## Average Lending Policy Rate Spread

## Source details

**Canonical URL:** [Average Lending Policy Rate Spread](https://www.imf.org/-/media/files/publications/cr/2026/english/1cpvea2026002-source-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2026/english/1cpvea2026002-source-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2026/english/1cpvea2026002-source-pdf.pdf.json)

---

### Interest Rate Channel of Monetary Policy
- Purpose: estimates the pass-through of the monetary policy rate to bank lending rates in Cabo Verde over 2001–May 2025 and tests for asymmetries between rate increases and decreases.
- Context and observations:
  - The Banco de Cabo Verde (BCV) embarked on a monetary policy tightening phase since May 2023.
  - Current tightening phase cumulative change: 225 bps in the policy rate; described as the steepest over the past two decades.
  - Emigrant deposits constitute an average of 30 percent of total deposits in the banking system.
  - Average lending and long-term rates have not adjusted noticeably to recent increases in the BCV policy rate.
  - Emigrant and resident deposit rates have not increased significantly, with a convergence among both rates since 2020.

### Empirical Assessment of Interest-Rate Pass-through
- Data and methods:
  - Monthly BCV data from January 2001 to May 2025.
  - Two models estimated:
    - Error-correction model (ECM) using the two-step Engle and Granger procedure.
    - Local projections (LP) estimating responses at horizons one to six months ahead.
  - Benchmark controls: reserve requirements ratio (RRR), CPI inflation, and a crisis dummy.
  - Robustness controls tested: non-performing loans (NPLs), excess liquidity, and loan-to-deposit ratio (LDR).
- Key quantitative findings:
  - Medium- to long-run policy rate pass-through estimates:
    - Average lending rate: range between 0.19-0.28.
    - Long-term rate: range between 0.13-0.23.
  - Robustness with additional controls (discarding first tightening episode):
    - Average lending rate: range between 0.17-0.3.
    - Long-term rate: range between 0.11-0.15.
  - Average interest rate betas across cycles:
    - Tightening cycles: pass-through falls into a range of -0.75−0.39 percentage points.
    - Loosening cycles: pass-through falls into a range of -0.49−0.66 percentage points.

### Banking System Structure and Transmission Drivers
- Bank structure (as of December 2024) — bank shares of total assets / credit / deposits:
  - BCA: 27.34% / 25.09% / 30.40% — High D-SIB
  - CECV: 26.41% / 29.04% / 30.09% — High D-SIB
  - BIA: 11.14% / 11.81% / 10.96% — Slight systemic importance
- Identified transmission frictions:
  - Bank concentration can weaken transmission by giving banks market power to set interest rates.
  - Possible drivers of weak pass-through: market concentration, risk premium, consumer search costs, excess liquidity, banks’ high funding costs, and the fact that banks do not rely on BCV refinancing.

### Asymmetric Pass-through During Loosening and Tightening Cycles
- Headline asymmetry:
  - Pass-through is larger during tightening episodes compared with easing episodes, but remains limited in absolute terms.
- Quantitative asymmetry (LP model, six-month horizon, long-term rates):
  - Tightening cycles: pass-through of 0.35.
  - Easing cycles: pass-through of 0.17.
  - Interpretation: pass-through during tightening cycles is roughly twice as large as during easing cycles.
- Heterogeneity across episodes:
  - Tightening episodes:
    - Apr 2001–Apr 2002: evidence of full/complete pass-through in the 2001 cycle.
    - Jan 2012–Feb 2014: limited pass-through of around 0.2 after six months.
    - May 2007–Dec 2009 and May 2023– (current) cycles: lending rates broadly insensitive to monetary policy.
  - Easing episodes:
    - Jan 2010–Dec 2012: stronger pass-through, though it took time to materialize.
    - Mar 2014–Apr 2023: pass-through estimated to be practically zero (banks did not decrease lending rates despite policy rate cuts).
- Time variation:
  - Time-varying pass-through estimates indicate the current tightening cycle that started in May 2023 appears broadly insensitive to monetary policy.

### Additional Relevant Quantities and Episodes
- BCV tightening cycles (duration, total increase, number of hikes):
  - Jan 2001 - April 2002: Duration (months) 15; Total increase 100bps; Number of hikes 1
  - May 2007 - Dec 2009: Duration (months) 32; Total increase 175bps; Number of hikes 5
  - Feb 2012 - Feb 2014: Duration (months) 26; Total increase 150bps; Number of hikes 1
  - May 2023 - Oct 2025: Duration (months) 30; Total increase 225bps; Number of hikes 6
- Spread behavior:
  - The average spread between the average lending rate and the policy rate is on average 113 basis points higher during policy easing phases compared to tightening phases.

### Conclusions on Pass-through (Annex summary)
- The policy rate pass-through in Cabo Verde is estimated to be rather limited:
  - Medium- to long-run pass-through ranges:
    - Average lending rate: 0.19 to 0.28.
    - Long-term rate: 0.13 to 0.23.
- Asymmetry:
  - Pass-through is roughly twice as large during tightening episodes compared to easing episodes, though still relatively low.
  - Larger pass-through during tightening cycles appears driven by the 2001 and 2012 hiking cycles; the 2007 and the current May 2023 cycles are broadly insensitive to monetary policy.

