## cr18199

## Source details

**Canonical URL:** [cr18199](https://www.imf.org/-/media/files/publications/cr/2018/cr18199.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2018/cr18199.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2018/cr18199.pdf.json)

---

### The IMF’s Global Integrated Monetary and Fiscal (GIMF) Model — calibration and policy package
- Model calibration and baseline assumptions:
  - Calibrated to Georgian stylized facts: structure of GDP, labor and capital income shares, government expenditure and revenues, trade structure, and April 2018 WEO components of GDP.
  - Fiscal block: debt-to-GDP anchored at 45 percent in the long run; long-run output elasticity of public investment = 0.25.
  - Monetary policy: inflation target = 3 percent.
  - Dynamics follow standard EME parameters (price/wage rigidities, investment adjustment costs).
  - Structural reforms affect economy-wide productivity; quantitative effects calibrated from macro-structural literature.
- Fiscal package quantitative assumptions:
  - Capital spending increases progressively by a maximum of 2.3 percentage points of GDP and then progressively declines.
  - Capital investments in steady state stabilize at around 5.3 percent of GDP (about 0.4 percent of GDP higher than under the baseline).
  - Current spending contracts by about 2.5 percent of GDP permanently.
  - Distributed-dividend taxation produces a permanent loss of revenue of around 1 percentage point of GDP.
  - Excise increases boost revenues by around 0.5 percentage point of GDP.
  - Baseline scenario updated from 2016 fiscal projections with latest GDP growth and lari/dollar exchange rate assumptions.
- Improvement in government efficiency (assumptions):
  - Georgia’s average distance to the frontier on governance/business environment indicators ≈ 10 to 20 percent.
  - Assumed reforms cut half the distance to the frontier over five to 10 years → 8-percent improvement in the governance indicator → gradual increase in TFP of 1 percent over a 10-year period.
- Education reform (assumptions and empirical approach):
  - Education spending would converge to ½ of the average OECD level (about 12 percent of GDP), implying a permanent increase in spending of about 2.5 percent of GDP.
  - Government covers 50 percent of the increase by higher consumption taxes; remaining 50 percent initially covered by higher borrowing then by progressively lowering capital spending to steady state.
  - Education reform estimated to increase TFP growth by about 0.5 percentage point over the medium run.
  - Empirical approach: cross-section regression (2000–2016) linking TFP growth to investment-to-GDP, CPI inflation, TFP distance to frontier, years of schooling, and a PISA-based quality index; preferred specification includes interaction (PISA Index)*(Years of Schooling).
  - With around 12 years of schooling in Georgia, closing half the gap to the frontier in education quality → productivity growth ↑ by about 0.5 percentage point.

### Selected calibration excerpts (reported figures)
- BASELINE SCENARIO (From 2016 Art IV):
  - Capital Spending: 5.4 (2017), 5.7 (2018), 5.5 (2019), 5.4 (2020), 5.4 (2021), 5.4 (2022), 5.4 (2023), Steady State 5.4
  - Current Spending: 22.9 (2017), 22.4 (2018), 22.1 (2019), 22.1 (2020), 22.0 (2021), 22.0 (2022), 22.0 (2023)
  - CIT Revenue Corporates: 2.8 (2017) … 2.9 (steady state)
  - Excises Revenue: 3.1 (2017) … 3.0 (steady state)
- FISCAL PACKAGE (Recent Framework):
  - Capital Spending: 6.1 (2017), 6.8 (2018), 7.0 (2019), 7.7 (2020), 7.9 (2021), 7.7 (2022), 7.6 (2023)
  - Current Spending: 22.9 (2017), 21.8 (2018), 21.2 (2019), 20.4 (2020), 19.4 (2021), 19.4 (2022), 19.4 (2023)
  - CIT Revenue Corporates: 2.0 (2017), 1.9 (2018–2023), Steady State 1.9
  - Excises Revenue: 3.8 (2017), 3.5 (2018), 3.5 (2019), 3.5 (2020), 3.4 (2021), 3.4 (2022), 3.3 (2023)
- DIFFERENCE (FISCAL PACKAGE minus BASELINE):
  - Capital Spending difference: 0.6 (2017), 1.1 (2018), 1.5 (2019), 2.3 (2020), 2.5 (2021), 2.3 (2022), 2.2 (2023), 0.3 (Steady State)
  - Current Spending difference: 0.1 (2017), -0.6 (2018), -0.9 (2019), -1.6 (2020), -2.6 (2021), -2.6 (2022), -2.6 (2023), -2.5 (Steady State)
  - CIT Revenue Corporates difference: -0.8 (2017), -0.9 (2018), -1.0 (2019), -1.0 (2020), -1.0 (2021), -1.0 (2022), -1.0 (2023), -1.0 (Steady State)
  - Excises Revenue difference: 0.7 (2017), 0.5 (2018), 0.6 (2019), 0.5 (2020), 0.5 (2021), 0.4 (2022), 0.4 (2023), 0.5 (Steady State)

### GIMF model structure and methodology
- Multi-country structural DSGE with Georgia, euro area, emerging Asia, United States, and rest of world.
- Features: households (OLG and LIQ), firms producing tradable/non-tradable goods, government (consumption, productive public investment, taxes, transfers) with fiscal rule stabilizing debt-to-GDP, monetary policy under inflation-forecast-targeting (Taylor-type rule).
- Fiscal and structural measures can change level of real output per capita but not exogenous long-term growth.

### Results from simulations — aggregate impacts
- Fiscal policy package effects:
  - Scaling up productive public investments and public capital stock raises productivity in private firms.
  - Replacing taxes on capital with taxes on consumption reduces financing distortions.
  - Real GDP increases by 3 percent in the new steady state (relative to baseline).
  - GDP growth increases by about 0.4 percentage point in the medium term during convergence to the new steady state.
  - In the long run, growth remains driven by productivity and population growth.
- Full reform package (fiscal rebalancing toward public investment, government efficiency improvements, and education reform):
  - Long-run real GDP about 5 percent higher than the baseline (paper’s summary).
  - Annual growth about 0.7 percentage points higher over the medium term along the path to the new equilibrium (paper’s summary).
  - Caveat: effects reported are an “upper bound”; rely on assumptions that public spending is employed productively and fiscal sustainability is preserved.

### Private sector dynamics, public finances, and external accounts
- Private investment and finance:
  - Private investment increases steadily in levels and as a share of GDP.
  - Investment share of output in the new steady state ≈ 1 percentage point of GDP higher than baseline.
  - External finance premium temporarily decreases (financial accelerator) leading to temporary decrease in cost of borrowed funds.
- Private consumption:
  - Reacts gradually; initially muted due to higher consumption taxes and need to increase savings to finance higher private investment.
  - Consumption initially declines as a share of GDP, then progressively increases by about 1.5 percent above baseline in the new steady state.
- Public finances and debt dynamics:
  - Higher consumption taxes do not fully cover higher government spending and revenue losses from lower capital taxes → public debt initially increases.
  - Public debt-to-GDP ratio increases by about 3 percentage points vs baseline, then progressively declines back to 45 percent of GDP in the new steady state (government assumed unwilling to raise debt-to-GDP).
  - With government investment declining to near-baseline levels and contained current spending, fiscal balance deteriorates only mildly in the medium run.
  - When education spending and higher taxes are added:
    - Private consumption may contract by up to 1 percent of GDP.
    - Overall GDP in new steady state ≈ 1 percent higher than with the simple fiscal package.
    - Public debt about 2 percent of GDP higher than with simple fiscal package, peaking at around 50 percent before declining to 45 percent.
- Exchange rate, interest rates, current account:
  - Increase in real wages → upward pressure on domestic prices → central bank moderately increases policy rate; interest rate increase consistent with initial appreciation of the lari.
  - Appreciation of the lari and higher import demand temporarily deteriorates the current account — worsens by a maximum of 1.5 percent of GDP.
  - As productivity rises and domestic prices fall, the real effective exchange rate depreciates and current account returns to baseline.

