## wp17270

## Source details

**Canonical URL:** [wp17270](https://www.imf.org/-/media/files/publications/wp/2017/wp17270.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2017/wp17270.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2017/wp17270.pdf.json)

---

### Acknowledgment and Overview
- Paper development supported by Daniela Gressani and Eric Mottu; contributions from Louis Dicks-Mireaux.
- Presented to Djiboutian authorities during staff visits in November 2016 and April 2017; participants included Central Bank Governor Ahmed Osman; Minister of Budget Boheh Ahmed Robleh; Minister of Economy and Finance Ilyas Moussa Dawaleh; Minister of Investment Ali Guelleh Aboubaker; Secretary General Simon Mibrathu; Director of the Debt Directorate Almis Mohamed Abdillahi; Director of the Budget Directorate Othman Sadik Ahmed.
- Research assistance from J. Chen and L. Zhao; administrative support from M. De Mesa.

- Country profile and recent estimates (2016, est.):
  - Nominal GDP: 1,894 million
  - Population: 993,000
  - GDP per capita: US $1,908
  - Currency: Djiboutian franc (DF)
  - Exchange rate: DF 177.7 = US$ 1
  - Unemployment rate: 39 percent
  - Poverty rate: 41 percent

- Key investment projects and costs:
  - Multipurpose port Doraleh: US$580 million
  - Addis-Ababa – Djibouti railway: US$490 million
  - Water pipeline from Ethiopia: US$322 million

- Vision Djibouti 2035 objectives:
  - Transform into a middle-income economy and logistics/commercial hub
  - Medium-term growth target: 7.5–10 percent per year
  - Target: tripling per capita income and reducing unemployment

- Recent and near-term growth:
  - Growth estimated at 6.5 percent in 2016
  - Growth projected to increase for 2017–19 supported by railroad, multipurpose port, and other public investment

- Fiscal and debt vulnerabilities:
  - Public external debt increased from 50 to 85 percent of GDP in 2015–16
  - All debt sustainability indicators projected to remain above vulnerability thresholds for a prolonged period

- Key policy priorities identified:
  - Strengthened debt policy and management
  - Tax reform to broaden tax bases
  - Structural reforms to improve efficiency and governance of public enterprises and the business climate

### Main Findings and Recommendations
- Long-term growth goal:
  - Transforming Djibouti into a middle-income economy within the next 20 years requires sustained annual GDP growth of about 7 percent
- Inclusiveness and poverty:
  - Poverty declined from 47 to 41 percent between 2002 and 2013; extreme poverty changed from 24 to 23 percent
  - Growth has mainly benefited the upper part of the income distribution
  - Policy priority: protect poor and vulnerable populations from transitional reform costs
- Fiscal policy balance:
  - Fiscal policy must support growth while restoring debt sustainability
  - Recommend growth-friendly tax reforms that increase revenue, reduce burden on the poor, and level the playing field across businesses
  - Rebalance expenditure toward items with high growth multipliers, such as capital expenditure
- Public finance reforms:
  - Medium-term budgeting integrated with a DSA
  - Better fiscal transparency and the introduction of the GFSM 2001/14
  - Use contingency mechanisms, such as a precautionary reserve envelope (PRE), to improve public investment efficiency
- Fiscal and debt rules:
  - Consider introducing fiscal and debt rules after broad national consensus and technical preparatory work
- Business environment reforms:
  - Priority areas: energy, getting credit, taxation, investor protection, property rights
  - Further reforms of public enterprises to reduce costs of energy and communication and improve the business environment

### Growth Inclusiveness and Labor Market (2002–2013 analysis)
- Data and scope:
  - Analysis covers 2002–2013 using EDAM-IS 2002 and 2013 and supplementary EDESIC 2015–16 and employment surveys
  - Surveys not fully comparable; 2013 better captures detailed food consumption and urban population
- Growth and demographics:
  - Average economic growth in the past decade: 4 percent
  - Growth fluctuation: 1.6 percent (2009 crisis) to over 6 percent in recent years
  - Average population growth: about 2.8 percent
  - 75 percent of population younger than 35
  - Economic participation rate: 62 percent
  - Women (age 15–64) labor force participation: 29 percent
  - Women with at least three years of university studies in the labor market: 54 percent (compared to 76 percent of men)
- Employment structure and unemployment:
  - Unemployment rate (2015 survey): 39 percent
  - Unemployment by group: women 49 percent; men 34 percent; rural 59 percent; urban 37 percent
  - Youth unemployment (15–35 years old): 76 percent
  - Employment-to-population ratio: about 25 percent
  - Public sector provides 60 percent of jobs: 46 percent in central government and 14 percent in public enterprises
  - Private sector employs 30 percent: 10 percent formal and 20 percent informal
  - Self-employed: 10 percent
  - Services and construction account for about 13 percent of formal employment
- Poverty lines (2013, per capita annualized):
  - Overall poverty line: DJF 147,936
  - Extreme poverty (food poverty) line: DJF 98,709
- Poverty rates (2013):
  - Overall poverty: 41 percent
  - Extreme poverty: 23 percent

### Regional and Household Poverty Patterns
- Regional poverty incidence (2013):
  - Djibouti-ville and other cities (82 percent of population):
    - Overall poverty: 34 percent
    - Extreme poverty: 17 percent
  - Rest of the country (mainly rural):
    - Overall poverty: about 63 percent
    - Extreme poverty: 45 percent
  - Highest regional poverty: Ali Sabieh and Obock
- Selected regional values (2002–13, as reported):
  - Overall poverty (selected values shown): 62.5; 34.2; 40.8; 73.7; 40.3; 46.7 (percent)
  - Extreme poverty (selected values shown): 44.5; 16.5; 23.0; 51.2; 17.6; 24.1 (percent)
  - Overall poverty gap (selected values shown): 30.7; 11.3; 15.9; 33.7; 12.9; 16.9 (percent)
  - Extreme poverty gap (selected values shown): 18.8; 4.7; 7.9; 18.9; 4.5; 7.4 (percent)
- Gender and household headship:
  - Female-headed households: about 19 percent of households
  - Overall poverty rate: female-headed households 21 percent; male-headed households 21 percent
  - Rural divergence: female-headed households 29 percent; male-headed households 17 percent
  - Average household expenditures: Male-headed households 105 percent of the average; Female-headed households 86 percent of the average; Male average 104.7; Female average 86.3; Overall 99.6; Djibouti-ville 101.4 (percent of the average)
- Inequality:
  - Gini index: 2013 Gini = 0.44; 2002 Gini = 0.39
  - Selected comparator Gini values: Senegal 0.403; Mauritania 0.405; Mozambique 0.457; Cape Verde 0.505; Comoros 0.643; selected comparators average 0.414 (2005–12)

### Decomposition of Poverty Change (2002–2013)
- Overall poverty rate dropped by 5.9 percentage points (ppts):
  - Growth effect contributed 8.8 ppts (poverty-reducing)
  - Distribution effect subtracted 2.9 ppts (poverty-increasing)
- Extreme poverty change marginal: about 1 ppts (not statistically significant)
- Poverty gap outcomes:
  - Overall poverty gap improvement: about 1 percentage point (not statistically significant)
  - Extreme poverty gap increased by about 0.5 percentage points; distribution effect strongly negative and statistically significant
- Tabulated decomposition excerpts (selected preserved values):
  - Parameter alpha = 0 (headcount)
    - Poverty line: 94,636 (2002); 147,936 (2013)
    - Distribution_1: 0.4670; St. error 0.0213; Lower bound 0.4251; Upper bound 0.5089
    - Distribution_2: 0.4081; St. error 0.0223; Lower bound 0.3641; Upper bound 0.4520
    - Difference (d2-d1): -0.0589; St. error 0.0454; Lower bound -0.1483; Upper bound 0.0304
    - Growth: -0.0908; St. error 0.0542; Lower bound -0.1976; Upper bound 0.0159
    - Distribution: 0.0256; St. error 0.0550; Lower bound -0.0825; Upper bound 0.1338
    - Residual: 0.0063
  - Parameter alpha = 1 (poverty gap)
    - Poverty line: 94,636 (2002); 147,936 (2013)
    - Distribution_1: 0.1689; St. error 0.0105; Lower bound 0.1482; Upper bound 0.1896
    - Distribution_2: 0.1586; St. error 0.0104; Lower bound 0.1383; Upper bound 0.1790
    - Difference (d2-d1): -0.0103; St. error 0.0210; Lower bound -0.0517; Upper bound 0.0311
    - Growth: -0.0434; St. error 0.0240; Lower bound -0.0907; Upper bound 0.0038
    - Distribution: 0.0326; St. error 0.0150; Lower bound 0.0031; Upper bound 0.0620
    - Residual: 0.0006
- Statistical significance: many estimates have relatively low statistical significance at the 0.05 level

