## 1dmaea2022002

## Source details

**Canonical URL:** [1dmaea2022002](https://www.imf.org/-/media/files/publications/cr/2022/english/1dmaea2022002.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2022/english/1dmaea2022002.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2022/english/1dmaea2022002.pdf.json)

---

### Overview and objective
- Paper develops a multi-sector general equilibrium model tailored to small developing states (SDS), calibrated to Dominica, to study macroeconomic and distributional implications of financing resilience-building (resilient public capital) using alternative fiscal instruments.
- Objective: quantify efficiency and distributional tradeoffs of alternative financing instruments for resilient public capital under SDS-specific features: formal and informal sectors, labor migration and remittances, high-skilled outward migration propensity, low output diversification, five household/agent types, granular fiscal instruments, and distinction between resilient and non-resilient public capital.

### Major model features and structure
- Agents / households:
  - Unskilled households — work for farmers or in informal service sector.
  - Skilled households — work in formal sector (manufacturing) or migrate.
  - Government workers — work for government or migrate.
  - Farmers — hire labor and invest in agriculture.
  - Entrepreneurs — hire labor in formal manufacturing and services and invest.
- Goods and production:
  - Three goods each period: agriculture, manufacturing, services.
  - Agriculture and manufacturing: Cobb–Douglas with public capital (k_p), private capital (k), and labor; productivity parameters z_a and z_m; public capital shares α_p and private capital shares α_*/α_m preserved.
  - Services: produced in formal or informal sectors using labor only with decreasing returns; y_s = z_s θ (h)^{1-η}.
- Capital dynamics and disasters:
  - Private capital depreciation δ; k_{t+1} = (1-δ) k_t + x.
  - Public capital splits into resilient k_R and non-resilient k_N. Natural disasters (NDs) destroy share δ_D of non-resilient capital with probability p_D each period.
  - Laws of motion preserved: k_{R,t+1} = (1-δ_p) k_{R,t} + I_R; k_{N,t+1} = [(1-δ_p) - p_D δ_D] k_{N,t} + I_N.
  - Resilient investment relative price p_{R,N} ≥ 1; I_N = λ I_R.
- Labor choice and migration:
  - Skilled and government workers choose domestic labor supply h and out-migration 1- h; out-migration earns foreign wage w_F and generates remittances (1- h)^{η} w_F.
  - Unskilled households choose labor split between agriculture h_a and informal services 1- h_a; informal sector not taxed.
- Government fiscal instruments and budget constraint (preserved exactly):
  - Tax instruments: τ_a, τ_m, τ_s (value-added taxes), τ_c (corporate), τ_l (personal income).
  - Nontax revenue NR (CBI, donor grants) and foreign grants GR.
  - Expenditures: public sector wages w_g, public investment I_R and I_N, transfers T_g, T_s, T_u, transfers to farmers and entrepreneurs, and debt service D.
  - Government dynamic budget constraint: 0 = R^{ct} + R^{corp} + R^{lab} + NR_t + GR_t - p_{R,N} I_{R,t} - I_{N,t} - I_t - T_t - CE_t - (1+r^*) D_t + D_{t+1}.
- Market clearing and equilibrium: decision rules for consumption, labor, investment and capital for each agent type; skilled and unskilled labor and services market clearing as in source.

