The instrumentalization of migration flows as state-sponsored pressure on the Belarus–Poland border: a model of coercion below the threshold of armed conflict

INTRODUCTION

A state that wishes to pressure a neighbour has, traditionally, reached for armies, sanctions, or subversion. In the summer and autumn of 2021 Belarus reached for people. Acting in evident retaliation for European Union sanctions imposed after the disputed 2020 election and the forced diversion of a Ryanair flight, the Belarusian authorities facilitated the arrival of thousands of migrants, mostly from the Middle East and Central Asia, and steered them toward the borders of Poland, Lithuania, and Latvia, presenting the European Union with a humanitarian emergency it had not chosen and could not ignore (Graban, 2024; Mészáros & Țoca, 2023). The instrument was novel in its bluntness but not without precedent: months earlier Morocco had relaxed its frontier controls and allowed thousands to swim into the Spanish enclave of Ceuta amid a diplomatic dispute, and in early 2020 Turkey had declared its border with Greece open and bused migrants to the Evros crossing to extract concessions over Syria (Álvarez-Miranda & Brey, 2025; Gabrielli & Garcés Mascareñas, 2025; İşleyen & Karadağ, 2023). A recognisable technique was taking shape across the European periphery.

The technique is coercive but it is not war, and that is precisely its appeal. It inflicts real costs, political, humanitarian, financial, and reputational, while remaining below the threshold that would justify a forceful response, and it does so using an instrument, displaced human beings, whose movement the coercer can plausibly disclaim and whose suffering the target cannot ignore without abandoning the very norms that define it (Ganty et al., 2024; Grześkowiak, 2023). The scholarship on weaponised or coercive engineered migration, inaugurated by Greenhill's foundational study of forced displacement as a tool of statecraft (Greenhill, 2010) and developed by a now-substantial body of work on migration diplomacy and migration power (Tsourapas, 2019; Fernández-Molina & Tsourapas, 2024; Tolay, 2022), has established that the phenomenon is real and recurrent. What it has been slower to provide is a model that specifies the mechanism: how, exactly, the movement of a comparatively small number of people generates strategic leverage over a far more powerful target.

Most analyses of the Belarus–Poland episode have proceeded descriptively. A securitisation literature has shown convincingly how the crisis was framed as a hybrid threat and an emergency that justified extraordinary measures, including pushbacks and a state of exception at the border (Graban, 2024; Straczuk, 2024; Liszkowska, 2024). A legal literature has dissected the European and Polish response, the suspension of asylum access, and the strained invocation of an instrumentalisation concept to excuse it (Ganty et al., 2024; Grześkowiak, 2023; Abbassi & Triculescu, 2025). Both are valuable, and this article draws on both, but neither offers a generative model of the coercion itself, one that explains why the instrument works, predicts when it will work better or worse, and identifies where a target can intervene. The central research question is therefore mechanistic: through what mechanism does the instrumentalisation of migration generate coercive leverage below the threshold of armed conflict, and what determines the magnitude of that leverage?

Three hypotheses structure the inquiry. The first holds that the coercive leverage of instrumentalised migration derives primarily from the target's normative entrapment, the bind created by its own legal and normative commitments, rather than from the physical magnitude of the flow, so that relatively small engineered flows can generate disproportionate pressure. The second holds that the instrument is deliberately calibrated to remain below the armed-conflict threshold, and that its non-military and deniable character is constitutive of its effectiveness rather than incidental to it. The third holds that effective counter-coercion works by lowering the coercive leverage, chiefly by reducing normative entrapment through lawful, legitimacy-preserving responses and by raising resilience, rather than by hardening the border alone.

The original contribution of this article is the model that these hypotheses presuppose. It introduces, for the first time to my knowledge in an integrated form, the Sub-Threshold Coercion (STC) model of weaponised migration, organised around the novel concept of the normative-entrapment dilemma and operationalised through a composite Coercive Leverage Index in which leverage is the product of cost asymmetry, deniability, and normative entrapment divided by the target's resilience. The model joins the coercion-theory, the migration-diplomacy, and the securitisation literatures that have addressed the phenomenon in isolation, and it converts their separate insights into a single mechanism that can be applied to a case and, in principle, measured. The concept of normative entrapment is the contribution's analytic core, because it names and locates the source of leverage that the existing accounts gesture toward without specifying.