---

### Fiscal and Distributional Simulations from VAT Reform Analysis (selected quantitative outcomes)
- Fiscal and macro context:
  - Government revenue expanded to about 20 percent of GDP.
  - Indirect tax exemptions total 4 percent of GDP and are regressive.
  - Top consumption quintile receives CVE 1.55 billion (about 0.8 percent of GDP) in indirect transfers through VAT exemptions, over five times more than the bottom quintile.
  - Financing needs for critical physical and human capital investments are estimated at around 1.5 percent of GDP per year.
  - Debt stock decreased to 110 percent of GDP in 2024 (from 148.5 percent in 2021).
- Model and scenario setup:
  - Model used: Debt Investment and Growth (DIG-Labor) model calibrated to Cabo Verde.
  - Baseline realized fiscal space from VAT streamlining: 1.2 percent of 2024 GDP; VAT rate remains 15 percent.
  - Upside scenario: removal of all indirect tax exemptions generates fiscal space equal to 4 percent of 2024 GDP; VAT rate remains 15 percent.
  - Downside scenario: administrative and political frictions reduce realized fiscal space to 0.6 percent of 2024 GDP; VAT rate remains 15 percent.
  - Public investment scale-up in upside and baseline: infrastructure +1.5 percentage points of GDP and education +0.5 percentage points of GDP relative to the 2024 baseline.
  - Grants assumption: Upside grant inflows 1.0 percent of GDP; Baseline and Downside grant inflows 0.8 percent of GDP.
  - Targeting assumptions for compensatory cash transfers: informal households comprise 60 percent of the population (휙 = 0.6).

### Key macro outcomes across scenarios (preserving reported figures)
- Short-run aggregate consumption (years 1 to 5):
  - Upside: Aggregate consumption increases by 1.1 percent relative to the no‑reform path.
  - Baseline: Aggregate consumption remains below the Upside path and recovers at a moderate pace.
  - Downside: Aggregate consumption remains subdued for longer under constrained transfers.
- Private investment during transition:
  - Upside: Private investment declines temporarily and then increases; at the steady state it stands 10.6 percent above the no‑reform path.
  - Baseline: Similar pattern but with a smaller and slower recovery.
  - Downside: Private investment remains weak for a longer period; crowding‑in occurs later.
- Long‑run aggregate consumption (steady state):
  - Upside: Aggregate consumption stands 12.6 percent above the no‑reform path.
  - Baseline: Aggregate consumption stands 1 percentage point below the Upside scenario.
  - Downside: Aggregate consumption remains below the Baseline scenario; convergence proceeds more slowly.
- Transfers feasible at the steady state (targeted):
  - Upside: Targeted transfers equal 4.2 percent of GDP, which corresponds to 7 percent of initial per‑capita income.
  - Baseline: Targeted transfers equal 1.8 percent of GDP, which corresponds to 3 percent of initial per‑capita income.
  - Downside: Targeted transfers peak at 1.6 percent of per‑capita income.
- Labor-market dynamics:
  - Upside: Formalization accelerates and wages increase earlier.
  - Baseline: Formal hiring proceeds more slowly despite the same public investment path.
  - Downside: Labor reallocation remains sluggish and unemployment improves later.
- Distributional outcomes at the steady state:
  - Gini coefficient declines to 0.378 across scenarios.
  - Poverty headcount:
    - Upside: declines to 0.5 percent of the population.
    - Baseline: declines to 1.9 percent of the population.
    - Downside: declines to 2.3 percent of the population.
- Quantified cash transfer schedules (selected exact figures reported)
  - Baseline scenario: Targeted (Informal Sector) monthly cash transfers (US$/Month) and share of initial per-capita income (%):
    - Year 1: 50.77; 0.2 percent (Universal: 0.46; 0.1 percent)
    - Year 6: 102.69; 0.7 percent (Universal: 1.61; 0.4 percent)
    - Year 11: 154.27; 1.2 percent (Universal: 2.55; 0.7 percent)
    - Year 16: 205.91; 1.6 percent (Universal: 3.53; 0.9 percent)
    - Year 21: 257.13; 1.9 percent (Universal: 4.26; 1.1 percent)
    - Steady state: 11.04; 3.0 percent (Universal: 6.60; 1.8 percent)
  - Upside scenario: Targeted (Informal Sector) monthly cash transfers (US$/Month) and share of initial per-capita income (%):
    - Year 1 - 5: 14.39; 3.9 percent (Universal: 8.60; 2.3 percent)
    - Year 6 - 10: 16.37; 4.4 percent (Universal: 9.78; 2.6 percent)
    - Year 11 - 15: 18.10; 4.9 percent (Universal: 10.82; 2.9 percent)
    - Year 16 - 20: 19.93; 5.3 percent (Universal: 11.91; 3.2 percent)
    - Year 21 - 25: 21.33; 5.7 percent (Universal: 12.75; 3.4 percent)
    - Steady state: 26.17; 7.0 percent (Universal: 15.64; 4.2 percent)

### Downside scenario specifics
- Technical estimate cited: eliminating all VAT exemptions could potentially expand fiscal space by about 1-1.5 percent of GDP, but implementation challenges may hinder achieving this potential.
- Downside assumption: revenue yield from the VAT reform is around 0.6 percent of GDP, due to administrative bottlenecks and political economy frictions.
- Under the downside policy choice:
  - Government maintains commitment to public investment despite underwhelming VAT revenue gains.
  - Adjustment burden falls on social transfers rather than investment.
- Macroeconomic implications:
  - Public investment remains on track, supporting modest productivity and formal sector output improvements.
  - Limited fiscal space forces a sharp contraction in transfers, especially in the early years, resulting in subdued aggregate consumption and private investment.
  - Targeted transfers to informal sector workers are delayed and peak, across steady state, at just 1.6 percent of per capita income.

