## LONG-TERM GROWTH PROSPECTS AND THE OUTPUT GAP IN MONTENEGRO

## Source details

**Canonical URL:** [LONG-TERM GROWTH PROSPECTS AND THE OUTPUT GAP IN MONTENEGRO](https://www.imf.org/-/media/files/publications/cr/2018/cr18122.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2018/cr18122.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2018/cr18122.pdf.json)

---

### Introduction and context
- Purpose: Provide long-term and medium-term growth projections and output gap estimates to inform fiscal sustainability, pensions, health care, education, and macroeconomic policy decisions.
- Methods used: production-function approach (Cobb-Douglas) and statistical filtering (two variations of the Hodrick-Prescott (HP) filter).
- Contextual points:
  - Montenegro: small, open, euroized economy, highly dependent on tourism and external financing, volatile growth since independence in 2006.
  - Reliable GDP series start in 2006 (ESA 2010 and NACE 2 consistent).
  - Large external shocks and structural breaks complicate potential output estimation.
  - Scenario note: EU accession and growth-enhancing reforms could yield growth above the illustrative “high-growth policy scenario.”
- Current output gap assessment: small positive output gap in 2017 and 2018 (less than 1 percent of GDP).

### Production-function approach — specification and baseline assumptions
- Production function: Cobb-Douglas with constant returns to scale: Yt = At Kt^(1−α) Lt^α.
  - Yt = output, Lt = employed labor, Kt = physical capital, At = Solow residual (TFP), α = labor share.
  - Capital accumulation: Kt = Kt−1 (1−δ) + It.
- Baseline parameter assumptions:
  - Labor share (α): 65 percent.
  - Capital depreciation rate (δ): 6 percent.
  - Initial capital stock ratio K/Y in 2005: 1.8.
- Sensitivity scenarios considered:
  - Labor share: 0.75 (high) and 0.55 (low).
  - Depreciation: 0.08 (high) and 0.04 (low).
  - Initial capital ratio K/Y: 2.3 (high) and 1.3 (low).

### Historical growth accounting (2006–17) — contributions and observations
- Growth statistics:
  - Growth over 2006-17: 2.5 percent on average and 2.4 percent annualized.
- Baseline contributions to annualized growth:
  - Capital: 2.3 percentage points.
  - Labor: 0.7 percentage points.
  - TFP: -0.5 percentage points.
- Observations on TFP:
  - Negative TFP contribution is unusual and near the 20th percentile of 10-year average TFP growth rates for emerging and advanced economies (2005-14).
  - Excluding crisis years 2009 and 2012, TFP growth was 0.6 on average.
- Sensitivity of component contributions across scenarios:
  - Labor contribution varies by 0.2 percentage points per year across scenarios.
  - Capital varies by 1.4 percentage points.
  - TFP varies by 1.2 percentage points.
- TFP sensitivity examples:
  - Labor share = 75 percent → average TFP growth rises to 0.0 percent (from -0.5 percent baseline).
  - Depreciation = 8 percent → average TFP growth increases to -0.1 percent.
  - Higher initial capital stock (K/Y = 2.3) → TFP growth rises to 0.0 percent.

### Initial capital stock, depreciation, and capital accumulation
- Data sources and 2006 capital-to-GDP ratios:
  - PWT (version 9.0): 2.3 (constant national prices); 2.8 in current US$ PPP terms.
  - IMF FAD database: 1.2.
  - Authorities' estimate (2005): 1.8 with assumed depreciation 5 percent.
- Implicit depreciation rates:
  - PWT implicit depreciation rate: 4 percent on average.
  - FAD implicit depreciation rate: 6.8 percent on average.
- Chosen analysis values:
  - Depreciation rate for analysis: 6 percent per year.
  - Initial capital stock ratio chosen: 1.8 for 2005.
- Implications for half-life of capital:
  - 6 percent depreciation → half-life of 11 years.
  - 4 percent depreciation → half-life of 17 years.
  - 8 percent depreciation → half-life of 8 years.
- Baseline investment assumptions and scenarios:
  - Investment ratio post-2023 set at 28 percent (based on staff macro-framework projections until 2023 and average historical investment-to-GDP ratio 2006-17).
  - Historical 5-year rolling averages between 2006-17: highest 32 percent and lowest 22 percent.
  - Nominal investment-to-GDP ratio in 2017: 30 percent.
  - Real investment-to-GDP ratio in 2017: 33 percent.
- Scenario impacts on capital accumulation:
  - High initial capital stock of 2.8 times GDP in 2005 (22 percent higher than baseline) → capital stock growth 0.2 percent lower than baseline over the projection period.
  - A 2 percentage point higher/lower depreciation rate → capital stock growth rate 0.34/0.67 percent lower/higher than baseline.
  - A 2 percentage point increase/decrease in the investment ratio → capital accumulation rate 0.2/0.2 percent higher/lower than baseline over 2018-50 (larger initial impact).
- Conclusion: initial capital stock assumption converges with baseline and is not crucial for long-run capital accumulation; depreciation rate assumption is important.

### Total Factor Productivity (TFP) — historical and projection assumptions
- Historical pattern: slow productivity growth due to structural and institutional reasons; high share of tourism and labor-intensive industries may constrain productivity growth.
- Baseline TFP assumption:
  - Baseline model assumes zero contribution from TFP over the projection period but adds 0.5 percentage point per year to reflect higher future TFP growth (text: "In the baseline model, we assume zero contribution from TFP over the projection period. Our baseline assumption is based on the historical experience, as explained in section B, but adds 0.5 percentage point per year to reflect higher future TFP growth.")
- Sensitivity scenarios:
  - High-growth scenario: assumes 0.5 percent TFP growth.
  - Low-growth scenario: assumes -0.5 percentage points contribution from TFP growth.
- Modeling note: TFP growth treated as residual; projections add 0.5 percentage points to historical averages for alternative models so baseline expected TFP boost is constant across models. Differences in historical TFP averages across models drive model-based growth projections.

### Long-term growth projections and contributions (2018–50)
- Projected potential output growth:
  - Average growth rate for 2018-23: 1.7 percent per year.
  - Projected average growth for 2041-50 decade: 0.1 percent per year.
- Baseline contributions to growth (Percent) — periods: 2018-23 / 2024-30 / 2031-40 / 2041-50 / 2018-50:
  - Labor: 0.10 / 0.05 / 0.05 / -0.19 / -0.01
  - Capital: 1.62 / 0.78 / 0.51 / 0.30 / 0.70
  - TFP: 0.00 / 0.00 / 0.00 / 0.00 / 0.00
  - Total: 1.71 / 0.83 / 0.56 / 0.11 / 0.69
- Interpretation:
  - Growth falls substantially in the baseline, reflecting a declining labor force and decreasing contributions from investment.
  - Even relatively high projected investment rates add less to long-run capital stock without TFP growth.
  - TFP is likely the main bottleneck for long-run growth prospects absent structural reforms.

### Model uncertainty, policy levers, and policy-based scenarios
- Model uncertainty:
  - For 2018-23, worst versus best model assumptions combined → difference of 1.1 percentage points per year on average.
  - Difference increases to 2.5 percentage points on average for 2041-50.
  - Maximum model average growth rate over 2018-50: 1.5 percent (assumes high depreciation rate, higher labor share, and high initial capital stock), largely driven by TFP (almost 1 percentage point per year from TFP in that model).
- Policy potential:
  - Best combination of policy levers could raise average growth over the projection period from 0.7 percent to 2 percent per year.
  - Most important policy levers: productivity growth (including human capital), labor force participation, unemployment, fertility, migration, and investment.
  - Government has limited influence over depreciation rate and labor share.
- Quantified policy impacts (simulated effects):
  - A 1/4 percentage point increase in the labor force participation rate per year → increases growth by 0.22 percentage points per year on average (with slightly higher initial impact).
  - A 1/4 percentage point reduction in the unemployment rate → increases growth by 0.17 percentage points per year on average (stated elsewhere as "0.17 percentage points per year, again with a higher initial impact").
  - Increasing the fertility rate to the UN high variant and no emigration, combined with labor market improvements and higher participation, are simulated in combined scenarios (see Employment section).
- Policy-lever aggregate impacts (selected values, percent contributions to growth by period):
  - High growth scenario (simple average): 2.92 (2018-23), 1.95 (2024-30), 1.83 (2031-40), 1.45 (2041-50), 1.94 (2018-50)
  - Sum positive policy contributions: 1.21 (2018-23), 1.12 (2024-30), 1.27 (2031-40), 1.35 (2041-50), 1.25 (2018-50)
  - No emigration: 0.09 (2018-23), 0.07 (2024-30), 0.08 (2031-40), 0.10 (2041-50), 0.09 (2018-50)
  - High investment ratio (+2%): 0.19 (2018-23), 0.13 (2024-30), 0.08 (2031-40), 0.04 (2041-50), 0.10 (2018-50)
  - Low unemployment rate (-0.25%): 0.19 (2018-23), 0.18 (2024-30), 0.17 (2031-40), 0.16 (2041-50), 0.17 (2018-50)
  - High fertility: 0.00 (2018-23), 0.00 (2024-30), 0.22 (2031-40), 0.35 (2041-50), 0.17 (2018-50)
  - High participation rate (+0.25%): 0.24 (2018-23), 0.23 (2024-30), 0.21 (2031-40), 0.19 (2041-50), 0.22 (2018-50)
  - High TFP (+0.5%): 0.50 (2018-23), 0.50 (2024-30), 0.50 (2031-40), 0.50 (2041-50), 0.50 (2018-50)
  - Baseline totals repeated for comparison: 1.71 (2018-23), 0.83 (2024-30), 0.56 (2031-40), 0.11 (2041-50), 0.69 (2018-50)
  - Low TFP (-0.5%): -0.50 across periods.
  - Sum negative policy contributions: -1.13 (2018-23), -1.07 (2024-30), -1.24 (2031-40), -1.35 (2041-50), -1.22 (2018-50)
  - Low growth scenario (simple average): 0.58 (2018-23), -0.24 (2024-30), -0.68 (2031-40), -1.25 (2041-50), -0.53 (2018-50)
- Notes:
  - Policy levers could improve growth by "1¼ percentage points or more per-year over the projection period."
  - Projected impact of policy levers likely underestimated because higher growth would induce higher investment, TFP growth, and immigration, which are not endogenously included.
  - Model uncertainty contributes between "0.5-0.8 percentage points" on the upside and "0.3-1.2 percentage points" on the downside.

