Executive Summary

On 15 September 2026 the European Parliament approved the EU's Autonomous Trade Measures (ATM) for Armenia by 583 votes to 49, with 39 abstentions (Public Radio of Armenia), and on 24 September the Council of the EU formally adopted the regulation, which enters into force the day after its publication in the EU's Official Journal (Council of the EU). For two years from entry into force, duties are suspended on a wide range of Armenian goods, with duty-free quotas for eight agricultural products. The measures answered a Russian campaign of import bans that began in late May 2026 (The Moscow Times), and the Council framed them plainly as a response to economic coercion (Council of the EU).

A companion analysis published by our centre on 24 August 2026 argued that these measures should be a bridge toward durable integration rather than a one-off crisis response. That premise is now widely shared and is not re-argued here. This brief answers the question the earlier analysis opened: what concrete instrument should replace the ATM before it expires, and how should it be built?

Its central finding is that Armenia cannot follow the route taken by Ukraine and Moldova, both of which upgraded lapsing emergency measures into durable frameworks because each already held a Deep and Comprehensive Free Trade Area (DCFTA) to fall back on and revise (European Commission). Armenia has none, and its membership of the Russia-led Eurasian Economic Union (EAEU), a customs union, rules one out. When its ATM expires in 2028, Armenia reverts to ordinary tariffs — a cliff, not a floor. The brief therefore proposes an EU–Armenia Trade Transition Facility: a longer-lasting arrangement in which the EU lowers tariffs on Armenian goods without asking Armenia to cut its own, and widens that access step by step as agreed milestones are met. Its design draws on the EU's long-running Autonomous Trade Preferences for the Western Balkans. Since 2000, those preferences have operated alongside the bilateral Stabilisation and Association Agreements, providing additional EU market access in selected agricultural sectors (European Commission, DG Trade). Adapted to Armenia, that instrument keeps EU integration moving within the EAEU constraint — and lets the EU show that partners who reform, and who are targeted for it, are rewarded rather than abandoned.

Point of departure

This brief accepts, rather than restates, the premises now common to Armenian and European commentary: that the ATM is a bridge, not a destination; that removing tariffs will not by itself turn Armenian producers into successful EU exporters, because the binding constraints are standards, certification, logistics, rules of origin and firm-level capacity; and that EAEU membership genuinely limits how far integration can go. The scale of the relief confirms the first two points. The EU expects to forgo only about €3 million in customs revenue a year (Council file) and estimates Armenian firms will save roughly €5.4 million over two years (JAMnews) — modest against the roughly €420 million of annual exports the Russian bans disrupted, per Financial Times reporting (Regional Post). The instrument's value lies in market opening and signal, not in the duties.

The design problem: why simply extending the ATM will not work

Three features make the current measure a poor foundation for durability.

First, it is bounded by design. It is an ordinary, time-limited EU trade regulation (adopted under Article 207(2) TFEU; EUR-Lex, COM(2026) 348) that runs for two years. Its benefits depend on Armenia proving that goods genuinely originate there, working with EU customs to prevent fraud, not introducing new barriers to EU imports, and respecting the core principles of its Partnership Agreement with the EU; the EU can also limit imports if they harm its own producers (Council of the EU). Coverage is incomplete — ores, pharmaceuticals, paper and works of art are excluded, and several major categories only partly liberalised (Public Radio of Armenia). It is scaffolding, not a building.

Second, unconditional emergency preferences are politically fragile — the real lesson of Ukraine. Ukraine's ATMs, which lifted all remaining volume limits and minimum-price rules on its farm exports, became contentious after three years; surging agri-food imports triggered farmer protests and unilateral bans in Poland, Hungary and Slovakia, and the emergency limits the Commission later added for seven sensitive products were judged insufficient (European Policy Centre). The measures lapsed in June 2025 and were replaced, from 29 October 2025, by a revised DCFTA that is in places more restrictive than the emergency regime (ORF). Durability had to be bought with conditionality and safeguards, not open-ended generosity.

