## wpiea2021100-print-pdf

## Source details

**Canonical URL:** [wpiea2021100-print-pdf](https://www.imf.org/-/media/files/publications/wp/2021/english/wpiea2021100-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2021/english/wpiea2021100-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2021/english/wpiea2021100-print-pdf.pdf.json)

---

### I. Key short- and long-run assessment of Ukraine’s economic performance
- Ukraine’s annual growth rate over 1990–2017: -0.2 percent (fifth lowest in the world over 1990–2017).
- Ukraine’s GDP per capita currently equals:
  - 30 percent of Poland,
  - 45 percent of Belarus,
  - 80 percent of Georgia.
- Demographics and investment:
  - Population growth averaged -0.5 percent per year since 1990 (tenth fastest decline globally).
  - Investment to GDP ratio averaged 20 percent in 1990–2017.
  - Benchmarks: lower middle-income countries average 25.5 percent; Central European countries average 23.1 percent; countries with adverse demographic trends average 22.2 percent.
  - Investment ratio fell to 16.2 percent between 2010–17 (bottom 10 percentile of world distribution).
- Projected labor force trend: labor force projected to shrink by 1.2 percent per annum over 2018–2030 (ILO-based).
- Baseline long-run GDP growth under current trends: 2.7 percent (potential GDP growth decelerating to 2.7 percent in the long run).
  - Under this baseline, Ukraine’s GDP per capita would be half of Poland’s by 2040.

### II. Institutional constraints and measured reform gaps
- Primary impediments to investment cited by foreign investors (2019 Q1): widespread corruption, lack of trust in the judiciary, monopolization/capture by oligarchs.
- Institutional features depressing investment:
  - Weak protection of property rights; weak court system susceptible to business and political influence.
  - State and oligarch ownership of substantial productive assets; concentrated rents from regulated markets.
  - Resultant underinvestment, limited foreign investment, and limited export reorientation away from commodities.
- Macrostructural areas analyzed (composite indicators built by principal component analysis):
  - Legal system (corruption, governance, crime, rule of law, property rights)
  - Financial system (financial development, access, banking soundness)
  - Product markets (competition, informality, regulatory burdens)
  - Labor markets (minimum wages, employment protection)
  - Taxation system (tax distortions)
  - Trade and openness (tariffs, non-tariff barriers)
  - Research and development (innovation and technology adoption)
- Comparative findings versus Poland:
  - Largest gap: legal system (gap twice as large as product markets).
  - Second largest gap: product markets.
  - Significant but smaller gaps: financial system, trade and openness.
  - Negligible shortfalls: R&D and taxation system.
  - Labor market performance: stronger in Ukraine than in Poland (in employment protection/dismissal rules).
- Correlations among composite indicators are large and positive, implying positive spillovers among reforms; causality not established but legal system likely to influence others (product markets, financial depth, trade facilitation).

### III. Methodology overview
- Data and aggregation:
  - Cross-country structural indicators assembled from the Fund’s Macrostructural Indicators Database (sources include World Bank, OECD, WEF, Transparency International).
  - Principal component analysis used to aggregate correlated indicators into composite indices per macrostructural area.
- Three analytical steps:
  1. Identify main structural gaps versus Poland (Section II).
  2. Use cross-country variance to estimate TFP impact of policy scenarios, controlling for correlations (Section III).
  3. Simulate macroeconomic outcomes with a small open economy New Keynesian DSGE model with financial frictions (Section IV).
- Cross-country regression sample for TFP estimation: 67 countries over 2000–2015.
- Regression specification links log(TFP) to composite structural indicators and controls; estimated long-run reform elasticities applied to Ukraine’s gaps versus Poland to compute TFP impacts.

### IV. Policy reform scenarios (structural and land reform design)
- Full reform scenario:
  - Catch-up to Poland in four macrostructural areas: legal system, product markets, financial system, trade and openness.
  - Full land market liberalization: no restrictions on non-resident buyers; land can be pledged as collateral.
- Partial reform scenario:
  - Catch-up to Poland in product markets, financial system, trade and openness.
  - Legal system gap remains constant (largest gap not addressed).
  - Partial land reform: access to land restricted to Ukrainian nationals; pledging as collateral allowed but land valuation reduced due to foreigner restrictions.
- Backsliding scenario:
  - Deterioration in legal system at same pace as 2010–2015 with negative spillovers to other areas.
  - No land reform (moratorium remains).

### V. Estimated TFP impacts of reforms (cross-country regression results and gaps)
- Long-run elasticities (per one-standard deviation improvement in institutional quality):
  - Partial reform: TFP growth increases by 5 percent (for a one-standard deviation improvement).
  - Full reform: TFP growth increases by 6 percent (for a one-standard deviation improvement).
  - Backsliding: TFP declines by 7 percent (for a one-standard deviation deterioration).
- Cumulative TFP impacts for Ukraine (catching up to Poland, over 20 years unless otherwise noted):
  - Full reform scenario: TFP increases by 58.78%.
  - Partial reform scenario (legal gap not addressed): TFP increases by 22.30 percent.
  - Backsliding scenario: TFP declines by 17.52%.

### VI. Economic impact of land reform (model-based local expert estimates)
- Land reform impacts treated separately and incorporated into scenarios:
  - Lifting the moratorium on land sales but restricting buyers to Ukrainian individuals/legal entities:
    - Cumulative GDP increase: 6.07% over 10 years (incorporated into partial reform scenario).
  - Removing restrictions on foreigners (part of full reform):
    - Cumulative GDP increase: 12.6 percent over 10 years.
- Cross-country price-change estimates (EasyBusiness (2019)):
  - Fully open land market could add up to 0.77 percentage point additional GDP growth per annum.
  - Scenario with no access to foreigners estimated at 0.40 percentage points additional GDP growth per annum.
- Collateral channel (macroeconomic model with financial accelerator):
  - Fully open land market collateral effects could raise annual GDP growth by a further 0.49 percentage points.
  - Restrictions on foreigners reduce collateral channel to 0.20 percentage points.

