Putting a Price on Economic Coercion

How the EU can build on its Anti-Coercion Instrument

This Policy Paper is published by GTPO in partnership with the Robert Schuman Foundation, which has selected it for its Schuman Papers series.

With contributions from Dimitri Zurstrassen, Adrian Forest and Rodman Afshar, GTPO

Executive Summary

Europe’s economic relationship with China is entering a more confrontational phase. In October 2026, negotiations with Beijing produced tangible results: China agreed to moderate hybrid-vehicle exports to Europe, improve market access for several European products and provide greater predictability on rare-earth export licences. But these concessions address only part of a much broader imbalance. The EU is simultaneously preparing greater use of safeguards and other trade instruments across additional sectors, while France and Germany have called for a new rapid-response mechanism against severe economic pressure. As Europe becomes more willing to defend its economic interests, it must also be prepared for the possibility that China responds not through negotiation, but through economic coercion.

The EU already has an instrument designed to respond to such pressure. The Anti-Coercion Instrument (ACI) combines a broad range of possible countermeasures with a deliberately gradual process of examination, engagement and, only where necessary, retaliation. This makes it capable of addressing different forms and degrees of serious economic coercion, from targeted pressure on individual Member States or sectors to acute confrontations requiring a rapid and forceful response. Its potential should therefore neither be exaggerated through the familiar “trade bazooka” rhetoric nor underestimated as too cumbersome or limited to matter. With targeted adaptation, the ACI could also provide much of the rapid-response function that France and Germany now appear to be seeking in a new instrument.

What limits the ACI is therefore less the range of measures it can deploy than the way its procedures are organised. The Commission retains broad discretion over whether to open an examination, while a formal determination of economic coercion subsequently requires a qualified majority in the Council – even though neither step imposes economic costs on the third country. This allows political concerns about eventual retaliation to enter the process before retaliation is actually on the table, a particular problem when coercion concentrates costs on only a few Member States or sectors. A newly compiled GTPO dataset of 27 publicly documented episodes of economic coercion against EU Member States since 2016 illustrates this problem. In none of these cases did the Member States directly targeted or particularly exposed constitute the qualified majority required for an ACI determination. Even the broadest case involved only seven Member States, against the 15 needed, while 12 of the EU’s 27 Member States do not appear in the dataset at all.

The paper therefore proposes a targeted reform of the ACI. The instrument should first be normalised as a specialised response to economic coercion. Where the relevant Article 2 factors lend themselves to quantitative assessment, its early stages should move towards a more data-driven and semi-automatic model akin to EU trade defence, with published indicators triggering an obligation to examine potentially coercive conduct rather than leaving initiation entirely to political discretion. The formal determination of coercion should move from the Council to the Commission, while political control should increase as the process approaches the imposition of actual economic costs. Greater European solidarity should help cushion the disproportionate costs borne by targeted Member States or sectors, and the procedure should be streamlined so that it can move considerably faster in genuinely urgent cases.

These lessons extend beyond the ACI. As economic security becomes a more central part of EU policy, Europe will need institutions that can act collectively even when the costs of external pressure are distributed unevenly across Member States. An effective European response to economic coercion therefore requires more than an expanding arsenal of instruments. Early procedural steps should be predictable and capable of moving without repeated political bargaining, while political discretion should increase as decisions approach the imposition of real economic costs. The same logic is relevant to the solidarity mechanism currently being prepared by the Commission. The experience of the ACI suggests that solidarity should be built into Europe’s response to economic pressure from the outset, so that highly concentrated costs do not undermine the capacity for collective action.

Acknowledgements. The author would like to thank Dimitri Zurstrassen for substantial input and feedback throughout the development of this paper, including on the recommendations developed in Section 5, and Rodman Afshar and Adrian Forest for their excellent work in compiling and analysing the data underlying Figures 1-3 and developing the accompanying visualisations.

1. Putting a price on economic coercion, not firing a bazooka

Europe’s economic relationship with China is entering a more confrontational phase. In October 2026, negotiations with Beijing produced tangible results, including an agreement to moderate Chinese hybrid-vehicle exports, improved market access for several European products and greater predictability on rare-earth export licences.¹ At the same time, the EU is preparing greater use of safeguards and other trade instruments, while France and Germany have called for a new rapid-response mechanism against severe economic pressure.² The Commission has also announced a new solidarity mechanism for Member States facing economic coercion or trade retaliation, to be presented alongside the DiversifyEU initiative in December 2026.³ As Europe becomes more willing to defend its economic interests, it must therefore also be prepared for the possibility that China responds not through negotiation, but through economic coercion.

The EU already has an instrument designed to respond to such pressure. Yet even before the Anti-Coercion Instrument (ACI) entered into force, it had acquired a nickname.⁴ During the legislative negotiations, the proposed instrument was already being referred to as the EU’s new “trade bazooka” by 2022.⁵ By September 2023, the metaphor had become sufficiently established for the European Parliament itself to ask, in an explainer published shortly before the instrument’s adoption, what exactly this new “bazooka” was.⁶

The metaphor is not entirely misplaced. The ACI gives the Union access to an unusually broad range of economic countermeasures (formally, Union response measures), ranging from tariffs and restrictions on trade in goods to measures affecting services, investment, public procurement and intellectual property.⁷ In that sense, the instrument provides the EU with considerable economic firepower. And it was perhaps even the emphasis that was put on this firepower back during the ACI’s adoption phase which may help to explain one of the most consequential changes made during the legislative negotiations. Unwilling to vest too much power in the Commission with an instrument capable of producing such far-reaching consequences, the Member States secured a substantially greater role for the Council in determining whether economic coercion exists in the first place.⁸

Yet the bazooka metaphor also creates a misleading picture of how the ACI actually works. A bazooka has a trigger that creates consequences which are hard to revoke once it is pulled.

The ACI does nothing of the sort. Opening an examination under Article 4 does not impose a tariff, exclude a foreign company from public procurement or restrict access to the EU’s internal market in any other way. It starts a regulatory process of legal and factual assessment. Even a subsequent formal determination that economic coercion exists imposes no economic costs and is followed by engagement with the coercing state, assessment of possible responses and repeated opportunities to reach a negotiated solution before Union response measures are eventually imposed as a last resort.⁹ Even then, the Regulation requires the Commission to consider effectiveness, proportionality and the wider Union interest.¹⁰ The instrument is consequently designed around gradual escalation rather than a single act of retaliation.

Recent confrontations also show how the ACI has been understood in practice. During the second Trump administration, tariff threats linked explicitly to the United States’ demands concerning Greenland generated renewed calls for the ACI, but ultimately no formal activation.¹¹ China’s restrictions on strategically important inputs, including rare earths, have likewise prompted debate about the instrument’s potential role.¹² In both contexts, the ACI has appeared in the political debate, but largely as an option of last resort, to be considered only once dialogue and less confrontational means of resolving the dispute have been exhausted.¹³ This makes sense if the EU views the ACI as a “bazooka” – nobody in their right mind wants to blow up the trading relationship with two of the EU’s most important partners. As these examples show, however, this misleading understanding of the ACI sets the threshold for “pulling the trigger” remarkably, perhaps even unachievably, high.

