## 2.   Comparison Between the Observed Taxable Income and the Theoretical Taxable Income

## Source details

**Canonical URL:** [2.   Comparison Between the Observed Taxable Income and the Theoretical Taxable Income](https://www.imf.org/-/media/files/publications/selected-issues-papers/2026/english/sipea2026016.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/selected-issues-papers/2026/english/sipea2026016.pdf.md)
- [Structured JSON version](/-/media/files/publications/selected-issues-papers/2026/english/sipea2026016.pdf.json)

---

### Key methods and validation
- Framework: Commitment to Equity (CEQ) approach using the 2020 household survey and microsimulation techniques to simulate Market Income (MI), Disposable Income (DI), and Consumable Income (CI).
- DI proxied by total consumption expenditure (including imputed rent and self-consumption).
- MI obtained by adding direct taxes (PIT and SPT) back to DI.
- CI computed by subtracting indirect taxes from DI and adding subsidies.
- Validation:
  - Equation (1) for imputing taxable wages reproduces the Excel tool used by the tax administration for computing personal income tax; Figure 2 shows an exact correspondence between the Excel tool and Equation (1) when applied to simulated net wages.
  - Formulas (3) and (4) used to replicate CIF values from tax-inclusive import values; Figure 3 compares observed CIF values and theoretical CIF values and provides internal consistency checks using comprehensive customs data.

### Taxable income imputation and tax schedule
- Tax schedule (Taxable income ranges, Net income ranges, Taxation rates (%), Cumulative tax (KMFs)):
  - r=1: Taxable income 0 – 150 000; Net income 0 – 214 285; Taxation rates 0; Cumulative tax 0
  - r=2: Taxable income 150 001 – 500 000; Net income 214 286 – 696 787; Taxation rates 5; Cumulative tax 17 500
  - r=3: Taxable income 500 001 – 1 000 000; Net income 696 788 – 1 361 073; Taxation rates 10; Cumulative tax 67 500
  - r=4: Taxable income 1 000 001-1 500 000; Net income 1 361 074 – 2 000 359; Taxation rates 15; Cumulative tax 142 500
  - r=5: Taxable income 1 500 001-2 500 000; Net income 2 000 360 – 3 228 931; Taxation rates 20; Cumulative tax 342 500
  - r=6: Taxable income 2 500 001-3 500 000; Net income 3 228 932 – 4 407 502; Taxation rates 25; Cumulative tax 592 499
  - r=7: Taxable income 3 500 001 and plus; Net income 4 407 503 and plus; Taxation rates 30

### Indirect taxes, specific unit taxes, and CIF validation
- Method: Compute pre-tax consumer expenditures by backing out ad-valorem and specific taxes consistent with 2024 tax code; indirect taxes include specific unit taxes on oil products and rice; implicit ad valorem rates derived from selling prices and applicable unit taxes.
- Implicit tax rates and national selling prices:
  - Ordinary rice (kmf/kg) specific tax: 25; National selling price: Ordinary rice 313.3; National implicit tax rate: Ordinary rice 8.7%
  - Premium rice (kmf/kg) specific tax: 200; National selling price: Premium rice 1000.8; National implicit tax rate: Premium rice 25.0%
  - Gasoline specific tax: 230; National selling price: Gasoline 750; National implicit tax rate: Gasoline 44.2%
  - Diesel specific tax: 115; National selling price: Diesel 650; National implicit tax rate: Diesel 21.5%
  - Kerosene specific tax: 0; National selling price: Kerosene 500; National implicit tax rate: Kerosene 0.0%
- Regional selling prices and implicit rates reported for Anjouan, Moheli, Grande-Comores (values preserved in Table 2).
- CIF validation: Formulas (3) and (4) replicate CIF values from tax-inclusive import values; Figure 3 compares observed and theoretical CIF values using comprehensive customs data.

