student achievement test programs. They are
measured in TIMMS-equivalent units, where 300 is minimal attainment and 625 is advanced attainment.
Most recent estimates are used.

6 According to World Bank (2022b), 40 percent of public teachers and 80 percent of FRAM teachers reported delays
in receiving their salary more than once in the last two years, sometimes reaching several months.
7The government is shifting gradually towards management by school grants (“Caisse Ecole”). These school grants
combined with school feeding transfers amounted to MGA 39 billion only in 2021 (0.07 percent of GDP).

INTERNATIONAL MONETARY FUND 7

©International Monetary Fund. Not for Redistribution

REPUBLIC OF MADAGASCAR

C. Health

6. Government spending and total spending on health are low compared with low-
income and SSA countries. Government
health spending decreased in recent years Figure 5. Health Spending, Central
(from 1.3 percent of GDP in 2016 to Government
1.0 percent of GDP in 2020) and remains low (Percent of GDP)
compared with LICs (1.8 percent of GDP over 3.0

2016-20), EMs (2.5 percent of GDP over 2016– 2.5

20), and SSAs (1.9 percent of GDP over 2016– 2.0

20) (Figure 5). Similarly to education spending, 1.5

government health spending as a percent of 1.0

total government spending is higher than in 0.5

peers (18.6 in Madagascar compared to 10.2 in 0.0

2004-07
2008-10
2011-15
2016-20

2004-07
2008-10
2011-15
2016-20

2004-07
2008-10
2011-15
2016-20

2004-07
2008-10
2011-15
2016-20
LICs), reflecting lower levels of government
spending. However, total health spending per EMs LICs SSA Madagascar

capita (both government and private) is about

Health

US$ 78 per person per year (on a PPP basis), Sources: IMF, Statistics Department
less than half the average in low-income
countries of $202 per person per year
(Figure 6).
Figure 6. Health Spending, Total
7. The public health system consists of 350

four types of health facilities with limited Madagascar 300

SSA
autonomy (World Bank PER, 2014). These LIDCs 250

include basic health centers for primary care

Value US$
200
implemented within communes, referral 150
hospitals within districts, referral and university
100
hospitals within regions, and specialized
50
centers at the regional and central levels.
0
Private and community-based health insurance Total health expenditure per Total health expenditure per
remains limited due to low incomes and a capita, current US$ (rhs) capita, PPP$-adjusted (rhs)

small formal employment base. Free services

Sources: IMF FAD Expenditure Assessment Tool (EAT),
and vouchers are limited to specific programs World Bank, World Health Organization.
and regions.

8. Madagascar has made progress on some key health indicators, but significant
challenges remain in malnutrition, immunization, and service delivery. According to World
Bank Development Indicators, Madagascar’s performance improved over the last decade on life
expectancy (67 years in 2020 compared to 62 in SSA), maternal mortality rates (335 deaths per
100,000 live births in 2017 compared to 534 in SSA), and infant mortality rates (36 deaths per 1,000
live births in 2020 compared to 50 in SSA). However, the prevalence of malnutrition has increased
significantly (from 28 percent of the total population in 2010 to 49 percent in 2020). Likewise, the
prevalence of stunting among children under five remains one of the highest in

8 INTERNATIONAL MONETARY FUND

©International Monetary Fund. Not for Redistribution

 REPUBLIC OF MADAGASCAR

the world (at 39.8 percent). Challenges remain on basic immunization coverage (Diphtheria, Tetanus,
Pertussis) which remains low at 55 percent among 12–23 months aged children in 2021 according to
World Bank indicators. Finally, the health system suffers from severe human resource shortages
(physicians, nurses, and midwives) and other resource shortages (e.g., hospital beds) compared to
other SSA and LIC countries (Figure 7).

9. Improving health outcomes will require the mobilization of additional domestic

resources and large-scale reforms. With government health spending at only 1 percent of GDP,
the health sector is clearly under-financed and under-staffed. Additional resources are needed—
most notably in primary healthcare services—
favoring the retention of qualified and Figure 7. Health Outcomes
motivated health workers while ensuring a
70 1.6
better distribution in rural areas (most notably
through strengthening decentralization and 60 1.4

increasing allocations to basic health centers). 50

1.2

PFM reforms are also needed to improve 1

budget execution and ensure sound public 40
Madagascar 0.8
investment management in the sector. The 30 SSA
LIDCs
government’s national social policy (adopted 0.6