### Education reform quantitative evidence (regression findings)
- Preferred regression specification (column 4) results:
  - Investment to GDP coefficient: 0.013
  - CPI Inflation coefficient: -0.000
  - TFP - Distance to the Frontier coefficient: 0.053***
  - (PISA Index)*(Years of Schooling) coefficient: 0.189**
  - Observations: 60
  - R-squared: 0.909
  - Significance: *** p<0.01, ** p<0.05, * p<0.1
- Marginal effects:
  - With 12 years of schooling, improving education quality from the lowest to the highest score → about 1.5 percentage points higher productivity growth (marginal effect from Figure 1).
  - Closing half the gap to the frontier in education quality → about 0.5 percentage point increase in productivity growth over the medium run.
- Thought-experiment and long-run range:
  - Education reform effect on output could translate into a level of output between 5 and 10 percent higher than under the baseline (subject to uncertainty and gradual human capital accumulation).

### Labor market, education, and inclusion — headline findings
- Growth and jobs:
  - GDP growth averaged 4.4 percent over 2006–16 while employment barely increased; around 99,000 net jobs created in 2006-16.
  - Unemployment = 14 percent; youth unemployment = 33.2 percent.
  - More than 40 percent of jobs remain in agriculture (low-productivity sector).
- Poverty and inequality:
  - One in every five Georgians is still poor.
  - Vulnerable to poverty share hovered between 47 and 54 percent during 2009–16.
  - Poverty headcount = 17.3 percent (global poverty line for lower-middle-income countries).
  - Consumption Gini coefficient = 36.5% in 2016.
- Skills mismatch and education outcomes:
  - University enrolment ≈ 50 percent of Georgian students (versus 60 percent in Central and Eastern Europe and 70 percent in Western Europe).
  - Vocational training is not popular.
  - PISA (2015): Georgia ranked 60th in mathematics, 63rd in science, 65th in reading out of 72 participating countries.
  - TIMSS results and STEP survey highlight mismatch between skills supplied and business demand; one in five firms cites lack of digital skills as a major/severe problem.
  - Over 50 percent of unemployed have secondary school diploma; 40 percent have higher education degree; in urban areas 46 percent of unemployed have higher education.
- Labor force composition and demographics:
  - Population expected to decline and age; share 65+ expected to increase significantly.
  - Female participation ≈ 20 percentage points below male participation but above peers for some measures.
  - Basic pension = GEL180 per month (marginally higher than the subsistence minimum).
- Active labor market programs and matching:
  - Government job-matching service: 140,000 job seekers registered (≈ 7 percent of labor force, ≈ 50 percent of unemployed).
  - Apprenticeship program and wage subsidies: around 1800 workers have benefited.
- Policy priorities to boost inclusive job creation:
  - Improve early childhood education, quality of general/basic/secondary education, strengthen VET and higher education, expand on-the-job training and lifelong learning.
  - Target vocational training to high-demand sectors; improve career guidance and labor market information systems.
  - Revamp insolvency framework to improve resource reallocation and business rehabilitation.
  - Encourage female labor force participation via childcare and flexible work arrangements.
  - Strengthen science & technology, PPPs in education, and incentives for firm-provided training.

### Exports, diversification, and product space diagnostics
- External balances and trade:
  - Current account deficit ≈ 12 percent of GDP over the past 15 years.
  - Trade deficit ≈ 25 percent of GDP on average.
  - Exports-to-GDP ≈ 22 percent of GDP during the past decade — low relative to peers.
- Export concentration and diversification:
  - Almost 60 percent of merchandise exports concentrated in a few products (ferrous products, copper, beverages, medicines).
  - Export market diversification improved: EU, China, and EFTA share increased from 22 percent to 34 percent during 2013–17.
  - Export product diversification remains low (Herfindahl product concentration and Theil’s index).
- Product space and upgrading prospects:
  - 1995: comparative advantage mostly on periphery of product space; 2016: expanded into middle/dense parts (cars, chemicals, plastics).
  - Some industries (vehicles) subject to high re-export content — caveat on true competitiveness.
  - Trade-off: products with high opportunity gain often at larger distance (harder to reach).
- Drivers of diversification (panel regression 2000–14; dependent variable = Theil’s Index):
  - Log GDP per capita: -0.156***; -0.083**; -0.080**  [robust s.e.: 0.038; 0.037; 0.037]
  - Human Capital Index: -0.178***; -0.143***; -0.147***  [robust s.e.: 0.049; 0.045; 0.044]
  - Terms of Trade (Growth): 0.008***; 0.006***; 0.006***  [robust s.e.: 0.002; 0.001; 0.001]
  - Population (Billions): -0.474***; -0.388***; -0.377***  [robust s.e.: 0.061; 0.041; 0.045]
  - Rule of Law (-2.5 to 2.5 = best): -0.231***; -0.234***  [robust s.e.: 0.048; 0.049]
  - Trade Openness: 0.026  [robust s.e.: 0.047] (insignificant)
  - Observations: 192; R-squared: 0.954; 0.962; 0.962
- Policy implications for diversification:
  - Prioritize structural reforms that expand human capital and improve institutions.
  - Choose diversification targets that are similar/closely linked to existing products, connected to many other products, and provide more economic complexity.

### Financial conditions, natural interest rate, and monetary stance
- Financial Conditions Index (FCI) approach:
  - VAR-based and factor-based FCIs constructed (sample period: 2003Q1–2017Q4).
  - Final variables: deposit growth, REER, loan spread (domestic); VIX, EUR-OIS spread (external).
  - Historical drivers: exchange rate most important; global factors (VIX, EUR-OIS) more important after GFC.
- Key correlations (Text Table 1; N = 60):
  - FCI correlation with Real GDP Growth: t = 0.544***, t+1 = 0.644***, t+2 = 0.629***, t+3 = 0.608***, t+4 = 0.533***
  - Deposit growth: t = 0.787***, t+1 = 0.763***, t+2 = 0.607***, t+3 = 0.493***, t+4 = 0.187
  - REER: t = -0.484***, t+1 = -0.553***, t+2 = -0.547***, t+3 = -0.505***, t+4 = -0.451***
  - VIX and EUR-OIS show negative correlations at various horizons.
- Recent assessment:
  - Financial conditions assessed as slightly loose (loose global monetary/liquidity conditions + cyclical recovery + stable exchange rates).
  - FCI has predictive power: adding FCI to AR GDP model increases R-squared; can explain up to 30 percent of variation in GDP growth not explained by lagged GDP.
- Natural interest rate estimates:
  - TVP-VAR approach (2013Q4 – 2018Q1) → Georgia’s real natural interest rate ≈ 2 percent and largely stable.
  - Using survey-based inflation expectations, Georgia’s real ex ante interest rate = 1.6 percent → lower than natural rate → slightly loose monetary policy.
  - Uncertainty: large standard errors and measurement challenges for inflation expectations.
- Policy message:
  - Monetary stance broadly adequate given lack of price pressures (2.5 percent inflation in April 2018), but National Bank of Georgia should continue close monitoring of inflation and credit dynamics given rapid credit growth.

