## Annex I— Measuring Infrastructure Gaps

## Source details

**Canonical URL:** [Annex I— Measuring Infrastructure Gaps](https://www.imf.org/-/media/files/publications/dp/2018/45547-western-balkans-public-infrastructure-020818.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/dp/2018/45547-western-balkans-public-infrastructure-020818.pdf.md)
- [Structured JSON version](/-/media/files/publications/dp/2018/45547-western-balkans-public-infrastructure-020818.pdf.json)

---

### Introduction: role of infrastructure in growth and convergence
- Western Balkan incomes now stand at about 30 percent of EU-15 incomes; this share "has not changed much since the onset of the global financial crisis, and has increased only by about 12 percentage points since the early 2000s."
- Infrastructure shortages constrain:
  - connectivity to global and regional markets (transportation networks);
  - production capacity and investment attractiveness (utilities: water and energy);
  - information and knowledge dissemination (communications networks);
  - productivity and competitiveness (underinvestment in human capital and innovation).
- 2016–17 Global Competitiveness Report: region average rank about 85th (out of 138) on infrastructure; 58th on health and primary education; 69th on higher education.
- Investment effects:
  - short term: boosts aggregate demand through fiscal multiplier effects and can crowd in private investment;
  - long term: raises productivity and potential output if public investment efficiency (project selection, implementation, monitoring) is high.
- Risks: weak institutions, inefficient governments, and widespread corruption can cause wasteful spending and high maintenance costs draining fiscal resources.

### Historical background and evolution
- Legacy:
  - Public infrastructure development in former Yugoslavia started later and proceeded more slowly than in Western Europe.
  - 1973–79: stock of gross fixed investment grew at 8.2 percent per year.
  - 1980–90: capital expenditures contracted at an average annual rate of about 5 percent.
  - 1990s conflicts constrained investment and destroyed capital stock.
- Post-1990s initiatives:
  - 1999 Stability Pact; 2008 Regional Co-operation Council.
  - EU Instrument for Pre-Accession Assistance (IPA).
  - Western Balkans Infrastructure Framework (WBIF): donor coordination, blending loans and grants, National Investment Committees, Single Project Pipelines.
  - 2014 Berlin Process; 2017 Trieste summit.
- Recent trend: Public investment accelerated significantly after 2007; accumulation of capital stock strong since then.

### Current stocks, quality, and sectoral issues
- Overall capital stock remains low compared with EU average and other neighboring regions (except CIS).
- Quality well below EU and EU new member states averages.
- Sector weaknesses:
  - Transport: severe underinvestment and inadequate maintenance; low railway density and low rail freight efficiency relative to EU average.
  - Energy: Albania, Kosovo, Montenegro, and FYR Macedonia suffer unstable supply and frequent outages; Bosnia and Herzegovina and Serbia have more secure supply.
- Measured reliability and loss indicators (2016):
  - Average duration of interruptions: about 97 hours in Albania, 62 hours in Kosovo, 27 hours in Montenegro, and 5.6 hours in FYR Macedonia; compared with EU‑NMS averages of about two hours.
  - Average number of interruptions per customer per year: about 43 times in Albania, 35 times in Kosovo, 20 times in Montenegro, and 13 times in FYR Macedonia; compared with EU‑NMS average of one time.
- Note: public capital stock estimates include non-infrastructure assets (residential dwellings, health institutions, government offices); roads and some assets are difficult to value; per capita measures supplement financial estimates.

### Identified constraints to scaling up investment
- Fiscal constraints and limited fiscal space for sustainable public investment.
- Weak public investment management: shortcomings in project selection, implementation, monitoring.
- Poor regional coordination slowing cross‑border projects.
- Financing and execution bottlenecks amid large planned capital budgets.

### Governance and institutional weaknesses (selected findings)
- Fragmented institutional frameworks with overlapping mandates and little coordination.
- Project selection criteria not systematically applied and often waived.
- Project pipelines (mainly WBIF-funded) often outside medium-term budget program; proliferation of unprepared projects included in the budget.
- Limited coordination between central government and municipalities, distorting capital spending allocation.
- Public procurement laws well designed, but implementation and compliance weak.
- Monitoring and disclosure of SOE financial performance, investment plans, and fiscal risks limited or nonexistent.
- Ex post assessments and audits of projects infrequent (mostly for donor-funded projects).

### Case studies and fiscal risks
- Montenegro: Bar–Boljare Highway
  - First phase costs about a quarter of GDP; crowds out other essential capital spending and poses major fiscal sustainability risks.
  - Without fiscal adjustment, public debt would have increased to over 90 percent of GDP by 2019.
  - Two remaining phases estimated additional cost about €1.2 billion; implementation conditional on mostly concessional financing.
  - Cost overruns: first phase cost increased by 25 percent (€1 billion versus €809 million) fully borne by the government.
  - Project awarded without competitive bidding; China ExIm bank financing with concessionality element of over 20 percent.
- Kosovo: Route 7 (Highway Kosovo–Albania)
  - Total cost close to 20 percent of GDP.
  - Usage to date less than one-third of capacity.
  - Project concentrated government capital budget in 2010–13 almost entirely on this highway.
  - Lacked attention to medium-term capacity needs, economic impact, and road safety; a less expensive option would have left resources for other roads.

### Fiscal space, debt, and financing constraints (key numbers)
- Public debt:
  - Public debt averaged 55 percent of GDP on average in 2016, with three countries above 70 percent of GDP.
  - IMF threshold for total public debt as vulnerability indicator: 65 percent of GDP.
  - Albania, Serbia, and Montenegro are already above this debt level.
  - Bosnia and Herzegovina and FYR Macedonia: total public debt in the range of 35–50 percent of GDP and increasing.
  - Kosovo: about 20 percent of GDP in 2016.
- Gross financing needs and liquidity:
  - IMF liquidity and solvency vulnerability indicator threshold: 20 percent of GDP for gross financing needs.
  - In 2016, gross financing needs in Albania and Montenegro were above 20 percent of GDP.
  - FYR Macedonia and Serbia had gross financing needs close to the 15 percent level.
  - Kosovo’s gross financing needs close to 10 percent of GDP largely due to very short-term maturity of outstanding debt.
- Several countries in the "risky zone" on fiscal sustainability:
  - Montenegro: above all three sustainability thresholds and substantial infrastructure gap.
  - Albania: large gross financing needs and high public debt.
  - Serbia: above or close to the limit in these two dimensions.
- Western Balkans Infrastructure Framework leverage:
  - Could leverage up to 1.5 percent of the regional GDP per year of external financing in the next five years.
  - Western Balkans: IFIs Annual Envelope (Lending and Grants) as percent of regional 2017 GDP:
    - World Bank Group, 0.326
    - EU grants, 0.253
    - Other IFIs, 0.856

### Efficiency of spending, revenues, and underexecution
- Underexecution of budgeted capital expenditures ranges from about 5 percent to 25 percent (or about 0.1–1.0 percent of GDP) of the budgeted capital expenditure at central government level.
- National authorities report higher underexecution at general government level when infrastructure funds and public utilities are included.
- Structure of spending:
  - Overall government size large (above 40 percent of GDP) in several countries (Montenegro, Bosnia and Herzegovina, Serbia), but spending mainly recurrent rather than investment.
  - Tax revenue efficiency below more advanced European countries; extensive tax exemptions and incentives common.
  - Property tax revenues low.
  - Energy subsidies (mostly for coal externalities) significant.
- Tax-efficiency weaknesses:
  - Serbia, Bosnia and Herzegovina, and Albania: low efficiency for corporate income tax (CIT).
  - FYR Macedonia and Albania: low efficiency for value added tax (VAT).
  - VAT C-efficiency ratio relatively good in the rest of the Western Balkan countries.