20
in 2015) and national social protection strategy 0.4

(2019–23) outline the goal to attain universal 10 0.2

health coverage (UHC) with a contributory
0 0
system and free healthcare for the poorest Number of Life expectancy Hospital beds, Nurses and Physicians, per
infant deaths, at birth per 1,000 midwives, per 1,000 people
households 8. However, the implementation per 1,000 people (rhs) 1,000 people (rhs)
people (rhs)
plan for the UHC strategy was never finalized. Sources: IMF FAD Expenditure Assessment Tool (EAT),
Therefore, a clear financing strategy is needed World Bank, World Health Organization.
to achieve the objective of UHC accompanied
with clear identification criteria of the poorest and most vulnerable households who will be eligible
for free healthcare—based on the social registry currently in development (World Bank, 2022a).

D. Social Assistance

10. Social assistance 9 spending in Madagascar remains among the lowest in the world
with limited coverage of the vulnerable population. Spending on social safety net programs (i.e.,
excluding social insurance such as pensions) averaged 0.2 percent of GDP over 2011–20 (according
to ASPIRE database). This is significantly below spending levels in peers (0.9 percent of GDP median
in SSA and 0.8 median in LIDCs, Figure 8). The government’s strategy document (2019–23) aims to

8A decree was passed in 2017 establishing the “Caisse Nationale de Solidarité en Santé (CNSS)” or national health
solidarity fund, a public entity with administrative and financial autonomy charged with collecting and managing
UHC contributions and paying benefits to health institutions. The CNSS has recently been abrogated.
9 Following the WB Aspire program classification, social assistance programs are defined as non-contributory

transfers in cash or in-kind. Social assistance spending therefore includes cash transfers (conditional and
unconditional), school feeding programs, targeted food assistance, and near cash benefits such as fee waivers and
food vouchers.

INTERNATIONAL MONETARY FUND 9

©International Monetary Fund. Not for Redistribution

REPUBLIC OF MADAGASCAR

increase social safety net coverage of extremely poor households to 15 percent by 2023 and
50 percent by 2030. However, coverage of social safety nets remains very limited (5 percent of the
population compared to 20 percent in peers, Figure 9).

11. The social safety net system

consists of two regular cash transfer Figure 8. Social Assistance Spending
programs mostly funded by donors: Total Social Assistance Spending (percent of GDP),
Total Social Assistance Spending (percent of
MostGDP),
RecentMost
Available Year
Recent Available Year
1.6 Total S
• A conditional cash transfer program Assista
1.4
(TMDH “Transfert Monétaire pour le
Spendi
1.2 (perce

Spending (% GDP)
Développement Humain”), providing 1.0
GDP)

regular cash transfers for families 0.8

with children under the age of 12 0.6

and conditional on primary school 0.4

0.2
attendance with a UNICEF-funded
0.0
top up for children transitioning to

(median)

(median)
Rwanda

Sierra Leone
Madagascar

Tanzania

LIDC
SSA
secondary school (LUL “Let us
Learn”). TMDH also includes
beneficiaries from the program
Sources: IMF FAD Social Protection & Labor - Assessment
“Fiavota”, an emergency cash Tool (SPL-AT), WB ASPIRE database.
transfer program designed to assist
families severely affected by drought
in the South of the country.
Figure 9. Social Assistance Coverage
Coverage for All Social Protection and Labor,
• A productive safety net program Coverage for All Social Protection and
Most Recent Available Year
Labor, Most Recent Available Year
(ACTP - Asa Avotra Mirindra “Argent 25
Coverage (% population)

contre Travail Productif”) providing 20

cash for work opportunities over a 15

minimum of three years for workers 10

assessed as poor in select districts 5

with an unconditional transfer 0

All Social All Labor Market All Social
component in favor of vulnerable Protection and Assistance
Labor
persons who are unable to work due Madagascar (median) SSA (median) LIDC
to disabilities.
Sources: IMF FAD Social Protection & Labor - Assessment Tool
These programs are supported by the World (SPL-AT), WB ASPIRE database.
Bank and implemented by the “Fonds
d’Intervention pour le Développment (FID)”. Combined, they currently cover about 309.000
households in extreme poverty mostly in rural areas (89 percent financed by donors). However,
according to UNICEF, geographical coverage of these programs is limited (only operating in 7 out of
22 regions), and coverage is rationed in the beneficiary regions through proxy means testing and
community verification (reaching only 30 percent of households in these regions despite pervasive
poverty).