Third — and this is what makes the cliff concrete rather than rhetorical — Armenia has already fallen off a version of it. Until 2022, Armenia enjoyed zero EU tariffs on a wide range of products under GSP+, the enhanced version of the EU's Generalised Scheme of Preferences (GSP) for developing countries. The Commission removed Armenia from both schemes in January 2022, after the World Bank classified Armenia as an upper-middle-income country for three years in a row (EUR-Lex, SWD(2023) 367). Since then, Armenian goods have faced the EU's standard tariffs (the "most-favoured-nation" rates applied to any partner without a special deal), because Armenia's Partnership Agreement with the EU contains no tariff cuts for goods (EU–Armenia Partnership Council, via PRISM). Independent modelling at the time projected that losing GSP+ would raise the average EU tariff on Armenian exports from roughly 0.8% to about 5.1% (CIVILNET), and Armenian officials have since cited the GSP+ exit, alongside logistics, as a barrier to selling into the EU (ARMENPRESS). The precise effect is hard to isolate, because it coincided with a war-driven surge in Armenia's re-exports to Russia — but the direction is not in doubt. 

The 2028 expiry would repeat that reversal on a larger base, and there is no safety net beneath it: the renewed GSP that applies from 2027 is reserved for lower-income countries, so Armenia does not qualify (European Commission, DG Trade). Nor can Armenia sign a DCFTA: a free-trade agreement would require Armenia to lower its own tariffs on EU goods, but as an EAEU member Armenia does not set its own tariffs; they are fixed jointly for the whole union (Armenia Ministry of Finance). Armenia fits none of the EU's existing templates; a new instrument must be built.

Design principles, and one honest trade-off

The revised Ukrainian and Moldovan DCFTAs supply the template for conditional market access: more duty-free access granted as the partner gradually adopts EU production standards, checked through annual reports, with an emergency brake if imports harm EU producers (DG Trade, Ukraine; AGRINFO, Moldova). The EU's own turn to gradual integration — invited by the European Council in June 2022 and elaborated in the CEPS and European Policy Centre staged-accession work — supplies the second: benefits granted in clear stages, earned through reform and withdrawable if reform stalls (European Policy Centre). From the Ukrainian episode comes a third: access that threatens EU producers is withdrawn, so visible progress on EU standards, backed by safeguards, is what makes deeper access politically survivable.

These yield five principles: access tied to reform (conditionality), the ability to scale back (reversibility), clear rules known in advance (predictability), no obligation on Armenia to cut its own tariffs (non-reciprocity), and reassurance for EU producers. Two of them, non-reciprocity and predictability, pull against each other. Non-reciprocity is what keeps the instrument EAEU-compatible, but a unilateral preference is inherently less credible than a treaty, because the EU can narrow it. The brief resolves this openly: the Facility is a second-best necessitated by the EAEU constraint, and it substitutes for treaty permanence in three ways — a fixed multi-year term rather than annual renewal, a milestone schedule the EU publishes and commits to in advance, and the precedent of the Western Balkans regime, which delivered stable unilateral access for over two decades.

The EU–Armenia Trade Transition Facility

Legal form. The Facility should take the form the EU already uses to give durable, one-way trade preferences to partners on a European path: a multi-year EU trade regulation modelled on the Western Balkans Autonomous Trade Preferences, granted since 2000 and renewed every five years. In practice, this means replacing the two-year measure with an initial five-year regulation that carries the same kind of checks. The Western Balkans preferences are not open-ended giveaways. They depend on beneficiaries proving where goods originate, cooperating with EU customs to prevent fraud, not raising new barriers to EU imports, and respecting human rights, democracy and the rule of law. The Commission can suspend them, fully or partly, for fraud, a breakdown in that cooperation, or a sudden surge of imports, and member states oversee how the rules are applied through a standing committee (Regulation 1215/2009, consolidated). Armenia's Facility should copy these checks. Because Armenia has no Association Agreement with the EU in which such review rules could sit, as Moldova does, it must write them into the regulation itself.

The WTO dimension must also be settled deliberately. WTO rules normally require the EU to offer all trading partners the same tariffs; special treatment for one country is an exception that needs a legal basis. A short emergency measure carries little legal risk, but a permanent arrangement cannot rely on that indefinitely. The EU's established route for durable preferences of this kind is a formal WTO waiver, meaning permission from the other WTO members, which the Western Balkans regime has long relied on (EUR-Lex). Securing that basis is a concrete legal task to address from the outset.

The Western Balkans precedent also settles the durability question directly. Although every Western Balkans partner has since signed a contractual free-trade agreement within its Stabilisation and Association Agreement, the autonomous preferences were not absorbed and retired: they still run alongside those FTAs — providing additional access for fruit, vegetables and wine — and were renewed again at the end of 2025 for a further five years, nearly twenty-five years after they began (European Commission, DG Trade). The one difference to state plainly is the instrument's role: for the Western Balkans it supplements an FTA the countries were able to sign, whereas for Armenia — for which that destination is blocked while it remains in the EAEU — the Facility would be the primary channel of preferential access rather than a top-up. The durability and the check architecture transfer intact; only the surrounding FTA is missing, which is exactly the gap the Facility exists to fill.