### Key policy recommendations and priorities
- Strengthen public investment management: improve project selection, appraisal, implementation, and monitoring.
- Prioritize investments enhancing regional connectivity and integration into European supply chains to maximize growth spillovers.
- Secure financing primarily from IFIs and donors where possible to preserve fiscal sustainability.
- Leverage private sector participation via well-designed PPPs while recognizing past PPP experiences and implementation challenges.
- Enhance regional cooperation frameworks (WBIF, Berlin Process, Regional Co‑operation Council) for preparation and execution of priority investments and improved donor coordination.
- Practical fiscal measures:
  - Integrate project pipelines into medium-term budget program.
  - Enforce procurement laws and e-procurement implementation.
  - Require monitoring, disclosure, and regular auditing of SOEs; include SOE capital spending in government budgets.
  - Conduct ex post assessments and audits routinely.
  - Create fiscal room via expenditure rationalization and better revenue mobilization (including property and energy taxation).
  - Manage external financing carefully: hedge exchange rate risks, build debt-management capacity, coordinate fiscal and monetary policy.
- Prioritize concessional and IFI financing for large projects with low expected economic returns unless concessional financing can secure viability.

### Regional initiatives (box 2.1) — commitments, status, and recommendations
- Berlin Process: accelerate implementation of priority connectivity projects in energy and transportation.
- July 2017 Trieste summit: EU Commission pledged an additional €190 million for connectivity projects.
- Action plan for establishing a regional economic area adopted at Trieste.
- Donor and recipient countries agreed on 10 priority projects (six transportation, four energy) to be implemented by 2020.
- EU grants and IFI loans amount to €1.4 billion for these priority projects.
- Summits: Berlin (2014), Vienna (2015), Paris (2016), Trieste (2017).
- Implementation status: progress limited; some projects soon to start; many remain in early preparation; capacity constraints limit preparation and execution.
- Recommended responses: delegate more to supranational entities (WBIF, IFIs) for preparation and execution given recipient-country capacity constraints.
- Expected benefits of timely implementation:
  - raise potential growth;
  - cement public consensus for greater regional cooperation;
  - promote political stability to advance EU integration.

### Diaspora bonds (box 5.1)
- Potential: mobilize diaspora wealth where low policy credibility and political instability hinder capital access.
- International precedents:
  - India raised over $11 billion.
  - Israel raised over $35 billion.
- Advantages: stable and cheap external finance; opportunity for diaspora to invest.
- Challenges:
  - Sizable fixed costs to set up a program.
  - Several countries (Ethiopia, Kenya) tried and failed.
  - Issuance in foreign currency or under foreign jurisdiction can mitigate devaluation and default risks.
  - May require institutional capacity building and credit enhancement from multilateral/bilateral agencies.
  - Success depends on diaspora trust that proceeds will be used as intended.
- Western Balkans specifics: could mobilize migrants in EU countries; higher success probability if proceeds directly finance key basic infrastructure or benefit diaspora/families.

### China’s involvement (box 5.2)
- Growing economic presence since November 2015 summit.
- Sectors: railways, motorways, power generation.
- Strategic rationale: geographic proximity to EU and prospects of EU accession; access to EU single market.
- Considerations and risks:
  - Chinese projects often carry concessional financing but may bypass normal project selection/procurement.
  - Examples: projects implemented without full attention to debt sustainability (Montenegro cited).
  - High reliance on Chinese contractors can limit domestic construction-phase economic impact.
- Selected aggregate value: total project value financed by China: €6.2 billion.

### PPPs: opportunities, risks, and experience (box 6.1)
- PPP rationale: mobilize private savings, increase efficiency, provide value for money if well planned and managed.
- Limitations:
  - May delay budgetary expenditures without changing net present value of government spending.
  - Fiscal risks at all stages; can create large contingent liabilities and encourage off-balance operations.
  - Public Investment Management Assessment studies show poor PPP management scores in Western Balkans.
- Five key elements for PPP success:
  1. Sound planning and project selection.
  2. Strong fiscal institutions with sufficient Ministry of Finance control at each stage.
  3. Strong legal frameworks.
  4. Strong budgeting, accounting, and reporting practices.
  5. Appropriate fiscal risk analysis at project level.
- Regional experience limited due to small markets, weak legal/institutional frameworks, perceived political risks, and governments’ limited capacity for credible long-term commitments.
- Example: Tirana International Airport (TIA)
  - 2005 concession: investment of €50 million for 20 years; monopoly on commercial air traffic during concession.
  - New terminal opened in 2007; passenger numbers tripled since concession signing.
  - 2016 renegotiation: rescinded monopoly to allow a second airport; ownership change with interests sold to Chinese investors.

### Model-based simulations: scenarios and main quantitative findings
- Simulation setup:
  - Public investment surge of 15 percent of GDP over eight years (surge corresponds to median size of top projects in Western Balkans project pipelines).
  - Model calibrated to average Western Balkan country; real risk free rate 3.5 percent.
  - Return on public investment: 20 percent (baseline) and 25 percent (improved policies).
  - Public investment efficiency: calibrated to average SEE-XEU (baseline) and CEE (high efficiency).
  - Tax rate responds by 10 percent of the fiscal gap in the previous period.
- Four scenarios and outcomes:
  - Baseline (financed only by domestic bank borrowing):
    - Growth dividends: 0.1 percentage point increase in annual growth rate compared with current projections.
    - Only half of public investment assumed to build capital stock (low efficiency).
    - Public debt peaks at about 60 percent of GDP (starting from Western Balkan average public debt of 51 percent of GDP).
  - Improved policies (higher efficiency and regional coordination):
    - Public investment builds three-quarters of expenditure into capital stock.
    - Growth dividends: 0.3 percentage point increase in annual growth rate compared with current projections.
    - Public debt ratios decline only slowly; vulnerabilities remain high medium term.
  - External financing (improved policies + access to external financing at lower costs and longer maturities):
    - Avoids crowding out of private investment; improves economic outlook by about 0.6 percentage point increase in the medium-term annual growth rate compared with current projections.
    - Short-term competition for labor persists due to closed output gap.
    - Lower external financing costs reduce required long-term tax increases; debt outlook improves but vulnerabilities remain.
  - Grants and IFI financing (sequenced with efficiency and regional coordination):
    - Treated as the most favorable scenario; eliminates crowding out of private investment and reduces the need to raise taxes materially; yields higher growth—0.8 percent at the peak—and prevents any significant increase in debt.
- Model output values preserved from figures and notes:
  - Simulated surge in public investment (Public Investment Percent of GDP) includes values: 5.0, 8.0, 5.5, 6.0, 6.5, 7.0, 7.5 for years 2017–29 and 2023–26 as displayed.
  - Initial Real GDP growth values visible in figure: 2.9, 3.8, 3.9, 3.0, 3.7, 3.5, 3.3, 3.6, 3.4, 3.2, 3.1 for years 2017–45.
  - Public debt levels (Percent of GDP) under scenarios include values: 50, 64, 52, 62, 58, 60, 56, 54 across timeline.
  - Tax rate on consumption (Percent of GDP) shows values including 18.0, 20.5, 21.0, 20.0, 19.0, 19.5, 18.5 across years.
- Key insights:
  - Financing modality critically affects outcomes: domestic bank financing leads to crowding out and large public debt increases; external concessional financing with improved policies yields substantially better growth outcomes.
  - Efficiency of public investment and regional connectivity improvements materially raise growth dividends.
  - Even with improved policy and financing, tax increases and debt vulnerabilities can persist in the medium term.

