## 1.  Public Transport Infrastructure Investment in Western Balkan Countries

## Source details

**Canonical URL:** [1.  Public Transport Infrastructure Investment in Western Balkan Countries](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2016/_cr16357.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2016/_cr16357.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2016/_cr16357.pdf.json)

---

### Abstract and introduction
- FYR Macedonia has a strong need for improvements in its transport infrastructure; exports are expected to play an important role in driving long-term economic growth.
- During 2010–15, net exports contributed on average to 0.04 percent in annual GDP growth.
- Historical public expenditure on transport infrastructure (including construction and maintenance) averaged 1 percent of GDP between 2005 and 2013.
- For 2014–2018, annual investments in road infrastructure are expected to reach 2–3 percent of GDP per year.
- Four ongoing major projects (three highway and one railway) are the focus of the analysis; these projects are mostly financed by the EBRD, the EIB and the China EXIM Bank through publicly guaranteed debt denominated in FX and are carried out by state-owned enterprises outside the central government budget.
- Key transport and infrastructure indicators cited (as presented):
  - Road Density (km per 100 square kilometer of land area): charted (WDI 2011).
  - Paved Roads in Percent of Total Road (percent): charted (WDI 2011).
  - Logistics Performance Score (higher score = better performance): charted (World Bank International Logistics Performance Report, 2014).
  - Transport Infrastructure Score (higher score = better performance): charted (World Bank Logistics Performance Report, 2014).
  - Quality of Roads (higher score = better performance): charted (World Economic Forum Global Competitiveness Report, 2015).

### Selected examples in Western Balkans (Box 1)
- Montenegro:
  - Building the Bar-Boljare highway; investment in the first 40 km amounts to €809 million, or almost a quarter of annual nominal GDP.
  - Authorities intend more projects including remaining 176 km of Bar-Boljare highway, the “Trans-Adriatic” highway, railways and port developments.
- Kosovo:
  - Completed the R7 highways project in 2014 with a total cost of €973 million, or around 18 percent of GDP.
  - Constructing the R6 Road connecting Pristina to the Macedonian border with total cost estimated at €668 million, or 12 percent of GDP; construction began in 2015 and is scheduled to be completed in 2018–19.
- Serbia:
  - In 2015 completed construction of a total of 91.1 km of motorways, refurbished some 190 km of roads, built 206 bridges and nine tunnels.
  - Planned completion in the year of a total of 90.7 km of motorways on the pan-European Corridor X and intensified works on Corridor XI with an estimated total annual investment of €115.5 million.

### Methodology
- Quantitative analysis uses a model-based simulation approach built on the standard neoclassical growth model augmented with public capital as a distinct factor of production (in addition to labor and private capital).
- New public capital is assumed to be formed and available for use only when a project is completed.
- Because many projects are financed outside the budget and are budget-neutral, the model abstracts away from fiscal policy considerations (assumes carrying out these projects will not oblige the government to reduce spending in other areas).
- The simulation treats the announcement of plans as an unanticipated shock at end-2013 and simulates the economy since 2014.
- Two types of growth effects captured:
  - Direct effect: additional investment on GDP, net of investment-related imports.
  - Indirect effect: private sector response (higher private investment and consumption) in anticipation of higher public capital stock.
- Public investment efficiency is modeled as the share of investment that actually goes into the formation of public capital (example: 80 percent efficiency implies every million dollars of investment adds 800,000 dollars’ worth of public capital stock).
- Efficiency calibration:
  - South Africa (best performer) overall score: 3.53 → defined as 100 percent efficiency.
  - FYR Macedonia overall score: 1.93 → implied public investment efficiency = 1.93/3.53*100 percent = 55 percent.
- The model allows for diminishing returns to public investment and uses country-specific calibration and investment efficiency rather than assuming linear cross-country effects.

