## 1. SDG Indicator 1.3.1 (Selected)

## Source details

**Canonical URL:** [1. SDG Indicator 1.3.1 (Selected)](https://www.imf.org/-/media/files/publications/cr/2026/english/1comea2026002-source-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2026/english/1comea2026002-source-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2026/english/1comea2026002-source-pdf.pdf.json)

---

### Executive summary
- Social protection coverage:
  - 3 percent of the population covered by at least one branch of social protection in 2022 (excluding health).
- Systemic diagnosis:
  - CRC (Caisse des Retraites des Comores) is main formal safety net but faces mounting financial imbalances, driven primarily by the civil servants’ scheme.
  - Actuarial analyses (ILO tools) show sizable financing gaps; even most comprehensive parametric adjustments still require transitional financing over the medium term.
- Operational recommendations (highlighted):
  - Improve data quality (individual member records).
  - Conduct external audits of financial statements.
  - Strengthen monitoring of key pension system indicators.

### Context and coverage
- Social protection and labor context:
  - Coverage: 3 percent (2022) vs regional LIC benchmark 9.7 percent.
  - Informal employment: 88 percent of the active population.
  - Regulatory gaps and limited institutional coordination constrain coverage expansion.

### Pension system structure and key parameters
- CRC schemes and coverage:
  - Three defined-benefit schemes: civil servants, non-civil servants, Members of Parliament (MPs); protections: old-age, disability, death.
  - Operates de facto on PAYG basis with no accumulated reserves and prevailing cross-financing practices.
- Key parameters (as reported):
  - Civil Servants: 17% contribution rate (12% employer, 5% employee); statutory retirement age 55–65 (category-dependent); minimum contribution period 15 years; benefit formula 2% × years of service × last 6 months’ average salary.
  - Non-Civil Servants: 8% contribution rate (5% employer, 3% employee); statutory retirement age 55–65; minimum contribution period 15 years; benefit formula 1.33% × years of service × last 6 months’ average salary.
  - Members of Parliament: 26% contribution rate (10% employer, 16% employee); statutory retirement age 55; minimum contribution period 5 years; benefit formula 5% × years of service × final parliamentary indemnity.
- Structural weakness:
  - Short salary reference period (last 6 months) inflates pension liabilities relative to contribution base.

### Demographic trends and sustainability metrics
- Contributors-per-beneficiary dynamics:
  - Civil servants’ scheme: fell from 4.1 in 2020 to 2.9 in 2024.
  - System-wide projection: 2.9 in 2024 → 1.6 by 2035.
- Drivers: stagnant contributory base (public hiring freezes), low enforcement and weak compliance among non-civil servants, rising life expectancy, declining fertility.

### Methodology and data limitations
- Methodological approach:
  - Scheme-by-scheme analysis consolidated into single framework.
  - Civil servants’ scheme: cohort-based projection using ILO actuarial modelling tool.
  - MPs and non-civil servants’ schemes: projections using aggregate assumptions calibrated on observed headcounts and cash flows.
- Data limitations and mitigation:
  - Significant administrative data gaps (missing date of birth, gender, contribution history).
  - Financial statements unaudited; reconstructed using best-estimated assumptions.
  - Supplementary data from Ministry of Civil Service.
- Interpretation caveat:
  - Results are indicative estimates to illustrate imbalance magnitude and support decision-making.

### Financial assessment and key fiscal figures
- 2024 system-wide outcome:
  - Reported system deficit of over KMF 118 million in 2024.
  - Without exceptional state support and arrears clearance, CRC’s annual deficit would have exceeded KMF 1.3 billion in 2024.
- Scheme-level 2024 deficits (theoretical figures):
  - Civil servants’ scheme: deficit of KMF 274 million in 2024 (theoretical; would have been over KMF 1.4 billion without exceptional state support and arrears clearance).
  - MPs’ scheme: deficit of KMF 69 million in 2024.
  - Non-civil servants’ scheme: the only scheme generating a surplus.
- Exceptional and temporary support provided:
  - Civil servants’ contribution rate increased from 14 percent to 17 percent in January 2025.
  - Government began settling past unpaid contribution arrears in 2023 through monthly transfers of KMF 60 million, with outstanding payments expected to continue until 2029.
  - Ad-hoc budget support: KMF 548 million in 2022 and KMF 527 million in 2024.
- Required PAYG vs legal rates (2024):
  - Civil servants’ scheme: PAYG required rate nearly 22 percent of salaries in 2024 versus legal rate of 14 percent in effect during the period.
  - MPs’ scheme: PAYG required rate exceeded 75 percent versus legal rate of 26 percent.

### Diagnosis and implications
- Central driver: mismatch between rising benefit expenditures and stagnating contributory base, compounded by inadequate contributions and weak governance.
- The required PAYG rate surpassed the legal contribution rate in 2022, meaning ordinary contributions no longer suffice for annual obligations.
- Reliance on exceptional transfers and arrears clearance is unpredictable and unsustainable; absent these, cash shortfalls could lead to delayed or disrupted benefit payments.

### Policy directions and recommended operational priorities
- Reform strategy emphasis:
  - Combine parametric adjustments to improve sustainability while preserving social protection objectives; comprehensive parametric reform scenarios still project transitional financing needs.
- Operational priorities:
  - Improve data quality (individual member records).
  - Conduct external audits of the Fund’s financial statements.
  - Strengthen monitoring of key pension system indicators.
  - Revisit contribution rates, benefit formulas (including last-6-months salary reference), accrual rates, and retirement age parameters as part of reform packages.

---

### 17. The CRC’s consolidated financial management masks a system of informal cross-financing

### Key findings on cross-financing and scheme interactions
- Surplus reuse and reserve erosion:
  - Surpluses from the non-civil servants’ scheme are used to offset large deficits in civil servants’ and MPs’ schemes, preventing reserve accumulation for non-civil servants’ scheme maturity phase.
  - Non-civil servants’ surplus no longer sufficient to offset rising public sector deficits; reliance unsustainable.
- Civil servants’ scheme:
  - Accounts for bulk of CRC’s future financing needs; on unsustainable trajectory.
- MPs’ scheme:
  - Small scale but significant structural deficit due to very limited contributor base (fewer than 35 active MPs depending on legislature) and growing beneficiaries.
- Non-civil servants’ scheme:
  - Currently in surplus under baseline but vulnerable to demographic/economic shocks and weakened by cross-financing practices.
- Contribution arrears across the Fund: nearly KMF 2 billion.

### Actuarial projections and drivers (civil servants’ scheme)
- Projection horizon: 2025–2035; baseline: no policy changes, stable contributor base, rising pensioner numbers.
- Contributors-to-pensioners ratio: 3.8 in 2025 → 2.2 by 2035.
- Scheme’s annual deficit:
  - KMF 364 million in 2025.
  - KMF 4.6 billion by 2035.
- Cumulative financing needs (civil servants’ scheme, 2025–2035): KMF 26.6 billion.
- PAYG required contribution rate (civil servants): 22.5 percent in 2025 → nearly 38 percent by 2035; statutory rate 17 percent.

### MPs’ and non-civil servants’ schemes (projections)
- MPs’ scheme annual deficit:
  - KMF 86 million in 2025.
  - Nearly KMF 290 million by 2035.
- MPs’ contribution to CRC’s total projected financing gap by 2035: less than 6 percent.
- System-wide position without reform:
  - Nearly balanced in 2025 (due to arrears clearance and non-civil servants’ surplus).
  - Annual system-wide deficit expected to exceed KMF 4.8 billion by 2035.
  - Cumulative financing requirement over 2025–2035: over KMF 25.2 billion (excluding interest).
- Non-civil servants’ scheme:
  - Statutory contribution rate 8 percent generates cumulative surplus under baseline; negative shocks or weaker compliance could erode surplus.

