## 1. Structural Indicators

## Source details

**Canonical URL:** [1. Structural Indicators](https://www.imf.org/-/media/files/publications/cr/2022/english/1stpea2022002.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2022/english/1stpea2022002.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2022/english/1stpea2022002.pdf.json)

---

### Introduction and context
- Uses the dynamic macroeconomic Sustainable Development Goals (SDGs) financing framework to estimate the infrastructure financing gap in São Tomé and Príncipe (STP).
- Paper objectives:
  - estimate STP’s human capital and infrastructure gap in 2030 based on the authorities’ current policies;
  - estimate how active policies—improved domestic revenue mobilization, enhanced spending efficiency, and efforts to attract private investment—could reduce this infrastructure gap;
  - determine the residual financing to be sought from STP’s development partners in grant financing.
- Key administrative and programmatic context:
  - National Development Plan (NDP) cost: US$84.3 million (15.8 percent of GDP).
  - Identified financing gap: US$57 million (10.7 percent of GDP).
  - Model draws on Benedek et al. (2021) dynamic macroeconomic SDG financing framework and uses Stochastic Frontier (SF) models for tax effort and public spending efficiency estimation.

### Stylized facts and vulnerabilities
- Demographics and structural features:
  - Population: about 200,000.
  - Close to 40 percent of the population do not have access to electricity.
  - More than 30 percent do not have access to drinking water.
- Economic and fiscal context:
  - Core sectors: subsistence agriculture and fisheries.
  - Key exports: cocoa products and tourism services.
  - Imports: food, fuel, and other essentials.
  - Weak tax revenues; grants remain an important source of financing.
  - Obsolete fossil fuel-based electricity generation and loss-making state-owned energy enterprises (SOEs) hinder efficient functioning.
- Recent trends:
  - GDP growth rate has declined steadily during the last fifteen years.
  - Both tax revenue and official development assistance (ODA) have declined steadily over the last decade.

### Human capital and infrastructure financing needs (to 2030)
- Aggregate additional investment need:
  - STP needs about 14 percent of GDP in additional annual investment to achieve the human capital and infrastructure SDGs by 2030.
- Sectoral additional annual investment needs (percent of 2030 GDP):
  - Health: 2.7 percent of GDP.
  - Education: 3.5 percent of GDP.
  - Electricity: 1.3 percent of GDP.
  - Water and sanitation: 2.0 percent of GDP.
  - Infrastructure (e.g., roads): 4.7 percent of GDP.
- Notes on estimates:
  - Reported in percent of 2030 GDP and account for investment spending efficiency, initial capital stock, and depreciation.
  - Benchmarked with STP’s comparison group and extrapolated from exercises in other countries.

### Domestic revenue mobilization (DRM) findings
- Current tax revenue performance:
  - Current tax revenue-to-GDP ratio: 13 percent.
  - Tax effort: more than 30 percent below the frontier.
- Tax potential:
  - Tax revenue-to-GDP ratio could increase by up to 6.6 percentage points to reach the frontier (from 13 percent to 19 percent).
  - SDG simulations assess increasing tax collection between 2023 and 2025 by the maximum tax potential (2 percentage points per year).
  - VAT implementation: a 15 percent rate with limited exemptions is expected to generate 4 percent of GDP.

### Public spending efficiency findings
- Method:
  - SF models estimate efficiency of public spending in health, education, and investment; efficiency parameters are incorporated into the SDG financing model.
- Efficiency and potential savings (Text Table 2 summary):
  - Investment: Efficiency range 86%–92%; Potential savings reported as 14.0 (percent of GDP in the source table).
  - Education spending: Efficiency range 86%–93%; Potential savings reported as 3.7 (percent of GDP in the source table).
  - Health spending: Efficiency 97%–97%; Potential savings reported as 2.7 (percent of GDP in the source table).
- Interpretation:
  - STP appears close to the frontier for health spending efficiency.
  - Educational outcomes could be improved by about 4 percent with the same level of public education spending.
  - STP could save about 14 percent of GDP on public investment spending (as reported).

### Private sector participation
- Potential contribution:
  - Scenario assumes government measures to improve regulatory framework and business environment lead to a one percentage point increase in private sector financing of SDGs.
  - Measures include public-private partnerships (PPPs), enhanced transparency, clarity in regulatory/legal frameworks, and governance strengthening.

### Simulation results and scenarios
- Baseline and aggregate results:
  - Baseline additional annual investment need to meet SDGs by 2030: about 14 percent of GDP.
  - Without active policies and additional ODA financing, STP would not meet the human capital and infrastructure SDGs before 2051.
- Impact of single measures:
  - Increasing tax collection by 2 percentage points of GDP each year between 2023 and 2025 lowers the additional infrastructure investment need per year to 9.4 percent of GDP to achieve the SDGs by 2030.
  - Boosting spending efficiency to the frontier for investment, education, and health lowers additional investment needs to 13.3 percent of GDP; this scenario improves the debt-to-GDP ratio in 2030 by 3.8 percentage points.
- Combined measures:
  - Revenue and spending efficiency measures together could cover more than 40 percent of SDG financing needs.
  - Gradually increasing private sector investment by one percentage point by 2023 lowers annual investment needs to close the gap by 2030 by 0.8 percentage points of GDP.
  - Altogether, active policies could reduce additional annual investment needs to 7.7 percent of GDP.

### Policy implications (summary of recommended actions)
- Implement VAT by end 2022 and complement it with broader tax policy and administration reforms.
- Improve tax compliance: boost registration, intensify use of technology, and improve management and strategic planning.
- Adopt modern compliance risk management practices, including audits using third-party information.
- Overhaul performance monitoring frameworks (KPIs and rewards).
- Improve public investment efficiency, especially in investment projects and education outcomes.
- Strengthen regulatory and legal frameworks and governance to attract private financing (incl. PPPs).
- Recognize that even with active domestic policies, additional grant financing from development partners will be needed to close residual financing gaps.

---

### Even with active policies, STP would need support from international development partners to close its human capital and infrastructure financing gap by 2030

### Key findings and financing gap estimates
- Additional grants needed from development partners: 7.7 percent of GDP per year if all active policies are implemented.
- Extreme no-action scenario:
  - Government would need to increase debt by at least 14 percent of GDP per year between 2022 and 2030.
- Under current policies:
  - Investment spending needed to achieve SDG targets in 2030: about 63 percent of GDP.
  - Adjusted estimate translates into an average additional annual investment need of about 14 percent of GDP between 2021 and 2030.
- Contribution breakdown of the 14 percent of GDP additional annual infrastructure financing need:
  - Enhancing domestic revenue mobilization could cover 4.6 percent of GDP.
  - Strengthening spending efficiency could cover about 1 percent of GDP.
  - Private sector involvement could cover about 1 percent of GDP.
- Potential domestic public and private resources total: 6.3 percent of GDP per year, leaving a residual gap of 7.7 percent of GDP per year requiring development partner contributions.

### Dynamic Infrastructure Financing Framework — scenario table (percent of GDP, as presented)
- Scenario entries (as presented in Text Table 3):
  - Baseline settings 2051: Infrastructur e SDG targets met 2051; Additional grants per year +0.06; Residual infrastruc ture gap in 2030 2.7; Additional investment need per year 0.0; Public debt in 2030 56.8
  - Baseline settings 2030: Infrastructur e SDG targets met 2030; Additional grants per year 0.00; Residual infrastruc ture gap in 2030 0.0; Additional investment need per year 14.0; Public debt in 2030 53.2
  - Domestic revenue mobilizat 2030: Additional grants per year 0.00; Residual gap 0.0; Additional investment need per year 9.4; Public debt in 2030 53.7
  - Increased spending efficency 2030: Additional grants per year 0.00; Residual gap 0.0; Additional investment need per year 13.3; Public debt in 2030 49.4
  - Combined active fiscal polic 2030: Additional grants per year 0.00; Residual gap 0.0; Additional investment need per year 8.6; Public debt in 2030 49.9
  - Private sector participation 2030: Additional grants per year 0.00; Residual gap 0.0; Additional investment need per year 13.2; Public debt in 2030 52.9
  - Active policies without grant 2030: Additional grants per year 0.00; Residual gap 0.0; Additional investment need per year 7.7; Public debt in 2030 49.8
  - Active policies plus grants 2030: Additional grants per year 0.00; Residual gap 7.7; Additional investment need per year 0.0; Public debt in 2030 49.8
- Source: Fund staff estimates (table entries reproduced exactly as presented).