*Source: IMF staff analysis and GIMF simulations as presented in cr18199.*

### 1. The IMF’s Global Integrated Monetary and Fiscal (GIMF) Model________________________11

### 1. The IMF’s Global Integrated Monetary and Fiscal (GIMF) Model

### Background
- Georgia’s average productivity growth (TFP) decreased from around 5 percent in 2004-08 to 1.5 percent in 2011-15.
- Georgia reduced the distance to the frontier in Doing Business Indicators to about 20 percent in 2017.
- Key opportunities to raise growth:
  - Broaden the production and export base.
  - Address high unemployment and employment concentration in low-productivity sectors.
  - Further improve the business environment and human capital quality and stock.
- Education and skill matching are critical:
  - University enrolment is about 50 percent of Georgian students (versus 60 percent in Central and Eastern Europe and 70 percent in Western Europe).
  - Vocational training is not popular, contributing to skill mismatch.
- The government’s 4-Point Plan comprises:
  1. Improvement in tax administration and the tax system.
  2. Increase in infrastructure investments.
  3. Improvement in government efficiency.
  4. Education reform (curriculum standards, new teacher framework, vocational training, adult learning).
- Planned government actions include:
  - Set up a Business House (one-stop shop for public services to enterprises).
  - Introduce International Financial Reporting Standards for corporations.
  - Reform insolvency law.
  - Scale-up infrastructure spending to transform Georgia into a transport and logistics hub.

### Modeling the Policy Package (GIMF calibration and assumptions)
- The IMF’s GIMF model is calibrated to Georgian stylized facts: structure of GDP, labor and capital income shares, government expenditure and revenues, trade structure, and April 2018 WEO components of GDP.
- Fiscal block calibration:
  - Debt-to-GDP anchored at 45 percent in the long run.
  - Long-run output elasticity of public investment calibrated to 0.25.
- Monetary policy calibration:
  - Inflation target calibrated to 3 percent (per the National Bank of Georgia “Main Directions of Monetary Policy”).
- Dynamics calibrated following standard emerging market economy (EME) parameters (price/wage rigidities, investment adjustment costs).
- Structural reforms assumed to affect economy-wide productivity; quantitative effects calibrated from macro-structural literature.

- The policy package components and quantitative assumptions:
  - Fiscal package:
    - Capital spending increases progressively by a maximum of 2.3 percentage points of GDP and then progressively declines.
    - Capital investments in steady state stabilize at around 5.3 percent of GDP (about 0.4 percent of GDP higher than under the baseline).
    - Current spending contracts by about 2.5 percent of GDP permanently (cuts mainly from wage bill reductions, administrative expenditure cuts, and health-efficiency gains).
    - Distributed-dividend taxation produces a permanent loss of revenue of around 1 percentage point of GDP.
    - Excise increases boost revenues by around 0.5 percentage point of GDP.
    - Baseline scenario constructed by updating 2016 fiscal projections with latest GDP growth and lari/dollar exchange rate assumptions.
  - Improvement in government efficiency:
    - Georgia’s average distance to the frontier on governance/business environment indicators is about 10 to 20 percent.
    - Assumption: reforms cut half the distance to the frontier over five to 10 years → 8-percent improvement in the governance indicator.
    - That 8-percent improvement translates into a gradual increase in TFP of 1 percent over a 10-year period.
  - Education reform:
    - Spending for education as a percentage of GDP would converge to ½ of the average level of OECD countries (about 12 percent of GDP), implying a permanent increase in spending of about 2.5 percent of GDP.
    - Government will cover 50 percent of the increase by higher consumption taxes; remaining 50 percent initially covered by higher borrowing and then by progressively lowering capital spending to its steady state level.
    - Education reform estimated to increase TFP growth by about 0.5 percentage point over the medium run.
    - Empirical approach: cross-section regression (2000–2016) linking TFP growth to investment-to-GDP, CPI inflation, TFP distance to frontier, years of schooling, and a PISA-based quality index; preferred specification allows for non-linear interaction (PISA Index)*(Years of Schooling).
    - With around 12 years of schooling in Georgia, closing half the gap to the frontier in education quality would raise productivity growth by about 0.5 percentage point.

- Table 1 (calibration excerpts presented in the source) — selected figures as reported:
  - BASELINE SCENARIO (From 2016 Art IV):
    - Capital Spending: 5.4 (2017), 5.7 (2018), 5.5 (2019), 5.4 (2020), 5.4 (2021), 5.4 (2022), 5.4 (2023), Steady State 5.4
    - Current Spending: 22.9 (2017), 22.4 (2018), 22.1 (2019), 22.1 (2020), 22.0 (2021), 22.0 (2022), 22.0 (2023)
    - CIT Revenue Corporates: 2.8 (2017) ... 2.9 (steady state)
    - Excises Revenue: 3.1 (2017) ... 3.0 (steady state)
  - FISCAL PACKAGE (Recent Framework):
    - Capital Spending: 6.1 (2017), 6.8 (2018), 7.0 (2019), 7.7 (2020), 7.9 (2021), 7.7 (2022), 7.6 (2023)
    - Current Spending: 22.9 (2017), 21.8 (2018), 21.2 (2019), 20.4 (2020), 19.4 (2021), 19.4 (2022), 19.4 (2023)
    - CIT Revenue Corporates: 2.0 (2017), 1.9 (2018–2023), Steady State 1.9
    - Excises Revenue: 3.8 (2017), 3.5 (2018), 3.5 (2019), 3.5 (2020), 3.4 (2021), 3.4 (2022), 3.3 (2023)
  - DIFFERENCE (FISCAL PACKAGE minus BASELINE):
    - Capital Spending difference: 0.6 (2017), 1.1 (2018), 1.5 (2019), 2.3 (2020), 2.5 (2021), 2.3 (2022), 2.2 (2023), 0.3 (Steady State)
    - Current Spending difference: 0.1 (2017), -0.6 (2018), -0.9 (2019), -1.6 (2020), -2.6 (2021), -2.6 (2022), -2.6 (2023), -2.5 (Steady State)
    - CIT Revenue Corporates difference: -0.8 (2017), -0.9 (2018), -1.0 (2019), -1.0 (2020), -1.0 (2021), -1.0 (2022), -1.0 (2023), -1.0 (Steady State)
    - Excises Revenue difference: 0.7 (2017), 0.5 (2018), 0.6 (2019), 0.5 (2020), 0.5 (2021), 0.4 (2022), 0.4 (2023), 0.5 (Steady State)

### Results
- Fiscal policy package:
  - Scaling up productive public investments and public capital stock raises productivity in private firms.
  - Replacing taxes on capital with taxes on consumption reduces distortions to financing.
  - Real GDP increases by 3 percent in the new steady state (relative to baseline).
  - GDP growth increases by about 0.4 percentage point in the medium term during convergence to the new steady state.
  - In the long run, growth remains driven by productivity and population growth.

- Education reform impacts (empirical findings summarized):
  - Regression results (preferred specification column 4) show:
    - Investment to GDP coefficient: 0.013
    - CPI Inflation coefficient: -0.000
    - TFP - Distance to the Frontier coefficient: 0.053***
    - (PISA Index)*(Years of Schooling) coefficient: 0.189**
    - Observations: 60
    - R-squared: 0.909
    - Standard errors and significance markers reported as in source (*** p<0.01, ** p<0.05, * p<0.1).
  - With 12 years of schooling, improving education quality from the lowest to the highest score translates into about 1.5 percentage points higher productivity growth (marginal effect from Figure 1).
  - Closing half the gap to the frontier in education quality is estimated to yield about 0.5 percentage point increase in productivity growth over the medium run.

