Economic Convergence Scoreboard for the Western Balkans 2026

Annex B. Methodology

Background and rationale

The Economic Convergence Scoreboard for the Western Balkans (referred to hereafter as the Scoreboard) provides a comparative assessment of the convergence of the six Western Balkan economies relative to the European Union (EU). It aims to help the Western Balkan economies identify the main policy areas requiring focused attention to accelerate economic convergence with the OECD and the EU. Building on the pilot Scoreboard first published in 2023 and the more comprehensive 2025 edition, this third edition further consolidates the Scoreboard as the annual monitoring tool for tracking economic convergence of the Western Balkan economies towards the EU, following its acknowledgement as such in the Chair's Conclusions of the 2025 Berlin Process Leaders' Summit (Berlin Process, 2025[1]).

GDP per capita is the primary indicator used to assess economic convergence, as it provides a snapshot of relative prosperity across regions and countries. However, on its own, it offers limited insights into the underlying factors driving economic growth. The Scoreboard bridges this gap by combining policy-relevant analysis and data to understand better and monitor the key drivers of economic performance.

Drawing primarily on the OECD’s Western Balkans Competitiveness Outlook 2024, the Scoreboard covers six main policy clusters that are key to long-term economic growth:

1. Macro-financial stability: Macro-financial stability reduces uncertainty and lowers the cost of capital, creating the conditions necessary for sustained economic growth and private investment.

2. Infrastructure and Connectivity: Infrastructure and connectivity enable the efficient movement of goods, services and people, improving access to markets and boosting economic activities.

3. Skills: As a key component of human capital, skills contribute to workforce productivity and adaptability, enabling economies to adopt and innovate with new technologies.

4. Business Environment: A competitive business environment promotes investment, fosters entrepreneurship and cultivates an ecosystem where businesses can operate efficiently.

5. Digital Transformation: Digital transformation enhances productivity, innovation and access to information, enabling economies to modernise and compete effectively.

6. Greening: Greening economies ensures the sustainability of economic growth, reducing environmental risks and fostering resource efficiency.

The Scoreboard assesses the degree of convergence among the Western Balkan economies across these six clusters, utilising 43 indicators, up from 35 in the 2025 edition. The indicators were selected based on relevance, reliability and availability:

• Relevance: While the Scoreboard does not exhaustively cover all areas relevant to economic growth, it prioritises indicators identified in the OECD’s work and academic literature as having an impact on economic performance. Statistical tests were then conducted to explore the causal relationship between the indicators and GDP per capita, informing the final selection of indicators for the Scoreboard.

• Reliability: The Scoreboard mainly draws on comparable data from national statistical offices, leveraging their extensive co-operation with the OECD within the framework of the Western Balkans Competitiveness Outlook and the SME Policy Index for the Western Balkans and Türkiye. It also draws on data from reputable international sources, such as the OECD, Eurostat and the World Bank.

• Availability: The limited statistical coverage in the Western Balkan economies, compared to OECD and EU members, influenced the selection of indicators. Indicators that lacked data for any economy across all periods, and for which no viable alternatives were available to allow for imputation, were excluded. For indicators with partial time-series coverage, missing data were addressed through imputations.

The entire list of indicators considered for the Scoreboard is shown in Table A B.1.