### Calibration and data inputs
- Calibrated to Dominica (date on contents: January 18, 2022).
- Consumption shares (CPI basket): 휓 Share of tradables in total consumption 35.9 (model = 35.9), 훾 Share of services in total consumption 42.2 (model = 42.2).
- Labor force allocation (Dominica Social Security Administration):
  - 휇௚ government workers 0.349
  - 휇௦ skilled workers 0.444
  - 휇௨ unskilled workers 0.105
  - 휇௘ entrepreneurs 0.049
  - 휇௙ big farmers 0.053
  - Emigration elasticity 훼 = 0.5; US wages 푤௨௦ calibrated so remittances match data (푤௨௦ Remittances 10.0 percent of GDP).
- Sector shares (model = data):
  - Agricultural output 푧௔ 16.8 percent of GDP
  - Services output 푧௘,௦ 50.9 percent of GDP
  - Informal sector output 푧௦ 3.7 percent of GDP
- Capital elasticities and depreciation:
  - 훼௠ tradable elasticity to private capital 0.24
  - 훼∗ agriculture elasticity to private capital 0.40
  - 훼௚ elasticity to public capital 0.14
  - 훿 private capital depreciation 0.14
  - 훿௚ public capital depreciation 0.109
- Fiscal parameters (percent of GDP):
  - Consumption tax 13.5, personal income tax 4.5, corporate taxes 4.0, Grants (Gr) 5.0, Non-tax revenues (NR) 10.0, Public sector wage bill 푤௚ 11, Public Capital investment over GDP in steady state 퐾଴௚ 12, Transfers to households 5.5.
  - Interest rate on government debt 푟∗ 0.029
  - Discount factor Β 0.96
  - Lower bound of agricultural consumption c̅ 0.20

### Natural disaster (ND) shock design
- ND shock assumptions:
  - Shock reduces stock of non-resilient public capital by 40 percent.
  - Temporary efficiency loss in capital and labor productivity θ and additional recovery loss in depreciation rate 훿ௗ.
  - Formal and informal services productivity impacted transitorily for one period (one year).
  - Productivity decline calibrated to obtain a 5 percent drop in GDP.

### Policy experiment design
- Baseline: economy starts from a non-resilient state; at time zero government begins investing in resilient infrastructure financed by one fiscal instrument.
- ND shock occurs in period ten; post-shock economy converges to a new steady state with higher resilient public capital.
- Debt financing not considered.
- Five financing instruments studied (each set to raise 1 percent of GDP in steady state):
  - Consumption tax (VAT)
  - Corporate tax (CIT)
  - Payroll tax (PIT)
  - Reduction in public sector wage
  - Increase in donor financing (grants)
- Fiscal revenue formulas preserved:
  - Consumption tax revenue: R^{ct}_t = τ_a p_a C_{t,a} + τ_m C_{t,m} + τ_s p_s C_{t,s}.
  - Corporate tax revenue: R^{corp}_t = τ_{c,t} (μ_a π_{t,a} + μ_e π_{t}^* + μ_e π_{t,s}).
  - Labor tax revenue: R^{lab}_t = τ_{l,t} (μ_s w h_{t,s} + μ_g w h_{t,g}).
- Table of policy changes required to raise 1 percent of GDP (Table 4, Benchmark Values / Policy Change (%)):
  - VAT 0.12 -> 5.2%
  - PIT 0.06 -> 15.0%
  - CIT 0.10 -> 17.2%
  - Public Sector Wages 0.27 -> 2.1%
  - Transfers 0.02 -> -9.0%

### Key quantitative outcomes and benchmark results
- Benchmark (no resilient investment):
  - Output declines by 16 percent with non-resilient public capital after the ND shock.
  - Capital-intensive sectors decline more: manufacturing most, then agriculture; formal and informal services decline less.
  - ND stimulates out-migration as skilled workers experience wage loss; remittances provide partial insurance.
- Building resilience:
  - Resilient investment reduces the impact of the ND shock on output by 2 to 2.5 percentage points depending on financing instrument.
- Efficiency ranking (impact on output and distortions):
  - VAT is most efficient — relatively less distortionary.
  - CIT is costliest — reduces investment and private capital; lowers labor productivity and wages in formal sectors; amplifies out-migration of skilled workers and output losses.
  - PIT has intermediate negative impact — increases out-migration, effect smaller than CIT.
  - Reduction in government wage expenditure is less distortionary under model assumptions (government workers assumed to have zero marginal contribution to output).
  - Increase in donor finance is most beneficial — causes no tax distortions.
- Policy tradeoffs (Table 5: rankings preserved):
  - OUTPUT ranking: VALUE-ADDED = 2, CORPORATE = 3, GOVERNMENT = 1, CONSUMPTION PUBLIC INVESTMENT = 4
  - INEQUALITY (GINI) ranking: VALUE-ADDED = 3, CORPORATE = 2, GOVERNMENT = 4, CONSUMPTION PUBLIC INVESTMENT = 1