### Policy implications and recommended actions (financing focus)
- Improve revenue mobilization:
  - Implement VAT in 2022 with a 15 percent rate and limited exemptions.
  - Achieve a tax-to-GDP ratio of 20 percent; increasing by 6 percentage points in the medium term is described as ambitious but achievable.
  - An efficient VAT would boost buoyancy of other taxes.
- Improve spending efficiency:
  - Strengthen institutions governing project appraisal, selection and management.
- Enable private investment:
  - Strengthen institutional frameworks for clarity and transparency; boost private finance to bridge SDG financing gap.
- Seek an increase in ODA financing:
  - Closing the residual gaps would require an increase in ODA from about 5.5 percent to about 13 percent of GDP per year.
  - This level of grant financing is stated to be achievable given the track record in the early 2010s.

### Macroeconomic and modeling framework (overview)
- Framework adapts Benedek et al. (2021) and uses accounting and behavioral equations across real, fiscal and external sectors over 2020-50.
- Real side uses IMF’s Debt, Investment and Growth (DIG) model addressing public-investment-growth nexus and fiscal adjustments.
- Key model elements and notation (as presented):
  - Y = A [K_G,nonbank + θ K_G,bank]^β K_P^α L (H/L)^σ (1−α)
  - Capital dynamics: K_i,t = (1−δ_i,k) K_i,(t−1) + ε_i I_i,t  for i = G,P with 0 < ε ≤ 1.
  - Human capital accumulation: H_t = (1−δ_h) H_(t−1) + ω ξ_(t−1), with ω ∈(0,1).
  - Education/health human-capital input dynamics: ξ_t = (1−ω) ξ_(t−1) + [(e*h)^φ * n^γ]_(t−1), with h = annual nominal spending on health and education; e>0 is efficiency; φ>0 and γ>0 are elasticities.
- Fiscal balance identity determines resources available for SDG spending.

### Annex — estimation methods
- Uses stochastic frontier analysis (SFA) for tax potential and spending efficiency.
- Tax potential models use panel datasets over 1996-2019 with explanatory variables including PPP-adjusted real GDP per capita, consumption, CPI inflation, financial deepening, urban population share, agricultural value-added, investment, and government effectiveness.
- Spending efficiency analyzed for education, health, and infrastructure using composite outcome indicators; explanatory variables include public and private spending, real GDP per capita, and other determinants.
- Estimation results for Public Investment Spending model presented in Text Table 2; results for Education and Health models available from authors upon request.

---

### Fiscal risks, energy sector, and SOEs

### Macroeconomic risks: overview and empirical evidence
- Macroeconomic shocks are relatively frequent and have major implications for public debt.
  - Bova et al (2016): shocks typically hit economies every ten-twelve years, average cost equivalent to 9 percent of GDP.
- Forecast errors and optimistic bias in STP projections:
  - Real GDP growth has been consistently overestimated in the pre-pandemic period.
  - Two factors: optimistic bias in forecasting methodology; materialization of risks causing unexpected slowdowns.
  - 2020 projection: revised to -6.5 percent due to COVID-19; actual real GDP increased around 3.0 percent (still a negative gap of 0.5 percentage points relative to the original budget projection).
  - Inflation is consistently underestimated; fiscal revenues are overestimated.
  - Higher probability actual performance will be lower than projected given historical biases.

### Policy priorities on macro-fiscal function
- Strengthen macro-fiscal unit (Ministry of Finance, Central Bank, Bureau of Statistics) and operationalize the macro-fiscal projections manual.
- Introduce a three-year medium-term framework in budget documentation for parliamentary and civil society scrutiny.

### SOEs: composition and fiscal risks
- SOEs pose high fiscal risk; four 100 percent government-owned SOEs with high risk: EMAE, ENAPORT, ENASA, CORREIOS.
- SOE list and reported government shares (selection):
  - EMAE — Empresa Nacional de Água e Electricidade; Year of creation 1979; Legal form EP; Government share 100.
  - ENAPORT — Empresa Nacional de Portos; Year of creation 1989; Legal form EP; Government share 100.
  - ENASA — Empresa Nacional de Aeroporto; Year of creation 1987; Legal form EP; Government share 100.
  - Correios — Empresa dos Correios; Year of creation 1982; Legal form EP; Government share 100.
  - CST — Companhia Sãotomense de Telecomunicações; Year of creation 1989; Legal form SARL; Government share 49.
  - BISTP — Banco Internacional de São Tomé e Príncipe; Year of creation 1993; Legal form SARL; Government share 48.
  - STP Airways — Transportes Aéreos de São Tomé e Príncipe; Year of creation 2008; Legal form SARL; Government share 38.
  - ENCO — Empresa Nacional de Combustíveis e Óleos; Year of creation 1997; Legal form SARL; Government share 16.

### EMAE: financial distress and government support
- EMAE operational issues:
  - High commercial and technical losses.
  - Electricity generation mainly from inefficient heavy-oil generators.
  - Insolvent: liabilities exceed assets by around 16 percent.
- Government capital transfers:
  - Dbs 33 million cash transfer in 2019.
  - Around Dbs 60 million cash payment in 2020 for arrears clearance and repairs.

### ENCO: quasi-fiscal activities and implicit subsidies
- ENCO accumulated losses from below-market pricing (2006–2014) estimated at around US$ 55 million by end-2015.
- Government agreed to pay this stock over 50 years at US$ 1 million per year.
- Automatic fuel price mechanism (introduced 2017) prevented further losses and reduced the accumulated stock by 47 percent.
- Debt to ENCO from quasi-fiscal activities (US$ million, reported values):
  - 2014: 55
  - 2015: 53
  - 2016: 43.9
  - 2017: 38.5
  - 2018: 37.8
  - 2019: 37.7
  - 2020: 31.8
  - 2021: 29

### ENASA: pre-pandemic and COVID-19 impact
- Pre-pandemic: profitable but liquidity concerns; net worth positive; overall fiscal risk moderate.
- Pandemic: border closures and flight suspensions severely affected operations; cash depletion and equity erosion; risk downgraded from moderate to high in 2020.

### Impact of COVID-19 on SOE tax debt and government measures
- Support measures included suspension of tax payments, exemptions from interest/penalties, moratoria, and suspension of foreclosures.
- Tax debt from SOEs to government increased by about 30 percent.
- Reported tax and social security debt (selected items, values):
  - Taxes/Customs TOTAL: 48.8 (2019) and 61.5 (2020)
  - EMAE taxes: 11.3 (2019) and 12.7 (2020)
  - ENASA taxes: 5.2 (2019) and 9.4 (2020)
  - ENAPORT taxes: 27.4 (2019) and 34.3 (2020)
  - TOTAL (taxes + social security as reported): 62.0 (2019) and 80.8 (2020)

### Quasi-Fiscal Activities (QFAs): definition and good practice
- QFAs: SOE operations for public policy objectives that are not strictly commercial (e.g., below-market fuel prices).
- Risks: weaken SOE performance and accountability; increase fiscal risks; reduce fiscal transparency.
- Good practice:
  - Expense and fully fund net costs of QFAs in the budget and disclose costs in government and SOE financial statements.
  - Disclose type of activity, rationale, opportunity cost, and compensation mechanism.

---

### Policy priorities: SOE restructuring, energy sector reforms, and fuel pricing

### Restructuring and temporary support for SOEs
- Restructure strategic/systemically important troubled SOEs using proper restructuring plans (example: Management Improvement Plan for EMAE).
- Recognize fiscal costs of restructurings explicitly (e.g., assumption of debt or employment liabilities).
- Sell or close non-viable SOEs without critical policy mandate (example: CORREIOS).
- Provide temporary support to otherwise viable SOEs (exhaust internal means first; use temporary guarantees where needed).
- Strengthen MoF capacity and mandate for SOE financial oversight: enhance legal mandate, improve oversight processes, require timely standardized financial information, and require strategic/investment/budget plans with prior review.