10 INTERNATIONAL MONETARY FUND

©International Monetary Fund. Not for Redistribution

 REPUBLIC OF MADAGASCAR

12. In addition, there are three types of shock-responsive social protection programs, two
of which were developed in response to the COVID-19 pandemic:

• An unconditional cash transfer program (TVA “Toseke Vonje Aigne”) launched in 2018
following the drought caused by El Nino in the South. This program includes specific triggers
to cover victims of areas affected by poor harvests, cyclones, and droughts in the South.
Coverage varies from year to year, but the total number of beneficiaries reached about
108,000 households receiving MGA 80.000 per month (around 20 U.S. dollars) for 5 months.

• An unconditional cash transfer program (Tosika Fameno) implemented following the

COVID-19 pandemic in coordination with donors to support vulnerable households in three
urban regions mostly affected by the negative effects of the national lockdown. Coverage
reached around 215,000 households receiving MGA 100.000 per month (around 25 U.S.
dollars) for two months (May and July 2020). Enrollment was done through self-registration
using an online survey at the community level and payments were done in cash (through
payment agencies) or through mobile money accounts.

• A domestically financed presidential project providing in-kind donations (Vatsy Tsinjo) to the
most vulnerable and to those whose activities were affected by COVID confinement in three
regions. It is estimated that 305,000 food packs were distributed (out of an initial objective
of 500.000 packs) with each household receiving two packs for two months (April and July
2020). While the strategy initially targeted the homeless, elderly, and those affected by
confinement, it was extended to cover university students, artists, persons with disabilities,
and public school teachers, among others.

13. Madagascar’s poverty rate worsened following the COVID-19 pandemic and is
projected to remain close to 80 percent over the next three years. The poverty rate based on
the international definition (percent of the population living on less than US$1.90/ day in 2011 PPP)
was estimated at around 78 percent in 2012, compared to 36 percent in LIDCs and 39 percent in
SSAs (Figure 10). World Bank (2022a) estimates that Madagascar has not been able to reduce
poverty over the last decade with the poverty rate now estimated at a record high of 81 percent in
2020. Furthermore, the poverty gap (a measure of the intensity of poverty given by the difference
between the poverty line and the mean income of the poor in percent of the poverty line) is around
40 percent, compared to 13 percent only in SSAs. As for inequality, the latest measures date back to
2012 10 and show that inequality is close to the average of SSA countries, with a GINI coefficient of
42.6.

10According to World Bank (2020), there is a significant risk that the COVID-19 crisis heightened existing inequalities,
given the projected increase in the extreme poverty rate in 2020.

INTERNATIONAL MONETARY FUND 11

©International Monetary Fund. Not for Redistribution

REPUBLIC OF MADAGASCAR

Figure 10. Social Protection Outcomes

Poverty Context: Madagascar and Comparators, Most Poverty Context: Madagascar and Comparators, Most
Recent Available Year Recent Available Year

(median) LIDC (median) LIDC

(median) SSA (median) SSA

Sierra Leone Sierra Leone

Rwanda Rwanda

Tanzania Tanzania

Madagascar Madagascar

0.0 20.0 40.0 60.0 80.0 100.0 0.0 10.0 20.0 30.0 40.0 50.0
Poverty Headcount (%) $1.90/day Poverty Gap (%) $1.90/day

Inequality/Distributional Indicators Context: Madagascar

Poverty Rate in Madagascar and Comparators
and Comparators, Most Recent Available Year p

(median) LIDC

(median) SSA

Sierra Leone

Rwanda

Tanzania

Madagascar

0.0 10.0 20.0 30.0 40.0 50.0

FAD Gini Index (0-100)
Madagascar
00 20

Sources: IMF FAD Social Protection & Labor - Assessment Tool (SPL-AT), WB ASPIRE, PovcalNet (World Bank), World
Bank, ILO.

14. Strengthening the social protection system and expanding existing social safety nets
should be a key policy priority in order to help reduce poverty and protect the most
vulnerable. Key areas for policy action include 11:

• Update the social protection strategy. The government adopted its National Social
Protection Policy (PNPS) in 2015 with a vision to ensure coverage of half of the vulnerable
population by 2030. The National Social Protection Strategy (SNPS) for 2019–23 in turn
states the objective to cover 15 percent of the population in extreme poverty by 2023. In
order to reach the PNPS objective, it will be important to update the medium-term strategy
including to link stated objectives with a clear funding strategy.

11These areas are in line with the Cour des Comptes audit report on social measures taken in 2020 to combat the
effects of the COVID-19 pandemic.