The article proceeds as follows. The next section reviews the three literatures and states the method. A results section presents the STC model, its diagram, and the index, and applies the index illustratively across three cases. Four analytical sections then develop the argument, treating normative entrapment as the source of leverage, the logic of staying below the threshold, the Belarus–Poland case read through the model, and the implications for counter-coercion. A conclusion returns to the three hypotheses, states the limitations, and identifies the empirical work the model now requires.

LITERATURE REVIEW AND METHODOLOGY

Literature Review

Three literatures bear on the problem, and the contribution of this article lies partly in joining them. The first is the scholarship on migration as an instrument of statecraft. Greenhill's foundational account established that states have repeatedly used the threat or reality of cross-border population movements to extract concessions from others, identifying the mechanism of hypocrisy costs by which liberal targets are punished for the gap between their professed values and their coercive impulses (Greenhill, 2010). A subsequent generation has broadened this into the study of migration diplomacy and migration power, mapping how states across the Global South and the European periphery convert control over mobility into bargaining leverage (Tsourapas, 2019; Fernández-Molina & Tsourapas, 2024), how refugee-hosting can be monetised into rent and political concession (Freier et al., 2021; Micinski, 2023), and how displacement has been leveraged in concrete negotiations from Greece's bailout to the European refugee crisis (Tsourapas & Zartaloudis, 2022). A critical strand cautions that the migration-diplomacy narrative can be overstated and urges precision about when leverage actually obtains (Tolay, 2022), and a recent reframing of coercive engineered migration theory through the Ceuta case sharpens the conditions under which the instrument succeeds (Álvarez-Miranda & Brey, 2025).

The second literature is the theory of coercion below the threshold of war. Work on weaponised interdependence has shown how asymmetries in networks, of finance, information, and infrastructure, can be turned into coercive leverage without firing a shot (Farrell & Newman, 2019), and the conceptual literature on hybrid and grey-zone interference has tried to specify what it means to coerce deliberately beneath the threshold that would trigger a collective or forceful response (Bergaust & Sellevåg, 2024). Migration fits this family closely, because it is a non-military instrument that imposes cost while supplying deniability, yet the coercion-theory and the migration literatures have largely developed in parallel, and one aim of the present model is to bring the leverage logic of the former to bear on the instrument studied by the latter.

The third literature is the securitisation of migration and the analysis of the European border. The Copenhagen-school account of how migration is constructed as an existential security threat, and how that construction licenses exceptional measures, has been applied extensively to the European Union and its agencies (Bello, 2022; Léonard & Kaunert, 2022), and to the externalisation and remote control by which the Union pushes migration management beyond its territory (FitzGerald, 2020; İşleyen, 2021). Applied to the recent border episodes, this literature has documented how Ceuta and the eastern border were framed as hybrid military threats that justified militarised governance and the suspension of protection (Gabrielli & Garcés Mascareñas, 2025; Kinacioglu, 2023). For the Belarus–Poland case specifically, a fast-growing body of work analyses the securitisation of the crisis and the construction of the eastern-borders route (Graban, 2024; Straczuk, 2024; Liszkowska, 2024), the polycentric and often informal governance of transit on the Union's eastern edge (Koinova, 2025a; Koinova, 2025b), the comparative responses of the frontline states (Bartasevičius et al., 2025; Braghiroli & Hagelin, 2024; Hagelin, 2024), the European Union's institutional and resilience response (Korosteleva & Petrova, 2021; Ceccorulli, 2025), and the legal contortions of the instrumentalisation concept and the crisis instruments that followed (Ganty et al., 2024; Grześkowiak, 2023; Abbassi & Triculescu, 2025).