### Quantitative results — growth effects and dynamics
- Planned public investment in the four mega projects studied amounts to 2–3 percent of GDP per year for 2014–2018.
- Estimated impacts:
  - Average annual growth rate increase of 0.5 percentage points annually for the period 2014–2020 from the projects studied.
  - Direct short-term contribution from these projects is estimated to average 0.25 percent of GDP annually for 2014–2018 (accounting for half of the total growth effect due to low import content).
  - The remaining half of the estimated annual growth effect comes from spillovers in higher private investment and consumption.
  - Contemporaneous effect of a one percent of GDP increase in public investment is about 0.3 percent increase in output in this model (compared with 0.25 reported in WEO 2014 cross-country regression).
  - In terms of the level of output: by end-2020 real GDP would be 4 percent higher than the hypothetical scenario without investments in the highway and railway projects.
  - Private sector accumulates 8 percent more capital stock by 2020 relative to the scenario without infrastructure investments.
- Transition dynamics:
  - After projects are completed in 2018, private sector investment adjustments are assumed to continue for another three years before the economy reaches its new growth trajectory and growth rates revert to long-run trend.
  - The growth increase is transient; the level effect is permanent (economy sits on a trajectory permanently higher by 4 percent relative to no-investment scenario).
- Non-linearity and limits:
  - Non-linearity due to decreasing returns to public capital implies results should not be linearly extrapolated for additional future infrastructure projects beyond those studied.

### Efficiency gains and sensitivity
- Raising FYR Macedonia’s public investment efficiency to the level of South Africa (top performer) would generate an annual growth effect of 0.8 percent, instead of 0.5 percent.
- Improved efficiency allows the same effective public capital formation with less nominal spending, reducing financing needs and accrued debt.
- FYR Macedonia’s performance across four sub-areas of public investment efficiency:
  - Scores relatively well in selection and evaluation.
  - Scores poorly in appraisal and managing.
- Priority reform areas recommended to enhance public investment efficiency:
  - Improve capabilities in project appraisal.
  - Improve capabilities in project managing.

### Costs, fiscal risks, and maintenance needs
- Publicly-guaranteed debt of the PESR, the main implementer of public infrastructure projects:
  - 1.1 percent of GDP when the entity was formed in 2013.
  - 3.9 percent of GDP by 2015.
  - Expected to grow to more than 7 percent of GDP by 2018.
- Public debt dynamics:
  - Public debt was 46 percent of GDP (current level referenced).
  - Public debt is expected to climb to 54 percent of GDP by 2021.
  - Public debt is double what it used to be in 2008.
- Currency-denomination and timing risks:
  - Debt is mostly denominated in FX; currency mismatch could present significant fiscal risks if the stabilized exchange rate arrangement comes under pressure.
  - Debt service is projected to pick up from 2019 onwards.
- Maintenance and recurrent fiscal needs:
  - Highways and railways will require fiscal resources for maintenance work that need to be budgeted to reap long-term benefits.
  - World Bank Public Expenditure Report estimates annual maintenance cost of 5000–7000 USD per kilometer will be required to reduce the existing maintenance backlog and to ensure that future maintenance needs are met.
- Policy trade-offs and conclusion:
  - While the estimated growth effect is sizable (0.5 percentage points on average for 2014–2020), plans for subsequent projects need to be weighed against the rising public debt level and the inevitable diminishing returns to investment.
  - Public investment efficiency in FYR Macedonia is decent by regional standards; improvements in this area can deliver sizable gains in growth.

### Technical annex — model description and calibration
- Model framework:
  - Based on the standard Neo-classical growth model with representative household maximizing lifetime utility.
  - Household problem: max U = Σ β^t u(c_t) subject to c_t + k_{t+1} = y_t + (1−δ) k_t, where y_t is household income (output of private firms).
  - Labor supply is inelastic and normalized to one.
  - Production function includes public infrastructure as a distinct factor of production (public capital K^G_t):
    - Y_t = K_t^α L_t^{1−α} (K^G_t)^γ.
- Steady state private capital stock expression (as presented in source notation):
  - K_ss = [1/(β^{−1}+δ) (K^G_ss)^γ / α]^{1/(α−1)}
- Calibration parameters (taken from literature and calibrated to match data moments):
  - β = 0.96
  - δ = 0.1
  - α calibrated to match long-run consumption to GDP ratio: α = 0.35
  - γ calibrated to the long-run elasticity of output to public capital (WEO October 2014 Chapter 3): γ = 0.12