### VII. Macroeconomic simulation framework (DSGE with financial frictions) and application
- Model features:
  - Small open economy New Keynesian DSGE (Gali and Monacelli (2005)) combined with a financial accelerator (Bernanke, Gertler and Gilchrist (1999)).
  - Liquidity premium in uncovered interest parity to reflect shallow FX markets; central bank FX interventions modeled to affect liquidity premium and exchange rate.
  - Central bank follows an FX intervention rule aimed at stabilizing the nominal exchange rate.
- Scenario simulation approach:
  - Reform scenarios simulated as a gradual increase in TFP over 20 years.
  - Cumulative TFP increases set to match empirically estimated TFP impacts, with adjustments for non-financial effects of land reform.
- Calibration:
  - Model calibrated to emerging market business-cycle literature and recent Ukraine data.
- Purpose:
  - To generate broader macroeconomic outcomes (growth, exchange rate, financial conditions, and welfare implications) under the full, partial, and backsliding reform scenarios.

### Macroeconomic outcomes of reform scenarios
- Analysis focuses on two main outcomes:
  - (i) the growth impact of structural reforms and the convergence path to Poland.
  - (ii) the macroeconomic adjustment and its implications for the accumulation of FX reserves.
- The financial effects of land reform are simulated as a relaxation in financial constraints proportionate to empirically estimated increases in land prices under reform scenarios.

### Scenario summaries and dynamics
- Full reform scenario
  - Average annual GDP growth over 20 years could be as high as 7.1 percent.
  - Could result in catching up with the current level of development of Poland by 2040.
  - Capital inflows finance a rise in consumption and investment in advance of productivity gains, leading initially to:
    - a decline in net exports,
    - a real appreciation,
    - the emergence of a positive output gap in the initial years.
  - FX reserves increase by 3.2 billion USD by 2024, and 11.2 billion by 2040.
  - Nominal interest rates only increase slightly due to the lower profile of inflation brought about by the gradual rise in total factor productivity.
  - Model leads to a steady depreciation in the real exchange rates and a rebound in net exports over 20 years.
  - Note: "Our model does not capture Balassa-Samuelson effects."
- Partial reform scenario
  - Macro dynamics similar to full reform but at a smaller scale.
  - Structural gap in the legal system is not addressed; productivity gains from reforms are lower.
  - Average annual GDP growth over 20 years amounts to 4.50 percent.
  - Projected GDP per capita reaches 77.28 percent of Poland’s current level by 2040.
  - FX reserves increase by 1.29 billion USD by 2024 and 4.22 billion by 2040.
- Backsliding scenario
  - Reform impetus ends and average GDP growth over the next 20 years drops to 1.67 percent per annum.
  - In 2040, GDP per capita remains less than half of current Polish GDP per capita.
  - Decline in capital inflows necessitates adjustment in net exports primarily through import compression.
  - Real appreciation driven by higher inflation erodes competitiveness and prevents the central bank from providing monetary stimulus.
  - Consumption and gross fixed capital formation decline relative to the baseline.
  - Stabilizing the nominal exchange rate leads to an erosion in FX reserves of 0.84 billion USD by 2024 and 2.30 billion by 2040, raising vulnerability to external shocks and adverse shifts in market sentiment.

### Quantitative results (selected)
- Table 5: Reform scenarios: Real GDP growth and catch-up
  - Baseline
    - Avg. GDP growth rate (annual): 2.76%
    - Cumulative growth by 2040: 72.40%
    - Projected GDP per capita (PPP) relative to current level in Poland: 54.63%
  - Full reform
    - Avg. GDP growth rate (annual): 7.10%
    - Cumulative growth by 2040: 293.07%
    - Projected GDP per capita (PPP) relative to current level in Poland: 130.15%
  - Partial reform
    - Avg. GDP growth rate (annual): 4.50%
    - Cumulative growth by 2040: 140.95%
    - Projected GDP per capita (PPP) relative to current level in Poland: 77.28%
  - Backsliding
    - Avg. GDP growth rate (annual): 1.67%
    - Cumulative growth by 2040: 39.25%
    - Projected GDP per capita (PPP) relative to current level in Poland: 43.92%

### Policy implications and conclusions
- Three key findings:
  - Institutional weaknesses: Ukraine is furthest behind Poland in the legal system, market competition, openness to trade and financial depth; labor market flexibility is stronger than in Poland; Ukraine performs relatively well in R&D and taxation.
  - Sequencing: Improving the legal system is an important prerequisite for other reforms to work. Tackling corruption and strengthening the rule of law would have significant positive spillover effects.
    - A "full reform scenario" that closes the gap vis-à-vis Poland across the legal system, product market deregulation, trade facilitation, financial market development, and gradual opening of the agricultural land market (including to foreigners) could result in an annual growth rate close to 7 percent.
    - If the legal system is not reformed while other macrostructural reforms proceed and access to land markets by foreigners is restricted, growth would be limited to 4–5 percent per annum.
  - Long-term commitment: Successful convergence to Poland or other Eastern European peers requires sustained reform commitment over decades.
    - Even under the full reform scenario, Ukraine would need to grow at close to 7 percent for 20 years to reach the level of development of Poland today.
    - Reform commitment must persist over five parliamentary cycles against vested interests and reform fatigue.
    - Reform reversals have significant economic and social costs: under backsliding where the legal system deteriorates further, GDP growth would drop to below 2 percent per annum.