### Growth Incidence Curve (GIC) and Distributional Dynamics (2002–13)
- GIC findings:
  - Mean of GIC above zero: average household consumption increased
  - Growth was driven by the middle of the distribution (2nd to 10th deciles)
  - GIC is positively sloped, indicating increased inequality
  - Lower deciles experienced lower consumption growth; 95 percent confidence interval for lower deciles is squarely below the horizontal axis
  - Mean growth rate for lower deciles: about 0.2 percent a year
  - Consumption of the poorest percentiles negative; consumption of deciles above the 8th clearly increased
- Overall interpretation:
  - Growth in 2002–13 was not inclusive: benefited mainly upper-income groups while poorest groups became relatively poorer
  - Data limitations and marginal statistical significance temper confidence in results

### Policy Implications to Increase Growth Inclusiveness
- Sustained high growth is necessary but not sufficient; combine with job-creating growth and distributional policies:
  - Invest in education and skills for youth; prioritize sectors that generate employment for less-qualified workers
  - Diversify the economy (tourism, fishing, port services)
  - Expand and target social safety nets; consider medium-term cash transfer system
  - Redirect resources from generalized subsidies to targeted measures (school lunches, public works)
  - Ensure basic utilities provision (electricity, water) with lower subsidized tariffs for basic consumption
  - Gender inclusion: reduce gender inequality to raise per capita growth potential
  - Deepen financial sector: operationalize Partial Credit Guarantee Fund; modernize payments and credit reporting; set up mobile payments; simplify taxation and mortgage procedures

### Debt Sustainability Analysis (DSA) and Fiscal Space
- DSA key result:
  - DSA indicates Djibouti may have some fiscal space after 2027—assuming implementation of the authorities’ reform program—when the debt trajectory drops below the sustainability threshold
- Timing and primary balance:
  - Primary surplus exceeds the debt-stabilizing primary balance beginning in 2023, implying potential fiscal space starting in those years
  - Because debt-to-GDP remains above the high-risk threshold initially, fiscal space should first be used to reduce rate of debt accumulation; only after 2027 consider additional spending
- Debt dynamics and financing:
  - Favorable debt dynamics if real growth > real interest rate on public debt; underscores importance of concessional financing
- Revenue mobilization and tax policy:
  - Tax-to-GDP ratio is the third highest among comparator countries but narrowing due to exemptions and special regimes
  - Measures to raise revenue: remove tax exemptions; improve compliance; make incentives part of common law; convert investment code into investor charter; strengthen tax and customs administration; consider rebalancing from income to consumption taxes
  - Growth-friendly reforms: reduce/simplify personal and corporate income taxes, shift toward consumption taxes, increase tax incentives for private R&D, reduce differentiated tax treatment across assets
- Fiscal multipliers (five-year cumulative based on MENAP template):
  - Capital expenditure: 1.11
  - Government consumption: 0.84
  - Subsidies and transfers: 0.52
  - Tax revenue: -0.43
- Fiscal transparency and public enterprises (PEs):
  - Public accounts cover only central government; over 40 public enterprises exist and are excluded from fiscal accounts
  - PEs are main conduit for external financing; some have difficulty servicing debt; government has made payments on their behalf
  - Implementing IMF’s Fiscal Transparency Code and GFSM 2001/14 recommended to improve reporting, forecasting, risk management, and coverage of PEs
- Institutional measures:
  - Medium-term fiscal frameworks (MTFs) recommended
  - Consider establishing an independent fiscal council to enhance credibility and oversight

### Precautionary Reserve Envelope (PRE) and Fiscal Rules
- PRE: mechanism to improve public investment quality by setting aside part of investment appropriations and releasing funds only after feasibility studies
  - Suggested PRE amount: at least 1 percent of GDP for domestically financed investment projects
  - Procedures: Ministry of Budget to release financing after approving feasibility studies; remove or postpone projects with unsatisfactory studies
- Fiscal rules rationale and types:
  - Debt rules (DR), Budget balance rules (BBR), Expenditure rules (ER), Revenue rules (RR)
  - Typical numerical benchmarks:
    - Debt rules in emerging markets: 40 to 60 percent of GDP
    - Budget balance targets: 1 to 3 percent of GDP
    - Examples: Cape Verde debt limit 60 percent of GDP; Mauritius public debt-to-GDP below 50 percent by 2018; EU SGP deficit limit 3 percent of GDP and debt limit 60 percent of GDP
- Institutional design features for success:
  - Enshrine rules in high-level legislation; top-down budgeting; fiscal responsibility law; independent bodies; automatic correction mechanisms; specified escape clauses; reliable data and forecasting capacity
- Recommended debt-law and debt-management measures for Djibouti:
  - Develop medium-term debt strategy appended to the annual budget
  - Announce explicit debt-to-GDP threshold for next five years and include it in the budget
  - Re-run DSA before any new loan exceeding 0.5 percent of GDP is contracted
  - Establish database to monitor debt contracted by PEs and guarantees; regularly publish comprehensive debt statistics
  - Limit public borrowing to projects with high rates of return; limit non-concessional borrowing to commercially viable projects
- Short-run constraints:
  - Very limited fiscal space due to high risk of debt distress; little scope for additional debt-financed fiscal support
  - Short-term options: growth-friendly tax base expansion; rebalance expenditure toward investment; target reforms to be revenue-neutral where possible

### Growth Model Estimation and Policy Scenarios
- Model estimation:
  - Panel sample: 109 countries with data for 1990–2012
  - Statistically significant coefficients (GDP growth per capita 1990–2012 regression):
    - Constant: Coef. 9.008; Std. Err. 2.149; t-stat. 4.190; P>|t| 0.000; 95% Conf. interval [4.736, 13.280]
    - GDP per capita in 1990: Coef. -0.978; Std. Err. 0.215; t-stat. -4.550; P>|t| 0.000; 95% Conf. interval [-1.405, -0.551]
    - Doing Business rank: Coef. -0.011; Std. Err. 0.003; t-stat. -3.740; P>|t| 0.000; 95% Conf. interval [-0.017, -0.005]
    - Investment (percent of GDP): Coef. 0.046; Std. Err. 0.020; t-stat. 2.290; P>|t| 0.025; 95% Conf. interval [0.006, 0.087]
    - Real exports growth (percent): Coef. 0.219; Std. Err. 0.042; t-stat. 5.180; P>|t| 0.000; 95% Conf. interval [0.135, 0.304]
  - Number of observations: 109
  - R-squared: 0.661
- Baseline projection for Djibouti (unchanged policies):
  - Given 1990 GDP per capita US$760; 2010 DB rank 163; average investment ratio 21 percent of GDP; annual real export growth 2.2 percent
  - Predicted average annual per capita growth rate in 2017-37: 0.4 percent
  - Predicted very long-run steady-state per capita GDP: about US$2,800 (current level about US$1,700)
- Growth needed to reach middle-income (steady-state per capita GDP ≥ US$5,000 from current US$1,700):
  - 2017–2021 per capita growth needed: in the range of 5–6 percent per year on average
  - With population growth ≈ 2 percent, required real growth ≈ 7–8 percent per year
- Policy simulations:
  - Improving Doing Business rank to Albania (from 154 to 65) would be sufficient to make Djibouti middle-income with steady-state per capita GDP over US$5,000
  - Investment-only: raising average investment ratio to Mongolia (33 percent of GDP) or Indonesia (31 percent of GDP) sustained for 20 years yields per capita income roughly US$2,500–3,300
  - Investment alone reaches required per capita income only if sustained at about 50 percent of GDP for two decades (not feasible if largely debt-financed)
  - Export-led: sustaining 10 percent annual real export growth for 20 years would reach middle-income status (unlikely given Djibouti’s structure)
  - Combined structural reform scenario (recommended):
    - Improve DB rank by about 70 points (to the 100–110 range) and raise investment ratio to about 25–30 percent of GDP → achieve middle-income status by 2035
    - Alternative: advance 40–50 DB ranks and raise investment to Indonesia/Vietnam levels (≈28 percent of GDP) plus double export growth to ≈4 percent annually
- Policy priorities:
  - Business environment reforms (priority)
  - Maintain relatively high investment-to-GDP ratio while shifting financing increasingly to private sector
  - Diversify exports toward tourism, telecommunications, advanced port services
- Caveats:
  - Model limitations: panel may not capture country-specific determinants; DB rank is partial measure; growth depends on exogenous demand for port services, exports to Gulf countries, financial inflows (notably from China)