### Technical annex — public capital stock construction and assumptions
- Construction approach:
  - Public infrastructure capital stock constructed from annual nominal investment amount for each project.
  - Nominal investment adjusted for inflation and investment efficiency to estimate real effective investment.
- Investment efficiency assumptions:
  - South Africa’s efficiency is assumed to be 1.
  - FYR Macedonia’s public investment efficiency score is 55 percent of South Africa’s, hence FYR Macedonia’s efficiency set at 0.55.
- Depreciation and baseline stock:
  - Assume new public capital is formed only when a project is completed.
  - Public infrastructure spending in 2013 assumed to be 1 percent of GDP spent on restoring depreciated capital.
  - Depreciation rate assumed to be 4 percent (WP/07/202).
  - Implied infrastructure capital stock in 2013 is 25 percent of GDP.
  - Infrastructure investments studied render public capital stock in 2017 to be 27 percent higher than that in 2013.

### FDI spillovers on growth and employment — key findings
- FDI inflows and policies:
  - FDI incentives and improved business environment helped attract significant FDI inflows over the past decade, though less so than other countries in the region.
  - By 2014, FDI stock had grown nearly tenfold from that of 2000 (from a low base).
- Composition and sources of FDI:
  - Largest FDI flows from EU countries; Germany, Austria, and the Netherlands accounted for about a third of total FDI inflows.
  - Flows were mostly equity during 2003–2013, after which they partly switched to inter-company loans.
  - Flows mostly for greenfield projects; largely into manufacturing and financial and insurance sectors.
- Direct contributions and fiscal impacts:
  - Exports from TIDZ about 42 percent of total goods exports; imports to the zone around 25 percent; overall net exports contributed to 2.6 percent of GDP during 2014–15 (NBRM estimates).
  - Domestic vendor purchases accounted for only 0.4 percent of GDP in 2015.
  - Employment of FDI companies about 2.5 percent of total employment as of September 2016; at full capacity employment would reach about 4.7 percent of total employment.
  - Companies within the TIDZ paid social security contributions and personal income tax amounting to MKD 1.1 billion during 2011–2015, or approximately 0.16 percent of total tax revenues and contributions per year.
  - A recent study shows companies in the TIDZ accounted for only one percent of total gross value added in 2013.
- Empirical estimates of FDI impact on growth and employment:
  - Cross-country regression (20 CESEE countries, 1991–2014) finds FDI inflows have a positive impact on growth in most specifications.
  - Estimated coefficient indicates real per capita GDP growth could rise by about 0.5 percentage point as FDI inflows increase by one percent of GDP.
  - Growth decomposition suggests FDI inflows likely contributed to annual real per capita GDP growth for FYR Macedonia of about 1.3 percent on average between 2010 and 2013.
  - Labor market evidence: an increase in FDI inflows per capita by 10 percent would reduce the unemployment rate by 2–3 percent for FYR Macedonia (based on multinomial logit estimates from Atoyan and Jankulov (2015) and micro-level labor force survey data).
- Caveats and heterogeneity:
  - Literature provides mixed evidence on FDI–growth and FDI–employment relationships; results depend on methodology, endogeneity, and type of FDI.
  - The study does not differentiate FDI by type or industry; high-end manufacturing FDI may have different impacts than real estate FDI.
  - Institutional quality matters: government effectiveness indicator ranges approximately from −2.5 (weak) to 2.5 (strong); FYR Macedonia’s government effectiveness was 0.15 in 2014, below the CESEE average of 0.25.