Table A B.1. List of indicators of the measurement framework

Cluster name	Sub-group	No.	Indicator
Infrastructure and Connectivity	Transport infrastructure	IC.01	Road infrastructure density (km per area of 100 km2)
		IC.02	Rail infrastructure density (km per area of 100 km2)
		IC.03	Rail freight intensity (tonnes-kilometres per thousand units of GDP expressed in 2021 USD in purchasing power parity PPP)
	Energy infrastructure	IC.04	Power losses (% of total energy injected into distribution and transmission grids)
		IC.05	Electricity cost for non-household consumers (2021 USD PPP per kilowatt-hour)
	Digital infrastructure	IC.06	4G penetration (% of total population)
		IC.07	Broadband speed (megabytes per second)
Skills	Putting skills to effective use	SK.01	Labour productivity (2021 USD in PPP)
		SK.02	Employment rate (% of population aged 20-64 years)
		SK.03	Gender gap in labour force participation (% of population aged 15-64 years)
	Skills development	SK.04	PISA performance: Mathematics
		SK.05	Youth NEET rate (% of 15-24 years old population)
		SK.06	Labour force with advanced education (% of total working-age population)
		SK.07	Lifelong learning (% of population aged 25-64 years)
	Leveraging skills for innovation	SK.08	Patent applications (per million inhabitants)
Business Environment	Enhancing access to finance and investment	BE.01	Domestic credit to the private sector (% of GDP)
		BE.02	Bank non-performing loans (% of total gross loans)
		BE.03	Private investment (% of GDP)
		BE.04	FDI net inflows in tradable sectors (% of GDP)
	Strengthening the level-playing field	BE.05	Control of corruption (composite score)
		BE.06	Estimated informal employment (% of total employment)
	Promoting trade and entrepreneurship	BE.07	Trade flows (2021 USD PPP per capita)
		BE.08	New business creation density (new business creations per million inhabitants)
Digital Transformation	Digital society	DG.01	Population with digital skills (% of population aged 16-74 years)
		DG.02	Internet use by individuals (% of total population)
		DG.03	Internet use for interacting with public authorities (% of population aged 16-74 years)
		DG.04	Internet use for purchases (% of population aged 16-74 years)
	Digital businesses	DG.05	Businesses with a website (% of enterprises with 10 employees or more)
		DG.06	Businesses using AI technologies (% of enterprises with 10 employees or more)
	Digital economy	DG.07	ICT services exports (% of GDP)
	Cybersecurity	DG.08	Number of secure Internet servers (per million inhabitants)
Greening	Accelerating decarbonisation	GR.01	CO2 emissions intensity (kg per unit of GDP in 2021 USD PPP)
		GR.02	Installed renewable power capacities (% of total installed power capacity)
	Enhancing resource use	GR.03	Energy productivity (GDP in 2021 USD PPP per kg of oil equivalent consumption)
		GR.04	Water productivity (GDP in 2021 USD PPP per cubic meter of total freshwater withdrawal)
		GR.05	Waste intensity (generated municipal waste per unit of GDP in 2021 USD PPP)
		GR.06	Municipal waste in landfills (% of municipal waste)
	Environmental quality	GR.07	Premature deaths from air pollution (deaths per million inhabitants)
Macro-financial stability	Monetary stability	MS.01	Inflation (% of consumer price index)
		MS.02	Nominal exchange-rate stability vis-à-vis the Euro (percentage points change range)
	Fiscal stability	MS.03	Government balance (% of GDP)
		MS.04	Long-term government interest rates (%)
	External sustainability	MS.05	Current account balance (% of GDP)

During the selection process for the Scoreboard indicators, several additional indicators were initially considered for inclusion due to their relevance to economic growth. However, many were ultimately excluded due to limited data availability. Consequently, alternative indicators were selected to capture comparable aspects of growth dynamics and to best explain variations in economic performance.

Convergence with the EU for each indicator was measured over four sub-periods, 2014-2016, 2017-2019, 2020-2022, and 2023-2024, rather than annually. Each data point represents the average value of the available years within each sub-period. This approach allowed for 1) integrating the inclusion of indicators not updated annually; 2) smoothing out cyclical variations or shocks; and 3) overcoming data inconsistencies or gaps.

Scoring

General scoring

The Scoreboard leveraged two methods to calculate scores and rankings:

1. Each individual indicator score represents the ratio of a specific Western Balkan economy's value to the EU average. The sample used to calculate the EU average includes the 27 EU member states. In addition, the 2026 edition introduces a supplementary informative benchmark comprising EU member states that share similar structural and geographical conditions with the Western Balkan economies, namely Bulgaria, Croatia, Greece, Hungary, Romania and Slovenia, providing a complementary perspective on the Western Balkan economies’ convergence performance.

2. Each cluster score was derived from a dedicated composite indicator using a three-step method:

• Outlier treatment: The individual values of the Western Balkan economies and the EU member states were considered, leading to a total sample size of 33 individual economies. Outliers were then detected using a skewness/kurtosis rule and treated with the Winsorisation method, adjusting extreme values to reduce their impact without removing them from the dataset. This step is crucial to ensure that the aggregation results are not distorted by extreme values.

• Normalisation: This process standardised all data points to a comparable scale, regardless of their original units or ranges. The distance-to-target (DTT) scales data based on the proximity of each value to a predefined target, ensuring that the resulting values reflect how close each observation is to the desired goal. The target values were the EU average values. The DTT score for the indicator $x$, ${DTTscore}_{x,i}$, was calculated as follows:

  ${DTTscore}_{x,i}=\frac{x_i-{MIN}_{x_i}}{{TARGET\_EU}_x-{MIN}_{x_i}}$

  where $x_i$ denotes the indicator value for the economy $i$ among Western Balkan and EU economies, ${MIN}_{x_i}$ denotes the minimum value for indicators among Western Balkan and EU economies, and ${TARGET\_EU}_x$ denotes the EU average’s value.