### Doing Business (DB) Topics: Variation, Constraints, and Reform Priorities
- Overall Doing Business 2018:
  - Djibouti rank: 154
  - Overall distance to frontier (DTF) score: 49.58
  - Population: 942,333
- Selected DB topic ranks and numeric indicators (preserved as reported):
  - Starting a business (rank): 115
    - DTF score: 83.38; Procedures: 7; Time (days): 14; Cost (% of income per capita): 35.0; Minimum capital (% of income per capita): 0.0
  - Getting credit (rank): 183
    - DTF score: 5.00; Strength of legal rights index (0–12): 1; Depth of credit information index (0–8): 0; Credit bureau coverage (% of adults): 0.0; Credit registry coverage (% of adults): 0.4
  - Trading across borders (rank): 159
    - DTF score: 51.87; Documentary compliance (hours): 72; Border compliance (hours): 109; Time to export (days): 14; Cost to export: 95; Border compliance (US$): 944; Time to import (days): 111; Cost to import (% of warehouse value): 5.4; Documentary compliance (US$): 100; Border compliance (US$): 1,209
  - Dealing with construction permits (rank): 84
    - Procedures: 17; Time (days): 125; Cost (% of income per capita): 5,979.9; Building quality control index (0–15): 11.0
  - Protecting minority investors (rank): 96
    - DTF score: 51.67; Extent of disclosure index: 7; Extent of director liability index: 7; Ease of shareholder suits index: 3; Extent of shareholder rights index: 6; Extent of ownership and control index: 5; Extent of corporate transparency index: 3
  - Getting electricity (rank): 169
    - DTF score: 40.75; Procedures: 4; Time (days): 1,025; Reliability index (0–8): 0; Time (hours per year): 76
  - Enforcing contracts (rank): 175
    - Time (days): 1,025; Cost (% of claim): 34.0; Quality of judicial processes index (0–18): 3
  - Paying taxes (rank): 108
    - DTF score: 68.91; Payments (number per year): 35; Total tax rate (% of profit): 37.7
  - Registering property (rank): 168
    - DTF score: 42.65; Time (days): 39; Procedures: 6; Cost (% of property value): 12.7; Quality of land administration index (0–30): 4.5
  - Resolving insolvency (rank): 73
    - Time (years): 2.3; Cost (% of estate): 11.0; Recovery rate (cents on the dollar): 37.5; Strength of insolvency framework index (0–16): 9.0; Postfiling index (0–100): 49.57

- Core access to credit constraints:
  - Strength of legal rights index: 1 (out of 0–12)
  - Depth of credit information index: 0 (out of 0–8)
  - Credit bureau coverage (% of adults): 0.0
  - Credit registry coverage (% of adults): 0.4
  - Practical consequence: only 0.4 percent of adult population has credit history in the registry

- Reform priorities and illustrative targets:
  - Broad-based improvements across DB topics required; illustrative improvement of 30–80 DTF points per category implied to reach overall gain of 70–80 points
  - Targeted areas: access to credit (credit bureaus, credit information, guarantee fund); contract enforcement; protecting minority investors; energy sector; paying taxes; property rights
  - Recent or ongoing reforms: electronic manifest system (trading across borders); new container terminal; property registration improvements; new commercial code (insolvency); strengthened secured transactions system; simplifications in starting a business; law creating new credit information system; improved shareholder protections
  - 2018 DB distance to frontier score improvement: 3.79 percentage points

- Feasibility and growth scenario:
  - Quick advances possible if a critical mass of deep reforms is implemented (examples: Vietnam, Mongolia, Armenia)
  - Under successful reforms, Djibouti could grow by 7 percent annually and reach US$5,000 per capita income subject to macroeconomic stability, debt sustainability, and frontloaded structural reforms
  - Caveat: high growth alone does not guarantee improved inclusiveness; complementary reforms (access to credit, contract enforcement) can support inclusiveness

*Source: wp17270 — https://www.imf.org/-/media/files/publications/wp/2017/wp17270.pdf*

### References .............................................................................................................

### References

### Acknowledgment
- Paper development supported by Daniela Gressani and Eric Mottu; contributions from Louis Dicks-Mireaux.
- Careful reading and comments by S. Bouna, M. Cangul, F. Gwenhamo, J. Karangwa, L. Kohler, and A. Olhaye.
- Presented to Djiboutian authorities during staff visits in November 2016 and April 2017; participants thanked include:
  - Central Bank Governor Ahmed Osman
  - Minister of Budget Boheh Ahmed Robleh
  - Minister of Economy and Finance Ilyas Moussa Dawaleh
  - Minister of Investment Ali Guelleh Aboubaker
  - Secretary General of the Ministry of Budget Simon Mibrathu
  - Director of the Debt Directorate Almis Mohamed Abdillahi
  - Director of the Budget Directorate Othman Sadik Ahmed
- Methodological input from World Bank colleagues J. Lohi and G. Lara.
- Research assistance from J. Chen and L. Zhao; administrative support from M. De Mesa.

### Overview
- Country profile and recent estimates (2016, est.):
  - Nominal GDP: 1,894 million
  - Population: 993,000
  - GDP per capita: US $1,908
  - Currency: Djiboutian franc (DF)
  - Exchange rate: DF 177.7 = US$ 1
  - Unemployment rate: 39 percent
  - Poverty rate: 41 percent
- Key investment projects and costs:
  - Multipurpose port Doraleh: US$580 million
  - Addis-Ababa – Djibouti railway: US$490 million
  - Water pipeline from Ethiopia: US$322 million
- Vision Djibouti 2035 objectives:
  - Transform into a middle-income economy and logistics/commercial hub
  - Medium-term growth target: 7.5–10 percent per year
  - Target: tripling per capita income and reducing unemployment
- Recent and near-term growth:
  - Growth estimated at 6.5 percent in 2016, driven by public sector investments
  - Growth projected to increase for 2017–19 supported by railroad, multipurpose port, and other public investment
- Fiscal and debt vulnerabilities:
  - Public external debt increased from 50 to 85 percent of GDP in 2015–16
  - All debt sustainability indicators projected to remain above vulnerability thresholds for a prolonged period
  - Rising debt service obligations risk crowding out priority social and poverty-reducing spending
- Key policy priorities identified:
  - Strengthened debt policy and management
  - Tax reform to broaden tax bases
  - Structural reforms to improve efficiency and governance of public enterprises and the business climate

### Main Findings and Recommendations
- Long-term growth goal:
  - Transforming Djibouti into a middle-income economy within the next 20 years requires sustained annual GDP growth of about 7 percent
- Inclusiveness and poverty:
  - Poverty declined in the past decade, but inclusiveness has not clearly improved
  - Growth has mainly benefited the upper part of the income distribution
  - Policies should protect poor and vulnerable populations from transitional reform costs
- Fiscal policy balance:
  - Fiscal policy must support growth while restoring debt sustainability
  - Recommend growth-friendly tax reforms that increase revenue, reduce burden on the poor, and level the playing field across businesses
  - Rebalance expenditure toward items with high growth multipliers, such as capital expenditure
- Public finance reforms:
  - Medium-term budgeting integrated with a DSA
  - Better fiscal transparency and the introduction of the GFSM 2001/14
  - Use contingency mechanisms, such as a precautionary reserve envelope, to improve public investment efficiency
- Fiscal and debt rules:
  - Consider introducing fiscal and debt rules, requiring broad national consensus and technical preparatory work drawing on Fund assistance and peer experience
- Business environment reforms:
  - Priority areas include energy, getting credit, taxation, investor protection, and property rights
  - Further reforms of public enterprises to reduce costs of energy and communication and improve the business environment

### I. Growth Inclusiveness in Djibouti
- Scope and data:
  - Analysis covers 2002–2013 using two household surveys (EDAM-IS 2002 and 2013) and supplementary EDESIC 2015–16 and employment surveys
  - Surveys are not fully comparable; 2013 survey collected more detailed food consumption data and focused more on urban population
- Summary finding:
  - Overall poverty fell from 47 to 41 percent between 2002 and 2013
  - Extreme poverty changed from 24 to 23 percent (may not be statistically significant at the 0.05 level)
  - Decomposition shows richest groups benefitted more from growth; the distributional component partly offset benefits of economic growth for the poor
  - To achieve greater poverty reduction and inclusiveness requires sustained high growth plus job creation in sectors with high earning potential for the poor and less-qualified workers; better targeted social policies and attention to regional distribution of spending

A. Introduction
- Definition used:
  - Inclusive growth defined as growth that reduces inequality across rich and poor, males and females, and urban and rural areas
  - Poverty reduction decomposed into growth and distributional components

