Macro-developments
We start this analysis with an overview of Serbia’s socio-economic convergence with the European Union over the last 10-15 years. This period starts when the unemployment rate in Serbia stood at 20% and employment at 47.2%. Investment over the previous 10 years had amounted to about 20% of GDP on average and the GDP still stood well below its pre-1990 level. The first years of the period were marked by a sharp fiscal consolidation that redressed sharp macroeconomic imbalances accumulated during the previous period. Between 2014 and 2018, the fiscal balance was improved from a deficit of 5.94% to a fiscal surplus of 0.61% of GDP. The public debt to GDP ratio was stabilized at first and has been declining since (with the exception of 2020) to 45% at end 2025.
The key to the success of the early period, other than the radically increased fiscal discipline, had been a strong pick up in FDI inflows, mainly from the European Union. Three factors were key in intensifying these flows from around 3% of GDP before 2015 to over 7% in 2021-2022. First, it is in this period that the outflow of European, especially German, capital intensified again in search of efficiency gains, as well as began nearshoring. Serbia was near, and with a deeply depressed economy, profit opportunities were extremely high. Second, more than a decade after the last political crisis and showing visible commitment to fiscal consolidation, Serbia finally was able to overcome its image of a politically risky destination. Last but not least, the authorities adopted an aggressive investment promotion policy that both facilitated entry, as we will argue by short-circuiting rather than cutting red-tape, and reinforced the high profit opportunities. However, as we will see, they also generated distortions that are increasingly costing the country today.
Table 1: Three macroeconomic periods — key drivers of growth and context
| Indicator | 2015-2017 | 2018-2021 | 2022-2024 | 2025 |
|---|---|---|---|---|
| GDP growth | 2.2 | 4.1 | 3.5 | 2.3 |
| Drivers (% of GDP) | ||||
| Investment contribution to GDP growth | 1.0 | 2.3 | 1.5 | 1.6 |
| Government CAPEX | 2.7 | 5.2 | 6.9 | 7.1 |
| FDI | 5.8 | 7.2 | 6.4 | 3.7 |
| Exports contribution to GDP growth | 3.7 | 3.7 | 4.5 | 4.3 |
| Context | ||||
| Inflation, average annual % change | 2.0 | 2.4 | 9.7 | 3.9 |
| Unemployment rate (%), end of period | 14.5 | 11.0 | 8.6 | 8.2 |
| Vacancies-unemployment ratio | 0.11 | 0.16 | 0.32 | 0.48 |
Note: Simple averages of annual growth rates. Primary fiscal balance shows change during period. Vacancies-unemployment ratio computed as number of vacancies / number of unemployed persons.
Soon after the acceleration of FDI, the government also succeeded in accelerating general government investment which peaked in 2024 as a share of GDP. By comparison, few were the years when they surpassed 3.5% in the entire previous transition period.
Exports have also been a major driver of growth. Domestic industry exports grew as well so that altogether they grew faster than those of CEU peers throughout the decade. They also grew faster than their destination markets, diversifying both in product and geographic structure. Eventually, the GDP share of goods and service exports caught up with those of the less integrated CEU countries.
Nearshoring capital came with ready markets, so it is not surprising that they would lead to Serbia’s market expansion — it was as if a “market acquisition.” However, Serbian SMEs also showed impressive competitive strength throughout most of this period, diversifying and conquering market share. They approximately kept pace with the rest of Serbia’s exports, and showed remarkable adaptability and resilience during the COVID crisis, which only mildly affected their performance.
Serbia’s FDI continued to perform strongly until 2024, despite Europe being badly hit by the energy crisis and Russia’s war on Ukraine. To understand this it is important to understand its structure. At first, this was due to the carryover effect from the intense FDI in manufacturing planned and implemented just before the crisis hit. The share of manufacturing in FDI peaked in 2021. However, a fast decline in manufacturing FDI was not perceived at first as just at the time three types of investment picked up. One was into mining and ore processing, following the Chinese acquisition of a major copper mine and smelter plant. Second was a pick up in IT investments, likely due to the Ukraine-Russian war. Finally, the acceleration of public investment has been accompanied by strong inflows into expansion of construction sector capacity and related knowledge-services, related to the arrival of foreign companies contracted to implement it.