### Labor market reforms, scenario analysis, and policy implications
- Macedonia performs well in labor market efficiency and flexibility but has room for improvement in female participation in the labor force and reliance on professional management.
- Scenario analysis (labor market outcome model in Annex II) shows:
  - Baseline, 2014
  - 10% increase in FDI per capita
  - 10% increase in FDI and labor market reforms
- Unemployment axis values illustrated: 24.0, 24.5, 25.0, 25.5, 26.0, 26.5, 27.0, 27.5, 28.0, 28.5 (Percent).
- Indirect impacts and vulnerabilities:
  - Backward linkages with domestic firms appear limited due to technical and safety requirements.
  - Outflows of high skilled labor could hinder movement up economic and technological ladders.
  - Costs of FDI incentives arise through forgone tax revenues (profit tax and personal income tax exemptions for ten years), subsidies and grants; precise costs are not calculated due to missing firm-level data.
  - Debt financing of FDI has been rising; a fast rise in debt-financed FDI could create external vulnerabilities.
- Policy implications:
  - For FDI to play a stronger role in tackling unemployment, financial benefits to investors need to be complemented by labor market and institutional reforms.
  - FDI incentives generate fiscal costs and vulnerabilities which need monitoring.

*Prepared by Hua Chai; IMF staff analysis as presented in the source PDF (content unit _cr16357).*

### 1.  Public Transport Infrastructure Investment in Western Balkan Countries ______________ 5

### 1.  Public Transport Infrastructure Investment in Western Balkan Countries

### Abstract and Introduction
- FYR Macedonia has a strong need for improvements in its transport infrastructure; exports are expected to play an important role in driving long-term economic growth.
- During 2010–15, net exports contributed on average to 0.04 percent in annual GDP growth.
- Historical public expenditure on transport infrastructure (including construction and maintenance) averaged 1 percent of GDP between 2005 and 2013.
- For 2014–2018, annual investments in road infrastructure are expected to reach 2–3 percent of GDP per year.
- Four ongoing major projects (three highway and one railway) form the focus of the analysis; these projects are mostly financed by the EBRD, the EIB and the China EXIM Bank through publicly guaranteed debt denominated in FX and are carried out by state-owned enterprises outside the central government budget.
- Key transport and infrastructure indicators cited:
  - Road Density (km per 100 square kilometer of land area): charted (WDI 2011).
  - Paved Roads in Percent of Total Road (percent): charted (WDI 2011).
  - Logistics Performance Score (higher score = better performance): charted (World Bank International Logistics Performance Report, 2014).
  - Transport Infrastructure Score (higher score = better performance): charted (World Bank Logistics Performance Report, 2014).
  - Quality of Roads (higher score = better performance): charted (World Economic Forum Global Competitiveness Report, 2015).

### Box 1 — Public Transport Infrastructure Investment in Western Balkan Countries (selected examples)
- Montenegro:
  - Building the Bar-Boljare highway; investment in the first 40 km amounts to €809 million, or almost a quarter of annual nominal GDP.
  - Authorities intend more projects including remaining 176 km of Bar-Boljare highway, the “Trans-Adriatic” highway, railways and port developments.
- Kosovo:
  - Completed the R7 highways project in 2014 with a total cost of €973 million, or around 18 percent of GDP.
  - Constructing the R6 Road connecting Pristina to the Macedonian border with total cost estimated at €668 million, or 12 percent of GDP; construction began in 2015 and is scheduled to be completed in 2018–19.
- Serbia:
  - In 2015 completed construction of a total of 91.1 km of motorways, refurbished some 190 km of roads, built 206 bridges and nine tunnels.
  - Planned completion in the year of a total of 90.7 km of motorways on the pan-European Corridor X and intensified works on Corridor XI with an estimated total annual investment of €115.5 million.