Read together, these literatures supply the parts but not the whole. The migrationstatecraft literature establishes that the instrument exists and that hypocrisy costs matter; the coercion literature explains how sub-threshold leverage works in the abstract; and the securitisation and legal literatures describe in rich detail how the target responds and at what normative cost. What is missing is a model that connects the coercer's calibration, the source of the leverage, and the target's dilemma into a single mechanism. The normative-entrapment concept and the index built around it are offered to fill that gap.

Research Methodology

This article is a conceptual, theory-building study combined with a structured, focused comparison, and I should be explicit about why this design fits the problem and what it cannot do. The phenomenon is recent, the universe of clear cases is small, and the key variables, deniability, the target's normative bind, the asymmetry of cost, are not yet measured in any dataset, so a large-N design is premature and a single descriptive case study would not yield a generalisable mechanism. Theory-building from a small set of structured comparisons is the appropriate intermediate, and it is the method by which much of the coercion and migration-diplomacy literature has advanced (Greenhill, 2010; Álvarez-Miranda & Brey, 2025). The model is derived inductively from the three literatures and from the documented features of the recent cases, and it is then applied deductively to organise the evidence on the Belarus–Poland episode.

The model is expressed as a mechanism in four phases and as a composite index. The mechanism, set out in Figure 1 and Table 1, runs from the coercer's engineering of a flow, through its channelling to the target's frontier, to the target's normative-entrapment dilemma, to the imposition of costs and the resulting outcome, with a feedback loop by which the coercer calibrates or escalates. The index, the Coercive Leverage Index, expresses the leverage generated as CLI = (C_a · D · N) / R, where C_a is the cost-asymmetry ratio of costs imposed on the target to costs borne by the coercer, D is a deniability factor on the unit interval capturing the difficulty of attributing the act to the sponsoring state, N is a normative-entrapment factor on the unit interval capturing the degree to which the target's legal and normative commitments constrain its response, and R is a resilience term not less than one capturing the target's capacity to absorb and counter the pressure. Leverage rises with asymmetry, deniability, and entrapment and falls with resilience.

The comparison applies the index to three cases, the 2021 Belarus–Poland episode as the focal case and the 2020 Turkey–European Union and 2021 Morocco–Ceuta episodes as illustrative contrasts, with the component values estimated ordinally from the secondary peer-reviewed literature rather than measured. I want to be candid that these values are heuristic: they are assigned to demonstrate the index's logic and to discriminate among the cases in a way the qualitative evidence supports, not to claim precision the present state of measurement cannot deliver. The data are entirely secondary, drawn from the peer-reviewed studies reviewed above, and the article makes no independent empirical claim about events beyond what those sources establish. Four limitations follow directly and are stated here rather than deferred. First, the index components are ordinally estimated and await an agreed operationalisation. Second, the case set is small and European-peripheral, so the model's transferability to other regions is asserted but not demonstrated. Third, the model treats the target as a unitary actor, whereas real targets are coalitions, the European Union and a frontline member state, whose internal division is itself part of the cost. Fourth, the focal case remains politically contested and partly opaque, so the deniability and intent attributed to the coercer rest on the published analyses rather than on adjudicated fact.

RESEARCH RESULTS

The results are presented in three parts: the STC model and its mechanism, the Coercive Leverage Index and its illustrative application, and the cross-case pattern that the index reveals. Interpretation is reserved for the analytical sections that follow.

The model represents weaponised migration as a four-phase coercive mechanism operating entirely below the threshold of armed conflict, shown in Figure 1. In the first phase the coercing state engineers a flow, recruiting or admitting third-country nationals and signalling an open route; in the second it channels that flow to the target's frontier through visa facilitation, transport, and direct guidance; in the third the arrival of the flow confronts the target with the normative-entrapment dilemma, the simultaneous pull of a security imperative to seal the border and a normative and legal commitment to asylum and non-refoulement; and in the fourth the dilemma imposes costs whichever way the target turns, producing an outcome that ranges from concession to resilient adaptation and feeding back into the coercer's calibration. The decisive feature, represented at the centre of the figure, is that the leverage is generated at the third phase, in the gap between the target's security and its norms, and not at the first, in the size of the flow.