### Appendix II: Description of the small open economy New Keynesian DSGE model with financial frictions — model structure and calibration (selected)
- Model structure (key elements):
  - CES aggregation of differentiated domestic and foreign goods with elasticities ε_ρ, ε_F, ε_H.
  - Exchange rate and terms of trade definitions: S_t = P_F,t / P_H,t; Θ_t = S_t (P_H,t / P_t) = P_F,t / P_t; ε_t = P_t / P_t^*.
  - Representative infinitely lived household with CRRA utility, incomplete financial markets (private, government, and foreign bonds).
  - Liquidity premium on foreign bonds decreasing in reserves: Λ_t = (ℛ_t / ℛ_{t−1})^{ε_ℛ}, ε_ℛ > 0.
  - Investment and capital accumulation with adjustment costs; capital producers’ FOCs (q_t and q_t').
  - Entrepreneurs subject to idiosyncratic shocks, costly state verification, and financial constraints; financial accelerator through net worth and collateral channels.
  - Monopolistically competitive final good firms with Cobb-Douglas production Y_H,t(j) = A_t K_t'(j)^a H_t(j)^{1−a} and Calvo pricing.
  - Aggregate demand and supply conditions, trade balance, debt-elastic interest rate spread, monetary and reserves policy rules, fiscal rule with lump-sum tax.
  - Rational expectations equilibrium defined over full set of endogenous variables given exogenous processes {ε_t^A, ε_t^μ}.
- Calibration (selected parameters and steady-state targets):
  - Period = quarter.
  - σ = 1.
  - H̄ = 0.2934.
  - β = 0.9975 (annualized steady state real interest rate = 1%).
  - Steady state inflation Π̄ calibrated to annual inflation rate of 5% (National Bank of Ukraine target) → nominal interest rate of 6% in annual terms.
  - ε_ℛ = 0.83 (liquidity premium responsiveness to reserves).
  - Production parameter: a = 0.3.
  - Price elasticity of demand among domestic goods: 휀휀P = 7.6 (implying a mark-up of 15% over marginal costs).
  - Calvo parameter: 휃휃 = 0.75.
  - Quarterly depreciation rate: 훿훿 = 0.05.
  - Investment adjustment cost parameter: 휎휎I = 4.602.
  - Trade elasticities: 휀휀F = 휀휀H = 휀휀I = 1.2.
  - Openness parameters: 훼훼 = 훼훼I = 0.36.
  - Debt-elasticity parameter: 휒휒 = 10−7.
  - Financial-frictions and monitoring calibration:
    - TTh,e set to 1% of steady-state income.
    - 훾훾 = 0.915, 휇휇 = 0.324, 휎휎entt = 0.125.
  - Monetary policy coefficients: 휙휙π = 1.5, 휙휙y = 0.56, 휙휙i = 0.99.
  - Reserve policy rule: 휐휐ε = 10, ℛ� = 0.18 (targets USD 20bn under 2019 exchange rates), 휐휐ℛ set extremely small.
  - Fiscal parameters: 휏휏y = 0.47, steady-state G/Y� = 0.32, b�g4Y� = 0.60.
- Innovation processes and simulated reform transition paths:
  - Reform transition paths {ε_t^A}t=1 80 and {ε_t^μ}t=1 80 gradually adjust A_t and monitoring costs to match cumulative impacts of land reform and other structural reforms per scenario.
  - Collateral-impact under the full reform scenario > partial reform scenario, consistent with differential land price estimates.

*IMF staff calculations and analysis as presented in the referenced chapter.*

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

### References (wpiea2021100-print-pdf)

### I. Key short- and long-run assessment of Ukraine’s economic performance
- Ukraine’s annual growth rate over 1990–2017: -0.2 percent (fifth lowest in the world over 1990–2017).
- Ukraine’s GDP per capita currently equals:
  - 30 percent of Poland,
  - 45 percent of Belarus,
  - 80 percent of Georgia.
- Demographics and investment:
  - Population growth averaged -0.5 percent per year since 1990 (tenth fastest decline globally).
  - Investment to GDP ratio averaged 20 percent in 1990–2017.
  - Benchmarks: lower middle-income countries average 25.5 percent; Central European countries average 23.1 percent; countries with adverse demographic trends average 22.2 percent.
  - Investment ratio fell to 16.2 percent between 2010–17 (bottom 10 percentile of world distribution).
- Projected labor force trend: labor force projected to shrink by 1.2 percent per annum over 2018–2030 (ILO-based).
- Baseline long-run GDP growth under current trends: 2.7 percent (potential GDP growth decelerating to 2.7 percent in the long run).
  - Under this baseline, Ukraine’s GDP per capita would be half of Poland’s by 2040.

### II. Institutional constraints and measured reform gaps
- Primary impediments to investment cited by foreign investors (2019 Q1): widespread corruption, lack of trust in the judiciary, monopolization/capture by oligarchs.
- Institutional features depressing investment:
  - Weak protection of property rights; weak court system susceptible to business and political influence.
  - State and oligarch ownership of substantial productive assets; concentrated rents from regulated markets.
  - Resultant underinvestment, limited foreign investment, and limited export reorientation away from commodities.
- Macrostructural areas analyzed (composite indicators built by principal component analysis):
  - Legal system (corruption, governance, crime, rule of law, property rights)
  - Financial system (financial development, access, banking soundness)
  - Product markets (competition, informality, regulatory burdens)
  - Labor markets (minimum wages, employment protection)
  - Taxation system (tax distortions)
  - Trade and openness (tariffs, non-tariff barriers)
  - Research and development (innovation and technology adoption)
- Comparative findings versus Poland:
  - Largest gap: legal system (gap twice as large as product markets).
  - Second largest gap: product markets.
  - Significant but smaller gaps: financial system, trade and openness.
  - Negligible shortfalls: R&D and taxation system.
  - Labor market performance: stronger in Ukraine than in Poland (in employment protection/dismissal rules).
- Correlations among composite indicators are large and positive, implying positive spillovers among reforms; causality not established but legal system likely to influence others (product markets, financial depth, trade facilitation).