### Distributional consequences and household welfare
- General ND effects on inequality:
  - ND shocks increase income inequality by disproportionately affecting lower-wage sectors.
  - Unskilled workers suffer more because they cannot migrate; skilled workers smooth via migration and remittances.
  - Government workers’ wage assumed unchanged.
- Fiscal instrument effects on inequality:
  - CIT increase reduces income inequality (levied on higher-income: skilled and entrepreneurs).
  - PIT increase reduces income inequality (affects higher-income government employees and skilled workers; informal workers not taxed).
  - VAT increase results in a small decline in income inequality due to model-specific mechanisms (informal sector labor supply responses; informal workers not paying labor taxes).
  - Reduction in government workers' wage expense reduces income inequality (government wages higher than most sectors).
  - Cut in government transfers increases income inequality (transfers are larger share of income for low-income households).
- Welfare by household type (consumption responses):
  - Unskilled workers:
    - Consumption declines under ND in benchmark; resilient investment reduces the concurrent decline.
    - Relative impacts of financing instruments:
      - CIT financing affects unskilled consumption least.
      - PIT financing has relatively small effect on unskilled wages.
      - VAT financing reduces unskilled consumption more significantly.
      - Decline in transfers has the highest negative impact on unskilled consumption.
  - Skilled workers:
    - Migration allows smoothing; model assumption implies no concurrent welfare loss at ND.
    - CIT reduces skilled consumption (lowers formal sector wages).
    - PIT increase is most detrimental to skilled consumption; out-migration mitigates but net loss occurs.
    - Cut in government transfers reduces skilled consumption less than for unskilled.

### Policy implications and recommendations (model-driven)
- Financing resilience with donors' support or by cutting unproductive government spending yields higher returns to building resilience than financing via distortionary taxes.
- Distortionary taxes (especially CIT) impose efficiency costs through lower investment, reduced private capital, lower productivity, and amplified skilled out-migration in SDSs.
- VAT is relatively less distortionary and more efficient for financing resilient investment, but it is less progressive than CIT or PIT.
- Improving targeting and efficiency of government transfers can create fiscal space for resilience investment while addressing income inequality.
- Grant financing is effective at containing income loss for low-income households after NDs and minimizes output/employment losses associated with tax distortions.

### Vulnerability to NDs and the role of resilient infrastructure
- Dominica is highly vulnerable to NDs with large human, economic, and social costs; resilient infrastructure lowers economic and social impact, speeds recovery, and provides continuity in public services.
- Public investment in resilience has high returns in output and employment but can increase inequality because formal sectors and skilled workers benefit most in the long run.
- Reducing unproductive government employment is an efficient financing option under model assumptions.

### Shadow economy measurement (MIMIC model findings)
- MIMIC methodology relates a latent shadow economy to drivers (tax rates, agriculture size, tourism share, Regulatory Quality Index) and indicators (real GDP growth, employment rate).
- Empirical relationships found:
  - Higher marginal corporate and personal income tax rates associated with larger shadow economy.
  - Larger tourism sector correlated with a smaller shadow economy.
  - Higher regulatory quality associated with a smaller shadow economy.
  - Agriculture positively related but not statistically significant for the sample.
  - Real GDP growth and employment rate negatively related to shadow economy.
- Key quantitative estimate:
  - The shadow economy of Dominica averaged 46 percent of GDP over the 2011–19 period (absolute size depends on the chosen benchmark).