### Quantitative impacts (regression and GE model results)
- General equilibrium and regression evidence:
  - Under the most favorable scenario, full implementation of regional connectivity projects implies a long-term improvement in the level of real GDP per capita in the range of 3.5 percentage points above steady state.
  - Closing the infrastructure gap by 20 percentage points (comparable to the investment surge) would generate higher annual real GDP growth rates by about 0.2–0.3 percentage point over the medium term.
  - The reduction in the infrastructure gap by 20 percentage points would translate into reductions in per capita income gaps of up to 6 percentage points over the long term.
  - Implied impact on real GDP level:
    - Near term: ranges between 0.1 and 0.4 percent, depending on policy and financing assumptions.
    - Medium term: ranges between 0.3 and 0.6 percent, depending on policy and financing assumptions.
- Comparative fiscal-multiplier evidence (reported):
  - IMF (2014) advanced economies: a 1 percent of GDP increase in investment spending increases output by about 0.4 percent in the same year and by 1.5 percent four years after the shock; with high public investment efficiency: 0.8 and 2.6 percent, respectively; during low growth: 1.5 and 3.0 percent, respectively.
  - Emerging market economies: about 0.3 percent in the same year and 0.5 percent four years after the shock.

### Measuring infrastructure gaps: indicators, aggregation, and data
- Six indicators:
  1. Telephone/cell phone lines per capita
  2. Broadband subscriptions per capita
  3. Installed capacity to generate electricity per capita
  4. Air passengers carried per capita
  5. Highways per km² after controlling for population density
  6. Railroad per km² after controlling for population density
- Indicator gap formula (structure preserved): Infrastructure gapi,j,t = (Indicator i,t / average(Indicator j)EU,t − 1) * 100
  - Example: Infrastructure gapElect., ALB,t = (Installed capacity to gen. electricityALB,t / Installed capacity to gen. electricityEU,t − 1) * 100
- Highways and railroads: gaps computed relative to EU-average adjusted for country population density via regression projection.
- Aggregate infrastructure gap: weighted average of indicator gaps with weights inversely related to volatility of each indicator gap across time:
  - Aggregate infrastructure gapi,t = Σj wj * Infrastructure gapi,j,t
  - wj = 1 / (Σi Stdi(Infrastructure gapi,j) / # of countries)
  - Intuition: indicators with high volatility receive low weight; low-volatility indicators receive high weight.
- Robustness: equal-weights aggregation produces similar results.
- Data sources:
  - Telephone/cell phone lines per capita: World Bank, WDI; completed with national statistics offices.
  - Broadband subscriptions per capita: World Bank, WDI; completed with national statistics offices.
  - Installed capacity to generate electricity per capita: International Energy Agency.
  - Air passengers per capita: World Bank, WDI; completed by national statistics offices.
  - Highways per km² after controlling for population density: International Road Federation; country authorities’ data.
  - Railroad per km² after controlling for population density: World Bank, WDI; completed by national statistics offices.

### Econometric approach, sample, and selected regression results
- Two-stage least squares to address endogeneity:
  - First stage instruments for infrastructure gap: log population; percent urban population; distance to Brussels; years since industrialization; indices of political stability and fighting corruption (World Governance Indicators).
  - Second stage: real GDP growth per capita regressed on instrumented infrastructure gaps and controls.
- Panel sample: 39 European countries, period 1997–2015.
  - Observations: 491
  - Countries: 39
  - Adj. R-squared (baseline IV): 0.12
- Key regression coefficients (Table AII.1, preserved values):
  - Infrastructure gap 6 sectors (weights = inverse standard deviations): 0.106 [0.025]***
    - Interpretation: Closing a (negative) infrastructure gap by 1 percentage point is estimated to be associated with 0.1 percent higher growth.
  - Infrastructure gap 6 sectors (average 6 sectors): 0.086 [0.023]***
  - Sector-specific:
    - Highways per 1000 square Km after controlling for density: 0.063 [0.029]**
    - Railways per square Km after controlling for density: 0.042 [0.015]***
    - Electricity generation installed capacity per capita relative to EU: 0.09 [0.022]***
    - Telephones (fixed + mobile) per 1000 inhabitants relative to EU: 0.003 [0.030]
    - Air passengers per capita relative to EU: 0.035 [0.022]
    - Broadband subscribers per 100 inhab. relative to EU: -0.016 [0.015]
  - Other controls (Table AII.1 column I):
    - Openness (exports + imports / GDP): 0.017 [0.007]**
    - Government consumption to GDP: -0.272 [0.073]***
    - Average inflation past 5 years: -6.99 [3.344]**
    - Log population: 0.757 [0.261]***
    - Log GDP per capita in ppp (t − 5): -6.569 [1.033]***
- Income-gap regressions (Table AII.2, preserved values):
  - Infrastructure gap 6 sectors (weights = inverse standard deviations): 0.280 [0.064]***
  - Infrastructure gap 6 sectors (average 6 sectors): 0.194 [0.050]***
  - Openness (exports + imports) / GDP: 0.023 [0.012]*
  - Government consumption to GDP: -0.434 [0.154]***
  - Average inflation past 5 years: -9.142 [5.445]*
  - Income gap with EU (t − 5): 0.763 [0.031]***
  - Observations: 491; Countries: 39; Adj. R-squared: 0.98
- Robustness notes:
  - Aggregate infrastructure gap shows consistent positive and significant effect on growth.
  - Physical infrastructure (highways, railways, electricity generation capacity) has the highest significant impact.
  - Telecommunications and broadband show weak or negative coefficients; quality not captured by indicators may explain this.

### General equilibrium model calibration (selected parameters)
- Per capita potential GDP growth: 3 percent
- Public debt-to-GDP ratio: 51 percent
- Average tax rate: 18 percent
- Public investment-to-GDP ratio: 5.2 percent
- Real average domestic interest rate: 7 percent
- Real average external interest rate: 5 percent
- Productivity of capital: 20 percent
- Model features: small open economy, two sectors, multiple public debt types, private and public capital, imported inputs, two consumer types (savers and hand-to-mouth), government financing via taxes, fees, domestic borrowing, external borrowing, concessional financing; accounts for inefficiencies/waste in public investment.