### Policy implications and overall quantitative summary
- The modeled reform package (fiscal rebalancing toward public investment, government efficiency improvements, and education reform) yields:
  - Long-run real GDP about 5 percent higher than the baseline (as stated in the paper’s summary).
  - Annual growth about 0.7 percentage points higher over the medium term along the path to the new equilibrium (as stated in the paper’s summary).
- Timing of effects:
  - Fiscal and efficiency reforms deliver substantial gains within the medium term.
  - Education reform produces sizeable effects primarily in the long run given Georgia’s relatively high years of schooling and the gradual nature of human capital accumulation.

*Source: IMF staff calculations and analysis as presented in the chapter "1. The IMF’s Global Integrated Monetary and Fiscal (GIMF) Model" (cr18199).*

### 10.      The lower tax burden on firms generates higher private investment.

### 10.      The lower tax burden on firms generates higher private investment.

### Private investment
- Private investment increases steadily both in levels and as a share of GDP.
- In the new steady state investment share to output ends up being about 1 percentage point of GDP higher than in the baseline.
- Dynamics are helped temporarily by a moderate decrease in the cost of borrowed funds.
- With increased profitability, the external finance premium temporarily decreases (financial accelerator).

### Private consumption
- Private consumption reacts more gradually to the fiscal package, but increases in the new steady state.
- Higher labor productivity and lower tax distortions increase output and demand for labor, pushing up real wages and supporting higher consumption.
- The initial response of consumption is muted because of higher consumption taxes and because consumption is compressed by the need to increase savings to finance higher private investment.
- As a result, consumption initially declines as a share of GDP, then progressively increases by about 1.5 percent above baseline in the new steady state.

### Public finances and fiscal balance
- Higher consumption taxes do not fully cover higher government spending and the revenue losses from lower capital taxes, so public debt initially goes up.
- The public debt-to-GDP ratio increases by about 3 percentage points with respect to the baseline, and then progressively declines back to 45 percent of GDP in the new steady state—it is assumed that the government does not want to increase its debt-to-GDP ratio.
- As government investment progressively declines to a level only slightly higher than in the baseline and current spending remains contained, the fiscal balance deteriorates only mildly in the medium run.
- Higher consumption taxes—together with a higher tax base—support the fiscal adjustment.

### Exchange rate, interest rates, and current account
- The increase in real wages puts upward pressure on prices of domestically produced goods, and the central bank moderately increases the policy rate.
- The increase in the interest rate is consistent with the initial appreciation of the lari.
- The appreciation of the lari, as well as the import component of demand (also from higher government and private investment), temporarily deteriorates the current account—which worsens by a maximum of 1.5 percent of GDP.
- As higher productivity allows for lower prices, the real effective exchange rate depreciates, and the current account returns to its baseline level.

### Improving government efficiency
- The improvement in government efficiency increases GDP permanently and growth temporarily.
- Because of the reforms, TFP increases progressively by 1 percent, translating into higher GDP in the new steady state, and temporarily higher growth on the path to the new equilibrium.
- The increase in productivity increases the marginal product of the factors of production, so that demand for capital and labor increases, translating into higher real wages and rental rate on capital.
- Investment responds faster than consumption to the gradual increase in productivity: investment to GDP increases, while the consumption-to-GDP ratio remains in line with the baseline and increases only in the long run.
- The current account worsens temporarily because of higher domestic demand (higher investment and consumption), but in the medium run the depreciation of the exchange rate corrects the current account imbalance; in the new steady state, the current account reverts to a small surplus.

### Fiscal package and education reform
- When the fiscal package is modified to accommodate higher education spending and higher taxes:
  - Effects on GDP are still positive, but private consumption contracts by up to 1 percent of GDP.
  - Overall GDP in the new steady state is about 1 percent higher than with the simple fiscal package analyzed above.
  - Consumption remains at the baseline level initially; as a share of GDP it declines throughout the transition to the new steady state.
  - Because the government finances some of the higher spending through borrowing, public debt is about 2 percent of GDP higher than with the simple fiscal package, peaking at around 50 percent, before declining to its steady state level of 45 percent.
  - Higher demand translates partly into higher imports, thus deteriorating the current account balance.
- The education reform has positive long-run effects, but these materialize only through the build-up of human capital and are inherently uncertain; quantitative analysis of the education reform is outside the DSGE model and is assessed via regression analysis.
- Regression results suggest that an education reform that improves the quality of education to the average level of OECD countries improves the growth rate of TFP by half a percentage point.
- Thought experiment assumptions:
  - Given elevated average years of schooling in Georgia, the effect on productivity starts to materialize after 12 years.
  - The growth rate of TFP then increases progressively by half a percentage point above the baseline (the full effect of the reform) and then progressively declines as Georgia closes the gap with the productivity frontier.
  - Overall, this translates into a level of output between 5 and 10 percent higher than under the baseline.

### Adding reforms to improve government efficiency (all reforms combined)
- When all reforms (fiscal reforms including education spending and improvement in government efficiency) are included simultaneously:
  - Real GDP increases in the long run by about 5 percent compared to the baseline.
  - Growth temporarily increases by about 0.7 percentage point in converging to the new steady state.
  - Private consumption gains mostly accrue in the long run; higher productivity induces more investment and increased real wages.
  - The productivity gains from the education reform could further boost output in the long run between 5 and 10 percent.
- Caveats and key assumptions:
  - Effects reported are to be interpreted as an “upper bound” of the reform package.
  - Results rely on crucial assumptions: that all public spending will be employed productively and that fiscal sustainability will be preserved.

### Model and methodology
- The Global Integrated Monetary and Fiscal (GIMF) model is used to quantify the impacts of fiscal and structural reforms.
- GIMF is a multi-country structural dynamic general equilibrium model featuring Georgia, the euro area, emerging Asia, the United States, and the rest of the world.
- GIMF links households, firms, and the government within and among countries, includes stock-flow budget constraints, and assumes exogenous long-term growth (fiscal or structural measures can change the level of real output per capita but not long-term growth).
- Households: optimizing overlapping-generations (OLG) households and liquidity-constrained (LIQ) households (breaking Ricardian equivalence).
- Firms produce tradable and non-tradable goods; government purchases final goods for consumption and productive public investments, levies taxes (consumption, labor income, capital, lump-sum), provides transfers, and follows a fiscal rule stabilizing the debt-to-GDP ratio.
- Monetary policy operates under an inflation-forecast-targeting framework with a standard Taylor-type rule.

*Source: IMF staff analysis (GIMF simulations) from the chapter "10.      The lower tax burden on firms generates higher private investment."*

### 1. Despite robust economic growth over 2006–16, job creation has been lackluster, and

### 1. Despite robust economic growth over 2006–16, job creation has been lackluster, and

### Overview and headline findings
- GDP growth averaged 4.4 percent over 2006–16 in Georgia, while employment barely increased.
- Between 2006-16, around 99,000 net jobs have been created.
- Unemployment remains high at 14 percent, with youth unemployment at 33.2 percent.
- More than 40 percent of jobs remain in agriculture, a low-productivity sector.
- The evidence points to growth that is insufficiently inclusive and not generating enough high‑productivity jobs to absorb the unemployed.