### Scenario comparisons and dynamics
- Role of compensatory transfers:
  - Compensatory transfers mitigate public education investment gestation lags and accelerate formal sector labor absorption.
  - Upside: back‑loaded compensatory transfers (averaging 2.3 percent of GDP during years 1 to 5 and converging to 4.2 percent of GDP at steady state) boost household disposable income during early years.
  - Baseline and Downside: minimal or delayed transfers weaken early demand for formal sector goods, delaying labor reallocation and private hiring.
- Fiscal-growth feedback:
  - Upside: earlier formal sector growth initiates a self-reinforcing fiscal-growth cycle—earlier output expansion boosts tax revenue, funding further transfers that accelerate demand and labor absorption.
  - Baseline: sluggish formal sector demand delays revenue mobilization, perpetuating low cash transfers and suboptimal labor allocation.

### Calibration and background parameters (selected exact values)
- Trend Per capita GDP growth: 0.04
- Real interest rate: 0.03
- Real interest rate on external commercial public debt: 0.05
- Real interest rate on external private debt: 0.03
- Gross (initial) return on public capital: 0.27
- Gross (initial) return on secondary education: 0.35
- Gross (initial) return on primary education: 0.35
- Initial unemployment: 12 (Average 2020-2023, World Bank data)
- Initial VAT rate: 0.15
- Public investment to GDP ratio: 0.04
- Share of remittances: 0.12
- Share of grants (% GDP): 0.01
- Share of external private debt: 0.11
- Share of domestic debt: 0.32
- Share of concessional debt: 0.61
- Share of external commercial public debt: 0.18

### Policy implications and recommendations (from simulations)
- Streamline VAT exemptions to expand the tax base while keeping the VAT rate at 15 percent to generate fiscal space for public investment without immediate rate changes.
- Prioritize a targeted, phased rollout of compensatory cash transfers to informal households to mitigate short‑term welfare losses:
  - Upside: back‑load transfers, averaging 2.3 percent of GDP during years 1 to 5 and converging to 4.2 percent of GDP at steady state.
  - Baseline: defer broad transfers until growth dividends materialize; transfers converge to 1.8 percent of GDP at steady state.
- Finance medium‑term public investment through a mix of base broadening, grants, and concessional borrowing to avoid sharp near‑term fiscal strain.
- Strengthen tax administration and communications to reduce administrative and political frictions that could limit realized fiscal space (downside risk: realized fiscal space reduced to 0.6 percent of GDP).
- Maintain attention to public investment absorptive capacity and operations and maintenance funding to realize productivity gains that support long‑run distributional improvements.
- Use projected productivity and formalization gains from public investment to create endogenous fiscal space for sustainable, targeted transfers that support poverty reduction and inequality decline.
- Strengthen compliance and the social registry to reduce leakage, promote formalization, and enable scaling-up benefits as revenues increase.

*Source: Cabo Verdean authorities and IMF staff estimates; annex prepared by Bruno Albuquerque and Nombulelo Gumata; staff simulations and staff computations as reported in the supplied content unit.*

### 1. Average Lending Policy Rate Spread _____________________________________________________ 5

### Interest Rate Pass-through in Cabo Verde

### A. Interest Rate Channel of Monetary Policy
- Purpose: estimates the pass-through of the monetary policy rate to bank lending rates in Cabo Verde over 2001–May 2025 and tests for asymmetries between rate increases and decreases.
- Context:
  - The Banco de Cabo Verde (BCV) embarked on a monetary policy tightening phase since May 2023.
  - The current tightening phase amounted to a cumulative change of 225 bps in the policy rate and is described as the steepest over the past two decades.
  - Emigrant deposits constitute an average of 30 percent of total deposits in the banking system.
- Observations:
  - Average lending and long-term rates have not adjusted noticeably to the recent increases in the BCV policy rate.
  - Emigrant and resident deposit rates have not increased significantly, with a convergence among both rates since 2020.