### Microsimulation calibration and revenue reconciliation
- CEQ-simulated government revenue for 2019–2021: 33.8–38.9 billion KMFs for 2019–2021 (Table 3).
- Micro-simulation reproduces 59 percent of total revenue in 2021 and 65 percent in 2020, excluding corporate income, capital, and export taxes, and certain non-tax revenues.
- Micro-simulated aggregates (billion KMFs):
  - Direct taxes (i): 3.2 (2019), 3.0 (2020), 3.5 (2021)
    - Personal income tax: 3.1 (2019), 2.9 (2020), 3.5 (2021)
    - Single professional tax: 0.1 (2019), 0.0 (2020), 0.0 (2021)
  - Indirect taxes (ii): 27.4 (2019), 30.9 (2020), 32.2 (2021)
    - Custom duties + Excise: 12.1 (2019), 13.8 (2020), 15.9 (2021)
      - Custom duties: 5.2 (2019)
      - Excise: 7.0 (2019)
    - Import license: 0.7 (2019), 0.6 (2020), 0.4 (2021)
    - Import prepayment tax: 0.7 (2019), 0.5 (2020), 0.2 (2021)
    - Special custom duties: 4.5 (2019), 8.8 (2020), 8.7 (2021)
      - Oil products: 1.4 (2019), 6.8 (2020), 7.2 (2021)
      - Rice: 2.8 (2019), 2.1 (2020), 1.5 (2021)
    - Administrative revenue: 5.7 (2019), 2.6 (2020), 2.9 (2021)
    - Consumption taxes: 3.7 (2019), 5.7 (2020), 4.9 (2021)
      - Imports: 2.8 (2019), 2.1 (2020), 1.0 (2021)
      - Domestic: 0.9 (2019), 3.6 (2020), 4.0 (2021)
  - Total (i) + (ii): 30.6 (2019), 33.8 (2020), 33.0 (2021)
  - Total government revenue (Government statement operations): 50.0 (2019), 49.4 (2020), 55.0 (2021)
- Explanations for discrepancies:
  - Customs duties differences stem from imports not directly consumed by households (e.g., oil used as intermediate input in electricity, transport, bakery).
  - Domestic consumption taxes underestimated because survey omits major taxable items (financial services, construction, incoming phone call charges).

### Fuel subsidies, SCH losses, and price structure
- Context: SCH monopolizes importation and distribution; selling prices to service stations set by ministerial decree; no clear formula for setting selling price; authorities maintain SCH selling prices to preserve household purchasing power.
- Financial losses and magnitude:
  - In 2024, financial losses associated with underpricing petroleum products to gas stations amounted to 3.0 billion KMFs (preliminary SCH financial statements for December 2024).
  - Two products underpriced relative to average cost:
    - Kerosene for household consumption underpriced by 19.9 percent (loss of -1.1 billion KMFs in 2024).
    - Diesel supplied to SONELEC underpriced by 12.2 percent (loss of -1.9 billion KMFs in 2024).
  - SCH losses reached 14.5 billion KMFs in 2022, equivalent to 2.5 percent of GDP (Ahamada, 2024); diesel sold to SONELEC accounted for 63 percent of losses in 2024 and 45 percent in 2022.
  - Exemption detail: SONELEC is exempt from the excise tax on diesel (115 KMF per liter), representing an additional tax expenditure by the government.
- SCH fuel price structure and margins in 2024 (Table 4 values):
  - Kerosene for households: 2024 Monthly average cost price [I] 393.4 KMF/liter; Selling price to gas station [II] 315 KMF/liter; SCH Margin [III] -1.1 (Billion KMFs); % change in price [(II – I)/I] -19.9
  - Diesel - SONELEC: [I] 393.1 KMF/liter; [II] 345 KMF/liter; [III] -1.9 (Billion KMFs); % change -12.2
  - Gasoline: [I] 638.5 KMF/liter; [II] 725 KMF/liter; [III] 1.8 (Billion KMFs); % change 13.5
  - Diesel - Tourism: [I] 508.4 KMF/liter; [II] 630 KMF/liter; [III] 3.0 (Billion KMFs); % change 23.9

### Policy-relevant findings and implications
- Validation results:
  - The imputation method for deriving taxable personal income from reported net wages is validated through exact correspondence with the tax administration Excel tool.
  - CIF reconstruction formulas provide internal consistency checks for customs valuation using comprehensive customs data.
- Revenue mobilization and microsimulation limits:
  - Micro-simulation reproduces a large share but not all government revenue (59 percent in 2021; 65 percent in 2020); exclusion of corporate, capital, export taxes, and some non-tax revenues explains remaining gaps.
  - Household surveys underrepresent imports and certain taxable domestic services, leading to underestimation of customs duties and domestic consumption taxes in microsimulations.
- Fuel subsidies:
  - Significant fiscal costs arise from underpriced petroleum products sold by SCH, with concrete losses in 2024 and historically large losses in 2022 (14.5 billion KMFs, 2.5 percent of GDP).
  - Diesel subsidy to SONELEC not only imposes direct SCH losses but also constitutes a tax expenditure via excise exemptions (115 KMF per liter).
  - Kerosene subsidy directly benefits household consumers but represents a sizeable implicit fiscal cost to the state-owned enterprise.

