## _cr06385

## Source details

**Canonical URL:** [_cr06385](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2006/_cr06385.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2006/_cr06385.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2006/_cr06385.pdf.json)

---

### Remittances: overview, origins, channels, and uses
- Remittances are central to the Comorian economy:
  - Remittances in 2005: 18½ percent of GDP.
  - Remittances share of current account receipts: almost 50 percent.
- Origins and scale:
  - Estimated diaspora living abroad: between 150,000 to 200,000 people.
  - Diaspora distribution: 80 percent in Mayotte or France (2003 household census).
- Transmission channels (1999–2005):
  - Remittances composition: about two-thirds cash; about one-fourth via wire transfers.
  - Remittances sent directly in the form of goods have also been increasing.
  - Grande Comore (Ngazidja) receives by far the highest share of remittances.
- Primary uses:
  - Remittances are mainly spent on imported goods.
  - Remittances finance about 70 percent of total imports of goods and more than half of all imports.
  - Anecdotal uses: Grand Marriage ceremonies and housing construction; annual data indicate most remittances finance consumption of imported goods.

### Determinants and macroeconomic effects
- Motives and cyclicality:
  - Main motive: altruistic transfers.
  - Inflows tend to be countercyclical to real GDP growth and export performance; remittances rose in 2001 and 2005 when they offset export sector losses.
- External balances and reserves:
  - Trade deficit in 2005: about 21 percent of GDP, with more than two-thirds explained by remittance-financed imports.
  - Gross international reserves move with remittances in the short term, but no evidence of a longer-term effect because remittances are eventually absorbed through higher imports.
- Investment and savings:
  - No evidence that remittances have boosted private investment; remittances show a negative correlation with private investment in recent years.
  - Negative correlation between remittances and imports of construction materials (cement and iron) in annual data.
  - National accounts: private savings tend to move with remittances, but banking deposits are not correlated with remittances in annual data.
  - Authorities’ estimate of conversion episode: CF 5 billion exchanged in 2001 (4 percent of GDP and 62 percent of time deposits).
  - Deposit interest rates: 3 percent (less than one percentage point higher than in France).

### Financial transmission, seasonality, and short-run dynamics (Annex I findings)
- Seasonality and short-term co-movement:
  - Most remittances flow into Comoros between June and August (the holiday season in France), creating pronounced seasonality in monetary variables.
  - Currency in circulation, broad money, and international reserves tend to move strongly with remittances in the short term.
- Long-run absorption:
  - Co-movement of remittances and monetary aggregates disappears in the long run as additional liquidity is absorbed through higher imports.
- Deposits and intermediation:
  - Remittances appear to have little impact on banking deposits; euros are largely exchanged into domestic cash or saved in euro cash outside the banking system.
  - Limited financial intermediation constrains benefits to domestic credit and long-term capital formation.
- Role for households and credit:
  - Remittances have been a source of liquidity for individuals unable to borrow from banks or microfinance institutions.
  - Private sector borrowing constrained: interest rates on borrowing are about 14 percent.
  - Monthly data show during peak remittance times there is a decline in credit to the private sector.
  - The initial effect of a remittance shock on credit is negative and marginally significant (at the 10 percent significance level), suggesting remittances may act as an alternative source of liquidity for households.
  - Little evidence remittances have created moral hazard in the local credit market (no strong co-movement between remittances and bad loans).

### Empirical analysis: SVAR and VECM (monthly data, January 1999—December 2005)
- SVAR setup:
  - Six-variable SVAR with ordering: changes in remittances => international reserves and monetary aggregates => inflation => real exchange rate => trade credit.
  - Four SVAR variants alternate the money-aggregate variable: broad money, currency in circulation, quasi money, and non-cash broad money.
  - Impulse: one standard deviation shock of remittances (about 40 percent).
- Short-run SVAR findings:
  - Effect of remittances on broad money and reserves is positive and statistically significant in the short run but dissipates in the long run.
  - A one standard deviation remittance shock causes a significant positive response of the percent change in the stock of broad money and international reserves instantaneously (more than 1½ percent) that dissipates over three months.
  - Replacing broad money with currency in circulation yields similar effects with higher magnitude, of about 2 percent.
  - With quasi-money or banking deposits, the effect is insignificant.
  - The impulse response shows no significant effect on the inflation rate (and thus on variations of the real exchange rate) in the short run.
- VECM (bivariate) long-run findings:
  - Four bivariate VECMs run for broad money, reserves, CPI and the REER confirm no long-run relationship with broad money or reserves, indicating absorption through higher imports.
  - No long-run impact of remittances on prices or on the REER with 12 lags.
- Elasticities to remittances (bivariate VECM):
  - With 1 lag (Optimal lag with Schwert criterion):
    - ∂(log money_broad)/∂(log remittances) = 1.064871**  (.1600133)
    - ∂(log reserves)/∂(log remittances) = 1.615073 **  (.2783712)
    - ∂(log CPI)/∂(log remittances) = .4952056 **  (.0789659)
    - ∂(log REER)/∂(log remittances) = 1.038971 **  (.1697837)
    - ** Significant at 5% significance level.
  - With 12 lags (Optimal lag with Akaike Information criterion):
    - ∂(log money_broad)/∂(log remittances) = -.0864541  ( .3246074)
    - ∂(log reserves)/∂(log remittances) = -.0901794  (.3753588)
    - ∂(log CPI)/∂(log remittances) = .043198  (.0242742)
    - ∂(log REER)/∂(log remittances) = .0335963  ( .0252072)
  - Note: one-lag VECM relationships for CPI and REER are significant, but both the 12-lag bivariate VECM and the multivariate SVAR do not show any effect in levels and differences.

