State aid control is an accession obligation and a governance test because it determines whether public support corrects market failures or entrenches privilege. This chapter shows that Serbia’s main alignment gaps sit in high-weight schemes in tax legislation, the free zones regime, and state guarantees, and it sets out what alignment would change through clearer classification, cumulation control, grant-equivalent reporting, regional differentiation, transparency, and evaluation. It also clarifies that de minimis is not the ceiling for SME policy under EU rules because GBER enables materially larger rules-based support.
State Aid in the Reform Agenda
State aid control occupies a strategic position in Serbia’s Reform Agenda because it goes to the heart of how public resources shape economic structure. Under the Stabilisation and Association Agreement, Serbia committed to align with Articles 106—108 TFEU on competition. This is also an opening benchmark for Chapter 8. The Growth Plan reinforced these obligations by linking disbursements to demonstrated progress. The Reform Agenda structures alignment as a sequenced process: complete an inventory of existing schemes (due end-2024, delivered late, unpublished); adopt a time-bound action plan for alignment (due June 2025, returned for revision); implement alignment (2025—2027); and achieve full acquis compatibility (December 2027). Serbia is currently behind on the first two steps.
Why does this matter beyond accession formality? EU state aid rules exist to ensure that public support corrects genuine market failures — financing gaps for SMEs, underinvestment in R&D, regional underdevelopment — without distorting competition or creating permanent dependencies. The rules require that aid be targeted to a clear objective, demonstrate an incentive effect (meaning it changes behavior rather than rewarding what would happen anyway), remain proportionate to the identified need, be tracked through cumulation control to prevent over-subsidization, and operate transparently with time limits and evaluation. When these principles are violated, support flows to whoever has connections rather than where market failures are binding, public resources reward investments that would occur anyway, a few large beneficiaries absorb disproportionate shares, and discretionary allocation becomes normalized. This is precisely Serbia’s current pattern.
How Serbia Uses Each Type of State Aid
Table 5 presents the standard categories of state aid recognized under EU rules and how Serbia actually deploys each. The first column identifies aid categories that EU practice recognizes as legitimate, each corresponding to a distinct policy objective and market failure. The second column shows the primary purpose each category is meant to serve. The third column lists the typical instruments through which such aid is delivered. The fourth column assesses Serbia’s actual practice against the reform objective of shifting from discretionary packages toward transparent, rules-based support. What emerges is a pattern of overreliance on regional aid for large FDI, systematic underfunding of horizontal and SME instruments, and formal compliance masking continued discretionary practice in high-risk categories.
Table 5. Types of State Aid and How Serbia Uses Them
| Aid Category | Primary Objective | Common Instruments | Serbia: Reality Check |
|---|---|---|---|
| Regional Aid | Attracting investment to less developed regions | Cash grants per job, land transfers, tax holidays | Overused. Applied at maximum intensities across the entire territory. Used primarily for large FDI at the expense of local value chains. Serbia is treated as a single aid region, so Belgrade projects receive the same terms as lagging areas. |
| Horizontal Aid | Supporting cross-cutting goals (green transition, R&D, innovation) | Direct grants, soft loans, R&D incentives | Policy-declared but underfunded. The Reform Agenda prioritizes shifting toward these schemes. Actual spending remains marginal compared to regional packages. A key area where alignment should drive rebalancing. |
| SME Aid | Strengthening small and medium enterprises | Equipment grants, credit guarantees, advisory support | Systemic crowding out. Formal schemes exist but are marginalized by massive negotiated packages for large investors. Administrative attention concentrates on big projects that promise visible job numbers. |
| Sectoral Aid | Support for specific industries (transport, agriculture) | Subsidies, debt write-offs | High risk. Most vulnerable area in EU negotiations due to ad hoc interventions. Agriculture subsidies lack clear limits; transport support often opaque. Requires careful restructuring. |
| Rescue & Restructuring | Preventing bankruptcy of distressed firms | Recapitalization, debt forgiveness | Formally constrained but bypassed. “One time, last time” rule exists on paper. Equivalent outcomes achieved via tolerated arrears and unpaid energy bills to SOEs, circumventing formal controls. |
| De Minimis | Small-scale aid presumed to cause no market distortion | Various grants, training support (EUR 300,000 cap over 3 years) | Aligned but misused as default. Serbia treats de minimis as the main SME channel — a domestic policy choice, not a legal requirement. GBER allows much larger rules-based SME schemes. |
Source: Authors’ compilation based on EU state aid guidelines, Serbian legislation, and state aid reporting.
Specific Schemes and Their Alignment Gaps
The most consequential alignment gaps are concentrated in high-weight schemes embedded in tax legislation, the free zones regime, and state guarantee frameworks — rather than in the State Aid Control Law itself. Serbia’s State Aid Control Law is assessed as largely aligned with the acquis. The problem lies in what sits outside its effective reach. These instruments can confer selective advantages without transparent grant-equivalent reporting, without credible tests of incentive effect, and without cumulation control across overlapping supports. They sit at the core of Serbia’s industrial policy model because they tilt incentives toward large, discrete projects rather than incremental domestic investment. Table 6 summarizes the main schemes, their alignment status, and the specific issues they raise from an EU state aid perspective.