*Source: 1dmaea2022002 - References (IMF PDF chapter/section), January 18, 2022.*

### References _____________________________________________________________________________ 19

### References (1dmaea2022002)

### Overview and objective
- Paper develops a multi-sector general equilibrium model tailored to small developing states (SDS), calibrated to Dominica, to study macroeconomic and distributional implications of financing resilience-building (resilient public capital) using alternative fiscal instruments.
- Key model features: formal and informal sectors, labor migration and remittances, high-skilled outward migration propensity, low output diversification, five household/agent types, granular fiscal instruments, and distinction between resilient and non-resilient public capital.

### Major model features and structure
- Agents / households:
  - Unskilled households — work for farmers or in informal service sector.
  - Skilled households — work in formal sector (manufacturing) or migrate.
  - Government workers — work for government or migrate.
  - Farmers — hire labor and invest in agriculture.
  - Entrepreneurs — hire labor in formal manufacturing and services and invest.
- Goods and production:
  - Three goods each period: agriculture, manufacturing, services.
  - Agriculture: Cobb–Douglas with inputs public capital (k_p), private capital (k), and labor h*; productivity z_a; public capital share α_p and private capital share α_*.
  - Manufacturing: Cobb–Douglas with public capital (k_p), private capital (k), and labor h_m; productivity z_m; public capital share α_p and private capital share α_m.
  - Services: produced in formal or informal sectors using labor only with decreasing returns; y_s = z_s θ (h)^{1-η} (notation preserved as in source).
- Capital dynamics:
  - Private capital depreciates at rate δ; k_{t+1} = (1-δ) k_t + x.
  - Public capital splits into resilient k_R and non-resilient k_N. Natural disasters (NDs) destroy share δ_D of non-resilient capital with probability p_D each period.
  - Law of motion: k_{R,t+1} = (1-δ_p) k_{R,t} + I_R; k_{N,t+1} = [(1-δ_p) - p_D δ_D] k_{N,t} + I_N.
  - Resilient investment is more costly: p_{R,N} ≥ 1 represents relative unit price of resilient to non-resilient capital; λ is relative share of investment in resilient to non-resilient capital with I_N = λ I_R.
- Disasters and productivity:
  - ND shocks reduce total factor productivity transitorily and destroy non-resilient capital stock; services sector includes parameter θ to capture ND disruption.
- Labor choice and migration:
  - Skilled and government workers choose domestic labor supply h and out-migration 1- h; out-migration earns foreign wage w_F and generates remittances (1- h)^{η} w_F (notation preserved from source).
  - Unskilled households choose labor split between agriculture h_a and informal services 1- h_a; informal sector does not pay taxes.
- Government fiscal instruments and expenditures:
  - Tax instruments: value-added taxes τ_a (agriculture), τ_m (manufacturing), τ_s (services), corporate tax τ_c, personal income tax τ_l.
  - Nontax revenue NR (CBI, donor grants) and foreign grants GR.
  - Expenditures: public sector wages w_g, public investment resilient I_R and non-resilient I_N, transfers T_g (government workers), T_s (skilled), T_u (unskilled), transfers to farmers and entrepreneurs, and debt service D.
  - Government borrowing only agent: borrows externally at risk-free rate r^*; dynamic budget constraint preserved exactly as in source:
    0 = R^{ct} + R^{corp} + R^{lab} + NR_t + GR_t - p_{R,N} I_{R,t} - I_{N,t} - I_t - T_t - CE_t - (1+r^*) D_t + D_{t+1}.
- Market clearing and equilibrium conditions:
  - Decision rules for consumption, labor, investment and capital for each agent type; market clearing for skilled and unskilled labor, services good market clearing condition as specified in source; government budget constraint (6) satisfied.
- Table 1: Model Structure (Good — Producer — Input — Use — Commerce) preserved in structure (Agriculture: Farmers — Unskilled Labor and Capital — Consumption — Tradable; Manufacturing: Entrepreneurs — Skilled Labor and Capital — Consumption and Investment — Tradable; Services: Unskilled Labor and Entrepreneur Labor — Consumption — Non-Tradable; Public Goods: Government — Public Workers and Manufacturing Goods — Infrastructure, Education, and Social Transfers — Non-Tradable).