How conditionality works without a treaty. Because the EU grants the preferences on its own, "conditionality" is not a contractual obligation but the EU's power to widen or narrow access. To make that power predictable for Armenian exporters, the regulation should publish in advance the milestones that unlock each additional step, so progress follows known rules rather than case-by-case decisions. The Facility works through three linked tracks: Track 3 changes Armenia's rules; Track 2 builds firms' capacity to use them; Track 1 is the reward both unlock.

Track 1 — Market access (the reward). A published list of tariff preferences that grows as milestones are met. It borrows the approach of the revised Ukrainian and Moldovan agreements, larger duty-free quotas in return for meeting EU standards, but without asking Armenia to cut its own tariffs, and with emergency brakes built in from the start. For example, a first step could remove all duties on products that are currently only partly covered, once Armenia runs an EU-recognised system for certifying plant health and tracing products back to the farm; a later step could enlarge agricultural quotas as Armenia adopts specific EU food-safety rules.

Track 2 — Export capacity (meeting the rules in practice). Existing EU funding — €270 million under the Resilience and Growth Plan for 2024–2027 (EUR-Lex), plus deployed experts and funding intended to help establish an export-promotion agency (Brussels Times) — is restructured from general assistance into targeted support for each sector's specific gaps: certification, cold-chain logistics, and export finance, so money follows the specific bottleneck rather than being spread thin.

Track 3 — Targeted regulatory alignment (changing the rules on paper). Rather than adopting the entire body of EU law (the acquis), the Facility sequences alignment on the few standards that decide whether the covered products can enter the EU — food safety, phytosanitary rules, product testing and certification — prioritised by export value. These are the milestones that release Track 1, mirroring the Ukrainian and Moldovan conditionality but scoped to what a non-candidate can achieve.

Governance. The tracks interlock: Track 2 and 3 milestones release Track 1 access, with progress assessed in a published annual scorecard, managed by the existing EU–Armenia committee that already oversees trade under the Partnership Agreement; reversibility is preserved through the safeguard clauses.

Handling the EAEU boundary

The Facility is engineered to function inside EAEU membership, not to force a choice. Its one-way preferences leave the shared external tariff untouched, avoiding the reciprocity that makes a DCFTA impossible. This is the honest ceiling: the Facility can carry Armenia far on access and standards, but it cannot become a DCFTA or resolve the strategic tension — which is real and measurable. Armenia still conducts roughly 35% of its foreign trade with Russia and buys most of its gas there (Reuters), against about 11.3% of trade with the EU (DG Trade). At Astana in May 2026 the EAEU's leaders said they would consider suspending Armenia and pressed for a referendum (Reuters), and Prime Minister Pashinyan has acknowledged that dual membership is impossible (Anews). The Facility neither triggers nor forecloses that decision; it keeps durable integration moving while the larger question stays open — its central advantage over waiting for a blocked DCFTA.

Recommendations

1. Commission (its trade department, DG TRADE, with the department for enlargement and Eastern neighbourhood, DG ENEST) — launch the successor process early enough to avoid a gap in tariff-free access after 2028. The Commission should begin preparing a successor to the temporary Armenian regime well before its expiry in 2028, with a draft legislative proposal targeted within twelve months, to preserve continuity of market access while giving Armenian exporters enough certainty to invest in EU-oriented production, certification and distribution. Ukraine illustrates the cost of letting a temporary regime lapse before its successor is operational: when its Autonomous Trade Measures ended in June 2025, trade fell back to the older, tighter quota limits of Ukraine's existing free-trade agreement while new terms were negotiated (European Policy Centre). The Western Balkans autonomous-preferences regime — unilateral, non-reciprocal and renewed in five-year tranches for nearly 25 years (DG Trade) — is a useful precedent for predictability and renewal, but should be treated as a model rather than a ready-made legal template. The Commission should settle the successor's WTO basis in parallel with drafting — most likely a waiver, as the Western Balkans regime uses (EUR-Lex) — rather than leaving it to the end.