*Annex I— Measuring Infrastructure Gaps, from "Public Infrastructure in the Western Balkans" (content unit provided).*

### Annex I� Measuring Infrastructure Gaps �����������������������������������������������������������������������51

### Annex I— Measuring Infrastructure Gaps

### Introduction: role of infrastructure in growth and convergence
- Western Balkan incomes now stand at about 30 percent of EU-15 incomes; this share "has not changed much since the onset of the global financial crisis, and has increased only by about 12 percentage points since the early 2000s."
- Shortages of core public infrastructure can constrain:
  - connectivity to global and regional markets via inadequate transportation networks;
  - production capacity and investment attractiveness via insufficient or unreliable utilities (for example, water and energy);
  - information and knowledge dissemination via underdeveloped communications networks;
  - productivity and competitiveness via underinvestment in human capital and innovation.
- The 2016–17 Global Competitiveness Report ranks countries in the region at the average rank of about 85th place (out of 138 countries) on infrastructure, compared with 58th and 69th positions on health and primary education and higher education.
- Infrastructure investment has both short- and long-term effects:
  - short term: boosts aggregate demand through fiscal multiplier effects and can crowd in private investment;
  - long term: raises productivity and potential output if public investment efficiency (project selection, implementation, monitoring) is high.
- Risks: weak institutions, inefficient governments, and widespread corruption can lead to wasteful spending and high maintenance costs that drain fiscal resources.

### Historical background and evolution of public infrastructure
- Legacy factors:
  - Public infrastructure development in the former Yugoslavia started later and proceeded more slowly than in Western Europe.
  - 1973–79: stock of gross fixed investment grew at 8.2 percent per year.
  - Early 1980s fiscal adjustment led to capital expenditure cuts; capital expenditures contracted at an average annual rate of about 5 percent in 1980–90.
  - 1990s conflicts constrained investment and destroyed part of the capital stock.
- Post-1990s initiatives and recent trends:
  - 1999 Stability Pact for the Balkans aimed to spur public investment; later replaced by the Regional Co-operation Council in 2008 to increase regional ownership.
  - EU Instrument for Pre-Accession Assistance (IPA) consolidated EU assistance.
  - Western Balkans Infrastructure Framework (WBIF) supports donor coordination, leveraging loans and grants via blending, and prioritizing projects with regional impact through country-led National Investment Committees and Single Project Pipelines.
  - 2014 Berlin Process focused on regional infrastructure projects and sustaining EU accession dialogue.
  - Public investment accelerated significantly after 2007, supported by international initiatives; the accumulation of capital stock has been strong since then.

### Current stocks and quality: measured gaps and sectoral issues
- Despite the recent surge in capital spending, overall capital stock remains low compared with the EU average and other neighboring regions (except CIS).
- Quality remains well below the EU and EU new member states averages; notable sectoral weaknesses include:
  - Transport: railroads and roads show legacy of severe underinvestment and inadequate maintenance; observable low railway density and rail freight efficiency relative to EU average.
  - Energy: Albania, Kosovo, Montenegro, and FYR Macedonia suffer from unstable energy supply and frequent outages, coupled with large distributional losses; Bosnia and Herzegovina and Serbia have more secure electricity supply.
- Measured reliability and loss indicators (2016):
  - Average duration of interruptions: about 97 hours in Albania, 62 hours in Kosovo, 27 hours in Montenegro, and 5.6 hours in FYR Macedonia; compared with EU‑NMS averages of about two hours.
  - Average number of interruptions per customer per year: about 43 times in Albania, 35 times in Kosovo, 20 times in Montenegro, and 13 times in FYR Macedonia; compared with EU‑NMS average of one time.
- Note on measurement: public capital stock estimates include assets beyond infrastructure (for example, residential dwellings, health institutions, government offices) and some assets (especially roads) are difficult to value; quantitative per capita measures supplement financial estimates.

### Identified constraints to scaling up infrastructure investment
- Fiscal constraints and limited fiscal space to ramp up public investment sustainably.
- Weak public investment management: shortcomings in project selection, implementation, and monitoring reduce efficiency and increase the risk of wasteful spending.
- Poor regional coordination slows realization of projects with cross‑border benefits.
- Potential financing and execution bottlenecks amid large planned capital budgets.

### Policy recommendations and priorities
- Strengthen public investment management to raise efficiency of public spending by improving project selection, implementation, and monitoring.
- Prioritize investments that enhance regional connectivity and facilitate integration into European supply chains to maximize growth spillovers.
- Secure financing primarily from international financial institutions (IFIs) and donors where possible to preserve fiscal sustainability.
- Leverage private sector participation, including through efficient use of public‑private partnerships, while recognizing past PPP experiences and implementation challenges.
- Enhance regional cooperation frameworks (WBIF, Berlin Process, Regional Co‑operation Council) to accelerate preparation and execution of priority investments and improve donor coordination.

### Expected payoffs from narrowing infrastructure gaps (qualitative summary)
- If institutional weaknesses are addressed and financing is appropriately sourced:
  - significant potential growth benefits are likely from closing infrastructure gaps;
  - short‑run demand effects and longer‑run supply effects (higher productivity and potential output) can be realized;
  - growth payoffs are larger when public spending efficiency is improved, investments are regionally integrated, and financing is channeled through IFIs/donors.

*Annex I— Measuring Infrastructure Gaps, from "Public Infrastructure in the Western Balkans" (content unit provided).*

### box 2.1. Western balkans—Regional Initiatives

### box 2.1. Western balkans—Regional Initiatives

### Regional political and financial commitments
- The Berlin Process calls for accelerating implementation of priority connectivity projects in the energy and transportation sectors.
- In the July 2017 Trieste summit, the EU Commission pledged an additional €190 million for connectivity projects.
- An action plan for establishing a regional economic area was adopted at the Trieste summit.
- Donor and recipient countries agreed on 10 priority projects (six transportation and four energy projects) to be implemented by 2020.
- EU grants and loans from international financial institutions amount to €1.4 billion for these priority projects.
- Summits were held in Berlin (2014), Vienna (2015), Paris (2016), and Trieste (2017).

### Implementation status and constraints
- Progress achieved so far on the 10 priority projects has been limited.
- Work on the ground is expected to start soon for a few projects; other projects remain in early preparation stages.
- Capacity constraints in recipient countries limit preparation and execution of regional projects.

### Recommended institutional responses
- Greater efforts are needed at both regional and national levels to advance implementation of priority projects.
- Given capacity constraints, recipient countries could delegate more to supranational entities, including the WBIF and international financial institutions, for preparation and execution of regional projects.

### Expected benefits of timely implementation
- Timely progress in implementing these regional projects:
  - will raise the potential growth of the Western Balkan economies,
  - will cement public consensus for greater regional cooperation,
  - and will promote political stability to bring the EU integration process forward.