### Energy sector risks and recent shocks
- Energy sector as main driver linking macroeconomic and SOE fiscal risks.
- Two major energy crises:
  - 2019: Sonangol reduced fuel supply to STP by one third due to circumstances in Angola and arrears.
  - 2021: Electricity shortages from non-operational oil-based generators due to depreciation and lack of maintenance.
- Complex interconnection:
  - Macroeconomic factors: international political/economic circumstances in Angola, international oil prices, electricity supply instability effects on GDP and revenue.
  - Operational factors: long-standing stock of arrears (government → EMAE/ENCO → Sonangol) and EMAE’s operational deficiencies.

### Energy sector reform agenda and impacts
- 2018 reform agenda (World Bank support) four objectives:
  - i) Reduce cost of electricity generation and shift towards renewables.
  - ii) Reduce peak demand and lower overall electricity consumption.
  - iii) Improve EMAE distribution management to reduce losses and improve collection.
  - iv) Reform tariff structure and strengthen regulatory framework.
- Least Cost Development Plan (LCDP) impacts:
  - Rehabilitation/expansion of hydropower and photovoltaic solar plants.
  - Full implementation would increase renewable share in generation from 5 to 45% by 2024.
  - Full implementation would reduce average cost of energy by around 40 per cent (from 0.25 to 0.14 USD/kWh).
- Management Improvement Plan (MIP) for EMAE:
  - Potential to reduce commercial and technical losses from 33 to 23 percent and raise collection rates from 85 to 95 percent.
- Implementation challenges: systematic delays, complexity, weak leadership, capacity constraints, internal restructuring debates.

### Fuel price policy, recent dynamics, and fiscal impact
- Since August 2021, the difference between administered retail prices and import prices became negative after positive differentials since early 2020.
- Cumulative balance of revenue inflows from price differentials reached Dbs 57.5 million by August 2021 and reduced to around Dbs 36 million by end-December.
- If negative differential persists throughout 2022 it will generate a cost (implicit subsidy) of around Dbs 70 million to ENCO.
- Recommended government actions:
  - Continue automatic fuel price adjustment mechanism and maintain retail prices aligned to international markets (continuous SB).
  - If full pass-through is socially difficult, partially increase retail prices and reduce fuel tax surcharges within formula; staff estimate forgoing around 60 to 80 million dobras in tax surcharges in 2022 to prevent implicit subsidies.
  - Option to avoid: maintain retail prices unchanged and reduce fuel tax surcharges (would forgo a higher portion of tax revenues).

### Automatic fuel price mechanism — components
- Pricing formula components:
  - Total applicable taxes levied on import CIF price: 5 percent duty on gasoline, EMAE and non-EMAE diesel and kerosene; surcharges of 106 percent on gasoline, 66 percent on non-EMAE diesel, and 15 percent on EMAE diesel.
  - Road fund fixed fees: 0.35 dobra/liter gasoline, 0.40 dobra/liter non-EMAE diesel, 0.20 dobra/liter EMAE diesel.
  - Margins and other costs: customs clearance percentage, distribution charges, wholesale and retail margins.
  - ENCO Surcharge (from 2019): 2.50 dobras per liter for gasoline and diesel, 0.50 dobra per liter of kerosene.

### Climate considerations and environmental pricing
- Reforms and pricing policy contribute to emissions reduction by:
  - Changing generation mix toward renewables.
  - Preventing implicit subsidies to ENCO.
  - Modernizing EMAE to achieve cost recovery.
- Assessment of current pricing vs environmental costs:
  - Automatic mechanism (tax surcharge) sufficient to cover global warming and local pollution externalities for gasoline and diesel.
  - Kerosene and electricity consumer prices fall below cost when environmental costs included:
    - For kerosene, an adjustment to the surcharge component would suffice.
    - Electricity prices are relatively high; comprehensive reform aims to reduce supply cost over medium term.
- Medium-term plan: implement VAT in 2022 (levied over fuel products) and modernize tax administration (reorganize Tax Authority (Direção dos Impostos - DI), adopt modern compliance risk management, overhaul performance monitoring/KPIs).

### Medium-term pricing and arrears management
- Once revenue mobilization package implemented and ENCO arrears cleared, authorities could gradually adjust pricing formula to reflect VAT and environmental costs while maintaining total revenue mobilization.
- Continue automatic fuel price mechanism and alignment with international markets to prevent implicit subsidies.

### SOE Health Check Tool: adoption guidance
- Tool outputs: risk rating (profitability, solvency, liquidity), financial ratios, inputs for public sector balance sheet.
- Required inputs: general country/SOE data, abridged financial statements, government-SOE transactions, choice of financial ratios and thresholds.
- Implementation steps: adopt matching table, collect historical audited statements, collect government-SOE transaction data, agree on common methodology and thresholds, update data as available.
- Default ratio example for return on equity thresholds provided (e.g., >15 percent = very low risk; -10 to 0 percent = high risk; worse than -10 percent = very high risk).

---

### Tourism shock and recovery (global and STP-specific)

### Global tourism shock and 2021 recovery
- International arrivals dropped by 73 percent in 2020 (1 billion fewer arrivals than in 2019).
- Estimated loss in export revenues from collapse in international travel: USD 1.3 trillion.
- Loss in 2020 was more than 11 times the loss recorded during the 2009 global economic crisis.
- Africa recorded a 77 percent drop in tourist arrivals in 2020.
- International tourist arrivals in 2021 were 72 percent down compared to 2019.

### Impact on São Tomé and Príncipe (STP)
- Tourist arrivals decreased by 70 percent in 2020 compared to 2019.
- Tourism revenue decreased by 74 percent in 2020 compared to 2019.
- Total number of travelers in 2021: 15,100 compared to 10,786 in 2020.
- Travel and Tourism direct contribution to employment:
  - Increased steadily since 2012 to account for more than 10 percent of total employment.
  - 13.8 percent of total employment in 2019.
  - Employment in Travel and Tourism suffered a 36 percent decline with a loss of more than five thousand jobs.
- Travel and Tourism direct contribution to GDP for STP:
  - 14.7 percent of total economy in 2019 (WWTC data).
  - Declined to 6.1 percent of total in 2020.
- Tourism receipts accounted for more than 70 percent of total service exports in 2019.

### Sector characteristics influencing recovery prospects
- High reliance on international spending: international spending accounted for 84 percent of total travel spending in 2019 and 77 percent in 2020.
- Full dependence on air travel increases susceptibility to travel restrictions.
- Low dependency on business travel: personal tourism contributed 70 percent of travelers and 73 percent of travel exports in 2019.
- 2021 arrivals show recovery to 42 percent of the 2019 level.
- Vaccination progress in source markets (reported):
  - Portugal ~90 percent fully vaccinated by February 2022.
  - France 77 percent fully vaccinated.
  - Angola around 32 percent completed at least one dose and 16 percent fully vaccinated as of February 2021.
- STP COVID-19 epidemiology and vaccination status:
  - 5,932 confirmed cases and 72 deaths reported to WHO from January 2020 to February 2022.
  - 182,000 vaccine doses administered as of February 2022.
  - Around 28 percent of total population fully vaccinated as of February 2022.

### Outlook and policy priorities for STP tourism recovery
- UNWTO and expert expectations:
  - Majority expect return of international arrivals to 2019 levels in 2024 or later.
  - Many expect a rebound in 2022 (mostly Q3) or 2023.
- Authorities expect full recovery to pre-pandemic travel exports by 2023 or 2024.
- Recommended measures:
  - Roll out vaccinations swiftly.
  - Facilitate transport and border entry via travel policies.
  - Promote tourism marketing.
  - Provide liquidity for businesses and protect jobs.
  - Recover confidence through safety and security.
  - Public-private collaboration and harmonized protocols.
  - Encourage innovation, new technologies, and sustainability.

_Italic: Source — 1stpea2022002 - 1. Structural Indicators; 7. Macroeconomic risks stem; 16. Even with active policies...; 18. Policy priorities...; 35–36. Once the revenue mobilization package...; and tourism sections (March 14, 2022)._

### 1. Structural Indicators __________________________________________________________________________ 11

### 1. Structural Indicators

### Introduction and context
- The Selected Issues Paper uses the dynamic macroeconomic Sustainable Development Goals (SDGs) financing framework to estimate the infrastructure financing gap in São Tomé and Príncipe (STP).
- The paper objectives:
  - estimate STP’s human capital and infrastructure gap in 2030 based on the authorities’ current policies;
  - estimate how active policies—improved domestic revenue mobilization, enhanced spending efficiency, and efforts to attract private investment—could reduce this infrastructure gap;
  - determine the residual financing to be sought from STP’s development partners in grant financing.
- Key administrative and programmatic context:
  - The National Development Plan (NDP) cost is estimated at US$84.3 million (15.8 percent of GDP), with an identified financing gap of US$57 million (10.7 percent of GDP).
  - The model and methods draw on the dynamic macroeconomic SDG financing framework (Benedek et al. (2021)) and use Stochastic Frontier (SF) models for tax effort and public spending efficiency estimation.