### Structural bottlenecks and skills mismatches
- The share of employment in higher-productivity and higher-wage sectors (industry and services) has increased, but agricultural employment remains large, indicating insufficient creation of high-productivity jobs.
- Skills mismatches are highlighted by multiple assessments:
  - The 2017–18 World Economic Forum (WEF) ranks workforce skills as the most problematic factor for doing business in Georgia.
  - The World Bank’s 2015 Skills Toward Employment and Productivity (STEP) survey indicates Georgians do not have the skills demanded by businesses.
  - One in five firms reports lack of workers with solid digital skills as a major or severe problem constraining growth.
- Business surveys by the Ministry of Economy and Sustainable Development show main hiring challenges are applicants’ lack of qualifications and experience and salary demands.
- Consequences noted include potential insufficient diversification of exports and production, low firm-level productivity, and need for innovation and R&D to break the low-productivity cycle.

### Poverty, inequality, and inclusiveness (Box 1)
- One in every five Georgians is still poor.
- The share of the population vulnerable to poverty hovered between 47 and 54 percent during 2009–16.
- Poverty headcount in Georgia is 17.3 percent (based on the global poverty line for lower-middle-income countries).
- Georgia’s consumption Gini coefficient reached 36.5% in 2016.
- Large urban-rural and intra-regional disparities: remote and mountainous regions bear the highest burden of poverty.

### Labor market demand, firms, and policy environment
- Job concentration: most jobs are in agriculture, retail and wholesale trade, and public and social services (such as education); new jobs reported in energy, tourism, and trade sectors.
- Doing Business ranking: Georgia is ranked 9th overall (from 112 countries in 2018?).
- Key areas for improvement in business environment: getting electricity, trading across borders, resolving insolvency.
- Georgia’s business insolvency law does not provide room for successful debt restructuring and business rehabilitation; authorities are revamping the law to bring it in line with international best practices.

### Demographics, labor force participation, and composition
- Georgia’s population is expected to continue shrinking and aging, with the share 65 and older expected to increase significantly; decline driven by low birthrates and sustained net emigration.
- Labor force participation:
  - Relatively high overall, especially for older people and women.
  - Female participation is about 20 percentage points below male counterparts but higher than in countries with similar development levels.
- The basic pension in Georgia amounts to GEL180 per month, marginally higher than the subsistence minimum.
- Labor force characteristics:
  - Relatively rural and educated; share of labor force in rural areas slightly higher than peers.
  - High share of educated people in the labor force may reflect social norms rather than labor market requirements.

### Employment dynamics and patterns
- Employment recovery has been slow since mid-2000s shocks; public-sector employment share declined significantly and has been absorbed by private sector.
- Labor market duality:
  - Self-employed are predominantly rural and less likely to access quality training and nonfarm employment.
  - High share of “contributing family workers” in agriculture (typically unpaid).
- Sectoral mobility has followed trends of industrialization and services expansion, but employment remains trapped in agriculture relative to peers.

### Unemployment, youth outcomes, and education mismatch
- High unemployment despite supportive business environment; male unemployment rate is relatively high compared to female.
- Youth outcomes:
  - Approximately one-third of those aged 15–19 are not in education, employment, or training (NEET); over 60 percent of these are women.
  - Among all 25-to-40-year-old non-working individuals who are neither studying nor actively looking for employment, almost 9 out of 10 are women.
  - NEET individuals differ from employed individuals in skill profiles (less proficient in reading, use computer skills less often).
- Education and unemployment paradox:
  - Over 50 percent of all unemployed Georgians have a secondary school diploma.
  - 40 percent of unemployed have a higher education degree.
  - In urban areas, 46 percent of the unemployed have higher education.
  - This suggests oversupply of educated labor and/or an education system not aligned with labor market needs.

### Policy focus and reform priorities identified in the paper
- Improve workforce skills to support more robust and inclusive growth through:
  - Enhancing quality of general, basic, and secondary education.
  - Strengthening vocational education and higher education.
  - Expanding on-the-job training and lifelong learning opportunities.
  - Expanding preschool education to support early cognitive and socio-emotional skill development and to enable greater female labor force participation.
  - Targeting vocational training to high-demand sectors and providing well-targeted courses.
- Address structural constraints to enable job creation in higher-productivity sectors: promote diversification, firm-level productivity improvements, innovation, and investment in research and development.
- Revamp insolvency framework to facilitate reallocation of resources from unproductive to productive firms, preserve jobs through restructuring, and improve access to finance.

_Prepared by Francois Painchaud (IMF), Lire Ersado (World Bank), and Jouko Sarvi (Asian Development Bank)._

### 20. Job creation in Georgia is not hindered by labor market institutions.

### 20. Job creation in Georgia is not hindered by labor market institutions.

### Labor market institutions and wages
- Georgia’s employment protection legislations are relatively unrestrictive and are thus unlikely to have a detrimental impact on aggregate employment or hinder structural changes in the economy.
- There is no unemployment insurance in Georgia, which could in principle lead to high reservation wages and ultimately lower employment and longer unemployment spells.
- Georgia’s statutory minimum wage is one of the lowest in the world, which makes it unlikely to prevent employment, including from typically disadvantaged groups (such as youth, women, and minorities).
- Ribe et al (2006) show that Georgia’s tax wedge is relatively low and thus unlikely to unduly affect employment.
- High public-sector wages and robust remittances could also translate into higher reservation wages.

### Labor matching and active labor market policies
- The government offers a service that lists vacancies by sector/region to facilitate matching labor demand (employers) and supply (employees).
- Registering for this service is mandatory for recipients of social security assistance.
- 140,000 job seekers have registered (roughly 7 percent of the labor force and about 50 percent of unemployed).
- The government regularly holds job fairs where employers can meet job seekers.
- Based on a survey of employers’ needs, the government supports job seekers who are willing to attend training courses provided by licensed facilities.
- The government has established an apprenticeship program and provides wage subsidies for vulnerable segments of the labor force.
- Around 1800 workers have benefitted from the government’s programs.

### Education and skills: current status and challenges
- Two decades of education reforms improved school governance, curriculum design, teaching methodology, grades assessment, and education financing practices; infrastructure upgrades and increased use of IT improved the learning environment.
- A recently adopted teachers’ evaluation via certification system creates ground for preparation, professional development, and career advancement of the teachers through pre-designed scheme, incentivized by salary grades.
- The quality of education and student learning outcomes in Georgia have improved but remain poor.
  - In PISA (2015), Georgia ranked 60th in mathematics, 63rd in science, and 65th in reading out of 72 participating countries.
  - In TIMSS, Georgia ranked 33rd and 35th in 4th grade math and science, respectively, out of 42 participant countries and 25th and 30th in 8th grade math and science out of 38 countries.
- Substantial in-country differences in performance exist by location, wealth, and availability of school resources; children growing up in rural areas perform poorly.
- While Georgians are highly educated, they do not have the relevant skills demanded by businesses:
  - The World Bank’s 2015 STEP survey highlights that businesses demand mostly low and middle technically skilled workers while the majority of job seekers are university graduates.
  - Employers report difficulties filling job vacancies due to a lack of required skills.
- Lack of skills constrains innovation, business expansion, structural transformation (from low- to higher-productivity sectors and from rural to urban areas), and the adoption/invention of new technologies.
- Despite a large share of the workforce having higher education degrees, there is a dearth of skills relevant for the labor market: socio-emotional skills, creative and critical thinking, problem solving, teamwork, and leadership and decision-making.
- The public financing model of higher education based on a fixed grant per student provides no incentives to improve the quality of teaching, learning, and research and development.
- The vocational education and training (VET) system challenges:
  - limited access in rural areas;
  - low overall enrollment rate;
  - low popularity among the Georgian youth;
  - no smooth transition to other higher levels of education;
  - limited private sector involvement, including in defining curriculum, delivering skills training, and sharing costs.
- Employers regularly report dissatisfaction with the supply of skills and unmet demand for job-relevant skills.