### Reform simulations and scenario impacts (civil servants’ scheme)
- Modeling tool: ILO-Pension model; simulations cover parametric reforms and combined packages for 2025–2035.
- Contribution rate increases:
  - Raising statutory rate to 23 percent by 2035 reduces cumulative financing need from KMF 26.6 billion to KMF 13.8 billion.
  - Even with higher statutory rates, PAYG required contribution remains well above statutory rates (baseline PAYG reaching 38 percent by 2035).
- Accrual rate adjustments (prospective, starting 2026):
  - Current accrual rate: 2 percent.
  - Considered reductions: 1.75 percent and 1.33 percent (ILO Convention 102 minimum).
  - Lowering accrual to 1.33 percent reduces cumulative need from KMF 26.6 billion to KMF 24.9 billion by 2035.
- Salary reference period extension:
  - From last 6 months to 60 months gradually reduces cumulative financing need by KMF 2.5 billion by 2035.
- Minimum retirement age reform:
  - Baseline minimum: 55 years.
  - Reform: increase by six months per year from 2026 to reach 60 years by 2035.
  - Cumulative financing need declines from KMF 26.6 billion to KMF 25.1 billion by 2035.
- Combined reform packages:
  - Package 3 (most ambitious): includes 2-percentage points annual increase in population coverage and accrual rate aligned with ILO minimum.
  - Package 3 reduces cumulative financing need from KMF 26.6 billion to KMF 2.5 billion by 2035 (over 90 percent reduction).
  - Under Package 3, PAYG required contribution stabilizes near 28 percent by 2035 (vs nearly 38 percent in baseline).
  - Even Package 3 does not fully eliminate transitional financing needs through 2035.

### MPs’ scheme reform insights
- Isolated parametric changes insufficient:
  - Increasing statutory contribution rate from baseline 26 percent to 44 percent fails to eliminate structural deficit.
- Structural options suggested:
  - Transition to budgetary financing, or
  - Integrate MPs’ scheme under civil servants’ regime with differentiated parameters.

### Policy recommendations and governance measures
- Differentiate reforms by scheme: tailor parametric and structural reforms to civil servants’, MPs’, and non-civil servants’ schemes.
- Civil servants’ scheme:
  - Coherent parametric package: phased increase in minimum retirement age, extend wage reference period, reduce accrual rates prospectively.
  - Recognize parametric reform alone insufficient; require complementary financing and governance measures.
- MPs’ scheme:
  - Require explicit political decision on financing (dedicated funding, integration, or specific budget allocation).
- Non-civil servants’ scheme:
  - Strengthen contribution collection and mandatory affiliation enforcement, particularly in private sector.
  - Reduce arrears (nearly KMF 2 billion) and improve contribution recovery to secure liquidity and build reserves.
- Formalize budget transfers:
  - Establish formal request mechanism integrated into Fund’s budget process to replace ad hoc interventions.
- Financial transparency:
  - Separate accounting for each scheme; decide explicit inter-scheme financing mechanisms.
- Stakeholder engagement:
  - Tripartite dialogue (government, employers, employees) to build consensus and social legitimacy.
- Administrative modernization:
  - Strengthen information systems, digitize processes, prepare regular audited financial statements.
- Legal reform:
  - Revise legislative framework to clarify responsibilities, embed parametric adjustments, and provide durable financing framework.
- Political commitment:
  - Sustained commitment essential to implement reforms and modernize the Fund; preserve social protection and meet ILO Convention No. 102 commitments.

### Key statistics and projections (preserve numeric values exactly)
- Projection horizon: 2025–2035.
- Contributors-to-pensioners ratio: 3.8 in 2025 → 2.2 by 2035.
- Civil servants’ scheme annual deficit: KMF 364 million in 2025; KMF 4.6 billion by 2035.
- Civil servants’ cumulative financing needs (2025–2035): KMF 26.6 billion (baseline).
- PAYG required contribution rate (civil servants): 22.5 percent in 2025 → nearly 38 percent by 2035.
- MPs’ scheme annual deficit: KMF 86 million in 2025 → nearly KMF 290 million by 2035.
- MPs’ baseline statutory contribution rate: 26 percent; simulated increase to 44 percent insufficient.
- System-wide annual deficit: expected to exceed KMF 4.8 billion by 2035.
- System-wide cumulative financing requirement (2025–2035): over KMF 25.2 billion (excluding interest).
- Non-civil servants’ statutory contribution rate: 8 percent (baseline) generating cumulative surplus under baseline assumptions.
- Contribution arrears across the Fund: nearly KMF 2 billion.
- Reform example impacts:
  - Raising statutory civil servants’ contribution rate to 23 percent by 2035 → cumulative need falls from KMF 26.6 billion to KMF 13.8 billion.
  - Lowering accrual rate to 1.33 percent → cumulative need falls from KMF 26.6 billion to KMF 24.9 billion by 2035.
  - Extending salary reference period to 60 months → reduces cumulative need by KMF 2.5 billion by 2035.
  - Increasing minimum retirement age to 60 years by 2035 (six months/year from 2026) → cumulative need falls from KMF 26.6 billion to KMF 25.1 billion by 2035.
  - Package 3 → cumulative need reduced from KMF 26.6 billion to KMF 2.5 billion by 2035; PAYG required contribution stabilizes near 28 percent by 2035.

---

### Tax and Expenditure Policies; Fuel Subsidies and Distributional Incidence

### Introduction: objectives and context
- Fiscal constraints and objectives:
  - Domestic revenue mobilization (DRM) averaged 8 percent of GDP over 2015–2022.
  - Government committed to fiscal consolidation under the ECF; reforms aim to streamline exemptions and broaden major tax bases.
- Core policy questions analyzed:
  - Incidence of (i) proper application of tax code at customs (broadening consumption taxes, excises, and income tax prepayments due to incorrect parametrization in SYDONIA) and (ii) removal of tax expenditures (consumption tax exemptions).
  - Distributional effects of removing border taxes post-WTO accession and incidence of fuel subsidies.
- Fuel subsidy preliminary figures:
  - Financial losses from underpricing petroleum products sold to gas stations in 2024 amounted to 3.0 billion KMFs (Preliminary SCH financial statements for December 2024).
  - Kerosene underpriced by 19.9 percent for household consumption; diesel supplied to SONELEC underpriced by 12.2 percent.

### Methodology: CEQ microsimulation approach and key assumptions
- Framework:
  - CEQ approach applied to 2020 household survey; microsimulation of Market Income (MI), Disposable Income (DI), Consumable Income (CI).
  - DI proxied by total consumption expenditure.
- Direct taxes and imputations:
  - MI computed by adding back direct taxes (PIT and SPT) to DI.
  - PIT taxpayers identified by employment characteristics (public sector or enterprise with pay slip).
  - Taxable wages imputed from reported net wages using Equation (1); exact correspondence with tax administration Excel tool reported.
- Tax schedule (selected brackets and rates):
  - r = 1: Taxable income 0 – 150 000; Net income 0 – 214 285; Taxation rate 0.
  - r = 2: 150 001 – 500 000; Net 214 286 – 696 787; Taxation rate 5; Cumulative tax 17 500.
  - r = 3: 500 001 – 1 000 000; Net 696 788 – 1 361 073; Taxation rate 10; Cumulative tax 67 500.
  - r = 4: 1 000 001-1 500 000; Net 1 361 074 – 2 000 359; Taxation rate 15; Cumulative tax 142 500.
  - r = 5: 1 500 001-2 500 000; Net 2 000 360 – 3 228 931; Taxation rate 20; Cumulative tax 342 500.
  - r = 6: 2 500 001-3 500 000; Net 3 228 932 – 4 407 502; Taxation rate 25; Cumulative tax 592 499.
  - r = 7: 3 500 001 and plus; Net 4 407 503 and plus; Taxation rate 30.
- Indirect taxes and specific unit taxes:
  - Specific unit taxes on oil products and rice treated by deriving implicit ad valorem rates from selling price per island and applicable unit taxes.
  - Petroleum Excise Tax (TIPP) specific taxes listed: Ordinary rice 25 (kmf/kg); Premium rice 200 (kmf/kg); Gasoline 230 (kmf/liter); Diesel 115 (kmf/liter); Kerosene 0 (kmf/liter).
  - National selling prices and national implicit tax rates:
    - Ordinary rice 313.3 → 8.7%.
    - Premium rice 1000.8 → 25.0%.
    - Gasoline 750 → 44.2%.
    - Diesel 650 → 21.5%.
    - Kerosene 500 → 0.0%.