*Source: box 2.1. Western balkans—Regional Initiatives, Public Infrastructure in the Western Balkans (excerpt).*

### 1. Institutional frameworks are fragmented with overlapping mandates and little coor-

### 45547-western-balkans-public-infrastructure-020818

### Major governance and institutional weaknesses
- Institutional frameworks are fragmented with overlapping mandates and little coordination of various public bodies.
- Project selection criteria are not systematically applied and are often waived.
- Project pipelines, primarily used for Western Balkans Infrastructure Framework–funded projects, are often outside the medium-term budget program, allowing a proliferation of other projects, which instead are included in the budget but are not ready for implementation.
- There is limited coordination between central government and municipalities, leading to a distorted allocation of capital spending.
- Public procurement laws, including on e-procurement, are well designed for competitive and transparent procedures, but implementation and compliance are weak and infrequent.
- Monitoring and disclosure of financial performance, investment plans, and fiscal risks of state-owned enterprises are limited or inexistent.
- There are substantial gaps in government budgets, largely due to state-owned enterprises’ capital spending, which is not included.
- Ex post assessments and audits of projects are not generally undertaken by the government—only infrequently in the cases of donor-funded projects.

### Case studies and project evidence
- Montenegro: Bar–Boljare Highway
  - The first phase of the project—which is the only one budgeted, contracted, and currently under implementation—will cost about a quarter of GDP, crowding out other essential capital spending and posing major fiscal sustainability risks.
  - In the absence of any fiscal adjustment, public debt would have increased to over 90 percent of GDP by 2019.
  - The two remaining phases have an estimated additional cost of about €1.2 billion; their implementation could be considered only if the authorities are able to secure mostly concessional financing for the project.
  - The estimated low economic return on the investment, due to a higher-than-projected cost per kilometer (because of geological challenges) and the lower expected traffic, calls for concessional financing to ensure the financial viability of the project.
  - Cost overruns due to currency risk: the cost of the first phase increased by 25 percent (€1 billion versus €809 million) fully borne by the government.
  - The project was given to a Chinese contractor without competitive bidding, but the China ExIm bank is providing financing with a concessionality element of over 20 percent.
- Kosovo: Route 7 (Highway Kosovo–Albania)
  - The total cost of the motorway was close to 20 percent of GDP.
  - Usage to date has been less than one-third of capacity.
  - The project concentrated the government’s capital budget in 2010–13 almost entirely on this highway.
  - The project was politically important but lacked sufficient attention to medium-term capacity needs, economic impact, and road safety; a less expensive option would have left resources for other roads.
  - Greater competition in bids, transparent procurement, and more robust monitoring and auditing could have helped prevent substantial cost increases.

### Fiscal space, debt, and financing constraints
- Closing infrastructure gaps requires substantial fiscal resources; most countries in the region face limited fiscal space.
- Public debt:
  - Public debt averaged 55 percent of GDP on average in 2016, with three countries above 70 percent of GDP.
  - IMF “norm” or threshold for total public debt as an economic vulnerability indicator: 65 percent of GDP.
  - Albania, Serbia, and Montenegro are already above this debt level.
  - Bosnia and Herzegovina and FYR Macedonia: total public debt in the range of 35–50 percent of GDP and increasing.
  - Kosovo: about 20 percent of GDP in 2016.
- Gross financing needs and liquidity:
  - IMF liquidity and solvency vulnerability indicator threshold: 20 percent of GDP for gross financing needs.
  - In 2016, gross financing needs in Albania and Montenegro were above 20 percent of GDP.
  - FYR Macedonia and Serbia had gross financing needs close to the 15 percent level.
  - Kosovo’s gross financing needs were close to 10 percent of GDP largely due to very short-term maturity of outstanding debt.
- Several Western Balkan countries are in the “risky zone” on fiscal sustainability indicators:
  - Montenegro is above all three sustainability thresholds and has a substantial infrastructure gap.
  - Albania has large gross financing needs and a high level of public debt.
  - Serbia is above or close to the limit in these two dimensions.
- Choice of financing tools:
  - Domestic funding options are unlikely to be sufficient given the scale of infrastructure gaps and underdeveloped banking systems.
  - External commercial borrowing can provide funding but introduces refinancing, interest, and exchange rate risks and can build up debt.
  - Eurobonds have been issued by Serbia, FYR Macedonia, Montenegro, and, to a lesser extent, Albania; interest rates have been moderate but could become more costly if global conditions deteriorate.
  - Diaspora bonds could leverage remittances but require specific expertise and cannot be a primary source for large projects.
  - IFI financing is most suitable for capital projects due to favorable interest costs, longer maturities, and grace periods; IFIs can help with project selection, preparation, and catalyzing private capital.
  - Bilateral financing (including Chinese financing) can provide concessionality but may come with lower project selection and procurement standards and greater reliance on donor-country contractors.
- Western Balkans Infrastructure Framework and IFI envelope:
  - The Western Balkans Infrastructure Framework could leverage up to 1.5 percent of the regional GDP per year of external financing in the next five years.
  - Western Balkans: IFIs Annual Envelope (Lending and Grants) as percent of regional 2017 GDP:
    - World Bank Group, 0.326
    - EU grants, 0.253
    - Other IFIs, 0.856

### Efficiency of government spending, revenues, and project implementation
- Underexecution of budgeted capital expenditures:
  - There is significant underexecution of budgeted capital expenditures in nearly all countries in the region, ranging from about 5 percent to 25 percent (or about 0.1–1.0 percent of GDP) of the budgeted capital expenditure at the central government level.
  - National authorities report even higher underexecution rates at the general government level when infrastructure funds and public utilities are included.
  - Strengthening project implementation capacity would improve utilization of fiscal space and donor financing absorption.
- Structure of spending and revenue mobilization:
  - The overall size of government is large (above 40 percent of GDP) in several countries (notably Montenegro, Bosnia and Herzegovina, and Serbia), but spending is mainly recurrent rather than investment.
  - Tax revenue efficiency is typically below more advanced European countries; many countries provide extensive tax exemptions and incentives.
  - Property tax revenues are low.
  - Energy subsidies (mostly for coal externalities) are significant, indicating room for revenue increases through energy taxation.
  - Specific tax-efficiency weaknesses noted:
    - Serbia, Bosnia and Herzegovina, and Albania: low efficiency for corporate income tax (CIT).
    - FYR Macedonia and Albania: low efficiency for value added tax (VAT).
    - VAT C-efficiency ratio is relatively good in the rest of the Western Balkan countries.

### Key policy implications and recommended priorities
- Prioritize concessional and IFI financing for large projects with low expected economic returns unless concessional financing can secure viability.
- Strengthen project selection, appraisal, and prioritization processes; ensure project selection criteria are systematically applied and not waived.
- Integrate project pipelines into the medium-term budget program to prevent proliferation of unprepared projects and ensure readiness for implementation.
- Improve coordination between central government and municipalities to avoid distorted capital spending allocation.
- Enforce procurement laws and e-procurement implementation to enhance competition, transparency, and compliance.
- Require monitoring, disclosure, and regular auditing of state-owned enterprises’ financial performance, investment plans, and fiscal risks; include SOE capital spending in government budgets.
- Conduct ex post assessments and audits of projects routinely, not only for donor-funded projects.
- Strengthen fiscal frameworks to address underexecution of capital budgets and to create room for productive capital spending through expenditure rationalization and better revenue mobilization (including property taxation and energy taxation).
- Manage external financing carefully: hedge exchange rate risks, build debt-management capacity, coordinate fiscal and monetary policy, and balance concessional versus commercial borrowing.
- Leverage the Western Balkans Infrastructure Framework and coordinate IFIs, bilateral donors, and governments to ensure rigorous project selection, appraisal, and procurement.