### IMF fuel subsidy toolkit: methodology and key formulas
- Toolkit assesses:
  - (1) the change in total spending required to maintain the same basket of goods, and
  - (2) the distribution of subsidies across different households.
- Calculations assume household consumption patterns remain unchanged.
- Direct impact (single good c):
  - dY/dbb / Y = T_b * (dP_b / P_b)
  - where T_b is the spending share on product c, and dP_i / P_i is the percentage change in the price of product c due to subsidy.
- Direct impact (multiple goods):
  - dY/dir / Y = Σ_b T_b * (dP_i / P_i)
- Indirect impact (through intermediate-input linkages):
  - dY_bbdd / Y = Σ_j T_j * (dP_j / P_j)
  - where j indexes goods whose prices vary due to use of good c as an intermediate input.
- Implementation detail: consumption categories from household survey are matched to corresponding industries from the input-output table.
- Caveat: assuming unchanged consumption patterns could distort results; if lower income households have higher price elasticity, unequal impacts between richer and poorer households are closer to an upper bound.

### Fiscal and tax-base broadening results
- Circular note implementation (September 2025) expected additional revenue: 1.2 billion KMFs equivalent to 0.18 percent of GDP.
  - Excise taxes: about 58 percent of the additional revenue = 0.7 billion KMFs.
  - Consumption taxes on imports: 33 percent.
  - Prepayment tax: 8 percent.
- Total fiscal cost of tax expenditures (defined as revenue foregone if products were subject to a 10 percent consumption tax and excise rate less than 10 percent): 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) would lead to revenue loss: 1.4 billion KMFs (0.24 percent of 2024 GDP), equally split between import licenses and prepayment tax.
- Comparable estimate by Sawadogo and Kett (2024): 1.7 billion KMFs.

### Table 5 key figures (Impact of Broadening the Tax Base and Tax Expenditure on Government Budget)
- Benchmark totals and impacts (billion KMFs):
  - Direct taxes (i): Benchmark 3.2; Impact 0.0
    - Personal income tax: Benchmark 3.1; Impact -
    - Single professional tax: Benchmark 0.1; Impact -
  - Indirect taxes (ii): Benchmark 27.4; Impact 1.2
    - Custom duties + Excise: Benchmark 12.2; Impact 0.7
      - Custom duties: Benchmark 5.2; Impact -
      - Excise: Benchmark 7.0; Impact 0.7
    - Import license (pt): Benchmark 0.7; Impact -
    - Import prepayment tax (ai): Benchmark 0.7; Impact 0.1
    - Special custom duties: Benchmark 4.5; Impact 0.0
      - Oil products (Gasoline, diesel): Benchmark 1.4; Impact -
      - Rice: Benchmark 2.8; Impact -
      - Vehicle: Benchmark 0.3; Impact -
  - Administrative revenues (rau, rci, cci): Benchmark 5.7; Impact 0.0
  - Consumption taxes: Benchmark 3.7; Impact 0.4
    - Imports: Benchmark 2.8; Impact 0.4
    - Domestic: Benchmark 0.9; Impact -
  - Total revenue (i) + (ii): Benchmark 30.7; Impact 1.2
  - Fiscal Cost of Tax Expenditures (iii): 1.6
    - Kerosene for household use (45 KMFs/liter): - 0.6
    - Exemption on 10% consumption tax: - 1.0

### Table 6 selected product impacts (unit price at customs %, CIF, and revenue impact)
- Food product examples:
  - Animal offal (0206): Unit price +1.0; CIF value 481.9 Million KMF; Tax revenue 98.8 Million KMF; Impact on revenue 5.6 Million KMF.
  - Poultry meat and offal (0207): Unit price +1.0; CIF value 6,442.8 Million KMF; Tax revenue 1,320.8 Million KMF; Impact on revenue 74.4 Million KMF.
  - Vegetables, roots (07): Unit price +5.8; CIF value 231.0 Million KMF; Tax revenue 54.4 Million KMF; Impact on revenue 16.5 Million KMF.
  - Root crops example: impact up to 5.8 percent.
- Non-food examples:
  - Alcohol (220300 + 220830): Unit price +15.0; CIF value 136.1 Million KMF; Tax revenue 632.5 Million KMF; Impact on revenue 115.0 Million KMF.
  - Tobacco (24): Unit price +14.3; CIF value 30.1 Million KMF; Tax revenue 121.9 Million KMF; Impact on revenue 24.5 Million KMF.
  - Cement (252329 + 252390): Unit price +0.4; CIF value 7,449.7 Million KMF; Tax revenue 1,811.8 Million KMF; Impact on revenue 36.2 Million KMF.
  - Butane (271113): Unit price +0.4; CIF value 293.4 Million KMF; Tax revenue 39.8 Million KMF; Impact on revenue 11.9 Million KMF.
- Source for Table 6: Staff calculation using customs data for 2024.