### Major Highway and Railway Projects under Construction (project details)
- Motorway Section / Total investment / Foreign creditor / Project start / Expected completion
  - Miladinovci - Sveti Nikole - Shtip: €226.7m / EXIM Bank of China / May-14 / May-17
  - Kichevo - Ohrid: €411.3m / EXIM Bank of China / May-14 / Jan-18
  - Demir Kapija - Smokvica: €245.0m / EBRD/EIB / Sep-12 / May-17
  - Kriva Palanka - Beljakovce: €140.0m / EBRD / May-14 / Jan-18

### Methodology
- Quantitative analysis uses a model-based simulation approach built on the standard neoclassical growth model augmented with public capital as a distinct factor of production (in addition to labor and private capital).
- New public capital is assumed to be formed and available for use only when a project is completed.
- Because many projects are financed outside the budget and are budget-neutral, the model abstracts away from fiscal policy considerations (assumes carrying out these projects will not oblige the government to reduce spending in other areas).
- The simulation treats the announcement of plans as an unanticipated shock at end-2013 and simulates the economy since 2014.
- Two types of growth effects captured:
  - Direct effect: additional investment on GDP, net of investment-related imports.
  - Indirect effect: private sector response (higher private investment and consumption) in anticipation of higher public capital stock.
- Public investment efficiency is modeled as the share of investment that actually goes into the formation of public capital (example: 80 percent efficiency implies every million dollars of investment adds 800,000 dollars’ worth of public capital stock).
- Efficiency calibration:
  - South Africa (best performer) overall score: 3.53 → defined as 100 percent efficiency.
  - FYR Macedonia overall score: 1.93 → implied public investment efficiency = 1.93/3.53*100 percent = 55 percent.
- The model allows for diminishing returns to public investment and uses country-specific calibration and investment efficiency rather than assuming linear cross-country effects.

### Quantitative Results — Growth Effects and Dynamics
- Planned public investment in the four mega projects studied amounts to 2–3 percent of GDP per year for 2014–2018.
- Estimated impacts:
  - Average annual growth rate increase of 0.5 percentage points annually for the period 2014–2020 from the projects studied.
  - Direct short-term contribution from these projects is estimated to average 0.25 percent of GDP annually for 2014–2018 (accounting for half of the total growth effect due to low import content).
  - The remaining half of the estimated annual growth effect comes from spillovers in higher private investment and consumption.
  - Contemporaneous effect of a one percent of GDP increase in public investment is about 0.3 percent increase in output in this model (compared with 0.25 reported in WEO 2014 cross-country regression).
  - In terms of the level of output: by end-2020 real GDP would be 4 percent higher than the hypothetical scenario without investments in the highway and railway projects.
  - Private sector accumulates 8 percent more capital stock by 2020 relative to the scenario without infrastructure investments.
- Transition dynamics:
  - After projects are completed in 2018, private sector investment adjustments are assumed to continue for another three years before the economy reaches its new growth trajectory and growth rates revert to long-run trend.
  - The growth increase is transient; the level effect is permanent (economy sits on a trajectory permanently higher by 4 percent relative to no-investment scenario).
- Non-linearity and limits:
  - Non-linearity due to decreasing returns to public capital implies results should not be linearly extrapolated for additional future infrastructure projects beyond those studied.

### Efficiency Gains and Sensitivity
- Raising FYR Macedonia’s public investment efficiency to the level of South Africa (top performer) would generate an annual growth effect of 0.8 percent, instead of 0.5 percent.
- Improved efficiency allows the same effective public capital formation with less nominal spending, reducing financing needs and accrued debt.
- FYR Macedonia’s performance across four sub-areas of public investment efficiency:
  - Scores relatively well in selection and evaluation.
  - Scores poorly in appraisal and managing.
- Priority reform areas recommended to enhance public investment efficiency: improve capabilities in project appraisal and managing.

### Costs and Risks
- Public infrastructure investments increase foreign currency denominated public debt and public liabilities, even when financed outside the central government budget.
- Potential negative growth channels (not judged dominant here):
  - Public borrowing could crowd out credit to the private sector; considered likely muted for FYR Macedonia because the public sector borrows primarily from foreign creditors that the private sector does not have access to.
- Fiscal and balance of payments implications:
  - Higher level of economic activity may boost exports and tourism, improving the balance of payments.