### CEQ model calibration and simulated revenue aggregates
- CEQ-simulated government revenue totals (micro-simulation, billion KMFs):
  - 2019: Total (i) + (ii) 30.6.
  - 2020: Total (i) + (ii) 33.8.
  - 2021: Total (i) + (ii) 33.0.
- Government statement operations (billion KMFs) totals:
  - 2019: Total government revenue 50.0.
  - 2020: 49.4.
  - 2021: 55.0.
- Direct taxes:
  - Micro-simulated direct taxes: 3.2 billion KMF (2019 baseline).
  - Government reported PIT: 3.0 (2019), 3.5 (2020), 3.9 (2021) billion KMF.
  - Micro-simulated PIT reproduced broadly: 3.1 (2019), 2.9 (2020), 3.5 (2021) billion KMF.
- Indirect taxes:
  - Micro-simulated indirect taxes: 27.4 (2019), 30.9 (2020), 32.2 (2021) billion KMF.
  - Government records: 30.9 (2019/20 aggregate) and 35.1 (2021).
  - Components (selected, billion KMFs, 2019):
    - Custom duties + Excise: 12.1.
      - Custom duties: 5.2.
      - Excise: 7.0.
    - Special custom duties: 4.5 (Oil products 1.4; Rice 2.8).
    - Administrative revenue: 5.7.
    - Consumption taxes: 3.7 (Imports 2.8; Domestic 0.9).
- Coverage and discrepancies:
  - Micro-simulation reproduces 59 percent of total revenue in 2021 and 65 percent in 2020 when excluding corporate income, capital, export taxes, and some non-tax revenues.
  - Discrepancies due to imports not directly consumed by households, survey omissions, and differences in special customs duties/domestic consumption tax coverage.

### Fuel market structure and SCH losses
- Market structure:
  - SCH monopolizes importation and distribution; retail prices set by ministerial decree; no clear pricing formula.
  - Authorities maintain SCH selling prices to preserve household purchasing power.
- SCH losses:
  - SCH incurred significant losses in 2024:
    - Diesel sold to SONELEC: -1.9 billion KMFs.
    - Kerosene for household consumption: -1.1 billion KMFs.
  - Total SCH losses reached 14.5 billion KMFs in 2022, equivalent to 2.5 percent of GDP.
  - Most losses stem from diesel sold to SONELEC (63 percent in 2024; 45 percent in 2022).
  - SONELEC exemption from excise tax on diesel (115 KMF per liter) increases fiscal cost via exemptions.
- Table 4 (selected entries, 2024):
  - Kerosene for households: Monthly average cost price [I] = 393.4 KMF/liter; Selling price [II] = 315 KMF/liter; SCH Margin [III] = -1.1 (Billion KMFs); % change = -19.9.
  - Diesel - SONELEC: [I] = 393.1; [II] = 345; [III] = -1.9; % change = -12.2.
  - Gasoline: [I] = 638.5; [II] = 725; [III] = 1.8; % change = 13.5.
  - Diesel - Tourism: [I] = 508.4; [II] = 630; [III] = 3.0; % change = 23.9.
  - Source: Staff calculations using preliminary SCH financial statement for December 2024.

### IMF fuel subsidy toolkit: methodology (summary)
- Toolkit assesses:
  1. Change in total spending required to maintain same consumption basket.
  2. Distribution of subsidies across households.
- Assumptions:
  - Household consumption patterns unchanged by subsidy (simplifying assumption).
  - Direct impact formula and aggregation across goods as presented.
  - Indirect impacts derived via input-output mapping; requires mapping household consumption categories to industries.
- Caveat:
  - If fuel price elasticity differs by income, toolkit results may overstate distributional impacts; reported results closer to upper bound if low-income elasticity higher.

### Fiscal and distribution incidence results (CEQ and related findings)
- Tax-code application fix:
  - Circular note issued in July (implementation September 2025 after moratorium) broadening bases for consumption, excise, prepayment taxes.
  - Expected additional revenue: 1.2 billion KMFs (0.18 percent of GDP).
    - Composition: Excise ~58% (0.7 billion KMFs); Consumption taxes on imports 33%; Prepayment tax 8%.
- Fiscal cost of tax expenditures:
  - Estimated at 1.0 billion KMFs (0.16 percent of GDP).
  - A 45 KMF/liter tax on kerosene would generate around 0.6 billion KMFs (0.8 percent of GDP).
- Post-WTO accession measures:
  - Removal of import licenses and prepayment tax per 2025 budget law → revenue loss of 1.4 billion KMFs (0.24 percent of 2024 GDP).
  - Comparable estimate by Sawadogo and Kett (2024): 1.7 billion KMFs.
- Broadening tax base impacts on unit prices:
  - Non-essential goods larger effects: alcohol +15.0 percent, tobacco +14.3 percent.
  - Basic foods: 1.0 percent (poultry or beef) to 5.8% (root crops).
  - Rice and fuel not affected (subject to single specific customs duty).

### CEQ results on inequality and incidence by decile
- Gini outcomes:
  - Three policy options (broadening base; accounting for tax expenditures; removing import licenses or prepayment tax) have insignificant effect on income distribution; Gini equals benchmark across options.
  - Benchmark Gini coefficients:
    - Market Income (MI): 33.18.
    - Disposable Income (DI): 32.99.
    - Consumable Income (CI): 33.08.
  - Overall redistributive impact of tax system negligible; 0.1-point decrease in Gini from MI to CI in some representations.
  - Direct taxes moderately progressive; indirect taxes mildly regressive.
- Revenue distribution from broadening the tax base (1.2 billion KMFs):
  - Top 10% (D10) contribute 0.34 billion KMFs; Bottom 10% (D1) contribute 0.04 billion KMFs.
  - Per capita tax increase by decile (KMFs): D10 = 4,384.3; D1 = 563.2.
  - Excise taxes account for 55–58% depending on decile.
  - Burden relative to disposable income broadly uniform across deciles, explaining unchanged Gini.
- Tax expenditures incidence:
  - Tax expenditures increase with income: Top 10% exempted benefit = 0.43 billion KMFs; Bottom 10% = 0.05 billion KMFs.
  - Per capita tax expenditure: Top 10% = 5,477.7 KMFs; Bottom 10% = 595.9 KMFs.
  - As share of disposable income: top 10% benefit 0.31 percent; bottom 10% 0.26 percent.

### Household survey and fuel consumption patterns
- Survey: 5,624 households.
- Household expenditure (selected figures):
  - Observations: Rural female-headed 1,426; Rural male-headed 2,270; Urban female-headed 715; Urban male-headed 1,213.
  - Average total household expenditure (Rural / Urban):
    - Female headed: 2,473,065 / 2,750,881.
    - Male headed: 2,742,845 / 3,253,449.
  - Average per capita household expenditure (Rural / Urban):
    - Female headed: 591,123.4 / 766,858.8.
    - Male headed: 552,228.3 / 785,726.8.
  - Women’s reported average wage income = 58 % of men’s income in survey.
- Fuel consumption patterns:
  - Richer households consume more fuel products per capita; kerosene largest budget share across income categories.
  - Richer households devote larger spending shares to vehicle fuel and gas for household use.

### Direct and indirect fuel subsidy incidence
- Kerosene direct subsidy:
  - Kerosene subsidy as share of disposable income higher among higher-income households; richer households receive larger share of total kerosene subsidy.
- Diesel subsidy (indirect via SONELEC):
  - Only diesel sold to SONELEC is subsidized; household impacts indirect through input-output linkages.
  - Simulated diesel subsidy: 12.2 percent.
  - Sector price impacts: electricity, gas and water most affected.
  - Distributional outcome: indirect subsidy benefits accrue disproportionately to richer households (top decile benefits most).

### Conclusion: fiscal and distributional implications
- Fiscal reforms effective at raising revenue but limited impact on inequality.
  - Broadening tax base and removing selected tax expenditures estimated to increase government revenues by approximately 2.8 billion KMF (about 0.35 percent of 2024 GDP); Gini largely unchanged.
  - Broadening base minimal effect on essential goods prices (1–5 percent) but larger on non-essential goods: alcohol +15 percent; tobacco +14.3 percent.
  - Reforms broadly uniform across income distribution when measured as share of disposable income; in absolute terms richer households pay more (top 10% pay 0.35 billion KMF vs bottom 10% 0.04 billion KMF under broadened base).
- Current tax exemptions and fuel subsidies are regressive:
  - Tax exemptions benefit higher-income households more: richest 10% receive 0.31 percent of disposable income in benefits vs poorest 10% 0.26 percent.
  - Fuel subsidies (direct and indirect) mainly benefit top income households.

### Policy recommendations (taxes and subsidies)
- Target tax exemptions:
  - Eliminate consumption tax exemptions for non-essential goods; redirect resources to improve equity and fiscal efficiency.
- Reform fuel subsidies:
  - Phase out fuel subsidies while mitigating adverse distributional effects through direct transfers.
  - Prefer targeted transfers to support vulnerable groups without distorting market prices or disproportionately benefiting the wealthy.