### Implications for entrepreneurship
- Lack of skills hinders entrepreneurship.
- The prevalence of entrepreneurship is low as measured by the percentage of total employment.
- Georgian entrepreneurs tend to have lower levels of educational attainment (compared to non-entrepreneurs) and lower income levels and scores in job-relevant, socio-emotional, and cognitive skills than wage earners, suggesting entrepreneurship is not often linked to innovation and higher productivity activities.
- Georgia has recently created the Georgia Innovation and Technology Agency (GITA) to facilitate technology transfer, commercialization of research results, promote innovative entrepreneurship, and act as an incubator and accelerator for startup companies.

### Reform priorities and policy recommendations
- Build the skills of the workforce; the inadequate education and training of the labor force is one of the most significant obstacles for expanding businesses.
- Get children off to the right start:
  - Expand effective participation in early childhood education (ECE) to improve cognitive and socio-emotional skills and later labor market outcomes.
  - Close the participation gap in ECE between urban and rural areas and among socioeconomic and ethnic groups.
  - Define central and local government roles in standards, financing, provision, and monitoring, and tap private sector potential to expand ECE provision.
- Ensure that all students learn:
  - Improve learning outcomes and skills formation by having a high-quality and effective teaching force and a strong quality assurance and accountability system for learning results.
  - Strengthen tertiary education with a close link to labor market needs and consider institutional twinning with foreign universities.
  - Narrow learning outcome and skills gaps, especially across socioeconomic and rural-urban divides.
- Build job-relevant skills:
  - Improve the VET system by greater employer involvement, prioritizing practical training, and supporting labor market-oriented professional trainings.
  - Develop and expand pathways from VET to polytechnic institutions and universities.
  - Use incentives to encourage firms to provide on-the-job training and apprenticeships in collaboration with schools and training institutions.
  - Encourage lifelong skills acquisition to maintain or upgrade workers’ skills with business needs and technological change.
- Strengthen science & technology by establishing centers of excellence in priority economic growth areas and sectors.
- Bolster education sector management and financial planning to ensure reforms are sustainable:
  - Improve resource planning and rationalize networks of schools and education institutions.
  - Undertake public-private partnerships (PPPs) in the education sector, in the context of the forthcoming PPP framework, to establish cost sharing with labor market stakeholders and improve labor market relevance and efficient delivery of education and skills training programs.
- Enhance communications about labor market trends and improve job matching services:
  - Enhance annual reports on labor market trends (for example, list marketable skills and professions).
  - Improve labor market information systems, including career guidance for skilled and low-skilled workers.
  - Enhance the matching services provided by the government to reduce unemployment.
- Encourage greater participation of women in the labor force by enhancing child care services and encouraging more flexible working hours and part-time employment.
- Encourage entrepreneurship and innovation by creating a business environment that supports entrepreneurship, particularly in innovative sectors, improving business-supporting services, and aligning some education and training courses at the tertiary level with entrepreneurship objectives.

### Conclusion: labor market outcomes and remaining challenges
- Georgia’s robust growth in the last decade has not been accompanied by job creation.
- Unemployment has remained high, and a large share of the labor force is trapped in low-productivity sectors, such as agriculture.
- A large share of Georgia’s unemployed have tertiary education, and an alarmingly high level of youth, especially women, remain outside the labor force.

*Source: cr18199 - 20. Job creation in Georgia is not hindered by labor market institutions.*

### 34. More robust and inclusive growth will require reforming the education sector and

### 34. More robust and inclusive growth will require reforming the education sector and strengthening some labor market policies

### Education reform: goals and recommended areas of action
- Education reform would entail:
  1. improving early childhood education;
  2. strengthening the education system to improve learning outcomes;
  3. enhancing the VET system, and supporting on-the-job training and lifelong learning;
  4. encouraging science and technology; and
  5. strengthening education sector management.

- Findings from comparative indicators (World Economic Forum, Global Competitiveness Report 2006–07 and 2017–18):
  - Georgia ranks well in terms of primary education enrollment, but the quality of primary education could improve.
  - Similar findings apply to higher education and on-the-job training (quality can be improved relative to peers).

### Labor market policy improvements: objectives and recommended actions
- Labor market policy improvement should aim to:
  1. provide more information on labor market needs;
  2. enhance labor matching services; and
  3. encourage a greater participation of women in the labor force.

### Demographics and participation
- Population trends:
  - Georgia’s population is expected to decline and to become older.
  - Population axis ticks indicated a range from 1.5 to 2.1 (million) across 2010–2030 in the figures.

- Share of the population by age — selected reported percentages:
  - 2010: <15 = 18.0%, 15–24 = 15.3%, 25–54 = 41.9%, 55–64 = 10.5%, 65+ = 14.2%.
  - 2030: <15 = 18.4%, 15–24 = 13.3%, 25–54 = 37.8%, 55–64 = 12.1%, 65+ = 18.4%.

- Labor force participation:
  - Georgia’s participation rate is relatively high, especially for the older generation.
  - Male participation rate is slightly above peers. Georgia’s female participation is also above peers, but well below male’s.
  - Participation rate axes referenced ticks: 2.75, 3.25, 3.75, 4.25, 4.75, 5.25 (log of real GDP per capita scale used in figures).

### Labor force composition and education attainment
- Age and residence composition:
  - Georgia’s labor force is relatively old and rural.

- Education levels in the labor force:
  - A very low share of the Georgian labor force has received only basic education (or less).
  - A large majority of the labor force has intermediate education.
  - A relatively high share of Georgians has received advanced education.
  - A greater share of women has pursued advanced education.

- Specific reported axes/labels used in figures:
  - Share of labor force: 65+ (percent), Share of labor force: Rural Areas (percent), Share of labor force: Basic Education (percent), Share of labor force: Intermediate Education (percent), Share of labor force: Advanced Education (percent).
  - Share of Male/Female Labor Force with Advanced Education (In percent) plotted.

### Employment patterns and sectoral structure
- Employment trends:
  - Employment has been recovering since 2008–09, while the share of public employment has been trending down.
  - The size of the public sector is now consistent with Georgia’s level of development.

- Employment by residence and status:
  - Employment is more likely to be found in rural areas and for self-employed workers.
  - Self-employed workers mostly live in rural areas, while “hired” workers are more prevalent in urban areas.
  - Self-employed workers are typically older.
  - Georgia has a high share of contributing family workers (typically unpaid and associated with the agriculture sector).

- Sectoral employment shares and recent shifts:
  - Georgia’s share of employment in agriculture remains high despite a marked reduction since 2006, consistent with a broad trend in labor moving out of agriculture.
  - Georgia’s share of employment in industry is lower than peers; it has increased slightly while a majority of countries have experienced a decline.
  - Georgia’s share of employment in services could be increased further after some gains in the last decade.
  - Figures compare 2006 vs 2016 shares and show labor moving out of agriculture and into services in many peer countries.

### Unemployment: levels and composition
- Aggregate and age patterns:
  - Georgia’s unemployment rate remains high despite recent decline.
  - Young Georgians have the highest unemployment rate.
  - Unemployment rate by age (2016) was shown across cohorts 15–19, 20–24, 25–29, 30–34, 35–39, 40–44, 45–49, 50–54, 55–59, 60–64, 65+.