*Prepared by Hua Chai; IMF staff analysis as presented in the source PDF.*

### 12. Even though FYR Macedonia’s public transport infrastructure investments is expected to

### _cr16357 - 12. Even though FYR Macedonia’s public transport infrastructure investments is expected to

### Costs and fiscal risks of large-scale public infrastructure projects
- Publicly-guaranteed debt of the PESR, the main implementer of public infrastructure projects:
  - 1.1 percent of GDP when the entity was formed in 2013.
  - 3.9 percent of GDP by 2015.
  - Expected to grow to more than 7 percent of GDP by 2018.
- Public debt dynamics:
  - Public debt was 46 percent of GDP (current level referenced).
  - Public debt is expected to climb to 54 percent of GDP by 2021.
  - Public debt is double what it used to be in 2008.
- Currency-denomination and timing risks:
  - Debt is mostly denominated in FX; currency mismatch could present significant fiscal risks if the stabilized exchange rate arrangement comes under pressure.
  - Debt service is projected to pick up from 2019 onwards.
- Maintenance and recurrent fiscal needs:
  - Highways and railways will require fiscal resources for maintenance work that need to be budgeted to reap long-term benefits.
  - World Bank Public Expenditure Report estimates annual maintenance cost of 5000–7000 USD per kilometer will be required to reduce the existing maintenance backlog and to ensure that future maintenance needs are met.

### Growth effects and policy trade-offs (Conclusion)
- Investment scale and timing studied:
  - Annual investments in the four highway and railway projects of 2–3 percent of GDP for 2014–18.
- Estimated growth impact:
  - Likely to raise real GDP growth rate by 0.5 percentage points on average for each year in 2014–2020.
- Policy considerations:
  - While the estimated growth effect is sizable, plans for subsequent projects need to be weighed against the rising public debt level and the inevitable diminishing returns to investment.
  - Public investment efficiency in FYR Macedonia is decent by regional standards; improvements in this area can deliver sizable gains in growth.

### Quantitative evidence and visualization (as presented)
- PESR Debt and Amortization (percent of GDP) series shown for years 2013–2021 indicating rising PESR debt stock and PESR amortization.
- Composition of Public Debt by Currency (percent of GDP) shown for years 2007–2021 with breakdown:
  - Foreign currency-denominated
  - Local currency-denominated

### Technical annex — model description
- Model framework:
  - Based on the standard Neo-classical growth model with representative household maximizing lifetime utility.
  - Household problem: max U = Σ β^t u(c_t) subject to c_t + k_{t+1} = y_t + (1−δ) k_t, where y_t is household income (output of private firms).
  - Labor supply is inelastic and normalized to one.
  - Production function includes public infrastructure as a distinct factor of production (public capital K^G_t):
    - Y_t = K_t^α L_t^{1−α} (K^G_t)^γ.
- Steady state private capital stock expression:
  - K_ss = [1/(β^{−1}+δ) (K^G_ss)^γ / α]^{1/(α−1)} (presented in source notation).

### Technical annex — calibration
- Parameters taken from literature and calibrated to match data moments:
  - β = 0.96
  - δ = 0.1
  - α calibrated to match long-run consumption to GDP ratio: α = 0.35
  - γ calibrated to the long-run elasticity of output to public capital (WEO October 2014 Chapter 3): γ = 0.12

### Technical annex — public capital stock construction and assumptions
- Construction approach:
  - Public infrastructure capital stock constructed from annual nominal investment amount for each project.
  - Nominal investment adjusted for inflation and investment efficiency to estimate real effective investment.
- Investment efficiency assumptions:
  - South Africa’s efficiency is assumed to be 1.
  - FYR Macedonia’s public investment efficiency score is 55 percent of South Africa’s, hence FYR Macedonia’s efficiency set at 0.55.
- Depreciation and baseline stock:
  - Assume new public capital is formed only when a project is completed.
  - Public infrastructure spending in 2013 assumed to be 1 percent of GDP spent on restoring depreciated capital.
  - Depreciation rate assumed to be 4 percent (WP/07/202).
  - Implied infrastructure capital stock in 2013 is 25 percent of GDP.
  - Infrastructure investments studied render public capital stock in 2017 to be 27 percent higher than that in 2013.