12 INTERNATIONAL MONETARY FUND

©International Monetary Fund. Not for Redistribution

 REPUBLIC OF MADAGASCAR

• Develop a national social registry. While the PNPS envisaged the creation of an inventory
of all social programs and a unique register of beneficiaries with a view to improve planning
and coordination of interventions, they are still under development and not yet operational.
Scaling up the existing register of beneficiaries to a national social registry 12 (starting for
example with a pilot phase with limited geographical coverage) will be key to facilitate
expansion of transfers and to respond rapidly to future crises. The social registry should be
managed and regularly updated over time by the Ministry of Population and serve as a basis
to identify potential beneficiaries of social programs, ensure a coordinated and harmonized
response to social needs, and avoid fragmentation of efforts by various donors.

• Ensure predictable and sufficient budget allocations and scale up social protection
programs sustainably. Social assistance spending (i.e., cash and in-kind transfers) is
extremely low and remains mostly financed by external donors. Given the poverty level and
immense needs, it is important to find fiscal space (through spending reallocation and/or
revenue mobilization) with a view to gradually expand coverage of existing social programs
and increase households’ resilience to future shocks. The successful experience gained from
the program “Tosika Fameno”, which was designed and implemented in just a few weeks,
could be used to scale up existing social programs.

E. Simulations

15. Four illustrative scenarios are considered to estimate the total budgetary cost of
raising the coverage of existing social assistance programs (Figure 11). In all four scenarios, it is
assumed that the transfer increases every year with average inflation and that administrative costs
are about five percent of the transfer size. Specifically:

• Scenario A assumes an average transfer of 50.000 MGA per household per month, with
coverage of the poor population immediately reaching 15 percent in 2023 (in line with the
authorities’ social protection strategy targets).

• Scenario B assumes an average transfer of 80.000 MGA per household per month, with
coverage of the poor population immediately reaching 15 percent in 2023.

• Scenario C assumes an average transfer of 50.000 MGA per household per month, with
coverage of the poor population only gradually reaching 15 percent by 2027 and continuing
to increase to 20 percent by 2030.

12 While a register of beneficiaries is a static database/list of existing beneficiaries of specific social programs, a social
registry is not just a database, but a full information system including potential beneficiaries (eligible or not). In
particular, it is a dynamic system that evolves over time and supports registration of applicants, determines potential
eligibility, stores, verifies, updates, and validates data. Scaling up from a register of beneficiaries to a social registry
would require pre-conditions such as establishing a unique identifier (e.g., biometric ID) and comprehensive surveys
to collect socioeconomic indicators allowing the calculation of a vulnerability index.

INTERNATIONAL MONETARY FUND 13

©International Monetary Fund. Not for Redistribution

REPUBLIC OF MADAGASCAR

• Scenario D assumes an average transfer of 80.000 MGA per household per month, with
coverage of the poor population only gradually reaching 15 percent by 2027 and continuing
to increase to 20 percent by 2030.

Figure 11. Increasing Social Assistance Coverage, Simulations 2022–30

Budgetary Cost Scenario A (50.000 MGA/m, 15 percent Coverage of Population in Poverty

(% of GDP)
immediate) Living on US$1.90/day (%)
Scenario B (80.000MGA/m, 15 percent Scenario A+ B (15 percent immediate)
1.6 25
immediate)
Scenario C + D (15 percent by 2027)
1.4
20
1.2
1.0
15
0.8
0.6 10

0.4
5
0.2
0.0 0
2022

2023

2024

2025

2026

2027

2028

2029

2030

2022

2023

2024

2025

2026

2027

2028

2029

2030
Sources: IMF Staff Estimates.

16. Significant budget allocations are needed to reach the authorities’ objective under all
four scenarios. The budgetary cost of social assistance spending would need to increase from the
current 0.1–0.2 percent of GDP to 0.7–1.5 percent of GDP by 2030 13. Under Scenario C, a gradual
increase in coverage accompanied by a moderate transfer amount (50.000 MGA per household per
month) would result in feasible budget increases of 0.1 percent of GDP per year, to reach the SSA
spending average of 0.9 percent of GDP by 2030.

17. Increasing education, health, and social assistance spending gradually to SSA levels by
2030 would require annual budget increases of 0.3 percent of GDP every year (Figure 12).
Specifically, government education spending is assumed to increase by 0.2 percent of GDP every
year (from 2.6 to reach SSA average of 3.8 by 2030). Similarly, government health spending is
assumed to increase by 0.1 percent of GDP every year (from 1.0 to reach SSA average of 1.9 by
2030). Social assistance spending is assumed to increase by 0.1 percent of GDP every year consistent
with Scenario C above. In total, social spending needs to increase by 0.5 percent of GDP in 2023
then by 0.3 percent of GDP every year during 2024–30 to reach SSA spending levels and cover
15 percent of the vulnerable population by 2027.