### Stylized facts and vulnerabilities
- Demographics and structural features:
  - Population of about 200,000.
  - Close to 40 percent of the population do not have access to electricity.
  - More than 30 percent do not have access to drinking water.
- Economic and fiscal context:
  - Subsistence agriculture and fisheries are core sectors; key exports: cocoa products and tourism services; imports: food, fuel, and other essentials.
  - Weak tax revenues; grants remain an important source of financing.
  - Obsolete fossil fuel-based electricity generation and loss-making state-owned energy enterprises (SOEs) hinder efficient functioning.
- Recent trends:
  - GDP growth rate has declined steadily during the last fifteen years.
  - Both tax revenue and official development assistance (ODA) have declined steadily over the last decade.
  - Tax revenue and ODA shown as declining in the authorities’ series (text figures referenced).

### Human capital and infrastructure financing needs (to 2030)
- Aggregate additional investment need:
  - STP needs about 14 percent of GDP in additional annual investment to achieve the human capital and infrastructure SDGs by 2030.
- Sectoral additional annual investment needs (percent of 2030 GDP):
  - Health: 2.7 percent of GDP.
  - Education: 3.5 percent of GDP.
  - Electricity: 1.3 percent of GDP.
  - Water and sanitation: 2.0 percent of GDP.
  - Infrastructure (e.g., roads): 4.7 percent of GDP.
- Notes on estimates:
  - Estimates are based on benchmarking STP with its comparison group and extrapolation from exercises conducted in other countries.
  - Estimates are reported in percent of 2030 GDP and account for investment spending efficiency, initial capital stock, and depreciation.

### Domestic revenue mobilization (DRM) findings
- Current tax revenue performance:
  - Current tax revenue-to-GDP ratio: 13 percent.
  - Tax effort is estimated to be more than 30 percent below the frontier.
- Tax potential:
  - Tax revenue-to-GDP ratio could increase by up to 6.6 percentage points to reach the frontier (i.e., from 13 percent to 19 percent).
  - In the SDG financing simulations, the impact of increasing tax collection is assessed as increasing tax collection between 2023 and 2025 by the maximum tax potential (i.e., 2 percentage points per year).
  - The VAT, if implemented with a 15 percent rate and limited exemptions, is expected to generate 4 percent of GDP.

### Public spending efficiency findings
- Method:
  - SF models estimate efficiency of public spending in health, education, and investment. Efficiency parameters are incorporated into the SDG financing model.
- Efficiency and potential savings (Text Table 2 summary):
  - Investment: Efficiency range 86%–92%; Potential savings reported as 14.0 (interpreted as percent of GDP in the source table).
  - Education spending: Efficiency range 86%–93%; Potential savings reported as 3.7 (percent of GDP in the source table).
  - Health spending: Efficiency 97%–97%; Potential savings reported as 2.7 (percent of GDP in the source table).
- Interpretation:
  - STP appears close to the frontier for health spending efficiency.
  - Educational outcomes could be improved by about 4 percent with the same level of public education spending, given initial conditions.
  - STP could save about 14 percent of GDP on public investment spending (as reported in the source).

### Private sector participation
- Potential contribution:
  - Scenario analyzed assumes government measures to improve the regulatory framework and business environment lead to a one percentage point increase in private sector financing of SDGs.
  - Such measures include public-private partnerships and enhanced transparency and clarity in regulatory/legal frameworks, alongside governance strengthening.

### Simulation results and scenarios
- Baseline:
  - Without active policies and additional ODA financing, STP would not meet the human capital and infrastructure SDGs before 2051 (Text Table 3).
  - Baseline additional annual investment need to meet SDGs by 2030: about 14 percent of GDP.
- Impact of single measures:
  - Increasing tax collection by 2 percentage points of GDP each year between 2023 and 2025 lowers the additional infrastructure investment need per year to 9.4 percent of GDP to achieve the SDGs by 2030.
  - Boosting spending efficiency to the frontier for investment, education, and health lowers the additional investment needs to 13.3 percent of GDP; this scenario improves the debt-to-GDP ratio in 2030 by 3.8 percentage points.
- Combined measures:
  - Revenue and spending efficiency measures together could cover more than 40 percent of the SDG financing needs.
  - Gradually increasing private sector investment by one percentage point by 2023 lowers the annual investment needs to close the gap by 2030 by 0.8 percentage points of GDP.
  - Altogether, active policies could reduce the additional annual investment needs to 7.7 percent of GDP.

### Policy implications (as presented in the source)
- Priorities highlighted by the authorities and the analysis:
  - Implement VAT by end 2022 (preparations at an advanced stage) and complement VAT with broader tax policy and administration reforms.
  - Improve tax compliance: boost registration, intensify use of technology, and improve management and strategic planning.
  - Adopt modern compliance risk management practices, including audits using third-party information.
  - Overhaul performance monitoring frameworks (KPIs and rewards).
  - Improve public investment efficiency, especially in investment projects and education outcomes.
  - Strengthen regulatory and legal frameworks and governance to attract private financing (incl. PPPs).
  - Recognize that even with active domestic policies, additional grant financing from development partners will be needed to close residual financing gaps.

*Source: 1stpea2022002 - 1. Structural Indicators (March 14, 2022).*

### 16.      Even with active policies, STP would need support from international development

### 16.      Even with active policies, STP would need support from international development partners to close its human capital and infrastructure financing gap by 2030

### Key findings and financing gap estimates
- Additional grants needed from development partners would amount to 7.7 percent of GDP per year if all active policies are implemented as described above.
- In the extreme scenario in which no policy action is taken, the government would need to increase debt by at least 14 percent of GDP per year between 2022 and 2030, which would lead to a significant increase in public debt, relative to the baseline scenario.
- Under current policies, the amount of investment spending needed to achieve the human capital and infrastructure SDG targets in 2030 amounts to about 63 percent of GDP.
- Considering the country’s investment spending efficiency, the initial capital stock, and the capital depreciation rate, the estimate translates into an average additional annual investment need of about 14 percent of GDP between 2021 and 2030.
- Out of the annual additional infrastructure financing need of 14 percent of GDP:
  - Enhancing domestic revenue mobilization could cover 4.6 percent of GDP.
  - Strengthening spending efficiency could cover about 1 percent of GDP.
  - Private sector involvement could cover about 1 percent of GDP.
- Potential domestic public and private resources total 6.3 percent of GDP per year, leaving a residual human capital and infrastructure investment gap of 7.7 percent of GDP per year that would need development partner contributions.

### Dynamic Infrastructure Financing Framework — Scenarios (text as presented)
- Text Table 3. Sao Tome and Principe: Dynamic Infrastructure Financing Framework Scenarios (in percent of GDP)
- Scenarios
  - Infrastructur e SDG targets met Additional grants per year Residual infrastruc ture gap in 2030 Additional investment need per year Public debt in 2030
  - Baseline settings2051 + 0.062.70.056.8
  - Baseline settings20300.00.014.053.2
  - Domestic revenue mobilizat20300.00.09.453.7
  - Increased spending efficienc20300.00.013.349.4
  - Combined active fiscal polic20300.00.08.649.9
  - Private sector participation20300.00.013.252.9
  - Active policies without grant20300.00.07.749.8
  - Active policies plus grants20307.70.00.049.8
- Source: Fund staff estimates.