Tellingly, neither episode resulted in a Commission proposal to the Council for a formal determination of coercion, nor was there any public indication that the Commission had opened an examination. This does not mean that either case necessarily warranted eventual response measures. In the Greenland case, the tariff threat was withdrawn only days after its announcement, while in the case of China’s rare-earth export controls their characterisation as economic coercion within the meaning of Article 2 would itself have been contestable, not least because the required nexus to an attempt to influence a particular Union or Member State act would have had to be established. Yet in both cases the EU could still have signalled at an early stage that the conduct was being assessed through the ACI framework.

The debate has since moved on. On 5 October 2026, President Macron and Chancellor Merz proposed a new instrument for a “decisive and systemic reaction” to market distortions, to be triggered by reverse qualified majority. Four days later, the Commission secured an understanding with China under which Chinese exports of hybrid vehicles to the EU are to be limited, reportedly through a quantitative ceiling of 350,000 vehicles in the first year, without triggering immediate Chinese retaliation.¹⁴ Both developments bear directly on the argument of this paper. The first confirms that the requirement of a Council majority is widely perceived as the main obstacle to a credible European response. The second suggests that a credible prospect of collective action can secure concessions before any measure is imposed.

These examples point to a different way of understanding the instrument’s main usefulness. Rather than a bazooka, the ACI should be better understood as a price-setting mechanism for economic coercion by third countries. While I happily admit that the bazooka metaphor is the catchier and more attention-grabbing label, I maintain that my characterisation is analytically more precise. Even a formal determination of economic coercion remains an early step in the process. It signals that the EU has identified economic coercion and starts a formal process through which possible responses can be prepared. Without imposing any costs on the EU’s end, it signals the expected costs of continuing the coercion for the third country. But it does not predetermine how far the Union will ultimately go. At several points, the coercing state can change course, negotiations can succeed and escalation can stop. What matters for deterrence is therefore not simply the destructive potential of the measures, but whether a third country believes that, once coercion occurs, the EU is willing and institutionally capable of setting this process in motion and making continued coercion progressively more costly. Importantly, this does not require either diluting the substantive definition of economic coercion or making escalation under the ACI automatic. It does, however, require the early stages of the process to operate with sufficient predictability that potentially coercive conduct is not left outside the ACI framework simply because of political hesitation.

The broader point is that the ACI combines deterrence with considerable flexibility. Its early stages can signal that continued coercion will carry progressively greater costs without predetermining whether those costs will ultimately be imposed. But the same architecture also leaves the Union with access to a broad range of powerful response measures where signalling and engagement fail. The ACI should therefore not be understood either as an instrument that must remain unused to avoid escalation or as one whose value lies only in gradual deterrence. Properly applied, it can provide both a structured path towards de-escalation and, where circumstances demand it, a route towards forceful economic action.

The paper develops this argument in four steps. First, it explains why the EU’s sui generis structure makes it particularly vulnerable to asymmetric economic coercion in the current geoeconomic environment. Second, it revisits the ACI’s legal architecture and shows that the instrument is more flexible than the ‘trade bazooka’ metaphor suggests, combining gradual escalation with the capacity for substantial economic countermeasures. Third, it examines how the current decision-making structure, both the Commission’s broad discretion over whether to open an examination and the Council’s role in determining economic coercion, can undermine the credibility of that process by introducing political friction too early. Finally, it sets out how the ACI could be normalised and strengthened, both through its operation under the existing Regulation and through targeted legislative changes that need not wait for the formal review due by 2028.¹⁵

2. Understanding the ACI’s role in geoeconomic times

Before turning to the technical details of the ACI, it is worth considering why the current geoeconomic context poses a particular problem for the EU. Economic coercion is, of course, not a new phenomenon. What has changed in recent years is the extent to which economic interdependence itself has become an instrument of geopolitical competition. In Babić’s terms, the relationship between markets and political power has increasingly been reversed: rather than politics primarily serving markets, the networks of trade, finance, technology and production created during decades of economic integration can themselves be mobilised in pursuit of political and security objectives.¹⁶ China’s restrictions on exports of strategically important raw materials illustrate the basic mechanism. The EU remains highly dependent on Chinese supplies for a number of critical inputs, including rare-earth magnets, while China has progressively introduced export controls affecting graphite, gallium, germanium, rare earths and associated technologies.¹⁷ Monitoring by the Commission’s DG GROW Chief Economist team confirms the scale of this exposure. Of the 204 products in sensitive industrial ecosystems for which the EU is identified as foreign-dependent, China is the main supplier of 64 and accounts for more than half of their combined import value. On DG GROW’s External Vulnerability Index (EXVI), the EU’s score is highest precisely in raw materials (0.28 on a 0–1 scale, where higher values indicate greater vulnerability, compared with 0.22 for all EU industrial products and with China’s own score of 0.24 in raw materials).¹⁸ Such dependencies create leverage because economic disruption can impose costs that are difficult to absorb or replace in the short term.

The EU is, however, not simply another economically interdependent actor. Its sui generis structure creates an additional vulnerability. The internal market and the common commercial policy provide the Union collectively with considerable market power, but the economic structures, external dependencies and geopolitical preferences of its 27 Member States remain highly heterogeneous. The same external shock can therefore generate very different levels of exposure across the Union. A DG GROW proof-of-concept stress test of a disruption to Chinese rare-earth extraction and processing, for example, found markedly different levels of resilience among upstream firms located in different Member States, within the supply chains of four leading EU wind-turbine manufacturers: the average resilience index was close to 0.5 for Ireland, above 0.9 for Czechia, Hungary, Italy, Poland and Sweden, and between 0.8 and 0.9 for Germany, Denmark and Spain. ¹⁹ A third country seeking to exercise pressure on the EU can exploit precisely these differences by targeting individual Member States or economically and politically salient sectors within them, while leaving others comparatively unaffected.

The recent Greenland episode provides an unusually clear illustration of how such differentiated exposure can translate into political fragmentation. The additional tariffs threatened by the Trump administration in January 2026 were not directed against the EU as a whole. They specifically targeted Denmark, Finland, France, Germany, the Netherlands and Sweden, alongside Norway and the United Kingdom, in response to their conduct concerning Greenland. ²⁰ Twenty-one EU Member States were therefore left outside the immediate scope of the threatened measures. The countries bearing the prospective economic costs thus constituted only a subset of those whose political support would ultimately be required for a common EU response. Moreover, attitudes towards the appropriate reaction differed: France strongly supported considering activation of the ACI, while several other Member States placed greater emphasis on dialogue and avoiding escalation with the United States.²¹ The episode illustrates the broader structural problem: economic pressure can be distributed asymmetrically across the Union, while recognising and responding to that pressure still requires collective political decision-making.