### Conclusions on remittances and policy implications
- Stylized facts:
  - Remittances are mainly altruistic, countercyclical, and help absorb external shocks.
  - Remittances enhance consumption and living conditions but are spent mostly on imports and do not significantly contribute to investment.
  - Comoros’s high trade deficit is largely structural, with more than half the goods imported financed by remittances.
  - No clear long-run Dutch-disease effect; remittances show at least a weak temporary impact on domestic prices and the REER.
  - Remittances help fill gaps in the underdeveloped financial sector but have so far not contributed to its development.
- Policy implications:
  - Monetary policy must develop analytical and policy tools to react to short-term fluctuations of reserves and monetary aggregates caused by remittances.
  - Policymakers need to prepare for possible structural shifts in remittances; large increases or decreases would be partly offset by higher or lower imports, but policy adjustments may be required to avoid adverse consequences on the balance of payments, inflation, or economic growth.
  - Remittances present opportunities for economic development if directed more toward investment and channeled through the local financial sector.
  - To increase developmental impact, improving the investment climate and opening up the financial sector and product markets for greater competition is important.

### Civil service payrolls, revenue sharing, and fiscal decentralization
- Civil service payrolls and wage bill:
  - Number of civil servants increased from 6,028 in 1994 to 10,027 in 2005.
  - Total wage bill was about 8 percent of GDP in 2005 and 54 percent of revenues.
- Revenue-sharing mechanism (practical observations):
  - In practice two parties have significant revenues (Union and Anjouan) and two have almost no revenue (Ngazidja and Moheli).
  - Major revenue sources: port of Moroni (effectively controlled by the Union), national telecommunications company (controlled by the Union), and port of Anjouan (controlled by the government of Anjouan).
  - Pace of revenue growth: 4 percent annually between 2002 and 2005.
- Arrears and capacity constraints:
  - Comoros has not been current on budgetary commitments for decades.
  - Civil service wage arrears in early 2006: on average four months overdue; examples of salaries overdue for over five years.
  - Average arrears by island in early 2006: Moheli 8 months, Anjouan 3 months.
  - Fiscal policy operates with four interlinked budgets; consolidated data on all fiscal operations are prepared only infrequently.
  - Data transmission is often manual and compilation requires expensive interisland travel; data often inaccurate and inconsistent.
- 2006 revenue-sharing arrangement (reforms and allocations):
  - All shared government revenues to go into a single special account at the central bank; multiple government revenue-receiving accounts abolished.
  - Central bank fixed deductions:
    - External debt service: 20.1%
    - Pension payments: 5.5%
  - Remainder to be shared as follows:
    - Union: 37.5%
    - Ngazidja: 27.4%
    - Anjouan: 25.7%
    - Moheli: 9.4%
  - Operational rules: balances due to be credited to each government’s account at the close of each day; each island’s own revenues to be deposited in its own account with full disclosure.
  - Payout rule: each entity is to pay out for salaries 60-80% of total receipts.
  - Transparency measures: daily reports on balances; island representatives have access to central bank information; Union placing representatives with each island government to monitor reporting.
  - Early outcomes (as of August 2006): parties visibly committed; revised 2006 budgets approved on consistent figures; four Ministers of Finance and interisland budget committee meeting monthly.
- Policy implications and options:
  - Decentralization is essential for reconciliation and service delivery, but has aggravated capacity constraints and created incentive problems for revenue collection.
  - Options include devolving more revenue and expenditure responsibilities to islands and classifying more revenues as recettes propres so each government retains more of its own revenue.
  - Implementation complexity: devolving education and health aligns with I-PRSP priorities but is complex given small scale and coordination difficulties.
  - External assistance: pledges at the 2005 donors’ conference could increase aid to implement I-PRSP, but increased aid had not yet materialized.