Table 1 sets out the four phases, the mechanism of each, and the way the Belarus– Poland episode of 2021 expressed them, drawing the case detail from the peer-reviewed studies of the crisis (Graban, 2024; Mészáros & Țoca, 2023; Ganty et al., 2024).

The Coercive Leverage Index expresses the magnitude of the leverage as the product of cost asymmetry, deniability, and normative entrapment divided by target resilience. Applied ordinally to the three cases, the index discriminates among them in a way the qualitative record supports, and the contrast is instructive. The Belarus–Poland episode scores highest, because it combined a high cost asymmetry with a relatively high deniability, the Belarusian state never formally admitting direction, and a very high normative entrapment, the European Union being constitutively committed to refugee protection, against a moderate resilience. The 2020 Turkey–European Union episode scores lower despite a comparable flow, chiefly because Ankara acted overtly, announcing the open border and thereby forfeiting deniability, which depresses the index even as entrapment remained high. The 2021 Morocco–Ceuta episode falls between, combining moderate deniability with high entrapment against a smaller and more localised target. Table 2 records the illustrative component values and the resulting index.

Figure 1. The Sub-Threshold Coercion (STC) model of weaponised migration. The mechanism runs in four phases (engineer, channel, entrap, impose costs) below the armed-conflict threshold; coercive leverage is generated at the normative-entrapment dilemma rather than by the magnitude of the flow. Source: author.

Table 1. The four phases of the Sub-Threshold Coercion mechanism, with the Belarus–Poland 2021 illustration.

• 1. Engineer — Mechanism: Recruit/admit third-country nationals; signal an open route; Belarus–Poland 2021 illustration: Tourist visas issued; Minsk presented as a transit hub to the EU
• 2. Channel — Mechanism: Facilitate transport and guide the flow to the frontier; Belarus–Poland 2021 illustration: State-linked transport and guidance of migrants to the Polish border line
• 3. Entrap — Mechanism: Confront target with the security-vs-norms dilemma; Belarus–Poland 2021 illustration: EU/Poland torn between sealing the border and asylum/non-refoulement duties
• 4. Impose costs — Mechanism: Extract political, humanitarian, financial, reputational costs; Belarus–Poland 2021 illustration: Pushbacks, state of emergency, reputational damage, internal EU contention

Table 2. Illustrative Coercive Leverage Index across three cases. Component values are ordinal, model-derived estimates assigned to demonstrate the index's logic, not measured quantities: (C_a = cost asymmetry; D = deniability, 0–1; N = normative entrapment, 0–1; R = target resilience, ≥1).

Case                      C_a  D    N     R    CLI
------------------------  ---  ---  ----  ---  ----
Belarus–Poland (2021)     8    0.7  0.8   2.0  2.24
Morocco–Ceuta (2021)      5    0.5  0.75  1.8  1.04
Turkey–EU / Evros (2020)  6    0.3  0.8   2.5  0.58

The cross-case pattern yields the central result. The index is dominated by deniability and normative entrapment, not by the cost-asymmetry term that tracks the scale of the flow, since the case with arguably the largest overt mobilisation, Turkey at Evros, generates the least leverage by this measure precisely because its overtness collapsed deniability. The result is consistent across the three cases and supports the model's core claim that leverage is a function of the coercer's ability to remain deniable and of the target's normative bind, rather than of the number of people moved. The analytical sections develop the three components of that claim in turn.