### B. Empirical Assessment of Interest-Rate Pass-through
- Data and models:
  - Monthly BCV data from January 2001 to May 2025.
  - Two models estimated:
    - Error-correction model (ECM) using the two-step Engle and Granger procedure.
    - Local projections (LP) estimating responses at horizons one to six months ahead.
  - Benchmark controls: reserve requirements ratio (RRR), CPI inflation, and a crisis dummy. Additional robustness controls tested: non-performing loans (NPLs), excess liquidity, and loan-to-deposit ratio (LDR).
- Key quantitative findings:
  - Medium- to long-run policy rate pass-through estimates:
    - Average lending rate: range between 0.19-0.28.
    - Long-term rate: range between 0.13-0.23.
  - Robustness with additional controls (discarding first tightening episode):
    - Average lending rate: range between 0.17-0.3.
    - Long-term rate: range between 0.11-0.15.
  - Average interest rate betas across cycles:
    - Tightening cycles: pass-through falls into a range of -0.75−0.39 percentage points.
    - Loosening cycles: pass-through falls into a range of -0.49−0.66 percentage points.
- Bank structure and market features influencing pass-through:
  - Banking system composition: eight banks, with three dominant banks.
  - Bank shares of total assets / credit / deposits (as of December 2024):
    - BCA: 27.34% / 25.09% / 30.40% — High D-SIB
    - CECV: 26.41% / 29.04% / 30.09% — High D-SIB
    - BIA: 11.14% / 11.81% / 10.96% — Slight systemic importance
  - Bank concentration can weaken transmission by giving banks market power to set interest rates.
  - Possible drivers of weak pass-through: market concentration, risk premium, consumer search costs, excess liquidity, banks’ high funding costs, and the fact that banks do not rely on BCV refinancing.

### C. Asymmetric Pass-through During Loosening and Tightening Cycles
- Asymmetry headline:
  - Pass-through is larger during tightening episodes compared with easing episodes, but remains limited in absolute terms.
- Quantitative asymmetry:
  - For long-term rates (LP model, six-month horizon):
    - Tightening cycles: pass-through of 0.35.
    - Easing cycles: pass-through of 0.17.
  - Thus, pass-through during tightening cycles is roughly twice as large as during easing cycles.
- Heterogeneity across episodes:
  - Tightening episodes:
    - Apr 2001–Apr 2002: evidence of full/complete pass-through in the 2001 cycle.
    - Jan 2012–Feb 2014: limited pass-through of around 0.2 after six months.
    - May 2007–Dec 2009 and May 2023– (current) cycles: lending rates broadly insensitive to monetary policy.
  - Easing episodes:
    - Jan 2010–Dec 2012: stronger pass-through, though it took time to materialize.
    - Mar 2014–Apr 2023: pass-through estimated to be practically zero (banks did not decrease lending rates despite policy rate cuts).
- Time variation:
  - Time-varying pass-through estimates indicate the current tightening cycle that started in May 2023 appears broadly insensitive to monetary policy.

### D. Additional Relevant Quantities and Episodes
- BCV tightening cycles (Table entries reproduced):
  - Jan 2001 - April 2002: Duration (months) 15; Total increase 100bps; Number of hikes 1
  - May 2007 - Dec 2009: Duration (months) 32; Total increase 175bps; Number of hikes 5
  - Feb 2012 - Feb 2014: Duration (months) 26; Total increase 150bps; Number of hikes 1
  - May 2023 - Oct 2025: Duration (months) 30; Total increase 225bps; Number of hikes 6
- Spread behavior:
  - The average spread between the average lending rate and the policy rate is on average 113 basis points higher during policy easing phases compared to tightening phases.

### E. Conclusions (Annex summary)
- The policy rate pass-through in Cabo Verde is estimated to be rather limited:
  - Medium- to long-run pass-through ranges:
    - Average lending rate: 0.19 to 0.28.
    - Long-term rate: 0.13 to 0.23.
- Asymmetry:
  - Pass-through is roughly twice as large during tightening episodes compared to easing episodes, though still relatively low.
  - Larger pass-through during tightening cycles appears driven by the 2001 and 2012 hiking cycles; the 2007 and the current May 2023 cycles are broadly insensitive to monetary policy.

*Source: Cabo Verdean authorities and IMF staff estimates; annex prepared by Bruno Albuquerque and Nombulelo Gumata.*

### References

### References

### Major empirical and methodological sources
- Chan, K. S. (1993), "Consistency and Limiting Distribution of the Least Squares Estimator of a Threshold Autoregressive Model," The Annals of Statistics.
- De Mello, João Manoel Pinho and Pedro Henrique Rosado de Castro. 2012. “Is the Bank Interest Rate Pass-Through of SELIC Rate Movements Asymmetric?” Brazilian Review of Econometrics 32(1): 3–30.
- Elias, Laura and Bernardo Guimaraes. 2024. “This is not America: The Credit Channel of Monetary Policy in Brazil.” Economía 23(1): 137–174.
- Francesco Grigoli and José M. Mota. 2015. Interest Rate Pass-Through in the Dominican Republic. IMF Working Paper WP/15/260.
- Jailson da Conceicao Teixeira de Oliveira, Bruno Ferreira Frascaroli and Osvaldo Candido da Silva Filho. 2015. Monetary policy in Cape Verde and macroeconomic changes: empirical evidence. African Journal of Business Management. Vol.9(3), pp.76-95.
- Leigh, Daniel and Rui Xu. 2025. Monetary Policy Transmission to Lending Rates: Evidence from Brazil.” IMF Working Paper WP/25/152.
- Niggle, C.J (1987) A Comment on the Markup Theory of Bank Loan Rates, Journal of Post Keynesian Economics ,9(4), 629-631
- Robert C M Beyer, Ruo Chen, Florian Misch, Claire Li, Ezgi O Ozturk, and Lev Ratnovski. 2024. Monetary Policy Pass-Through to Interest Rates: Stylized Facts from 30 European Countries. IMF Working Paper WP/24/9.
- Rocha, Carlos Manuel da Luz Delgado. 2007. Macroeconomic stabilization and monetary policy in Cape Verde. Doctoral Thesis. Technical University of Lisbon. Higher Institute of Economics and Management.
- Valadkhani, A and Anwar, S. 2012. Interest rate pass-through and the asymmetric relationship between cash rate and the mortgage rate, Economic record, 88(282), 341-350
- Walter Enders, Pierre L. Siklos. 2001. Cointegration and Threshold Adjustment. Journal of Business & Economic Statistics, Vol. 19, No. 2 (Apr. 2001), pp. 166-176.
- Yildrim, D 2012. Interest rate pass-through to Turkish lending rates: A threshold cointegration analysis, EC working chapters in economics 12/7.