### Distributional impacts and inequality metrics
- Gini coefficients across scenarios:
  - Benchmark Gini (Disposable Income): 33.1
  - Broadening the tax base, accounting for tax expenditures, or removing import licenses or prepayment tax: Gini remains 33.1
  - Reported values across components:
    - Market Income (MI): 33.18
    - Disposable Income (DI): 32.99
    - Consumable Income (CI): 33.08
  - Overall redistributive impact of the tax system: 0.1-point decrease in the Gini coefficient (from 33.2 for MI to 33.1 for CI).
  - 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 income:
  - Top 10% contributes 0.34 billion KMFs.
  - Bottom 10% contributes 0.04 billion KMFs.
  - Per capita tax increase:
    - Top 10%: 4,384 KMFs per capita.
    - Bottom 10%: 563 KMFs per capita.
  - Excise taxes contribution: 55–58% depending on decile.
- Tax expenditures incidence:
  - Volume by decile (billion KMFs): bottom deciles low up to top 10% = 0.43 billion KMFs (Panel A).
  - Per capita tax expenditure:
    - Top 10%: 5,477 KMFs per capita.
    - Bottom 10%: 596 KMFs per capita.
  - In percent of disposable income:
    - Top 10% benefit 0.31 percent of disposable income.
    - Bottom 10% benefit 0.26 percent of disposable income.
  - Conclusion: tax expenditures are regressive; higher-income groups benefit disproportionately more.

### Household expenditure and fuel consumption patterns
- Household survey:
  - Number of households in survey: 5,624 households.
  - Household distribution:
    - Rural female-headed observations: 1,426
    - Rural male-headed observations: 2,270
    - Urban female-headed observations: 715
    - Urban male-headed observations: 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
  - Finding: urban households spend more than rural households; male-headed households spend more than female-headed households; per capita gap between female- and male-headed households is less pronounced.
  - Data limitation: unable to conduct individual-level analysis; women’s reported average wage income is 58 % of men’s income.
- Fuel consumption by income:
  - Richer households consume more fuel products than poorer ones.
  - Kerosene represents the largest budget share among fuel products across income categories.
  - Richer households devote a larger share of spending to vehicle fuel and gas for household use.

### Fuel subsidy simulations and incidence
- Kerosene subsidy:
  - Kerosene direct subsidy as a share of disposable income is higher among higher-income households.
  - Allocation of total kerosene subsidy: richer households receive a larger share.
- Diesel subsidy scenario:
  - Simulated scenario: only diesel used as an intermediate input is subsidized by 12.2 percent.
  - Sectoral price impacts: Electricity, gas and water is the most affected sector, since it uses diesel as an intermediate input more than other sectors.
  - Distributional impact: indirect subsidy benefits households in the top decile most; richer households benefit more in absolute terms because they have larger consumption budgets.

### Conclusions (section D highlights)
- Fiscal reforms effectiveness:
  - 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).
  - These reforms have limited effect on reducing inequality; Gini coefficient remains largely unchanged across reform scenarios.
- Effects on essential vs. non-essential goods:
  - Unit price increases for essential goods: range from 1 to 5 percent (examples: poultry/beef 1 percent; root crops up to 5.8 percent).
  - Non-essential goods experience much higher price increases: alcohol +15.0 percent; tobacco +14.3 percent.
  - Interpretation: reforms spare necessities while targeting items with negative externalities, strengthening Pigouvian properties of the tax system.
- Contribution to reforms by income group:
  - When measured as a share of disposable income, reforms affect households uniformly across income distribution.
  - In absolute terms, top 10 percent pay 0.35 billion KMF vs bottom 10 percent pay 0.04 billion KMF under broadened tax base.
- Regressivity of current exemptions and subsidies:
  - Tax exemptions disproportionately benefit higher-income population; richest 10 percent receive 0.31 percent of disposable income in benefits vs poorest 10 percent receive 0.26 percent.
  - Fuel subsidies (direct and indirect) mainly accrue to top income households.