*Source: Extracted from the provided IMF PDF content unit.*

### 1. SDG Indicator 1.3.1 (Selected) _______________________________________________________ 3

### 1. SDG Indicator 1.3.1 (Selected)

### Executive summary
- Social protection coverage in Comoros is very limited: only about 3 percent of the population was covered by at least one branch of the social protection system in 2022.
- The CRC (Caisse des Retraites des Comores) is the main formal safety net but faces mounting financial imbalances, driven primarily by the civil servants’ scheme.
- Actuarial analyses (conducted with ILO tools and methodologies) quantify sizable financing gaps and show that, even under the most comprehensive combination of parametric adjustments, the system is projected to continue requiring some level of transitional financing over the medium term.
- Key operational recommendations highlighted in the analysis include improving data quality (individual member records), conducting external audits of the financial statements, and strengthening monitoring of key pension system indicators.

### Context and coverage
- Social protection coverage:
  - 3 percent of the population covered by at least one branch of social protection in 2022 (excluding health).
  - Regional benchmark: 9.7 percent average for low-income countries (LIC).
- Labor market and institutional context:
  - Informal employment estimated at 88 percent of the active population.
  - Regulatory gaps and limited coordination across institutions constrain coverage expansion.

### Pension system structure and key parameters
- CRC administers three defined-benefit schemes: civil servants, non-civil servants, and Members of Parliament (MPs). Schemes protect against old-age, disability, and death.
- The system operates de facto on a pay-as-you-go (PAYG) basis with no accumulated reserves and prevailing cross-financing practices.
- Key parameters (as reported):
  - Civil Servants: 17% (12% employer, 5% employee) contribution rate; statutory retirement age 55–65 (category-dependent); minimum contribution period 15 years; benefit formula 2% × years of service × last 6 months’ average salary.
  - Non-Civil Servants: 8% (5% employer, 3% employee) contribution rate; statutory retirement age 55–65 (category-dependent); minimum contribution period 15 years; benefit formula 1.33% × years of service × last 6 months’ average salary.
  - Members of Parliament: 26% (10% employer, 16% employee) contribution rate; statutory retirement age 55; minimum contribution period 5 years; benefit formula 5% × years of service × final parliamentary indemnity.
- Structural weakness: short salary reference period (last 6 months’ average salary) inflates pension liabilities relative to the contribution base.

### Demographic trends and sustainability metrics
- Demographic ratios (contributors per beneficiary):
  - Civil servants’ scheme: fell from 4.1 in 2020 to 2.9 in 2024.
  - System-wide projection: from 2.9 in 2024 to 1.6 by 2035.
- Drivers: stagnant contributory base (including public sector hiring freezes), low enforcement and weak compliance among non-civil servants, rising life expectancy, and declining fertility rates.

### Methodology and data limitations
- Methodological approach:
  - Scheme-by-scheme analysis consolidated into a single framework.
  - Civil servants’ scheme: cohort-based projection using the ILO actuarial modelling tool.
  - MPs and non-civil servants’ schemes: projections using aggregate assumptions calibrated on observed headcounts and cash flows.
- Data limitations and mitigation:
  - Significant gaps and inconsistencies in administrative data (missing date of birth, gender, contribution history).
  - Financial statements unaudited; data reconstruction carried out using best-estimated assumptions.
  - Supplementary data sourced from the Ministry of Civil Service to fill gaps.
- Interpretation caveat: results are indicative estimates intended to illustrate sources and magnitude of imbalance and to support decision-making.

### Financial assessment and key fiscal figures
- 2024 system-wide outcome:
  - Reported system deficit of over KMF 118 million in 2024.
  - Without exceptional state support and arrears clearance, the CRC’s annual deficit would have exceeded KMF 1.3 billion in 2024.
- Scheme-level 2024 deficits (theoretical figures):
  - Civil servants’ scheme: deficit of KMF 274 million in 2024 (theoretical; would have been over KMF 1.4 billion without exceptional state support and arrears clearance).
  - MPs’ scheme: deficit of KMF 69 million in 2024.
  - Non-civil servants’ scheme: the only scheme generating a surplus.
- Exceptional and temporary support provided:
  - Civil servants’ contribution rate increased from 14 percent to 17 percent in January 2025.
  - Government began settling past unpaid contribution arrears in 2023 through monthly transfers of KMF 60 million, with outstanding payments expected to continue until 2029.
  - Ad-hoc budget support to cover payment shortfalls: KMF 548 million in 2022 and KMF 527 million in 2024.
- PAYG required contribution rates versus legal rates (2024):
  - Civil servants’ scheme: PAYG required rate nearly 22 percent of salaries in 2024 versus legal rate of 14 percent in effect during the period.
  - MPs’ scheme: PAYG required rate exceeded 75 percent versus legal rate of 26 percent.

### Diagnosis and implications
- The central driver of financial deterioration is the mismatch between rising benefit expenditures and a stagnating contributory base, exacerbated by inadequate contribution levels and weak governance.
- The crossover where the required PAYG rate surpassed the legal contribution rate occurred in 2022, marking the point where ordinary contributions no longer sufficed to meet annual obligations.
- Reliance on exceptional transfers and arrears clearance is unpredictable and unsustainable; without these measures, cash shortfalls would be substantially larger and could result in delayed or disrupted benefit payments.

### Policy directions and recommended operational priorities (as presented)
- Reform strategy emphasis:
  - Combine parametric adjustments to improve sustainability while preserving social protection objectives; even comprehensive parametric reform scenarios are projected to require transitional financing over the medium term.
- Operational priorities highlighted by the analysis:
  - Improve data quality, especially individual member records.
  - Conduct external audits of the Fund’s financial statements.
  - Strengthen monitoring of key pension system indicators.
  - Address structural drivers: revisit contribution rates, benefit formulas (including the last-6-months salary reference), accrual rates, and retirement age parameters as part of reform packages.

*Source: Pension Reform in Comoros (Appendix, January 27, 2026).*

### 17. The CRC’s consolidated financial management masks a system of informal cross-

### 17. The CRC’s consolidated financial management masks a system of informal cross-

### Key findings
- Surpluses generated by the non-civil servants’ scheme are currently used to partially offset the large deficits of the civil servants’ and MPs’ schemes, preventing the non-civil servants' scheme from building reserves for its eventual maturity phase.
- The surplus from the non-civil servants’ scheme is no longer sufficient to offset rising deficits in the public sector schemes; continued reliance on these diminishing surpluses is unsustainable and will further exacerbate the CRC’s overall financing gap.
- The civil servants’ scheme accounts for the bulk of the CRC’s future financing needs and is on an unsustainable trajectory.
- The MPs’ scheme, while smaller in scale, has a significant structural deficit driven by a very limited contributor base (fewer than 35 active MPs depending on the legislature) and growing beneficiaries.
- The non-civil servants’ scheme is currently in surplus but remains vulnerable to demographic and economic shifts and weakened by cross-financing practices.

### Actuarial projections and drivers (civil servants’ scheme)
- Projection horizon: 2025–2035; baseline assumes no policy changes, a stable contributor base, and a rising number of pensioners.
- Contributors-to-pensioners ratio projected to decline from 3.8 in 2025 to 2.2 by 2035.
- Scheme’s annual deficit:
  - KMF 364 million in 2025
  - KMF 4.6 billion by 2035
- Cumulative financing needs for civil servants’ scheme over 2025–2035: KMF 26.6 billion.
- PAYG required contribution rate projected to rise from 22.5 percent in 2025 to nearly 38 percent by 2035, exceeding the statutory rate of 17 percent.

### MPs’ and non-civil servants’ schemes (projections)
- MPs’ scheme annual deficit:
  - KMF 86 million in 2025
  - Nearly KMF 290 million by 2035
- MPs’ scheme contribution to CRC’s total projected financing gap by 2035: less than 6 percent.
- System-wide position without reform:
  - Nearly balanced in 2025 (due to arrears clearance and non-civil servants’ surplus).
  - Annual system-wide deficit expected to exceed KMF 4.8 billion by 2035.
  - Cumulative financing requirement over 2025–2035: over KMF 25.2 billion (calculated excluding interest).
- Non-civil servants’ scheme: statutory contribution rate of 8 percent generates a cumulative surplus under baseline assumptions, but negative demographic/economic shocks or weaker contribution compliance could erode the surplus. Contribution arrears across the Fund are currently estimated at nearly KMF 2 billion.