### Calibration, experiments, and policy scenarios
- Calibration:
  - Model calibrated to Dominica (date on contents: January 18, 2022).
  - Captures SDS-specific features: large informal sector, high skilled worker propensity to migrate (source notes United Nations data with "about 80 percent" of Caribbean nationals living in the diaspora being relatively "high skilled"), dependence on imported manufacturing inputs for investment.
- Policy experiments (counterfactuals):
  - Finance relatively more expensive resilient public capital using alternative instruments:
    - Value-added taxes (VAT) on goods (τ_a, τ_m, τ_s).
    - Labor income tax τ_l.
    - Corporate income tax τ_c.
    - Donors' support / foreign grants (NR, GR, CBI).
    - Cutting unproductive government spending (reducing transfers or other expenditures).
  - Fiscal revenue definitions preserved:
    - Consumption tax revenue: R^{ct}_t = τ_a p_a C_{t,a} + τ_m C_{t,m} + τ_s p_s C_{t,s}.
    - Corporate tax revenue: R^{corp}_t = τ_{c,t} (μ_a π_{t,a} + μ_e π_{t}^* + μ_e π_{t,s}).
    - Labor tax revenue: R^{lab}_t = τ_{l,t} (μ_s w h_{t,s} + μ_g w h_{t,g}).
  - Public investment choices allow comparing resilient (not destroyed by NDs) vs non-resilient capital accumulation and costs (p_{R,N} > = 1).

### Key findings and distributional implications (as reported)
- Efficiency and returns:
  - Investing in resilient capital is efficient despite its higher economic cost.
  - Returns depend critically on the financing instrument used.
- Ranking of financing instruments (qualitative outcomes):
  - Financing resilience with donors' support or by cutting unproductive government spending yields higher returns to building resilience compared with financing via distortionary taxes.
  - Distortionary taxes impose efficiency costs: corporate income tax and labor income tax reduce resource allocation efficiency; corporate tax is the most costly.
- Mechanisms and macroeconomic effects:
  - Corporate income tax effects:
    - Reduces investment and the stock of private capital.
    - Lowers labor productivity and wages in formal sectors.
    - Generates greater out-migration of skilled workers in SDSs, amplifying output losses.
  - Labor income tax effects:
    - Intermediate negative impact on output.
    - Increases out-migration of skilled workers, but effect smaller than corporate taxes.
  - Value-added taxes and other consumption taxes implied as distortionary but specific quantitative ranking preserved in paper (Table 5: Policy Instruments Rankings referenced).
- Distributional outcomes:
  - Corporate and labor income taxes reduce income inequality because these taxes are levied on higher-income workers.
  - Cutting government transfers increases income inequality since low-income workers typically benefit more from transfers.
- Additional model-specific notes:
  - The model includes adjustment costs for entrepreneurs' and farmers' investment decisions parameterized by ε and σ as in source equations.
  - Entrepreneurs’ profit definitions and farmer profit functions include pricing, wages, capital adjustment terms and retained notation (π definitions preserved).

### Figures and tables (listed in source)
- Figures:
  - 1. The Impact of ND on the Benchmark Economy (page reference 14).
  - 2. Fiscal Instruments (page reference 17).
- Tables:
  - 1. Model Structure (page reference 7).
  - 2. Parameters (page reference 11).
  - 3. Moment Calibration Summary (page reference 12).
  - 4. Tax Changes in the Fiscal Policy Experiments (page reference 14).
  - 5. Policy Instruments Rankings (page reference 18).