B. Factors of Growth Inclusiveness
- Growth performance:
  - Average economic growth in the past decade: 4 percent
  - Growth fluctuation: 1.6 percent during the crisis of 2009 to over 6 percent in recent years
  - Average population growth: about 2.8 percent
  - National accounts and demographic statistics are not sufficiently developed; growth and population data rely on estimates
- Labor market and demographics:
  - 75 percent of population younger than 35
  - Economic participation rate: 62 percent
  - Low labor market participation concentrated among young, urban, low-education cohorts
  - Women (age 15–64) labor force participation: 29 percent
  - Women with at least three years of university studies in the labor market: 54 percent (compared to 76 percent of men)
  - Migration pressures, including refugees from Yemen, likely raise migration above reported 13 percent
- Unemployment and employment structure:
  - Unemployment rate (2015 survey): 39 percent
  - Unemployment by group: women 49 percent; men 34 percent; rural 59 percent; urban 37 percent
  - Youth unemployment (15–35 years old): 76 percent
  - Employment-to-population ratio: about 25 percent
  - Public sector provides 60 percent of jobs: 46 percent in central government and 14 percent in public enterprises
  - Private sector employs 30 percent: 10 percent formal and 20 percent informal
  - Self-employed: 10 percent
  - Services and construction account for about 13 percent of formal employment
  - Underdeveloped private sector and capital-intensive investment have limited trickle-down effects; many jobs filled by expatriates due to low domestic skills base
- Consumption and poverty measurement:
  - Poverty lines (2013, per capita annualized):
    - Overall poverty line: DJF 147,936
    - Extreme poverty (food poverty) line: DJF 98,709
  - Poverty rates (2013):
    - Overall poverty: 41 percent
    - Extreme poverty: 23 percent

*Source: wp17270 - References — https://www.imf.org/-/media/files/publications/wp/2017/wp17270.pdf*

### 10.      Regional poverty incidence varies widely, with poverty rates higher in rural

### wp17270 - 10.      Regional poverty incidence varies widely, with poverty rates higher in rural

### Regional poverty incidence
- In 2013, in Djibouti-ville and other cities, where 82 percent of the population reside:
  - Overall poverty: 34 percent
  - Extreme poverty: 17 percent
- In the rest of the country (mainly rural regions):
  - About 63 percent of the population lived in poverty
  - 45 percent in extreme poverty
- The Ali Sabieh and Obock regions have the highest levels of poverty.
- Figure 3 (2002–13) reported:
  - Overall poverty (selected values shown): 62.5; 34.2; 40.8; 73.7; 40.3; 46.7 (percent)
  - Extreme poverty (selected values shown): 44.5; 16.5; 23.0; 51.2; 17.6; 24.1 (percent)
  - Overall poverty gap (selected values shown): 30.7; 11.3; 15.9; 33.7; 12.9; 16.9 (percent)
  - Extreme poverty gap (selected values shown): 18.8; 4.7; 7.9; 18.9; 4.5; 7.4 (percent)

### Gender and poverty
- Female-headed households are about 19 percent of total households.
- Overall poverty rate:
  - Female-headed households: 21 percent
  - Male-headed households: 21 percent
- Rural divergence:
  - Poverty rate of female-headed households in rural areas: 29 percent
  - Poverty rate of male-headed households in rural areas: 17 percent
- Expenditure indicators:
  - Average expenditure per person is broadly equal in both household types.
  - Average household expenditures:
    - Male-headed households: 105 percent of the average
    - Female-headed households: 86 percent of the average
  - Other expenditure ratios reported: Male average 104.7; Female average 86.3; Overall 99.6; Djibouti-ville 101.4 (percent of the average)

### Inequality and distribution
- Gini index estimates:
  - 2013 government household survey: Gini = 0.44
  - 2002: Gini = 0.39
- Inequality is more pronounced in rural areas than in the capital (differences not large).
- Comparative Gini values (selected comparators):
  - Senegal: 0.403
  - Mauritania: 0.405
  - Mozambique: 0.457
  - Cape Verde: 0.505
  - Comoros: 0.643
  - Selected comparators average noted: 0.414 for 2005–12

### Inclusive growth diagnostics — data and concepts
- Inclusive growth criteria used: high, sustained, broad-based growth; creates productive employment; benefits large parts of population; considers gender, regional differences, empowerment, and inclusive institutions.
- Data limitations:
  - Need at least two household surveys with comparable methodology.
  - A substantial share of the population is employed in the informal sector; the 2016 survey estimates informal sector employment at about 20 percent of the economically active population, compared with 46 percent employed by the public administration and 14 percent by public enterprises (DISED, 2016).
  - Differences in survey methods between 2002 and 2013 (food consumption questionnaire detail, housing expenditure estimation) complicate comparisons.
- DISED 2017 household income and expenditure survey (EDAM) aims to apply best practices to improve reliability and disaggregation by region, gender, vulnerability status.

### Change in poverty: decomposition into growth and distribution effects (2002–2013)
- The poverty rate change was decomposed into growth effect (G), distribution effect (D), and residual (R) following Datt and Ravallion (1992).
- Main decomposition results (2002–2013):
  - Overall poverty rate dropped by 5.9 percentage points (ppts).
    - Growth effect contributed 8.8 ppts (poverty-reducing).
    - Distribution effect subtracted 2.9 ppts (poverty-increasing).
  - Improvement in extreme poverty was marginal: 1 ppts (not statistically significant), as distributional losses offset growth gains.
- Poverty gap results:
  - Overall poverty gap improvement: about 1 percentage point (not statistically significant); growth effect almost entirely offset by negative distribution effect.
  - Extreme poverty gap increased by about 0.5 percentage points; distribution effect strongly negative and statistically significant, overwhelming the growth effect.
- Alternative approaches (Datt & Ravallion reference period t1, reference period t2, Shapley) yield broadly similar results across poverty measures:
  - Overall poverty headcount reduction: about 5.9 ppts
  - Overall poverty gap reduction: about 1 ppt
  - Extreme poverty headcount reduction: about 1.1 ppts
  - Extreme poverty gap: increase of about 0.4 percent (distribution effect dominates)
- Statistical significance: confidence intervals suggest relatively low statistical significance at the 0.05 level for many findings.
- Tabulated decomposition excerpts (selected values preserved as reported):
  - Parameter alpha = 0 (headcount)
    - Poverty line: 94,636 (2002); 147,936 (2013)
    - Distribution_1: 0.4670; St. error 0.0213; Lower bound 0.4251; Upper bound 0.5089
    - Distribution_2: 0.4081; St. error 0.0223; Lower bound 0.3641; Upper bound 0.4520
    - Difference (d2-d1): -0.0589; St. error 0.0454; Lower bound -0.1483; Upper bound 0.0304
    - Growth: -0.0908; St. error 0.0542; Lower bound -0.1976; Upper bound 0.0159
    - Distribution: 0.0256; St. error 0.0550; Lower bound -0.0825; Upper bound 0.1338
    - Residual: 0.0063
  - Parameter alpha = 1 (poverty gap)
    - Poverty line: 94,636 (2002); 147,936 (2013)
    - Distribution_1: 0.1689; St. error 0.0105; Lower bound 0.1482; Upper bound 0.1896
    - Distribution_2: 0.1586; St. error 0.0104; Lower bound 0.1383; Upper bound 0.1790
    - Difference (d2-d1): -0.0103; St. error 0.0210; Lower bound -0.0517; Upper bound 0.0311
    - Growth: -0.0434; St. error 0.0240; Lower bound -0.0907; Upper bound 0.0038
    - Distribution: 0.0326; St. error 0.0150; Lower bound 0.0031; Upper bound 0.0620
    - Residual: 0.0006
  - Additional reported threshold and estimation cells for other poverty-line specifications are retained in the source.

### Growth incidence curves (GIC) and distributional characteristics of growth (2002–13)
- GIC concept: plots growth rate of consumption expenditure by consumption expenditure percentiles; informs which deciles benefit from growth.
- Djibouti 2002–13 GIC findings:
  - Household consumption increased on average (mean of the GIC is above zero).
  - Growth was driven by the middle of the distribution (from the 2nd to the 10th deciles).
  - The GIC is positively sloped, suggesting an increase in inequality during this period.
  - Lower deciles experienced lower consumption growth relative to an average Djiboutian; the 95 percent confidence interval for lower deciles is squarely below the horizontal axis.
  - Mean growth rate for lower deciles: about 0.2 percent a year.
  - For roughly the 2nd to the 8th decile, the trend points at a worsening of growth inclusiveness.
- Interpretation:
  - A positively sloped GIC indicates growth has not been inclusive.
  - Improvements in inclusiveness would be signaled by the GIC changing slope from positive to negative and the mean of the GIC and the curve moving up.

### Policy implications and recommendations (as discussed in text)
- The distribution effect working against the growth effect implies the need to promote greater equality alongside strong and sustained growth.
- Policies suggested to ensure both effects reduce poverty:
  - Promote fair taxation.
  - Remove regressive expenditures such as energy subsidies.
  - Level the playing field for business.
  - Support the development of human capital.
- Section 2 of the source is indicated to elaborate on these issues.