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

*Source: sipea2026016 - 2.   Comparison Between the Observed Taxable Income and the Theoretical Taxable Income*

### 2.   Comparison Between the Observed Taxable Income and the Theoretical Taxable

### 2.   Comparison Between the Observed Taxable Income and the Theoretical Taxable Income

### Key methods and validation
- Framework: Commitment to Equity (CEQ) approach using the 2020 household survey and microsimulation techniques to simulate income concepts: Market Income (MI), Disposable Income (DI), and Consumable Income (CI).
- DI is proxied by total consumption expenditure (including imputed rent and self-consumption).
- MI is obtained by adding direct taxes (PIT and SPT) back to DI.
- CI is computed by subtracting indirect taxes from DI and adding subsidies.
- Validation: Equation (1) for imputing taxable wages reproduces the Excel tool used by the tax administration for computing personal income tax; Figure 2 shows an exact correspondence between the Excel tool and Equation (1) when applied to simulated net wages.

### Taxable income imputation and tax schedule (exact values preserved)
- Tax schedule used (Taxable income ranges, Net income ranges, Taxation rates (%), Cumulative tax (KMFs)):
  - r=1: Taxable income 0 – 150 000; Net income 0 – 214 285; Taxation rates 0; Cumulative tax 0
  - r=2: Taxable income 150 001 – 500 000; Net income 214 286 – 696 787; Taxation rates 5; Cumulative tax 17 500
  - r=3: Taxable income 500 001 – 1 000 000; Net income 696 788 – 1 361 073; Taxation rates 10; Cumulative tax 67 500
  - r=4: Taxable income 1 000 001-1 500 000; Net income 1 361 074 – 2 000 359; Taxation rates 15; Cumulative tax 142 500
  - r=5: Taxable income 1 500 001-2 500 000; Net income 2 000 360 – 3 228 931; Taxation rates 20; Cumulative tax 342 500
  - r=6: Taxable income 2 500 001-3 500 000; Net income 3 228 932 – 4 407 502; Taxation rates 25; Cumulative tax 592 499
  - r=7: Taxable income 3 500 001 and plus; Net income 4 407 503 and plus; Taxation rates 30

### Indirect taxes, specific unit taxes, and CIF validation
- Method: Compute pre-tax consumer expenditures by backing out ad-valorem and specific taxes consistent with 2024 tax code. Indirect taxes include specific unit taxes on oil products and rice; implicit ad valorem rates derived from selling prices and applicable unit taxes.
- Implicit tax rates and selling prices (exact values preserved):
  - Ordinary rice (kmf/kg) specific tax: 25; Premium rice (kmf/kg) specific tax: 200; Gasoline specific tax: 230; Diesel specific tax: 115; Kerosene specific tax: 0
  - 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%
  - Regional selling prices and implicit rates reported for Anjouan, Moheli, Grande-Comores (values preserved in Table 2).
- CIF validation: Formulas (3) and (4) used to replicate CIF values from tax-inclusive import values; Figure 3 compares observed CIF values and theoretical CIF values and provides internal consistency checks using comprehensive customs data.

### Microsimulation calibration and revenue reconciliation (exact figures)
- CEQ-simulated government revenue for 2019–2021: 33.8–38.9 billion KMFs for 2019–2021 (Table 3).
- Micro-simulation reproduces 59 percent of total revenue in 2021 and 65 percent in 2020, excluding corporate income, capital, and export taxes, and certain non-tax revenues.
- Micro-simulated aggregates (billion KMFs, exact values from Table 3):
  - Direct taxes (i): 3.2 (2019), 3.0 (2020), 3.5 (2021)
    - Personal income tax: 3.1 (2019), 2.9 (2020), 3.5 (2021)
    - Single professional tax: 0.1 (2019), 0.0 (2020), 0.0 (2021)
  - Indirect taxes (ii): 27.4 (2019), 30.9 (2020), 32.2 (2021)
    - Custom duties + Excise: 12.1 (2019), 13.8 (2020), 15.9 (2021)
      - Custom duties: 5.2 (2019)
      - Excise: 7.0 (2019)
    - Import license: 0.7 (2019), 0.6 (2020), 0.4 (2021)
    - Import prepayment tax: 0.7 (2019), 0.5 (2020), 0.2 (2021)
    - Special custom duties: 4.5 (2019), 8.8 (2020), 8.7 (2021)
      - Oil products: 1.4 (2019), 6.8 (2020), 7.2 (2021)
      - Rice: 2.8 (2019), 2.1 (2020), 1.5 (2021)
    - Administrative revenue: 5.7 (2019), 2.6 (2020), 2.9 (2021)
    - Consumption taxes: 3.7 (2019), 5.7 (2020), 4.9 (2021)
      - Imports: 2.8 (2019), 2.1 (2020), 1.0 (2021)
      - Domestic: 0.9 (2019), 3.6 (2020), 4.0 (2021)
  - Total (i) + (ii): 30.6 (2019), 33.8 (2020), 33.0 (2021)
  - Total government revenue (Government statement operations): 50.0 (2019), 49.4 (2020), 55.0 (2021)