*Italic: Public Infrastructure in the Western Balkans (selected chapter content).*

### 1.5 percent of regional

### 1.5 percent of regional GDP per year

### Diaspora bonds: mobilizing diaspora wealth
- Diaspora bonds can tap into diaspora wealth to finance infrastructure where low policy credibility and political instability hinder capital access.
- International precedents:
  - India raised over $11 billion.
  - Israel raised over $35 billion.
- Advantages:
  - Represent a stable and cheap source of external finance.
  - Offer diaspora investors an opportunity to help their country of origin while providing an investment opportunity.
- Challenges and design considerations:
  - Sizable fixed costs to establish a diaspora bond program (assessing risk profile, liquidity preferences, expected return of diaspora).
  - Several countries, including Ethiopia and Kenya, have tried but failed to issue diaspora bonds.
  - Issuance in foreign currency or under foreign jurisdiction could mitigate devaluation and default risks for risk-averse migrants.
  - Issuance from countries with weak governance and high sovereign risk may require institutional capacity building and credit enhancement from multilateral or bilateral agencies.
  - Success hinges on diaspora trust that proceeds will be used for intended purposes.
- Western Balkans specifics:
  - Diaspora bonds could mobilize wealth of migrants in EU countries by leveraging emotional ties.
  - Success probability rises if proceeds directly finance key basic infrastructure projects or benefit diaspora/families.
  - Diaspora communities also facilitate significant FDI inflows by providing information on investment opportunities and compliance with domestic regulation and legislation.

*box 5.1. Diaspora bonds*

### China’s involvement in the Western Balkans
- China’s economic presence in the Western Balkans is growing following the November 2015 summit with Southeastern Europe counterparts.
- Sectors with investment opportunities: railways, motorways, power generation.
- Strategic rationale: geographical proximity to the EU and prospects of EU accession provide Chinese operators access to the EU single market.
- Strengthening trade corridors used by Chinese companies will improve regional connectivity and facilitate transport of Chinese goods into the EU single market.
- Considerations and risks:
  - Chinese projects often come with concessional financing but may be treated outside normal project selection or procurement procedures.
  - Examples of concerns: projects implemented without full attention to debt sustainability considerations (Montenegro cited).
  - High reliance on Chinese contractors can limit domestic economic impact during construction.
- Selected projects and aggregate value:
  - Total project value financed by China: €6.2 billion.
  - Country project examples include Danube bridge, Belgrade-Budapest high-speed railway, 350 MW unit at Kostolac thermal power plant, sections of European motorway XI, 45 MW unit at Tuzla thermal power plant, 350 MW Banovici thermal power plant, 300 MW Stanari thermal power plant, renewal ship fleet, railways modernization-European corridor X, Industrial park I Durres.

*box 5.2. China’s Involvement in the Western balkans*

### Public-Private Partnerships (PPPs): opportunities and risks
- Rationale:
  - Insufficient fiscal space and public sector inefficiencies make private financing of infrastructure attractive.
  - PPPs can mobilize private savings, increase efficiency, and provide value for money if well planned and managed.
- Limitations:
  - PPPs may only delay budgetary expenditures and not change total net present value of government spending.
  - PPPs involve fiscal risks at all stages: budget preparation, procurement, financing, and managing performance-based contracts.
  - PPPs can generate large explicit and implicit contingent liabilities (for example, guarantees) and encourage off-balance operations that reduce transparency.
  - Public Investment Management Assessment studies for the Western Balkans show poor scores for PPP management.
- Five key elements for government success in PPPs:
  1. Sound planning and project selection.
  2. Strong fiscal institutions with sufficient control of the ministry of finance at each stage of the PPP process, including possible contract renegotiation.
  3. Strong legal frameworks.
  4. Strong budgeting, accounting, and reporting practices.
  5. Appropriate fiscal risk analysis at the project level.
- Regional experience and prospects:
  - Limited PPP experience due to small national markets, inadequate legal and institutional frameworks, perceived regional political risks.
  - Governments’ limited capacity for credible long-term commitments contributes to a “high PPP mortality rate.”
  - EU integration and regional investment planning could reduce political risks and attract private investors.
- Example: Tirana International Airport (TIA) concession
  - 2005 concession: investment of €50 million for 20 years; monopoly on all commercial air traffic in Albania during concession.
  - New terminal opened in 2007; passenger numbers tripled since concession signing.
  - Outcomes:
    - PPP achieved construction and operation of a modern airport with high service standards and minimal fiscal risks.
    - Business model emphasized relatively high landing fees and European legacy carriers; low-cost carriers initially absent.
    - Government renegotiated the concession in 2016 to rescind the monopoly and allow a second airport to operate.
    - Ownership change: Tirana Airport Partners agreed to sell interests to Chinese investors, who adopted a more expansionist business model; low-cost carriers became frequent at TIA.

*box 6.1. The PPP Experience for the Development of the Tirana International Airport*

### Infrastructure investment, growth, and debt sustainability — model-based simulations
- Simulation setup:
  - General equilibrium model simulates an increase in public investment in infrastructure of 15 percent of GDP over eight years (surge corresponds to median size of top projects in Western Balkans project pipelines).
  - Key assumptions:
    - Permanent increase in public investment requires maintaining higher capital stocks.
    - Model calibrated to average Western Balkan country; external borrowing costs reflect region’s current spreads; domestic financing costs higher than external financing.
    - Economy operates at long-term equilibrium with closed output gap.
    - Endogenous tax response aimed at ensuring debt sustainability over the long term.
    - Real risk free rate is 3.5 percent.
    - Return on public investment: 20 percent (baseline) and 25 percent (improved policies).
    - Public investment efficiency calibrated to average SEE-XEU (baseline) and CEE (high efficiency) levels.
    - Tax rate responds by 10 percent of the fiscal gap in the previous period.
- Four scenarios and main outcomes:
  - Baseline (financed only by domestic bank borrowing):
    - Growth dividends muted: 0.1 percentage point increase in annual growth rate compared with current projections.
    - Only half of public investment expenditure assumed to build capital stock (low efficiency).
    - Public debt burden increases rapidly, peaking at about 60 percent of GDP (starting from Western Balkan average public debt of 51 percent of GDP), leaving debt vulnerabilities high.
  - Improved policies (higher efficiency and regional coordination):
    - Public investment channels three-quarters of expenditure into capital stock.
    - Growth dividends: 0.3 percentage point increase in annual growth rate compared with current projections; about half of improvement due to efficiency gains and half to productivity gains.
    - Public debt ratios decline only slowly; debt vulnerabilities remain high throughout the medium term.
  - External financing (improved policies + access to external financing at lower costs and longer maturities):
    - Avoids crowding out of private investment; improves economic outlook by about 0.6 percentage point increase in the medium-term annual growth rate compared with current projections.
    - Short-term: public investments still compete with private sector for labor inputs due to closed output gap.
    - Lower external financing costs require lower long-term tax increases; debt outlook improves further but vulnerabilities remain.
  - Grants and IFI financing (implied in model sequencing):
    - Not explicitly quantified in the supplied excerpt but treated as the final scenario in the sequence with efficiency and regional coordination.
- Model outputs illustrated (selected numeric references preserved from figures and notes):
  - Simulated surge in public investment: shown in the timeline (Public Investment Percent of GDP) with values including 5.0, 8.0, 5.5, 6.0, 6.5, 7.0, 7.5 for years 2017–29 and 2023–26 as displayed in Figure 7.1.
  - Real GDP growth per capita trajectories and percent deviations from steady state are reported across scenarios (examples of values visible in Figure 7.2: Initial growth values include 2.9, 3.8, 3.9, 3.0, 3.7, 3.5, 3.3, 3.6, 3.4, 3.2, 3.1 for years 2017–45).
  - Public debt levels (Percent of GDP) under scenarios show values such as 50, 64, 52, 62, 58, 60, 56, 54 across the timeline.
  - Tax rate on consumption (Percent of GDP) shows values including 18.0, 20.5, 21.0, 20.0, 19.0, 19.5, 18.5 across years.
- Key insights:
  - Financing modality matters critically: domestic bank financing leads to crowding out and large increases in public debt; external financing with improved policies yields substantially better growth outcomes.
  - Efficiency of public investment and regional connectivity improvements materially raise growth dividends.
  - Even under improved policy and financing scenarios, tax increases and debt vulnerabilities can persist in the medium term.