### Tax system highlights (selected rates and rules)
- Scope and key rates (selected items only):
  - Profit tax (IBD): levied on companies and self-employed with turnover exceeds CF 20 million; rates: 35 percent for individuals and companies with turnover less than CF 500 million, 50 percent otherwise.
  - Single professional tax (TPU): levied on companies and self-employed with turnover below CF 20 million.
  - Advance on income taxes (Acompte sur IBD et TPU): 1 percent of customs value of imports for importers with taxpayer number; 50 percent for all others (as surcharge on imports, deductible).
  - Taxes on capital income: 15 percent.
  - Payroll tax brackets and rates: incomes under FC150,000 per year (exempt); rates by bracket include 5, 10, 15, 20, 25, 30 (bracket cutoffs listed in source).
  - Business license (Patente d’exploitation): rates 3, 5, or 10 percent of annual rental value plus fixed CF 5,000 to CF 750,000.
  - Consumption tax (Taxe de consommation): five rates with base rate 10 percent; basic necessities 0 percent; water/private schools/inter-island airfares 3 percent; electricity/telephone/hotel/restaurants/banking services/international airfares 5 percent; casino revenues 25 percent.
  - Visa fees: Visas CF500-2000 depending on duration and number of entries; Residency card CF 60,000.
  - Motor vehicle tax (Vignette): CF4,000-22,500 per cylinder depending on age and engine capacity.
  - Single import tax (TUI): rates 0 percent, 15 percent, 21 percent, and 30 percent of import value, c.i.f.
  - Accompte sur IBD et TPU (AIT): 1 percent of c.i.f. value (levied on all commercial imports; deductible).
  - Single specific tax on imports (TUS): CF 10-150 per kilogram, CF 8-200 per liter, up to CF 1 million for buses.
  - Single tax on rice: CF 150/kg for high-grade rice; CF 45/kg for ordinary rice.
  - Single tax on petroleum (TPP): Gasoline (normal rate) CF 230/l; Gasoline (reduced rate) CF 211/l; Diesel oil (normal rate) CF 115/l; Diesel oil (reduced rate) CF 95/l; Kerosene (normal rate) CF 10/l.
  - Excises, Special fiscal duty (TSVF): Spirits 300 percent; wine and beer 200 percent; tobacco products 300 percent; Cement 15 percent of c.i.f. value.
  - Forfait regime (containers): 20 ft container FC 2.1 million; 40 ft container FC 3.5 million.
  - Additional centimes: 1 percent of export value, f.o.b., less export duties (levied on behalf of the Chamber of Commerce UCCIA).

*Source: Annex I: Econometric Analysis, chapter on "THE ROLE OF REMITTANCES IN THE COMORIAN ECONOMY" (text provided).*

### Annex I: Econometric Analysis .......................................................................13

### Annex I: Econometric Analysis .......................................................................13

### Overview
- Remittances are central to the Comorian economy:
  - In 2005, remittances were estimated at 18½ percent of GDP.
  - The share of remittances in current account receipts was almost 50 percent.
- The chapter uses annual time series correlations since 1999 and monthly time series regressions (Annex I) to analyze determinants, macroeconomic effects, and financial transmission of remittances.

### Origins, Channels, and Uses of Remittances
- Origins and scale:
  - It is estimated that between 150,000 to 200,000 people of Comorian origin live outside the Union of the Comoros.
  - The 2003 household census shows that 80 percent of the diaspora lives in Mayotte or France.
- Transmission channels (1999–2005):
  - About two-thirds of remittances flow in as cash, mainly euros.
  - Wire transfers (after rapid growth) have reached about one-fourth of total remittances.
  - Remittances sent directly in the form of goods have also been increasing.
  - The island of Grande Comore (Ngazidja) receives by far the highest share of remittances.
- Primary uses:
  - Remittances are mainly spent on imported goods.
  - It is estimated that remittances amount to about 70 percent of total imports of goods and finance more than half of all imports to Comoros.
  - Anecdotal evidence indicates a significant share is spent on Grand Marriage ceremonies and housing construction, but annual data suggest most remittances finance consumption of imported goods.

### Determinants and Macroeconomic Effects
- Motives and cyclicality:
  - Altruistic motives are clearly the main reason for remittances to Comoros.
  - Inflows tend to be countercyclical to real GDP growth and export performance; remittances rose in 2001 and 2005 when they offset export sector losses.
- External balances and reserves:
  - Comoros’s trade deficit amounted to about 21 percent of GDP in 2005, of which more than two-thirds can be explained by imports financed by remittances.
  - Gross international reserves have tended to move with remittances in the short term, but there is no evidence of a longer-term effect because remittances are eventually absorbed through higher imports.
- Investment and savings:
  - There is no evidence that remittances have boosted private investment; remittances show a negative correlation with private investment in recent years.
  - Despite anecdotal evidence of housing construction financed by remittances, annual data show a negative correlation between remittances and imports of construction materials (cement and iron).
  - National accounts data show private savings tend to move with remittances, but banking deposits are not correlated with remittances in annual data.
  - Monthly time series regression analysis (Annex I) shows a change in remittances does not cause a statistically significant change in deposits, implying most savings from remittances remain outside the financial system.
  - The authorities estimate that conversion of the French franc into euros in 2001 involved the equivalent of CF 5 billion (4 percent of GDP and 62 percent of time deposits), indicating substantial informal cash holdings.
  - Deposit interest rates at 3 percent are less than one percentage point higher than in France.
- Dutch-disease assessment:
  - There is no conclusive evidence of a Dutch-disease effect:
    - Annual changes in remittances have not been strongly correlated with movements in inflation or the real exchange rate.
    - Monthly data show a weak but positive short-run correlation between remittances and monthly changes in the CPI, suggesting a possible small short-run impact on the real exchange rate, but no long-run relationship in the monthly data (Annex I).
  - Remittances may reduce demand for domestic goods if they finance imported substitutes, but given their large size, even a small share used domestically could affect domestic prices.