### III. Methodology overview
- Data and aggregation:
  - Cross-country structural indicators assembled from the Fund’s Macrostructural Indicators Database (sources include World Bank, OECD, WEF, Transparency International).
  - Principal component analysis used to aggregate correlated indicators into composite indices per macrostructural area.
- Three analytical steps:
  1. Identify main structural gaps versus Poland (Section II).
  2. Use cross-country variance to estimate TFP impact of policy scenarios, controlling for correlations (Section III).
  3. Simulate macroeconomic outcomes with a small open economy New Keynesian DSGE model with financial frictions (Section IV).
- Cross-country regression sample for TFP estimation: 67 countries over 2000–2015.
- Regression specification links log(TFP) to composite structural indicators and controls; estimated long-run reform elasticities applied to Ukraine’s gaps versus Poland to compute TFP impacts.

### IV. Policy reform scenarios (structural and land reform design)
- Full reform scenario:
  - Catch-up to Poland in four macrostructural areas: legal system, product markets, financial system, trade and openness.
  - Full land market liberalization: no restrictions on non-resident buyers; land can be pledged as collateral.
- Partial reform scenario:
  - Catch-up to Poland in product markets, financial system, trade and openness.
  - Legal system gap remains constant (largest gap not addressed).
  - Partial land reform: access to land restricted to Ukrainian nationals; pledging as collateral allowed but land valuation reduced due to foreigner restrictions.
- Backsliding scenario:
  - Deterioration in legal system at same pace as 2010–2015 with negative spillovers to other areas.
  - No land reform (moratorium remains).

### V. Estimated TFP impacts of reforms (cross-country regression results and gaps)
- Long-run elasticities (per one-standard deviation improvement in institutional quality):
  - Partial reform: TFP growth increases by 5 percent (for a one-standard deviation improvement).
  - Full reform: TFP growth increases by 6 percent (for a one-standard deviation improvement).
  - Backsliding: TFP declines by 7 percent (for a one-standard deviation deterioration).
- Cumulative TFP impacts for Ukraine (catching up to Poland, over 20 years unless otherwise noted):
  - Full reform scenario: TFP increases by 58.78%.
  - Partial reform scenario (legal gap not addressed): TFP increases by 22.30 percent.
  - Backsliding scenario: TFP declines by 17.52%.

### VI. Economic impact of land reform (model-based local expert estimates)
- Land reform impacts treated separately and incorporated into scenarios:
  - Lifting the moratorium on land sales but restricting buyers to Ukrainian individuals/legal entities:
    - Cumulative GDP increase: 6.07% over 10 years (incorporated into partial reform scenario).
  - Removing restrictions on foreigners (part of full reform):
    - Cumulative GDP increase: 12.6 percent over 10 years.
- Cross-country price-change estimates (EasyBusiness (2019)):
  - Fully open land market could add up to 0.77 percentage point additional GDP growth per annum.
  - Scenario with no access to foreigners estimated at 0.40 percentage points additional GDP growth per annum.
- Collateral channel (macroeconomic model with financial accelerator):
  - Fully open land market collateral effects could raise annual GDP growth by a further 0.49 percentage points.
  - Restrictions on foreigners reduce collateral channel to 0.20 percentage points.

### VII. Macroeconomic simulation framework (DSGE with financial frictions) and application
- Model features:
  - Small open economy New Keynesian DSGE (Gali and Monacelli (2005)) combined with a financial accelerator (Bernanke, Gertler and Gilchrist (1999)).
  - Liquidity premium in uncovered interest parity to reflect shallow FX markets; central bank FX interventions modeled to affect liquidity premium and exchange rate.
  - Central bank follows an FX intervention rule aimed at stabilizing the nominal exchange rate.
- Scenario simulation approach:
  - Reform scenarios simulated as a gradual increase in TFP over 20 years.
  - Cumulative TFP increases set to match empirically estimated TFP impacts, with adjustments for non-financial effects of land reform.
- Calibration:
  - Model calibrated to emerging market business-cycle literature and recent Ukraine data.
- Purpose:
  - To generate broader macroeconomic outcomes (growth, exchange rate, financial conditions, and welfare implications) under the full, partial, and backsliding reform scenarios.

*Italic: IMF staff calculations and analysis as presented in the referenced chapter.*

### Section III.B above). The financial effects of land reform are simulated as  a relaxation in

### wpiea2021100-print-pdf - Section III.B above). The financial effects of land reform are simulated as  a relaxation in 

### Macroeconomic outcomes of reform scenarios
- Analysis focuses on two main outcomes:
  - (i) the growth impact of structural reforms and the convergence path to Poland.
  - (ii) the macroeconomic adjustment and its implications for the accumulation of FX reserves.
- The financial effects of land reform are simulated as a relaxation in financial constraints proportionate to empirically estimated increases in land prices under reform scenarios.

### Scenario summaries and dynamics
- Full reform scenario
  - Average annual GDP growth over 20 years could be as high as 7.1 percent.
  - Could result in catching up with the current level of development of Poland by 2040.
  - Capital inflows finance a rise in consumption and investment in advance of productivity gains, leading initially to:
    - a decline in net exports,
    - a real appreciation,
    - the emergence of a positive output gap in the initial years.
  - FX reserves increase by 3.2 billion USD by 2024, and 11.2 billion by 2040.
  - Nominal interest rates only increase slightly due to the lower profile of inflation brought about by the gradual rise in total factor productivity.
  - Model leads to a steady depreciation in the real exchange rates and a rebound in net exports over 20 years.
  - Note: "Our model does not capture Balassa-Samuelson effects."

- Partial reform scenario
  - Macro dynamics similar to full reform but at a smaller scale.
  - Structural gap in the legal system is not addressed; productivity gains from reforms are lower.
  - Average annual GDP growth over 20 years amounts to 4.50 percent.
  - Projected GDP per capita reaches 77.28 percent of Poland’s current level by 2040.
  - FX reserves increase by 1.29 billion USD by 2024 and 4.22 billion by 2040.