- Youth and NEET:
  - Youth unemployment rate in Georgia is much higher than the world average.
  - Georgia’s share of youth Not in Employment, Education or Training (NEET) is alarmingly high given Georgia’s adverse demographics.

- Education and unemployment:
  - Georgians with advanced education have a relatively high unemployment rate, consistent with an over-supply of advanced education in low and lower middle-income countries.
  - A figure plotted “Unemployment Rate: Advanced Education, Latest Year Available (In percent)” and another showed the “Share of Labor Force and Unemployed with Tertiary Education by Country (In percent)” across income groups, noting that education generally protects against unemployment but Georgia shows relatively high unemployment among the highly educated.

### Implications and summary messages
- Structural reforms that support human capital accumulation and improvements in institutions can help Georgia expand its export potential and support more robust and inclusive growth.
- Reforms should focus on:
  - Raising quality across education levels (including early childhood through tertiary and VET),
  - Strengthening on-the-job training and lifelong learning,
  - Better aligning education outputs with labor market needs through improved information and matching services,
  - Encouraging greater female labor force participation to utilize human capital more fully.

*Source: cr18199 - 34. More robust and inclusive growth will require reforming the education sector and strengthening some labor market policies (IMF).*

### 1.      Relative to its peers

### 1.      Relative to its peers

### Georgia’s current account and trade balance
- Georgia’s current account (CA) deficit has been persistently large.
- The current account deficit hovered around 12 percent of GDP for each of the past 15 years.
- Georgia’s trade deficit has been a major contributor to the CA deficit and has stayed, on average, at about 25 percent of GDP, reflecting both high imports and low export growth.

### Merchandise exports: levels and concentration
- Merchandise exports have been relatively low compared to peer countries.
- Georgia’s exports to GDP—averaging around 22 percent of GDP during the past decade—are among the lowest in its peers.
- Almost 60 percent of total merchandise exports are concentrated on a few products, including ferrous products, copper, beverages, and, recently, medicines.
- Peers are selected based on population size, income, and geographical location and include Albania, Armenia, Belarus, Bosnia and Herzegovina, Bulgaria, Macedonia, FYR, Moldova, Serbia, and Ukraine.

### Export market diversification
- According to the Herfindahl market concentration index, Georgia’s export markets are relatively diversified.
- The share of Georgia’s exports to the European Union, China, and EFTA increased from 22 percent to 34 percent during 2013–17, thanks to free trade agreements with those markets.

### Export product diversification (performance and indices)
- Georgia’s export product diversification is relatively low and has not improved over time.
- Both Herfindahl product concentration and Theil’s index indicate Georgia is less diversified in export products compared to peers.
- Theil’s overall index shows Georgia is among the worst performers among peers in terms of product diversification (Theil’s index used to capture extensive and intensive margins).

### Export potentials: product space analysis and prospects
- The paper uses product space analysis to assess Georgia’s potential to diversify its export products based on its current comparative advantage (RCA>1).
- Two key insights from the product space:
  - A country’s diversification prospects depend in part on what it currently produces and exports; products located in the middle and dense part of the product space have many linkages and make diversification easier.
  - What a country produces and exports reflects underlying productive factors, including skills, technological know-how, and institutions.
- Georgia’s potential to diversify has improved but is still limited:
  - In 1995, Georgia’s merchandise exports with comparative advantage were mostly located on the periphery of the product space, implying limited links to new products.
  - By 2016, products with comparative advantage expanded into the dense and middle part of the product space; these include cars, chemicals, and plastics, which have many linkages to other products.
- Caveat: For Georgia, some industries (i.e., vehicles) are subject to high re-export content and therefore may not fully reflect true competitiveness of the sectors.

### Ability to upgrade to more complex products
- Mapping “distance” and “economic complexity” shows challenges in upgrading to more complex products:
  - Distance measures how easy it is for a country to produce new products given its current mix of exports; the larger the distance, the more difficult it is to produce that product.
  - Economic complexity ranks the complexity of a country’s export basket.
  - Potential products closely linked to Georgia’s current export basket (shorter distance) such as sugar, fish, and vegetables are mostly less complex than Georgia’s current economic complexity.
- Opportunity for future diversification is also limited:
  - Opportunity gain measures how producing a given product facilitates future diversification.
  - There is a tradeoff between distance and products with more opportunity: products with higher opportunity gain tend to be at larger distance (harder to reach).

### Drivers of export product diversification (literature and empirical analysis)
- Literature identifies key drivers: per capita income, human capital, population, terms of trade, exchange rate, and institutional and governance factors.
  - Higher GDP per capita and population provide incentives for firms to produce a wider range of products.
  - Favorable terms of trade are theoretically and empirically shown to increase export concentration (lower economic diversification).
  - Human capital accumulation and good governance are prerequisites for attracting investment and promoting diversification in high-value sectors.
  - Exchange rate overvaluation is associated with lower economic diversification.
- Empirical approach:
  - Panel regression analysis for middle income countries (excluding small states and islands) during 2000-14.
  - Dependent variable: Theil index (measure of export product diversification).
  - Sample divided into 3 clusters—average Theil indices during 2000-04, 2005-09, and 2010-14.
  - Independent variables: GDP per capita, human capital, terms of trade, population, rule of law, and trade openness.
  - To address endogeneity, regressors at time 2000, 2004, and 2010 are regressed onto 5-year averages of Theil indices.
- Key regression results (Text Table 1. Drivers of Export Product Diversification; Dependent Variable: Theil’s Index):
  - Log GDP per capita: -0.156***; -0.083**; -0.080**  [robust standard errors: 0.038; 0.037; 0.037]
  - Human Capital Index: -0.178***; -0.143***; -0.147***  [robust standard errors: 0.049; 0.045; 0.044]
  - Terms of Trade (Growth): 0.008***; 0.006***; 0.006***  [robust standard errors: 0.002; 0.001; 0.001]
  - Population (Billions): -0.474***; -0.388***; -0.377***  [robust standard errors: 0.061; 0.041; 0.045]
  - Rule of Law (-2.5 to 2.5 = best): -0.231***; -0.234***  [robust standard errors: 0.048; 0.049]
  - Trade Openness: 0.026  [robust standard error: 0.047] (statistically insignificant)
  - Year FE: Yes; Observations: 192; R-squared: 0.954; 0.962; 0.962
  - Robust standard errors in brackets; *** p<0.01, ** p<0.05, * p<0.1
- Interpretation of signs in the table: lower Theil’s index implies greater diversification; negative coefficients on Log GDP per capita, Human Capital Index, Population, and Rule of Law imply these factors are associated with greater diversification; positive coefficient on Terms of Trade implies favorable terms of trade are associated with less diversification.

### Structural weaknesses and constraints
- Georgia’s limited opportunity to diversify underlines structural weaknesses:
  - Skills mismatches and an inadequately educated workforce (cited as a problematic factor for doing business).
  - Low technological know-how and lagging innovation and business sophistication relative to comparators.
  - Limited quality of institutions in some areas despite improvements.
- Doing Business (2018) assessment: Georgia fares well in ease of doing business, starting a business, registering property, protecting minority investors, and enforcing contracts; further improvements could be made in easing insolvency procedures, trading across borders, and innovation.