In 2025 GDP growth sharply slowed down to an estimated 2.0%. This was primarily due to a near halving of FDI inflows. Many observers attributed this to the political protests that had been gathering steam since the tragic fall of the Novi Sad Railroad Station Canopy the previous November. While there is no doubt that political unrest eventually dampened investor spirits, a careful analysis of the FDI structure and timing suggest the drop should have been expected. Manufacturing was already down, and the expansion of construction and related service capacities could not last indefinitely. As public investment also slowed down (in line with plans) growth was supported by exports and by an unusually high increase in household consumption that was the result of very large real public and private wage increases.
Socio-economic Convergence
The results of the above policies are reflected in a gradual convergence in key economic indicators such as GDP per capita, income per capita and productivity. Albeit the rate of employment can be said to have converged faster, aided by a 16% decline in the size of the working age population, in the other economic dimensions the convergence has been slow and uneven. At the current rate, productivity is projected to converge only in 25 years. Given that in the observed period Serbia was the recipient of very high levels of much needed capital, this is puzzling. Labor productivity should have grown faster.
This economic picture is complemented by an even more mixed picture in regard to governance and social indicators. These indicators generally started the period at relatively higher levels than the economic ones, and a slower convergence was to be expected. However, a number of them, especially those measuring the quality of governance, have been diverging, and especially in more recent years. There are also strong indications that important aspects of Serbia’s infrastructure quality have been diverging as well. This should not be unexpected, given that in recent years Serbia has succeeded in raising its public investment to commendably high levels, unlike any other WB country.
While aggregate indicators point to gradual improvement, convergence outcomes are uneven across policy domains and economic systems. Advances in integration and digitalization coexist with persistent gaps in productivity, human capital formation, governance, and core infrastructure, limiting the economy’s capacity to translate growth into durable efficiency gains. At the same time, convergence challenges are not only external but also internal: pronounced regional disparities and uneven social outcomes shape aggregate performance and reflect differences in access to opportunities, public services, and institutional capacity. Together, these patterns indicate that Serbia’s convergence path is shaped as much by governance quality and capability-building as by macroeconomic conditions, and that internal dispersion is an integral part of the economic governance challenge rather than a secondary distributional issue.
Table 2: Convergence Dashboard
| Area | Indicator | Serbia 2015 | Serbia 2023/24 | Unit | Comparison to EU | Time to converge | Structural impact |
|---|---|---|---|---|---|---|---|
| Economic | GDP per capita | 11,300 | 20,370 | Euro, current prices at PPS | 51% | 29 years | Income, wages, and productivity determine unit labour costs and the ability of firms to compete and upgrade in tradable sectors. |
| Average Gross Wage | 506 | 1,156 | Euro, current | 38% | 26 years | ||
| Output per Worker | 30,810 | 40,260 | Euro, current prices at PPS | 57% | 25 years | ||
| Social & Employment | Total employment | 18.23 | 66.3 | %, age group 15-64 | 94% | 6 years | Weak labour market and social outcomes reduce effective labour supply and limit inclusive productivity growth. |
| Poverty | 0.04 | 0.042 | Share of post-tax income of bottom 20% | 59% | 77 years | ||
| Gender employment gap | 14.23 | 10.9 | Percentage points | 83% | 100+ years | ||
| Health | Public spending on health | 5.08 | 5.95 | % of GDP | 91% | 4 years | Health outcomes and system capacity affect workforce productivity, labour participation, and long-term cost competitiveness. |
| Life expectancy | 75.29 | 76.22 | Years | 94% | Diverging | ||
| Number of physicians | 29.26 | 32.88 | Per 10,000 inhabitants | 79% | Diverging | ||