### Employment projections and demographic scenarios — key figures
- Population projections (UN-based):
  - Estimated population in 2017: 629 thousand.
  - Baseline medium fertility variant projected population in 2050: 588 thousand (decline of almost 7 percent).
  - High fertility variant in 2050: 653 thousand.
  - Zero net emigration between 2016-50 scenario: population in 2050 = 610 thousand (3 percent decrease from 2017).
- Working-age population (15-65) projections:
  - Baseline: decline from 421 thousand to 352 thousand by 2050 (decline of 16 percent).
  - Fertility variant effect on working-age population for 2050: +/- 30 thousand (+/- 9 percent).
  - Zero-migration scenario: raises employment by 9 thousand by 2050 (4 percent) relative to baseline.
- Baseline employment assumptions:
  - Employment grows broadly at the rate of the working-age population.
  - Labor force participation (LFP) ratio (relative to 15-65 cohort) increases by 0.2 percentage points per year (reflecting higher female participation and increases in effective pension age).
  - Unemployment rate improves by 0.2 percentage points per year (gradual reduction in structural unemployment).
  - Under these assumptions, employment would stay essentially stable in the 225-230 thousand range through 2050.
- No-change scenario:
  - If LFP and unemployment rates remain constant, employment would fall from 229 thousand in 2017 to 192 thousand in 2050 (16 percent decline).
- Sensitivity of employment to policy/assumption changes (effects by 2050):
  - A 0.25 percentage point per year increase/decrease in LFP → raise/lower employment by some 25 thousand (+/- 11 percent).
  - A 0.25 percentage point per year decrease/increase in unemployment → raise/lower employment by some 20 thousand (+/- 9 percent).
  - Using UN high/low fertility variants → raise/lower employment by some 20 thousand (+/- 9 percent).
  - Zero-migration variant → raise employment by 9 thousand (4 percent) by itself; about 20 thousand in a combined scenario.
- Combined scenario outcomes by 2050:
  - High-growth upside scenario (0.25 p.p./yr increase in LFP, 0.25 p.p./yr decrease in unemployment, UN high fertility, no net outward migration): employment increases by 87 thousand to 316 by 2050 (38 percent).
  - Worse-case scenario (low LFP, high unemployment, low fertility): employment falls by 63 thousand to 160 thousand by 2050 (28 percent).

### Policy recommendations — labor market, human capital, and capital accumulation
- Labor market policies to increase employment and lift potential growth:
  1. Make hiring and firing processes more flexible while preserving adequate employee protection.
  2. Reduce the effective labor tax wedge.
  3. Support policies that increase female participation (parental leave, child care).
  4. Increase the effective pension age through pension reform.
  5. Reduce the skill mismatch by improving the education system.
  6. Discourage emigration and encourage immigration by improving domestic opportunities.
- Policies to improve capital accumulation:
  1. Create conditions conducive for domestic or FDI projects with significant impacts on economic activity.
  2. Increase domestic savings.
  3. Create fiscal space to maintain/increase public investment.
  4. Improve public investment selection and management frameworks.
  5. Strengthen financial markets to facilitate domestic and foreign investment.
- TFP and structural reform priority:
  - Structural reforms to raise TFP (human capital accumulation, business environment, institutions) are key to prevent growth rates from falling towards zero.
  - Productivity growth (TFP) has the largest potential impact; high TFP (+0.5%) contributes 0.50 percentage points to growth across periods in simulations.

### Output gap estimation — HP filter analysis and production-function feasibility
- HP filter approach:
  - One-sided and two-sided HP filters estimated using historical real GDP (2006-17) and projection path (2018-23), with lambda = 100 and lambda = 6.25 (Ravn and Uhlig (2002) guidance).
  - Lambda = 6.25 "fits Montenegro’s growth experience better" given short time series and greater uncertainty.
  - Two-sided HP with lambda = 6.25 (HP2-6.25) produces a more plausible result for 2017 and 2018 and projects the output gap closed by 2023.
- Production-function feasibility:
  - Difference between macro-framework real GDP growth and baseline production function projections largely explained by framework TFP contribution concentrated in 2020-23 (e.g., highway supply effect).
  - Selected feasibility figures (growth and TFP contributions):
    - Real GDP growth (framework): 3.0 (2006-17), 3.0 (2018), 2.4 (2019), 3.0 (2020), 3.1 (2021), 3.4 (2022), 3.0 (2023)
    - Real growth - production function (baseline): 3.0 (2006-17), 2.5 (2018), 1.9 (2019), 1.3 (2020), 1.5 (2021), 1.6 (2022), 1.6 (2023)
    - TFP contribution (framework): 0.0 (2006-17), 0.6 (2018), 0.5 (2019), 1.7 (2020), 1.6 (2021), 1.8 (2022), 1.4 (2023)
    - TFP contribution (production function baseline): -0.2 (2006-17), 0.0 (2018–23)
  - Constructed potential output using annualized TFP growth from 2006-23 yields plausible behavior through 2017 but diverging values for 2018-23, leading to increasing output gaps under that method.
  - Output gap in 2017 is almost identical across methodologies, indicating substantial judgment required.

### Structural fiscal balances, cyclically adjusted balance, and fiscal impulse — methods and findings
- CAB methodology:
  - CAB = R_CA – G_CA; revenue elasticity estimated by regression: ln(Revenue/TrendRevenue) = ∝ + β1 ln(Y/Y*) + u.
  - Trend revenues produced with two-sided HP filter with λ=100.
  - Estimated revenue elasticity εR = 1.14 (high statistical significance).
  - Estimated expenditure elasticity εG ≈ 0.6 (statistically significant at 90 percent), but policy view applied: εG = 0 for CAB calculation (expenditures treated as not automatically varying with output).
  - Chosen CAB formula: CAB = R (Y*/Y)^{εR} − G.
- One-off adjustments for structural balance:
  - One-off revenues excluded: tax debt rescheduling revenues in 2017; one-time telecommunications fee in 2016; large one-time VAT, PIT, social contributions by EPCG in 2014; repayment of loan to Prva Bank in 2009.
  - One-off expenditures excluded: 2008 government loan to Prva Bank (1.4 percent of GDP), repayments of guarantees (2009–2014 payments 0.1–3.2 percent of GDP), 2016 transfers to universities (0.8 percent of GDP), 2017 transfer to Montenegro Airlines (0.1 percent of GDP).
  - Bar-Boljare highway treated separately: structural balance presented with and without highway spending (excluding highway spending shows much stronger structural balance 2015-19).
- Fiscal impulse history and near-term stance:
  - Fiscal impulse: change in annual structural fiscal balance (indicates whether discretion is expansionary/neutral/contractionary).
  - After Global Financial Crisis: fiscal impulse neutral on average 2009-11; contractionary 2012-14; expansionary from 2015 with highway construction through 2017.
  - Excluding highway spending: non-highway budget neutral in 2016 and contractionary in 2017.
  - Near-term (2018-20): overall fiscal impulse negative in 2019 and 2020 as highway spending ends; non-highway budget very contractionary in 2018-20.
- Fiscal adjustment and debt projections:
  - General government debt (including guarantees) at end-2017: 75 percent of GDP.
  - Staff projection under full implementation of fiscal adjustment and a strong primary surplus:
    - Primary surplus peaking at 4½ percent of GDP in 2020 and falling towards 3 percent of GDP by 2023.
    - Debt lowered to 53 percent of GDP by 2023.
  - Recommendation: implement reforms to lower public-sector wage bill and restrict early retirements to relieve expenditure pressures.

### Reserve adequacy, ELA, and liquidity buffers
- Central Bank reserves and metrics:
  - At end-2017, CBM held nearly €900 million (21 percent of GDP) in reserves (up from €424 million or 13 percent of GDP at end-2013).
  - Reserves largely reflect banks’ required and excess reserves and central government deposits.
- Traditional reserve metrics (end-2017) — targets and results:
  - Months of import cover (target 3.0): Gross reserves 3.7; Net of bank excess reserves 1.6.
  - Reserves to short-term external debt (target 100.0): Gross reserves 295.0; Net 124.7.
  - Bank Deposits (percent, target 20.0): Gross reserves 34.8; Net 14.7.
- ARA metric (modified for Montenegro):
  - Modified ARA: Reserves = 10% of X + 30% of STD + 15% of Deposits (broad money substituted by deposits and weight increased to 15%).
  - Adequacy range: 100–150 percent of the metric.
  - ARA results (end-2017): end-2017 reserve levels are 170 percent (standard) and 137 percent (modified) of the metric.
  - Conservative scenario net of banks’ excess reserves: coverage 72 percent (standard ARA) and 58 percent (modified concept).
  - Excluding deposits of Euro Area banks substantially improves coverage (gross 226–253 percent; net 95–107 percent).
- Projections and policy implications:
  - Staff baseline: reserves projected to increase by 5 percentage points to 26 percent of GDP in 2018 (driven by government pre-financing of Eurobond amortizations), averaging 21 percent of GDP over 2020-23 after drawdowns.
  - Reserves in the range of 12-15 percent of GDP would be broadly adequate over the medium term per ARA metrics.
  - Recommendations: adjust required reserves, increase government deposits at the CBM, create a government sub-account at the central bank for ELA purposes.

### Banking sector liquidity, ELA capacity, and policy options
- Banking liquidity and CBM role:
  - Banks very liquid: overall ratio of liquid assets to short-term liabilities = 36 percent.
  - Banks' liquid assets plus 50 percent of required reserves and CBM intraday/overnight lending arrangements form buffers (~40 percent of short-term liabilities at end-2017).
  - CBM capital small: only 2 percent of banks’ short-term liabilities; ELA capacity limited.
- ELA policy options:
  - Potential viable options:
    - Explore standing line of credit from an international financial institution (consider cost and disbursement speed).
    - Create dedicated sub-account at CBM for ELA with safeguards and government reimbursement commitment after independent audit.
    - Consider prudently increasing reserve requirements to augment ELA buffers.
  - Pooled liquidity arrangements and liquid asset contributions assessed as having moral hazard or perverse incentive issues.

### Labor market outcomes, regulations, and reform priorities
- Stylized facts and key indicators:
  - Inactivity rate: about 37 percent (gap with EU average nearly 14 percentage points for women).
  - Unemployment rate: about 16.1 percent; youth (15–24 years) unemployment: 32 percent.
  - Long-term unemployment: about 80 percent of unemployed out of work for more than a year on average.
  - Informal employment: ratio of informal employment to labor force ~33 percent since 2009 (just above 35 percent in 2016); about 10 percent of population (15+) works in informal sector and half is inactive.
- Labor regulation observations:
  - Employment Protection Legislation (EPL) in Montenegro above OECD average; strict EPL can increase long-term unemployment and informal employment.
  - Draft new Labor Law proposals include: simplify dismissal procedures, extend duration of fixed-term contracts to 36 months, internal organization and classification of jobs Act, mechanism for amicable resolution.
  - Areas for further revision: allow employees to work for more than one employer; tailor internal organization requirements for SMEs; avoid provisions that eliminate contract extensions (concern for aging population).
- Wage costs, minimum wage, and labor taxation:
  - Real unit labor costs increased by 3 percent (2008-17); average wage real decline 5 percent (2010-17); labor productivity increased 4 percent (same period).
  - Minimum wage: just below 40 percent of average wage; relative to median/average debates cited.
  - Tax wedge on labor around 40 percent; personal income tax flat at 9 percent with no personal allowance; social security contributions add to combined tax burden of 33.8 percent of gross wages.
  - High effective tax wedge and design of social assistance/family benefits create disincentives to formal work and large effective marginal tax rates at low earnings.
- Informal employment specifics:
  - 78 percent of employees formally employed (67 percent with full salary declared, 10 percent with part declared); 22 percent informally employed.
  - Of those employed, 23 percent informally employed; 10 percent paid partially in cash.
  - 70 percent of informally employed are self-employed.
  - Age/education profile: dominant group aged 46–64; 37 percent of self-employed have no education or primary only; 6 percent have a college degree.
  - Informal incomes lower: informally employed people earn almost 30 percent less on average than formally employed; undeclared self-employed earn half of registered self-employed.
- Active labor market policies (ALMPs) and recommendations:
  - ALMPs include training, employment incentives, supported employment, public works, first-employment stimulation, start-up grants.
  - Average coverage of unemployed by programs 2014-16: 6 percent.
  - Recommendations: strengthen monitoring and evaluation, target measures based on demand-supply mismatch, require ex-ante evaluations and performance indicators.
- Policy recommendations (labor taxation, benefits, ALMPs, family policies):
  - Reduce tax wedge, especially for low-income earners; consider re-introduction of a basic allowance, targeted earned income tax credits, lower employer contributions, or finance health insurance from general budget.
  - Gradual withdrawal of social assistance and family benefits to avoid poverty traps and high effective marginal tax rates at low incomes.
  - Reduce length of paid parental leave toward OECD norms, require minimum father leave, expand childcare availability and subsidies to boost female participation.
  - Improve education-employer links, reorient ALMPs to address mismatches, and strengthen enforcement plus incentive-based strategies to reduce informality.