### Financial Transmission and Seasonality (Findings from Annex I)
- Short-term co-movement:
  - Currency in circulation, broad money, and international reserves tend to move strongly with remittances in the short term.
  - Most remittances flow into Comoros between June and August (the holiday season in France), creating pronounced seasonality in monetary variables.
- Long-run absorption:
  - The co-movement of remittances and monetary aggregates disappears in the long run as additional liquidity is absorbed through higher imports.
- Deposits and formal financial intermediation:
  - Remittances appear to have little impact on banking deposits; euros are largely exchanged into domestic cash or saved in euro cash outside the banking system.
  - The limited financial intermediation of remittances implies constrained benefits to domestic credit and long-term capital formation.

### Key Quantitative Statements and Indicators (as reported)
- Remittances in 2005: 18½ percent of GDP.
- Remittances share of current account receipts: almost 50 percent.
- Remittances composition: about two-thirds cash; about one-fourth via wire transfers.
- Diaspora distribution: 80 percent in Mayotte or France.
- Estimated diaspora size living abroad: between 150,000 to 200,000 people.
- Remittances finance about 70 percent of total imports of goods and more than half of all imports.
- Trade deficit in 2005: about 21 percent of GDP, with more than two-thirds explained by remittance-financed imports.
- Conversion episode estimate: CF 5 billion exchanged in 2001 (4 percent of GDP and 62 percent of time deposits).
- Deposit interest rates: 3 percent (less than one percentage point higher than in France).
- Seasonality: majority of inflows between June and August.

*Source: Annex I: Econometric Analysis, chapter on "THE ROLE OF REMITTANCES IN THE COMORIAN ECONOMY" (text provided).*

### 14.      Remittances have been a source of liquidity for individuals who are not able to

### _cr06385 - 14.      Remittances have been a source of liquidity for individuals who are not able to

### Role and characteristics of remittances
- Remittances have been a source of liquidity for individuals who are not able to borrow from a bank or microfinance institution.
- Private sector borrowing is constrained by the monopolistic structure of the banking system; interest rates on borrowing are about 14 percent.
- A large share of remittances flows into Comoros through informal channels, limiting their contribution to the development of the financial system and formal financial intermediation.
- Remittances are seasonal and cause short-term fluctuations in monetary aggregates (in monthly percent changes).
- Remittances are mainly transferred based on altruistic motives. They have recently been countercyclical and helped to absorb external shocks.
- Remittances are spent mostly on imports and do not significantly contribute to investment; more than half the goods imported are financed by remittances.
- The main impact of remittances has been to enhance consumption and improve living conditions; it is not clear to what extent they have contributed to long-run economic growth.
- There is no clear evidence for a significant Dutch-disease effect, as most remittances are absorbed through imports; however, remittances appear to have at least a weak temporary impact on domestic prices and the real exchange rate.

### Observed interactions with credit and crises
- Monthly data show that during the peak time for remittances there is also a decline in credit to the private sector.
- In the third quarter of 2005, remittances increased sharply after the vanilla crisis had led to a strong increase in non-performing loans in mid-2005. Remittances spiked after an increase in non-performing loans to the vanilla sector in 2005 (in monthly percent changes).
- There is otherwise little evidence of co-movement between remittances and bad loans, and it is therefore unlikely that remittances have created moral hazard in the local credit market.
- The initial effect of a remittance shock on credit is negative and marginally significant (at the 10 percent significance level), suggesting remittances may act as an alternative source of liquidity for households.

### Empirical analysis: SVAR and VECM results (monthly data, January 1999—December 2005)
- SVAR setup and assumptions:
  - Six-variable SVAR model with ordering: changes in remittances => international reserves and monetary aggregates => inflation => real exchange rate => trade credit.
  - Four SVAR variants alternate the money-aggregate variable: broad money, currency in circulation, quasi money, and non-cash broad money.
  - The impulse analyzed is a one standard deviation shock of remittances (about 40 percent).
- Short-run SVAR findings:
  - The effect of remittances on broad money and reserves is positive and statistically significant in the short run but dissipates in the long run.
  - A one standard deviation remittance shock causes a significant positive response of the percent change in the stock of broad money and international reserves instantaneously (more than 1½ percent) that dissipates over three months.
  - Replacing broad money with currency in circulation yields similar effects with even higher magnitude, of about 2 percent.
  - With quasi-money or banking deposits, the effect is insignificant, confirming remittances are mainly used for consumption or hoarding outside the financial system rather than savings in the local banking sector.
  - The impulse response shows no significant effect on the inflation rate (and thus on variations of the real exchange rate), pointing to lack of evidence of a Dutch-disease effect in the short run.
- VECM (bivariate) long-run findings:
  - Four bivariate vector error correction models were run for broad money, reserves, CPI and the real effective exchange rate (REER) to calculate elasticities to remittances.
  - The VECM confirm that remittances have no long-run relationship with broad money or reserves, indicating they are likely absorbed through higher imports.
  - There is also no long-run impact of remittances on prices or on the REER with 12 lags.
- Elasticities of selected variables to remittances (bivariate VECM)
  - With 1 lag (Optimal lag with Schwert criterion):
    - ∂(log money_broad)/∂(log remittances) = 1.064871**  (.1600133)
    - ∂(log reserves)/∂(log remittances) = 1.615073 **  (.2783712)
    - ∂(log CPI)/∂(log remittances) = .4952056 **  (.0789659)
    - ∂(log REER)/∂(log remittances) = 1.038971 **  (.1697837)
    - ** Significant at 5% significance level.
  - With 12 lags (Optimal lag with Akaike Information criterion):
    - ∂(log money_broad)/∂(log remittances) = -.0864541  ( .3246074)
    - ∂(log reserves)/∂(log remittances) = -.0901794  (.3753588)
    - ∂(log CPI)/∂(log remittances) = .043198  (.0242742)
    - ∂(log REER)/∂(log remittances) = .0335963  ( .0252072)
  - Note: although one-lag VECM relationships for CPI and REER are significant, both the 12-lag bivariate VECM and the multivariate SVAR do not show any effect in levels and differences.