### Reform simulations and scenario impacts (civil servants’ scheme)
- Modelling tool: ILO-Pension model; simulations focus on parametric reforms and combined reform packages over 2025–2035.
- Contribution rate increases (statutory rate from baseline 17 percent to alternatives):
  - Raising statutory rate to 23 percent by 2035 reduces cumulative financing need from KMF 26.6 billion to KMF 13.8 billion.
  - Even with higher statutory rates, PAYG required contribution remains well above statutory rates, reaching 38 percent by 2035 (baseline PAYG projection reference).
- Accrual rate adjustments (starting 2026; applied prospectively):
  - Current accrual rate: 2 percent.
  - Reductions considered: 1.75 percent and 1.33 percent (aligns with ILO Convention 102 minimum).
  - Lowering accrual rate to 1.33 percent reduces cumulative financing need from KMF 26.6 billion to KMF 24.9 billion by 2035.
- Salary reference period extension:
  - Current reference: last 6 months of salary.
  - Gradual extension to 60 months reduces cumulative financing need by KMF 2.5 billion by 2035.
- Minimum retirement age reform:
  - Baseline minimum retirement age: 55 years.
  - Reform: increase by six months per year starting in 2026, reaching 60 years by 2035.
  - Cumulative financing need declines from KMF 26.6 billion to KMF 25.1 billion by 2035.
- Combined reform packages:
  - Three packages assessed; Package 3 (most ambitious) includes, among other measures, a 2-percentage points annual increase in population coverage and alignment of the accrual rate with the ILO minimum standard.
  - Package 3 reduces cumulative financing need from KMF 26.6 billion to KMF 2.5 billion by 2035 (over 90 percent reduction).
  - Under Package 3, the PAYG required contribution rate stabilizes near 28 percent by 2035 (compared to nearly 38 percent in baseline).
  - Even the most ambitious package does not fully eliminate transitional financing needs through 2035.

### MPs’ scheme reform insights
- Isolated parametric changes are insufficient:
  - Increasing statutory contribution rate from baseline 26 percent to 44 percent fails to eliminate the structural deficit.
- Structural options suggested:
  - Transition to budgetary financing, or
  - Integrate MPs’ scheme under the civil servants’ regime with differentiated parameters,
  - These are technical options to ensure long-term sustainability while addressing equity concerns.

### Policy recommendations and governance measures
- Differentiate reforms by scheme: tailor parametric and structural reforms to the civil servants’, MPs’, and non-civil servants’ schemes based on their specific trajectories and parameters.
- Civil servants’ scheme:
  - Implement a coherent package of parametric reforms: increase minimum retirement age (phased), extend wage reference period, and reduce accrual rates (applied prospectively).
  - Recognize that parametric reform alone will not fully eliminate medium-term deficits; complementary financing and governance measures will be required.
- MPs’ scheme:
  - Require an explicit political decision on financing: options include dedicated funding, integration as a complementary component to another scheme, or specific budget allocation.
- Non-civil servants’ scheme:
  - Strengthen contribution collection and mandatory affiliation enforcement, particularly in the private sector.
  - Reduce arrears (currently nearly KMF 2 billion) and improve contribution recovery to secure short-term liquidity and build reserves.
- Formalize budget transfers:
  - Establish a formal request mechanism integrated into the Fund’s budget process to replace ad hoc interventions and stabilize cash flows.
- Financial transparency:
  - Separate and transparent accounting for each scheme to reveal true imbalances and enable targeted adjustments; decide on explicit inter-scheme financing mechanisms.
- Stakeholder engagement:
  - Conduct tripartite dialogue among government, employers, and employee representatives to build consensus and enhance social legitimacy for reforms.
- Administrative modernization:
  - Strengthen information systems, digitize processes, and prepare regular official audited financial statements to improve data reliability and governance.
- Legal reform:
  - Revise the legislative framework to clarify institutional responsibilities, embed parametric adjustments, and provide a durable financing framework.
- Political commitment:
  - Sustained political and institutional commitment is essential to implement reforms and modernize the Fund, preserving social protection and meeting international commitments (notably ILO Convention No. 102).

### Key statistics and projections (preserve numeric values exactly)
- Projection horizon: 2025–2035.
- Contributors-to-pensioners ratio: 3.8 in 2025 → 2.2 by 2035.
- Civil servants’ scheme annual deficit: KMF 364 million in 2025; KMF 4.6 billion by 2035.
- Civil servants’ cumulative financing needs (2025–2035): KMF 26.6 billion (baseline).
- PAYG required contribution rate (civil servants): 22.5 percent in 2025 → nearly 38 percent by 2035.
- MPs’ scheme annual deficit: KMF 86 million in 2025 → nearly KMF 290 million by 2035.
- MPs’ baseline statutory contribution rate: 26 percent; simulated increase to 44 percent insufficient.
- System-wide annual deficit: expected to exceed KMF 4.8 billion by 2035.
- System-wide cumulative financing requirement (2025–2035): over KMF 25.2 billion (excluding interest).
- Non-civil servants’ statutory contribution rate: 8 percent (baseline) generating cumulative surplus under baseline assumptions.
- Contribution arrears across the Fund: nearly KMF 2 billion.
- Reform example impacts:
  - Raising statutory civil servants’ contribution rate to 23 percent by 2035 → cumulative need falls from KMF 26.6 billion to KMF 13.8 billion.
  - Lowering accrual rate to 1.33 percent → cumulative need falls from KMF 26.6 billion to KMF 24.9 billion by 2035.
  - Extending salary reference period to 60 months → reduces cumulative need by KMF 2.5 billion by 2035.
  - Increasing minimum retirement age to 60 years by 2035 (six months/year from 2026) → cumulative need falls from KMF 26.6 billion to KMF 25.1 billion by 2035.
  - Package 3 → cumulative need reduced from KMF 26.6 billion to KMF 2.5 billion by 2035; PAYG required contribution stabilizes near 28 percent by 2035.

*Source: ILO and IMF staff calculations as presented in the source.*

### References

### References and Appendix I: Pension Reform; Tax and Expenditure Policies in Comoros — Methodology and Calibration

### Key references cited
- Drissi-Bourhanbour, L. 2024. Union of the Comoros, Selected Issue Paper: Costing and financing of social development goals (SDG) in Comoros. IMF Country Report No. 2024/005.
- D’Harcourt, J. 2019. Les régimes des retraites des Comores: Étude de soutenabilité et scénarios d’évolution. AFRITAC South, April. International Monetary Fund.
- International Monetary Fund. 2022. IMF Engagement on Pension Issues in Surveillance and Program Work. IMF Technical Notes and Manuals 2022/004. Fiscal Affairs Department and Strategy, Policy, and Review Department.
- International Monetary Fund, 2025. Union of the Comoros: Fourth Review under the Extended Credit Facility Arrangement and Request for Modification of Performance Criteria—Staff Report. IMF Country Report No. 2025/186.
- International Labour Organization. Forthcoming. Rapport à la Caisse des Retraites des Comores: Étude de Soutenabilité. Geneva: International Labour Organization.

### Appendix I — Pension reform: scenarios and model outputs
- Figures referenced (sources: ILO calculations; ILO and IMF staff calculations) include:
  - Legal Contribution Rate vs. Required PAYG Rate (Percent of Pensionable Earnings) for A. Civil Servants’ Scheme and B. MPs Scheme.
  - Civil Servants Scheme - Net Result, 2020–2024 (KMF x 1000).
  - Projections of Legal Contribution Rate vs. Required PAYG Rate (Percent of Pensionable Earnings) for A. Civil Servants’ Scheme and B. System-wide.
  - Required Contribution Rates Under Different Reforms, 2025–2035 (Percent of Pensionable Earnings) panels for A. Contribution Rate; B. Accrual Rate; C. Salary Reference Period; D. Retirement Age.
  - Required vs. Legal Contribution Rate Under Different Reform Packages (Percent of Pensionable Earnings).
  - Projected Financing Needs Under Different Contribution Rates (KMF Million) for A. MPs’ Scheme and B. Non-Civil Servants’ Scheme.
- Table 1. Union of the Comoros: Summary of Reform Scenarios (Source: ILO calculations)
  - Scenario: Baseline
    - Population Growth (%): 0.00
    - Contribution Rate (%): 17.00
    - Accrual Rate (%): 2
    - Reference Salary (months): 6
    - Minimum Retirement Age (years): 55
  - Scenario: Package 1
    - Population Growth (%): 1.00
    - Contribution Rate (%): 19.00
    - Accrual Rate (%): 1.75
    - Reference Salary (months): 36
    - Minimum Retirement Age (years): 60
  - Scenario: Package 2
    - Population Growth (%): 2.00
    - Contribution Rate (%): 21.00
    - Accrual Rate (%): 1.5
    - Reference Salary (months): 36
    - Minimum Retirement Age (years): 60
  - Scenario: Package 3
    - Population Growth (%): 2.00
    - Contribution Rate (%): 23.00
    - Accrual Rate (%): 1.33
    - Reference Salary (months): 60
    - Minimum Retirement Age (years): 60