| Education | Public spending | 3.85 | 3.24 | % of GDP | — | — | Education spending, learning outcomes, and skills shape human capacity. |
| Average PISA scores | 439.87 | 447.46 | Score | 93% | 13 years | ||
| Enrolment in tertiary education | 58.95 | 73.21 | % of tertiary-age population | 92% | Diverging | ||
| Governance | Control of Corruption | 43.63 | 41.06 | Estimate | 60% | Diverging | Governance quality influences investment predictability, transaction costs, and the effectiveness of market competition. |
| Government Effectiveness | 49.95 | 50.15 | Estimate | 73% | Diverging | ||
| Regulatory Quality | 52.73 | 52.76 | Estimate | 74% | 56 years | ||
| Environment | Premature deaths due to PM 2.5 | 271 | 228 | Rate | 32% | 98 years | Environmental performance affects competitiveness through health costs, energy efficiency, and compliance with EU climate standards. |
| Renewable energy share | 33.41 | 38.79 | % of total capacity | 67% | Diverging | ||
| Carbon intensity | 1,176.19 | 864.26 | kg per thousand euro of GDP | 21% | Diverging | ||
| Digitalization | Households with internet access | 63.8 | 88.85 | Percent | 94% | 6 years | Digital access and skills determine firms’ efficiency, scalability, and integration into modern value chains. |
| Use of e-government services | 17.88 | 45.98 | Percent | 82% | 3 years | ||
| Value added in ICT | 5.96 | 9.55 | % of total value added | 181% | Converged | ||
| Infrastructure | Length of motorways | 78.31 | 108.81 | Km per 10,000 sq km | 52% | 22 years | Transport and energy infrastructure shape connectivity, logistics costs, and the feasibility of industrial upgrading. |
| Railway tracks | 609.27 | 564.41 | Km per 10,000 sq km | 63% | Diverging | ||
| Electricity installed capacity | 825.32 | 994.11 | MW per 10,000 sq km | 24% | Diverging |
Source: Converge2EU: Interactive database and dashboard. Retrieved January 23rd, 2026, from https://converge2.eu/
Serbia’s aggregate economic convergence with the European Union has progressed over the past decade, but the pace of catch-up remains slow and structurally constrained. GDP per capita expressed in purchasing power standards has reached around 51% of the EU average, with the implied time to convergence remaining close to three decades. Productivity indicators convey a consistent message: output per worker in PPS terms remains well below the EU average, in fact below all EU member levels. Gross value added per employee in PPS terms places Serbia at roughly one half of the EU average in the latest year. Although Serbia has recorded positive productivity growth over the period from 2010 to 2024, the speed of convergence has remained modest when compared with several Central and Eastern European economies, including Romania, Bulgaria, Lithuania, Poland, and Czechia, which achieved faster growth despite starting from higher productivity levels. The current convergence rate therefore remains insufficient, particularly given that the observed catch-up is measured in PPS terms; under these conditions, convergence in nominal terms appears increasingly distant if Serbia’s growth dynamics remain unchanged.
Aggregate convergence indicators conceal a renewed and intensifying spatial concentration of economic development within Serbia. Regional disparities increased during the 2000s, declined slowly after the global financial crisis, stabilising for several years after 2013, and widening again since 2019, diverging from the broader European trend of declining regional concentration. By the latest available year, its level of regional dispersion is comparable to Romania and Bulgaria, though below Ireland, and substantially higher than in neighbouring countries such as Croatia and Slovenia. This renewed divergence has occurred alongside sustained population decline across nearly all regions between 2017 and 2025, with Belgrade remaining broadly demographically stable. Rather than being offset by demographic contraction, regional disparities appear to have been reinforced, reflecting the increasing concentration of economic activity, skills, and services in the capital and a limited number of larger urban centres. Under such conditions, economic development that is weakly anchored in regional structures may face growing challenges in maintaining momentum over the medium to long term, particularly with respect to labour supply, productivity diffusion, and social cohesion outside the core urban areas.