### Conclusions and policy implications
- Summary of stylized facts:
  - Remittances are mainly altruistic, countercyclical, and have helped absorb external shocks.
  - Remittances enhance consumption and living conditions but are spent mostly on imports and do not significantly contribute to investment.
  - Comoros’s high trade deficit is largely structural, with more than half the goods imported financed by remittances.
  - No clear long-run Dutch-disease effect; remittances show at least a weak temporary impact on domestic prices and the REER.
  - Remittances help to fill gaps in the underdeveloped financial sector but have so far not contributed to its development.
- Implications for policymakers:
  - Monetary policy faces challenges in developing analytical and policy tools to react appropriately to short-term fluctuations of reserves and monetary aggregates caused by remittances.
  - Policymakers need to prepare for possible structural shifts in remittances; large increases or decreases would be partly offset by higher or lower imports, but policy adjustments may still be required to avoid adverse consequences on the balance of payments, inflation, or economic growth.
  - Remittances present opportunities to support economic development in line with Comoros’s Interim Poverty Reduction Strategy Paper if directed more toward investment and channeled through the local financial sector.
  - To achieve greater developmental impact, it will be important to improve the investment climate and to open up the financial sector and product markets for greater competition.

*Source: Fund staff analysis and calculations as presented in the chapter.*

### 14.      The disorderly process of decentralization and creation of new government

### 14.      The disorderly process of decentralization and creation of new government structures after 2001 have caused civil service payrolls to swell

### Civil service payrolls and wage bill
- Number of civil servants increased from 6,028 in 1994 to 10,027 in 2005.
- Total wage bill was about 8 percent of GDP in 2005 and 54 percent of revenues.
- The wage bill appears to have declined somewhat from the early 1990s, suggesting average salaries were cut in response to swelling payrolls and constant financial resources.

### Revenue-sharing mechanism: structure and asymmetries
- In theory: customs and tax authorities for each of the three islands transfer all revenues designated for distribution to the revenue-sharing account; all four governments (Union and islands) receive allocations based on fixed percentages.
- In practice: two parties have significant revenues (Union and Anjouan) and two have almost no revenue (Ngazidja and Moheli).
- Major revenue sources in practice: port of Moroni (effectively controlled by the Union), national telecommunications company (controlled by the Union), and port of Anjouan (controlled by the government of Anjouan).
- Virtually all revenue collected by the four governments is designated for sharing; only a marginal share is designated as recettes propres for retention at the island level.
- Pace of revenue growth was 4 percent annually between 2002 and 2005.

### Incentive problems, arrears, and capacity constraints
- The revenue-sharing mechanism suffers from free rider and moral hazard problems:
  - With key revenues designated for sharing, incentives to maximize revenue collection are reduced, particularly for islands with relatively low redistribution percentages.
  - Incentives to hide revenues have created mutual mistrust between the Union and island governments.
  - Some discretion in application of redistribution reduces incentives for island governments to cut spending commitments.
- Domestic arrears dominate fiscal policy:
  - Comoros has not been current on budgetary commitments for decades.
  - In early 2006, civil service wage arrears were on average four months overdue; instances of salaries overdue for over five years are noted.
  - Average arrears by island in early 2006: Moheli 8 months, Anjouan 3 months.
  - Settling arrears is a constant social and political priority because civil service wages are the sole source of cash earnings for many families.
- Capacity and information constraints:
  - Fiscal policy effectively operates with four interlinked budgets; consolidated data on all fiscal operations are prepared only infrequently.
  - Data transmission is often manual; compilation requires expensive interisland travel.
  - Data are often inaccurate and inconsistent because the four governments have different methodologies and there is insufficient information-sharing.
  - Consequences: chronic arrears, poor expenditure targeting, volatile revenues, and relatively low levels of foreign aid.
- Political cycle effects:
  - Progress in interisland cooperation made in drafting the 2005 budget collapsed before the 2006 presidential elections.
  - Collection of customs revenues in the port of Moroni plunged before the transfer of power; Anjouan ceased participation in joint meetings and early-year contributions were well below expectations.