  • Lower-is-better indicators: It should be noted that for indicators where higher values are not desirable, such as IC.04 (power losses), SK.03 (gender gap in labour force participation), SK.04 (youth NEET), BE.01 (non-performing loans), BE.06 (estimated informal employment), GR.01 (CO2 emissions intensity), GR.05 (waste intensity), GR.06 (municipal waste in landfills) and GR.07 (premature deaths from air pollution), the inverse value of $x_i$, $\frac{1}{x_i}$ has be en considered.

  • Converged indicators: A consequence of the DTT approach is that a Western Balkan economy whose performance is close to the EU average may nonetheless receive a low DTT-based cluster score if the minimum observed in the sample lies close to its own value, an outcome that would misrepresent the degree of convergence achieved. To address this, the Scoreboard sets aside the DTT normalisation for indicators where the ratio of all Western Balkan economies to the EU average falls between 90% and 150%, deriving the cluster score contribution directly from that ratio, scaled to an EU average of 100. The list of indicators treated as converged indicators is the following: “4G penetration” (IC.06) for the periods 2020-2022 and 2023-2024, “Labour force with advanced education” (SK.06) for the periods 2017-2019 and 2023-2024, Private investment (BE.03) for the period 2020-2022 and Internet use by individuals (DG.02) for the period 2023-2024.

• Weighting and aggregation: Indicators were then aggregated into dimension scores. The average score for each cluster was computed as the simple average, giving each indicator equal weight. Each DTT score ${DTTscore}_{x,i}$ was assigned to cluster $c$, leading to ${DTTscore}_{x,c,i}$.The cluster $c$ average score for economy $i$ was calculated as:

  ${CLUSTERscore}_{c,i}=\frac{\sum _i^{X_c}{DTTscore}_{x,c,i}}{X_c}$

  where $X_c$ denotes the total number of indicators $x$ in cluster $c$.

Macro-financial stability cluster scoring

When it comes to the Macro-financial stability cluster, the Scoreboard leveraged a specific method to calculate scores and rankings:

Each individual indicator score reflects the degree of compliance with the reference value defined for each criterion, drawing on the convergence criteria established in the context of EU accession. For each economy and each year, the score is set at 100 if the indicator value meets or does not exceed the applicable threshold. Where the threshold is breached, the score falls below 100 in proportion to the magnitude of the deviation. Five threshold-based scoring formulas are applied, one per indicator, as follows:

a. MS.01 (Inflation rate): If the inflation rate is below or equal to the reference value from the European Central Bank (ECB)’s Convergence Reports (see Table A B.2), the score is set at 100. If the inflation rate exceeds the reference value, the score progressively decreases. The scoring formula is defined as:

${Score}_{MS.01,i}=\frac{100}{1+Max(0;{Inflation\mathrm{ }rate}_{\mathrm{ }i}-{Reference\mathrm{ }value}_{x_i})}$

The reference value is defined by the ECB in its Convergence Reports as the average inflation rate of the three best-performing EU member states in terms of price stability, plus 1.5 percentage points.

Table A B.2. Reference values for MS.01 (2014-2024)

	2014	2015	2016	2017	2018	2019	2020	2021	2022	2023	2024
Reference values	1.7		0.7		1.9		1.8		4.9		3.3

Note: Reference values are published biannually.

Source: (European Central Bank, 2025[2])

b. MS.02 (Nominal exchange-rate stability): For the exchange-rate stability criterion, the reference threshold is derived from the fluctuation margin used under the ERM II (European Commission, 2020[3]), under which participating currencies are expected to remain within a 15% band around a central exchange rate. As Western Balkan economies cannot participate in ERM II and therefore do not have a central exchange rate against the euro, exchange-rate stability is assessed using the observed fluctuation range of the exchange rate vis-à-vis the euro over a given reference period, defined as:

${Range}_{i,t}=\left(\frac{{max}_{i,t}-{min}_{i,t}}{min,it}\right)\times 100$

If the fluctuation range does not exceed 15%, the score is set at 100, and as the fluctuation range exceeds the reference value, the score progressively declines towards 0. The scoring formula is defined as:

${Score}_{MS.02,i}=\frac{100}{1+Max(0;{Range}_{i,t}-15)}$

c. MS.03 (Government balance): The ECB Convergence Reports assess compliance with the excessive deficit criterion established under the Maastricht Treaty. The reference value is set at a general government deficit of 3% of GDP (European Council, 2026[4]). If the government balance is equal to or above the reference value, the score is set at 100. As the deficit exceeds the reference value, the score progressively declines towards 0. The scoring formula is defined as:

${Score}_{MS.03,i,t}=\frac{100}{1+Max(0;-3-{Government balance rate}_{i,t})}$

d. MS.04 (Long-term government bond interest rate): The ECB Convergence Reports assess whether an economy’s average nominal long-term interest rate exceeds by more than 2 percentage points that of the three best-performing EU member states in terms of price stability. If the long-term rate is below or equal to the reference value (see Table A B.3), the score is set at 100. As the interest rate exceeds the reference value, the score declines progressively toward 0. The scoring formula is defined as:

${Score}_{MS.04,i,t}=\frac{100}{1+Max(0;{Long term government bond interest rate}_{i,t}-{Reference value}_{x_{i,t}})}$

Table A B.3. Reference values for MS.04 (2014-2024)

	2014	2015	2016	2017	2018	2019	2020	2021	2022	2023	2024
Reference values	6.2		4		3.2		2.9		2.6		4.8

Note: Reference values are published biannually.