### FDI spillovers on growth and employment — key findings
- FDI inflows and policies:
  - FDI incentives and improved business environment helped attract significant FDI inflows over the past decade, though less so than other countries in the region.
  - By 2014, FDI stock had grown nearly tenfold from that of 2000 (from a low base).
- Composition and sources of FDI:
  - Largest FDI flows from EU countries; Germany, Austria, and the Netherlands accounted for about a third of total FDI inflows.
  - Flows were mostly equity during 2003–2013, after which they partly switched to inter-company loans.
  - Flows mostly for greenfield projects; largely into manufacturing and financial and insurance sectors.
- Direct contributions and fiscal impacts:
  - Exports from TIDZ about 42 percent of total goods exports; imports to the zone around 25 percent; overall net exports contributed to 2.6 percent of GDP during 2014–15 (NBRM estimates).
  - Domestic vendor purchases accounted for only 0.4 percent of GDP in 2015.
  - Employment of FDI companies about 2.5 percent of total employment as of September 2016; at full capacity employment would reach about 4.7 percent of total employment.
  - Companies within the TIDZ paid social security contributions and personal income tax amounting to MKD 1.1 billion during 2011–2015, or approximately 0.16 percent of total tax revenues and contributions per year.
  - A recent study shows companies in the TIDZ accounted for only one percent of total gross value added in 2013.
- Empirical estimates of FDI impact on growth and employment:
  - Cross-country regression (20 CESEE countries, 1991–2014) finds FDI inflows have a positive impact on growth in most specifications.
  - Estimated coefficient indicates real per capita GDP growth could rise by about 0.5 percentage point as FDI inflows increase by one percent of GDP.
  - Growth decomposition suggests FDI inflows likely contributed to annual real per capita GDP growth for FYR Macedonia of about 1.3 percent on average between 2010 and 2013.
  - Labor market evidence: an increase in FDI inflows per capita by 10 percent would reduce the unemployment rate by 2–3 percent for FYR Macedonia (based on multinomial logit estimates from Atoyan and Jankulov (2015) and micro-level labor force survey data).
- Caveats and heterogeneity:
  - Literature provides mixed evidence on FDI–growth and FDI–employment relationships; results depend on methodology, endogeneity, and type of FDI.
  - This study does not differentiate FDI by type or industry; high-end manufacturing FDI may have different impacts than real estate FDI.
  - Institutional quality matters: government effectiveness indicator ranges approximately from −2.5 (weak) to 2.5 (strong); FYR Macedonia’s government effectiveness was 0.15 in 2014, below the CESEE average of 0.25.

*Source: IMF staff report (content unit _cr16357).*

### 11. Further labor market reforms could enhance the impact of FDI on employment. FYR

### 11. Further labor market reforms could enhance the impact of FDI on employment. FYR

### Labor market reforms and FDI — scenario analysis and findings
- Macedonia performs well in labor market efficiency and flexibility but has room for improvement in female participation in the labor force and reliance on professional management.
- Scenario analysis (based on the labor market outcome model in Annex II) lifted labor market structural indicators to the average level of the new member states while maintaining FYR Macedonia’s advantages; combined labor market reforms and an increase in FDI inflows would significantly help to reduce the unemployment rate.
- Scenario labels shown in the source:
  - Baseline, 2014
  - 10% increase in FDI per capita
  - 10% increase in FDI and labor market reforms
- Unemployment axis values illustrated in the source chart: 24.0, 24.5, 25.0, 25.5, 26.0, 26.5, 27.0, 27.5, 28.0, 28.5 (Percent).