NORMATIVE ENTRAPMENT AS THE SOURCE OF LEVERAGE

The first and most important claim of the model is that the leverage of weaponised migration is generated by the target, not by the coercer. The coercer supplies the occasion, a flow of people at the frontier, but the pressure that flow exerts is produced by the target's own commitments, because a state bound by asylum and non-refoulement obligations cannot simply repel the arrivals without violating the norms that constitute its identity and underwrite its legitimacy, and cannot admit them without conceding the very entry the coercer sought to force. This is the normative-entrapment dilemma, and it is the analytic heart of the contribution. Greenhill's foundational insight about hypocrisy costs anticipated it, locating the punishment of liberal targets in the gap between professed values and coercive behaviour (Greenhill, 2010), and the present model formalises that insight by making the size of the gap, rather than the size of the flow, the operative variable.

The Belarus–Poland case displays the mechanism with unusual clarity. The European Union and Poland confronted a small flow by the standards of the 2015 crisis, yet the pressure was acute because the Union's legal order and self-understanding are saturated with refugee-protection commitments that the arrivals activated automatically (Ganty et al., 2024; Grześkowiak, 2023). The response, suspension of asylum access, pushbacks, and a state of emergency that sealed the border zone, did seal the frontier, but at a normative cost the legal literature has documented in detail, including the strained invention of an instrumentalisation rationale to excuse measures otherwise difficult to reconcile with European and international law (Ganty et al., 2024; Abbassi & Triculescu, 2025). The coercer did not have to inflict that cost directly; the target inflicted it on itself by choosing security over norms, which is exactly what entrapment means.

The securitisation literature illuminates the other horn of the dilemma, the discursive work required to make the security choice acceptable. Framing the crisis as a hybrid threat and an emergency was the move that licensed the exceptional measures, transforming a humanitarian situation into a security one in which pushbacks could be presented as defence rather than as the abandonment of protection (Graban, 2024; Straczuk, 2024; Gabrielli & Garcés Mascareñas, 2025). That framing reduced the immediate political cost of the security choice, but it did not eliminate the normative cost, and it generated its own long-run liabilities, including the erosion of the asylum acquis and the precedent it set, which the legal literature treats as a lasting consequence of the episode (Grześkowiak, 2023; Ganty et al., 2024). Securitisation, in the model's terms, is the target's attempt to lower its own normative-entrapment factor by redefining the situation, and its partial success and partial failure are both visible in the European response.

The comparative cases confirm that entrapment, not flow, is doing the analytic work. At Ceuta the flow was large and the target was a single enclave, yet the leverage operated through the same bind, Spain's obligations toward arrivals including unaccompanied minors, and the reframing of the episode as a hybrid military threat was again the device by which the security response was legitimised (Gabrielli & Garcés Mascareñas, 2025; Álvarez-Miranda & Brey, 2025). At Evros the entrapment was equally high, Greece and the Union being no less bound, yet the leverage was lower, and the model attributes the difference not to the flow, which was substantial, but to Turkey's overtness, which is the subject of the next section. The consistent finding across the cases is that wherever the target's normative commitments are deep, a modest engineered flow suffices to generate acute pressure, which is the first hypothesis stated as a result.

STAYING BELOW THE THRESHOLD: DENIABILITY AND THE NON-MILITARY INSTRUMENT

The second claim of the model concerns calibration. Weaponised migration is attractive precisely because it coerces without crossing the threshold that would license a forceful or collective-defence response, and the model treats this sub-threshold character as constitutive rather than incidental. The instrument is non-military by nature, it moves civilians rather than munitions, so it does not register as an armed attack, and it is deniable by design, because the sponsoring state can attribute the movement to the migrants' own decisions, to smugglers, or to humanitarian circumstance even as it organises the channel (Bergaust & Sellevåg, 2024; Tsourapas, 2019). The literature on coercion below the threshold of war explains why this matters: an instrument that stays beneath the threshold denies the target the justification for the decisive response that might otherwise deter it, leaving the target to choose among options that all lie within the permissive zone the coercer has constructed (Farrell & Newman, 2019; Bergaust & Sellevåg, 2024).