### Tax and Expenditure Policies in Comoros — Introduction: objectives and context
- Fiscal constraints and objectives:
  - Domestic revenue mobilization (DRM) averaged only 8 percent of GDP over 2015–2022.
  - Government committed to fiscal consolidation under the ECF; reforms aim to streamline exemptions and broaden major tax bases.
- Core policy questions analyzed:
  - Incidence on income distribution of (i) proper application of the tax code at customs (broadening consumption taxes, excises, and income tax prepayments due to incorrect parametrization in SYDONIA) and (ii) removal of tax expenditures, particularly consumption tax exemptions.
  - Distributional effects of removing border taxes post-WTO accession, including abolishing income tax advance payments and business license taxes collected at the customs cordon.
  - Incidence of fuel subsidies on income distribution: in 2024 financial losses from underpricing petroleum products sold to gas stations amounted to 3.0 billion KMFs (Preliminary SCH financial statements for December 2024).
    - Two petroleum products significantly underpriced relative to their average cost: kerosene for household consumption underpriced by 19.9 percent; diesel supplied to the state-owned electricity company SONELEC underpriced by 12.2 percent.

### Methodology: CEQ microsimulation approach and key assumptions
- Methodological framework:
  - Commitment to Equity (CEQ) approach (Lustig et al., 2023) applied to the 2020 household survey; microsimulation of income concepts: Market Income (MI), Disposable Income (DI), and Consumable Income (CI).
  - DI proxied by total consumption expenditure (including implied value of self-consumption and imputed rent); acknowledges potential underestimation of DI by the amount saved, especially for upper-decile households.
- Direct taxes and imputations:
  - Market Income (MI) computed by adding back direct taxes (Personal Income Tax—PIT—and Single Professional Tax—SPT) to DI.
  - PIT taxpayers identified by employment characteristics: an adult over 15 years old employed in public sector or enterprise and receiving a pay slip is treated as a PIT payer.
  - Taxable wages imputed from reported net wages using Equation (1) (as presented) and cross-checked against the tax administration Excel tool; exact correspondence reported between Equation (1) and the Excel tool.
- Tax schedule used (Table 1 in text; Union of the Comoros: Tax Schedule)
  - r = 1: Taxable income ranges (KMFs) 0 – 150 000; Net income ranges (KMFs) 0 – 214 285; Taxation rates (%) 0; Cumulative tax (KMFs) 0
  - r = 2: Taxable income ranges 150 001 – 500 000; Net income ranges 214 286 – 696 787; Taxation rates (%) 5; Cumulative tax (KMFs) 17 500
  - r = 3: Taxable income ranges 500 001 – 1 000 000; Net income ranges 696 788 – 1 361 073; Taxation rates (%) 10; Cumulative tax (KMFs) 67 500
  - r = 4: Taxable income ranges 1 000 001-1 500 000; Net income ranges 1 361 074 – 2 000 359; Taxation rates (%) 15; Cumulative tax (KMFs) 142 500
  - r = 5: Taxable income ranges 1 500 001-2 500 000; Net income ranges 2 000 360 – 3 228 931; Taxation rates (%) 20; Cumulative tax (KMFs) 342 500
  - r = 6: Taxable income ranges 2 500 001-3 500 000; Net income ranges 3 228 932 – 4 407 502; Taxation rates (%) 25; Cumulative tax (KMFs) 592 499
  - r = 7: Taxable income ranges 3 500 001 and plus; Net income ranges 4 407 503 and plus; Taxation rates (%) 30
- Indirect taxes and subsidies:
  - CI computed by subtracting indirect taxes from DI and adding subsidies; methodology works backward from survey-reported tax-inclusive expenditures T_k to pre-tax expenditures c_k consistent with the 2024 tax code.
  - Specific unit taxes on oil products and rice treated by deriving implicit ad valorem rates from selling price on each island and applicable unit taxes (Table 2: Implicit Tax Rates for Specific Taxes on Oil Products and Rice).
    - Petroleum Excise Tax (TIPP) specific taxes listed: Ordinary rice 25 (kmf/kg); Premium rice 200 (kmf/kg); Gasoline 230 (kmf/liter); Diesel 115 (kmf/liter); Kerosene 0 (kmf/liter) [as presented in the table].
    - National selling prices: Ordinary rice 313.3; Premium rice 1000.8; Gasoline 750; Diesel 650; Kerosene 500.
    - National implicit tax rates: Ordinary rice 8.7%; Premium rice 25.0%; Gasoline 44.2%; Diesel 21.5%; Kerosene 0.0%.
    - Island-level implicit tax rates also provided for Anjouan, Moheli, Grande-Comores (values as shown in table).
  - Validation checks:
    - Formulas for deriving pre-tax consumption and theoretical CIF values are cross-checked against comprehensive customs data and the tax-inclusive import universe; Figure 3 compares observed CIF values with theoretical values (noted consistency check).

### CEQ model calibration and simulated revenue aggregates
- CEQ-simulated government revenue for 2019–2021 (Table 3: Category of Tax Revenues Included in the Microsimulation Model)
  - Micro-simulation (billion KMFs) totals:
    - 2019: Total (i) + (ii) 30.6
    - 2020: Total (i) + (ii) 33.8
    - 2021: Total (i) + (ii) 33.0
  - Government statement operations (billion KMFs) totals:
    - 2019: Total government revenue 50.0
    - 2020: Total government revenue 49.4
    - 2021: Total government revenue 55.0
- Direct taxes (micro-simulation vs. government statements)
  - Micro-simulated direct taxes: 3.2 billion KMF (2019 baseline in table).
  - Government reported personal income tax: 3.0 (2019), 3.5 (2020), 3.9 (2021) billion KMF.
  - Micro-simulation reproduces direct taxes broadly: simulated personal income tax 3.1 (2019), 2.9 (2020), 3.5 (2021) billion KMF (per table cell breakdown).
- Indirect taxes (micro-simulation vs. government statements)
  - Micro-simulated indirect taxes: 27.4 billion KMF (2019) and 30.9 (2020) and 32.2 (2021).
  - Government records: 30.9 (2019/20 aggregate reference) and 35.1 (2021).
  - Components in micro-simulation (selected rows, billion KMFs):
    - Custom duties + Excise: 12.1 (2019)
      - Custom duties: 5.2 (2019)
      - Excise: 7.0 (2019)
    - Import license: 0.7 (2019)
    - Import prepayment tax: 0.7 (2019)
    - Special custom duties: 4.5 (2019); breakdown includes Oil products 1.4 (2019) and Rice 2.8 (2019)
    - Administrative revenue: 5.7 (2019)
    - Consumption taxes: 3.7 (2019) with Imports 2.8 and Domestic 0.9 (2019)
- Coverage and discrepancies:
  - Micro-simulation reproduces 59 percent of total revenue in 2021 and 65 percent in 2020 when excluding corporate income, capital, and export taxes, and non-tax revenues from stamps and registration fees.
  - Discrepancies attributed to:
    - Imports not directly consumed by households (e.g., oil products used as intermediate inputs in electricity, transport, bakery).
    - Survey omission of major taxable domestic items (financial services, construction, incoming phone call charges).
    - Differences in special customs duties and domestic consumption taxes due to survey coverage limits.

*Source: Extracted from the provided IMF PDF content unit.*

### 12.      We also analyze the distribution of fuel subsidy benefits using the IMF subsidy toolkit.

### 12.      We also analyze the distribution of fuel subsidy benefits using the IMF subsidy toolkit.

### Fuel market structure and SCH losses
- Importation and distribution of petroleum products are monopolized by the state-owned Comorian Hydrocarbons Company (SCH); retail prices are set and regulated by ministerial decree.
- The decree sets the selling price per liter charged by SCH to service stations and the margin service stations apply to end consumers. There is no clear formula in setting the selling price.
- Until 2022, the selling price per liter imposed by the government covered the SCH price cost per liter allowing the SCH to generate profits. However, the selling price per liter may be below the cost price per liter resulting in financial losses for the company.
- Authorities intend to preserve households’ purchasing power by maintaining SCH selling prices regardless of changes in company costs.
- SCH incurred significant losses in 2024:
  - Diesel sold to the state-owned electricity company SONELEC: -1.9 billion KMFs.
  - Kerosene used for household consumption: -1.1 billion KMFs.
- Total SCH losses reached 14.5 billion KMFs in 2022, equivalent to 2.5 percent of GDP (Ahamada, 2024). Most losses stem from diesel sold to SONELEC (63 percent in 2024 and 45 percent in 2022).
- SONELEC is exempt from the excise tax on diesel (115 KMF per liter); the government thus covers underpricing losses and incurs additional tax expenditure via exemptions.