The asymmetry is potentially even greater once differences in economic and geopolitical exposure are taken into account. Member States do not trade with, receive investment from, or depend on the United States for security to the same extent or in the same way. A government assessing whether to support an EU response consequently considers not only solidarity with the directly targeted Member States, but also the possible costs of retaliation for its own economy and the wider transatlantic relationship. The significance of the Greenland episode therefore lies less in proving that the United States deliberately designed its tariff threat to divide the EU than in showing how easily differentiated economic pressure can produce such an effect. A coercing state can concentrate immediate costs on those governments whose conduct it wishes to change while the political and economic risks associated with an EU response are distributed very differently across the Union.

China’s response to the EU investigation into subsidised electric vehicles demonstrates a second and more granular version of the same problem. Pressure can be concentrated on politically salient sectors within particular Member States. As the EU’s anti-subsidy investigation into Chinese electric vehicles progressed towards the imposition of countervailing duties, China initiated trade defence investigations affecting European brandy, pork and dairy products.²² The potential costs were distributed highly unevenly. The brandy investigation was overwhelmingly relevant to France, while exposure to the pork investigation was concentrated particularly in Spain, followed by the Netherlands and Denmark.²³ The dairy investigation again affected a different configuration of national interests, where the Member States mainly concerned were France (€179 million), Italy (€57 million), the Netherlands (€53 million), Denmark (€52 million), Spain (€46 million), Ireland (€45 million) and Germany (€33 million).²⁴ At the same time, Germany’s automotive industry was highly exposed to the Chinese market and strongly concerned about the possibility of wider retaliation.²⁵ The Commission itself subsequently characterised China’s investigations as part of a broader pattern of retaliatory trade action surrounding the EV dispute.²⁶

This points towards an important political economy of economic coercion. The most effective target does not necessarily have to be the sector whose disruption maximises aggregate economic damage. Pressure may instead be especially effective where it falls on an industry that is geographically concentrated, politically organised, highly visible and dependent on access to the coercing state’s market, while the costs to the coercing country of restricting that trade remain manageable. French cognac or Spanish pork potentially fit this logic better than measures aimed indiscriminately at the European economy. In 2024, for example, Spain was the EU’s largest pork exporter to China and had supported the proposed EV tariffs in the Member States’ initial advisory vote; during a subsequent visit to China, however, Prime Minister Pedro Sánchez called for the EU to reconsider its position and emphasised the need to avoid a trade war.²⁷ This does not establish that Chinese pressure caused Spain’s position to change, but the relevant point is that concentrating economic costs on politically influential domestic constituencies can plausibly change the incentives facing national governments and thereby affect the coalition from which a common EU position must emerge for a potential use of the ACI.

These examples reveal a vulnerability that goes beyond the EU simply being economically dependent on major powers. Economic coercion can differentiate directly between Member States and target politically influential domestic constituencies whose exposure varies across the Union. The Member States bearing the immediate costs of coercion, those most concerned about the costs of responding to it, and those whose support is required for a collective EU response need not be the same. In this sense, the economic integration that gives the EU considerable collective market power coexists with a political fragmentation that third countries can plausibly exploit by targeting it asymmetrically.

The pattern is not confined to these prominent examples. Figure 1 maps 27 publicly documented episodes of targeted economic pressure affecting identifiable EU Member States since 2016. The dataset is intended to capture the distribution of such pressure across the Union rather than to make a legal determination that each episode constitutes economic coercion within the meaning of Article 2 ACI. It nevertheless provides an indication of how unevenly the immediate costs of geoeconomic pressure have been distributed across the Union.

Figure 1 – Publicly documented episodes of targeted economic pressure on EU Member States, 2016–2026

‡ Nominally EU-wide probe; Member States most exposed according to the Commission (SWD(2025) 231).

† Borderline case (contractors mostly non-EU; wartime measure; border restrictions).

* Estimate from Eurostat Comext (dataset DS-045409): goods exports of the targeted Member State(s) to the coercing country in the year before the episode; all goods where no products were named, cars (HS 8703) for No. 3, wine (HS 2204) for No. 25.

Scope and sources. The dataset covers publicly documented threats or measures affecting trade or other economic interests, directed at identifiable EU Member States and linked to a specific policy choice of the Union or a Member State. It was compiled and analysed by Adrian Forest and Rodman Afshar for this paper from publicly available official, institutional, company and press sources, supplemented by Eurostat Comext data for selected exposure estimates. The dataset is not intended as an exhaustive catalogue of economic coercion, and inclusion does not constitute a legal determination of economic coercion under Article 2 ACI.

Values and source. Unmarked values as reported by the source (government, Commission, company and press figures). Dataset: GTPO, compiled and analysed by Adrian Forest and Rodman Afshar (2026).

The distribution is strikingly uneven. Only 15 of the EU’s 27 Member States appear at least once in the dataset, while 12 do not appear at all. France, Germany, the Netherlands and Spain each appear in six episodes. Importantly, this concentration persists even after widening the dataset beyond China to include pressure originating from the United States, Belarus, Turkey, Russia, Iran and Algeria. The asymmetry is therefore not simply a product of one coercing country’s targeting strategy.

This uneven exposure matters institutionally. The Member States bearing the immediate costs of economic pressure are not necessarily those whose support is required for a collective response. Figure 3 illustrates this problem by comparing, for each episode, the Member States directly targeted or particularly exposed with the two components of a qualified majority in the Council. Even comparatively broad episodes generally leave the immediately affected coalition well short of the qualified majority required for an Article 5 determination

The hybrid vehicle negotiations of October 2026 point in the same direction from the opposite angle. This time, France, Germany and Spain, the Member States most exposed to Chinese retaliation in the earlier dispute over electric vehicles, aligned their positions, and Beijing accepted a negotiated limit on hybrid exports instead of opening new trade defence investigations against politically salient European products.²⁸ Coordination among the most exposed Member States appears to have neutralised the strategy of asymmetric targeting described above. The episode should not be overstated: the concession also served Chinese objectives, notably curbing destructive price competition among Chinese carmakers, and the Chinese readout mentions neither volumes nor a timetable. It nonetheless illustrates the mechanism on which this paper relies: a credible and collective prospect of costs can change a third country’s calculation before any measure is adopted.

The ACI’s graduated structure responds to this problem by separating the recognition of coercion and the signalling of prospective costs from the later political question of whether those costs must actually be imposed. At the same time, its broad menu of response measures gives the Union the capacity to escalate substantially where engagement fails. In a geoeconomic environment in which such pressure is becoming more common, however, the EU needs to be capable of identifying and processing serious economic coercion without requiring agreement from the outset on whether the Union is ultimately prepared to enter into a trade confrontation. The next sections show that the legal architecture of the ACI is already much closer to this logic but also where its regulatory design allows the political fragmentation described above to interfere with that process unnecessarily early.

3. Revisiting the ACI’s functionality

What matters, then, is not simply what the ACI allows the Union to do at the end of the process, but how credibly it can move between its different stages. The ACI is not a one-off retaliatory weapon, but a staged process with robust legal guardrails. It begins with examination and signalling, moves through a formal determination of coercion and structured engagement and, where necessary, proceeds towards Union response measures.²⁹ This architecture allows the EU both to make the prospective costs of continued coercion increasingly visible and to escalate towards substantial economic measures where those signals fail.