Social indicators
| Indicator | Serbia | EU27 | Serbia as % of EU | Convergence | Source and note |
|---|---|---|---|---|---|
| GINI | 31.90 | 29.40 | 92.2% | Converging | Inequality measure (0 = equal, 100 = highly unequal) — Eurostat |
| PISA Inequality | 233.21 | 251.81 | 108.0% | Converged | Average difference in score across all three categories between students at the top 10% and the bottom 10% of the socio-economic status distribution — OECD PISA |
| Unmet needs | 6.80 | 2.50 | 36.8% | Diverging | Self-reported unmet needs for medical examination, too expensive or too far to travel or waiting list — 16 years and over |
| Inequality | 0.33 | 0.28 | 84.2% | Converging | Share of post-tax income of the top 10% of equal-split adults — World Inequality Database |
Source: Converge2EU: Interactive database and dashboard, Eurostat, OECD
Progress in social outcomes has been slower — which is understandable given that they generally stood at higher levels — but in many ways has actually reversed. Social convergence is a goal in its own right, but its absence raises questions about the sustainability of economic convergence in the absence of comparable social convergence. Labour market participation has converged relatively quickly, with total employment close to the EU average, and income inequality and poverty indicators have improved in recent years. By contrast, underlying access and capability constraints remain pronounced. In the health sector, public spending as a share of GDP is close to EU levels, yet outcomes and access indicators remain weak. Self-reported unmet medical needs are more than twice the EU average and continue to diverge, while households finance an unusually large share of healthcare costs out of pocket, accounting for over 32% of total current health expenditure in 2023, compared with around 15% in the EU. This pattern points to persistent inefficiencies in service provision and a system oriented primarily towards curative care rather than prevention, resulting in high private costs despite substantial public spending. In education, public expenditure has fallen further below EU benchmarks, while learning outcomes and tertiary enrolment have remained relatively resilient, reflecting legacy effects rather than sustained investment. These patterns seem to suggest that recent social improvements have not translated into durable gains in human capital, constraining productivity growth and weakening the foundations of long-term convergence.
Governance and digitalisation outcomes increasingly shape Serbia’s convergence prospects, with advances in digital capacity occurring alongside a deterioration in institutional performance. Effective convergence requires stable institutions that support investment and productivity growth; continued weakening of governance therefore represents a direct constraint on economic progress. Indicators of control of corruption and government effectiveness continue to diverge from EU benchmarks, reflecting persistent weaknesses in accountability, policy implementation, and administrative capacity. These trends have been reinforced by prolonged political instability and elevated institutional uncertainty since late 2024, following the canopy collapse in Novi Sad that resulted in 16 fatalities. Digitalisation presents a more mixed picture. Access-related indicators and the use of e-government services have converged relatively quickly and function in practice, but they face significant shortfalls and require continuous upgrading to support more complex administrative and business processes. The ICT sector generates a high share of value added, largely reflecting outsourced activities, while the presence of major international firms (Microsoft, Ubisoft, among others) points to some progress in higher-value activities. However, uptake of digital technologies across the wider business sector and public administration remains limited.
Environmental and infrastructure indicators point to a growing misalignment between Serbia’s convergence path and the EU’s ongoing transition towards lower-carbon energy systems, rail-based transport, and stricter environmental standards. Environmental outcomes remain particularly weak, with pollution-related health impacts and carbon intensity diverging further from EU benchmarks, indicating slow progress in decarbonisation and energy efficiency. Infrastructure development shows an uneven pattern. Investment in road infrastructure has supported improvements in connectivity, yet rail infrastructure and electricity generation capacity continue to lag behind EU levels and follow diverging trajectories. This asymmetry limits the contribution of infrastructure to productivity growth, regional integration, and the green transition. The evidence suggests that the challenge lies less in the overall volume of public investment than in its structure and execution, as spending has not been sufficiently aligned with infrastructure segments that carry the highest productivity and transition effects.