- Table 4: SCH Fuel Price Structure and Profit in 2024 (selected entries)
  - Kerosene for households: 2024 Monthly average cost price (KMF/liter) [I] = 393.4; Selling price to gas station (KMF/liter) [II] = 315; SCH Margin (Billion KMFs) [III] = -1.1; % change [(II – I)/I] = -19.9
  - Diesel - SONELEC: [I] = 393.1; [II] = 345; [III] = -1.9; % change = -12.2
  - Gasoline: [I] = 638.5; [II] = 725; [III] = 1.8; % change = 13.5
  - Diesel - Tourism: [I] = 508.4; [II] = 630; [III] = 3.0; % change = 23.9
  - Source: Staff calculations using preliminary SCH financial statement for December 2024.

### IMF fuel subsidy toolkit: methodology
- The toolkit assesses fuel subsidies on two dimensions:
  1. The change in total spending required to maintain the same basket of goods.
  2. The distribution of subsidies across different households.
- Calculations assume household consumption patterns remain unchanged in response to subsidies (simplifying assumption).
- Direct impact formula (as presented):
  - 푑푑푌푌 / 푑푑푏푏푏푏  (푌푌) = 푇푇_b * (푑푑푃푃_i / 푃푃_i)
    - 푇푇_b is the spending share on product c.
    - 푑푑푃푃_i / 푃푃_i is the percentage change in the price of product c due to subsidy.
- If multiple goods change price: 푑푑푌푌 / 푑푑푖푖푟푟 (푌푌) = Σ_b (푇푇_b * (푑푑푃푃_i / 푃푃_i))
- Indirect impact accounts for price changes in goods that use subsidized goods as intermediate inputs:
  - 푑푑푌푌 / 푑푑푏푏푏푏 (푌푌) = Σ_j (푇푇_j * (푑푑푃푃_j / 푃푃_j)), where j indexes affected goods.
- Implementation requires mapping household consumption categories to industries in the input-output table.
- Caveat: simplifying assumption may distort results if fuel price elasticity differs by income; if lower-income households have higher elasticity, unequal impacts reported by the toolkit are closer to an upper bound.
- Note: The losses are not systematically reflected on the government fiscal accounts.

### Fiscal and distribution incidence results (CEQ model and related findings)
- Tax-code application fix: A FAD TA mission supporting the Tax Policy Unit identified inconsistencies between the tax code and customs application. Authorities issued a circular note in July (implementation in September 2025 after a moratorium) broadening bases used to compute consumption, excise, and prepayment taxes.
  - Expected additional revenue: 1.2 billion KMFs, equivalent to 0.18 percent of GDP (Table 5).
  - Revenue composition of the 1.2 billion KMFs:
    - Excise taxes: about 58 percent (0.7 billion KMFs).
    - Consumption taxes on imports: 33 percent.
    - Prepayment tax: 8 percent.
- Fiscal cost of tax expenditures: estimated at 1.0 billion KMFs (0.16 percent of GDP).
  - A 45 kmf/liter tax on kerosene would generate around 0.6 billion KMFs (0.8 percent of GDP).
- Post-WTO accession measures (removal of import licenses and prepayment tax per 2025 budget law) would lead to revenue loss of 1.4 billion KMFs (0.24 percent of 2024 GDP), equally spread between import licenses and the prepayment tax.
  - Comparable estimate by Sawadogo and Kett (2024): 1.7 billion KMFs.
- Broadening the tax base mainly affects alcohol and tobacco:
  - Unit price increases at customs: tobacco products +14.3 percent; alcoholic beverages +15.0 percent.
  - Impact on basic food commodities ranges from 1 percent (poultry or beef) to 5.8% (root crops).
  - Rice and fuel are not affected, as they are subject to a single tax (specific customs duty).
- Table 5: Impact of Broadening the Tax Base and the Tax Expenditure on the Government Budget (selected figures)
  - Benchmark total revenue = 30.7 billion KMFs; Impact (billion KMFs) = 1.2
  - Indirect taxes (ii): Benchmark 27.4; Impact 1.2
    - Custom duties + Excise: Benchmark 12.2; Impact 0.7
      - Excise: Benchmark 7.0; Impact 0.7
    - Import prepayment tax (ai): Benchmark 0.7; Impact 0.1
  - Fiscal Cost of Tax Expenditures (iii): 1.6
    - Kerosene for household use (45 KMFs/liter): - 0.6
    - Exemption on 10% consumption tax: - 1.0
  - Source: Staff calculations. Benchmark is reference before circular note.

- Table 6 (selected sensitive products): Impact on unit price at customs (%)
  - Animal offal (beef, goat, etc.) (0206): Unit price impact = 1.0
  - Poultry meat and offal (0207): Unit price impact = 1.0
  - Vegetables, roots (07): Unit price impact = 5.8
  - Wheat or meslin flour (1101): Unit price impact = 0.0
  - Vegetable oil (151620): Unit price impact = 1.7
  - Sugar (1701): Unit price impact = 0.9
  - Water, non-alcoholic beverages (220210): Unit price impact = 2.9
  - Alcohol (220300 + 220830): Unit price impact = 15.0
  - Tobacco (24): Unit price impact = 14.3
  - Butane (271113): Unit price impact = 0.4
  - Source: Staff calculation using customs data for 2024.

### Effects on inequality and incidence by income decile
- Gini outcomes:
  - The three policy options (broadening tax base; accounting for tax expenditures; removing import licenses or prepayment tax) have insignificant effect on population income distribution; Gini coefficient for each option equals the benchmark.
  - Benchmark Gini coefficients reported:
    - Market Income (MI): 33.18
    - Disposable Income (DI): 32.99
    - Consumable Income (CI): 33.08 (values shown across scenarios)
  - Overall redistributive impact of the tax system is negligible, with a 0.1-point decrease in the Gini coefficient—from 33.2 for MI to 33.1 for CI in some representations.
  - Direct taxes show moderate progressivity: DI Gini (33.0) lower than MI Gini (33.18).
  - Indirect taxes mildly regressive: CI Gini (33.08) slightly higher than DI Gini (32.99).

- Revenue from broadening the tax base (1.2 billion KMFs) by decile:
  - Top 10% (D10) contribute 0.34 billion KMFs; Bottom 10% (D1) contribute 0.04 billion KMFs.
  - Per capita tax increase by decile (KMFs): D10 = 4,384.3; D1 = 563.2 (intermediate deciles listed in figure).
  - Excise taxes account for 55–58% depending on decile, followed by consumption and prepayment taxes.
  - Although absolute tax contributions are higher for richer deciles, the burden relative to disposable income and percentage change remains broadly uniform across deciles, explaining unchanged Gini.

- Tax expenditures (consumption tax exemptions) incidence:
  - Tax expenditures increase with income; top 10% exempted benefit = 0.43 billion KMFs; bottom 10% = 0.05 billion KMFs.
  - Per capita tax expenditure: Top 10% = 5,477.7 KMFs; Bottom 10% = 595.9 KMFs.
  - Measured as share of disposable income, top deciles benefit proportionally more: 0.31 percent for the top 10 versus 0.26 percent for the bottom 10 (figures reported in text and panels).

### Household survey and fuel consumption patterns
- Household expenditure survey: 5,624 households.
- Table 7: Household Expenditure (selected figures)
  - Number of observations:
    - Rural female-headed: 1,426; Rural male-headed: 2,270
    - Urban female-headed: 715; Urban male-headed: 1,213
  - Average total household expenditure (Rural / Urban):
    - Female headed: 2,473,065 / 2,750,881
    - Male headed: 2,742,845 / 3,253,449
  - Average per capita household expenditure (Rural / Urban):
    - Female headed: 591,123.4 / 766,858.8
    - Male headed: 552,228.3 / 785,726.8
  - Findings:
    - Urban households spend more than rural households.
    - Male-headed households spend more than female-headed households in total household expenditure; per capita gap is less pronounced.
  - Data limitation: unable to conduct individual-level analysis (e.g., compare consumption across individual members or by gender) due to household-level survey design; women’s reported average wage income is 58 % of men’s income in the survey.

