Serbia’s labour market has reached a turning point that is more important than the headline employment numbers suggest. After years in which rising employment helped drive economic growth, the country is now facing a period in which job creation can no longer be counted on as a major source of expansion. The shift is not dramatic in statistical terms, but it is structurally significant: Serbia’s old model of growth, built around cheap labour, large foreign direct investment inflows and labour-intensive manufacturing, has reached its natural limit.
The latest warning comes from the Fiscal Council, which has noted that Serbia recorded a mild decline in employment in 2025 and that the government’s own fiscal projections no longer assume meaningful employment growth over the next three years. According to the Labour Force Survey, the number of employed people fell by about 1% in 2025, while registered-employment data point to a smaller decline of around 0.2%. The unemployment rate edged up only slightly, from 8.6% to 8.7%, which means Serbia is not entering a classic unemployment crisis. The problem is deeper: the labour pool that supported the previous growth cycle is shrinking, ageing and no longer cheap enough to sustain the same investment model.
This is the point that matters for investors and policymakers. Serbia’s earlier growth formula worked because it combined relatively low wages, high unemployment, public capital investment and large foreign investments in sectors such as mining, basic manufacturing, automotive components, textiles, footwear, cable production and other labour-intensive industries. That model was rational at the time. It absorbed unemployed workers, brought production into weaker regions, increased exports and helped Serbia grow faster than some regional peers between 2018 and 2024. But the conditions that made it work have changed.
The working-age population has contracted sharply. The number of people aged 20 to 64 has fallen by an estimated 540,000 over the past decade, or 12.5%, reflecting demographic decline and continued emigration. At the same time, employment expanded, informal workers moved into formal jobs, and employers were forced to offer higher wages and better conditions. What once looked like a deep reservoir of available labour has become a much tighter market. Serbia is no longer the low-wage platform it was a decade ago.
Wage dynamics have accelerated the shift. The Fiscal Council has warned that salaries have for years grown faster than productivity, partly because of strong administrative increases in the minimum wage. That did not immediately destroy competitiveness because Serbia started from a low wage base. A decade ago, labour in Serbia was still around 15% cheaper than in Central and Eastern European peers for the same value of output. Today, that cost advantage has largely disappeared. Labour costs relative to production have effectively converged with comparable economies, but Serbia has not yet moved far enough into higher-productivity, technology-intensive activities.
That is the uncomfortable middle ground in which Serbia now finds itself. The country is becoming too expensive for the simplest forms of labour-intensive production, but it has not yet become sufficiently productive, technologically advanced or institutionally predictable to replace those activities at scale. This is the classic middle-income trap, not as an abstract academic concept, but as an operating problem visible in factories, local labour markets and investment decisions.
Several recent closures illustrate the point. Low-value foreign manufacturers that came to Serbia for cheap labour, subsidies and manageable operating costs are reassessing their position as wages and energy costs rise and subsidy periods expire. Benetton closed operations in Niš. Leoni shut a plant in Malošište, affecting almost 1,900 workers. Turkish jeans manufacturer Džinsi left Leskovac after years of operations and state incentives. Danish company Kentaur withdrew from Vranje. These cases are not isolated scandals; they are early signals of a wider adjustment in labour-intensive manufacturing, especially in southern Serbia, where such investors were often central to local employment.
The departure of low-productivity employers is not automatically negative. In a successful upgrading process, cheaper industries leave as wages rise, while more advanced industries arrive to replace them. That is how economic convergence should work. The problem in Serbia is that replacement capacity remains insufficient. The country has strong pockets of higher-value activity in IT, engineering services, advanced business services, some parts of automotive supply, energy, construction engineering and professional services, but these sectors are not yet large enough to absorb all workers leaving lower-value manufacturing or to lift national productivity decisively.
This creates a difficult labour-market split. Low-skilled workers are caught between two pressures. On one side, they have become too expensive for the simplest and lowest-paid production jobs when compared with cheaper destinations. On the other side, many do not have the qualifications needed for more complex, better-paid work. Serbia is therefore reaching labour-market limits while still having unemployment of around 8.5%, whereas many Central and Eastern European countries reached similar constraints closer to full employment, with unemployment near 4%. That difference matters. Serbia has not exhausted unemployment because everyone found good jobs; it has exhausted the part of the labour force that fits the old investment model.
The beginning of 2026 confirms that the shift is continuing. According to the government’s fiscal-strategy references to social-insurance registry data, average employment in the first quarter was around 2.312mn, down 0.2% from the same period of 2025. The decline was linked to weaker activity in labour-intensive manufacturing branches, especially clothing and leather, electrical equipment, metal products and trade. Growth in services partly offset the fall, with stronger employment contributions from information and communications, administrative and professional services, accommodation and social protection. The structure of employment is changing, but not yet fast enough to remove the risks.
For Serbia’s industrial policy, the implications are direct. The country can no longer rely on the simple promise of low-cost labour to attract foreign investors. Subsidies for large employers may still bring projects, but the quality of those projects matters more than before. Paying public money to attract firms whose only competitive logic is cheap labour now carries higher risk. If wages continue rising and productivity does not follow, such investors may leave once incentives expire. The state then inherits local unemployment, political pressure and empty industrial facilities.