### Policy implications and recommended actions
- Improve revenue mobilization.
  - Implement the VAT in 2022 as scheduled, with a 15 percent rate and limited exemptions.
  - Achieve a tax-to-GDP ratio of 20 percent; increasing the tax-to-GDP ratio by 6 percentage points in the medium term is described as an ambitious but achievable aspiration for STP.
  - An efficient VAT would also boost the buoyancy of other taxes.
- Improve spending efficiency.
  - Strengthen institutions governing project appraisal, selection and management to enhance effectiveness of spending and reduce resources required to achieve the SDGs.
- Enable private investment.
  - Strengthen institutional frameworks to enhance clarity and transparency of regulatory and legal frameworks.
  - Boost private finance to bridge the SDG financing gap, increasing tax revenues, bringing efficiency gains, and enhancing risk sharing between the public and private sectors.
- Seek an increase in ODA financing.
  - Closing the residual human capital and infrastructure financing gaps would require an increase in ODA from the current level of about 5.5 percent to about 13 percent of GDP per year.
  - This level of grant financing for projects is stated to be achievable, given the track record in the early 2010s.

### Macroeconomic and modeling framework (overview)
- The macroeconomic framework adapts Benedek et al. (2021) and relies on a set of accounting and behavioral equations covering real, fiscal and external sectors to evaluate financing strategies to achieve the SDGs over the 2020-50 horizon.
- The real side uses the IMF’s Debt, Investment and Growth (DIG) model addressing the public-investment-growth nexus and fiscal adjustments.
- Key modeling elements and notation (as presented):
  - Y = A [K_G,nonbank + θ K_G,bank]^β K_P^α L (H/L)^σ (1−α)
  - H is human capital; A is total factor productivity; K_G,nb and K_G,nbk are public bankable and public non-bankable capital stocks respectively; K_P is private capital stock; L is labor force; H/L is human capital per worker.
  - Investment I_i,t and depreciation δ_i,k determine capital dynamics: K_i,t = (1−δ_i,k) K_i,(t−1) + ε_i I_i,t  for i = G,P
  - 0 < ε ≤ 1 is the efficiency with which investment spending is transformed into effective capital.
  - Human capital accumulates: H_t = (1−δ_h) H_(t−1) + ω ξ_(t−1), with ω ∈(0,1).
  - ξ_t = (1−ω) ξ_(t−1) + [(e*h)^φ * n^γ]_(t−1), where h is annual nominal spending on health and education; e>0 is efficiency; φ>0 is elasticity; n is share of school-age population; γ>0 is elasticity.
- Fiscal balance identity determines resources available for SDG spending (identity shown in text).

### Annexes — Estimation methods for tax potential and spending efficiency
- A stochastic frontier analysis (SFA) is used for tax potential and spending efficiency estimation.
- Tax potential analysis:
  - Models use economic and socio-political variables to explain tax revenue indicators (Total tax, Goods & Services, Income, Trade) for subgroups STP belongs to (Fragile states, Low-income countries, Sub-Saharan Africa).
  - Panel datasets over 1996-2019; explanatory variables include PPP-adjusted real GDP per capita, consumption, CPI inflation, financial deepening index, share of urban population, agricultural value-added, investment, and government effectiveness.
  - Tax effort (TTE_i) = ε_i; tax frontier (TTF_i) = y_i_TTF_i; tax potential (TTP_i) = TTF_i − y_i.
- Spending efficiency analysis:
  - Efficiency analyzed for education, health, and infrastructure (investment).
  - Education outcome indicator combines out-of-school children, mean schooling years, enrollment and attainment.
  - Health outcome indicator combines life expectancy, infant/child/maternal mortality, and treatment outcomes for tuberculosis, diphtheria and measles.
  - Public investment efficiency includes measures of coverage (access to electricity, water, telecommunications) and infrastructure quality (World Economic Forum).
  - Explanatory variables include public spending, private spending (where available), real GDP per capita, and other determinants.
  - Estimation results for Public Investment Spending model are presented in Text Table 2; results for Education and Health models are available from the authors upon request.

### Fiscal risks and the energy sector
- STP is prone to multiple fiscal risks due to small island, import-dependent economy and dependency on grants; risks include changes in international political conditions and commodity prices (fuel and cocoa).
- The energy sector is highlighted as the main vein for materialization of fiscal risks in STP:
  - Two largest energy SOEs: EMAE (electricity) and ENCO (fuel).
  - EMAE uses obsolete and expensive fossil fuel-based power plants and faces significant commercial and technical losses in distribution.
  - ENCO imports fuel for electricity generation; accumulated losses from below-market fuel pricing were addressed by an automatic fuel pricing adjustment mechanism, but delays in adjusting prices to reflect rising international fuel prices could increase ENCO losses.
- Fiscal Risk Assessment Tool (FRAT) findings:
  - Macroeconomic risks and risks from public corporations are high priority risks for STP.
  - Only 50 percent of the information and data required to support risk quantification and mitigation for macroeconomic risks is available; 25 percent is partially available.
  - For risks stemming from SOEs, the information gap is larger.
- Institutional implications:
  - A macro-fiscal unit has been created and is being strengthened in collaboration with the central bank and the statistics agency to improve data collection and forecasting methodologies.
  - Managing fiscal risks could benefit from a dedicated fiscal risks unit with specialized staff and skills; STP needs to move in this direction.

*Source: Fund staff estimates.*

### 7.       Macroeconomic  risks  stem

### 7.       Macroeconomic  risks  stem

### Macroeconomic risks: overview and empirical evidence
- Macroeconomic risks arise from variations in the macroeconomic parameters supporting fiscal projections, which impact fiscal outcomes.
- Empirical evidence shows a significant impact of macroeconomic shocks on fiscal indicators and targets.
- Macroeconomic shocks are relatively frequent and have major implications for public debt.
  - According to Bova et al (2016), these shocks typically hit economies every ten-twelve years, with an average cost equivalent to 9 percent of GDP.
- Examples of situations that may materialize macroeconomic risks:
  - Slowdown in economic activity.
  - Fluctuations in commodity prices, such as oil, with strong impact in the budget.

### Forecast errors and optimistic bias in STP projections
- Initial assessment of macroeconomic risks in STP is performed by examining forecast errors in recent years in key macroeconomic variables.
- Optimistic bias in projection methodologies:
  - The real GDP growth has been consistently overestimated in the pre-pandemic period.
  - Two factors explain the consistent overestimation:
    - i) an optimistic bias in the current methodology to forecast GDP growth, which should be addressed by the newly created macro-fiscal unit;
    - ii) the materialization of risks which cause unexpected slowdown in economic activity.
  - 2020 projection impacts:
    - The 2020 projection was impacted by the COVID-19 pandemic when, at the early stages of the health crisis, the 2020 GDP projection was revised to -6.5 percent due to the pandemic and all emergency measures implemented to contain the spread of the virus.
    - Given the strong financial support from the international community (international organizations and bilateral donors) implemented during the year, the real GDP actually increased by around 3.0 percent, still a negative gap of 0.5 percentage points relative to the original budget projection.
- Inflation forecasting:
  - Inflation is consistently underestimated.
  - Underestimation is detrimental when expenditure items (payroll, social benefits, health services) are indexed by inflation, leading to underbudgeting and potential arrears or need to find additional funds during budget execution.
  - Underestimation of inflation may also affect revenue side toward higher revenue collection than budgeted.
- Fiscal revenue forecasting:
  - Fiscal revenues are overestimated.
  - As revenues fall short of budget targets, expenditure envelope is adjusted to meet the fiscal deficit target, undermining proper expenditure planning.
- Probability of underperformance:
  - Given the central projection scenario, there is a higher probability that actual performance will be lower than originally projected because of the observed optimistic bias in the historical data.
  - Fan charts for current projections of real GDP growth show evolution of predictive densities for probabilistic deviations of the central scenario (not adjusted and adjusted for historical forecast biases).

### Policy priorities on macro-fiscal function
- Strengthening macro-fiscal function:
  - The recently implemented macro-fiscal unit comprising representatives from the Ministry of Finance, Central Bank and Bureau of Statistics should continue developing processes and methodologies to enhance forecast capacity.
  - The recently developed manual for macro-fiscal projections should be made operational by the agencies involved in projecting macroeconomic parameters and fiscal aggregates.
  - The three-year medium-term framework should be introduced in the budget documentation to allow for scrutiny by the parliament and civil society institutions, generating positive incentives for strengthening macro-fiscal processes over time.