*Source: DISED, 2013a; DISED, 2013b; DISED (2015); DISED (2016); content extracted from wp17270 chapter text.*

### 29.      The GIC suggests that consumption of the poorest parts of the population

### wp17270 - 29.      The GIC suggests that consumption of the poorest parts of the population

### Distributional analysis and Growth Incidence Curve (GIC) findings (2002–13)
- The GIC is located below the horizontal axis for the lowest percentile, indicating that the poorest groups experienced a negative growth rate of their living standards in 2002–13.
- The depth of poverty among the poorest of the Djiboutian population increased further.
- The 95 percent confidence interval around the GIC touches the X-axis several times, in particular the 5th to the 8th percentiles, indicating the results are marginally statistically significant.
- As a result, the change of consumption of middle-income groups could have been also negative.
- Consumption of the high-income deciles above the 8th decile clearly increased.
- Overall conclusions from the distributional analysis:
  - (i) the overall poverty incidence in Djibouti has declined in 2002–13, although there has been no significant changes in extreme poverty;
  - (ii) inequality in Djibouti remains high, in particular between different income groups, urban and rural areas, and men and women, and there are indications that inequality worsened;
  - (iii) in 2002–13, growth has not been inclusive, as it benefitted mainly people in the upper side of the income distribution, while the poorest groups became even poorer in relative terms and the very poorest may have become poorer in absolute terms;
  - (iv) the underlying data is marginally sufficient for the growth inclusiveness analysis and is not entirely comparable between surveys, and the statistical significance of most estimates is low.

### Policy options to increase growth inclusiveness
- Sustained overall economic growth is a precondition for poverty reduction:
  - Kraay (2004): growth of average income explains 70 percent of the variation in poverty reduction across countries in the short run.
  - Berg and Ostry (2011): longer growth spells are robustly associated with more equality in the income distribution.
  - Lopez and Servén (2006): for a given inequality level, the poorer the country, the more important is the growth component in explaining poverty reduction.
  - Affandi and Peiris (2012): growth is in general pro-poor; a 1 percent increase in real per capita income leads to about a 2 percent decline in the poverty headcount ratio.
  - Implication for Djibouti: supplement debt-financed capital-intensive growth with job creating growth.
- Distributional dimensions matter:
  - About two-thirds of poverty reduction within a country comes from growth, and greater equality contributes the other third.
  - A 1 percent increase in incomes in the most unequal countries produces a mere 0.6 percent reduction in poverty, while in the most equal countries it yields a 4.3 percent cut (Ravallion, 2013).
  - In Djibouti, with a majority of the population under 35 and services as the driving sector, investment in education to improve youth employment opportunities is likely very important.
  - Emphasis: ensure labor force trained for job market needs and prepare Djibouti nationals—rather than expatriates—to take jobs created during the investment boom.
- Economic diversification:
  - Diversification is essential to develop opportunities in tourism and fishing and reduce reliance on a single sector (services) catering mainly to one client (Ethiopia).
  - Improving the business climate is indispensable.
- Well-designed social policies:
  - Protect the poor and vulnerable from high costs of living by expanding social safety net coverage to the poorest population, extending compulsory health insurance available to government employees to the poorest population, and improving health insurance and social housing.
  - Redirect resources from generalized subsidies to better-targeted measures (e.g., school lunches, public works programs). In the medium term, a well-targeted cash transfer system is the best option for assistance for the poorest.
  - Ensure broad-based and competitive provision of basic utilities; prioritize better electricity and water supply. Lower, subsidized tariffs could be charged for basic consumption levels of water and electricity to benefit the poor.
- Gender inclusion:
  - Gender inequality can impede economic growth; Hakura et al. (2016) find gender inequalities are negatively associated with per capita GDP growth, mainly in low-income countries.
  - Per capita income growth in sub-Saharan Africa could be higher by as much as 0.9 percentage points on average if gender inequality could be reduced to levels observed in fast-growing emerging Asian countries.
  - Well-designed policies to influence women’s opportunities to participate in economic activities can play a major role in alleviating inequalities.
- Inclusive institutions:
  - Institutions that ensure the rule of law, provide adequate access to public services, and protect property and freedom to contract without discrimination accelerate growth (Acemoglu and Robinson, 2012).
- Labor market policies:
  - Creating productive employment throughout the economy is important for inclusive growth (IMF, 2013).
  - In Djibouti, increasing employment opportunities and productivity in rural areas would prompt higher consumption growth among poorer households, though major agricultural employment growth is unlikely given the predominantly desert environment.
  - Policies should focus on upgrading in-demand skills, especially for young adults, and improving conditions for small and medium enterprises to enter the formal economy.
- Deepening the financial sector:
  - Financial development generally increases incomes of the poorest households (Claessens, 2005).
  - Barriers to formal financial access are widespread in Djibouti; microfinance, rural finance initiatives, and credit information sharing could expand credit availability.
  - Specific policy actions for Djibouti:
    - Operationalize the Partial Credit Guarantee Fund;
    - Accelerate implementation of the national strategy for modernization of payments and credit reporting systems;
    - Set up a framework for mobile payments to make financial services accessible to low-income groups;
    - Simplify the taxation system to encourage small enterprises to migrate to the formal sector;
    - Simplify access to land and improve mortgage procedures.

*Source: wp17270 - 29.      The GIC suggests that consumption of the poorest parts of the population (IMF staff analysis and selected citations as presented in the source content).*

### 47.      The DSA indicates that Djibouti may have some fiscal space after 2027—

### wp17270 - 47.      The DSA indicates that Djibouti may have some fiscal space after 2027—

### DSA findings on fiscal space and timing
- The DSA indicates that Djibouti may have some fiscal space after 2027—assuming implementation of the authorities’ reform program—when the debt trajectory drops below the sustainability threshold.
- When the primary surplus is greater than the debt-stabilizing primary balance, a country can increase spending without increasing its debt-to-GDP ratio; the difference between the primary balance and its debt-stabilizing level provides an indicator of fiscal space.
- The DSA shows the primary surplus exceeds the debt-stabilizing primary balance beginning in 2023, implying potential fiscal space starting in those years.
- Because the debt-to-GDP ratio remains above its high-risk threshold initially, this fiscal space should first be used to reduce the rate of debt accumulation so that the debt-to-GDP ratio falls.
- Only after 2027, when debt-to-GDP reaches a more moderate level, should this indicator of fiscal space be considered as possible additional spending.

*Source: DSA, staff estimates.*

### Debt dynamics, interest-growth differential, and financing implications
- If projections of growth higher than the real interest rate on public debt materialize, Djibouti would benefit from favorable debt dynamics.
- This underscores the importance of seeking continued concessional financing with low interest rates.
- Given robust real growth rates and debt contracted primarily on concessional terms, the positive impact of the real growth rate exceeds the impact of the real interest rate (Figure 13).

### Growth-friendly fiscal policies — Tax policies and revenue mobilization
- Fiscal space could be generated by additional efforts to raise domestic revenue.
- Statutory tax rates in Djibouti appear broadly in line with international practices; economic costs of tax rate hikes are likely to be high.
- Djibouti’s tax-to-GDP ratio is the third highest among comparator countries but is narrowing due to widespread tax exemptions and special regimes (e.g., tax-free zones).
- Authorities’ actions to date:
  - A government-sponsored Tax Conference in 2015 launched debate on tax reform pillars: simplifying the fiscal regime, enhancing fiscal equity, improving tax efficiency, and securing fiscal revenues.
  - The 2016 budget included measures to streamline and simplify the tax structure; further tax policy and tax administration measures were planned in 2017.
- Potential revenue-raising measures:
  - Removing tax exemptions and improving taxpayer compliance (broadening the tax base by streamlining tax expenditure, exemptions, and tax regimes for free zones).
  - Making tax incentives part of common law and converting the investment code into an investor charter to level the playing field.
  - Strengthening tax and customs administration (online tax payment, better control systems, risk analysis, transparent tax dispute resolution).
- Considerations on tax structure:
  - Rebalancing taxes from income toward consumption could be considered to support growth.
  - Growth-friendly reforms typically reduce and simplify personal income taxation and corporate income taxes and shift toward consumption taxes, which are generally less distortive for saving and investment decisions.
  - Increasing tax incentives for private research and development spending is recommended to support innovation and productivity growth.
  - Recent research suggests improving tax design to minimize differentiated tax treatment across assets, reduce tax compliance costs, and target tax relief to new firms rather than small firms to reduce distortions and boost aggregate productivity and growth.

### Growth-supporting tax policy examples (Box 1 highlights)
- Portugal:
  - Tax rebalancing from income to consumption: corporate income tax lowered from 25 to 21 percent with a reduced rate of 17 percent for SMEs; VAT increased to 23 percent; removal of some tax incentives.
  - Reduction of incentives for debt financing: deductibility of interest for companies progressively limited.
- Belgium:
  - Reduction of personal income tax offset by increases in environmental taxes and elimination of deductions and exemptions.
  - Reduction of capital income taxation as part of broader revenue-neutral reform.
  - Reallocation of saving towards long-term instruments by extending tax benefits to all financial instruments.
  - Rebalancing of property taxes from transaction to recurrent taxes.