- Fuel consumption patterns:
  - Richer households consume more fuel products than poorer ones (average per capita expenditure by income quintile higher for richer households).
  - Kerosene represents the largest budget share among fuel products across income categories.
  - Richer households devote larger spending shares to vehicle fuel and gas for household use.

### Direct and indirect fuel subsidy incidence
- Kerosene direct subsidy:
  - On average, kerosene subsidy as a share of disposable income is higher among higher-income households.
  - In allocation of total kerosene subsidy, richer households receive a larger share.

- Diesel subsidy (indirect effects):
  - Only diesel sold to SONELEC is subsidized; household impacts are indirect via input-output linkages.
  - Simulated scenario: diesel used as intermediate input subsidized by 12.2 percent.
  - Impact on sector prices: electricity, gas and water is the most affected sector (uses diesel as intermediate input more than others).
  - Distributional outcome: indirect subsidy benefits accrue disproportionately to richer households (top decile benefits most), because richer households have larger consumption budgets and consume more of the affected goods.

### Conclusion (summary of main findings)
- Fiscal reforms proposed by the Comorian government are effective at raising revenue but have limited impact on reducing inequality.
  - Broadening the tax base and removing selected tax expenditures estimated to increase government revenues by approximately 2.8 billion KMF (about 0.35 percent of 2024 GDP); modest effect on income inequality (Gini largely unchanged).
  - Broadening the tax base has minimal effect on prices of essential goods (unit price increases generally 1 to 5 percent) but larger effects on non-essential goods: alcohol +15 percent, tobacco +14.3 percent.
  - Reforms are broadly uniform across income distribution when measured as share of disposable income; in absolute terms richer households pay more (e.g., top 10% pay 0.35 billion KMF vs. bottom 10% 0.04 billion KMF under broadened base).
- Current tax exemptions and fuel subsidies are regressive:
  - Tax exemptions disproportionately benefit higher-income households; richest 10 percent receive 0.31 percent of disposable income in benefits vs. poorest 10 percent 0.26 percent.
  - Fuel subsidies (direct and indirect) mainly benefit top income households.

### Policy recommendations
- Targeting tax exemptions:
  - Eliminate consumption tax exemptions for non-essential goods, since these primarily benefit affluent households.
  - Redirect resources from such exemptions to improve equity and fiscal efficiency.
- Reforming fuel subsidies:
  - Phase out fuel subsidies while mitigating adverse distributional effects through direct transfers.
  - International experience suggests targeted transfers are more effective in supporting vulnerable groups without distorting market prices or disproportionately benefiting the wealthy.

*Source: Staff calculations and analysis from the IMF application of the fuel subsidy toolkit and CEQ model as presented in the supplied text.*

### References

### References

### Bibliographic references
- Ahamada Ibrahim, 2024, “State-Owned Enterprises: Oil and Gas in Comoros”, IMF selected Issue paper for Union of Comoros, IMF Country Report No. 24/5.
- Akim Al-Mouksit, Ben Jelloul Mahdi, Czajka Leo, Robilliard Anne-Sophie, 2020, “Collect More, Spend Better? Assessing the Incidence Of Fiscal Systems And Public Spending In Three Francophone West African Countries”, AFD research papers, No. 190.
- Fabrizio Stefania, Goumilevski Alexi, and Kpodar Kangni, 2015, "A New Tool for Distributional Incidence Analysis: An Application to Fuel Subsidy Reform", Technical Notes and Manuals, International Monetary Fund, Washington DC.
- Mokhtari Mohammad Ali and Ghoddusi Hamed, 2025, Fueling Inequality: A Novel Estimate From Large-Scale Reforms, Journal of Environmental Economics and Management.
- Moss Todd, Lambert Caroline, and Majerowicz Stephanie, 2015, Oil to Cash: Fighting the Resource Curse through Cash Transfers”, Center for Global Development, Washington DC.
- Nora Lustig, Jon Jellema, and Valentina Martinez Pabon, 2023, “Are Budget Neutral Income Floors fiscally viable in Sub-Saharan Africa?”, Journal of African Economies.
- Sawadogo Pegdewende Nestor and Kett Benjamin Robert 2024, “Trade Liberalization and Tax Revenue Mobilization in Comoros”, IMF selected Issue paper for Union of Comoros, IMF Country Report No. 24/5.
- World Bank, 2017, “Comoros Poverty Assessment”, Report, Washington DC.

### Figures documented in the Appendix I excerpt
- Figure 1. Revenue in Percent of GDP
  - Caption: Revenue as a Percentage of GDP, 5-year Average
  - Source: Staff calculations using UNU-WIDER Government Revenue Dataset (GDR)
  - Note: Small islands include Cabo Verde, Sao Tome and Principe.
  - Shown series (labels visible): Sub-saharan Africa; Comparable Small Island; Comoros; Asia. Periods shown: 2000–2004, 2005–2009, 2010–2014, 2015–2019, 2020–2022. Visible measures: Total revenue; Tax revenue.
- Figure 2. Contribution of Each Tax to the Impact of Broadening the Tax Base
  - Source: A Staff calculations using UNU-WIDER Government Revenue Dataset (GDR)
  - Note: Small islands include Cabo Verde, Sao Tome and Principe.
  - Vertical axis percent scale shown from 0% to 100%.
  - Tax components or labels visible: Consumption Tax (tc); Excise (dac); Prepayment Tax (ai).
  - Horizontal axis: Disposable Income Deciles labeled D1 D2 D3 D4 D5 D6 D7 D8 D9 D10.

### Table 1 — Union of the Comoros: Average Tax Rates for Selected Sensitive Products
- Source: Staff calculation using customs data for 2024.
- Column headings (as presented): Code SH | Labels | Average tax rate (%) | Custom duties | Consumption tax | RCI | FMG | AI | CCI | PT | Excise
- Entries (preserve exact numeric values and formatting):
  - 0206 | Animal offal (beef, goat, etc.) | 0.0 | 0.0 | 2.5 | 5.0 | 1.0 | 1.0 | 1.0 | 10.0
  - 0207 | Poultry meat and offal | 0.0 | 0.0 | 2.5 | 5.0 | 1.0 | 1.0 | 1.0 | 10.0
  - 04029910 | Dairy products (milk, cream, etc.) | 0.0 | 0.0 | 2.5 | 5.0 | 1.0 | 1.0 | 1.0 | 10.0
  - 07 | Vegetables, roots, tubers | 5.0 | 0.0 | 0.3 | 5.0 | 1.0 | 1.0 | 1.0 | 15.0
  - 1101 | Wheat or meslin flour | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 3.0
  - 151620 | Vegetable oil | 5.0 | 10.0 | 2.5 | 5.0 | 1.0 | 1.0 | 1.0 | 5.0
  - 1604 | Canned sardines | 11.3 | 10.0 | 1.5 | 5.0 | 1.0 | 1.0 | 1.0 | 10.4
  - 1701 | Sugar | 0.0 | 0.0 | 2.2 | 5.0 | 1.0 | 1.0 | 1.0 | 10.0
  - 220210 | Water, non-alcoholic beverages | 5.0 | 10.0 | 0.8 | 5.0 | 1.0 | 1.0 | 1.0 | 15.0
  - 250100 | Salt | (blank) | Non-food 20.0 | 10.0 | 2.3 | 5.0 | 1.0 | 1.0 | 1.0 | 352.0
  - 220300+220830 | Alcohol | 20.0 | 9.4 | 0.3 | 5.0 | 1.0 | 1.0 | 1.0 | 302.0
  - 24 | Tobacco | 4.1 | 0.0 | 1.4 | 4.6 | 0.9 | 0.9 | 0.8 | 11.7
  - 252329 + 252390 | Ordinary or colored cement | 0.0 | 5.0 | 1.9 | 0.0 | 0.0 | 0.0 | 1.0 | 15.0
  - 271113 | Butane | 0.0 | 0.0 | 1.3 | 5.0 | 1.0 | 1.0 | 1.0 | 5.0
  - 30 | Pharmaceutical products | 0.0 | 0.0 | 2.0 | 5.0 | 0.0 | 0.0 | 0.0 | 5.0

*UNION OF THE COMOROS — INTERNATIONAL MONETARY FUND*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2026/english/1comea2026002-source-pdf.pdf_