### Structural change, emigration, and long-run convergence prospects
- Structural shifts:
  - Deindustrialization and fast expansion of services; by 2013 services employment reached 77 percent, industry employment declined to 18 percent.
  - Growth relied on labor-intensive activities; structural change reduced demand for industrial occupations and lengthened job search.
- Emigration and demographics:
  - Western Balkans experienced population losses up to 18 percent via emigration (1995–2010); emigrants often men of prime age and above-average education.
  - Remittances may raise reservation wages, affect employment choices, and contribute to informal employment and female inactivity.
- Convergence prospects:
  - Baseline: Montenegro would likely not converge with EU without significant policy effort.
  - Under best policy-based scenario, per-capita growth could increase to almost 2 percent on average and could help reach EU income levels over time; real GDP per-worker may grow less in such scenarios if higher fertility and no net immigration lower per-worker output.

*Prepared by Marzie Taheri Sanjani and Min Song; content based on IMF staff analysis in the chapter "LONG-TERM GROWTH PROSPECTS AND THE OUTPUT GAP IN MONTENEGRO."*

### References ____________________________________________________________________________ 24

### LONG-TERM GROWTH PROSPECTS AND THE OUTPUT GAP IN MONTENEGRO

### Introduction
- Purpose: Provide long-term and medium-term growth projections and output gap estimates to inform fiscal sustainability, pensions, health care, education, and macroeconomic policy decisions.
- Methods used: production-function approach and statistical filtering (two variations of the Hodrick-Prescott (HP) filter). Focus is on production function and filtering due to data availability.
- Context and caveats:
  - Montenegro is a small, open, euroized economy highly dependent on tourism and external financing, with volatile growth since independence in 2006.
  - Reliable GDP data series start in 2006 (methodology consistent with ESA 2010 and NACE 2).
  - Large external shocks and structural breaks complicate potential output estimation.
  - Scenario note: If Montenegro joins the EU and implements growth-enhancing reforms, growth could exceed the illustrative “high-growth policy scenario.”

### Production Function Approach (Methodology and Baseline Assumptions)
- Production function specification: Cobb-Douglas with constant returns to scale: Yt = At Kt^(1−α) Lt^α.
- Definitions:
  - Yt = output, Lt = employed labor, Kt = physical capital, At = Solow residual (TFP), α = labor share.
  - Capital accumulation: Kt = Kt−1 (1−δ) + It.
- Baseline parameter assumptions:
  - Labor share (α): 65 percent (EU-average).
  - Capital depreciation rate (δ): 6 percent.
  - Initial capital stock ratio K/Y in 2005: 1.8.
- Sensitivity scenarios considered:
  - Labor share: 0.75 (high) and 0.55 (low).
  - Depreciation: 0.08 (high) and 0.04 (low).
  - Initial capital ratio K/Y: 2.3 (high) and 1.3 (low).

### Historical Growth Accounting (2006–17) — Key Statistics
- Growth:
  - Growth over 2006-17: 2.5 percent on average and 2.4 percent annualized.
- Contributions to annualized growth (baseline):
  - Capital: 2.3 percentage points.
  - Labor: 0.7 percentage points.
  - TFP: -0.5 percentage points (negative contribution).
- Observations on TFP:
  - Negative TFP contribution is unusual and near the 20th percentile of 10-year average TFP growth rates for emerging and advanced economies over 2005-14.
  - Excluding crisis years 2009 and 2012, TFP growth was 0.6 on average.
  - TFP sensitivity:
    - Labor share = 75 percent → average TFP growth rises to 0.0 percent (from -0.5 percent baseline).
    - Depreciation = 8 percent → average TFP growth increases to -0.1 percent.
    - Higher initial capital stock (K/Y = 2.3) → TFP growth rises to 0.0 percent.
  - Historical variation magnitudes:
    - Labor contribution varies by 0.2 percentage points per year across scenarios.
    - Capital varies by 1.4 percentage points.
    - TFP varies by 1.2 percentage points.

### Long-Term Growth Projections and Scenarios
- Baseline projection:
  - Projected growth declining from about 3 percent in 2018 to around zero in 2050.
- Policy-based scenarios:
  - Range of long-run growth paths of 1¼ percentage points above or below the baseline estimates; upside could be larger if productivity improves.
- Main drivers for potential improvement:
  - Productivity growth (TFP) has largest potential impact — can be raised via education and business environment improvements.
  - Labor market improvements and incentives for capital accumulation also have potentially large positive effects.
- Output gap:
  - The analysis suggests that there currently is a small positive output gap.

### Employment Projections and Demographics — Key Figures and Scenarios
- Population projections (UN-based):
  - Estimated population in 2017: 629 thousand.
  - Baseline medium fertility variant projected population in 2050: 588 thousand (decline of almost 7 percent).
  - High fertility variant in 2050: 653 thousand.
  - Zero net emigration between 2016-50 scenario: population in 2050 = 610 thousand (3 percent decrease from 2017).
- Working-age population (15-65) projections:
  - Baseline: decline from 421 thousand to 352 thousand by 2050 (decline of 16 percent).
  - Fertility variant effect on working-age population for 2050: +/- 30 thousand (+/- 9 percent).
  - Zero-migration scenario: raises employment by 9 thousand by 2050 (4 percent) relative to baseline.
- Baseline employment assumptions:
  - Employment grows broadly at the rate of the working-age population.
  - Labor force participation (LFP) ratio (relative to 15-65 cohort) increases by 0.2 percentage points per year (reflecting higher female participation and increases in effective pension age).
  - Unemployment rate improves by 0.2 percentage points per year (gradual reduction in structural unemployment).
  - Under these assumptions, employment would stay essentially stable in the 225-230 thousand range through 2050.
- No-change scenario:
  - If LFP and unemployment rates remain constant, employment would fall from 229 thousand in 2017 to 192 thousand in 2050 (16 percent decline).
- Sensitivity of employment to policy/assumption changes (effects by 2050):
  - A 0.25 percentage point per year increase/decrease in LFP → raise/lower employment by some 25 thousand (+/- 11 percent).
  - A 0.25 percentage point per year decrease/increase in unemployment → raise/lower employment by some 20 thousand (+/- 9 percent).
  - Using UN high/low fertility variants → raise/lower employment by some 20 thousand (+/- 9 percent).
  - Zero-migration variant → raise employment by 9 thousand (4 percent) by itself; about 20 thousand in a combined scenario.
- Combined scenario outcomes:
  - High-growth upside scenario (0.25 p.p./yr increase in LFP, 0.25 p.p./yr decrease in unemployment, UN high fertility, no net outward migration): employment increases by 87 thousand to 316 by 2050 (38 percent).
  - Worse-case scenario (low LFP, high unemployment, low fertility): employment falls by 63 thousand to 160 thousand by 2050 (28 percent).

### Policy Recommendations — Labor Market Policies
- To increase employment and lift potential growth, recommended labor market policies include:
  1. Make hiring and firing processes more flexible while preserving adequate employee protection;
  2. Reduce the effective labor tax wedge;
  3. Support policies that increase female participation (parental leave, child care);
  4. Increase the effective pension age through pension reform;
  5. Reduce the skill mismatch by improving the education system;
  6. Discourage emigration and encourage immigration by improving domestic opportunities.

*Prepared by Marzie Taheri Sanjani and Min Song; content based on IMF staff analysis in the chapter "LONG-TERM GROWTH PROSPECTS AND THE OUTPUT GAP IN MONTENEGRO."*

### 16.      The initial capital stock and the

### 16.      The initial capital stock and the

### Estimates of initial capital stock and depreciation
- Two publicly available estimates for Montenegro: Penn World Tables (PWT, version 9.0) and IMF FAD Investment and Capital Stock Dataset.
- PWT capital-to-GDP ratio in 2006: 2.3 (calculated in constant national prices); 2.8 in current US$ PPP terms.
- FAD database capital-to-GDP ratio in 2006: 1.2.
- Authorities estimate capital-to-GDP ratio in 2005: 1.8 and assume a depreciation rate of 5 percent.
- Implicit depreciation rates:
  - PWT implicit depreciation rate: 4 percent on average.
  - FAD implicit depreciation rate: 6.8 percent on average.
  - Chosen depreciation rate for analysis: 6 percent per year.
- Implications of chosen depreciation rates for half-life of average capital good:
  - 6 percent depreciation → half-life of 11 years.
  - 4 percent depreciation → half-life of 17 years.
  - 8 percent depreciation → half-life of 8 years.
- Chosen initial capital stock ratio: 1.8 for 2005 (matches authorities’ numbers and provides reasonable capital evolution).

### Baseline projections and scenario design
- Baseline projections of capital stock:
  - Based on staff macro-framework investment projections until 2023 and the average historical investment-to-GDP ratio (2006-17) for the long-term.
  - Investment ratio post-2023 set at 28 percent.
  - Historical 5-year rolling averages between 2006-17: highest 32 percent and lowest 22 percent.
- Real versus nominal investment-to-GDP ratio (2017):
  - Nominal ratio in 2017: 30 percent.
  - Real ratio in 2017: 33 percent.
- Sensitivity scenarios considered:
  - High and low depreciation rates (±2 percentage points).
  - High and low initial capital ratios.
  - High investment case: investment 2 percentage points of GDP higher over the projection period.
  - Low investment case: investment 2 percentage points of GDP lower over the projection period.
- Key scenario impacts (using the same real investment path as baseline):
  - High initial capital stock of 2.8 times GDP in 2005 (22 percent higher than baseline) → capital stock growth 0.2 percent lower than baseline over the projection period (due to higher depreciation).
  - A 2 percentage point higher/lower depreciation rate → capital stock growth rate 0.34/0.67 percent lower/higher than baseline.
  - A 2 percentage point increase/decrease in the investment ratio → capital accumulation rate 0.2/0.2 percent higher/lower than baseline over 2018-50 (with larger initial impact).
- Conclusion: initial capital stock assumption converges with baseline and is not very crucial for long-run capital accumulation; depreciation rate assumption is important.

### Domestic savings and financing of investment
- Montenegro’s private saving is well below the average of Western Balkan peers but has trended upward since 2008 (driven by remittances, growth rates, and income).
- Currently, a large part of investment in Montenegro is financed by the government, which has temporarily endangered debt sustainability and cannot continue at the high present level.
- Concern: gap between investment needs and total saving ratios raises sustainability concerns and should be addressed by increasing domestic savings and improving prospects for domestic investment.