Deniability is the component of the index most sensitive to the coercer's conduct, and the contrast between Belarus and Turkey demonstrates its weight. The Belarusian authorities never formally acknowledged orchestrating the flow, maintaining a posture of injured innocence that, however implausible, preserved enough ambiguity to complicate attribution and to divide the target's response between those who saw a hybrid attack and those who saw a humanitarian emergency (Mészáros & Țoca, 2023; Graban, 2024). Turkey, by contrast, announced in early 2020 that it would no longer hold migrants back and openly facilitated movement to the Evros border, a transparency that converted the act into an unambiguous instance of state coercion and thereby forfeited the deniability that protects such operations (İşleyen & Karadağ, 2023). In the model, this is why the Turkish case scores a low deniability factor and a correspondingly lower index despite a comparable flow and an equally entrapped target, a result the qualitative literature supports.

The non-military character of the instrument also shapes the target's available responses and, through them, the feedback loop. Because the flow is not an armed attack, the target's counter-measures are confined to the domains of border control, law, and diplomacy, and each of these imposes costs the coercer can observe and exploit. Border hardening generates humanitarian and reputational costs; legal suspension of protection generates normative costs; diplomatic concession generates the strategic cost the coercer sought in the first place (Ganty et al., 2024; Korosteleva & Petrova, 2021). The coercer reads the target's response and calibrates, sustaining, escalating, or relaxing the flow to keep the pressure within the band that maximises leverage without provoking the threshold-crossing response that would change the game. This calibration, represented by the feedback arrow in the model, is what makes the instrument controllable, and it is the second hypothesis stated as a result: the sub-threshold and deniable character of the instrument is not a limitation on its effectiveness but the condition of it.

There is a boundary to the technique that the model also makes visible, and intellectual honesty requires noting it. Deniability is fragile and decays with repetition, since a state that reaches for the instrument repeatedly erodes the ambiguity that protects it, and a target that has learned to read the pattern attributes the next episode more quickly and more confidently (Fernández-Molina & Tsourapas, 2024; Bergaust & Sellevåg, 2024). The Belarus–Poland episode, by being so widely analysed and so confidently attributed in the scholarship, may have lowered the deniability available for the next such operation in the region, which would, in the model's terms, lower the index and blunt the instrument. Whether that erosion is durable is an open question the model can pose but not yet answer.

THE BELARUS–POLAND CASE READ THROUGH THE MODEL

Applying the full model to the focal case shows how the components combine, and it also tests whether the model adds anything to the descriptive accounts. The engineering and channelling phases are well documented: Belarus eased entry for third-country nationals, its state-linked apparatus presented Minsk as a gateway to the European Union, and migrants were transported and guided to the border line, where Belarusian forces are widely reported to have prevented their return (Mészáros & Țoca, 2023; Graban, 2024). The model's value is not in re-describing these phases but in locating where the leverage was generated, and it directs attention away from the flow, which was modest, and toward the entrapment, which was severe, because the European Union could neither absorb the arrivals without conceding the coercer's aim nor repel them without damaging the legal order that the same Union exists to uphold (Ganty et al., 2024; Grześkowiak, 2023).

The cost-imposition phase is where the model's account becomes sharpest. The costs fell on the target across every domain the index anticipates. Politically, the crisis divided the European Union and strained the relationship between frontline states and the Commission over the legality of the response (Ceccorulli, 2025; Abbassi & Triculescu, 2025). Humanitarian and reputational costs accrued as pushbacks and the border-zone emergency produced suffering and deaths that drew condemnation and litigation, and that the securitising frame could contain but not erase (Straczuk, 2024; Grześkowiak, 2023). Operationally and financially, the frontline states mounted expensive border deployments and physical barriers, and the comparative literature shows how Poland, Lithuania, and Latvia each absorbed these costs differently according to their capacity and politics (Bartasevičius et al., 2025; Braghiroli & Hagelin, 2024; Hagelin, 2024). Societally, the crisis polarised domestic publics over the competing claims of security and protection, a cost the coercer could observe and that fed the feedback loop. The model predicts costs in exactly these domains, and the case supplies them.