Table 6. Overview of Existing State Aid Schemes and Alignment Gaps
| Scheme / Legal Basis | Alignment Status | Key Issue from EU State Aid Perspective |
|---|---|---|
| State Aid Control Law (Serbia) | Largely aligned | Requires consistent implementation and further alignment of implementing by-laws and secondary legislation. |
| PPP incentives (Corporate Income Tax Law, Art. 25a(3) and 30a) | Not aligned | Highly selective (high threshold). Unclear policy objective. Weak legal qualification and no effective cumulation control. |
| Aid for employing persons with disabilities (Corporate Income Tax Law, Art. 46) | Partially aligned | Legitimate social objective, but no linkage to detailed horizontal aid rules. Risk of exceeding permissible aid intensities and missing transparency conditions. |
| Large investment and job creation tax relief (Corporate Income Tax Law, Art. 50a) | Not aligned | Unclear legal basis (regional, employment, or other). Undermines cumulation control. Thresholds (EUR 8.5m, 100 jobs) effectively exclude most SMEs. |
| Tax credit for investing in start-ups (Corporate Income Tax Law, Art. 50j) | Not aligned | Does not follow the dedicated EU-style framework for risk finance aid. Risk of prohibited forms or excessive aid amounts. |
| Employment-related tax relief package (multiple provisions in tax laws) | Not aligned | Operates outside horizontal aid safeguards. Lacks transparency and incentive effect tests. No cumulation control across overlapping supports. |
| Customs incentives in free zones (Law on Free Zones, Art. 19 and 29) | Not aligned | Overly broad and undefined objective. Difficult to classify under recognized aid categories. Creates scope for discretionary and uncontrolled support. |
| State guarantees (Public Debt Law, Art. 16 and 18) | Not aligned | Market-conform pricing is not mandatory. Selective advantage for state-owned entities. Significant competition distortion and fiscal risk. Weak integration with state aid control framework. |
Source: Ministry of European Integration 2024, Corporate Income Tax Law 2024, Law on State Aid Control 2019, Law on Free Zones 2006, Public Debt Law 2020.
The core weakness of these schemes lies in their legal architecture rather than their stated economic objectives. Support for investment, innovation, and employment is legitimate. But many measures rely on automatic eligibility rules — such as fixed investment thresholds — without clearly defining the aid objective or the applicable EU compatibility category. This prevents appropriate safeguards from being applied. The corporate profit tax holiday (Art. 50a) is the clearest example: it grants exemption to firms that invest one billion dinars and create 100 jobs, but does not specify whether this is regional aid, employment aid, or something else. Without this classification, there is no way to determine which maximum intensities apply, which cumulation rules govern, or how the incentive effect should be tested. The scheme functions as an “empty frame” — it describes conditions for eligibility but provides no legal basis for compatibility control.
What Alignment Would Concretely Change
Alignment would not eliminate support for large investments — EU rules explicitly permit regional aid for major projects. But it would change how that support operates in specific, consequential ways.
First, the corporate profit tax holiday would need to specify which aid category it falls under — regional, employment, or other — and comply with that category’s rules. If classified as regional aid, it would need to demonstrate that the investment would not occur without support (incentive effect), calculate the tax benefit’s grant-equivalent value in advance, and count toward cumulation limits with other support the beneficiary receives. The current automatic eligibility based solely on investment thresholds (EUR 8.5 million, 100 jobs) would be incompatible because it cannot verify incentive effect — aid is granted whether or not it actually influenced the investment decision. Either the scheme would be redesigned with application-based eligibility that allows ex ante assessment, or it would need to be phased out.
Second, state guarantees to SOEs would need to be priced at market-conform rates or counted as state aid. Currently, guarantees under the Public Debt Law do not require market pricing, conferring selective advantages on state-owned entities that private competitors cannot access. Under alignment, below-market guarantees would need to be notified, assessed for compatibility, and counted toward cumulation limits. This would make visible the implicit subsidies to entities like EPS and Srbijagas that currently flow through the guarantee framework without scrutiny. Making these visible is the first step toward disciplining them.
Third, free zone customs incentives would require clear classification and beneficiary-level tracking. The current scheme under the Law on Free Zones has undefined objectives — it is genuinely impossible to say whether it qualifies as regional aid, export support, investment aid, or something else. This ambiguity creates scope for discretionary, uncontrolled benefits that cannot be properly assessed against any compatibility standard. Alignment would require either defining clear eligible costs with appropriate safeguards under a recognized aid category, or discontinuing the scheme and replacing it with something that can actually be controlled.