### Policy implications and recommendations
- Two key policy implications derived from product space analysis and drivers:
  - Choosing potential products to diversify:
    - Strategically consider producing products that 1) are similar or closely linked to existing products because they employ similar know-how and hence are more efficient to produce; 2) are connected to many other potential products or at the core of the product space; and 3) provide more economic complexity.
  - Expanding underlying productive factors:
    - Structural reforms to advance human capital accumulation and improve quality of institutions are warranted to increase product diversification.
- Conclusion summary:
  - Human capital accumulation and better quality of institutions can help expand Georgia’s capacity to diversify its exports.
  - Georgia fares well in export market diversification thanks to free trade agreements with the European Union and China.
  - Georgia’s export product diversification is relatively low due to a narrow production base.
  - Product space analysis shows improving prospects for diversification but underlying capacity limits potential to upgrade to higher value-added products.
  - GDP per capita, human capital, quality of institutions, terms of trade, and population are significantly associated with economic diversification.
  - Structural reforms supporting human capital (better quality of education) and improvements to the quality of institutions can help expand Georgia’s capacity.

### Annex: Product space analysis and revealed comparative advantage (RCA)
- Product space concept: products in the core dense part of the product space display many links to nearby products (metal, machinery, chemicals); peripheral products (fishing, tropical, cereal and agriculture) have fewer links and hinder diversification.
- Revealed comparative advantage (RCA) definition and formula:
  - A country has Revealed Comparative Advantage in a product if it exports more than its “fair” share.
  - RCA = (X_cp / ∑_c X_cp) / (∑_p X_cp / ∑_p ∑_c X_cp)
  - Country c has revealed comparative advantage in product p if RCA >= 1.
- Note: The product space and RCA calculations are based on gross exports; for Georgia, high re-export content in some industries may not fully reflect true competitiveness.

*Source: IMF staff analysis (chapter: "1. Relative to its peers").*

### 1.      Financial conditions suggest that the monetary stance is slightly loose, but there is no

### 1. Financial conditions suggest that the monetary stance is slightly loose, but there is no evidence of price pressures.

### Assessment framework and methodology
- Two complementary methods used to assess monetary stance for Georgia:
  - Financial Conditions Index (FCI): provides a historical perspective and allows assessment of relative tightness or looseness of financial conditions.
  - Natural interest rate: estimated for Georgia and compared to central bank policy rate to assess the stance of monetary policy.
- VAR-based FCI specification:
  - FCIt = Σi=1..n ωi (xit − x̄i)
  - Variables enter as deviations from sample means (sample period: 2003Q1–2017Q4).
  - Weights (ωi) are estimated as the cumulative 5-quarter impulse response of GDP growth to a unit shock of xit from a recursive VAR including all financial variables, real GDP growth and the GDP deflator; structural shocks identified via Cholesky decomposition.
- Factor-based FCI specification:
  - Xt − μ = βFt + Et, with Ft purged of past economic activity by regressing Ft on current and lagged output growth (Ft = B(L) yt + υt).
- Variable selection criteria for FCI:
  - Time span should cover at least the last 10 years.
  - Impulse response of GDP growth to a unit shock must be economically meaningful.
- Final variables included in the FCI: deposit growth, REER, loan spread (domestic variables), and VIX, EUR-OIS spread (external variables).
  - Equity prices and house price index were dropped due to data limitations or lack of significant relationship with GDP growth.

### Key findings on FCI construction and drivers
- The weight of each variable in the VAR-based FCI is given by the cumulative five-quarter response of GDP growth to a unit shock in that variable.
- Contributions and correlations (Text Table 1: correlation between Real GDP Growth and Financial Variables; N = 60):
  - FCI: t = 0.544***, t+1 = 0.644***, t+2 = 0.629***, t+3 = 0.608***, t+4 = 0.533***
  - VIX: t = -0.363**, t+1 = -0.425***, t+2 = -0.476***, t+3 = -0.320*, t+4 = -0.305*
  - EUR-OIS spread: t = -0.0815, t+1 = -0.247, t+2 = -0.294*, t+3 = -0.446***, t+4 = -0.479***
  - Deposit growth: t = 0.787***, t+1 = 0.763***, t+2 = 0.607***, t+3 = 0.493***, t+4 = 0.187
  - Loan spread: t = -0.0961, t+1 = -0.097, t+2 = -0.0834, t+3 = -0.0779, t+4 = -0.0907
  - REER: t = -0.484***, t+1 = -0.553***, t+2 = -0.547***, t+3 = -0.505***, t+4 = -0.451***
  - p < 0.05, ** p < 0.01, *** p < 0.001
- Qualitative signs from impulse responses (Figure 1):
  - GDP growth falls in response to exchange rate appreciation, higher loan spread, higher VIX and higher EUR-OIS spread.
  - Higher deposit growth increases GDP growth.
- Factor analysis results:
  - Same set of variables used; correlation of individual financial variables with the common factor is greater than 50 percent except for loan-spread.
  - Factor loadings: positive correlation for deposit growth; negative correlation for REER, VIX, EUR-OIS spread and loan spread.

### Recent assessment of financial conditions
- Financial conditions are assessed to be slightly loose in the recent period.
  - Driver: loose global monetary and liquidity conditions coupled with cyclical recovery in Georgia and relatively stable exchange rates.
- Historically:
  - Exchange rate has been the most important driver of financial conditions.
  - Global factors (VIX and EUR-OIS spread) became increasingly important after the global financial crisis (GFC).
  - Financial conditions were supportive prior to the GFC, tightened significantly in 2008 (GFC and armed conflict with Russia), partially recovered, and deteriorated again in 2011 during rapid monetary contraction in response to a food price shock.

### Predictive power of FCI for GDP growth (Forecast evaluation)
- FCI has meaningful predictive power for GDP growth and is useful for forecasting:
  - Adding FCI to an autoregressive GDP model significantly increases R-squared.
  - FCI’s can explain up to 30 percent of the variation in GDP growth not explained by lagged GDP growth.
- Regression used for evaluation:
  - yt+h = β0 + Σi=1..k β1i yt+1−i + γ Xt + εt, where Xt denotes the FCI and lag length chosen by AIC.
- Partial R-squared (marginal contribution of adding FCI) for VAR- and factor-based FCIs is positive for horizons t+1 to t+4 (Figure 3).

### Estimating the natural interest rate for Georgia
- Methodology:
  - Time-varying parameter VAR (TVP-VAR) following Lubik and Matthes (2015) for three variables: growth rate of real GDP, inflation rate, and a measure of real interest rate.
  - Natural interest rate extracted using a long-horizon (5-year) forecast of the observed real rate as the measure of the natural rate.
- Estimates and comparisons:
  - Georgia’s real natural interest rate is estimated to be around 2 percent and largely stable over the sample period 2013Q4 – 2018Q1.
  - Using survey-based inflation expectations, Georgia’s real interest rate (ex ante) at 1.6 percent is lower than the natural rate, reflecting slightly loose monetary policy.
- Uncertainty and caveats:
  - Natural rate estimates contain uncertainty due to:
    - Large standard errors typical of natural rate estimation (Laubach and Williams, 2003; Lubik and Matthes, 2015).
    - Difficulty in accurately measuring inflation expectations for emerging market economies.

### Conclusion and policy implications
- Overall assessment:
  - Both the FCI-based analysis and the natural interest rate approach suggest a slightly loose monetary stance.
  - In the absence of price pressures (2.5 percent inflation in April 2018) despite cyclical upswing and narrowing output gap, the monetary stance is assessed to be broadly adequate at this point.
- Policy recommendation:
  - Given rapid credit growth, the National Bank of Georgia (NBG) should continue to monitor inflation and credit dynamics closely to support that the monetary policy stance remains adequate.

*IMF staff calculations; National Bank of Georgia data.*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2018/cr18199.pdf_