### Indirect economic impacts and labor supply
- Backward linkages with domestic firms appear limited, largely due to the inability of local producers to meet necessary technical and safety requirements (IMF, 2015).
- Anecdotal evidence indicates domestic firms are making progress in improving capacity, but this will take time.
- Outflows of high skilled labor, particularly technical and engineering occupations, could hinder or delay movement up economic and technological ladders.
- IMF (2016) finds emigration lowered the working-age population, affecting labor supply, skill composition, and growth; it is estimated that emigration might have shaved off 0.6–0.9 percentage points of annual growth rates in some countries in CESEE.

### Costs and vulnerabilities of FDI incentives
- Costs to the country from FDI incentives arise through forgone tax revenues from exemptions including profit tax and personal income tax for the salaries of employees for the period of ten years, and subsidies and grants associated with FDI incentives.
- Precise calculation of costs would require detailed data from FDI firms (tax returns, subsidies and grants received, investment, exports and imports, employment), which are missing.
- Debt financing of FDI has been on the rise; some foreign companies view intercompany lending as initial investment likely to be converted into equity later if projects succeed.
- A fast and continued rise in debt-financed FDI would expose the country to various shocks faced by parent companies or home countries and could create external vulnerabilities.

### Empirical evidence on FDI and growth (Annex I)
- Cross-country regression framework for 20 CESEE countries, 1991–2014, finds FDI has contributed positively to economic growth in FYR Macedonia.
- Selected estimated coefficients from Table 1 (Dependent variable: Real per capita GDP growth):
  - Foreign direct investment: 0.391, 0.442, 0.422, 0.490, 0.414 (reported across specifications)
  - Initial income per capita: -0.221, -0.033, -0.034, -0.034, -0.016
  - Schooling: 0.002, 0.001, 0.000, 0.001, 0.001
  - R-squared adjusted: 0.580, 0.211, 0.215, 0.200, 0.210
- Notes: FDI inflows are in percent of GDP (UNCTAD); model includes controls such as inflation, size of government, financial depth, trade openness, and institutional indicators.

### Empirical evidence on labor market outcomes (Annex II)
- Labor market outcomes estimated using a micro-econometric multinomial logit model linking transitions between employment, unemployment, and non-participation to demographic, macroeconomic, and structural indicators; pooled microdata cover labor force surveys from 2006 to 2013 for multiple countries.
- Selected structural and FDI-related coefficients from the pooled results table:
  - Labor market structural indicators (examples reported as significant):
    - Cooperation in labor-employer relations: negative effect on inactivity (coefficients reported in table as negative and significant across specifications)
    - Flexibility of wage determination: negative effect on inactivity
    - Redundancy costs: negative effect on inactivity
    - Pay and productivity: positive effect on employment outcomes
    - Reliance on professional management: negative effect on inactivity (significant in some specifications)
    - Women in labor force: small positive coefficients in employment specifications (reported as 0, 0.001, 0.001 *, 0.003 ***, 0.004 ***, 0.001 **)
  - FDI per capita:
    - For 0=Inactive outcome: FDI per capita.................................-0.401 ***
    - For 1=Employed outcome: FDI per capita.................................0.825 ***
- Model performance notes:
  - Pseudo R2 values reported per country/specification range (examples): 0.310, 0.352, 0.303, 0.368, 0.279, 0.294, 0.316, 0.339, 0.348, 0.350, 0.351, 0.351
  - Number of observations and years vary by country and availability; pooled data cover 2006–13 depending on availability.

### Conclusions and policy implications
- The study finds FDI has contributed positively to economic growth in FYR Macedonia; FDI inflows have direct positive impacts on growth and exports, but impacts on employment and value added remain limited.
- For FDI to play a stronger role in tackling unemployment, financial benefits to investors need to be complemented by labor market and institutional reforms going forward.
- FDI incentives generate fiscal costs and vulnerabilities which need monitoring.

*Source: IMF staff estimates and supporting annexes in the provided IMF chapter.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2016/_cr16357.pdf_