At the outset lies the substantive definition of economic coercion. Under Article 2 of the Regulation, economic coercion exists where a third country applies or threatens to apply a measure affecting trade or investment in order to prevent or obtain the cessation, modification or adoption of a particular act by the Union or a Member State, thereby interfering in their legitimate sovereign choices. The Commission and the Council must take into account a number of contextual factors, including the intensity and impact of the measure, whether it forms part of a broader pattern of interference, the extent to which it encroaches upon Union or Member State sovereignty, whether the third country is acting on the basis of a legitimate internationally recognised concern, and whether it previously made serious good-faith efforts to resolve the matter through coordination or adjudication.³⁰

The definition is therefore broad in the sense that it is capable of capturing very different forms of economic coercion, rather than being confined to a particular type of measure, but it does not establish a low substantive threshold. Interference must reach a sufficient qualitative or quantitative degree of seriousness so that only economic coercion with a sufficiently serious impact or a sufficiently credible threat falls within the Regulation. This ensures that the ACI is directed at serious instances of coercion without confining its application to only the most exceptional geopolitical confrontations. The standard is therefore broad and flexible in the range of coercive conduct it can capture, while retaining a meaningful threshold of seriousness before such conduct qualifies as interference with the legitimate sovereign choices of the Union or its Member States. The Regulation does not, however, say how that seriousness is to be measured. The first of the Article 2(2) factors – intensity, severity, frequency, duration, breadth and magnitude of the measure and its impact on trade or investment – is in essence quantitative, yet no indicator or benchmark is attached to it.

Conducting this assessment may be described as the ACI’s examination and signalling phase. Under Article 4, the Commission may, either on its own initiative or upon a duly substantiated request, examine a third-country measure in order to determine whether the conditions of Article 2(1) are met. The Commission must act “expeditiously”, while the examination should “normally not exceed four months”. ³¹ Importantly, this is an indicative timeframe rather than a fixed deadline. Where the relevant conduct and evidence are clear, the Commission can complete its examination substantially faster, but it may also take longer where the circumstances warrant it. This flexibility allows the Union to manage politically sensitive cases, although prolonged examinations risk weakening the instrument’s signalling effect. Crucially, initiating an examination merely sets in motion a process of legal and factual assessment, rather than committing the Union to an irreversible escalation.

More fundamentally, Article 4 does not oblige the Commission to open an examination at all, even upon a duly substantiated request by a Member State. The contrast with the Union’s trade defence instruments is instructive. Under the Basic Anti-Dumping Regulation, the Commission must assess whether a complaint lodged by or on behalf of the Union industry contains sufficient evidence and, where that threshold is met, initiate proceedings within 45 days.³² Under the ACI, by contrast, even a duly substantiated request does not trigger an equivalent obligation to proceed. The Commission therefore retains broad discretion over whether the price-setting process begins at all. The first institutional gatekeeper of the mechanism is consequently not the Council but the Commission itself.

From a regulatory perspective, such flexible timelines cut both ways. On the one hand, they make the ACI escalation-averse by giving the Commission time to gather evidence, assess the economic effects of the third-country conduct and leave room for political engagement. On the other hand, too much flexibility may weaken what Ayres and Braithwaite describe as the regulator’s ability to project an “image of invincibility”: the expectation that, once a regulatory process has begun, failure to change the relevant conduct will lead to further escalation, or – in the words of this contribution – carry a price tag.³³ Open-ended timing can weaken this expectation by signalling to third countries that the process may slow down or stall as political circumstances change. As a consequence, the ACI manages a tension between preserving flexibility and projecting resolve. If anything, however, the examination phase appears designed to avoid rather than produce escalation, as initiating it carries no consequences, while the procedural flexibility may be detrimental for signalling resolve.

As further parts of the examination and signalling phase, the Commission may gather information from any reliable source, including Member States, the European Parliament, economic operators and trade unions, and may seek information specifically on the impact of the third-country measure.³⁴ Where appropriate, it may also publish a notice inviting stakeholders to submit information and if it does so, the third country concerned must be notified that the examination has begun.³⁵ These public elements of the investigation can themselves contribute to deterrence. Once an examination is opened, firms, affected industries and even subnational governments in the coercing country have an incentive to assess where the EU’s eventual response might fall and what costs they could bear. If they conclude that they are likely to be on the receiving end of Union response measures, they may in turn press their own government to cease or at least de-escalate the coercive conduct.³⁶ The examination phase can therefore begin to alter the domestic political economy of the coercing country before the EU has imposed any countermeasure at all. This also mirrors, in reverse, the collective-action problem identified for the EU in the previous section. While third countries may be less institutionally fragmented than the EU, they are rarely unitary economic actors, and divergent interests among firms, sectors and regions can likewise create domestic pressure to de-escalate.³⁷

Only if the Commission concludes, following its examination, that the conditions of Article 2(1) are met does the process move to the next stage. Under Article 5, the Commission must submit a proposal to the Council for an implementing act formally determining that economic coercion exists. In doing so, it must explain how the Article 2 conditions are fulfilled and set out an indicative period, normally not exceeding six months, for assessing whether the conditions for Union response measures under Article 8(1) are met.³⁸ Even at this stage, however, the Regulation creates further opportunities for de-escalation. The Commission may invite the third country to submit observations before making its proposal,³⁹ and may propose that the Council request reparation for the injury caused to the Union.⁴⁰ In functional terms, this allows the EU to communicate a clear message before any response measures are contemplated: The Commission considers the conduct coercive and a formal determination is approaching.

It is only after the Commission has reached this conclusion that the Council enters the regulatory process directly. Acting by qualified majority, it must decide whether to adopt the proposed determination, normally within eight weeks and, absent a justified delay, within no more than ten weeks.⁴¹ These periods are again outer procedural benchmarks rather than waiting periods – where a qualified majority exists, nothing in the Regulation prevents the Council from acting considerably sooner. Such a qualified majority requires the support of at least 55 per cent of the Member States, currently at least 15 out of 27, representing at least 65 per cent of the EU population.⁴² This is the most politically sensitive point in the procedure and, as argued below, its most visible institutional bottleneck. Importantly, however, even a positive Council determination imposes no economic cost on the third country. It does not activate tariffs, restrict market access or otherwise prejudge the eventual Union response. It merely establishes that economic coercion exists and allows the process to proceed to the subsequent stages of engagement, assessment and, only if those efforts fail, possible countermeasures. The political threshold attached to the Article 5 determination therefore sits somewhat uneasily with the relatively limited consequences that the determination itself produces.