### Policy recommendations to improve capital accumulation
- Recommended policies:
  1. Create conditions conducive for domestic or FDI projects with significant impacts on economic activity.
  2. Increase domestic savings.
  3. Create fiscal space to maintain/increase public investment.
  4. Improve public investment selection and management frameworks.
  5. Strengthen financial markets to facilitate domestic and foreign investment.

### Total Factor Productivity (TFP) projections and scenarios
- Historical productivity growth has been slow due to structural and institutional reasons; structural features (high share of tourism and labor-intensive industries) may contribute to low productivity growth.
- Baseline TFP assumption:
  - "In the baseline model, we assume zero contribution from TFP over the projection period. Our baseline assumption is based on the historical experience, as explained in section B, but adds 0.5 percentage point per year to reflect higher future TFP growth."
- Sensitivity analysis:
  - High-growth scenario: assumes 0.5 percent TFP growth (in line with experience of other emerging market economies).
  - Low-growth scenario: assumes minus 0.5 percentage points contribution from TFP growth (in line with historical baseline estimation).
- Modeling approach:
  - TFP growth is treated as a residual; projections add 0.5 percentage points to historical averages for all alternative models so that the baseline expected TFP boost is constant across models.
  - Differences in historical TFP averages across models drive model-based growth projections more than other assumptions.

### Long-term growth projections and contributions (2018-50)
- Projected potential output growth:
  - Average growth rate for 2018-23: 1.7 percent per year.
  - Projected average growth for 2041-50 decade: 0.1 percent per year.
- Baseline contributions to growth (Percent) — periods: 2018-23 / 2024-30 / 2031-40 / 2041-50 / 2018-50:
  - Labor: 0.10 / 0.05 / 0.05 / -0.19 / -0.01
  - Capital: 1.62 / 0.78 / 0.51 / 0.30 / 0.70
  - TFP: 0.00 / 0.00 / 0.00 / 0.00 / 0.00
  - Total: 1.71 / 0.83 / 0.56 / 0.11 / 0.69
- Interpretation:
  - Growth falls substantially in the baseline, reflecting a declining labor force and decreasing contributions from investment.
  - Even relatively high projected investment rates add less to long-run capital stock without TFP growth.
  - TFP is likely the main bottleneck for long-run growth prospects absent structural reforms.

### Model uncertainty and policy levers
- Model uncertainty effects:
  - For 2018-23, worst model assumptions combined versus best result in a difference of 1.1 percentage points per year on average; difference increases to 2.5 percentage points on average for 2041-50.
  - Maximum model average growth rate over 2018-50: 1.5 percent (assumes high depreciation rate, higher labor share, and high initial capital stock), largely driven by TFP (almost 1 percentage point per year from TFP in that model).
- Equalizing TFP assumptions across models:
  - Results in smaller and more intuitive model differences; depreciation rate would have the highest impact and initial capital stock the smallest, with impacts declining over time.
- Policy potential:
  - Best combination of policy levers could raise average growth over the projection period from 0.7 percent to 2 percent per year.
  - Most important policy levers: productivity growth (including human capital), labor force participation, unemployment, fertility, migration, and investment.
  - Government has limited influence over depreciation rate and labor share.
  - Specific quantified policy impacts:
    - A 1/4 percentage point increase in the labor force participation rate per year → increases growth by 0.22 percentage points per year on average (with slightly higher initial impact).
    - A 1/4 percentage point reduction in the unemployment rate would increase growth by (statement incomplete in source text; no numeric value supplied beyond this point).

*Source: IMF staff estimates and projections (content unit: cr18122 - 16.      The initial capital stock and the).*

### 0.17 percentage points per year, again with a higher initial impact. Increasing the fertility rate to the

### cr18122 - 0.17 percentage points per year, again with a higher initial impact. Increasing the fertility rate to the

### Policy contributions to growth — simulated effects and aggregates
- Policy levers could improve growth by "1¼ percentage points or more per-year over the projection period."
- Projected impact of policy levers likely underestimated because higher growth would induce higher investment and capital accumulation, TFP growth, and immigration, which is not included in the projections.
- Model uncertainty contributes between "0.5-0.8 percentage points" on the upside and a somewhat wider range on the downside "(0.3-1.2 percentage points)."
- Policy levers considered (individual contributions by scenario and period):
  - High growth scenario (simple average): 2.92 (2018-23), 1.95 (2024-30), 1.83 (2031-40), 1.45 (2041-50), 1.94 (2018-50)
  - Sum positive policy contributions: 1.21 (2018-23), 1.12 (2024-30), 1.27 (2031-40), 1.35 (2041-50), 1.25 (2018-50)
  - No emigration: 0.09 (2018-23), 0.07 (2024-30), 0.08 (2031-40), 0.10 (2041-50), 0.09 (2018-50)
  - High investment ratio (+2%): 0.19 (2018-23), 0.13 (2024-30), 0.08 (2031-40), 0.04 (2041-50), 0.10 (2018-50)
  - Low unemployment rate (-0.25%): 0.19 (2018-23), 0.18 (2024-30), 0.17 (2031-40), 0.16 (2041-50), 0.17 (2018-50)
  - High fertility: 0.00 (2018-23), 0.00 (2024-30), 0.22 (2031-40), 0.35 (2041-50), 0.17 (2018-50)
  - High participation rate (+0.25%): 0.24 (2018-23), 0.23 (2024-30), 0.21 (2031-40), 0.19 (2041-50), 0.22 (2018-50)
  - High TFP (+0.5%): 0.50 (2018-23), 0.50 (2024-30), 0.50 (2031-40), 0.50 (2041-50), 0.50 (2018-50)
  - Baseline: 1.71 (2018-23), 0.83 (2024-30), 0.56 (2031-40), 0.11 (2041-50), 0.69 (2018-50)
  - Low TFP (-0.5%): -0.50 (2018-23), -0.50 (2024-30), -0.50 (2031-40), -0.50 (2041-50), -0.50 (2018-50)
  - Low participation rate (-0.25%): -0.25 (2018-23), -0.25 (2024-30), -0.24 (2031-40), -0.23 (2041-50), -0.24 (2018-50)
  - Low fertility: 0.00 (2018-23), 0.00 (2024-30), -0.23 (2031-40), -0.39 (2041-50), -0.19 (2018-50)
  - High unemployment rate (+0.25%): -0.19 (2018-23), -0.19 (2024-30), -0.19 (2031-40), -0.18 (2041-50), -0.19 (2018-50)
  - Low investment ratio (-2%): -0.19 (2018-23), -0.13 (2024-30), -0.08 (2031-40), -0.04 (2041-50), -0.10 (2018-50)
  - Sum negative policy contributions: -1.13 (2018-23), -1.07 (2024-30), -1.24 (2031-40), -1.35 (2041-50), -1.22 (2018-50)
  - Low growth scenario (simple average): 0.58 (2018-23), -0.24 (2024-30), -0.68 (2031-40), -1.25 (2041-50), -0.53 (2018-50)
- Specific note on investment ratio: increasing the investment ratio by 2 percentage points per year would raise growth by "0.10 percent per year" on average, though likely underestimated because permanently higher investment would probably endogenously increase TFP. Short-term impact is "0.2 percentage points over the 2018-23 period."

### Convergence, long-run prospects, and role of institutions
- Baseline: Montenegro would likely not converge with the rest of EU without significant policy efforts.
- Convergence hypothesis: economies with lower per capita GDP should converge to higher income levels assuming similar technologies, saving, demographic features, and human capital.
- Policy actions that could help convergence: EU accession, adoption of new technologies, enhanced institutions, higher FDI, increased overall efficiency, and reforms to raise productivity.
- Montenegro’s legal, judiciary, and regulatory institutions "are still catching up with EU standards."
- Under the best policy-based scenario, per-capita growth "would increase to almost 2 percent on average" and could help reach EU level incomes over time.
- Real GDP per-worker would grow less in the best policy scenarios because higher fertility and no net immigration lower per-worker output.

### Potential output and output gap — HP filter analysis
- Univariate filters like the Hodrick-Prescott (HP) filter assume potential output is a smoothed trend around actual output; they have conceptual shortcomings including the "end-point problem."
- The analysis estimates one-sided and two-sided HP filters using historical real GDP data (2006-17) and the projection path (2018-23), with smoothing parameters lambda = 100 and lambda = 6.25 (Ravn and Uhlig (2002) discussion).
- A lower lambda value provides more rapid adjustment of estimated potential output with respect to actual data.
- Results indicate lambda=6.25 "fits Montenegro’s growth experience better" given the relatively short available time series and greater uncertainty.
- One-sided filters use past data only; two-sided filters use projected values as well. Two-sided HP with lambda=6.25 (HP2-6.25) produces a more plausible result for the latest observation, 2017, and for the projection period 2018-23.
- The HP2-6.25 filter:
  - Captures periods of growth downturns well.
  - Possibly underestimates booms due to the "fast corrections" assumption.
  - Estimates the output gap in 2017 and 2018 to be "small and positive" and the output gap is closed at the end of the projection period, 2023.

### Production function approach and feasibility of framework projections
- The difference between real GDP growth rates in the macro-economic framework and the baseline production function (PF) projections is largely explained by the TFP contribution.
- Table of selected feasibility figures (Real GDP growth and TFP contributions):
  - Real GDP growth: 3.0 (2006-17), 3.0 (2018), 2.4 (2019), 3.0 (2020), 3.1 (2021), 3.4 (2022), 3.0 (2023)
  - Real growth - production function (baseline): 3.0 (2006-17), 2.5 (2018), 1.9 (2019), 1.3 (2020), 1.5 (2021), 1.6 (2022), 1.6 (2023)
  - TFP contribution (framework): 0.0 (2006-17), 0.6 (2018), 0.5 (2019), 1.7 (2020), 1.6 (2021), 1.8 (2022), 1.4 (2023)
  - TFP contribution (production function baseline): -0.2 (2006-17), 0.0 (2018), 0.0 (2019), 0.0 (2020), 0.0 (2021), 0.0 (2022), 0.0 (2023)
- The larger TFP contribution in the framework relative to the baseline PF approach is concentrated in 2020-23 and could reflect the highway supply effect; however, this could imply the macro-economic framework assumptions are "somewhat optimistic."
- Production function method can derive potential GDP and output gap by using smoothed TFP as underlying TFP growth; subtracting annualized TFP growth or a trailing moving average from estimated TFP provides an estimate of growth contribution from the change in the output gap.
- Constructed potential output series using annualized TFP growth from 2006-23 produces plausible behavior through 2017 but diverging values for 2018-23, leading to increasing output gaps for 2018-23 in this method.
- Output gap in 2017 is "almost identical" across methodologies, suggesting substantial judgment is needed when estimating the output gap.

### Conclusions and growth drivers
- Long-run growth estimated using a production function with employment, capital, and TFP growth.
- Historical drivers: growth was driven mostly by capital with some contribution from labor, while TFP contributed negatively.
- Baseline outlook (no reforms):
  - Employment will likely have a slightly negative contribution because of demographic dynamics unless both labor force participation and unemployment improve significantly.
  - The highway project will contribute to capital accumulation in the near term, but the contribution from capital accumulation will likely fall despite relatively high investment ratios.
  - Based on historical performance, TFP contribution is likely limited and "constitutes the main bottleneck for long-run growth prospects in the no-reform baseline."
- Policy implication: structural reform efforts to increase TFP through human capital accumulation and improving the efficiency of the economy are "key for preventing growth rates from falling towards zero."
- Policy-based scenarios show growth rates could increase by "2 percentage points per year" in a high growth scenario and perhaps more if this high growth scenario leads to immigration and other positive feedback loops.
- Output gap analysis: the output gap in both 2017 and 2018 is positive but small at "less than 1 percent of GDP."