Fourth, Serbia would need to establish a functional regional aid map with differentiated intensities. EU rules permit higher aid intensities in less developed regions precisely to steer investment toward areas that need it most. Currently, Serbia foregoes this policy instrument entirely by treating the entire territory uniformly at maximum intensity. A proper regional map would mean that aid for a project in Bor or Pirot could legitimately exceed aid for an equivalent project in Belgrade or Novi Sad — channeling incentives toward territorial cohesion rather than reinforcing concentration. This is not merely a compliance matter; it is a genuine industrial policy tool that Serbia currently does not use.
Fifth, ex post evaluation would become mandatory for major schemes, with results feeding into continuation decisions. EU practice requires time-limited schemes with evaluation clauses. Serbia’s main investment instruments have no sunset dates and no systematic assessment of whether they achieve stated objectives. Alignment would force questions that are currently not asked: did the jobs created persist after the subsidy period ended? Did investments generate productivity spillovers to domestic firms? Was the fiscal cost proportionate to the developmental benefit? Was aid decisive or did it reward investments that would have happened anyway? Currently, no institution is required to answer these questions, and no scheme faces discontinuation for poor results.
Finally, publication of significant awards would enable public scrutiny of who gets what. EU transparency rules require disclosure of awards above EUR 100,000—500,000 depending on the category. This allows competitors to identify potential distortions, civil society to monitor resource allocation, and researchers to assess distributional patterns. Serbia currently operates in opacity — the completed aid inventory has not been published, and there is no systematic disclosure of individual awards above any threshold. Alignment would make the distribution of significant awards — and the balance between large-investor packages and rules-based horizontal/SME instruments — an ongoing matter of public record rather than something that requires forensic reconstruction from scattered budget documents.
De Minimis Is a Choice, Not a Constraint
A persistent misconception in Serbian policy discourse holds that EU alignment would confine SME support to de minimis aid. This fundamentally misreads how the EU framework operates. De minimis (currently capped at EUR 300,000 over three years) is designed for small interventions presumed to cause no market distortion and therefore requiring no notification. It is one narrow channel for administrative simplicity, not the main template for SME policy. The belief that EU alignment means SMEs can only get EUR 300,000 every three years reflects either unfamiliarity with EU practice or convenient justification for the status quo.
The General Block Exemption Regulation (GBER) provides the actual framework for substantial SME support. It explicitly allows — without requiring individual notification to the Commission — investment aid for SMEs at 20% intensity for small enterprises and 10% for medium enterprises, plus regional bonuses that can raise intensities significantly in less developed areas. It permits aid for SME advisory services, trade fair participation, innovation support, environmental investments, and training at meaningful levels. These are not de minimis caps; they are material instruments for firm upgrading. Italy’s Nuova Sabatini programme has supported over 100,000 SME investment projects using interest subsidies and guarantees under GBER. Poland’s operational programmes routinely channel hundreds of millions of euros to SME investment and innovation under block exemption. Germany’s Mittelstandsrichtlinie provides systematic support for SME modernization under the same framework.
Serbia’s reliance on de minimis as the default SME channel is a domestic policy choice, not an SAA requirement. The SAA requires alignment with state aid principles — targeting, incentive effect, proportionality, cumulation control, transparency — not adoption of the most restrictive instruments. Serbia could implement a GBER-style SME investment scheme tomorrow that combines diagnostics, conditional co-financing, and implementation support, providing multiples of current de minimis ceilings while remaining fully compatible with EU requirements. The constraint is not Brussels. The constraint is domestic design choices that reserve substantial instruments for large investors while channeling SMEs through fragmented, low-ceiling programmes whose main virtue is that they require minimal administrative capacity to implement.
Indicators and Monitoring
To assess whether alignment changes outcomes rather than only formal compliance, monitoring should track allocation patterns, concentration, and transparency — not merely the number of schemes deemed compatible:
-
SME share of total state aid volume disaggregated by instrument type (tax expenditure, grants, guarantees) and objective (investment, innovation, employment, regional). This tests whether fiscal rebalancing is actually occurring.
-
Concentration ratio: share absorbed by top 10 and top 50 beneficiaries. High concentration indicates continued reliance on negotiated packages regardless of how many schemes exist on paper.
-
Grant-equivalent reporting coverage: percentage of total aid volume expressed in comparable GGE terms. This tests whether the largest instruments are becoming auditable and comparable.
-
Publication of significant awards above EU thresholds, enabling public tracking of distribution patterns over time.
-
Ex post evaluation completion rate for major schemes, testing whether time limits and assessment requirements are operational or merely nominal.
State aid alignment is a test case for Serbia’s governance reform capacity. The instruments involved shape economic structure more directly than almost any other policy lever. The principles at stake — targeting, incentive effect, proportionality, cumulation control, transparency — are precisely the governance disciplines that earlier chapters found lacking across Serbian institutions: unclear objectives, missing feedback loops, weak coordination across agencies, preference for discretion over rules. If Serbia aligns substantively, it will have demonstrated capacity for systematic, rules-based implementation in a domain that matters. If alignment remains nominal while discretionary practices continue through legislative loopholes that the Commission for State Aid Control cannot close, it will signal that governance reform is not genuinely underway regardless of what laws say on paper. The Reform Agenda, rightly, treats this as a core test.