Once these hurdles have been cleared and a determination of economic coercion has been made, the ACI enters what I term the structured engagement phase. Under Article 5(9) and (10), the Commission must inform the third country and request that it cease the economic coercion immediately and, where appropriate, repair the injury caused to the Union. Article 6 then requires the Commission to provide an adequate opportunity for consultations with the third country with a view to obtaining the cessation of coercion and, where relevant, reparation. The Regulation expressly contemplates direct negotiations, international adjudication and other forms of third-party assistance, such as mediation or conciliation.⁴³ In parallel, the Commission must seek to raise the matter in relevant international fora, such as the WTO, after consulting the Council where appropriate,⁴⁴ and may cooperate with third countries affected by the same or similar coercion in order to coordinate responses.⁴⁵

Importantly, the Regulation imposes no hard deadline on this structured engagement phase. Although the Commission must, already in its Article 5 proposal, indicate a period (normally not exceeding six months) for assessing whether the conditions for Union response measures are met, Article 6 itself merely requires an “adequate opportunity” for consultations, while Article 8 refers to the failure of such efforts within a “reasonable period”. The Commission therefore retains considerable discretion over the pace of escalation even after economic coercion has formally been determined. This flexibility again cuts both ways, as it weakens the image of regulatory invincibility, but at the same time reinforces the instrument’s fundamentally escalation-averse design. As a result, this part of the instrument’s architecture makes clear that a formal determination of economic coercion is still far from equivalent to starting a trade war. Rather, the ACI continues to escalate gradually, now signalling more directly and through several channels that retaliatory measures are moving closer.

As noted above, Union response measures only become available where the earlier efforts under Articles 5 and 6 have not resulted, within a reasonable period, in the cessation of the coercion and, where requested, reparation of the injury to the Union has not been provided. Article 8(1) then adds two further cumulative conditions: the measures must be necessary to protect the interests and rights of the Union and its Member States in the particular case, taking into account the options available, and their adoption must be in the Union interest as determined under Article 9. The latter requires the Commission to balance the preservation of the Union’s and Member States’ sovereign choices against the other interests at stake, including those of Member States, economic operators, upstream and downstream industries, and consumers. This provides an additional safeguard against disproportionate escalation because relatively minor instances of economic coercion would not justify highly disruptive countermeasures simply because the Article 2 threshold has been met.⁴⁶ At the same time, the deliberately broad Union-interest assessment leaves room for wider strategic considerations.⁴⁷ Where, for example, retaliation against the United States would jeopardise security interests considered more important in the circumstances, the Regulation does not require the EU to impose response measures merely because economic coercion has previously been determined.

Only once these conditions are satisfied does the ACI enter its third and most clearly escalatory phase, what I term its response phase, in which concrete Union response measures are selected and adopted. Under Article 8, the Commission selects the appropriate measures from those available under Annex I of the Regulation and must explain both why the conditions for taking action are fulfilled and why the particular response is appropriate in light of the criteria set out in Article 11. The measures are adopted by Commission implementing act under the examination procedure provided for by Regulation (EU) 182/2011, which governs Member State control over the Commission’s exercise of implementing powers.

In practice, this means that a committee composed of representatives of the Member States delivers an opinion on the Commission’s draft by qualified majority.⁴⁸ If a qualified majority supports the draft, the Commission must adopt it; if a qualified majority opposes it, the Commission may not do so.⁴⁹ The ACI also opts for the more restrictive form of the examination procedure available under Regulation 182/2011. Where no qualified majority forms either for or against the draft and the committee therefore delivers no opinion, the Commission may not adopt the draft at that stage.⁵⁰ It may, however, submit an amended version to the committee or refer the matter to the appeal committee in accordance with Regulation 182/2011.⁵¹ Member State involvement therefore remains substantial even at this later stage.⁵² The difference is that, unlike the Article 5 determination, this is the point at which the political caution built into the decision-making structure corresponds to a genuinely escalatory decision, as the Union is now deciding whether, and through which measures, to impose concrete economic costs on the coercing third country.

The Regulation, however, places significant legal guardrails around this price-setting exercise. Response measures must be proportionate and may not exceed the level of injury to the Union, taking into account the gravity of the coercion, its economic impact and the rights of the Union and its Member States.⁵³ This is further evidence that the ACI is not designed to “blow up” the economic relationship with a coercing country, but to impose a calibrated price for the coercive conduct. In selecting measures, the Commission must prioritise their effectiveness in inducing the cessation of coercion while minimising negative effects on Union actors, the investment environment, innovation, consumers and other EU policies.⁵⁴ Put differently, the Regulation builds in the fairly intuitive requirement that retaliation should not inflict greater costs on the EU than on the third country it is intended to influence. The ACI is therefore not designed to maximise pain, but its objective is to make the costs of continued coercion sufficiently credible that a rationally calculating third country concludes that the coercive conduct is no longer worth pursuing, while ensuring that the additional costs imposed on the Union by its own response remain proportionate to the injury it seeks to address.

In addition, the ACI’s process remains reversible even at this late stage. Under Article 8(6), Union response measures normally have a deferred date of application of up to three months, creating another window in which the third country can de-escalate before any economic costs are actually imposed. This also allows the pressure generated by the announced measures to work through the domestic political economy of the coercing country, giving potentially affected firms, sectors and other actors an incentive to push for a change of course. If the Commission receives credible information that the coercion has ceased, or that concrete steps towards cessation have been taken, application may be deferred further. If the coercion ends before the measures take effect, the implementing act must be repealed.⁵⁵ Even after measures enter into force, Article 12 allows for their suspension, amendment or termination as circumstances change.

The ACI therefore preserves off-ramps throughout the process, making it difficult to characterise even the adoption of response measures as an irreversible act of escalation. This de-escalatory design is not, however, absolute. In duly justified cases, Article 8(10) permits response measures to apply without a prior demand to cease coercion or advance warning of their application, where necessary to protect the rights and interests of the Union or a Member State, particularly the effectiveness of the measures. Moreover, where coercion takes the form of a threat, Article 8(11) provides that Union response measures apply from the date on which the threatened third-country measure is implemented.

Taken together, these different stages reveal an instrument that is considerably more cautious, gradual and reversible than the “trade bazooka” metaphor suggests. The ACI does not move directly from a finding of coercion to retaliation. It first investigates and signals, then formally determines, subsequently engages and negotiates, and only as a last resort moves towards the imposition of carefully calibrated economic costs. Even at that point, the Regulation subjects retaliation to necessity, the Union interest, Member State involvement, proportionality and repeated opportunities for de-escalation. In regulatory terms, the ACI therefore already contains most of the ingredients of a credible price-setting mechanism, as it allows the EU to make the prospective costs of continued coercion increasingly visible before those costs are imposed. Its real-world impact consequently depends less on the severity of the measures available at the end of the process than on whether the Union can navigate the preceding stages with sufficient credibility to convince third countries that continued coercion will indeed carry a price.⁵⁶

4. Making the price for coercion more credible

The ACI’s potential therefore depends not only on the economic costs that the Union is ultimately capable of imposing, but also on its ability to move credibly through the process that precedes them. Regulatory theory offers a useful way of understanding this dynamic. Responsive regulation conceives enforcement as a graduated and adaptable process. Regulators should ordinarily begin with less coercive forms of intervention, such as dialogue and persuasion, escalate when these fail to produce a change in conduct, and de-escalate again when cooperation is forthcoming.⁵⁷ Consequently, more coercive measures can be used only rarely, but their availability, and the regulator’s demonstrated willingness to move towards them when necessary, is what gives the preceding stages much of their force. Countering coercion with coercion is costly and therefore generally secondary to dialogue, but the threat of sanctions must occasionally be followed through if it is to remain credible.⁵⁸

Put in the terms of the price-setting mechanism developed here, the price attached to economic coercion must therefore not only be sufficiently high; the coercing country must also believe that continued coercion creates a credible prospect that it will eventually have to pay it.⁵⁹ A potentially severe response at the end of the process has little deterrent value if the institutional hurdles preceding it make reaching that point appear unlikely. The relevant expectation is not that every examination will inevitably end in retaliation, but that continued coercion will move the process forward unless the third country changes its conduct.⁶⁰ A third country should be able to avoid paying the price by ceasing the coercion; if it chooses to maintain the coercive conduct, it should face an appropriate price for doing so. What it should not be able to do is rely on political difficulties within the EU to prevent the price-setting process from advancing in the first place.

