Serbia’s 2025 real-economy profile is best described as stable but uneven: the system did not break, but it did not accelerate either. Official statistics show industrial production rising 1.0% and manufacturing 1.2% in 2025, while gross fixed capital formation grew only 0.9% in real terms and agricultural output edged down 0.3%. These are “low single-digit” numbers that typically characterize economies in a transition phase between an investment-driven expansion and a maturity period where productivity and skills become the binding constraints.
That mix already tells you where the internal tension sits. Manufacturing is still expanding, suggesting existing industrial platforms remain competitive and continue to find demand. Yet investment growth is barely positive, which implies that capacity expansion is happening slowly and perhaps selectively. In such a setting, the role of retail and household demand becomes more important, because consumption can temporarily mask investment weakness. But relying on consumption is a risky substitution if it is not backed by productivity growth, because it can regenerate inflation pressure or widen the trade deficit through import-heavy spending.
The NBS’s own interpretation of growth drivers adds granularity. In its monetary policy reporting, it highlighted that 2025 growth was driven by services and manufacturing, with visible effects from expanding serial production of electric vehicles at Stellantis in Kragujevac and accelerated tyre production, while construction and energy made negative contributions in parts of the year. This is an economically meaningful decomposition: a modern manufacturing node tied into European automotive chains is carrying weight, while domestically anchored project sectors are weaker. It implies that Serbia’s tradable sector competitiveness is not the problem; the problem is the translation of that competitiveness into a broader capex cycle that lifts productivity and supports consistent wage growth without inflation.
Construction softness is particularly important because it has multiplier effects. When construction slows, it affects materials, logistics, subcontractor employment, and municipal finance. It also delays the productivity payoffs from infrastructure improvements. If investment confidence was damaged and some investments were deferred in 2025—as the NBS inflation reporting suggests—then construction is likely the most immediate transmission channel for that deferral.
The productivity dimension is where Serbia’s next stage will be won or lost. A year in which industrial output grows by 1.0% while investment rises only 0.9% is not a year of large productivity jumps; it is a year of incremental gains. Over time, incrementalism becomes a constraint when the labor market tightens or when wages rise faster than productivity. Serbia already faces structural pressures common to the region—migration, demographic decline, and skills mismatch—which do not appear suddenly in one data release but show up as persistent friction in hiring, training, and the ability to scale complex production.
That friction matters most in the sectors Serbia wants to expand: higher-value manufacturing, engineering services, ICT and embedded software, and energy infrastructure. These sectors require not just labor, but specific competencies. When skills are scarce, firms respond in predictable ways: they either slow expansion, relocate marginal growth elsewhere, automate more aggressively, or accept lower margins. In a country where FDI has historically been a key driver, skills constraints can also change investor selection: investors with training capacity and long time horizons remain, while those looking for quick, labor-arbitrage wins may move on.
The investment climate literature for Serbia recognizes this transition problem. Analyses of Serbia’s investment and business climate have pointed out that the contribution of fixed investment declined in 2025, primarily due to a drop in FDI inflows, but anticipate stronger growth in 2026–2027 supported by easing global inflation pressures and the implementation of transport, energy, and communal infrastructure projects. That is essentially a statement that Serbia is between cycles: it needs the next infrastructure and energy buildout to restart capex and unlock productivity, but it is not yet assured that financing conditions and project execution will align quickly enough.
Retail endurance, in this context, is not automatically “good news”. It can indicate resilience of household incomes and consumer confidence, which supports political and social stability. But if retail strength is partly import-driven and not matched by export growth, it can worsen the external position. This is why Serbia’s real-economy story cannot be separated from the trade and current-account story: a stable consumer sector is an asset only if the tradable sector can keep earning and if investment keeps improving productivity so wage growth is sustainable.
The macro overlay reinforces the message of a modest-growth environment. The IMF estimated Serbia’s real GDP growth at around 2% in 2025, with a recovery to about 3% in 2026, while noting that the current account deficit may widen in 2026 due to fuel import costs and export restrictions affecting steel. For the real economy, this signals that the near-term growth “ceiling” is not purely domestic—it is also shaped by external costs and external market access. A modest growth ceiling makes productivity and skills even more central: you cannot grow fast by simply adding more inputs; you need output per worker to rise.
The country therefore enters 2026 with a clear internal agenda even if politics and geopolitics remain noisy: keep manufacturing competitive, restart construction and infrastructure execution, and treat skills formation as an economic-growth instrument rather than as a social policy. If those elements align, Serbia can raise its medium-term growth path without needing inflationary stimulus. If they do not, the economy risks a prolonged phase of “stable but slow” where headline stability hides a gradual erosion of convergence speed.