*Source: 1dmaea2022002 - References (IMF PDF chapter/section), January 18, 2022.*

### 14.  The model is used to analyze different fiscal instruments to finance the buidling of

### 14.  The model is used to analyze different fiscal instruments to finance the buidling of 

### Calibration and data inputs
- Economy calibrated to match quantitative parameters of the Dominica economy (a SDS island in the Caribbean), matching sector sizes, use of labor and capital, inter-sector linkages, and the distribution of income.
- Preferences:
  - Households’ consumption shares calibrated to CPI basket: 휓 Share of tradables in total consumption 35.9 (model = 35.9), 훾 Share of services in total consumption 42.2 (model = 42.2).
- Labor force allocation (from Dominica Social Security Administration):
  - 휇௚ share of government workers 0.349
  - 휇௦ share of skilled workers 0.444
  - 휇௨ share of unskilled workers 0.105
  - 휇௘ share of entrepreneur 0.049
  - 휇௙ share of big farmers 0.053
  - Emigrants allocated to skilled workers and government workers; emigration elasticity 훼 = 0.5; US wages 푤௨௦ calibrated so remittances match data (푤௨௦ Remittances 10.0 percent of GDP).
- Economic sectors and productivity shares (model = data):
  - Agricultural output 푧௔ 16.8 percent of GDP
  - Services output 푧௘,௦ 50.9 percent of GDP
  - Informal sector output 푧௦ 3.7 percent of GDP
- Capital elasticities and depreciation:
  - 훼௠ tradable elasticity to private capital 0.24
  - 훼∗ agriculture elasticity to private capital 0.40
  - 훼௚ elasticity to public capital 0.14
  - 훿 private capital depreciation 0.14
  - 훿௚ public capital depreciation 0.109
- Fiscal parameters calibrated to central government data:
  - Revenue shares (percent of GDP): consumption tax 13.5, personal income tax 4.5, corporate taxes 4.0, Grants (Gr) 5.0, Non-tax revenues (NR) 10.0, Public sector wage bill 푤௚ 11, Public Capital investment over GDP in steady state 퐾଴௚ 12, Transfers to households 5.5.
  - Interest rate on government debt 푟∗ 0.029
  - Discount factor Β 0.96
  - Lower bound of agricultural consumption 푐̅ 0.20

### Natural disaster (ND) shock design
- ND shock assumptions:
  - Shock reduces stock of non-resilient public capital by 40 percent, impacting manufacturing and agriculture sectors.
  - Temporary efficiency loss in capital and labor productivity 휃 and additional recovery loss in depreciation rate 훿ௗ.
  - Formal and informal services productivity impacted transitorily for one period (one year).
  - Productivity decline calibrated to obtain a 5 percent drop in GDP (capturing a relatively large ND, about one standard deviation of destruction as a share of GDP by a hurricane, based on the ND Annual Loss curve for Dominica).

### Policy experiment design
- Baseline: economy starts from a non-resilient state (non-resilient infrastructure).
- At time zero, government begins investing in resilient infrastructure; additional investment financed by one fiscal instrument chosen among five alternatives.
- Five financing instruments studied (each set to raise 1 percent of GDP in steady state):
  - Consumption tax (value-added tax, VAT)
  - Corporate tax (profits, CIT)
  - Payroll tax (personal income tax, PIT)
  - Reduction in public sector wage
  - Increase in donor financing (grants)
- In period ten, ND shock occurs as specified above.
- Post-shock: economy converges to a new steady state with higher resilient public capital.
- Note: debt financing not considered (because debt service would require subsequent fiscal adjustment and revert analysis to choice of financing instrument).

### Benchmark (no resilient investment) results
- Output decline:
  - Output declines by 16 percent with non-resilient public capital after the ND shock.
  - Capital-intensive sectors experience larger declines: manufacturing declines most, followed by agriculture; formal and informal services decline less.
- Migration and labor effects:
  - ND shock stimulates out-migration as skilled workers experience wage loss from manufacturing shock.
  - Out-migration operates as de-facto insurance by reducing domestic labor supply, containing wage declines for migrants and generating remittances.
- Table of policy changes required to raise 1 percent of GDP (Table 4, Benchmark Values / Policy Change (%)):
  - VAT 0.12 -> 5.2%
  - PIT 0.06 -> 15.0%
  - CIT 0.10 -> 17.2%
  - Public Sector Wages 0.27 -> 2.1%
  - Transfers 0.02 -> -9.0%