### The new revenue-sharing arrangement (2006 reforms)
- Reaffirmed that all shared government revenues should go into a single special account at the central bank; multiple government revenue-receiving accounts abolished.
- Central bank fixed deductions:
  - External debt service: 20.1%
  - Pension payments: 5.5%
- Remainder to be shared as follows:
  - Union: 37.5%
  - Ngazidja: 27.4%
  - Anjouan: 25.7%
  - Moheli: 9.4%
- Operational rules:
  - Balances due are to be credited to each government’s account at the close of each day.
  - Each island’s own revenues are to be deposited in its own account with full disclosure to the Union and other islands.
  - The arrangement stipulates that each entity is to pay out for salaries 60-80% of total receipts to prevent neglect of salary payments or minimum operating expenditures.
- Transparency and mutual control measures:
  - Daily reports on balances are made available and island representatives have access to central bank information at all times.
  - Union government will place representatives with each island government to monitor reporting and assist with communications.
- Early outcomes:
  - As of August 2006, all parties were visibly committed to functioning of the arrangement.
  - A revised budget for 2006 for the Union and the islands was approved on the basis of figures consistent with smooth functioning and full disclosure.
  - The four Ministers of Finance and the interisland budget committee have been meeting regularly each month to coordinate fiscal policy.

### Policy implications and challenges
- Fiscal decentralization is essential to national reconciliation and economically rational given geographical dispersion; devolving significant fiscal administration and policy decisions to the island level is warranted despite small country size.
- Decentralization has aggravated capacity constraints:
  - Large civil service and multilayered public administration have coexisted with limited fiscal revenues, preventing adequate salaries for qualified technical staff and leading to chronic wage arrears.
  - Institutional capacity has remained very low and government services highly constrained.
  - Decentralization increases needs for information sharing and data consolidation, complicated by poor communications infrastructure and insufficient interisland cooperation.
  - Improving interisland coordination and information sharing, and unifying certain government functions such as accounting and statistics, is critical to strengthen institutional capacity and policy implementation.
- Options to improve incentives and local responsibility:
  - Devolve more revenue and expenditure responsibilities to the islands to address free rider and incentive problems.
  - One option: classify more revenues as recettes propres, allowing each government to retain more of their own revenue and assume more responsibility for financing expenditures.
  - This option aligns with Comoros’s I-PRSP, which identifies education and health as first priorities to be devolved to the islands; implementation will be complex given small scale of island-level expenditures and coordination difficulties.
- External assistance considerations:
  - Pledges at the 2005 donors’ conference could lead to significantly higher external assistance to help implement the I-PRSP, although increased aid had not yet materialized.
  - Donors historically implement projects in any island, with government co-financing symbolic amounts; project execution is largely left to donors and maintenance is rarely budgeted after completion.
  - As project execution shifts from donors to country authorities, a mechanism will be needed to balance competing wishes of four governments with the need to channel external funds into projects with highest economic and social returns.
- Summary judgment:
  - Fiscal decentralization has been a mixed blessing: vital to reconciliation and service delivery, but creating incentive problems for revenue collection, aggravating data and staffing problems, and imposing additional administrative layers and costs.
  - Benefits from flexible, locally focused public administration are possible, but devolving responsibilities demands better monitoring, information sharing, continuous close policy dialogue, and a national consensus.

*Source: Comorian authorities; data and narrative as presented in the chapter.*

### References

### References

### Cited works
- Bolton, Patrick, and Gerard, Roland, 1997, “The Breakup of Nations: A Political Economy Analysis”, The Quarterly Journal of Economics, Vol. 112, No. 4., pp. 1057-1090.  
- Breton, Albert, 2002, “An Introduction to Decentralization Failure,” Managing Fiscal Decentralization, ed. by Ethisham Ahmed and Vito Tanzi (London: Routledge Studies in the Modern World Economy)  
- Brosio, Girogio, 2000, “Decentralization in Africa”, mimeo, prepared for the Conference on Fiscal Decentralization International Monetary Fund, Washington, DC November 20-21, 2000.  
- Lienert, Ian and Jitendra, Modi, 1997, "A Decade of Civil Service Reform in Sub-Saharan Africa," IMF Working Paper 97/179, Washington: International Monetary Fund.  
- Musgrave, Robert A. 1969, The Theory of Public Finance (New York, McGraw-Hill).  
- Ndegwa, Stephen N.,  2002, “Decentralization in Africa: A Stocktaking Survey,” Africa Region Working Paper Series, No 40 (Washington: World Bank).  
- Person and Tabellini, 1996, “Federal Fiscal Constitutions: Risk Sharing and Moral Hazard” Econometrica, Vol. 64, No. 3, pp. 623-646.  
- Smoke, Peter, 2000, “Fiscal Decentralization in East and Southern Africa: A Selective Review of Experience and Thoughts on Moving Forward”, mimeo, prepared for the Conference on Fiscal Decentralization International Monetary Fund, Washington, DC November 20-21, 2000.  
- Tanzi, Vito, 1996, “Fiscal Federalism and decentralization: A review of Some Efficiency and Macroeconomic Aspects”, in Annual Wordl Bank Conference on Development Economics, 1995 (Washington: World Bank).  
- Tanzi, Vito 2002, “Pitfalls on the Road to Fiscal Decentralization,” in Managing Fiscal Decentralization, ed. by Ethisham Ahmed and Vito Tanzi (London: Routledge Studies in the Modern World Economy).