Source: (European Central Bank, 2025[2])

e. MS.05 (Current account balance): For the current account criterion, the ECB Convergence Reports assess external sustainability using the thresholds defined under the MIP. According to the MIP scoreboard, the three-year backwards-moving average of the current account balance should remain within the range of -4% to +6% of GDP (European Commission, 2026[5]). If the three-year backwards-moving average of the current account balance falls within this range, the score is set at 100. As the balance moves beyond either reference value, the score declines progressively toward 0. The scoring formula is defined as:

${Score}_{MS.05,i,t}=\frac{100}{1+Max(0;-4-{Current account balance}_{x_{i,t}}; {Current account balance}_{x_{i,t}}-6)}$

The macro-financial stability cluster score is the simple average of the five individual indicator scores, computed for each economy and each sub-period (2014-2016, 2017-2019, 2020-2022 and 2023-2024), with each indicator assigned equal weight. The macro-financial stability cluster score is expressed as:

${CLUSTERscore}_{MS,i}=\frac{\sum _i^5{Score}_{MS.x,i}}{5}$

Where ${Score}_{MS.x,i}$ denotes the individual score of indicator x (MS.01, MS.02, MS.03, MS.04, MS.05). A cluster score of 100 indicates that an economy complies with all the macro-financial stability criteria.

Limitations

The Scoreboard is a valuable tool for Western Balkan policymakers, helping them identify the bottlenecks that hinder faster economic convergence. Yet, the method has certain limitations, as outlined in Table A B.4.

Table A B.4. Limitations of the Economic Convergence Scoreboard for the Western Balkan economies 2026

Analytical framework	While the Scoreboard focuses on areas crucial to accelerating the economic convergence of the Western Balkan economies, it does not cover all relevant areas in exhaustive detail.
	Constraints in data availability and comparability limited the choice of indicators.
	The Scoreboard data cover a decade, preventing analysis over a longer time frame.
Methodology	Despite the benefits of measuring convergence in four time periods, the Scoreboard does not fully capture the most recent developments and state-of-play.
	The DTT method measures proximity to the target without explaining the underlying causes of disparities.
	The DTT method assumes a linear relationship between the distance to the target and the degree of convergence. This assumption fails to account for non-linear relationships, in which the impact of improvements could vary across stages. The treatment applied to converged indicators partly addresses this limitation for economies close to the EU average, while introducing a threshold effect between the 90% and 150% marks.
	Assigning equal weight to indicators when calculating cluster scores may not adequately reflect their varying importance in driving economic growth.
	Assigning equal weight to the EU's 27 economies to calculate EU averages may have introduced bias in the EU's actual performance by over-weighting smaller economies.
	The macro-financial stability cluster relies on a threshold-based scoring approach, drawing on the Maastricht convergence criteria, rather than the DTT normalisation applied to the other five clusters. This methodological distinction limits the direct comparability of cluster scores across the Scoreboard.

References

[1] Berlin Process (2025), Western Balkans Summit on the Berlin Process, 22 October 2025, Chair’s Conclusions, https://www.berlinprocess.de/uploads/documents/berlin-process-2025-chairs-conclusion_1764661015.pdf.

[2] European Central Bank (2025), Convergence Report, https://www.ecb.europa.eu/press/other-publications/convergence/html/index.en.html (accessed on 1 June  2026).

[5] European Commission (2026), Macroeconomic imbalance procedure scoreboard, https://economy-finance.ec.europa.eu/economic-governance-framework/macroeconomic-imbalance-procedure/scoreboard_en (accessed on 12 August 2026).

[3] European Commission (2020), ERM II – the EU’s Exchange Rate Mechanism, https://economy-finance.ec.europa.eu/euro/enlargement-euro-area/adoption-fixed-euro-conversion-rate/erm-ii-eus-exchange-rate-mechanism_en (accessed on 12 August 2026).

[4] European Council (2026), Excessive deficit procedure, https://www.consilium.europa.eu/en/policies/excessive-deficit-procedure/ (accessed on 13 August 2026).