### Timing, political economy, and complementary measures
- Increasing tax revenue is a long-term endeavor requiring strong political and social support.
- Tax reforms should be accompanied by public expenditure reforms focused on social and development spending to build support for a balanced package of revenue and expenditure reforms.

### Expenditure policies — rebalancing and efficiency
- Fiscal space could be created by rebalancing public expenditures toward items that enhance long-term growth and human capital formation (education, healthcare) and directing capital expenditure to projects with highest positive impacts on growth.
- Expenditure consolidation recommendations:
  - Ring-fence growth-enhancing spending.
  - Focus consolidation mainly on growth-neutral spending.
  - Ensure efficiency of pro-growth current expenditure and public investment; positive impact depends strongly on efficiency.
- Constraints and priorities:
  - Administrative capacity constraints limit feasibility of launching new high-impact projects given many complex public investments already under way.
  - High public debt limits scope to increase borrowing for further investment without large adverse impacts on debt sustainability.
  - Priority should be given to maintaining the quality of the capital stock: ensure maintenance spending—financed mainly from domestic resources—is sufficient to maintain productivity of recently created infrastructure and prevent deterioration.
- Rebalancing expenditure toward areas with high multipliers can help create additional fiscal space: investment expenditures are expected to be growth enhancing via demand effects (short term) and supply effects (long term).

### Fiscal multipliers (Box 3) — Djibouti estimates based on MENAP template
- Cumulative five-year multipliers applied:
  - Capital expenditure: 1.11
  - Government consumption: 0.84
  - Subsidies and transfers: 0.52
  - Tax revenue: -0.43
- Five-year dynamic persistence assumptions:
  - Government Consumption: Year 1 0.7; Year 2 0.8; Year 3 0.4; Year 4 0.0; Year 5 0.0
  - Subsidies and Transfers: Year 1 0.4; Year 2 0.5; Year 3 0.3; Year 4 0.0; Year 5 0.0
  - Capital Expenditure: Year 1 0.9; Year 2 1.1; Year 3 0.7; Year 4 0.3; Year 5 0.0
  - Tax Revenue: Year 1 -0.3; Year 2 -0.4; Year 3 -0.2; Year 4 0.0; Year 5 0.0
- Interpretation:
  - Highest multiplier: public investment (capital expenditure), stimulating domestic output directly and indirectly.
  - Government consumption multipliers are relatively high in the short run.
  - Subsidies and transfers have relatively low multipliers.
  - Tax multiplier is low and negative, indicating that raising taxes would depress growth in the short term and tax cuts are less effective than spending increases in stimulating the economy.

### Structural reforms of public finances and fiscal transparency
- Medium-term fiscal frameworks (MTFs) could improve decision making and reduce fiscal risks by increasing visibility about policy intentions:
  - An MTF would articulate fiscal policies into a consistent and sustainable framework and bring multi-year considerations into decision making.
  - An MTF could enhance credibility with explicit commitments on medium-term fiscal objectives and contingency planning and serve as a core communication tool.
- Fiscal transparency shortcomings:
  - Public accounts cover only the central government and do not include public enterprises (PEs), even those posing significant risks; public enterprises are the main conduit for external financing and are excluded from fiscal accounts.
  - There are currently over 40 public enterprises; several new PEs created with the investment boom (railroad management, oil pipeline, multipurpose port).
  - PEs have embarked on large investment programs financed with government-guaranteed external debt, creating substantial contingent external debt liabilities.
  - Some PEs already have difficulties servicing their debt; government has on occasion made payments on their behalf.
- Implementing the IMF’s Fiscal Transparency Code would help increase fiscal space by improving:
  - Fiscal reporting (implement comprehensive Government Financial Statistics Manual 2001/14 for the public sector).
  - Fiscal forecasting and budgeting (factor fiscal space into budget forecasting).
  - Risk analysis and management (ensure contingency reserve to cover calls on government guarantees to PEs).
  - Resource revenue management (apply best practices for potential transit natural gas revenues).
  - Fiscal transparency evaluation (assess impact of increased fiscal transparency on fiscal space and produce sequenced action plan).

### Independent oversight
- Establishing an independent fiscal council could help create space for growth-enhancing expenditure:
  - In countries where fiscal councils are independent and credible, they contribute to fiscal stability and growth.
  - Fiscal councils typically assess public finances independently and provide public information; modalities range from reporting compliance to preparing alternative fiscal projections.

*Source: IMF Country Report No. 17/87, DSA and staff estimates.*

### 61.      Finally, a precautionary reserve envelope (PRE) could be introduced to both

### Finally, a precautionary reserve envelope (PRE) could be introduced to both

### Precautionary Reserve Envelope (PRE): purpose and mechanics
- Objective: improve the quality of public investment and control expenditure by setting aside part of investment budget appropriations and defining procedures to release this envelope during the fiscal year.
- Mechanism and expected fiscal effect:
  - If part of the appropriations is not released during the fiscal years because the underlying investment projects are not ready for financing, this would create additional fiscal space for other expenditure.
- Typical PRE system steps (authorities would):
  - (i) establish a PRE for domestically financed investment projects in the amount of at least 1 percent of GDP;
  - (ii) specify in the budget that the Ministry of Budget will release financing for these projects only after it has approved feasibility studies prepared by line ministries charged with project execution;
  - (iii) ensure that feasibility studies are prepared in line with the best international practices included in the World Bank Investment Assessment Guide; and
  - (iv) remove the projects with no or unsatisfactory feasibility studies from the next budget, or postpone their financing until the study is ready.

### Fiscal Rules for Djibouti: rationale and types
- Rationale:
  - Fiscal rules could be useful to generate and protect the long-term fiscal space needed to support growth.
  - Rules limit options for discretionary measures but can foster a prudent and sustainable fiscal stance and help create fiscal space over the medium term.
- Types of fiscal rules (most countries use combinations):
  - Debt rules (DR): set explicit limit or target for public debt, in percent of GDP.
  - Budget balance rules (BBR): constrain budgetary balances (overall, structural, cyclically adjusted, or balance over the cycle).
  - Expenditure rules (ER): set limits on total, primary, or current spending (absolute terms, growth rates, or percent of GDP over three to five years).
  - Revenue rules (RR): set floors or ceilings on revenues to boost revenue collection or prevent excessive tax burden.
- Usage notes:
  - About 80 percent of all fiscal rules in the world constrain public debt or the budget balance.
  - Expenditure rules are frequently used, mostly in advanced economies; revenue rules are much less common.
  - About 60 percent of countries that use a combination of rules adopt a debt rule that caps overall public debt and a fiscal balance rule.

### International experience and numerical benchmarks
- Debt rules in emerging markets and developing economies typically range from 40 to 60 percent of GDP.
- Coverage: public debt rules mostly encompass general government debt (central government, local government, and entities where central government is the source of 50 percent of revenues).
- Countries with debt limits set at 60 percent of GDP typically start putting in debt brakes at 50 percent.
- Numerical target for budget balance rules typically ranges from 1 to 3 percent of GDP.
- Selected country practices and thresholds (international examples cited):
  - Cape Verde: Short-term debt limited to 60 percent of GDP; soft benchmark domestic debt-to-GDP ratio at 25 percent; domestic borrowing limited to 3 percent of GDP.
  - Mauritius: Public debt-to-GDP ratio below 50 percent to be reached by 2018.
  - Swaziland: Public debt ceiling of 60 percent of GDP, domestic debt ceiling of 25 percent and external debt ceiling of 35 percent.
  - Vanuatu: General government debt below 40 percent of GDP; ex ante balanced budget.
  - The Slovak Republic: Triggers and automatic corrective measures at 50, 53, 55, and 57 percent of GDP with progressively stronger actions.
  - Poland: Corrective actions triggered at 50, 55, and 60 percent of GDP; measures triggered when debt ratio exceeds 55 percent.
  - Hungary: Parliament may not adopt a State Budget Act that allows state debt to exceed 50 percent of GDP; if state debt exceeds 50 percent of GDP, Parliament may only adopt a State Budget Act that contains a reduction in state debt to GDP.
  - EU SGP: fiscal deficit limit of 3 percent of GDP; Maastricht debt limit of 60 percent of GDP with an annual pace of debt reduction introduced in November 2011 governance reform.

### Institutional design features and prerequisites for success
- Enshrine rules in high-level legislation to increase durability across governments.
- Introduce top-down budgeting where aggregate expenditure limits are decided before distribution.
- Support rules with a fiscal responsibility law outlining procedural and transparency responsibilities toward parliament.
- Establish independent bodies (e.g., independent fiscal councils) to enhance credibility.
- Use automatic correction mechanisms (debt brakes) to prevent temporary deviations leading to systematic debt buildup.
- Specify escape clauses that clearly define circumstances for temporary suspension, with a limited range of triggering factors.
- Ensure reliable data availability and technical forecasting capacity to ensure credibility.
- Maintain a budget reporting system and timely release of fiscal data for internal and external monitoring.