*Economic Dividends of Infrastructure Development: Quantitative Evidence*

*Italic: IMF document (content unit: 45547-western-balkans-public-infrastructure-020818 - 1.5 percent of regional)*

### 1. Real GDP Growth per capita

### 45547-western-balkans-public-infrastructure-020818 - 1. Real GDP Growth per capita

### Model-based simulations: assumptions and scenarios
- Assumes a cumulative increase in public investment of 15 percent of GDP spread over eight years.
- Macroeconomic calibration: average Western Balkans’ levels.
- Public investment efficiency: calibrated to average SEE-XEU (baseline) and CEE (high efficiency) levels.
- Return on public investment: 20 percent (baseline) and 25 percent (improved policies).
- Tax rate responds by 10 percent of the fiscal gap in the previous period.
- Real risk free rate: 3.5 percent.
- IFI = International financial institutions.
- IFI and grant financing scenario: surge financed by an equal mix of grants and IFI financing; described as the most favorable scenario because it eliminates crowding out of private investment and reduces the need to raise taxes materially; yields higher growth—0.8 percent at the peak—and prevents any significant increase in debt.

### Quantitative impacts reported (model- and regression-based)
- Under the most favorable scenario, full implementation of regional connectivity projects would imply a long-term improvement in the level of real GDP per capita in the range of 3.5 percentage points above steady state.
- Closing the infrastructure gap by 20 percentage points (comparable to the investment surge) would generate higher annual real GDP growth rates by about 0.2–0.3 percentage point over the medium term.
- The reduction in the infrastructure gap by 20 percentage points would translate into reductions in per capita income gaps of up to 6 percentage points over the long term.
- Implied impact on the real GDP level (summary statements):
  - Near term: ranges between 0.1 and 0.4 percent, depending on policy and financing assumptions.
  - Medium term: ranges between 0.3 and 0.6 percent, depending on policy and financing assumptions.
- Comparative fiscal-multiplier evidence cited:
  - IMF (2014) advanced economies: a 1 percent of GDP increase in investment spending increases output by about 0.4 percent in the same year and by 1.5 percent four years after the shock; with high public investment efficiency: 0.8 and 2.6 percent, respectively; during low growth: 1.5 and 3.0 percent, respectively.
  - Emerging market economies: about 0.3 percent in the same year and 0.5 percent four years after the shock.

### Table 7.1 (reported figures from source)
- Reported figures correspond to percentage points increases in real GDP levels compared with the initial GDP level that arises from a 1 percent of GDP public investment shock.
- For comparability, simulations for the Western Balkans assume autocorrelation of 0.5 and removal of a natural trend growth of 3 percent in the general equilibrium model.
- Raw reported table entries (preserved exactly as in source):
  - Row labeled "T0.1": 30.010.170.360.360.40.81.50.25
  - Row labeled "T + 40.250.060.410.620.641.52.63.00.50

### Econometric approach and identification
- Panel sample: advanced and emerging European countries, period 1997–2014.
- Two-step approach to address endogeneity:
  1. Instrument the infrastructure gap using geographic, historic, and demographic variables believed correlated with the infrastructure gap but not with the error term.
  2. Use predicted infrastructure gaps from the first stage as instruments in the second stage.
- Regression-based simulation finding: an infrastructure shock of 15 percent of GDP in eight years, calibrated to country-specific effects matching priority projects in the Single Project Pipeline, produces country-specific impacts on growth and income gaps.

### Key policy issues and recommendations
- Need to mobilize domestic resources to create fiscal room for critical infrastructure spending and cofinancing of projects.
- Strengthened project implementation to improve utilization of fiscal space and absorption of donor financing.
- Create additional fiscal space by:
  - Containing current spending and increasing capital spending.
  - Stronger revenue mobilization through broadening the tax base via elimination of exemptions and tax incentives.
  - Increasing revenue intake from property taxation where room exists.
  - Strengthening tax administrations to improve compliance.
- Bolster public investment management frameworks to improve planning, allocation, and implementation capacities; recommended measures include:
  1. Greater coordination of involved public bodies by clarifying roles and responsibilities.
  2. Preparation and publication of a national development strategy covering all capital and current spending.
  3. Enhanced appraisal processes for key public investment projects.
  4. Comprehensive government oversight over PPPs and state-owned enterprises.
  5. Transparency of budget documentation by covering PPP operations and all off-budget contingent liabilities.
  6. Independent ex post assessments and audits conducted on a regular basis.
- Prioritization of infrastructure projects should be insulated from politicization via a defined infrastructure pipeline anchored in robust economic-efficiency analysis and transparent cost-benefit justification for canceling previously approved priority projects.
- Financing strategy: scale up public infrastructure largely through external, concessional official donor and multilateral financing; leverage stronger regional coordination to maximize growth returns, improve investment attractiveness, and secure financing from the EU, international financial institutions, and bilateral donors.
- Consider trade-offs: infrastructure investment increases imports of investment goods and can deteriorate current account balances; sharp current account widening could pose macroeconomic challenges if payoffs to higher potential GDP take time to materialize.
- Complement infrastructure investments with strong policies and renewed reform momentum; infrastructure is necessary but not sufficient to close large income gaps with the EU.