- Backsliding scenario
  - Reform impetus ends and average GDP growth over the next 20 years drops to 1.67 percent per annum.
  - In 2040, GDP per capita remains less than half of current Polish GDP per capita.
  - Decline in capital inflows necessitates adjustment in net exports primarily through import compression.
  - Real appreciation driven by higher inflation erodes competitiveness and prevents the central bank from providing monetary stimulus.
  - Consumption and gross fixed capital formation decline relative to the baseline.
  - Stabilizing the nominal exchange rate leads to an erosion in FX reserves of 0.84 billion USD by 2024 and 2.30 billion by 2040, raising vulnerability to external shocks and adverse shifts in market sentiment.

### Quantitative results (selected)
- Table 5: Reform scenarios: Real GDP growth and catch-up
  - Baseline
    - Avg. GDP growth rate (annual): 2.76%
    - Cumulative growth by 2040: 72.40%
    - Projected GDP per capita (PPP) relative to current level in Poland: 54.63%
  - Full reform
    - Avg. GDP growth rate (annual): 7.10%
    - Cumulative growth by 2040: 293.07%
    - Projected GDP per capita (PPP) relative to current level in Poland: 130.15%
  - Partial reform
    - Avg. GDP growth rate (annual): 4.50%
    - Cumulative growth by 2040: 140.95%
    - Projected GDP per capita (PPP) relative to current level in Poland: 77.28%
  - Backsliding
    - Avg. GDP growth rate (annual): 1.67%
    - Cumulative growth by 2040: 39.25%
    - Projected GDP per capita (PPP) relative to current level in Poland: 43.92%

### Policy implications and conclusions
- Three key findings:
  - Institutional weaknesses: Ukraine is furthest behind Poland in the legal system, market competition, openness to trade and financial depth; labor market flexibility is stronger than in Poland; Ukraine performs relatively well in R&D and taxation.
  - Sequencing: Improving the legal system is an important prerequisite for other reforms to work. Tackling corruption and strengthening the rule of law would have significant positive spillover effects.
    - A "full reform scenario" that closes the gap vis-à-vis Poland across the legal system, product market deregulation, trade facilitation, financial market development, and gradual opening of the agricultural land market (including to foreigners) could result in an annual growth rate close to 7 percent.
    - If the legal system is not reformed while other macrostructural reforms proceed and access to land markets by foreigners is restricted, growth would be limited to 4–5 percent per annum.
  - Long-term commitment: Successful convergence to Poland or other Eastern European peers requires sustained reform commitment over decades.
    - Even under the full reform scenario, Ukraine would need to grow at close to 7 percent for 20 years to reach the level of development of Poland today.
    - Reform commitment must persist over five parliamentary cycles against vested interests and reform fatigue.
    - Reform reversals have significant economic and social costs: under backsliding where the legal system deteriorates further, GDP growth would drop to below 2 percent per annum.

*IMF staff calculations based on model simulations of reform scenarios and baseline forecasts. Canonical source: https://www.imf.org/-/media/files/publications/wp/2021/english/wpiea2021100-print-pdf.pdf*

### Appendix II.

### Appendix II. Description of the small open economy New Keynesian DSGE model with financial frictions

### II.1 Small open economy setting and CES aggregation
- World composed of continuum of small open economies i ∈ [0,1] with identical preferences, technology and market structure.
- Final good firms j ∈ [0,1] produce differentiated goods aggregated via CES at multiple levels:
  - Domestic goods basket C_H,t and price index P_H,t with elasticity ε_ρ > 1:
    - (1) C_H,t ≡ ( ∫_0^1 C_H,t(j)^{ε_ρ/(ε_ρ−1)} dj )^{(ε_ρ−1)/ε_ρ}
    - (2) P_H,t ≡ ( ∫_0^1 P_H,t(j)^{1−ε_ρ} dj )^{1/(1−ε_ρ)}
  - Country-specific import baskets C_i,t and P_i,t; aggregated foreign basket C_F,t and P_F,t (trade elasticity ε_F).
  - Aggregate consumption basket C_t and CPI P_t with openness parameter α ∈ [0,1] and trade elasticity ε_H > 0:
    - (3) C_t = [ (1−α)^{1/ε_H} C_H,t^{(ε_H−1)/ε_H} + α^{1/ε_H} C_F,t^{(ε_H−1)/ε_H} ]^{ε_H/(ε_H−1)}
    - (4) P_t = [ (1−α) P_H,t^{1−ε_H} + α P_F,t^{1−ε_H} ]^{1/(1−ε_H)}
  - Demand relationships (price-dependent shares):
    - (5) C_H,t(j) = C_H,t [ P_H,t(j) / P_H,t ]^{−ε_ρ}
    - (6) C_i,t(j) = C_i,t [ P_i,t(j) / P_i,t ]^{−ε_ρ}
    - (7) C_i,t(j) = C_F,t [ P_i,t / P_F,t ]^{−ε_F}
    - (8) C_H,t = (1−α) [ P_H,t / P_t ]^{−ε_H} C_t
    - (9) C_F,t = α [ P_F,t / P_t ]^{−ε_H} C_t
- Exchange rate and terms of trade:
  - (10) S_t = P_F,t / P_H,t  (nominal exchange rate)
  - (11) Θ_t = S_t (P_H,t / P_t) = P_F,t / P_t  (effective terms of trade / real exchange rate)
  - (12) ε_t = P_t / P_t^*  (nominal exchange rate relative to world price level P_t^*)
  - CPI inflation relation (non-linear):
    - (13) Π_t−1,t = Π_H,t−1,t [ (1−α + α S_t^{1−ε_H}) / (1−α + α S_{t−1}^{1−ε_H}) ]^{1/(1−ε_H)}
  - Real exchange rate expression:
    - (14) Θ_t = [ (1−α) S_t^{ε_H−1} + α ]^{1/(ε_H−1)}
    - Dynamic expression linking real exchange rate to nominal depreciation and international inflation:
    - (15) Θ_t / Θ_{t−1} = (ε_t / ε_{t−1}) Π_t^* Π_{t−1,t}^{−1}