*Source: IMF staff estimates and projections.*

### References

### References and Structural Fiscal Balances

### References (selected bibliography)
- Acemoglu, A. and Johnson S., 2005 “Unbundling Institutions” Journal of Political Economy, 113, 949-995.
- Alberola, E., Estrada, A., and D. Santabárbara, 2014, “Growth and Imbalances in Spain: A Reassessment of the Output Gap,” Journal of the Spanish Economic Association, Vol. 5(2–3), p. 333–356.
- Blagrave, P., R. Garcia-Saltos, D. Laxton, and F. Zhang, 2015, "A Simple Multivariate Filter for Estimating Potential Output," IMF Working Papers 15/79 (Washington: International Monetary Fund).
- Benes, J., Johnson M., Clinton K., Matheson T., Manchev P., Garcia-Saltos R., and Laxton D., 2010, “Estimating Potential Output with a Multivariate Filter,” IMF Working Paper 10/285.
- Berger, H., Dowling, T., Lanau, S., Lian, W., Mrkaic, M., Rabanal, P., and Taheri Sanjani T., 2015, “Steady as She Goes–Estimating Potential Output During Financial “Booms and Busts”, IMF Working Paper 15/233 (Washington: International Monetary Fund).
- Borio, C., Piti D., and Mikael K., 2013, “Rethinking Potential Output: Embedding Information About the Financial Cycle,” BIS Working Paper No. 404.
- Böwer, U. and Turrini, A., 2009. "EU accession: A road to fast-track convergence?", European Economy - Economic Papers 2008 - 2015 393, Directorate General Economic and Financial Affairs (DG ECFIN), European Commission.
- Bosworth, B. P., and S.M. Collins, 2003. "The Empirics of Growth: An Update," Brookings Papers on Economic Activity, vol. 34(2003–2), pages 113–206.
- Campos, N., Coricelli, F., and Moretti, L., 2014. "Economic Growth and Political Integration: Estimating the Benefits from Membership in the European Union Using the Synthetic Counterfactuals Method," IZA Discussion Papers 8162, Institute for the Study of Labor (IZA).
- Christiano, Lawrence J, and Terry J Fitzgerald, 2003, “The Band Pass Filter,” International Economic Review, Vol. 44(2), p. 435–465.
- D'Auria, F., Denis, C., Havik, K., Mc Morrow, K., Planas, C., Raciborski, R., Werner, R. and Rossi, A., 2010, “The Production Function Methodology for Calculating Potential Growth and Output gaps” European Economy - Economic Papers 2008 - 2015 420, Directorate General Economic and Financial Affairs (DG ECFIN), European Commission.
- Dobrinsky, R. and Havlik, P., 2014. "Economic Convergence and Structural Change: The Role of Transition and EU Accession," wiiw Research Reports 395, The Vienna Institute for International Economic Studies, wiiw.
- Fernald, J., 2014, “Productivity and Potential Output before, During, and after the Great Recession,” NBER Working Paper No. 20248.
- Havik, Karel, Kieran Mc Morrow, Fabrice Orlandi, Christophe Planas, Rafal Raciborski, Werner Röger, Alessandro Rossi, Anna Thum-Thysen, Valerie Vandermeulen., 2014, “The Production Function Methodology for Calculating Potential Growth and Output Gaps” European Economy - Economic Papers 2008 - 2015 535, Directorate General Economic and Financial Affairs (DG ECFIN), European Commission.
- IMF, World Economic Outlook, April 2017, Chapter 2, “Road Less Traveled: Growth in Emerging Markets and Developing Economies in a Complicated External Environment”.
- IMF, “The Western Balkans: 15 years of Economic Transition”, Regional Economic Issues Special Report, March 2015.
- IMF, Regional Economic Issues (REI): Central, Eastern, and Southeastern Europe, November 2016, “Effective Government for Stronger Growth”, Washington, DC.
- IMF, “Potential Output in Albania”, Selected Issues Paper, 2016.
- IMF (2017), “FAD Investment and Capital Stock Database 2017: Manual & FAQ – Estimating Public, Private, And PPP Capital Stocks.” Washington, DC.
- Hausmann, R., Pritchett, L., and Rodrik, R., 2005. “Growth Accelerations.” Journal of Economic Growth 10 (4): 303–29.
- Lenain, P., and Rawdanowicz, L., 2004. "Enhancing Income Convergence in Central Europe after EU Accession," OECD Economics Department Working Papers 392, OECD Publishing.
- Martin, C., Velazquez, F., and Funck, B., 2001. "European Integration and Income Convergence: Lessons for Central and Eastern European Countries," World Bank Publications, The World Bank, number 13968, December.
- Okun, A., 1962, “Potential GNP: Its Measurement and Significance,” Cowles Foundation Paper 190.
- Podpiera, J, Raei, F, and Stepanyan, A, 2017, “A Fresh Look at Potential Output in Central, Eastern, and Southeastern European Countries,” IMF Working Paper 17/37 (Washington: International Monetary Fund).
- Rabanal, P., and Taheri Sanjani, M., 2015, “Financial Factors: Implications for Output Gaps,” IMF Working Paper 15/153 (Washington: International Monetary Fund).
- Rapacki, R. and Próchniak, M., 2009. "The EU enlargement and economic growth in the CEE new member countries," European Economy - Economic Papers 2008 - 2015 367, Directorate General Economic and Financial Affairs (DG ECFIN), European Commission.
- Ravn, M., and H. Uhlig, 2002, "On Adjusting the Hodrick-Prescott Filter for the Frequency of Observations," The Review of Economics and Statistics, Vol. 84(2), pp. 371–375.
- Svirydzenka, K. and Petri, M., 2014, “Mauritius: The Drivers of Growth—Can the Past be Extended?”, IMF working paper WP/14/134.

### Structural Fiscal Balances — Major findings and methods
- High-level finding:
  - Fiscal policies in Montenegro have generally been pro-cyclical (expansionary during upswings and contractionary during downturns). Starting in 2018, fiscal policies should be appropriately counter-cyclical as Montenegro undertakes fiscal adjustment during an economic expansion.
  - Structural fiscal balance excluding spending on the Bar-Boljare highway indicates that the necessary fiscal adjustment to restore sustainability is not as large as the headline fiscal balance suggests.
  - Recommendation: implement reforms to strengthen budget institutions, in particular medium-term budgetary frameworks, and maintain sound fiscal policies to facilitate strong financial market access.

- A. Introduction
  - Distinguishes discretionary fiscal decisions from cyclical effects and one-off items.
  - Cyclically adjusted fiscal balance (CAB) removes cyclical effects; structural fiscal balance further removes one-off revenues and expenditures.
  - The analysis estimates cyclically adjusted and structural fiscal balances and calculates the fiscal impulse to assess whether discretionary fiscal policy changes are expansionary or contractionary.
  - Prepared by William Lindquist.

- B. Cyclically Adjusted Fiscal Balance — methodology and estimates
  - Decomposition: OB = CB + CAB where OB is overall fiscal balance, CB is cyclical balance, CAB is cyclically adjusted balance.
  - CAB computed as CAB = R_CA – G_CA.
  - Revenue elasticity estimation:
    - Regression: ln(Revenue/TrendRevenue) = ∝ + β1 ln(Y/Y*) + u estimates elasticity β1.
    - Trend revenues produced with a two-sided HP filter with λ=100 (for trend revenues in revenue regression).
    - Adjustments to revenues before regression include: tax debt rescheduling revenues received in 2017; a one-time telecommunications fee in 2016; large one-time VAT, PIT, and social contributions made by EPCG in 2014; repayment of a loan to Prva Bank in 2009.
    - Revenues deflated by the GDP deflator (real terms).
    - Tax policy changes were not adjusted for due to data limitations.
  - Estimated revenue elasticity:
    - Estimated elasticity of revenues to the output gap: 1.14 with high statistical significance.
    - Table details (OLS estimation of elasticity of revenues to the output gap):
      - Log of output gap coefficient: 1/1.136 0.209 5.426 0.000
      - Constant: -0.624 0.963 -0.648 0.528
      - Number of observations: 2/15
      - R-squared: 0.694
      - Note: output gap is ratio of observed to potential output; annual data from 2002 to 2016.
  - Expenditure elasticity estimation:
    - Estimated elasticity of expenditures to the output gap: around 0.6 with a p-value indicating statistical significance at the 90 percent confidence level.
    - Table details (OLS estimation of elasticity of expenditures to the output gap):
      - Log of output gap coefficient: 2/0.574 0.279 2.054 0.061
      - Constant: 1.961 1.286 1.525 0.151
      - Number of observations: 3/15
      - R-squared: 0.245
      - Note: includes highway expenditures; output gap defined as ratio of observed to potential output; annual data from 2002 to 2016.
    - Interpretation: the estimated expenditure elasticity does not imply automatic expenditure variation with output; rather expenditures have been pro-cyclical due to policy decisions and financing constraints. Automatic stabilizers are small (example: unemployment benefits of 0.3 percent of GDP in 2017).
  - Choice of elasticities for CAB calculation:
    - Use estimated revenue elasticity εR = 1.14.
    - Use expenditure elasticity εG = 0 (policy view that expenditures do not automatically change with output).
    - CAB formula used: CAB = R (Y*/Y)^{εR} − G.
  - Output gap estimation:
    - Output gap estimates produced using a two-sided HP filter with a lambda smoothing parameter of 6.25 (resulting in relatively small output gap estimates).
    - Consequence: cyclically adjusted fiscal balance does not exhibit strong variations from observed balance given small output gap estimates.

- C. Structural Fiscal Balance — adjustments and interpretation
  - Structural balance adjusts for non-recurring revenues and expenditures in addition to cyclical effects; useful for assessing necessity and size of fiscal adjustment.
  - One-off revenue/expenditure adjustments applied:
    - One-off revenues: tax debt rescheduling revenues in 2017; one-time telecommunications fee in 2016; large one-time VAT, PIT, and social contributions by EPCG in 2014; repayment of loan to Prva Bank in 2009; omitted expected dividend payment of 0.7 percent of GDP from EPCG in 2018 and related corporate income tax payment of 0.1 percent of GDP.
    - One-off expenditures excluded (per EC and Bornhorst et al (2011) guidance): loan to Prva Bank (2008 government loan to Prva Bank for 1.4 percent of GDP, reversed in 2009); repayment of guarantees (2009–2014 payments ranging from 0.1 to 3.2 percent of GDP each year); transfers to public institutions (2016 one-off payment 0.8 percent of GDP to universities); transfers to public enterprises (2017 transfer to Montenegro Airlines of 0.1 percent of GDP).
  - Bar-Boljare highway treatment:
    - Highway expenditures concentrated in 2015-19; presented structural fiscal balance with and without highway spending.
    - Including highway spending, structural fiscal balance broadly similar to cyclically adjusted balance.
    - Excluding highway spending, structural fiscal balance from 2015-19 is much stronger than headline balance — implying the size of needed fiscal adjustment is much smaller than suggested by overall balance, conditional on highway spending being truly one-off and no further highway phases financed through the budget.