2. Commission (DG TRADE, DG ENEST and the health and food-safety department, DG SANTE) — convert conditionality into a transparent, product-specific market-access roadmap. The successor should identify three to five priority product groups and publish, for each, the conditions for any additional preference, distinguishing tariff conditions, which the EU sets in the trade instrument, from food-safety and plant- and animal-health conditions (SPS rules), which depend on meeting EU import rules and, for some products, on the EU approving Armenia as an exporting country and approving individual producers. Each roadmap should specify the responsible Armenian institution, the EU requirement, the evidence needed, the verification procedure and the review date. This makes a unilateral preference more predictable and "bankable" without creating the false expectation that meeting an Armenian certification requirement automatically confers EU access — EU rules already make entry for many plant products and products of animal origin conditional on checks by the exporting country's state inspectors and, for animal products, on each producer being on the EU's approved list (European Commission, Food Safety).

3. Both sides — sequence the funding to the milestone gates. The Resilience and Growth Plan and export-promotion support should be tied to the readiness gates rather than spread as general assistance, so money flows to the specific bottleneck that keeps each sector from meeting an EU requirement — and disburses as those milestones are met. This closes the loop between the three tracks: the capacity funding (Track 2) becomes the instrument that helps Armenia clear the regulatory gates (Track 3) that in turn unlock market access (Track 1).

4. Armenia (Food Safety Inspection Body, Ministry of Economy and relevant sectoral authorities) — turn existing certification capacity into reliably EU-compliant export capacity. Armenia should not build a system from scratch. It already has competent food-safety, veterinary and plant-health authorities, pesticide-residue monitoring and phytosanitary certification, and it secured EU approval to export aquaculture products in 2025 (AGRINFO) — evidence the system can meet EU standards when resourced. The priority is to close the implementation gaps that keep that capacity from supporting consistent access for the covered plant sectors: records of pesticide use and tracing of products from farm to export, internationally accredited testing laboratories, regular checks for pesticide residues, a register of producers and processing facilities, electronic export certificates, more inspectors, and fast action when problems are found. Its own inspectorate reports that it cannot monitor on-farm pesticide use and cites weak traceability as a major obstacle, and its 2024 testing found residues above permitted limits in several fruits and vegetables (CIVILNET). The benchmark should be not whether a certificate can be issued, but whether Armenia can demonstrate consistent compliance across an entire export chain — since EU entry can require robust official controls, traceability, certification and, for certain animal-origin products, EU approval of the country and of individual producers (European Commission, Food Safety).

5. Both sides — measure diversification through a five-indicator trade-access dashboard. The EU and Armenia should set a baseline and publish annual progress against five indicators: (i) the value of Armenian agri-food exports to the EU; (ii) the EU share of Armenia's agri-food exports; (iii) the number of Armenian products and establishments exporting to the EU; (iv) the share of eligible exports that actually use the tariff preferences; and (v) how far Armenian agri-food exports depend on a small number of markets. Ukraine is a comparator, not a target: the EU share of its agri-food exports rose from 28% in 2021 to 52% in 2024, but amid wartime disruption and rerouting (European Policy Centre). The Armenian objective should be defined as sustained growth and diversification of commercially viable exports, not replication of the Ukrainian percentage.

6. Commission — embed a predictable safeguard mechanism in the successor regime from the outset. The successor should include an emergency safeguard with clear rules on when it can be used, what evidence is needed, who must be consulted, how long it lasts and when it is reviewed. Ukraine shows why timing matters: safeguards were added only after political pressure had intensified and were then judged insufficient by neighbouring Member States, whose objections helped end the preferences (European Policy Centre). For Armenia, the aim should be a credible backstop in place before pressure emerges, reducing the incentive for Member States to seek unilateral national bans and giving EU producers confidence that import surges can be handled through a common mechanism.

Conclusion

The Autonomous Trade Measures bought Armenia two years of access to a market of 450 million consumers (JAMnews), and the prior analysis set the direction. The risk now is arriving at 2028 with a direction but no instrument — and, unlike Ukraine and Moldova, with no free-trade floor beneath the cliff Armenia already fell off once in 2022. The EU has built the answer before, for the Western Balkans: a durable, non-reciprocal, milestone-based preferential regime. Adapted to Armenia's EAEU constraints and engineered against the pressures that curtailed similar measures elsewhere, such a Trade Transition Facility would convert a temporary act of solidarity into a lasting relationship — and let the EU demonstrate that economic coercion of a reforming partner is met with commitment, not a concession that quietly lapses.