### II.2 Households
- Representative infinitely lived household maximizes expected lifetime utility:
  - (16) E_0 Σ_{t=0}^∞ β^t u(C_t), with β ∈ (0,1), u(C_t) CRRA, IES = 1/σ.
- Financial markets incomplete; asset space: private bonds B_t, government bonds B_t^G, foreign bonds B_t^F.
  - Domestic bonds price = 1, pay gross nominal returns R_t and R_t^G.
  - Foreign bond B_t^F denominated in foreign currency with price ε_t and return ε_{t+1} R_t^F in home currency.
- Period budget constraint (nominal):
  - (17) P_t C_t + B_t + B_t^G + ε_t B_t^F ≤ P_t ∫_0^1 H_t(j) w_t(j) dj + R_{t−1} B_{t−1} + R_{t−1}^G B_{t−1}^G + ε_{t−1} Λ_{t−1} R_{t−1}^F B_{t−1}^F + T_t^{lump}
- Labour supply inelastic: H_t = H̄ with calibrated value below.
- Liquidity premium on foreign bonds decreasing in reserves R_t:
  - (18) Λ_t = (ℛ_t / ℛ_{t−1})^{ε_ℛ}, ε_ℛ > 0
- Euler conditions (first order) for assets:
  - (19) 1 = β E_t [ (C_{t+1}/C_t)^{−σ} (1/Π_{t,t+1}) R_t ]
  - (20) 1 = β E_t [ (C_{t+1}/C_t)^{−σ} (1/Π_{t,t+1}) R_t^G ]
  - (21) 1 = β E_t [ (C_{t+1}/C_t)^{−σ} (ε_{t+1}/ε_t) (1/Π_{t,t+1}) Λ_t R_t^F ]

### II.3 Investment and Capital Producers
- Investment composite (Dixit-Stiglitz) with openness α_I ∈ [0,1] and trade elasticity ε_I > 0:
  - (22) I_t = [ (1−α_I)^{1/ε_I} I_H,t^{(ε_I−1)/ε_I} + α_I^{1/ε_I} I_F,t^{(ε_I−1)/ε_I} ]^{ε_I/(ε_I−1)}
  - (23) I_H,t = (1−α_I) [ P_H,t / P_t^I ]^{−ε_I} I_t
  - (24) I_F,t = α_I [ P_F,t / P_t^I ]^{−ε_I} I_t
  - Investment price index P_t^I defined analogously; real price of investment:
    - (25) p_t^I ≡ P_t^I / P_t = [ (1−α_I) (Θ_t S_t)^{1−ε_I} + α_I Θ_t^{1−ε_I} ]^{1/(1−ε_I)}
- Capital production and adjustment costs:
  - (26) K_t = (1−δ) K_t' + I_t − (σ_I / 2) [ (I_t / K_t') − δ ]^2 K_t'
- Capital producers buy effective capital K_t' at real price q_t' and sell K_t at q_t, yield FOCs:
  - (27) q_t = p_t^I [ 1 − σ_I ((I_t / K_t') − δ) ]^{−1}
  - (28) q_t' = [ 1 − δ + σ_I (I_t / K_t') ] (I_t / K_t')^{−1} q_t

### II.4 Entrepreneurs and Financial Intermediaries
- Continuum of risk-neutral entrepreneurs e ∈ [0,1]:
  - Financing constraint prior to period t: P_t q_t K_t(e) = N_t(e) + B_t(e).
  - Idiosyncratic shock ω_{t+1}(e) ∈ [0,∞], unit mean, affects conversion K_t' = ω_{t+1} K_t.
  - Gross nominal payoff: R_{t+1}^K P_t q_t ω_{t+1}(e) K_t(e), where R_{t+1}^K ≡ Π_{t,t+1} r_{t+1}^k + q_{t+1}' / q_t.
- Costly state verification: monitoring cost proportional to realized payoff, time-varying:
  - (29) log(μ_t) = log(μ̄) + ε_t^μ, μ ∈ [0,1]
- Optimal contract: state-contingent nominal interest rates Z_{t+1}(e); cutoff ω̄_{t+1}(e) for default such that:
  - R_{t+1}^K P_t q_t ω̄_{t+1}(e) K_t(e) = Z_{t+1}(e) B_t(e)
- Group entrepreneurs by net worth N; aggregated financing constraint and conditions:
  - (30) P_t q_t K_{t,N} = N_N + B_{t,N}
  - (31) R_{t+1}^K P_t q_t ω̄_{t+1,N} K_{t,N} = Z_{t,N} B_{t,N}
- Zero profit condition (ZPC) for intermediaries holds in each aggregate state and for each net worth group:
  - (32) [ Γ(ω̄_{t+1,N}) − μ_t G(ω̄_{t+1,N}) ] R_{t+1}^K R_t = k_{t,N}^{−1} k_{t,N}
    - where k_{t,N} ≡ P_t q_t K_{t,N} / N_N, Γ(ω̄) ≡ [1−T(ω̄)] ω̄ + G(ω̄) ∈ [0,1], G(ω̄) ≡ ∫_0^{ω̄} ω dF(ω)
- Aggregated relationships:
  - (34) B_t = N_t (k_t − 1)
  - (35) Z_{t+1} = R_{t+1}^K P_t q_t ω̄_{t+1} K_t / B_t
  - (36) k_t = P_t q_t K_t / N_t
  - (37) [ Γ(ω̄_{t+1}) − μ_t G(ω̄_{t+1}) ] R_{t+1}^K R_t = k_t^{−1} k_t
  - (38) First order condition determining k_t (integrated FOC)
  - (39) V_{t+1} = [1 − Γ(ω̄_{t+1})] R_{t+1}^K k_t N_t (entrepreneur payoff)
  - Effective capital supplied by entrepreneurs:
    - (40) K_{t+1}' = N_t / (P_t q_t) [ 1 − R_{t+1}^K R_t Γ(ω̄_{t+1}) − μ_t G(ω̄_{t+1}) ]

- Large family setup:
  - Aggregate net worth evolution:
    - (41) N_t = P_t T_{h,e} + γ V_{t−1}, with entrepreneur transfers T_{h,e} and retention fraction γ.