### Recommended debt-law and debt-management measures for Djibouti
- Suggested steps the authorities could take:
  - (i) develop a medium-term debt strategy integrated with the macroeconomic framework, appended to the annual budget;
  - (ii) announce an explicit debt-to-GDP threshold deemed sustainable over the next five years and include it in the budget;
  - (iii) commit to implementing corrective measures in subsequent budgets in case of a breach of the threshold;
  - (iv) re-run the DSA before any new loan exceeding 0.5 percent of GDP is contracted;
  - (v) establish a database to monitor external and domestic debt contracted by public enterprises, and the guarantees and collateral set aside by the government on such debt; and
  - (vi) regularly publish comprehensive debt statistics on the public and private sector.
- Additional recommended practices:
  - include a public debt sustainability framework in the macroeconomic framework;
  - limit public borrowing to projects with high rates of return;
  - limit non-concessional borrowing to projects that can be run on a commercial basis.

### Conclusions and policy implications for fiscal space and growth
- Overall assessment:
  - Fiscal policy has substantial potential to support growth in Djibouti, even in a high debt environment.
  - Creating and maintaining fiscal space are main preconditions for continuing to support growth given limited scope to increase borrowing sustainably.
  - A multidimensional approach to fiscal space suggests that, with prudent policies and continued reform efforts, some fiscal space may open in the medium to long run.
  - Additional fiscal space would support projected growth, which if sustained would contribute to a reduction of the level of debt—a rate of growth greater than the rate of debt accumulation.
- Short-run constraints and options:
  - In the short run, the country has very limited fiscal space due to the high risk of debt distress signaled by the DSA.
  - There is very little space for providing additional, debt-financed fiscal support to growth.
  - Targeted policies can open pockets of fiscal space in the short term:
    - Growth-friendly tax reforms to expand the tax base and rebalance toward more growth-friendly consumption taxes—designed and implemented to avoid excessive burden on lower income groups.
    - Rebalance expenditure composition toward items with high growth multipliers, such as investment expenditure.
    - All such reforms should target at least a revenue-neutral outcome to sustain current and upcoming budgetary needs.
- Structural reforms:
  - Medium-term budgeting integrated with a public debt sustainability framework to improve fiscal decision-making, especially for investment.
  - Improve fiscal transparency using the IMF’s Fiscal Transparency Code as a reference.
  - Introduce GFSM 2001/14 (international standard for public accounting) to provide a complete overview of public sector financial flows, assets, and liabilities.
  - Use contingency mechanisms, such as the PRE, to prevent fiscal slippages, improve public investment efficiency, and open additional fiscal space.
- Fiscal and debt rules:
  - Consider introducing fiscal and debt rules, possibly after building broad national consensus through a National Fiscal Rule Conference or Debt Conference.
  - Technical preparatory work could draw on Fund technical assistance and peer experience, including from other countries in similar high debt situations.
- Recent Djibouti actions:
  - Prepared a Debt Law and are working on establishing a national public debt committee.
  - Established a database of all debts contracted by the central government and public enterprises, as well as guaranteed by the government.
  - Considering the appropriate level and configuration of a debt anchor.
  - Following a broadly balanced budget policy targeting small and easily financeable overall deficits in the medium term.
  - Strengthening monitoring of large investment projects.
  - Reduced the tax burden on small and medium-sized enterprises to allow them to contribute better to growth.

*Source: wp17270, extracted content on PRE, fiscal rules, and conclusions for Djibouti.*

### 78.      The parameters of the growth model were estimated using a panel sample of

### wp17270 - 78.      The parameters of the growth model were estimated using a panel sample of

### Model estimation and explanatory variables
- Panel sample: 109 countries with data for 1990–2012.
- Explanatory variables considered (as in Sachs and Warner (1997)): World Bank Doing Business (DB) rank, real export growth, economic complexity, economic diversification, investment rate, percentage of population with some secondary education, average years of schooling, adjusted net saving, mortality rate, life expectancy at birth, log GDP per person in 1990, average annual growth in GDP per capita, percentage of GDP in natural resources, natural resource value added per person, costs of exports, and non-market economy status.
- DB rank: standardized measure of ease of doing business across 190 economies across ten sets of indicators; only harmonized cross-country business climate measure available among the explanatory variables.

### Regression results (GDP growth per capita 1990-2012)
- Variables found statistically significant:
  - Constant: Coef. 9.008; Std. Err. 2.149; t-stat. 4.190; P>|t| 0.000; 95% Conf. interval [4.736, 13.280]
  - GDP per capita in 1990: Coef. -0.978; Std. Err. 0.215; t-stat. -4.550; P>|t| 0.000; 95% Conf. interval [-1.405, -0.551]
  - Doing Business rank: Coef. -0.011; Std. Err. 0.003; t-stat. -3.740; P>|t| 0.000; 95% Conf. interval [-0.017, -0.005]
  - Investment (percent of GDP): Coef. 0.046; Std. Err. 0.020; t-stat. 2.290; P>|t| 0.025; 95% Conf. interval [0.006, 0.087]
  - Real exports growth (percent): Coef. 0.219; Std. Err. 0.042; t-stat. 5.180; P>|t| 0.000; 95% Conf. interval [0.135, 0.304]
- Number of observations: 109
- R-squared: 0.661
- Interpretation of signs:
  - Negative coefficient on initial GDP per capita (1990) implies convergence: higher initial GDP associated with lower subsequent growth.
  - Negative coefficient on DB rank: improvements (lower rank) associated with higher growth.
  - Positive coefficients on investment ratio and real export growth as expected.

### Implications and scenarios for Djibouti (empirical results)
- Comparator performance (selected figures):
  - Djibouti: Doing Business 2018 rank 154; 1990-2014 average investment (percent of GDP) 21.1; 1990-2014 average export growth (percent) 2.2
  - Examples of comparators (Doing business 2018 rank; 1990-2014 investment; 1990-2014 export growth): Armenia 47; 21.1; 8.5. Peru 58; 20.7; 6.2. Albania 65; 28.1; 9.1. Vietnam 68; 28.1; 11.8. Mongolia 62; 32.6; 10.0. Cape Verde 127; 18.3; 10.2.
- Baseline (unchanged policies) projection:
  - Given: 1990 GDP per capita US$760; 2010 DB rank 163; average investment ratio 21 percent of GDP; annual real export growth 2.2 percent.
  - Predicted average annual per capita growth rate in 2017-37: 0.4 percent.
  - Predicted per capita steady state GDP in the very long run: about US$2,800 (current level about US$1,700).
- Growth needed to reach middle-income status (target steady-state per capita GDP at least US$5,000 from current US$1,700):
  - During 2017–2021 per capita growth rate needed: in the range of 5-6 percent per year on average.
  - With population growth of about 2 percent, this translates to a real growth rate exceeding 7-8 percent per year.
  - Comparable countries that have maintained such rates for at least a decade: Armenia, Cape Verde, Vietnam.

### Policy simulations and quantitative policy options
- Static simulation approach: set each explanatory variable to the level of comparator countries and project Djibouti’s per capita income in 2035.
- Improving Doing Business (DB) rank:
  - Improving DB rank to Albania (from 154 to 65) would be sufficient to make Djibouti middle-income with steady-state per capita GDP over US$5,000.
  - Djibouti needs to improve DB rank twofold or gain at least about 100 points (suggested target: gain at least 5 ranks per year for the next 20 years — characterized as very ambitious).
  - Partial-equilibrium caveat: other countries may also improve their DB rankings during this period.
- Investment-only scenarios:
  - Raising average investment ratio to Mongolia (33 percent of GDP) or Indonesia (31 percent of GDP) and maintaining for 20 years would yield per capita income roughly US$2,500–3,300.
  - Investment alone could reach required per capita income only if investment ratio sustained at about 50 percent of GDP for two decades (equivalent to the 2015 investment boom peak) — noted as not feasible if investment is largely debt-financed given high risk of debt distress.
- Export-led scenarios:
  - Sustaining 10 percent annual real export growth for 20 years (as in Mongolia, Vietnam, Cape Verde, Lao PDR) would take Djibouti to middle-income status.
  - Likelihood low because rapid export growth in comparators often driven by natural resource booms; Djibouti lacks comparable large exportable natural resources.
  - Djibouti’s service exports constitute about 30 percent of GDP and 80 percent of overall exports; recent annual growth rate slightly above 2 percent. Achieving fivefold increase in export growth and sustaining it would require major structural transformation and diversification (e.g., expanding port and transshipment services from planned new ports).
- Combined structural reform scenario (recommended):
  - A strong improvement in business climate by at least 70 points in the DB ranking (placing Djibouti in the 100–110 range), supported by a permanent increase of the investment ratio to about 25–30 percent of GDP (levels of Dominican Republic, Ghana, Sri Lanka), would allow Djibouti to achieve middle-income status by 2035.
  - Advancing only 40–50 DB ranks (to the level of Lao PDR and Sri Lanka, 130–140) would require raising investment to at least Indonesia or Vietnam levels (28 percent of GDP) and doubling export growth to about 4 percent annually (levels of Bhutan and Guyana).
  - Success likely depends on a critical mass of structural reforms generating a virtuous cycle: business environment improvements → higher private investment → higher export diversification.