- Explanations for discrepancies:
  - Customs duties differences stem from imports not directly consumed by households (e.g., oil used as intermediate input in electricity, transport, bakery).
  - Domestic consumption taxes underestimated because survey omits major taxable items (financial services, construction, incoming phone call charges).

### Fuel subsidies, SCH losses, and price structure (exact figures)
- Context: SCH (state-owned Comorian Hydrocarbons Company) monopolizes importation and distribution; selling prices to service stations set by ministerial decree; no clear formula for setting selling price; authorities maintain SCH selling prices to preserve household purchasing power.
- Financial losses and magnitude:
  - In 2024, financial losses associated with underpricing petroleum products to gas stations amounted to 3.0 billion KMFs (preliminary SCH financial statements for December 2024).
  - Two products underpriced relative to average cost:
    - Kerosene for household consumption underpriced by 19.9 percent (loss of -1.1 billion KMFs in 2024).
    - Diesel supplied to SONELEC underpriced by 12.2 percent (loss of -1.9 billion KMFs in 2024).
  - SCH losses reached 14.5 billion KMFs in 2022, equivalent to 2.5 percent of GDP (Ahamada, 2024); diesel sold to SONELEC accounted for 63 percent of losses in 2024 and 45 percent in 2022.
  - Exemption detail: SONELEC is exempt from the excise tax on diesel (115 KMF per liter), representing an additional tax expenditure by the government.
- SCH fuel price structure and margins in 2024 (exact values from Table 4):
  - Kerosene for households: 2024 Monthly average cost price [I] 393.4 KMF/liter; Selling price to gas station [II] 315 KMF/liter; SCH Margin [III] -1.1 (Billion KMFs); % change in price [(II – I)/I] -19.9
  - Diesel - SONELEC: [I] 393.1 KMF/liter; [II] 345 KMF/liter; [III] -1.9 (Billion KMFs); % change -12.2
  - Gasoline: [I] 638.5 KMF/liter; [II] 725 KMF/liter; [III] 1.8 (Billion KMFs); % change 13.5
  - Diesel - Tourism: [I] 508.4 KMF/liter; [II] 630 KMF/liter; [III] 3.0 (Billion KMFs); % change 23.9

### Policy-relevant findings and implications (drawn from analysis in the chapter)
- Validation results:
  - The imputation method for deriving taxable personal income from reported net wages is validated through exact correspondence with the tax administration Excel tool.
  - CIF reconstruction formulas provide internal consistency checks for customs valuation using comprehensive customs data.
- Revenue mobilization and microsimulation limits:
  - Micro-simulation reproduces a large share but not all government revenue (59 percent in 2021; 65 percent in 2020); exclusion of corporate, capital, export taxes, and some non-tax revenues explains remaining gaps.
  - Household surveys underrepresent imports and certain taxable domestic services, leading to underestimation of customs duties and domestic consumption taxes in microsimulations.
- Fuel subsidies:
  - Significant fiscal costs arise from underpriced petroleum products sold by SCH, with concrete losses in 2024 and historically large losses in 2022 (14.5 billion KMFs, 2.5 percent of GDP).
  - Diesel subsidy to SONELEC not only imposes direct SCH losses but also constitutes a tax expenditure via excise exemptions (115 KMF per liter).
  - Kerosene subsidy directly benefits household consumers but represents a sizeable implicit fiscal cost to the state-owned enterprise.