### Building resilience — aggregate and sectoral impacts
- Resilient investment reduces the impact of the ND shock on output by 2 to 2.5 percentage points depending on financing instrument.
- Efficiency ranking (impact on output):
  - Value-added tax (VAT) is most efficient — relatively less distortionary, broader tax base, stimulates saving and investment.
  - Corporate income tax (CIT) is costliest — reduces investment and private capital, lowering labor productivity and wages in formal sectors; amplified by skilled out-migration.
  - Labor income tax (PIT) has intermediate negative impact — increases out-migration of skilled workers.
  - Reduction in government wage expenditure is less distortionary under model assumptions (government workers assumed to have zero marginal contribution to output).
  - Increase in donor finance is most beneficial instrument — a “gift” that causes no tax distortions.
- Financing resilience by cutting unproductive government spending yields higher return to building resilience than financing via distortionary CIT or PIT.

### Distributional consequences
- General effects of ND on inequality:
  - ND shocks increase income inequality by disproportionately affecting lower-wage sectors.
  - Unskilled workers suffer more because they cannot migrate; skilled workers can migrate and send remittances, cushioning welfare losses.
  - Government workers’ wage assumed unchanged (empirical regularity).
- Fiscal instrument effects on inequality:
  - CIT increase reduces income inequality (levied on higher-income households: skilled workers and entrepreneurs).
  - PIT increase reduces income inequality (affects relatively high-income government employees and skilled workers; informal workers not taxed).
  - VAT increase results in a small decline in income inequality due to SDS model-specific mechanisms:
    - Unskilled workers increase working hours in informal service sector (formal sector reduced by out-migration), helping maintain wage earnings; informal workers are not paying labor taxes.
  - Reduction in government workers' wage expense reduces income inequality (government wages higher than most sectors).
  - Cut in government transfers increases income inequality (transfers are a larger share of income for low-income households; proportional cut implies larger decline in household income for low-income households).

### Welfare by household type (consumption responses)
- Unskilled workers (hand-to-mouth, cannot migrate):
  - Consumption declines when ND hits under benchmark.
  - Resilient investment reduces the concurrent decline in consumption at ND.
  - Relative impacts of financing instruments on unskilled consumption:
    - CIT financing affects consumption least compared with benchmark of no resilient investment.
    - PIT financing has relatively small effect on unskilled workers’ wage (informal workers pay no labor tax).
    - VAT financing reduces unskilled consumption more significantly (high propensity to consume).
    - Financing with a decline in transfers has the highest negative impact on unskilled consumption.
- Skilled workers (can migrate):
  - Migration allows smoothing of income and consumption when ND occurs, producing no concurrent welfare loss in the model (due to assumption of no delays or costs of migration).
  - Ranking of instruments’ impact similar to unskilled, except:
    - CIT reduces skilled household consumption because it lowers formal sector wages via reduced investment.
    - PIT increase is most detrimental to skilled households’ consumption; these workers out-migrate and send remittances but face net income and consumption loss.
    - Cut in government transfers reduces skilled household consumption less than for unskilled due to migration ability.

### Policy trade-offs and rankings
- Table 5: Dominica: Policy Instruments Rankings (efficiency = impact on output; inequality = GINI, higher value means more equality)
  - POLICY COLUMNS: VALUE-ADDED TAXES | CORPORATE TAXES | GOVERNMENT CONSUMPTION PUBLIC INVESTMENT
  - OUTPUT ranking: VALUE-ADDED = 2, CORPORATE = 3, GOVERNMENT = 1, CONSUMPTION PUBLIC INVESTMENT = 4
  - INEQUALITY (GINI) ranking: VALUE-ADDED = 3, CORPORATE = 2, GOVERNMENT = 4, CONSUMPTION PUBLIC INVESTMENT = 1
- Key tradeoff: instruments that are more efficient (less output cost) are not necessarily those that are most progressive; CIT and PIT reduce inequality more but at higher efficiency cost; VAT is most efficient but only slightly reduces inequality in this model due to sectoral and migration dynamics.
- Model contribution: quantifies the efficiency and distributional tradeoffs of alternative financing instruments under SDS-specific features (migration, sectoral structure, informality, remittances).