### Policy guidance and priorities
- Three broad policy areas for reforms to accelerate growth:
  - Business environment reforms (priority).
  - Investment (maintain relatively high investment-to-GDP ratio; shift financing increasingly to private sector to avoid worsening public debt burden).
  - Exports (diversify exports toward services such as tourism, telecommunications, and advanced port services).
- Specific fiscal and investment cautions:
  - Public financing of large-scale investment increases public debt service obligations and raises operations and maintenance requirements for a rising public capital stock, implying the need for adequate medium-term fiscal provisions.
- Role of model and limitations:
  - Model useful to identify key long-term growth factors, estimate required growth to reach steady-state GDP per capita targets, and run policy simulations to prioritize structural reforms.
  - Limitations: panel may not capture all country-specific determinants; DB rank captures business climate only partially; growth in Djibouti depends heavily on exogenous factors (demand for port services by Ethiopia, exports to Gulf countries, level of financial inflows in particular from China, state of international trade) that can produce positive or negative spillovers.

### Progress and reform actions on Doing Business indicators (selected recent measures)
- Reforms implemented in recent years:
  - Trading across borders: implemented electronic manifest system; reduced documentation requirements; improved port administration; built a new container terminal.
  - Property registration: improved efficiency at the department of property of the Ministry of Finance.
  - Resolving insolvency: adopted new commercial code allowing preventive settlement, legal redress, or liquidation with clear rules.
  - Getting credit: strengthened secured transactions system via new commercial code broadening movable assets usable as collateral.
  - Starting a business: simplified company name search; eliminated minimum capital requirement; removed requirement to publish notice of commencement; opened a single window for company registration (2016-17).
- 2018 DB report additional improvements:
  - Starting a business: exempted new companies from professional license fees; reduced registration and publication fees.
  - Dealing with construction permits: reduced cost of concrete inspections; implemented decennial liability for professionals in construction projects.
  - Registering property: increased transparency of land administration system.
  - Getting credit: adopted law creating a new credit information system.
  - Protecting minority investors: increased disclosure and remedies, extended access to corporate information before trial, increased shareholder rights and corporate transparency.
  - Distance to frontier score improvement: 3.79 percentage points.
- 2018 DB topic ranks (out of 10 topics; ranks where shown):
  - Better than overall rank 154: starting a business (115), dealing with construction permits (84), protecting minority investors (96), paying taxes (108), resolving insolvency (73).
  - Worse than overall rank 154: getting electricity (169), registering property (168), getting credit (183), trading across borders (159), enforcing contracts (175).

*Source: IMF staff estimates and World Bank Doing Business database as summarized in the supplied content.*

### 96.      Djibouti’s DB environment in individual topics varies substantially (Figure 21).

### wp17270 - 96.      Djibouti’s DB environment in individual topics varies substantially (Figure 21)

### Variation across Doing Business (DB) topics — key findings
- In the 2018 Doing Business ranking, Djibouti’s ease of doing business rank is 154 and overall distance to frontier (DTF) score is 49.58.
- High relative ranks that improve the overall score:
  - Resolving insolvency: rank 73; DTF score 48.32.
  - Dealing with construction permits: rank 84; DTF score 68.48.
- Low relative ranks that worsen the overall score:
  - Enforcing contracts: rank 175; DTF score 34.78.
  - Getting credit: rank 183; DTF score 5.00.

### Detailed Doing Business indicators (preserve original numeric values)
- Population: 942,333
- Starting a business (rank): 115
  - DTF score for starting a business (0–100): 83.38
  - Procedures (number): 7
  - Time (days): 14
  - Cost (% of income per capita): 35.0
  - Minimum capital (% of income per capita): 0.0
- Getting credit (rank): 183
  - DTF score for getting credit (0–100): 5.00
  - Strength of legal rights index (0–12): 1
  - Depth of credit information index (0–8): 0
  - Credit bureau coverage (% of adults): 0.0
  - Credit registry coverage (% of adults): 0.4
- Trading across borders (rank): 159
  - DTF score for trading across borders (0–100): 51.87
  - Documentary compliance (hours): 72
  - Border compliance (hours): 109
  - Time to export: Time (days) 14; Cost to export 95; Border compliance (US$) 944
  - Time to import: Time (days) 111; Cost to import Cost (% of warehouse value) 5.4; Documentary compliance (US$) 100; Border compliance (US$) 1,209
- Dealing with construction permits (rank): 84
  - Procedures (number): 17
  - Time (days): 125
  - Cost (% of income per capita): 5,979.9
  - Building quality control index (0–15): 11.0
- Protecting minority investors (rank): 96
  - DTF score for protecting minority investors (0–100): 51.67
  - Extent of disclosure index (0–10): 7
  - Extent of director liability index (0–10): 7
  - Ease of shareholder suits index (0–10): 3
  - Extent of shareholder rights index (0–10): 6
  - Extent of ownership and control index (0–10): 5
  - Extent of corporate transparency index (0–10): 3
- Getting electricity (rank): 169
  - DTF score for getting electricity (0–100): 40.75
  - Procedures (number): 4
  - Time (days): 1,025
  - Reliability of supply and transparency of tariffs index (0–8): 0
  - Time (hours per year): 76
- Enforcing contracts (rank): 175
  - Time (days): 1,025
  - Cost (% of claim): 34.0
  - Quality of judicial processes index (0–18): 3
- Paying taxes (rank): 108
  - DTF score for paying taxes (0–100): 68.91
  - Payments (number per year): 35
  - Total tax rate (% of profit): 37.7
- Registering property (rank): 168
  - DTF score for registering property (0–100): 42.65
  - Time (days): 39
  - Procedures (number): 6
  - Cost (% of property value): 12.7
  - Quality of land administration index (0–30): 4.5
- Resolving insolvency (rank): 73
  - Time (years): 2.3
  - Cost (% of estate): 11.0
  - Recovery rate (cents on the dollar): 37.5
  - Strength of insolvency framework index (0–16): 9.0
  - Postfiling index (0–100): 49.57

### Access to credit — diagnosis
- Core constraints identified:
  - Strength of legal rights index is 1 (out of 0–12).
  - Depth of credit information index is 0 (out of 0–8).
  - Credit bureau coverage (% of adults) is 0.0.
  - Credit registry coverage (% of adults) is 0.4.
- Legal framework features:
  - The law allows businesses to grant security rights without requiring a specific description of collateral.
  - On 11 categories of the strength of legal rights index, Djibouti is ranked at zero.
- Practical consequence:
  - Only 0.4 percent of the adult population has credit history and is included in the credit registry.
  - Credit registries are not yet operational; the index on the depth of credit information is zero.

### Reform priorities and illustrative targets
- Broad-based reforms across all DB areas are necessary; focusing only on one area will be insufficient because other countries are also improving their frameworks.
- Under the assumption that Djibouti should gain at least 70–80 points in the overall DB rank to raise per capita income to the level of a middle-income economy, an illustrative calculation suggests:
  - An improvement on each individual DB category should be in the order of 30–80 points, depending on its current DTF score.
- Targeted reform areas explicitly noted:
  - Access to credit (credit bureaus, credit information, guarantee fund).
  - Contract enforcement.
  - Protecting minority investors.
  - Energy sector (getting electricity; trans-border energy supply and geothermal projects).
  - Paying taxes (major reforms under consideration following the 2015 National Tax Conference).
  - Property rights and other deep legal and institutional reforms.

### Feasibility, timing, and growth scenario
- Quick advances are possible and have been observed in Vietnam, Mongolia, and Armenia when a critical mass of deep reforms was implemented.
- In Djibouti, several reforms are already under consideration or nearing completion in specific DB areas (paying taxes, access to credit, access to electricity).
- Growth scenario under successful reform implementation:
  - Djibouti could grow by 7 percent annually, with particularly high growth in the next few years—as needed to reach its per capital income target.
  - The goal of reaching $5,000 per capita income is achievable, subject to strong macroeconomic policy implementation, debt sustainability, and a critical mass of deep frontloaded macro-structural reforms focused on improving the business environment.
- Caveat:
  - Such growth rates cannot be achieved under unchanged policies, even with very high investment rates.
  - Achieving high growth is not a guarantee of improved inclusiveness, though many business-climate reforms (easier access to credit, better contract enforcement) can help improve inclusiveness.

*Source: IMF staff analysis in wp17270 (figures and text as provided).*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2017/wp17270.pdf_