A smarter investment strategy would focus less on headline job numbers and more on value added per employee, supplier depth, technology transfer, export sophistication and training intensity. A factory employing 500 workers in low-margin assembly may look attractive in a local announcement, but its long-term contribution can be weaker than a smaller engineering, software, machinery, energy-technology or industrial-services investment with higher productivity and stronger domestic linkages. Serbia’s labour constraint makes this distinction unavoidable.
Education is central to the transition. Serbia does not lag dramatically in the expected number of years of schooling. The expected duration is around 13.3 years, close to the developed EU figure of 13.6 years. But when adjusted for actual acquired knowledge, the effective level falls to below 10 years. That gap is damaging. It means the formal duration of education overstates the real capabilities available to the economy. International PISA testing points in the same direction, with Serbian pupils lagging behind EU and regional benchmarks. For a country trying to move into technology-intensive activities, weak learning outcomes are not a social-policy detail; they are an economic-growth constraint.
This is especially important because advanced sectors require more than basic literacy and formal diplomas. They need technicians who can work with industrial automation, electricians trained for modern energy systems, mechanical engineers familiar with quality documentation, software developers, data specialists, environmental engineers, process-control technicians, logistics planners, compliance professionals and managers capable of operating under international standards. Serbia has some of these capabilities, but not enough of them, and not distributed evenly across the country.
The institutional environment is the other major constraint. Productivity growth is not produced only by buying newer machines or attracting more sophisticated investors. It depends on legal certainty, predictable regulation, efficient courts, low corruption, fair competition, stable tax policy and credible public administration. These are precisely the areas where Serbia has struggled for years. A technology-intensive private sector needs an environment in which firms can invest, innovate, defend contracts, compete fairly and plan beyond the next political or regulatory shock.
The government appears to recognise that something has changed. Its fiscal strategy refers to weaker employment momentum, labour-supply and demand mismatches, changes in demand for skills, and the need for education reform, dual education, retraining programmes and support for harder-to-employ groups. It also acknowledges that medium-term labour-market outcomes will depend on productivity growth and the economy’s ability to adjust its workforce structure toward more technologically intensive and productive activities. But recognition is not yet a policy programme. The missing piece is a detailed analysis of causes, consequences and concrete measures.
Active labour-market policy will matter more in the next phase. Retraining, targeted support for marginalised groups, better links between employers and vocational schools, regional training centres and sector-specific skills programmes should become central economic tools rather than secondary employment-office measures. But these programmes must be practical, measurable and tied to real demand. Retraining people for jobs that do not exist, or for sectors that cannot absorb them, only creates statistical activity without structural impact.
The tax and social-protection system also needs adjustment. If low-skilled workers are squeezed between declining demand for simple jobs and insufficient qualifications for better ones, the state must reduce the cost of formal work where possible, strengthen incentives for training and design social assistance that does not trap people outside the labour market. A somewhat more progressive tax approach, combined with targeted support and activation measures, could help cushion the transition without preserving unproductive employment models artificially.
For companies operating in Serbia, the message is equally clear. Labour availability can no longer be treated as an unlimited advantage. Employers will have to invest more in training, retention, automation, process improvement and productivity. Wage increases that are not matched by better organisation and higher value added will erode margins. Companies that can raise productivity will survive the new labour environment; those that depend only on cheap manual work will face growing pressure.
For foreign investors, Serbia remains attractive, but for different reasons than before. Its location, trade links, industrial zones, engineering tradition, regional access and growing service sectors still matter. But investors looking only for cheap labour will increasingly find better options elsewhere. The more durable opportunity lies in projects that use Serbia as a platform for engineering, regional management, specialised manufacturing, energy transition, digital services, logistics, food processing, mining-related services, environmental compliance and industrial technology. That requires a different investment-promotion narrative from the one Serbia used successfully in the previous decade.
The political risk is that employment stagnation may be misread as a temporary weakness rather than a structural warning. Because unemployment may not rise sharply, the pressure to respond could appear limited. Demography can produce a paradoxical situation: employment stagnates or falls slightly, while unemployment remains stable or even declines because the working-age population is shrinking. That does not mean the economy is healthy. It means the labour base is contracting at the same time as the employment model is losing momentum.
Serbia’s next growth phase therefore cannot be built on more of the same. Public infrastructure investment and foreign direct investment can still support GDP, but without stronger productivity they will deliver diminishing returns. The economy needs more value per worker, not only more workers. That means stronger education, better institutions, deeper domestic supplier networks, more demanding investment selection, greater support for innovative private firms and a serious shift away from subsidising low-value employment as a default development tool.
The old model produced results because it matched Serbia’s conditions at the time: high unemployment, low wages and available labour. Those conditions no longer exist in the same form. The country is now entering a harder stage of convergence, where growth depends less on absorbing idle labour and more on improving what each worker, company and institution can produce. That is a more difficult model, but also the only one that can sustain higher wages without pushing investors away.
The stagnation of employment is therefore not just a labour-market statistic. It is a signal that Serbia has reached the end of its low-cost development cycle. The factories leaving the south, the shrinking working-age population, the rise in wages ahead of productivity and the government’s own projections all point in the same direction. Serbia can no longer rely on employment growth to carry the economy. The next decade will be decided by whether the country can replace cheap labour with skilled labour, low-value assembly with productive industry, and subsidy-led attraction with a business environment capable of supporting genuinely higher-value growth.