Seen from this perspective, two points in the procedure become particularly important. The first, and least visible, is the Commission’s discretion over whether to open an examination under Article 4 at all. The second is the decision-making procedure under Article 5. The formal determination that economic coercion exists requires an affirmative qualified majority in the Council even though, as explained above, this determination neither imposes response measures nor commits the Union to adopting them later.⁶¹ Yet it is precisely at this comparatively early stage that the ACI introduces one of its most demanding political hurdles. This front-loads geopolitical caution to a stage that is primarily concerned with what should be a technocratic assessment of third country conduct. The possibility of eventually imposing substantial economic costs may therefore shape Member State positions before the Union has reached the point at which those costs actually need to be considered.⁶² The two filters reinforce each other. A Commission that anticipates a difficult vote in the Council has every reason not to open an examination in the first place, so that the expected politics of Article 5 are already priced into the non-decision under Article 4.

There is an important distinction here between two questions that the institutional design of the ACI risks collapsing. The first is whether a third-country measure constitutes economic coercion within the meaning of Article 2. This is principally a legal and evidentiary assessment based on the criteria laid down in the Regulation. The second is whether, after having made that determination and exhausted opportunities for engagement, it is appropriate for the EU to retaliate in the circumstances of the particular case. The latter is unavoidably political. It may require difficult judgments about economic exposure, relations with the coercing country, wider foreign and security policy interests, and the costs that Member States are prepared to bear. But the Regulation already provides considerable space for precisely these judgments at later stages. The necessity and Union-interest requirements under Articles 8 and 9, the criteria governing the design of measures under Article 11, Member State involvement through comitology, and the repeated opportunities for suspension or termination all ensure that recognising coercion does not make retaliation “automatic” in any sense.⁶³

Neither the Commission’s discretion at the examination stage nor the Council’s role under Article 5 needs to carry the full burden of geopolitical caution. To the contrary, doing so undermines an important advantage of graduated enforcement, where low-cost steps at the beginning of the process can themselves alter the calculations of the regulated actor. The purpose of an examination or a formal determination is not yet to impose the price of coercion, but to make that price progressively visible and credible. Each step communicates that the EU has recognised the conduct, that the regulatory process is continuing and that, unless the coercion ceases, more consequential action may follow. For such signalling to work, however, these early steps must themselves be sufficiently easy and predictable to take. If an Article 5 determination already requires Member States to resolve the political questions associated with an eventual trade confrontation, the Union is effectively forced to negotiate the consequences of retaliation before retaliation is actually on the table.⁶⁴ The result is paradoxical, as concerns about costs that may arise at the end of the enforcement process can prevent the EU from taking precisely those lower-cost steps that might make incurring those costs unnecessary. A third country observing such hesitation can rationally discount the price attached to continued coercion, regardless of how formidable the measures listed in Annex I may appear.⁶⁵

The more effective sequence would be almost the reverse. The EU should be able to move through the early signalling stages without unnecessary institutional friction, allowing the prospect of progressively increasing costs to work on the calculations of the coercing country and on the domestic actors in the coercing country that stand to bear those costs. Ideally, that signal will be sufficient and the process stops before the politically difficult question of exactly how much economic pain the Member States are willing to bear ever needs to be resolved. Only where signalling and engagement fail should the Union have to undertake that more difficult political bargaining. At that point, the ACI already contains extensive safeguards, including necessity, the Union-interest test, proportionality, Member State involvement in the adoption of response measures and repeated opportunities for de-escalation.⁶⁶ Together, these safeguards are sufficient to prevent an economically or geopolitically harmful escalation. The institutional problem is therefore not insufficient political control over retaliation, but the premature exercise of that control. Making the institutional process for determining coercion simpler to navigate would not lower the substantive threshold for coercion or make retaliation automatic. Instead, it would strengthen the expectation that continued coercion will move the process forward, while leaving the necessity and design of any eventual response to the later stages of the ACI.

A technocratic determination, however, presupposes technocratic inputs. If the Article 2 assessment is to be treated as a legal and evidentiary finding, the evidence must be generated in a way that Member States, economic operators and third countries can anticipate. Much of it already exists. DG GROW’s Supply Chain Analytics Hub combines a structural mapping of the EU’s strategic dependencies, a composite External Vulnerability Index (EXVI) and a high-frequency Supply Chain Alert Notification (SCAN) system that tracks import prices and quantities at product level on the basis of customs surveillance data; the same methodology informs DG TRADE’s Import Surveillance Task Force, which monitors trade-diversion risks.⁶⁷ The connection with economic security policy is already established at the level of strategy. The 2023 European Economic Security Strategy identified the weaponisation of economic dependencies and economic coercion as one of four categories of risk, and SCAN has since been used in the resulting supply-chain risk assessments. The Joint Communication of December 2025 on strengthening EU economic security goes further: it lists the ACI among the tools meant to deter coercion, sets out indicators of high-risk dependencies and announces an Economic Security Information Hub that will build on existing monitoring tools such as SCAN, including a mechanism for rapid information on trade flows.⁶⁸ What is still missing is a formal link between this monitoring capacity and the opening of an examination under Article 4. None of these tools was designed for the ACI. Together, however, they could measure the first and most quantitative of the Article 2(2) factors – the intensity and impact of a third-country measure – far more systematically than an ad hoc examination does today. The next section sets out recommendations for a potential review of the Regulation.

5. Recommendations for a more credible ACI

The analysis above suggests that the ACI does not principally suffer from a lack of economic power.⁶⁹ Likewise, its legal architecture does not require a fundamental redesign. On the contrary, the Regulation already provides a carefully graduated process through which the Union can identify coercion, signal that continued conduct will carry costs, pursue negotiated solutions and, only where necessary, impose proportionate response measures. The challenge is to make this architecture operate more credibly and predictably when sufficiently serious economic coercion occurs. Normalisation in this sense does not mean routinely reaching for the ACI in ordinary trade disputes. Article 2, read together with recital 15, already imposes a meaningful substantive threshold, where only coercion reaching a sufficient qualitative or quantitative seriousness falls within the Regulation. Rather, where conduct plausibly raises economic coercion of that significance, opening an examination and, where the legal conditions are subsequently met, formally determining that coercion exists should not itself be treated as an extraordinary act of escalation.