### Central findings, projections, and quantitative results (from Cabo Verde VAT reform analysis)
- Fiscal and macro context:
  - Government revenue expanded to about 20 percent of GDP.
  - Indirect tax exemptions total 4 percent of GDP and are regressive.
  - Top consumption quintile receives CVE 1.55 billion (about 0.8 percent of GDP) in indirect transfers through VAT exemptions, over five times more than the bottom quintile.
  - Financing needs for critical physical and human capital investments are estimated at around 1.5 percent of GDP per year.
  - Debt stock decreased to 110 percent of GDP in 2024 (from 148.5 percent in 2021).

- Model and scenario setup:
  - Model used: Debt Investment and Growth (DIG-Labor) model calibrated to Cabo Verde.
  - Baseline realized fiscal space from VAT streamlining: 1.2 percent of 2024 GDP; VAT rate remains 15 percent.
  - Upside scenario: removal of all indirect tax exemptions generates fiscal space equal to 4 percent of 2024 GDP; VAT rate remains 15 percent.
  - Downside scenario: administrative and political frictions reduce realized fiscal space to 0.6 percent of 2024 GDP; VAT rate remains 15 percent.
  - Public investment scale-up in upside and baseline: infrastructure +1.5 percentage points of GDP and education +0.5 percentage points of GDP relative to the 2024 baseline.
  - Grants assumption: Upside grant inflows 1.0 percent of GDP; Baseline and Downside grant inflows 0.8 percent of GDP.
  - Targeting assumptions for compensatory cash transfers: informal households comprise 60 percent of the population (휙 = 0.6).

- Key macro outcomes across scenarios (relative to the no-reform path unless stated):
  - Short-run aggregate consumption (years 1 to 5):
    - Upside: Aggregate consumption increases by 1.1 percent relative to the no‑reform path.
    - Baseline: Aggregate consumption remains below the Upside path and recovers at a moderate pace.
    - Downside: Aggregate consumption remains subdued for longer under constrained transfers.
  - Private investment during transition:
    - Upside: Private investment declines temporarily and then increases; at the steady state it stands 10.6 percent above the no‑reform path.
    - Baseline: Similar pattern but with a smaller and slower recovery.
    - Downside: Private investment remains weak for a longer period; crowding‑in occurs later.
  - Long‑run aggregate consumption (steady state):
    - Upside: Aggregate consumption stands 12.6 percent above the no‑reform path.
    - Baseline: Aggregate consumption stands 1 percentage point below the Upside scenario.
    - Downside: Aggregate consumption remains below the Baseline scenario; convergence proceeds more slowly.
  - Transfers feasible at the steady state (targeted):
    - Upside: Targeted transfers equal 4.2 percent of GDP, which corresponds to 7 percent of initial per‑capita income.
    - Baseline: Targeted transfers equal 1.8 percent of GDP, which corresponds to 3 percent of initial per‑capita income.
    - Downside: Targeted transfers peak at 1.6 percent of per‑capita income.
  - Labor-market dynamics:
    - Upside: Formalization accelerates and wages increase earlier.
    - Baseline: Formal hiring proceeds more slowly despite the same public investment path.
    - Downside: Labor reallocation remains sluggish and unemployment improves later.
  - Fiscal stance and VAT:
    - Debt stabilizes while the VAT rate remains at 15 percent in the Upside scenario; authorities rely on concessional financing in the medium term.
    - Baseline: Investment scale-up remains feasible while the VAT rate remains at 15 percent.
    - Downside: VAT rate remains at 15 percent; tighter fiscal space requires sharper prioritization.
  - Distributional outcomes at the steady state:
    - Gini coefficient declines to 0.378 across scenarios.
    - Poverty headcount:
      - Upside: declines to 0.5 percent of the population.
      - Baseline: declines to 1.9 percent of the population.
      - Downside: declines to 2.3 percent of the population.