*Source: _cr06385 - References*

### 1. Taxes on net income and profits 1.1 Profit tax (Impôt sur les Bénéfices Divers, IBD)

### 1. Taxes on net income and profits 1.1 Profit tax (Impôt sur les Bénéfices Divers, IBD)

### Scope and exemptions
- Tax levied on companies and self-employed individuals whose turnover exceeds CF 20 million.
- Newly created enterprises eligible under the Investment Code, agricultural cooperatives and credit unions, government agencies and offices are exempt.

### Rates and structure
- 35 percent for individuals and companies with turnover less than CF 500 million, 50 percent otherwise.
- Multiples of the cost of a business licence (see item 3.1), depending on turnover, with x equal to:
  - Turnover (T, in CF millions)
    - 4
    - 15 < T < 20
    - 3
    - 5 < T < 15
    - 2
    - 3< T < 5

### 1.2 Single professional tax (Taxe Professionnelle Unique, TPU)

### Scope and exemptions
- Tax levied on companies and self-employed individuals with turnover below CF 20 million.
- Provisions of the investment code apply.

### Rates
- 1                                      T<                                      3

### 1.3 Advance on income taxes (Acompte sur IBD et TPU)

### Nature and treatment
- Surcharge on imports, deductible from IBD or TPU. See item 4.2.
- Same as for customs duties.

### Rates and application
- 1 percent of customs value of imports for importers with taxpayer number; 50 percent for all others.

### 1.4 Taxes on capital income (Impôt sur les revenus de valeur et capitaux mobiliaire)

### Scope
- Tax on distributed dividends and interest paid.
- Interest on loans contracted or granted by microcredit agencies, housing cooperatives, and agricultural credit unions.

### Rates
- 15 percent Tax bracket (in CF)
- Tax rate (in percent)

### 1.5 Payroll Tax (Taxe sur les salaires, formerly ‘Impôt general sur le revenu’)

### Scope and treatment
- Paid by employees on wages and salaries.
- Withheld at source for civil servants; settled within one month for others.
- Diplomats exempt.

### Exemptions and brackets
- Incomes under FC150,000 per year
- 150,001 – 500,000
- 500,001 – 1,000,000
- 1,000,001 – 1,500,000
- 1,500,001 – 2,500,000
- 2,500,001 – 3,500,000
- over 3,500,000

### Rates by bracket
- 5
- 10
- 15
- 20
- 25
- 30

---

### 2. Property taxes

### 2.1 Tax on rentals (Impôt sur la proprieté batie et louée)
- Annual tax on rental dwellings.
- Exemptions: New properties used as dwellings for the first two years, and properties belonging to the government or to communes.
- Rates:
  - Dwellings: 20 percent of rental value
  - Commercial and industrial premises: 30 percent

### 2.2 Tax on registered property (droits d’enrigestrement)
- Transactions tax on real estate transmissions (sales, donations, inheritances).
- Exemptions: Government transactions (that include Muslim religion buildings).
- Rates and fixed taxes:
  - Judicial acts: fixed taxes of  CF1,000 to CF20,000.
  - Sales: 2-9 percent of price.
  - Donations and inheritances: 5-60 percent of assessed value.

### 2.3 Property recording fees (Taxe de publicité fonciére)
- Tax on recording acts concerning property transactions.
- Exemptions: Government transactions (that include Muslim religion buildings).
- Rates:
  - 2 percent property rights and motgages; 1 percent baux.

### 2.4 Tax on real estate gains (Taxe sur la plus value immobiliere)
- Tax on capital gains for property sales.
- Exemptions: Government transactions (that include Muslim religion buildings).
- The capital gain calculation and applicable tax are determined in the art. 65.2 of the Tax Code.

---

### 3. Taxes on goods and services

### 3.1 Business license (Patente d’exploitation)
- Tax on any company or self-employed individual involved in trade, industry, or a profession.
- Exemptions: Craftsmen, farmers, agricultural cooperatives, agricultural credit unions.
- Rates:
  - 3, 5, or 10 percent of the annual rental value of the business premises depending on location; plus fixed amount of between CF 5,000 and CF 750,000, depending on the type of business

### 3.2 Consumption tax (Taxe de consommation)
- Consumption tax on domestic transactions: Turnover tax levied on domestic services (hotels, restaurants, banks, electricity, water, telecommunica-tions, import trade,...). Consumption tax on imports: Levied at customs, based on c.i.f. value plus single import tax, credited against domestic consumption tax if applicable.
- Exemptions: Medical services; Enterprises with turnover of CF 20 million or less; Exports, petroleum products,and publishing.
- Rates:
  - Five rates: base rate, 10 percent; basic necessities, 0 percent; water, private schools, and inter-island airfares, 3 percent; electricity, telephone, hotel and restaurants, banking services, and international airfares, 5 percent; casino revenues, 25 percent.