The recommendations that follow operate at several levels. First, the EU should normalise the ACI conceptually as a specialised response to serious economic coercion rather than an exceptional “trade bazooka”. Second, the opening of an examination should become more rules-based and predictable. Third, the formal determination of coercion should be returned to the Commission, while political control increases as the process moves closer to actual countermeasures. Fourth, mechanisms for European solidarity should cushion the asymmetric costs that coercion and the Union’s response may impose across Member States. Finally, the ACI should make fuller use of the procedural flexibility it already contains and, through targeted amendment, acquire a clearer accelerated route for exceptionally urgent cases.

Recommendation 1: Normalise the ACI as an instrument for serious economic coercion

The EU should finally leave the “trade bazooka” behind. The metaphor may have helped communicate the significance of the instrument during the legislative process, but it now risks becoming an obstacle to its effective use. Anyone firing a metaphorical trade bazooka must be prepared for destructive consequences. As this paper has shown, this bears little resemblance to the ACI’s actual operation. Opening an examination does not impose retaliation. Determining that economic coercion exists does not impose retaliation. Even after such a determination, the Regulation requires further engagement and provides considerable space for the coercing country to change course before response measures are adopted or applied. The ACI is better understood as a specialised trade instrument that puts a price on economic coercion and whose effectiveness depends on the predictable operation of these different stages.

Changing the language surrounding the instrument is more than a matter of presentation. How the ACI is understood shapes the political threshold for using it. If opening the procedure is routinely described in drastic terms, both Member States and affected economic actors have reason to evaluate even its initial stages by reference to the potential consequences of a full-scale trade confrontation. The same signal is sent externally, where initiating an examination can be interpreted by the third country as an extraordinary act of escalation rather than the beginning of a structured regulatory process. A graduated-enforcement conception offers a different expectation. An examination or determination communicates that the EU has identified potentially coercive conduct, and that continued coercion will progressively increase the likelihood of economic consequences. It does not imply that those consequences are inevitable. Normalising this understanding would make it easier for both the EU and third countries to treat the early stages of the ACI as part of ordinary economic diplomacy rather than as the opening move of a trade war.

Recommendation 2: Make the opening of an examination rules-based and predictable

A targeted amendment should narrow the Commission’s discretion at the very beginning of the procedure. Transferring the determination to the Commission while leaving it complete discretion over whether to look at a measure at all would simply move the political bottleneck from the Council to the Berlaymont. A more robust design borrows the logic of the trade defence instruments. A duly substantiated request from an affected Member State or from a representative group of Union operators, or an alert generated by the Commission’s own monitoring, would oblige the Commission to conclude within a fixed period – for instance 45 days, as in anti-dumping – whether the conditions of Article 2(1) appear to be met and, if so, to open an examination; a decision not to proceed would have to be reasoned and published.⁷⁰ This would not make escalation automatic. The Council’s role at the determination stage, if retained, and Member State control over response measures would remain untouched. What it would make predictable is the start of the price-setting process, which is precisely what a third country needs to believe for the price to deter.

The triggers for such an obligation need not be invented. The Commission already applies explicit, published thresholds to identify strategic dependencies and supply-chain distress, and the December 2025 Joint Communication on economic security has added a set of non-cumulative indicators of high-risk dependencies, including the control of 60 per cent or more of EU supply by a single third country or operators and the fact that the third country has already weaponised economic dependencies or threatened to do so.⁷¹ Table 1 below sets out how these tools could be mapped onto the Article 2(2) factors, for instance in an implementing act or guidelines adopted under a revised Article 4.

Table 1 – Possible indicator-based triggers for opening an ACI examination.

Factors (c) to (e) of Article 2(2) – encroachment on sovereignty, legitimate international concern and prior good-faith attempts at settlement – remain qualitative legal assessments and are deliberately left outside the indicator set.⁷²

Four design principles follow. First, as in the December 2025 Joint Communication, the indicators should be non-cumulative but hierarchical: a robust signal, such as a SCAN alert on a product where the coercing country controls most of EU supply, should suffice on its own, whereas weaker signals, such as recurrence, should only count in combination. Second, the indicators should trigger an obligation to examine, not a determination: they speak to intensity and impact, while the question of whether a measure seeks to interfere with a legitimate sovereign choice remains a separate judgment. Third, where the Commission already uses published thresholds, the ACI should draw on them; where new thresholds are required, they should be calibrated transparently and reviewed periodically. Fourth, the triggers should be plugged into the economic-security risk-assessment cycle launched in 2023 rather than create a parallel one: the same data-driven analysis, complemented by the qualitative, intelligence-based assessment the Commission already carries out, would serve both to map exposure ex ante and to justify opening an examination when a measure materialises.

The limits of this approach should be acknowledged. The tools described above were built to detect supply shocks, not coercive intent, and they are strongest for Chinese-style export restrictions. They are much weaker for threats that have not yet materialised in trade data, such as the Greenland tariff threat, and for measures targeting EU exports rather than imports; for these, the trigger has to rest on a formally announced measure combined with export-side exposure data, which the Commission does not yet publish systematically. Poorly calibrated thresholds could also lead to over-triggering and dilute the signal, while a third country could try to calibrate its measures just below them. Neither risk is decisive. The Commission would retain its power to act on its own initiative below the thresholds, and a de minimis clause can prevent minor disputes from being drawn into the procedure. What the indicators change is the default, as once they are reached, not acting would require a public justification.

Indicators disaggregated by Member State would also make visible, before any crisis, how the costs of both coercion and a Union response are likely to fall across the Union, the very asymmetry that third countries exploit. They would thus provide the factual basis for negotiating in advance how those costs could be shared, rather than discovering the distribution of burdens in the middle of a confrontation. For European firms, the benefit is predictability: a published set of triggers tells operators which exposures are monitored and when a request by their Member State or sector association would oblige the Commission to act. Evidence from the DG GROW-EIB survey conducted in 2024 (published in 2025) suggests that the demand is there, as firms relying on essential inputs from China were markedly more likely than others to report difficulties in accessing raw materials (20% against 12%) and semiconductors (19% against 11%).⁷³

Recommendation 3: Return the formal determination of coercion to the Commission

An amendment of the ACI should also reconsider the institutional design of the determination stage. The most direct reform would be to move closer to the Commission’s original legislative proposal. Under that proposal, following its examination the Commission itself would have adopted a decision determining whether the conditions for economic coercion were fulfilled and could, before doing so, invite the third country to submit observations.⁷⁴ The subsequent involvement of the Council that ultimately emerged from the legislative negotiations considerably increased the political threshold for taking this first formal determination under what is now Article 5 of the Regulation. Recent policy proposals have similarly identified the ACI’s institutional complexity as a potential obstacle to credible deterrence. Tobias Gehrke, for example, has called for an “ACI 2.0” combining a fast-track mechanism with a broader conception of coercion.⁷⁵ The proposal developed here differs in emphasis. It would leave the substantive scope and threshold of Article 2 unchanged and instead address the sequencing of political control within the existing regulatory process. As discussed below, greater procedural speed may also be warranted in particularly urgent cases, but this need not necessarily entail broadening the conduct that qualifies as economic coercion.