- D. Fiscal Impulse — stance and history
  - Fiscal impulse defined as change in annual structural fiscal balance; indicates whether fiscal policy changes are expansionary, neutral, or contractionary (does not by itself estimate impact on growth absent fiscal multiplier).
  - Historical patterns:
    - After the Global Financial Crisis, fiscal impulse was on average neutral over 2009-11.
    - Fiscal impulse turned contractionary over 2012-14 during fiscal consolidation.
    - Beginning of highway construction in 2015 marked return of expansionary policies, which continued through 2017 when initial fiscal adjustment measures were offset by acceleration of highway spending.
  - Non-highway budget perspective:
    - Excluding highway spending, non-highway budget was neutral in 2016 and contractionary in 2017 as fiscal adjustment measures were implemented.
    - In 2015 the non-highway budget was more expansionary due to a decline in the revenue/GDP ratio in 2015, partly caused by a reduction in the “crisis tax” personal income tax rate on higher incomes.
  - Near-term stance:
    - Fiscal stance will be tight in 2018-20 as highway spending ends. Because highway spending is not projected to accelerate significantly in 2018, it is relatively neutral on economic growth, but the non-highway budget will be very contractionary.

### Policy implications and recommendations (extracted)
- Fiscal policies should be appropriately counter-cyclical starting in 2018 as Montenegro undertakes fiscal adjustment during an economic expansion.
- To avoid pro-cyclical fiscal policymaking in the future, the authorities should:
  - Implement reforms to strengthen budget institutions, in particular medium-term budgetary frameworks.
  - Maintain sound fiscal policies that would facilitate strong financial market access.
- Interpretation caveat: conclusions about the magnitude of needed fiscal adjustment depend on whether highway spending is truly one-off and whether further phases are financed through the budget.

_International Monetary Fund — cr18122 References and Structural Fiscal Balances content (as provided)._

### 2018. In 2019 and 2020, the overall fiscal

### 2018. In 2019 and 2020, the overall fiscal

### Fiscal impulse and growth outlook
- Overall fiscal impulse (including the highway) will be negative in 2019 and 2020 as highway spending ends; the negative impulse is especially large in 2020.
- Authorities phased fiscal adjustment so the non-highway budget adjusts while the highway budget provides stimulus over 2017-20.
- If adjustment had been delayed until 2019 or 2020, economic growth would likely be much weaker in those years.
- Staff’s current economic growth projection: 3.0 percent.
- Staff notes downside risk to the 3.0 percent growth projection from the large contractionary fiscal impulse in 2020.
- Supply-side effects from the completed highway section will partly offset the drag, but staff estimates these effects will be relatively small because the first section of the highway will not provide economically significant transportation links.

### Policy recommendations (fiscal institutions and spending)
- Tax expenditures:
  - Ensure tax expenditures do not sever the link between GDP and revenue growth.
  - Examples of tax exemptions that could undermine revenue buoyancy: goods and services for the construction of five-star hotels and energy facilities are zero rated for VAT; marina services taxed at the lower VAT rate; newly established businesses in underdeveloped areas enjoy corporate income tax holidays.
  - Establish a framework to evaluate the cost of tax expenditures and consider whether policies provide sufficient economic benefits to justify their cost.
- Strengthen budget institutions to control pro-cyclical spending:
  - Medium-term budgetary framework:
    - Currently budget is adopted for one year with non-binding estimates for two more years.
    - Starting in 2019, the budget will be presented for a three-year period, including more details on expenditures, but expenditures remain only indicative for years two and three.
    - Recommendation: make medium-term expenditure levels more binding and force reconciliation/justification of changes relative to previous budget projections.
  - Medium-term projections:
    - Ministry of Finance needs to improve economic and demographic parameters for baseline revenue and expenditure projections (example: high-quality projections for pension expenditures).
  - Process for costing new expenditures:
    - Law on Budget and Fiscal Responsibility requires proposers of laws that lower tax receipts or increase expenditures to propose financing or define fiscal impact, but Law does not specifically mention the role of the Ministry of Finance.
    - Recommendation: any proposed law with fiscal implications should not be considered without analysis from the Ministry of Finance.
    - Example: social benefit for mothers approved in 2015 without analysis from the Ministry of Finance.
  - Investment spending:
    - Strengthen procedures to evaluate economic and social returns of public investment projects to choose projects with highest rates of return.
    - With consideration of completing the Bar-Boljare highway via PPP, strengthen PPP framework to avoid large contingent fiscal liabilities.
  - Expenditure ceiling:
    - Current rule: growth rate of current spending must be less than projected real GDP growth rate; capital spending growth cannot exceed nominal GDP growth.
    - If followed, current spending would continuously fall as a share of nominal GDP; in practice rules have not been consistently followed.
    - Recommendation: reform rules over the medium term; consider setting medium-term expenditure ceilings with automatic stabilizers excluded to avoid pro-cyclical spending.

### Fiscal adjustment, debt, and market access
- General government debt (including guarantees) at end-2017: 75 percent of GDP.
- Staff projection under full implementation of fiscal adjustment strategy and maintenance of a strong primary surplus:
  - Primary surplus peaking at 4½ percent of GDP in 2020 and falling towards 3 percent of GDP by 2023.
  - Debt lowered to 53 percent of GDP by 2023.
- Recommendation: implement reforms to lower public-sector wage bill and restrict early retirements to relieve expenditure pressures.
- Successful fiscal adjustment and medium-term reforms would facilitate strong market access, supporting counter-cyclical fiscal policies during downturns and potentially easing financing constraints to avoid pro-cyclical expenditure cuts.

### Reserve adequacy — overview
- Montenegro is a unilaterally euroized economy without a lender of last resort; reserve adequacy should consider balance of payments needs, buffers for fiscal financing, and bank emergency liquidity assistance (ELA).
- International reserves appear adequate relative to standard metrics for balance of payments purposes but central bank resources for ELA are limited and government deposits historically small.
- Recommendation: increase fiscal buffers and resources for ELA by building up government deposits and creating a government sub-account at the central bank for ELA purposes.

### International reserve levels and composition
- At end-2017, the Central Bank of Montenegro (CBM) held nearly €900 million (21 percent of GDP) in reserves, more than double end-2013 (€424 million or 13 percent of GDP).
- High growth in reserves traced mainly to an increase in domestic banks’ excess reserves held at the CBM.
- International reserves correspond mainly to banks’ required and excess reserves and central government deposits; CBM assets are principally marketable foreign securities, deposits in foreign banks and central banks, and cash.

### Traditional reserve metrics (end-2017)
- Standard rules of thumb: reserves should cover at least three months of imports, 100 percent of short-term external debt, and 20 percent of broad money (M2). Bank deposits substituted for broad money because of euroization.
- Table 1 (selected metrics and results):
  - Months of import cover (target): 3.0
    - Gross reserves: 3.7
    - Net of bank excess reserves: 1.6
  - Reserves to short-term external debt (target): 100.0
    - Gross reserves: 295.0
    - Net of bank excess reserves: 124.7
  - Bank Deposits (percent, target): 20.0
    - Gross reserves: 34.8
    - Net of bank excess reserves: 14.7
- Observation: reserves comfortably exceed these metrics when including bank excess reserves; net of excess reserves the coverage falls short on some metrics.

### ARA (multidimensional) metric assessment
- ARA metric specification used:
  - Original IMF form: Reserves = 10% of X + 30% of STD + 10% of BM + 20% of OPL (OPL omitted due to data gaps).
  - Modified for Montenegro: Reserves = 10% of X + 30% of STD + 15% of Deposits (broad money substituted by deposits and weight increased to 15%).
- Adequacy ranges: reserves in the range of 100 to 150 percent of the metric considered adequate.
- ARA results (end-2017):
  - Against standard and modified ARA concepts, end-2017 reserve levels are within or above the 100–150 percent range at 170 percent and 137 percent, respectively.
  - In a conservative scenario without banks’ excess reserves, coverage is between 72 percent (standard ARA) and 58 percent (modified Montenegro-specific concept) of the metric.
- Detailed breakdown from the "Potential sources of liquidity need" (ARA Concept / Modified ARA Concept):
  - Exports: 4.1 / 4.1
  - Short-term external debt: 2.3 / 2.3
  - Base money (bank deposits): 6.1 / 9.2
  - Target: 12.5 / 15.5
  - Without deposits of euro area banks: 8.4 / 9.4
  - Available resources — Gross international reserves: 21.3
  - Percent of metric: 170% / 137%
  - Percent of metric, without euro area banks: 253% / 226%
  - Net of bank excess reserves: 9.0
  - Percent of metric (net): 72% / 58%
  - Percent of metric, w/o euro area banks (net): 107% / 95%

### Effects of Euro Area bank presence
- Euro Area banks dominate the banking system and hold two-thirds of system deposits; they likely have access to ECB liquidity via parent banks.
- Excluding deposits of Euro Area banks from metrics substantially improves coverage: gross reserves coverage becomes 226–253 percent of the metric; net-of-excess-reserves coverage becomes 95–107 percent.

### Reserve projections and policy implications
- Staff baseline projections:
  - International reserves projected to increase by 5 percentage points to 26 percent of GDP in 2018, driven by increased government deposits at the CBM as it pre-finances coming Eurobond amortizations.
  - Over 2020-23, reserves may average 21 percent of GDP after deposits are drawn down in 2020 to pay Eurobonds.
  - Reserves in the range of 12-15 percent of GDP would be broadly adequate over the medium term according to both ARA metrics.
- Policy options if needed:
  - Adjust required reserves so that the modified ARA concept for Montenegro net of bank excess reserves but without banks with Euro Area parents is at least 100 percent (this condition holds for 2018 under staff’s assessment).
  - Increase government deposits at the CBM and create a government sub-account at the central bank for ELA purposes.

*International Monetary Fund*

### 11.      With all liabilities in foreign currency, fully euroized economies may also need greater

### 11.      With all liabilities in foreign currency, fully euroized economies may also need greater buffers for government financing.

### Fiscal buffers and government deposits
- Governments in fully euroized economies may wish to maintain additional fiscal savings as a buffer against fluctuations in revenue or spending, as funding in the adopted currency may be difficult during times of stress.
- In Montenegro, the central government holds the vast majority of its deposits at the CBM, and these deposits fund about 7 percent and 29 percent (projected) of the CBM’s international reserves in 2017 and 2018, respectively.
- A useful rule of thumb for the minimum size of fiscal reserves is one month of central government expenditures.
- The government’s deposits at the central bank have generally fallen short of the benchmark of one month of expenditures:
  - Average monthly deposit balance in 2017: €70 million (1.7 percent of GDP).
  - Benchmark one month of spending: €150 million (3.6 percent of GDP).
  - Average deposits in 2017 comprised less than half of the benchmark.
- Pre-financing in 2018 of coming Eurobond amortizations provides an opportunity to build larger fiscal buffers:
  - Successful liability management operation concluded in April 2018 and projected future borrowing with a World Bank Policy-Based Guarantee in 2019 should allow pre-financing of Eurobond amortizations in 2019 and 2020.
  - Government deposits may increase to nearly 8 percent of GDP in 2018 and 2019.
  - In 2020, when authorities draw down deposits to amortize the Eurobond, they should retain deposits worth one month of expenditures (3 percent of GDP).
- Policy recommendations on fiscal buffers:
  - Maintain a standing balance of deposits equivalent to at least one month of expenditures, weighing carrying costs against reduced risk of a financing crisis.
  - Encourage development of medium- to longer-term domestic government bond markets:
    - Authorities regularly issue T-bills with maturity up to one year (held mainly by banks); issuance of longer-term domestic government bonds has been infrequent and trading activity is limited.
    - A more regular offering of longer-maturity domestic bonds would diversify financing sources and tap high liquidity in the banking sector.
  - Maintain sound fiscal policies and fully implement the medium-term fiscal adjustment strategy to ensure government debt falls to safer levels and financing costs decline.