### Risks stemming from SOEs: overview and sector composition
- SOEs pose a high level of fiscal risk to the government.
- Four SOEs with 100 percent government ownership have high risks for profitability, liquidity, and solvency ratios:
  - Electricity and water supply company (EMAE).
  - Port management corporation ENAPORT.
  - Aviation regulation/airport company ENASA.
  - Mail company CORREIOS.
- SOE Health Check Tool applied to audited annual financial statements over 2017-2020; risks were present pre-pandemic and were exacerbated by COVID-19.

- Composition of the SOE sector (as reported by STP Authorities):
  - EMAE — Empresa Nacional de Água e Electricidade; Year of creation 1979; Legal form EP; Government share 100; Sector: Generation, transmission and distribution of electricity, and provision of water and sanitary services.
  - ENAPORT — Empresa Nacional de Portos; Year of creation 1989; Legal form EP; Government share 100; Sector: Regulation and operation of ports on the islands of São Tomé and Príncipe.
  - ENASA — Empresa Nacional de Aeroporto; Year of creation 1987; Legal form EP; Government share 100; Sector: Operation of the international airport on the island of São Tomé.
  - Correios — Empresa dos Correios; Year of creation 1982; Legal form EP; Government share 100; Sector: Postal services.
  - CST — Companhia Sãotomense de Telecomunicações; Year of creation 1989; Legal form SARL; Government share 49; Sector: Terrestrial and wireless access to telephone and internet services.
  - BISTP — Banco Internacional de São Tomé e Príncipe; Year of creation 1993; Legal form SARL; Government share 48; Sector: Commercial banking activity.
  - STP Airways — Transportes Aéreos de São Tomé e Príncipe; Year of creation 2008; Legal form SARL; Government share 38; Sector: Domestic and international air transport.
  - ENCO — Empresa Nacional de Combustíveis e Óleos; Year of creation 1997; Legal form SARL; Government share 16; Sector: Import, wholesale distribution, and retail distribution of gasoline and liquid petroleum gas (LPG).

### EMAE: financial distress and government support
- EMAE is facing severe operational and technical difficulties that undermine its financial performance:
  - High operational (commercial and technical) losses.
  - Electricity generation matrix mostly based on inefficient and obsolete heavy-oil-based generators.
  - Internal management deficiencies.
- EMAE is currently insolvent, and its liabilities exceed its assets by around 16 percent.
- Government capital transfers to support EMAE:
  - Dbs 33 million cash transfer in 2019 to support cash flow.
  - Around Dbs 60 million cash payment in 2020 used in the context of clearance of arrears and to repair some non-operational oil generators and water supply system.

### ENCO: quasi-fiscal activities and implicit subsidies
- ENCO’s quasi-fiscal activities (implicit subsidies) from below-market pricing of oil products have been very detrimental to the company’s financial position.
  - In 2015 the government estimated the accumulated losses incurred by ENCO resulting from below-market prices in place in the period 2006-2014 to have reached around US$ 55 million (approximately Dbs 1.1 billion) by end-2015.
  - This stock was recognized by the government to be paid during 50 years at a pace of US$ 1 million per year.
- Automatic fuel price mechanism introduced in 2017:
  - Prevented further incurrence of losses and reduced the accumulated stock by 47 percent.
  - Under the mechanism, retail prices for all fuel products (gasoline, diesel and kerosene) are set administratively, using pricing formulas based on import CIF prices, applicable taxes, margins and other costs.
  - During historically low international prices in recent years the government kept administrated prices unchanged, which surpassed formula prices and generated positive differentials used to reduce the debt stock.
- Debt to ENCO from quasi-fiscal activities (implicit subsidies), by reported values (US$ million):
  - 2014: 55
  - 2015: 53
  - 2016: 43.9
  - 2017: 38.5
  - 2018: 37.8
  - 2019: 37.7
  - 2020: 31.8
  - 2021: 29

### ENASA: pre-pandemic position and COVID-19 impact
- ENASA had been generating profits pre-pandemic, but with some liquidity challenges:
  - Current assets were sufficient to meet current liabilities.
  - Debt dynamics showed a sustainable path relative to assets.
  - Net worth was positive and overall fiscal risk level was moderate.
  - Concerns existed related to liquidity turnover indicators (debtor turnover days) and accumulation of arrears to suppliers and delayed collection from customers.
- Pandemic impact:
  - Closure of borders, banning of international and domestic flights, and suspension of operations at São Tomé international airport severely affected ENASA’s activities.
  - Impact absorbed by depleting cash resources and eroding equity.
  - No additional government support in 2020; continued business impact may require government support (capital transfers or guarantees).
  - ENASA’s risk rating downgraded to high risk in 2020 from moderate risk over 2017–2019.

### Impact of COVID-19 on SOE tax debt and government measures
- Government policy measures to support companies during the pandemic included temporary suspension of tax payments, exemptions from interest and additional penalties on tax arrears accumulated during the state of emergency, moratoria without penalties on non-complying with tax payment plans established on previous years, and suspension of tax foreclosures.
- Under this policy, tax debt from SOEs to the government increased by about 30 percent.
- Reported tax debt from SOEs to the government (selected items, as reported in SOE financial statements):
  - Taxes/Customs: 48.8 (2019) and 61.5 (2020)
  - EMAE: 11.3 (2019) and 12.7 (2020)
  - ENASA: 5.2 (2019) and 9.4 (2020)
  - ENAPORT: 27.4 (2019) and 34.3 (2020)
  - CORREIOS: 4.9 (2019) and 5.1 (2020)
  - Social Security: 13.2 (2019) and 19.3 (2020)
  - EMAE social security: 0.9 (2019) and 1.0 (2020)
  - ENASA social security: 1.2 (2019) and 4.0 (2020)
  - ENAPORT social security: 10.1 (2019) and 13.0 (2020)
  - CORREIOS social security: 1.0 (2019) and 1.3 (2020)
  - TOTAL: 62.0 (2019) and 80.8 (2020)

### Quasi-Fiscal Activities (QFAs): definition and good practice
- QFAs are operations undertaken by an SOE in pursuit of a public policy objective that are not strictly commercial in nature (for example, below-market administrated fuel prices).
- Risks from QFAs:
  - Weaken SOE performance and accountability.
  - Increase likelihood of fiscal risks materializing and reduce fiscal transparency.
  - Costs of QFAs are often not disclosed or are masked when support is provided via capitalizations, guarantees and loans.
- Good practice:
  - Governments should expense and fully fund the net costs of QFAs in the budget and disclose costs in financial statements of the government and SOEs.
  - Disclosed information should describe the type of activity, rationale for performing it through the state budget, opportunity cost, and compensation mechanism for the SOE.

_Italic: Source — 1stpea2022002 - 7.       Macroeconomic  risks  stem (PDF chapter/section)._

### 18.      Policy priorities should focus on:

### Policy priorities should focus on:

### Restructuring and temporary support for SOEs
- Restructure state-owned enterprises (SOEs) that were already in trouble pre-COVID and that are strategic or systemically important, using proper restructuring plans (example: Management Improvement Plan for EMAE).
- Restructuring plans should be informed by a clear understanding of the drivers of the losses pre- and post-COVID.
- Corresponding fiscal costs should be explicitly recognized by the state (e.g., assumption of debt or employment-related liabilities).
- Recognize that reforms will likely take time to yield results.
- SOEs that are not viable and do not have a critical policy mandate should be sold or closed (example: CORREIOS).
- Provide temporary support to SOEs that, prior to the COVID-19 pandemic, were healthy and whose performance is expected to recover once the crisis subsides:
  - These SOEs should first exhaust other means of managing the turbulence (example: using their own balance sheet as in ENASA).
  - Temporary government support should be provided only where required and, ideally, in a form that limits the immediate impact on the budget (example: temporary guarantees).
- Strengthen capacity and mandate within the Ministry of Finance (MoF) for financial oversight of SOEs:
  - Current unit within the Treasury Directorate oversees SOEs, but legal mandate should be enhanced and overseeing processes improved.
  - The unit should be empowered to regularly collect financial information from SOEs on a timely and standardized manner to regularly analyze financial performance and provide timely advice to the minister.
  - Require strategic plans, investment plans and budget documents for SOEs to be made available and require prior option from the oversight unit before being operationalized by the corporations.