- Real resource cost of monitoring (paid in period t):
  - (42) M_t ≡ μ_t G(ω̄_t) R_t^K q_{t−1}^{−1} K_{t−1} Π_{t−1,t}

### II.5 Final Good Firms
- Continuum of monopolistically competitive firms j ∈ [0,1] with Cobb-Douglas production:
  - (43) Y_H,t(j) = A_t K_t'(j)^a H_t(j)^{1−a}, a = capital share, log(A_t) = log(Ā) + ε_t^A
- Factor market FOCs:
  - (44) MC_{t}^n(j) = P_t w_t (1−a) [ Y_H,t(j) / H_t(j) ]^{−1}
  - (45) K_t'(j) / H_t(j) = a/(1−a) w_t / r_t^k
- Aggregate real marginal cost:
  - (46) MC_t ≡ MC_t^n(j) P_H,t = S_t Θ_t w_t (1−a) [ Y_t^H ]^{−1}
- Calvo pricing with fraction θ ∈ (0,1) unable to re-optimize; optimal re-optimizing price p_H,t^{opt}(j) satisfies recursive conditions:
  - (47) Firm maximization (forward-looking expected discounted profits) subject to demand constraint (48).
  - (49)-(51) Recursive definitions for optimal real price p_H,t^{opt}, Ω_t, and T_t.
- Domestic product inflation evolution:
  - (52) (1−θ) p_H,t^{opt}^{1−ε_ρ} = 1−θ [ Π_H,t−1,t ]^{1−ε_ρ}

### II.6 Aggregate Demand and Aggregate Supply
- Goods market clearing (firm-level):
  - (53) Y_H,t(j) = C_H,t(j) + I_H,t(j) + M_H,t(j) + G_H,t(j) + ∫_0^1 [ C_H,t^i(j) + I_H,t^i(j) + M_H,t^i(j) + G_H,t^i(j) ] di
- Aggregate demand (after algebra):
  - (54) Y_t = (1−α) (S_t Θ_t)^{ε_H} (C_t + G_t + M_t) + (1−α_I) (S_t Θ_t)^{ε_I} (p_t^I)^{ε_I} I_t + S_t^{ε_F} W_t^*
    - where W_t^* = α (C^* + G^* + M^*) + α_I I^*
- Aggregate supply (production aggregation):
  - (55) Y_t = A_t K_t'^{a} H̄^{1−a} Δ_t
  - Price dispersion recursion:
    - (56) Δ_t = (1−θ) p_H,t^{opt}^{−ε_ρ} + θ [ Π_H,t−1,t / Π̄ ]^{ε_ρ} Δ_{t−1}

### II.7 Trade Balance
- Net exports (nominal):
  - (57) P_H,t NX_t = P_H,t Y_t − P_t (C_t + M_t + G_t) − P_t^I I_t
- Real terms:
  - (58) NX_t = Y_t − S_t Θ_t (C_t + M_t + G_t − p_t^I I_t)
- Balance of payments identity and real expression:
  - (59) NX_t = S_t ( b_t^F + ℛ_t − R_{t−1}^F Π^* ( b_{t−1}^F + ℛ_{t−1} ) )
    - where b_t^F ≡ B_t^F / P_t^*
- Debt-elastic interest rate spread (Schmitt-Grohe and Uribe, 2003):
  - (60) R_t^F − R^* = −χ ( ε B_t^F / (4 P_H,t Y_t) )
    - χ > 0 small so B_t^F reverts to steady-state zero.

### II.8 Monetary Policy
- Policy interest rate (Taylor rule with smoothing):
  - (61) log(R_t) = log(R̄) + φ_i log(R_{t−1} / R̄) + (1−φ_i) [ φ_π log( Π_{t−1,t} / Π̄ ) + φ_y log( Y_t / Ȳ ) ]
    - φ_i ∈ [0,1], φ_π > 1, φ_y ≥ 0
- Exchange rate flexibility with foreign exchange interventions; reserves rule:
  - (62) log( ℛ_t / ℛ_{t−1} ) = υ_ℛ log( ℛ̄ / ℛ_{t−1} ) + υ_ε log( ε̄ / ε_t )
    - υ_ℛ > 0, υ_ε > 0

### II.9 Fiscal Policy
- Real government spending G constant; consolidated government budget constraint (real):
  - (63) b_t^G = G + (R_t^G / Π_{t−1,t}) b_{t−1}^G − T_t  ∀ t
    - b_t^G ≡ B_t^G / P_t
- Fiscal rule for lump-sum real tax T_t:
  - (64) T_t = τ̄ + τ_y ( Y_t − Ȳ ) + τ_b ( b_{t−1}^G − b̄_g )
    - τ_b > 0 sufficiently high for solvency; τ_y ∈ [0,1]

### II.10 Rational Expectations Equilibrium
- Equilibrium is a collection of endogenous processes for the full set of variables listed in (65) that satisfy the model's equilibrium conditions given exogenous values {G, H̄, Π^*, C^*, G^*, M^*, I^*} and exogenous shock processes { ε_t^A, ε_t^μ }.