- Quantified cash transfer schedules (selected exact figures reported)
  - Baseline scenario: Targeted (Informal Sector) monthly cash transfers (US$/Month) and share of initial per-capita income (%):
    - Year 1: 50.77; 0.2 percent (Universal: 0.46; 0.1 percent)
    - Year 6: 102.69; 0.7 percent (Universal: 1.61; 0.4 percent)
    - Year 11: 154.27; 1.2 percent (Universal: 2.55; 0.7 percent)
    - Year 16: 205.91; 1.6 percent (Universal: 3.53; 0.9 percent)
    - Year 21: 257.13; 1.9 percent (Universal: 4.26; 1.1 percent)
    - Steady state: 11.04; 3.0 percent (Universal: 6.60; 1.8 percent)
  - Upside scenario: Targeted (Informal Sector) monthly cash transfers (US$/Month) and share of initial per-capita income (%):
    - Year 1 - 5: 14.39; 3.9 percent (Universal: 8.60; 2.3 percent)
    - Year 6 - 10: 16.37; 4.4 percent (Universal: 9.78; 2.6 percent)
    - Year 11 - 15: 18.10; 4.9 percent (Universal: 10.82; 2.9 percent)
    - Year 16 - 20: 19.93; 5.3 percent (Universal: 11.91; 3.2 percent)
    - Year 21 - 25: 21.33; 5.7 percent (Universal: 12.75; 3.4 percent)
    - Steady state: 26.17; 7.0 percent (Universal: 15.64; 4.2 percent)

### Policy implications and recommendations drawn from simulations
- Streamline VAT exemptions to expand the tax base while keeping the VAT rate at 15 percent to generate fiscal space for public investment without immediate rate changes.
- Prioritize a targeted, phased rollout of compensatory cash transfers to informal households to mitigate short‑term welfare losses:
  - Upside scenario: back‑load transfers, averaging 2.3 percent of GDP during years 1 to 5 and converging to 4.2 percent of GDP at steady state.
  - Baseline scenario: defer broad transfers until growth dividends materialize; transfers converge to 1.8 percent of GDP at steady state.
- Finance medium‑term public investment through a mix of base broadening, grants, and concessional borrowing to avoid sharp near‑term fiscal strain.
- Strengthen tax administration and communications to reduce administrative and political frictions that could limit realized fiscal space (downside risk: realized fiscal space reduced to 0.6 percent of GDP).
- Maintain attention to public investment absorptive capacity and operations and maintenance funding to realize productivity gains that support long‑run distributional improvements.
- Use the projected productivity and formalization gains from public investment to create endogenous fiscal space for sustainable, targeted transfers that support poverty reduction and inequality decline.

*Source: Staff simulations and references contained in the provided PDF content.*

### 15. While Table 2 illustrates the scale of cash transfers under targeted and universal

### 15. While Table 2 illustrates the scale of cash transfers under targeted and universal coverage schemes, it serves purely as an analytical exposition

### E. Downside Scenario: Underperformance of VAT Base-expansion
- Context: Emerging market economies like Cabo Verde encounter difficulties in increasing revenue beyond initial gains.
- Technical estimate cited: eliminating all VAT exemptions could potentially expand fiscal space by about 1-1.5 percent of GDP, but implementation challenges may hinder achieving this potential.
- Downside scenario assumption: revenue yield from the VAT reform is around 0.6 percent of GDP, due to administrative bottlenecks and political economy frictions.
- Policy choice under downside:
  - Government maintains commitment to public investment despite underwhelming VAT revenue gains.
  - Adjustment burden falls on social transfers rather than investment, reflecting a prioritization to protect long-term growth drivers (infrastructure and education).
- Macroeconomic implications:
  - Public investment remains on track, supporting modest improvements in productivity and formal sector output over time.
  - Limited fiscal space forces a sharp contraction in transfers, especially in the early years, resulting in subdued aggregate consumption and private investment.
  - Targeted transfers to informal sector workers are delayed and peak, across steady state, at just 1.6 percent of per capita income, about half the baseline level.
  - Retracted early transfers undermine redistributive potential, delay recovery for vulnerable groups, weaken demand-side stimulus, slow formal sector expansion, and depress labor market absorption.

### F. Scenario Comparisons – Macroeconomic Outcomes
- Role of compensatory transfers:
  - Compensatory transfers mitigate public education investment gestation lags and accelerate formal sector labor absorption.
  - Upside scenario: significant compensatory transfers financed through savings from removal of all indirect tax exemptions boost household disposable income during early years of education investment gestation.
  - Result: sustained demand for formal sector goods enables firms to expand employment via efficiency wage mechanisms simultaneously with public capital accumulation.
- Contrasts across scenarios:
  - Minimal transfers in baseline and downside scenarios delay demand growth for formal sector goods, lengthening the period before labor markets absorb more productive workers.
  - Identical public infrastructure and education investments yield differing labor market outcomes due to demand-mediated complementarities.
- Dynamics of private investment and labor utilization:
  - Private capital accumulation follows similar paths in baseline and upside scenarios, but early formal sector demand in the upside scenario drives faster labor reallocation toward the formal sector.
  - In baseline and downside, subdued formal sector demand causes firms to defer hiring despite a rising pool of productive low-skill labor.
- Fiscal-growth feedback and path dependence:
  - Upside scenario: earlier formal sector growth initiates a self-reinforcing fiscal-growth cycle—earlier output expansion boosts tax revenue, funding further transfers that accelerate demand and labor absorption.
  - Baseline scenario: sluggish formal sector demand delays revenue mobilization, perpetuating low cash transfers and prolonging suboptimal labor allocation.