*Source: authors’ calculations based on Dominica government data.*

### 24. Dominica is highly vulnerable to NDs that can have large human, economic, and social

### 24. Dominica is highly vulnerable to NDs that can have large human, economic, and social costs

### Vulnerability to natural disasters and the importance of resilient infrastructure
- Dominica is highly vulnerable to NDs that can have large human, economic, and social costs, making it critical to invest in resilient infrastructure.
- Benefits of building resilience for NDs include lowering the economic and social impact, speeding up recovery, and providing greater continuity in public services.

### Fiscal instruments, efficiency, and equity trade-offs (findings from the multi-sector GE model with SDS features)
- Public investment in resilience:
  - Has high returns in terms of output and employment.
  - Increases inequality because the formal sector and skilled workers benefit the most in the long-term.
  - Is critical to containing the income losses of poor households and informal workers in the aftermath of a ND, showing smaller increases in inequality after those events when more resilient investment is in place.
- Corporate taxes and labor income taxes:
  - Reduce income inequality.
  - Can have more negative impact in the long-run, especially in SDSs, due to amplification effects from out-migration, limited labor mobility across sectors because of skills’ mismatches and labor informality, and limited economic diversification.
- Transfers and value-added taxes:
  - Are relatively more efficient, with a lower distortionary impact.
  - Are less effective in reducing income inequality compared with corporate and labor income taxes.
- Government spending inefficiencies:
  - Reducing unproductive government employment to contain the wage bill is identified as an efficient way to finance costly resilient investment.
  - Increasing the efficiency of government transfers with better targeting can provide fiscal space for resilience investment while addressing income inequality.

### Role of foreign grants
- Grant financing:
  - Contains income loss of low-income households after NDs.
  - Minimizes output and employment loss associated with tax distortions.
  - Helps address both efficiency and income distribution considerations, containing increases in inequality given the disproportionate impact of NDs on workers with relatively lower wages.

### Measuring the shadow economy in Dominica (MIMIC model application)
- Motivation:
  - Knowing the size of the shadow economy matters for designing public policies, observing trends of economic activity and employment, and determining growth–employment linkages.
  - The shadow economy includes economic activities hidden from official authorities that, if recorded, would contribute to national GDP.
- Methodology:
  - The MIMIC model relates a latent unobserved variable (the shadow economy) to its drivers (multiple causes) and observable outcomes (multiple indicators) through a structural equation model.
  - Indicator variables used: rate of real GDP growth and rate of employment (both hypothesized to be negatively related to the size of unreported activities).
  - Causal variables used: tax rates (top marginal tax rates for corporate taxes and personal income taxes), size of agriculture (agricultural value added), size of tourism (tourism as a percent of GDP), and Regulatory Quality Index from the World Governance Indicators.
- Empirical findings for Dominica and the sample:
  - The burden of taxation seems to increase the size of the shadow economy (marginal corporate and personal income tax rates show the expected positive relationship).
  - A larger tourism sector is correlated with a smaller informal sector (tourism percent of GDP is negatively related to the size of the shadow economy).
  - Higher regulatory quality tends to decrease the size of the shadow economy (negative relationship).
  - Agriculture, while positively related to the size of the shadow economy, is not statistically significant for the sample.
  - Real GDP growth and employment rate are negatively related to the size of the shadow economy, consistent with theory.
- Key quantitative estimate:
  - The shadow economy of Dominica averaged 46 percent of GDP over the 2011–19 period (absolute size depends on the chosen benchmark).
- Benchmark note:
  - The absolute size estimate depends on the chosen benchmark; as the benchmark the analysis uses the size of informal employment in Jamaica in 2015, measured by the Labor Force Survey.

*International Monetary Fund — Chapter: Dominica (extracted content).*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2022/english/1dmaea2022002.pdf_