### II.11 Calibration
- Calibration targets characteristics of Ukraine and emerging markets. Period = quarter.
- Household and preference parameters:
  - σ = 1
  - H̄ = 0.2934
  - β = 0.9975  (annualized steady state real interest rate = 1%)
  - Steady state inflation Π̄ calibrated to annual inflation rate of 5% (National Bank of Ukraine target) → nominal interest rate of 6% in annual terms.
  - ε_ℛ = 0.83 (liquidity premium responsiveness to reserves, per Adler, Lisack and Rui (2015))
- Production parameters:
  - a = 0.3 (capital income share)
- Steady-state normalization:
  - Technology parameter A calibrated to normalize steady-state output to Ȳ = [value truncated in source]

*Source: Appendix II. Description of the small open economy New Keynesian DSGE model with financial frictions (wpiea2021100-print-pdf - Appendix II).*

### 1. We set the price elasticity of demand among domestic goods to 휀휀

### 1. We set the price elasticity of demand among domestic goods to 휀휀

### Calibration: price, Calvo, capital producers, and adjustment costs
- Price elasticity of demand among domestic goods: 휀휀P = 7.6, implying a mark-up of 15% over marginal costs.
- Calvo (1983) parameter: 휃휃 = 0.75 (consistent with an average period of one year between price adjustments).
- Capital producers:
  - Quarterly depreciation rate: 훿훿 = 0.05.
  - Investment adjustment cost parameter: 휎휎I = 4.602 (from Fernandez and Gulan (2015)).

### Open-economy and trade elasticities; debt elasticity
- Consumption and investment trade elasticities and related parameters:
  - 휀휀F = 휀휀H = 휀휀I = 1.2.
  - 훼훼 = 훼훼I = 0.36.
- Trade elasticity calibration note: (휀휀F, 휀휀H, 휀휀I) calibrated at the lower end of literature ranges to compensate for lack of trade flow frictions.
- Debt-elasticity parameter: 휒휒 = 10−7 (negligible impact over the simulation horizon).

### Financial accelerator: idiosyncratic shock distribution and analytical expressions
- Idiosyncratic shock assumed log-normal: ln(휔휔) ~ 푁푁(−휎휎푒푒푒푒푡푡2/2, 휎휎푒푒푛푛푡푡2).
- Analytical expressions for G(휔휔�푡푡+1), Γ(휔휔�푡푡+1), and derivatives:
  - G(휔휔�푡푡+1) = Φ( (log(휔휔�푡푡+1) / 휎휎푒푒푛푛푡푡) − 휎휎푒푒푛푛푡푡2 )
    - (Equation (66) in source)
  - Γ(휔휔�푡푡+1) = Φ( (log(휔휔�푡푡+1) / 휎휎푒푒푛푛푡푡) − 휎휎푒푒푛푛푡푡2 ) + 휔휔�푡푡+1 (1 − Φ( (log(휔휔�푡푡+1) / 휎휎푒푒푛푛푡푡) + 휎휎푒푒푛푛푡푡2 ))
    - (Equation (67) in source)
  - Γ′(휔휔�푡푡+1) = 1 − Φ( (log(휔휔�푡푡+1) / 휎휎푒푒푛푛푡푡) + 휎휎푒푒푛푛푡푡2 )
    - (Equation (68) in source)
  - G′(휔휔�푡푡+1) = (1 / 휎휎푒푒푛푛푡푡) · 휙휙( (log(휔휔�푡푡+1) / 휎휎푒푒푛푛푡푡) + 휎휎푒푒푛푛푡푡2 )
    - (Equation (69) in source)
- Φ(.) and 휙휙(.) denote the standard normal cdf and pdf respectively.
- Monitoring/financial-frictions calibration:
  - TTh,e set to 1% of steady-state income.
  - Remaining parameters: 훾훾 = 0.915, 휇휇 = 0.324, 휎휎entt = 0.125 (estimates from Fernandez and Gulan (2015) for emerging market economies).

### Government policy and policy-rule calibrations
- Monetary policy coefficients:
  - 휙휙π = 1.5.
  - 휙휙y = 0.56.
  - 휙휙i = 0.99 (high smoothing parameter justified by large shocks associated with structural reforms).
- Reserve policy rule:
  - Exchange rate responsiveness parameter: 휐휐ε = 10.
  - Target steady-state level of reserves: ℛ� = 0.18 (targets USD 20bn under 2019 exchange rates).
  - Stabilization parameter for reserves 휐휐ℛ set to an extremely small value (negligible impact over the simulation horizon).
- Fiscal policy:
  - Pro-cyclicality parameter of the tax rule: 휏휏y = 0.47 (regional estimates from Girouard and Andre (2005)).
  - Steady-state lump-sum tax parameter 휏휏̅ calibrated to sustain:
    - Government spending share of output: G/Y� = 0.32.
    - Public debt-to-GDP ratio: b�g4Y� = 0.60.
  - Targets reflect average values for Ukraine over recent years.

### Innovation processes and simulated reform transition paths
- Innovation processes calibrated for gradual but permanent shocks reflecting structural reforms: (휀휀tA, 휀휀tμ).
- For each reform scenario:
  - Simulate transition path {휀휀tA}t=1 80 which gradually adjusts At to a long-term value consistent with cumulative impact of land reform and other structural reforms per estimates in Table 4 and 5 for that reform scenario.
  - Simulated transition path {휀휀tμ}t=1 80 gradually reduces monitoring costs (relaxes financial frictions) proportionate to the collateral impact of the corresponding land reform scenario.
- Collateral-impact note: collateral impact under the full reform scenario is greater than under the partial reform scenario, consistent with differential land price estimates from EasyBusiness (2019).

*Source: wpiea2021100-print-pdf - 1. We set the price elasticity of demand among domestic goods to 휀휀*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2021/english/wpiea2021100-print-pdf.pdf_