*Source: sipea2026016 - 2.   Comparison Between the Observed Taxable Income and the Theoretical Taxable Income*

### 13.      The IMF fuel subsidy toolkit assesses the impact of fuel subsidies on two dimensions:

### 13.      The IMF fuel subsidy toolkit assesses the impact of fuel subsidies on two dimensions:

### Methodology and Key Formulas
- The toolkit assesses two dimensions: (1) the change in total spending required to maintain the same basket of goods, and (2) the distribution of subsidies across different households. These calculations assume that household consumption patterns remain unchanged in response to subsidies.
- Direct impact (single good c):
  - dY/dbb / Y = T_b * (dP_b / P_b)
  - where T_b is the spending share on product c, and dP_i / P_i is the percentage change in the price of product c due to subsidy.
- Direct impact (multiple goods):
  - dY/dir / Y = Σ_b T_b * (dP_i / P_i)
- Indirect impact (through intermediate-input linkages):
  - dY_bbdd / Y = Σ_j T_j * (dP_j / P_j)
  - where j indexes goods whose prices vary due to use of good c as an intermediate input.
- Implementation detail: consumption categories from household survey are matched to corresponding industries from the input-output table.
- Caveat noted: the simplifying assumption that household consumption patterns remain unchanged could potentially distort results; if lower income households have higher price elasticity, unequal impacts between richer and poorer households are closer to an upper bound.

### Fiscal and Tax-Base Broadening Results (selected findings)
- Circular note implementation (September 2025) expected additional revenue: 1.2 billion KMFs equivalent to 0.18 percent of GDP.
  - Excise taxes: about 58 percent of the additional revenue = 0.7 billion KMFs.
  - Consumption taxes on imports: 33 percent.
  - Prepayment tax: 8 percent.
- Total fiscal cost of tax expenditures (defined as revenue foregone if products were subject to a 10 percent consumption tax and excise rate less than 10 percent): 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) would lead to revenue loss: 1.4 billion KMFs (0.24 percent of 2024 GDP), equally split between import licenses and prepayment tax.
- Comparable estimate by Sawadogo and Kett (2024): 1.7 billion KMFs.

### Table 5 key figures (Impact of Broadening the Tax Base and Tax Expenditure on Government Budget)
- Benchmark totals and impacts (billion KMFs):
  - Direct taxes (i): Benchmark 3.2; Impact 0.0
    - Personal income tax: Benchmark 3.1; Impact -
    - Single professional tax: Benchmark 0.1; Impact -
  - Indirect taxes (ii): Benchmark 27.4; Impact 1.2
    - Custom duties + Excise: Benchmark 12.2; Impact 0.7
      - Custom duties: Benchmark 5.2; Impact -
      - Excise: Benchmark 7.0; Impact 0.7
    - Import license (pt): Benchmark 0.7; Impact -
    - Import prepayment tax (ai): Benchmark 0.7; Impact 0.1
    - Special custom duties: Benchmark 4.5; Impact 0.0
      - Oil products (Gasoline, diesel): Benchmark 1.4; Impact -
      - Rice: Benchmark 2.8; Impact -
      - Vehicle: Benchmark 0.3; Impact -
  - Administrative revenues (rau, rci, cci): Benchmark 5.7; Impact 0.0
  - Consumption taxes: Benchmark 3.7; Impact 0.4
    - Imports: Benchmark 2.8; Impact 0.4
    - Domestic: Benchmark 0.9; Impact -
  - Total revenue (i) + (ii): Benchmark 30.7; Impact 1.2
  - Fiscal Cost of Tax Expenditures (iii): 1.6
    - Kerosene for household use (45 KMFs/liter): - 0.6
    - Exemption on 10% consumption tax: - 1.0

### Table 6 selected product impacts (impact on unit price at customs and revenue)
- Food product examples (Unit price impact at customs %):
  - Animal offal (beef, goat, etc.) (0206): Unit price +1.0; CIF value 481.9 Million KMF; Tax revenue 98.8 Million KMF; Impact on revenue 5.6 Million KMF.
  - Poultry meat and offal (0207): Unit price +1.0; CIF value 6,442.8 Million KMF; Tax revenue 1,320.8 Million KMF; Impact on revenue 74.4 Million KMF.
  - Vegetables, roots (07): Unit price +5.8; CIF value 231.0 Million KMF; Tax revenue 54.4 Million KMF; Impact on revenue 16.5 Million KMF.
  - Root crops example: impact up to 5.8 percent (as noted in prose).
- Non-food examples:
  - Alcohol (220300 + 220830): Unit price +15.0; CIF value 136.1 Million KMF; Tax revenue 632.5 Million KMF; Impact on revenue 115.0 Million KMF.
  - Tobacco (24): Unit price +14.3; CIF value 30.1 Million KMF; Tax revenue 121.9 Million KMF; Impact on revenue 24.5 Million KMF.
  - Cement (252329 + 252390): Unit price +0.4; CIF value 7,449.7 Million KMF; Tax revenue 1,811.8 Million KMF; Impact on revenue 36.2 Million KMF.
  - Butane (271113): Unit price +0.4; CIF value 293.4 Million KMF; Tax revenue 39.8 Million KMF; Impact on revenue 11.9 Million KMF.
- Source for Table 6: Staff calculation using customs data for 2024.