### Banking sector liquidity buffers and CBM role
- Banks in Montenegro are currently very liquid:
  - Overall ratio of liquid assets to short-term liabilities: 36 percent, in line with banks in other dollarized/euroized economies.
- Banks’ own liquid assets should be the first line of defense; however, individual bank outflows may exceed own liquid assets (which include excess reserves).
- CBM framework for granting liquidity loans to banks:
  - CBM has authority to grant liquidity loans against adequate collateral and has defined conditions; loans may only be granted to solvent banks.
  - Banks may draw up to 50 percent of their required reserves for intraday liquidity needs.
  - For further needs, CBM may grant intraday, overnight, or short-term liquidity loans (up to 180 days) with increasing interest rates for each type.
  - CBM generally expects subsidiaries of foreign banks to first turn to their parents; for a subsidiary to receive a short-term liquidity loan from the CBM, the parent must confirm inability to provide liquidity itself.
  - Allowing banks to use required reserves for a limited time is reasonable if the bank continues to be considered solvent.
- Composition of overall liquidity buffers at end-2017:
  - (1) Banks' liquid assets (including excess reserves and 50 percent of required reserves at the CBM).
  - (2) The 50 percent of banks’ required reserves that they cannot withdraw freely and without penalty.
  - In total, these buffers equaled 40 percent of short-term liabilities at the end of 2017.
- CBM’s own resources for Emergency Liquidity Assistance (ELA) are limited:
  - CBM’s own capital position is small at only 2 percent of banks’ short-term liabilities.
  - Even if short-term liabilities of banks with Euro Area parents are excluded, ELA resources would still only cover 8 percent of short-term liabilities for the remaining banks.
- Risks:
  - In the future, banks could become less liquid if credit growth accelerates.
  - If a bank exhausts its own liquid assets, collateral, and required reserves, CBM’s ELA capacity is constrained by limited capital.

### ELA policy options and recommendations
- Strengthening of ELA buffers will ultimately need to be funded and backstopped by the fiscal authorities because the CBM cannot issue domestic currency.
- Options considered and assessed:
  - Pooled liquidity arrangement (example: Ecuador’s liquidity fund with target of 10 percent of deposits financed by banks) has coordination and moral hazard issues in Montenegro; subsidiaries of foreign banks would likely not access ELA and thus would effectively provide liquidity to domestic banks, raising moral hazard concerns.
  - Liquid asset contribution (higher contributions from banks with larger liquid assets) would create perverse incentives against self-insurance.
- Potential viable options:
  - Explore feasibility of a standing line of credit from an international financial institution:
    - Such a line could augment existing ELA resources during a liquidity event.
    - Authorities should consider cost of the commitment fee and the institution’s ability to disburse quickly after a request.
    - Montenegro’s deposit insurance fund currently has a credit line with the EBRD until the fund is fully funded; authorities could explore whether the EBRD could provide a similar credit line for ELA purposes.
    - A foreign commercial bank could, in principle, provide a credit line but may resist disbursing during a broader liquidity event.
    - Euro Area central banks will not grant credit lines to the CBM because Montenegro’s euroization is not endorsed by the ECB.
  - Create a dedicated sub-account at the CBM for ELA purposes and maintain prudent fiscal buffers:
    - CBM would be able to use this account at its discretion with strong safeguards; government should create arrangement to reimburse the CBM for losses stemming from ELA upon an independent audit report.
    - Kosovo has a Special Reserve Fund at the Central Bank of Kosovo for ELA purposes; Montenegro should create such an account and commit to reimburse CBM for ELA losses.
    - The coming build-up of government deposits at the CBM provides an opportunity to fund such an account in principle, but given fungibility and cost, it may not be optimal to fund until need arises.
    - Careful consideration required whether banks need liquidity support, capital injections, or other bank resolution actions.
    - Authorities have drafted a law to implement the EU framework for bank resolution and recovery.
  - Follow best practices for banking supervision as the first line of defense and consider increasing reserve requirements to augment ELA buffers:
    - In 2017, the CBM lowered reserve requirements by two percentage points, from 9.5 to 7.5 percent and from 8.5 to 6.5 percent, depending on the type of deposits.
    - CBM could consider increasing reserve requirements in the future if it wished to augment buffers.
    - Most importantly, closely monitor banks and maintain high prudential standards to minimize risks of emergency liquidity situations.

### Labor market outcomes: stylized facts and policy implications
- Study overview:
  - Provides explanations for observed labor market outcomes in Montenegro, which are weak by European standards.
  - Discusses inactivity rate, (long-term) unemployment rate, and informal employment across dimensions: employment protection, labor taxation, pace of wage growth including minimum wage, and non-employment income support and activation policies.
  - Main conclusions:
    - Room to improve the current draft Labor Law to increase flexibility, formalization, and labor participation by modifying proposals related to establishment of employment relationship, internal organization and systematization act, and termination of employment.
    - Labor taxation reform options should consider reduction of the tax wedge, especially on low income earners, and a gradual withdrawal of social assistance and family benefits rather than a sudden decline after reaching a certain level of net income.
    - Active labor market policies (ALMPs) can reduce labor market mismatch in the short to medium term; education system is key for medium to long term reduction of mismatches.
- Inactivity rate:
  - Inactivity rate is about 37 percent.
  - This is lower than the average for Western Balkan countries (excluding Montenegro), but much higher than the average for EU member countries and New Member States.
  - Gap is larger for women; difference with EU average is nearly 14 percentage points.
  - Inactivity rate is more critical among youth (15–24 years) and people with low education attainment.
- Unemployment rate:
  - Current unemployment rate: about 16.1 percent.
  - Youth (15–24 years) unemployment rate: 32 percent.
  - Unemployment more critical for less educated people and the youth.
- Long-term unemployment:
  - About 80 percent of the unemployed in Montenegro were out of work for more than a year on average, reflecting structural rather than cyclical characteristics.
  - Low outflows from unemployment are the main source of high joblessness, indicating difficult labor market entry associated with rigidities, frictions, and education and training gaps.
  - Consequences of long-term unemployment include erosion of human capital, decline of productivity, and increased risk of social exclusion.
- Policy implications:
  - Improve Labor Law to increase flexibility and formalization of work.
  - Consider tax wedge reductions for low income earners and gradual withdrawal of social assistance and family benefits.
  - Use ALMPs to reduce short- to medium-term mismatches and strengthen education to address medium- to long-term mismatches.

*Source: IMF staff analysis as presented in the supplied content unit.*

### 4. Negative demographic trends

### 4. Negative demographic trends

### Demographic drivers of labor outcomes
- Most Western Balkan countries face demographic contraction or stagnation driven by outward migration and declining birth rates (World Bank, 2017).
- Population aging compounded these developments: decline more pronounced in the under 25 age group, while the number of those aged 55-64 increased significantly.
- Working age population has declined in recent years, especially for the age groups between 15 and 24 and between 25 and 54 years old.
- The decline in total population, 4.3 percent, was larger than the one in working age population, 4.0 percent, between 2010 and 2017.
- These trends are expected to continue and will pose further challenges.

### Implications
- Declining youth cohorts and rising shares of older age groups constrain labor supply, affect youth employment prospects, and increase pressures on pension and social systems.

---

### Informal Employment in Montenegro

### Key findings from UNDP National Human Development Report (2016)
- Based on Labor Force Survey data, staff calculations suggest that the ratio of informal employment to the labor force is around 33 percent since 2009, showing a slight increase in recent years (just above 35 percent in 2016).
- About 10 percent of the population (15+) works in the informal sector and half is inactive.
- Structure of employees by status:
  - 78 percent of employees are formally employed (67 percent with their full salary declared and 10 percent with only part of their salary declared).
  - 22 percent of employees are informally employed.
- Of those employed, 23 percent are informally employed, and 10 percent are paid partially in cash.
- About 70 percent of informally employed people are self-employed.
- Age and education profile:
  - Dominant group: informally employed people aged between 46 and 64.
  - 37 percent of self-employed are people without any education or primary school education only.
  - 6 percent have a college degree.
- The probability of engagement in the informal economy is higher for young people, less educated people, and the elderly. Women are less engaged in informal activities than men.
- Sectors with largest informal employment: agriculture and households as employers.

### Income and inequality within informal employment
- Income of informally employed people is, on average, almost 30 percent lower than the incomes of formally employed individuals.
- People employed informally with an employer earn 5 percent less than formally employed people.
- Undeclared self-employed people earn half of what registered self-employed earn.
- By income deciles, a high degree of inequality exists within lower-income deciles: income of informally employed people in the first decile would be equivalent to 36 percent of the income of formally employed in the first decile. The difference decreases for higher deciles.
- The report documents that the average net income of informally self-employed people is at the level of the national net minimum wage.

---

### How to explain these labor outcomes in Montenegro?

### Analytical framework
- OECD jobs strategy elements to include in employment policy analysis: (i) level of employment protection; (ii) level of labor taxation; (iii) pace of wage growth, including the minimum wage; and (iv) non-employment income support, activation policies, education, and training.

---

### Labor Market Regulations

### Employment Protection Legislation (EPL)
- Strict EPL reduces inflows into unemployment and outflows from it, contributes to long-term unemployment, affects youth and women employment prospects, and stimulates the use of fixed-term or temporary contracts (OECD evidence cited).
- Countries with stricter EPL tend to have a larger informal sector (Johnson, Kaufmann, and Zoido-Lobaton, 1998; Loayza, Oviedo, and Serven, 2005; and Fialova and Schneider, 2011).
- Strict EPL appears to undermine productivity growth (Scarpetta and Tressel, 2004; Micco and Pages, 2006; and Bassanini and Venn, 2007).

### EPL in Montenegro
- EPL in Montenegro is above the average of OECD countries.
- 2008 labor legislation reform increased flexibility primarily through regulations related to fixed-term contracts.
- 2011 reform led to a decline in flexibility due to limitation on the use of fixed-term contracts, although severance payments improved.
- Latest OECD’s EPL rigidity index: individual and collective strictness indicators are higher than the average for OECD and European countries.

### Proposed features of the new Labor Law (draft)
- Simplification of procedures for dismissal of employees.
- Extension of the duration of fixed-term contracts to 36 months (calculated in a cumulative way).
- Adoption of a legal Act for the internal organization and classification of jobs.
- Introduction of a mechanism for an amicable resolution of employment.

### Areas for further discussion and revision (to increase flexibility, declaration of work, and labor participation)
- Establishing employment relationships:
  - Allow an employee to work for more than one employer even if additional work is full time.
  - Current registration limitations under the Labor Law appear to hamper additional self-employment activity, thus providing “incentives” to under-report self-employment.
- Internal organization and systematization act:
  - Requirement that all firms need a Rulebook regarding internal organization and classification of jobs could burden SMEs; small firms could suffice with a brief description/specification of work in the employment contract.
- Termination of employment:
  - Draft provision: “Employment shall terminate by virtue of law when the employee reaches the age of 67 and minimum 15 years of pension insurance...” eliminates possibility to extend a contract even if employer and employee agree; counterproductive in a declining and aging population.
- Addressing informal employment:
  - Draft Law: if a job inspector finds a worker without a formal employment contract, the worker automatically receives an open-ended contract.
  - Compliance and enforcement are necessary but not sufficient to reduce informal employment; it is questionable whether this penalty alone will provide right incentives to comply.
  - Authorities should devise a system of incentives focused on formalization of economic activities, involving employment and tax policies (comprehensive strategy).