### Addressing fiscal risks from the energy sector in STP — key risks and recent shocks
- The inefficient energy sector set up is the main driver linking macroeconomic risks with specific fiscal risks stemming from SOEs in STP.
- Two major energy crises in recent years:
  - In 2019 Sonangol unilaterally reduced the supply of fuel to STP by one third, influenced by economic circumstances in Angola and the large stock of arrears owed by STP to its main supplier.
  - In 2021 shortages in electricity supply caused by depreciation and lack of maintenance in oil-based electricity generators which became non-operational.
- Complex interconnection of macroeconomic and operational factors (staff figure reference):
  - Macroeconomic: international political and economic circumstances in Angola (via Sonangol), international oil prices, and the effect of electricity supply instability on GDP growth and revenue collection.
  - Operational: a long-standing stock of arrears (government → EMAE and ENCO; EMAE → ENCO; ENCO → Sonangol) and EMAE’s operational difficulties (technical and operational losses, management deficiencies, major dependence on expensive thermal generation).

### Energy sector reform agenda and expected impacts
- 2018 comprehensive reform agenda (with World Bank support) to modernize the Energy Sector with a five-year timeline focused around four objectives:
  - i) Reduce cost of electricity generation and change the generation mix towards renewable sources.
  - ii) Reduce peak electricity demand and lower overall electricity consumption.
  - iii) Improve management of electricity distribution by EMAE to reduce losses and improve collection rates.
  - iv) Reform tariff structure and strengthen regulatory framework.
- Least Cost Development Plan (LCDP) actions and impacts:
  - Rehabilitation and expansion of the only operating hydropower plant on São Tomé Island, development of other hydropower locations, and implementation of photovoltaic solar plants.
  - Full implementation would increase the share of renewable source in the electricity generation mix from 5 to 45% by 2024.
  - Full implementation would lead to a reduction in the average cost of energy in STP by around 40 per cent (from 0.25 to 0.14 USD/kWh), in line with the production cost in the Sub-Saharan region.
- Management Improvement Plan (MIP) for EMAE:
  - Aims to reduce grid losses and raise collection rates, reorganize institutional arrangements, and modernize management processes and capacities.
  - MIP implementation would potentially reduce commercial and technical losses from 33 to 23 percent and raise collection rates from 85 to 95 percent, enabling EMAE to achieve cost recovery in the medium-term.
- Implementation challenges:
  - Systematic delays due to complexity, multiple pillars, weak leadership, capacity constraints at EMAE, and internal discussions on restructuring frameworks.

### Fuel price policy, recent dynamics, and fiscal impact
- Domestic fuel pricing policy is crucial to prevent implicit subsidies to ENCO and contain fiscal risks.
- Since August 2021 the difference between administered retail prices for gasoline, diesel and kerosene and import prices became negative after positive differentials since early 2020:
  - Cumulative balance of revenue inflows from price differentials reached Dbs 57.5 million by August 2021 and was reduced to around Dbs 36 million by end-December.
  - If the negative price differential persists throughout 2022 it will generate a cost (implicit subsidy) of around Dbs 70 million to ENCO.
- Government action recommended to prevent implicit subsidies:
  - Continue relying on the automatic fuel price adjustment mechanism and maintain retail prices aligned to international markets (continuous SB) to prevent implicit fuel subsidies and contain fiscal risks. Ideally, government should increase retail prices to ensure full pass-through of international price changes to domestic prices (under this option, government will not incur implicit subsidies to ENCO).
  - If social considerations do not allow full pass-through, consider partially increasing retail prices and reducing fuel tax surcharges within the formula to offset the negative price differential and prevent implicit subsidies to ENCO:
    - Under this option, staff estimates the government will forgo around 60 to 80 million dobras in tax surcharges in 2022, equivalent to the budget costing of preventing implicit subsidies to ENCO throughout the year.
  - An alternative to be avoided: maintain retail fuel prices unchanged and reduce fuel tax surcharges within the formula to offset the negative price differential. Under this option, the government will need to forgo a higher portion of tax revenues than in the second-best option to offset implicit subsidies to ENCO in 2022.

### Automatic fuel price mechanism — components and fiscal/environmental role
- The automatic fuel price mechanism (introduced in 2017) operationalizes full pass-through of changes in international fuel prices to domestic retail prices and prevents implicit subsidies and build-up of payment arrears to ENCO.
- Pricing formula components:
  - Total applicable taxes levied on the import CIF price: a 5 percent duty charged on gasoline, EMAE and non-EMAE diesel and kerosene and surcharges of 106 percent on gasoline, 66 percent on non-EMAE diesel, and 15 percent on EMAE diesel.
  - Road fund: a fixed fee of 0.35 dobra per liter of gasoline, 0.40 dobra per liter of non-EMAE diesel, and 0.20 dobra per liter of EMAE diesel earmarked for road maintenance.
  - Margins and other costs: (i) cost associated with customs clearance assessed as a percentage of the import CIF price; (ii) distribution charges differentiated per product; (iii) wholesale and retail margins.
  - ENCO Surcharge: effective from 2019 to pay long-lasting accumulated debt to ENCO; currently set at 2.50 dobras per liter for gasoline and EMAE and non-EMAE diesel, and 0.50 dobra per liter of kerosene.
- The government complied with the pricing mechanism and maintained retail fuel prices aligned to international markets, keeping retail prices fixed throughout 2020 and 2021 and benefiting from the large price differential generated by lower international prices in 2020.

### Climate considerations and environmental pricing
- Energy reforms and fuel pricing policy contribute to global emissions reduction efforts by:
  - Changing the electricity generation matrix towards renewable sources.
  - Preventing implicit (and explicit) subsidies to ENCO.
  - Modernizing EMAE to achieve cost-recovery.
- Incorporating environmental costs (CO2 emissions, local air pollution, broader externalities such as road congestion) helps define a socially efficient price for fossil fuel and electricity (methodology references: Parry and others 2021 and earlier IMF work).
- Assessment of current pricing against environmental costs:
  - The automatic fuel price mechanism, particularly the tax surcharge component, is sufficient to cover the costs associated with global warming and local pollution and externalities for gasoline and diesel.
  - Kerosene and electricity current consumer prices fall below cost when environmental costs are included:
    - For kerosene, current retail prices cover supply costs and an adjustment to the surcharge component of the price formula would suffice to offset environmental costs.
    - Electricity prices are already relatively high in STP; the comprehensive reform aims to reduce supply cost over the medium-term.
- The mechanism also generates tax revenues and can be better adjusted once revenue mobilization capacity is strengthened:
  - Medium-term action plan to strengthen revenue mobilization includes effective implementation of a value-added tax (VAT) in 2022 (which will also be levied over fuel products) and modernization of tax administration services:
    - (i) reorganization of the Tax Authority (Direção dos Impostos - DI) to improve management and strategic planning focused on tax compliance;
    - (ii) adoption of modern compliance risk management practices, including audit programs that make use of information from third parties;
    - (iii) overhaul of the current performance monitoring framework including key performance indicators and a rewards program.

*Source: IMF staff report excerpt.*

### 35.      Once the revenue mobilization package has been fully implemented and the stock of

### 35–36. Once the revenue mobilization package has been fully implemented and the stock of arrears to ENCO has been cleared

### Findings on fuel and electricity pricing
- Once the revenue mobilization package has been fully implemented and the stock of arrears to ENCO has been cleared, authorities could gradually adjust the price formula.
- The applicable tax components (duty charge + tax surcharge) of the pricing formula can be adjusted to reflect VAT and other applicable taxes and the corresponding environmental cost (global and local pollution costs and externalities).
- Moving into this direction in the medium-term will enable STP to achieve the socially efficient price for fossil fuel and electricity without forgoing budget revenues.
- Continuing to rely on the automatic fuel price adjustment mechanism and maintaining retail fuel prices aligned to international markets (continuous SB) helps prevent implicit fuel subsidies and contain fiscal risks.

### Policy recommendations (fuel/electricity)
- Continue relying on the automatic fuel price adjustment mechanism and maintaining retail fuel prices aligned to international markets (continuous SB) to prevent implicit fuel subsidies and contain fiscal risks.
- In the medium-term, and making sure the total revenue mobilization is not reduced, the applicable taxes component (duty charge + surcharge) of the pricing formula could be gradually adjusted to:
  - reflect VAT and other applicable taxes; and
  - reflect the corresponding environmental cost (global and local pollution costs and externalities).
- Moving in this direction will enable STP to achieve the socially efficient price for fossil fuel and electricity without forgoing budget revenues.