### 3.3 Visa fee (Droits de visa)
- Entry and residency fees to be paid by foreigners coming into Comoros.
- Exemptions: Diplomats.
- Rates:
  - Visas         CF500-2000 depending on duration and number of entries
  - Residency card             CF 60,000

### 3.4 Motor vehicle tax (Vignette)
- Annual tax on vehicles in use.
- Exemptions: Vehicles owned by the central government or local governments, diplomats and cooperant missions. Cars in Comoros for more than 20 years.
- Rates:
  - CF4,000-22,500 per cylinder, depending on the age and of the vehicle and engine capacity.

### 3.5 Special tax on diesel engine vehicles (Taxe sur les vehicules à moteur diesel)
- Annual tax on vehicles in use.
- Exemptions: Vehicles owned by the central government or local governments, diplomats and cooperant missions; and vehicles used in agricultural activities.
- Rates:
  - CF 25,000 per ton of load capacity.

### 3.6 Parking fee (Droit de stationement)
- Annual tax on vehicles in use.
- Same exemptions as special tax on diesel engine vehicles.
- Rates:
  - CF 1,000 per vehicle.

### 3.7 Stamp duties (Droit de timbre)
- Tax charge on official acts.
- Exemptions: Government, diplomats cooperant missions.
- Rates:
  - CF500-CF25,000 depending on transaction

### 3.8 Tax on insurance policies (Taxe sur les contrats d’assurances)
- Tax charge on insurance policies.
- Exemptions: Government, diplomats cooperant missions.
- Rates:
  - 3 percent of the policy value for life, maritime, and rental insurance.
  - 15 percent for fire insurance.
  - 4 percent for all other modalities.

### 3.9 Tax on TV equipment (Impôt sur les récepteurs de télévisions, les magnétoscopes et les vidéoscopes)
- Annual tax charged on use of TV, magnetoscope and videoscope sets.
- Exemptions: None.
- Rates:
  - FC20,000 per set

### 3.10 Alcohol sale license (License de vente de boissons alcoholiques)
- One-time tax for newly licensed distributors and retailers of alcoholic beverages.
- Exemptions: None.
- Rates:
  - CF 750,000 for distributors
  - CF 500,000 for retailers

---

### 4. Taxes on international trade

### 4.1 Single import tax (TUI)
- Tax on imported goods.
- Exemptions: Goods imported by the central government under grants, and goods imported by diplomatic missions.
- Rates:
  - Rates of 0 percent, 15 percent, 21 percent, and 30 percent of import value, c.i.f.

### 4.2 Accompte sur IBD et TPU (AIT)
- Levied on all commercial imports.
- Deductible from payment of IBD and TPU, see items 1.1 and 1.2.
- Rates:
  - 1 percent of c.i.f. value

### 4.3 Administrative levy (redevance administrative)
- Tax on imported goods.
- Exemptions: Goods imported by the central government under grants, and imports by diplomatic missions. Cement, fertilizers, rice, petroleum products, flour.
- Rates:
  - 1 percent of customs receipts on taxable imports; 3 percent of c.i.f. value of exempt imports.

### 4.4 Single specific tax on imports (TUS)
- Applies to selected imports, mainly meat and buses.
- Exemptions: Goods imported by the central government under grants, and goods imported by diplomatic missions.
- Rates:
  - CF 10-150 per kilogram, CF 8-200 per liter, up to CF 1 million for buses.

### 4.5 Single tax on rice
- Rates:
  - CF 150/kg for high-grade rice; CF 45/kg for ordinary rice;

### 4.6 Single tax on petroleum (TPP)
- Reduced rate applies to diplomatic missions, ships, and aircrafts (zero rate for aviation fuel Jet A1).
- Rates:
  - Gasoline (normal rate)   CF 230/l
  - Gasoline (reduced rate) CF 211/l
  - Diesel oil (normal rate)  CF 115/l
  - Diesel oil (reduced rate) CF 95/l
  - Kerosene (normal rate)  CF 10/l

### 4.7 Excises, or Special fiscal duty (TSVF)
- Levied on alcoholic beverages, tobacco products, and cement.
- Rates:
  - Spirits: 300 percent
  - wine and beer:  200 percent
  - tobacco products, 300 percent
  - Cement: 15 percent of c.i.f. value

### 4.8 Export duty
- Applies to cloves, vanilla, and Ylang-ylang.
- Rates:
  - Set annually, dependent on market conditions

### 4.9 Forfait regime (regime forfaitaire)
- Specific tax rate levied on all containers (20 foot and 40 foot). Not applied to one imported good if more than 60 percent of content.
- Exemptions/notes: Tabac, Alcool, Rice, Soda beverages Printed textiles (Kiromani, Lesso, Pagne, Mamuwa, and Msoutrou)
- Rates:
  - 20 ft container: FC 2.1 million
  - 40 ft container: FC 3.5 million

---

### 5.0 Additional centimes
- Levied on behalf of the Chamber of Commerce (UCCIA).
- Rates:
  - 1 percent of export value, f.o.b., less  export duties.

*Source: Comoros tax system summary (Table 37).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2006/_cr06385.pdf_