### General Collective Agreement (GCA) issues
- Under the GCA, negotiated provisions cover all workers in the economy; best practice is collective agreements apply only to employers and employees who agree to be bound.
- Collective agreements should be time bound, with possibility of extension with consent of parties.
- GCA currently links minimum wage to a set of coefficients depending on education level, potentially implying multiple minimum wages and modifying minimum wage via the basic amount or coefficients.

---

### Wage Costs

### Unit labor costs and productivity
- Measured in PPP in 2016, the highest wage levels were found in Montenegro and Bosnia and Herzegovina (World Bank, 2018).
- Real unit labor costs (ULCs) increased by 3 percent over the last 10 years (2008-17), but have declined in recent years.
- Average wage, in real terms, has declined by 5 percent between 2010 and 2017.
- Labor productivity increased 4 percent during the same period.
- Productivity increase is modest from a medium- to long-term perspective (since 2008) and lags growth observed in other Western Balkan countries such as Albania and Bosnia and Herzegovina.

---

### Minimum Wage

### Levels and incidence
- The minimum wage in Montenegro appears moderate by some measures.
- Currently, the minimum wage is just below 40 percent of the average wage in Montenegro.
- A joint ILO, OECD, IMF, and World Bank report concludes that a minimum wage of 30-40 percent of the median wage would strike a suitable balance, which corresponds to a range of 25-35 percent below the average wage; i.e., the minimum wage is slightly below this range.
- Relative to GDP per capita, Montenegro’s minimum wage ranks high across European countries (denominator affected by large proportion of inactivity).

### Distributional and labor market impacts
- Different minimum wage levels based on education (coefficients range from 1.00, unqualified labor, to 4.12, Ph.D. education level) may have adverse impact on hiring decisions, especially for new labor market entrants with higher education; may create multiple minimum wages.
- Minimum wage mainly influences workers in the public services sector; in the private sector it is not binding across the whole formal sector.
- Some employers pay the minimum wage officially along with an additional undeclared amount; an increase in the minimum wage could raise the effective tax burden and lead to reshuffling between official and unofficial payment without affecting actual remuneration.
- Workers in the informal sector at lower income deciles would not be protected by the minimum wage: gap between formal and informal income at lower-income deciles is between 50 and 65 percent difference for the lowest three income deciles.
- Minimum wage might price out low-productivity informal workers rather than protect them.

---

### Taxes on labor

### Levels and structure
- Tax wedge on labor is around 40 percent in Montenegro and largely independent of income level (lack of progressivity) despite a low personal income tax rate.
- Personal income tax is flat at 9 percent with no personal allowance and only a marginal degree of progressivity.
- Social security contributions are also flat, adding to a combined tax burden of 33.8 percent of gross wages (there are virtually no deductions from wage taxes, but certain sources are exempt).
- Absence of a basic allowance and very little progressivity of the PIT result in little variation in the total tax wedge, which stands slightly higher than the EU average.
- Para-fiscal and local charges widen the effective labor tax wedge as perceived by employers (compulsory levels for the chamber of commerce, earmarked labor fund for covering severance pay in bankruptcies, professional rehabilitation fund). Health insurance contributions are phased out at certain income levels.

### Incentive effects and interactions with social benefits
- A relatively high effective tax wedge on labor exerts significantly adverse labor market incentives: discourages formal market labor supply and/or labor demand and contributes to involuntary unemployment and inactivity.
- Social benefits and transfers are often completely foregone or severely taxed away when workers enter the formal market, implying high effective marginal tax rates for low-income individuals.
- Example: for a family that receives family benefits, declaring any income of more than around €1,300 annually (around 15 percent of average wage) could result in the complete loss of family benefits – a loss in net income of about €280.
- Design of social assistance and family benefits does not provide incentives to declare work or to work because social assistance benefits are withdrawn at a high rate as formal gross income increases.

### Reform options recommended
- Consider reduction of the tax wedge, especially on low income earners, and gradual withdrawal of social assistance and family benefits as part of a comprehensive reform.
- Options to reduce incentives for under-declaration of income:
  - Re-introduction of a basic allowance for wage income.
  - Targeted earned income tax credits.
  - Lower employer contributions (e.g., finance health insurance through the general budget rather than by contributions).
- Beneficiaries of social assistance and family benefits should not be penalized at low income levels when entering the formal labor market; they should gain from joining the formal market (any additional formal wage should increase their net income including benefits).

---

*Source: IMF staff summary of “4. Negative demographic trends” (cr18122 - 4. Negative demographic trends).*

### 22. The structure of the Montenegrin economy has experienced significant changes over

### 22. The structure of the Montenegrin economy has experienced significant changes over the fifteen years

### Structural change and employment composition
- The structure of the economy has gone through "profound changes, characterized by a continuous process of deindustrialization and fast expansion of services."  
- By 2013, the share of employees in the service sector had reached 77 percent, while the share of employees in the industrial sector had declined to 18 percent.
- Growth relied mainly on labor-intensive activities, implying that rising employment rates did not necessarily reflect a structural shift toward higher productivity jobs.
- Structural change reduced demand for occupations related to industry and lengthened job search, particularly in labor-intensive sectors.

### Unemployment of higher-education graduates and job matching
- The overall unemployment rate and the unemployment rate of higher-education graduates in 2014 were about twice as high as in the EU-28.
- Currently unemployed graduates have been unemployed on average for over a year: 13.5 months.
- Graduates receive little assistance in job search beyond family and friends and often lack relevant work experience gained during higher education.
- To mitigate mismatch, the Employment Office implemented a program for professional training of university graduates in 2013 involving 4,000 participants (UNDP, 2016).
- Recommendation: Foster cooperation between employers and higher education institutions to involve employers in curricula development and communicate labor market priorities.

### Active labor market policies (ALMPs) and coverage
- Over the last decade, the Employment Office implemented ALMPs with objectives of reducing unemployment and mismatches; programs include:
  - professional rehabilitation of persons with disabilities and other hard-to-employ persons;
  - adult education and training;
  - training programs for independent work;
  - training programs for work with the employer;
  - public works;
  - activation of users of family social benefits;
  - stimulating the first employment of persons with acquired secondary education;
  - a project financing/grant scheme.
- These programs target training, employment incentives, supported employment and professional rehabilitation, direct job creation, and start-up incentives.
- On average, 6 percent of unemployed people were covered by these programs between 2014 and 2016.
- According to Drobnič et al. (2017), the percentage of program participants that were employed in 2016 ranged from 5 percent to 43 percent.

### Monitoring, evaluation, and recommendations for ALMPs
- Implementation will benefit from strengthened monitoring and assessment to determine whether primary objectives are met or whether targeting, scope, funding, or coordination should change.
- Drobnič et al. (2017) recommendations specific to Montenegro:
  - Active labor market measures of training (by occupation or by skills/knowledge/competency) should be based on mismatch between demand and supply.
  - Analytical reports on positioning of certain groups in the labor market should be obligatory when designing active measures.
  - An ex-ante evaluation is required to establish adequate scope in creating and designing measures.
  - Comprehensive evaluations (process, results, preliminary assessments) should be carried out.
  - There should be indicators for measuring performance of results (both quantitative and qualitative) versus the goals.

### Inactivity, retirement, and disability pensions
- Main explanations for labor inactivity (per LFS) are retirement and education or training.
- The inactivity rate for the 55–64 age group is about 56 percent, higher than other European countries and slightly above the Western Balkan average of about 54 percent.
- Authorities should assess whether loopholes allow early retirement via disability pension schemes; strict eligibility criteria are important to provide incentives to remain in the labor force.

### Family benefits, parental leave, and female labor force participation
- Publicly financed maternal/parental leave programs can help keep mothers connected to the labor market but very long paid leave can cause skill loss and jeopardize future employment.
- Jaumotte (2004) finds parental leaves (up to a duration of 20 weeks) and child care subsidies raise female labor force participation.
- In most OECD countries, the sum of paid maternity and paternity leave is below 30 weeks.
- In Montenegro, parental leave allows one parent to be paid while absent from work for 365 days (52 weeks) from the birth of the child if caring for or nursing a child.
- An employed woman may start paid maternity leave 45 days prior to delivery, and 28 days prior to delivery as mandatory leave.
- Atoyan and Rahman (2017) note family leave policies should (i) not create incentives for women to stay too long away from work causing skills loss and (ii) ensure possibility or mandate for fathers to take leave.
- Policy considerations:
  - Consider reducing the length of paid parental leave to bring it closer to most OECD countries.
  - Require a minimum leave period for fathers.
  - Allow the prospective mother discretion to take part of maternity leave prior to delivery.

### Child care availability
- Public child care units increased from 99 to 122 between 2010 and 2017.
- Private child care units increased from 9 to 25 during the same period, possibly indicating unmet demand.
- Expanding child care options, together with subsidies and affordable services, could increase labor force participation of new parents, mainly affecting female participation.

### Emigration and remittances
- During 1995–2010, countries in the Western Balkans have lost up to 18 percent of their population to emigration, mostly men of prime age with average or above average educational levels (brain drain).
- Remittances could negatively impact employment and labor force participation of those who stay by increasing reservation wages, pushing men into informal employment and women into inactivity due to family responsibilities.

### Conclusions and reform priorities
- Neither the new Labor Law nor changes in labor taxation alone would significantly improve labor market outcomes; a comprehensive and unified strategy is needed to address inactivity, long-term unemployment, and informality.
- Close coordination is needed between the Ministry of Finance and the Ministry of Labor on employment policy, enforcement of the Labor Code, labor taxes, and social benefits.
- Possible improvements to the Labor Law include:
  - Allowing an employee to work for more than one employer even if the additional work is full time (establishing an employment relationship).
  - Tailoring internal organization and systematization acts for SMEs to reduce burdens, particularly for small firms.
  - Addressing informal employment through incentives for formalization in addition to compliance and enforcement.
- Reform options on taxation and benefits:
  - Reduce the tax wedge, especially on low income earners, with a gradual withdrawal of social assistance and family benefits to avoid poverty traps.
  - Re-introduce a basic allowance for wage income and/or targeted earned income tax credits or lower employer contributions.
  - Consider financing health insurance through the general budget rather than by employee and employer contributions.
  - Ensure beneficiaries of social assistance and family benefits are not penalized when entering the formal labor market.
- Continued labor market reforms should:
  - Maintain National Strategies for Employment and Human Resources Development to increase human capital investment and employment.
  - Assess medium-term labor market needs to better align education with future labor demands and avoid costly retraining.
  - Strengthen cooperation between higher education institutions and employers for curricula design.
  - Use ALMPs to reduce mismatches in the short to medium term while relying on the education system for medium to long-term mismatch reduction.
  - Reduce paid parental leave and increase child care options to raise young parents' labor market participation, particularly for women.

*International Monetary Fund — Chapter 22, Montenegro Country Report*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2018/cr18122.pdf_