### SOE Health Check Tool — purpose and key outputs
- The IMF’s SOE Health Check Tool offers a standardized methodology to assess risks from SOEs.
- Key outputs include:
  - a risk rating of SOEs based on financial ratios for profitability, solvency, and liquidity;
  - financial information and ratios of individual SOEs and the SOE sector in aggregate;
  - inputs for the compilation of a public sector balance sheet.
- Risk rating categories: "Very low risk, low risk, moderate risk, high risk, and very high risk." Overall risk ratings are arrived at by taking the arithmetic average of risk ratings of individual financial ratios.

### SOE Health Check Tool — required inputs
- General information (country, currency, number of SOEs, years of analysis);
- Key information about the SOEs (name, sector, type, ownership structure);
- Abridged financial statements (balance sheet and income statement);
- Transactions between government and SOEs (guarantees, on-lending, subsidies, capital injections); and
- A choice of financial ratios and risk rating thresholds.

### Adoption steps and implementation guidance for Ministry of Finance (MoF)
- Adopt a matching table between SOE financial statements and the abridged balance sheet and income statement in the SOE Health Check Tool; staff developed a matching table mapping items from SOEs’ chart of accounts and 2020 financial statements to the tool’s line items.
- Collect and input information from SOE financial statements, ideally including historical audited financial statements for the last five to ten years, as well as provisional data for 2021 and budget information for 2021.
- Collect and input information on financial transactions between SOEs and the government on an SOE-level to assess fiscal risks and arms-length relationships.
- Review the tool’s default settings for financial ratios used and risk thresholds; the tool allows users to choose from a set of 13 indicators and contains preset risk rating thresholds for each ratio.
- Ensure the use of a common methodology across SOEs, agreeing on the choice of ratios and thresholds and implementing version control; over time, consider adapting the tool to sector-specific factors.
- Update data as new information becomes available (e.g., quarterly budget execution reports or submission of audited financial statements).

### Default ratio example and thresholds (as provided)
- Example for return on equity:
  - more than 15 percent corresponds to a very low risk rating;
  - between 8 and 15 percent corresponds to a low risk rating;
  - 0 to 8 percent corresponds to a moderate risk rating;
  - -10 to 0 percent corresponds to a high risk rating;
  - anything worse than -10 percent corresponds to a very high risk rating.

*Source: IMF staff text in the provided content unit.*

### 1.      Global tourism has suffered the worst year on record in 2020 following the

### 1stpea2022002 - 1.      Global tourism has suffered the worst year on record in 2020 following the

### Overview of global tourism shock and 2021 recovery
- International arrivals dropped by 73 percent in 2020 (1 billion fewer arrivals) than in 2019.  
- Estimated loss in export revenues from the collapse in international travel: USD 1.3 trillion.  
- Loss in 2020 was more than 11 times the loss recorded during the 2009 global economic crisis.  
- Africa recorded a 77 percent drop in tourist arrivals in 2020.  
- International tourist arrivals (overnight visitors) in 2021 were 72 percent down compared to the same period of pre-pandemic year 2019.  
- Modest improvement in June and July 2021 driven mainly by vaccinations and reopening, mostly in Europe and the Americas.  
- In Africa, international travel dropped by 77 percent in the first seven months of 2021 compared to the same period of 2019.  
- Annual arrivals to Africa showed a 12 percent increase in 2021 compared to 2020.

### Impact on São Tomé and Príncipe (STP)
- Tourist arrivals decreased by 70 percent in 2020 compared to 2019.  
- Tourism revenue decreased by 74 percent in 2020 compared to 2019.  
- Total number of travelers in 2021: 15,100 compared to 10,786 in 2020.  
- Travel and Tourism direct contribution to employment increased steadily since 2012 to account for more than 10 percent of total employment; Travel and Tourism contributed to 13.8 percent of total employment in 2019.  
- Employment in Travel and Tourism suffered a 36 percent decline with around a loss of more than five thousand jobs.  
- Economy services exports decreased by around 46 percent in the second to fourth quarters of 2020 compared to the same periods of 2019, mainly reflecting the fall in demand for tourism.  
- Travel and Tourism direct contribution to GDP for STP: 14.7 percent of total economy in 2019 (WWTC data), declined to 6.1 percent of total in 2020.  
- Tourism receipts accounted for more than 70 percent of total service exports in 2019.

### Sector characteristics influencing recovery prospects
- High reliance on international spending: international spending accounted for 84 percent of total travel spending in 2019 and 77 percent in 2020; domestic spending was 16 percent of total spending in 2019.  
- Full dependence on air travel increases susceptibility to travel restrictions.  
- Low dependency on business travel: in 2019 personal tourism contributed to 70 percent of total number of travelers and 73 percent of total travel exports.  
- In 2021 tourist arrivals data shows a recovery to 42 percent of the 2019 level.  
- Progress in vaccination among key source markets (as reported): Portugal ~90 percent fully vaccinated by February 2022; France 77 percent fully vaccinated; Angola around 32 percent completed at least one dose and 16 percent fully vaccinated as of February 2021.  
- São Tomé and Príncipe COVID-19 epidemiology and vaccination status: 5,932 confirmed cases and 72 deaths reported to WHO from January 2020 to February 2022; 182,000 vaccine doses administered as of February 2022; around 28 percent of total population fully vaccinated as of February 2022.

### Outlook for tourism recovery
- UNWTO: international tourism saw a modest improvement in 2021 with 15 million more international tourist arrivals (4 percent increase) compared to 2020, but arrivals were still 72 percent lower than 2019.  
- Recovery is uneven across regions; small islands in the Caribbean, Southern Mediterranean Europe, and Central America recorded better performance in 2021.  
- WEO Oct 2021: trade volumes expected to grow almost 10 percent in 2021, moderating to about 7 percent in 2022 and 3.5 percent in medium term.  
- Expert expectations (Nov 2021 and Jan 2022 UNWTO reports): more than half expect a rebound in 2022 (mostly Q3 2022); nearly forty percent expect a potential rebound in 2023; majority expect return of international arrivals to 2019 levels only in 2024 or later.  
- Eurocontrol (Oct 2021) baseline forecast: air traffic in Europe would return to 2019 levels by end of 2023, conditional on vaccine roll-out reaching herd immunity within Europe in 2021, limited travel restriction, and coordinated European approach.  
- Authorities of STP shared views of a full recovery in travel exports to pre-pandemic level expected by 2023 or 2024.  
- Demand-side shifts and opportunities: movement toward familiar, predictable, trusted, low-risk travel; domestic and regional vacations, outdoors, health and safety, flexible policies, collaboration across value chains, innovation and new technologies, and emphasis on social, environmental, and institutional sustainability.

### Policy priorities and measures for STP
- Measures adopted (summarized):  
  - Fiscal measures to increase spending on epidemic prevention and readiness, provide social assistance and protect small businesses.  
  - Monetary policies to ease pressure on the banks and increase liquidity.  
  - External support from international communities including the World Bank and the IMF.  
  - State of emergency declared in March 2020 including suspension of all international passenger flights; three phases of reopening started from June 2020; reopening of hotels, restaurants, commercial flights, and extended operation of commerce and public service to normal hours; in-person classes at universities resumed.  
  - All air passengers arriving in São Tomé and Príncipe are required to have proof of a negative coronavirus PCR test taken in the 72 hours prior to their arrival.  
  - Jobs and skills measures: expansion of social assistance to the most vulnerable and increased support to the disadvantaged; protecting small businesses and employment through salary contributions and state-guaranteed loans; financial assistance to workers who lost their jobs in formal or informal sectors.  
- Recommended continued short-term targeted support given tourism’s systemic importance:  
  - Rolling out vaccinations swiftly.  
  - Travel policies to facilitate transport and border entry.  
  - Promoting marketing for tourism to boost tourism revenues.  
  - Providing liquidity for businesses and protecting jobs.  
  - Recovering confidence through safety and security.  
  - Public and private sector collaboration and harmonizing/coordinating protocols and procedures.  
  - Adding value jobs through new technologies, encouraging innovation and sustainability as the new normal.

*Prepared by Ke Wang; International Monetary Fund.*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2022/english/1stpea2022002.pdf_