### G. Scenario Comparisons – Poverty & Distributional outcomes
- Methodology note: simulations use the 2015 household income and expenditure survey as benchmark for initial poverty headcount, Gini coefficient, and consumption share by decile; forward-looking income distribution inferred via growth-incidence curves.
- Long-run (steady state) distributional outcomes:
  - Gini coefficient declines from 0.423 in 2015 to about 0.378 across all scenarios.
  - Baseline scenario: bottom decile experiences an elevenfold increase in income across steady states; top decile rises 3.3-fold.
  - Poverty headcount rate changes (reported in two passages):
    - From 14.6 percent in 2015 to 1.9 percent under the baseline scenario, 0.5 percent in the upside scenario, and 2.3 percent in the downside scenario.
    - Later summary: long-run distributional gains with Gini declining to 0.378 and poverty head counts falling to 0.5 percent under the upside scenario, 1.9 percent under the baseline scenario, and 2.3 percent under the downside scenario.
- Medium-term (around ten years) dynamics:
  - Upside and baseline scenarios: faster income growth among lower deciles and moderated growth among upper deciles (except the very top).
  - Structural and distributional adjustments are driven by integrating agriculture and informal sector participants into the tax net and reinvesting proceeds in infrastructure and primary education.
  - Gini coefficient declines from .423 to about .388 in both the upside and baseline scenarios over the medium term; poverty headcount drops to 2.2 percent and 3 percent, respectively.
- Downside scenario distributional constraints:
  - Weaker fiscal capacity constrains gains: income growth across deciles is statistically indistinguishable from zero, the Gini coefficient remains around 41, and the poverty headcount stays elevated at roughly 9.9 percent.
  - Without sustained revenue mobilization, public investment risks delivering limited benefits to the poor in the medium term.

### H. Conclusion
- Model and policy insights:
  - Structural macroeconomic model assesses trade-offs in phasing out VAT exemptions and reallocating fiscal resources toward public investment.
  - The model assists policymakers in creating fiscal space for essential investments, supporting vulnerable populations through targeted cash transfers in the short term, and evaluating how investment can promote GDP growth, jobs, and reduce income inequality over medium and long term.
- Illustrative simulation results (numeric outcomes preserved):
  - Upside scenario: base broadening of 4 percent of 2024 GDP lifts aggregate consumption by 1.1 percent in the first five years and by 12.6 percent in the long run; raises private investment by 10.6 percent at the steady state.
  - Baseline scenario: same investment path yields a weaker near-term demand response and a long-run consumption level that is 1 percentage point below the Upside path.
  - Downside scenario: slower revenue gains delay crowding-in and lower macro gains, although the authorities maintain the investment envelope.
- Fiscal and financing anchors:
  - Keeping the VAT rate at 15 percent, anchoring grants within the scenario envelope, and relying on concessional financing could help sustain debt stabilization while the growth dividend accrues.
- Role and sizing of targeted cash transfers:
  - Targeted cash transfers are essential to protect real incomes during adjustment periods and support demand in line with realized fiscal resources.
  - Upside scenario: transfers to informal households could be back‑loaded and increase gradually, averaging 2.3 percent of GDP in the first 5 years and reaching 4.2 percent of GDP at steady state.
  - Baseline scenario: may require delaying broad transfers until revenues improve, ultimately converging to 1.8 percent of GDP at steady state.
  - Downside scenario: maintaining investment envelope while providing a lean safety net with transfers peaking at 1.6 percent of per‑capita income is critical.
  - Strengthening compliance and the social registry will help reduce leakage, promote formalization, and enable scaling-up benefits as revenues increase.
- Aggregate long-run projections reiterated:
  - All scenarios project long‑run distributional gains, with the Gini coefficient declining to 0.378 and poverty head counts falling to 0.5 percent under the upside scenario, 1.9 percent under the baseline scenario, and 2.3 percent under the downside scenario.

*International Monetary Fund — Cabo Verde (excerpts from the supplied content unit).*

### Annex II. Macroeconomic and Distributional Effects of Tax Policy

### Annex II. Macroeconomic and Distributional Effects of Tax Policy

### Scenario outcomes: Poverty, mean income, and distribution (Figure 7)
- Initial Headcount (HC): 14.3 percent
- Upside scenario (Medium-term):
  - Final HC: 2.2 percent
  - Mean Income: $4475 → $5106
- Baseline scenario (Medium-term):
  - Final HC: 3.0 percent
  - Mean Income: $4475 → $4610
- Downside scenario (Medium-term):
  - Final HC: 9.9 percent
  - Mean Income: $4475 → $4580

### Distributional metrics (Figure 5 and Figure 6)
- Gini: plotted range shown between 0.30 and 0.55 (Bootstrapped Gini, Initial Gini, Simulated Median displayed for Upside, Baseline, Downside)
- Income growth by decile: presented graphically for Upside, Baseline, Downside (no additional numerical decile values provided in text)

### Macroeconomic and sectoral indicators across scenarios (Figures 1–4)
- Charts provided for each scenario (Baseline: Limited Base Expansion (VAT); Upside: With Removal of all Indirect Tax Exemptions; Downside: VAT Base Expansion Underperforms) include time series or scenario comparisons for:
  - Public Investment (% of GDP) with subcomponents: Infrastructure Investment, Education Investment
  - Real GDP
  - Unemployment Rate
  - Sectoral Output: Agric, Formal, Non-Agric Inf
  - Real Wages: Agric, Formal, Non-Agric Inf
  - Sectoral Labor: Agric, Formal, Non-Agric Inf
  - Private Capital
  - Aggregate Consumption
  - Cash Transfers
  - VAT rate (axes showing values in ranges such as 13 to 17 and 14 to 32 depending on chart)
  - Public external and Domestic Debt components: External commercial debt to GDP, Domestic Debt to GDP
  - Total Public Debt to GDP (axes showing values in the range 95 to 120)
- Figures are based on "Staff computations"