### Measuring infrastructure gaps (Annex I summary)
- Six indicators used to approximate infrastructure gaps:
  1. Telephone/cell phone lines per capita
  2. Broadband subscriptions per capita
  3. Installed capacity to generate electricity per capita
  4. Air passengers carried per capita
  5. Highways per km² after controlling for population density
  6. Railroad per km² after controlling for population density
- Indicator gap formula (preserved structure): Infrastructure gapi,j,t = (Indicator i,t / average(Indicator j)EU,t − 1) * 100, where j enumerates the six indicators.
- Example (as in source): Infrastructure gapElect., ALB,t = (Installed capacity to gen. electricityALB,t / Installed capacity to gen. electricityEU,t − 1) * 100
- Highways and railroads: gaps computed relative to an EU-average adjusted for country population density via regression projection.
- Aggregate infrastructure gap: weighted average of indicator gaps with weights inversely related to the volatility of each indicator gap across time:
  - Aggregate infrastructure gapi,t = Σj wj * Infrastructure gapi,j,t
  - wj = 1 / (Σi Stdi(Infrastructure gapi,j) / # of countries)
  - Intuition: indicators with high volatility receive low weight; low-volatility indicators receive high weight.
- Robustness check: equal-weights aggregation produces similar results.
- Data sources for indicators:
  - Telephone/cell phone lines per capita: World Bank, WDI; completed with national statistics offices.
  - Broadband subscriptions per capita: World Bank, WDI; completed with national statistics offices.
  - Installed capacity to generate electricity per capita: International Energy Agency.
  - Air passengers carried per capita: World Bank, WDI; completed by national statistics offices.
  - Highways per km² after controlling for population density: International Road Federation; and country authorities’ data.
  - Railroad per km² after controlling for population density: World Bank, WDI; completed by national statistics offices.

*Source: IMF staff calculations and text from Public Infrastructure in the Western Balkans chapter provided in the source content.*

### Annex I. Measuring Infrastructure Gaps

### Annex I. Measuring Infrastructure Gaps

### Regression approach and controls
- Impact estimated with a simple convergence regression including an infrastructure gap index and standard controls.
- Control variables:
  - FDI-to-GDP ratio
  - Openness ratio
  - Government consumption-to-GDP ratio
  - Average inflation for the previous five years
  - Log population
  - Log of GDP per capita five years earlier
- Presence of lagged income allows for some dynamics analysis; with time, improvement in growth fades as country reaches higher income reflecting decreasing marginal return on capital.

### Endogeneity and instrumentation
- To address endogeneity of infrastructure, a two-stage least squares regression was used:
  - First stage: infrastructure gap instrumented with exogenous variables:
    - Log population
    - Percent of urban population
    - Distance to Brussels
    - Years since industrialization (based on Holzner, Stehrer, and Vidovic (2015))
    - Indices of political stability and fighting corruption from the World Governance Indicators database
  - Second stage: real GDP growth per capita regressed on instrumented infrastructure gaps and control variables.

### Sample
- Data sample: 39 European countries for the period 1997–2015.
  - Countries listed in the sample include: Albania, Austria, Belarus, Belgium, Bosnia and Herzegovina, Bulgaria, Croatia, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Kosovo, Latvia, Lithuania, Luxembourg, FYR Macedonia, Moldova, Montenegro, Netherlands, Norway, Poland, Portugal, Romania, Russia, Serbia, Slovak Republic, Slovenia, Spain, Sweden, Switzerland, Turkey, Ukraine, and United Kingdom.
- Observations: 491
- Countries: 39
- Adj. R-squared (baseline instrumental variables): 0.12

### Key regression findings (selected coefficients from Table AII.1)
- Infrastructure gap 6 sectors (weights = inverse standard deviations): 0.106 [0.025]***
  - Interpretation reported in text: Closing a (negative) infrastructure gap by 1 percentage point is estimated to be associated with 0.1 percent higher growth.
- Infrastructure gap 6 sectors (average 6 sectors): 0.086 [0.023]***
- Sector-specific coefficients:
  - Highways per 1000 square Km after controlling for density: 0.063 [0.029]**
  - Railways per square Km after controlling for density: 0.042 [0.015]***
  - Electricity generation installed capacity per capita relative to EU: 0.09 [0.022]***
  - Telephones (fixed + mobile) per 1000 inhabitants relative to EU: 0.003 [0.030]
  - Air passengers per capita relative to EU: 0.035 [0.022]
  - Broadband subscribers per 100 inhab. relative to EU: -0.016 [0.015]
- First-difference estimates of infrastructure gaps (not significant):
  - Diff. infrastructure gap 6 sectors (weights = inverse standard deviations): 0.461 [0.312]
  - Diff. infrastructure gap 6 sectors (average 6 sectors): 0.194 [0.123]
- Other notable control coefficients (baseline instrumental variables, Table AII.1 column I):
  - Openness (exports + imports / GDP): 0.017 [0.007]**
  - Government consumption to GDP: -0.272 [0.073]***
  - Average inflation past 5 years: -6.99 [3.344]**
  - Log population: 0.757 [0.261]***
  - Log GDP per capita in ppp (t − 5): -6.569 [1.033]***

### Robustness and interpretation notes
- Baseline and alternative specifications show consistent positive and significant effect of aggregate infrastructure gap on growth.
- Physical infrastructure (highways, railways, electricity generation capacity) shows the highest significant impact on growth among sectors.
- Telecommunications and broadband show weak or negative coefficients; possible explanation: gaps do not account for quality, a key feature of telecommunication infrastructure.
- Estimates based on first differences present positive but not significant impacts.
- Estimates relating income gaps and infrastructure gaps are presented in Table AII.2 (selected results below).

### Income-gap regressions (selected coefficients from Table AII.2)
- Dependent variable: Income gap relative to EU
- Infrastructure gap 6 sectors (weights = inverse standard deviations): 0.280 [0.064]***
- Infrastructure gap 6 sectors (average 6 sectors): 0.194 [0.050]***
- Openness (exports + imports) / GDP: 0.023 [0.012]*
- Government consumption to GDP: -0.434 [0.154]***
- Average inflation past 5 years: -9.142 [5.445]*
- Income gap with EU (t − 5): 0.763 [0.031]***
- Observations: 491; Countries: 39; Adj. R-squared: 0.98

### General equilibrium model calibration (summary from Annex III)
- Model used to simulate a public investment surge and interactions among GDP growth, public investment, and public debt; follows Berg and others (2015).
- Main calibrated parameters to reflect an average Western Balkan country:
  - Per capita potential GDP growth: 3 percent
  - Public debt-to-GDP ratio: 51 percent
  - Average tax rate: 18 percent
  - Public investment-to-GDP ratio: 5.2 percent
  - Real average domestic interest rate: 7 percent
  - Real average external interest rate: 5 percent
  - Productivity of capital: 20 percent
- Model features:
  - Small open economy, two sectors, multiple kinds of public debt
  - Private sector produces tradable and nontraded goods using private capital, public infrastructure, and labor
  - Public and private capital produced using imported inputs and nontraded goods
  - Two types of consumers: savers and hand-to-mouth
  - Government can finance investment via taxes, fees, domestic borrowing, external borrowing, or concessional external financing; tax rates can respond to public debt to ensure debt sustainability
  - Model accounts for inefficiencies/waste in public investment where only a fraction of expenditures augment public capital

*Source: 45547-western-balkans-public-infrastructure-020818 - Annex I. Measuring Infrastructure Gaps (IMF staff calculations and annex text).*

---


_Source: https://www.imf.org/-/media/files/publications/dp/2018/45547-western-balkans-public-infrastructure-020818.pdf_