13 In all four scenarios, there is a steep increase in 2023 relative to 2022 due to the increase in the average transfer

amount per household in 2023. The average transfer amount for all current social transfer programs is not known for
2022 but is assumed to be 22.500 MGA per household for the purpose of these simulations (which is implicitly
derived by taking current household coverage and the estimated total budget cost).

14 INTERNATIONAL MONETARY FUND

©International Monetary Fund. Not for Redistribution

 REPUBLIC OF MADAGASCAR

Figure 12. Increasing Social Spending to Average SSA Levels, Simulations 2023–30

Education and Health Spending Education, Health, Social Assistance Spending

(Government, % of GDP) Education Health (Government, change relative to previous year, in pp
of GDP) Social Assistance, Scenario C
4.0 0.5
Health
3.5 0.5
Education
0.4
3.0
0.4
2.5
0.3
2.0 0.3
1.5 0.2
0.2
1.0
0.1
0.5
0.1
0.0 0.0
2022

2023

2024

2025

2026

2027

2028

2029

2030

2023

2024

2025

2026

2027

2028

2029

2030
Sources: IMF Staff Estimates.

F. Conclusions

18. Finding fiscal space to allocate more public resources to the education, health, and
social protection sectors should be a key government priority. The resources currently budgeted
for these sectors remain much lower than in other SSA countries and insufficient to improve
development outcomes. Madagascar made some progress in improving access to primary education
and basic health services, but the quality of the education system has deteriorated, significant
human resource gaps remain in the health sector, and the poverty rate has increased. Continued
efforts are needed to scale up social services, identify additional sources of financing, and allocate
resources efficiently given the immense development needs of the country.

19. Large scale institutional and structural reforms are needed to raise social outcomes
and alleviate poverty:

• In all three sectors, undertake PFM reforms to address significant budget under-execution
(including implementation of expenditure commitment plans by social ministries in line with
sectoral strategies) and to strengthen public investment prioritization, budgeting, and
management.

• In the education sector, ensure quality teachers by gradually integrating community teachers
into the civil service through a transparent and merit-based recruitment system and increase
decentralized management of education resources.

• In the health sector, mobilize additional resources to address equipment and medical staff
shortages, improve merit-based recruitment processes to ensure the integration of qualified
health workers, and clearly identify criteria for free basic healthcare coverage of the most
vulnerable households. For households not eligible for free healthcare, encourage
formalization or facilitate voluntary participation in contributory health insurance schemes.

INTERNATIONAL MONETARY FUND 15

©International Monetary Fund. Not for Redistribution

REPUBLIC OF MADAGASCAR

• In the area of social protection, identify clear and predictable funding sources with a view to
gradually scale up existing social programs, while developing a national social registry to
harmonize the social response among different interventions and actors and set a strong
basis to gradually increase social assistance coverage of the vulnerable population.

16 INTERNATIONAL MONETARY FUND

©International Monetary Fund. Not for Redistribution

 REPUBLIC OF MADAGASCAR

References
Cour des Comptes Audit Report, 2020, “Audit De Performance Des Mesures D’urgence Sociales Dans
Le Cadre De La Lutte Contre La COVID-19”

Politique Nationale de Protection Sociale (PNPS), 2015,

[Link]
20Sociale%20(PNPS).pdf

Public Expenditure Tracking Survey (PETS) in the Education Sector at the Primary Level, 2021,
[Link]

Stratégie Nationale de la Protection Sociale (SNPS) 2019-2023,

[Link]

UNICEF, 2021, “Towards a Universal and Inclusive Social Protection for the Children of Madagascar”

UNICEF, 2021, “Enquête sur le suivi des dépenses publiques dans le secteur de l’éducation au niveau
primaire à Madagascar (enquête PETS)”

UNICEF, 2020, Budget Briefs, [Link]

2020

World Bank, 2014, “Public Expenditure Review (PER) Health Sector Background Paper”

World Bank, 2020, “Madagascar Economic Update: Setting a Course for Recovery, December 2020”

World Bank, 2022 (a), “Systematic Country Diagnostic Update for Madagascar, April 2022”

World Bank, 2022 (b), “Madagascar Economic Update: Navigating through the Storm, May 2022”

INTERNATIONAL MONETARY FUND 17

©International Monetary Fund. Not for Redistribution