### Background macro charts (Figure 4)
- Narrative captions shown:
  - "Rapid growth in the past has moderated... ...in line with frequent shocks and slowdown in public investment."
  - "Public debt is decreasing but vulnerabilities remain... ... requiring continued commitment to fiscal adjustment."
  - "Adjustment is supported by strong tax revenue mobilization..."
  - "... and can be complemented by efficiency gains."

### Calibration of model variables and parameters (Table 1)
- Trend Per capita GDP growth: 0.04 (Median of the trend component 1981-2024, World Bank)
- Real interest rate: 0.03 (Based on T-Bond rates: CVOTELOSI009-OTE3,0625%OT28JUL2031, CVOTELOSH001-OT3%12JUL2029)
- Real interest rate on external commercial public debt: 0.05 (Assuming EM Sovereign bond yields are around 8-10% and WEO global inflation 4.2%)
- Real interest rate on external private debt: 0.03 (Based on a baseline a blue bond issued in Nov 2023)
- Cost share of private capital in sector "n=formal sector": 0.15 (Estimates based on core parameters)
- Cost share of private capital in sector "j=informal sector": 0.35 (Estimates based on core parameters)
- Cost share of public capital in sector "n": 0.15 (Estimates based on core parameters)
- Cost share of public capital in sector "j": 0.35 (Estimates based on core parameters)
- Cost share of skilled labor in sector "n": 0.15 (Estimates based on core parameters)
- Cost share of skilled labor in sector "j": 0.35 (Estimates based on core parameters)
- Output elasticity to private capital in sector "x=agriculture sector": 0.60 (Estimated using data for CPV from USDA)
- Output elasticity to private capital in sector "n": 0.40 (Estimates based on literature)
- Output elasticity to private capital in sector "j": 0.20 (Estimates based on literature)
- Output elasticity to skilled labor in sector "x": 0.36 (Estimated using data for CPV from USDA)
- Output elasticity to skilled labor in sector "n": 0.26 (Estimates based on literature)
- Output elasticity to skilled labor in sector "j": 0.20 (Estimates based on literature)
- Output elasticity to land in sector "x": 0.30 (Estimated using data for CPV from USDA)
- Share of consumption "n": 0.20 (Estimated from HH survey on consumption expenditure)
- Share of consumption "x": 0.30 (Estimated from HH survey on consumption expenditure)
- Share of consumption "m": 0.40 (Estimated from HH survey on consumption expenditure)
- Initial unemployment: 12 (Average 2020-2023, World Bank data)
- Broadening VAT parameter for good "x": 0.55 (Assumed C-Efficiency levels based on WP CPV PFR 2025)
- Broadening VAT parameter for good "j": 0.55 (Assumed C-Efficiency levels based on WP CPV PFR 2025)
- Broadening VAT parameter for good "m": 1 (Assumed C-Efficiency levels reflect import exemptions)
- Parameter of wage curve: 0.1 (From literature)
- Effort elasticity of unemployment: 0.5 (From literature)
- Gross (initial) return on public capital: 0.27 (From literature)
- Gross (initial) return on secondary education: 0.35 (From literature)
- Gross (initial) return on primary education: 0.35 (From literature)
- Fraction recovered of recurrent costs: 0.50 (Assumed levels)
- Marginal efficiency of public investment: 1.0 (Assumed levels)
- Historical efficiency of public investment: 1.0 (Assumed levels)
- Degree of absorptive capacity constraints: 0.0 (Assumed levels)
- Public investment to GDP ratio: 0.04 (IMF Capital Stock Dataset)
- Public investment in primary education/GDP: 0.04 (World Bank)
- Public investment in upper-level education/GDP: 0.01 (Assumed level)
- Share of external private debt: 0.11 (CPV 2024 DSA Data)
- Share of domestic debt: 0.32 (CPV 2024 DSA Data)
- Share of concessional debt: 0.61 (CPV 2024 DSA Data)
- Share of external commercial public debt: 0.18 (CPV 2024 DSA Data)
- Share of remmittances: 0.12
- Share of grants (% GDP): 0.01
- Initial VAT rate: 0.15 (CPV Revenue Administration)
- Tax rate on labor supplied to sector "n = formal sector": 0.03 (Estimated effective tax rate)
- Labor tax rate on labor supplied to sector "x=agriculture": 0.00
- Labor tax rate on labor supplied to sector "j=informal sector": 0.00
- Capital tax rate in sector "n": 0.21 (CPV Revenue Administration)
- Capital tax rate in sector "x": 0.21 (CPV Revenue Administration)
- Capital tax rate in sector "j": 0.21 (CPV Revenue Administration)

*Source: Staff computations and Table 1 calibration (Annex II. Macroeconomic and Distributional Effects of Tax Policy).*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2026/english/1cpvea2026002-source-pdf.pdf_