The analysis in this paper suggests that this political safeguard is located at the wrong point in the procedure. Determining whether conduct satisfies the criteria in Article 2 is predominantly a legal and evidentiary exercise. It certainly has political significance, but it neither selects response measures nor commits the EU to any action down the line. The Commission is institutionally better placed to conduct this assessment consistently across cases, whereas requiring an affirmative qualified majority in the Council invites Member States to bring considerations about the possible future costs of retaliation into a decision concerning whether coercion exists in the first place. Restoring the Commission’s responsibility for the determination would therefore not eliminate political scrutiny from the ACI. Importantly, it would merely change who applies the Article 2 test, but not the content or the substantive threshold of the test itself. As a result, it would separate more clearly the assessment of coercion from the subsequent political question of what, if anything, the EU should ultimately do about it.

Importantly, shifting the Article 5 determination towards the Commission would not require reducing Member State influence where the stakes become genuinely higher. If anything, the logic developed in this paper points in the opposite direction. Political involvement should increase as the process approaches substantive escalation. The present procedure already reflects this to a considerable extent. Before response measures can be adopted, the Commission must establish necessity and the Union interest, select and calibrate measures in accordance with Article 11 and subject the proposed measures to Member State control through the examination procedure. These are precisely the stages at which questions about the distribution of economic costs, relations with the third country and, potentially, wider foreign policy interests become unavoidable.

If Member States nevertheless considered stronger political control necessary as part of any changes following an amendment to the regulation, there would be a more coherent place to locate it. This would be at the decision to adopt Union response measures rather than the prior determination that coercion exists. A Council implementing act at that stage could be contemplated as a politically feasible compromise. This paper does not advocate such a change, as the existing Commission implementing act, adopted through comitology with Member State involvement, already provides political control. But if the choice were between direct Council involvement at the determination stage and direct Council involvement when economic costs are actually about to be imposed, the latter would better reflect the logic of the instrument. The closer the EU moves towards actual retaliation, the stronger the case for political judgment becomes.

Recommendation 4: Share the costs of resisting coercion

The greater visibility over the distribution of costs created by the indicator framework should be matched by a mechanism for sharing them. The problem is not new. The Commission’s original impact assessment for the ACI considered a financial compensation mechanism under which economic operators could receive support for losses caused either by the coercive measures themselves or by Union countermeasures.⁷⁶ The option was ultimately not retained, principally because of its budgetary and administrative implications. The analysis developed here provides a reason to revisit it. Where coercion works precisely by concentrating costs on particular Member States, regions or sectors, some capacity to redistribute those costs would make it harder for a third country to translate asymmetric economic pressure into opposition to a collective EU response. The solidarity mechanism announced by the Commission in October 2026 for Member States facing economic coercion or trade retaliation offers a natural vehicle for this, provided it includes a financial dimension.

Such a mechanism need not compensate firms euro for euro. Its logic could instead resemble that of the European Globalisation Adjustment Fund: targeted Union solidarity with workers, regions or sectors bearing unusually concentrated adjustment costs.⁷⁷ The indicator framework proposed above could help identify where those burdens are likely to fall and provide objective criteria for accessing support. The ongoing negotiations over the 2028–2034 MFF offer an opportunity to consider such a facility alongside a review of the ACI. Where Union response measures themselves generate revenue, in particular through additional customs duties, one option would be to reflect part of that revenue politically in the resources available for such solidarity.⁷⁸ Direct earmarking would require careful consideration under the EU budgetary framework, but the basic principle is straightforward: part of the economic value generated by the Union’s collective response could be used to cushion those actors bearing a disproportionate share of its costs.

A useful illustration is the EU’s response to China’s economic pressure on Lithuania following the opening of a Taiwanese Representative Office in Vilnius in 2021. As Lithuanian exports and firms connected to Lithuanian supply chains encountered restrictions in the Chinese market, the Commission approved a €130 million Lithuanian loan scheme specifically for affected companies, allowing firms to finance adaptation, diversification and access to new markets.⁷⁹ The episode demonstrated that targeted economic support can complement the Union’s diplomatic and legal response to coercion. Yet the financing remained national: EU solidarity principally took the form of permitting Lithuania to use its own fiscal resources under the State aid rules. A permanent Union-level mechanism would go one step further by allowing the costs of resisting coercion to be shared across the Union rather than falling primarily on the budget of the Member State that happens to be targeted.

Recommendation 5: Streamline the ACI’s timelines

A potential review of the Regulation should also streamline the timing of the ACI procedure. The Regulation already contains important temporal benchmarks: examinations should normally not exceed four months, the Council is generally expected to act within eight weeks and normally no later than ten weeks, and the Commission’s Article 5 proposal must identify an indicative period – ordinarily no longer than six months – for assessing whether the conditions for Union response measures are met.⁸⁰ Yet significant flexibility remains, particularly through concepts such as providing an “adequate opportunity” for consultations and allowing a “reasonable period” for coercion to cease. Some flexibility is necessary, especially where genuine negotiations are taking place. But it should not allow delay itself to become a strategy for defeating the instrument. The default expectation should be that the procedure continues to advance, subject at every stage to the substantive conditions laid down in the Regulation, unless the third country takes credible steps towards cessation or meaningful engagement provides a reason to slow the process. Greater discipline around extensions, including clearer reasons for significant delays, would reinforce the expectation that continued coercion moves the process forward.

The case for greater procedural discipline has become still more salient in light of the emerging debate over a rapid-response “red button” trade instrument.⁸¹ The Franco-German non-paper of 5 October 2026 calls for a mechanism capable of acting rapidly in response to severe and systematic market distortions, including through a decision-making procedure designed to reduce the risk of political blockage.⁸² The proposed instrument is broader than the ACI and appears intended to address conduct that may not constitute economic coercion within the meaning of Article 2. Yet where an acute confrontation does fall within the ACI’s substantive scope, the same concern about speed arises. Rather than necessarily creating an overlapping instrument for such cases, a review could consider an accelerated procedure within the ACI itself, for example through substantially shortened deadlines or, in narrowly defined circumstances of particularly severe and imminent harm, provision for provisional measures pending completion of the ordinary procedure. Such an option would require carefully defined thresholds and safeguards. The point would not be to lower the substantive threshold for economic coercion, but to ensure that the ACI remains capable of acting where delay would deprive the instrument of its practical effect.

Conclusion

The objective of these reforms would not be to turn the ACI into a more aggressive instrument, but to make it more credible across the full range of situations for which it was designed. Early procedural steps should be comparatively easy, predictable and technocratic, and grounded in indicators the Commission already produces, because their principal function is to signal that coercion has been recognised and that continued conduct will carry an increasingly credible price. The procedure should also be capable of moving more quickly where urgency requires it, while a solidarity mechanism would reduce the ability of third countries to exploit the uneven distribution of the costs of either coercion or the Union’s response. Difficult political bargaining over the precise costs that the Union is prepared to impose should occur later and, if the signalling function of the ACI works as intended, may never be necessary at all. A credible ACI should therefore be capable both of inducing de-escalation before countermeasures become necessary and of moving decisively towards proportionate retaliation when they do.

References