### Distributional Impacts and Inequality Metrics
- Gini coefficients across scenarios:
  - Benchmark Gini (Disposable Income): 33.1
  - Broadening the tax base, accounting for tax expenditures, or removing import licenses or prepayment tax: Gini remains 33.1 (i.e., no change).
  - Reported values across components:
    - Market Income (MI): 33.18
    - Disposable Income (DI): 32.99
    - Consumable Income (CI): 33.08 (examples in Figure 4)
  - Overall redistributive impact of the tax system: 0.1-point decrease in the Gini coefficient (from 33.2 for MI to 33.1 for CI).
  - 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 income:
  - Top 10% contributes 0.34 billion KMFs.
  - Bottom 10% contributes 0.04 billion KMFs.
  - Per capita tax increase:
    - Top 10%: 4,384 KMFs per capita.
    - Bottom 10%: 563 KMFs per capita.
  - Excise taxes contribution: 55–58% depending on decile.

- Tax expenditures incidence:
  - Volume by decile (billion KMFs): bottom deciles low up to top 10% = 0.43 billion KMFs (Panel A).
  - Per capita tax expenditure:
    - Top 10%: 5,477 KMFs per capita.
    - Bottom 10%: 596 KMFs per capita.
  - In percent of disposable income:
    - Top 10% benefit 0.31 percent of disposable income.
    - Bottom 10% benefit 0.26 percent of disposable income.
  - Conclusion: tax expenditures are regressive; higher-income groups benefit disproportionately more.

### Household Expenditure and Fuel Consumption Patterns
- Household expenditure survey details:
  - Number of households in survey: 5,624 households.
  - Household distribution (Table 7):
    - Rural female-headed observations: 1,426
    - Rural male-headed observations: 2,270
    - Urban female-headed observations: 715
    - Urban male-headed observations: 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
  - Finding: urban households spend more than rural households; male-headed households spend more than female-headed households; per capita gap between female- and male-headed households is less pronounced.
  - Data limitation: unable to conduct individual-level analysis; women’s reported average wage income is 58 % of men’s income.

- Fuel consumption by income:
  - Richer households consume more fuel products than poorer ones.
  - Kerosene represents the largest budget share among fuel products across income categories.
  - Richer households devote a larger share of spending to vehicle fuel and gas for household use.

### Fuel Subsidy Simulations and Incidence
- Kerosene subsidy:
  - Kerosene direct subsidy as a share of disposable income is higher among higher-income households.
  - Allocation of total kerosene subsidy: richer households receive a larger share.

- Diesel subsidy scenario:
  - Simulated scenario: only diesel used as an intermediate input is subsidized by 12.2 percent.
  - Sectoral price impacts (Figure 9, left panel): Electricity, gas and water is the most affected sector, since it uses diesel as an intermediate input more than other sectors.
  - Distributional impact (Figure 9, right panel): indirect subsidy benefits households in the top decile most; richer households benefit more in absolute terms because they have larger consumption budgets.

### Conclusions (section D highlights)
- Fiscal reforms effectiveness:
  - 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).
  - These reforms have limited effect on reducing inequality; Gini coefficient remains largely unchanged across reform scenarios.
- Effects on essential vs. non-essential goods:
  - Unit price increases for essential goods: range from 1 to 5 percent (examples: poultry/beef 1 percent; root crops up to 5.8 percent).
  - Non-essential goods experience much higher price increases: alcohol +15.0 percent; tobacco +14.3 percent.
  - Interpretation: reforms spare necessities while targeting items with negative externalities, strengthening Pigouvian properties of the tax system.
- Contribution to reforms by income group:
  - When measured as a share of disposable income, reforms affect households uniformly across income distribution.
  - In absolute terms, top 10 percent pay 0.35 billion KMF vs bottom 10 percent pay 0.04 billion KMF under broadened tax base.

- Regressivity of current exemptions and subsidies:
  - Tax exemptions disproportionately benefit higher-income population; richest 10 percent receive 0.31 percent of disposable income in benefits vs poorest 10 percent receive 0.26 percent.
  - Fuel subsidies (direct and indirect) mainly accrue to top income households.

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

*Source: Staff calculations.*

---


_Source: https://www.imf.org/-/media/files/publications/selected-issues-papers/2026/english/sipea2026016.pdf_