The outcome phase reveals a result the descriptive accounts tend to underplay. The European response was not a clean concession, and in that sense the coercion did not fully succeed: the border was held, the flow was eventually reduced, and the Union did not lift the sanctions that had provoked the episode (Korosteleva & Petrova, 2021; Ceccorulli, 2025). Yet the coercion was not a clean failure either, because the price of holding the border was paid in the currency the model identifies, the partial sacrifice of the asylum acquis and the normalisation of exceptional measures, which the legal literature treats as a durable cost that outlasts the crisis (Ganty et al., 2024; Grześkowiak, 2023; Abbassi & Triculescu, 2025). The model captures this ambiguous outcome better than a binary success-or-failure reading, because it treats the target's resilience as reducing but not eliminating the leverage, leaving a residue of cost that is the coercion's real yield. The high index value assigned to the case reflects this: not that Belarus achieved its stated aim, but that it imposed disproportionate cost at low cost to itself while remaining below the threshold, which is what the index measures.

Reading the case through the model also exposes where the governance of the eastern border was weakest, in the informal and polycentric arrangements through which transit was managed and contested (Koinova, 2025a; Koinova, 2025b). The coercer exploited not only the Union's normative commitments but the unevenness of its eastern-border governance, finding in the seams between national and Union competence, and between security and protection mandates, the space in which the instrument operates most freely. That observation, which the model frames as a feature of the target's low resilience, points directly to the counter-coercion implications taken up next.

COUNTERING SUB-THRESHOLD MIGRATION COERCION: LOWERING THE INDEX

If the model is correct that leverage is the product of cost asymmetry, deniability, and normative entrapment divided by resilience, then it follows that effective counter-coercion works by lowering that quantity, and the model's practical value is that it specifies which terms a target can actually move. The instinctive response, hardening the border to stop the flow, attacks the cost-asymmetry term by reducing the flow, but the analysis above shows that the flow is the weakest determinant of leverage, so border hardening alone addresses the least important variable while incurring the humanitarian and normative costs that raise the very entrapment it ignores (Graban, 2024; Ganty et al., 2024). The model therefore predicts that a border-only response is self-defeating, which is consistent with the Belarus–Poland record, where sealing the frontier held the line but paid the normative price that constituted the coercion's yield.

The two terms a target can move with greater profit are normative entrapment and resilience. Reducing normative entrapment does not mean abandoning protection norms, which would hand the coercer a strategic victory and corrode the target's identity, but closing the gap between security and norms by responding through lawful, legitimacy-preserving channels, processing claims rapidly and fairly at the border rather than suspending them, so that the target can both control entry and honour its commitments and thereby deny the coercer the hypocrisy costs the instrument feeds on (Grześkowiak, 2023; Ganty et al., 2024). The European Union's post-crisis legislative response, the crisis and force-majeure instruments built to manage exactly such situations within the legal order rather than outside it, can be read as an attempt to lower normative entrapment institutionally, though the legal literature is divided on whether these instruments reduce the gap or merely codify a lower standard of protection (Abbassi & Triculescu, 2025; European Union, 2024; European Commission, 2021).

Raising resilience, the denominator of the index, is the other lever, and it operates by increasing the target's capacity to absorb the pressure without conceding. Resilience here is partly material, the reception capacity, surveillance, and border infrastructure that reduce the operational shock of an engineered flow (Korosteleva & Petrova, 2021; Bartasevičius et al., 2025), and partly political and institutional, the unity of the target coalition and the coherence of its eastern-border governance that deny the coercer the internal seams it exploits (Ceccorulli, 2025; Koinova, 2025a). A target that can process arrivals lawfully and quickly, that presents a united rather than a divided front, and that has rehearsed the contingency converts a strategic crisis into a manageable administrative event, which is precisely the reduction in leverage the model describes. Counter-coercion, on this account, is less about the border than about the institutions behind it.

Two further measures attack the deniability term, and the model clarifies their logic. Rapid, public, and evidentially grounded attribution raises the political cost to the coercer and erodes the ambiguity that protects the operation, while a coordinated rather than a fragmented response among targets denies the coercer the divided reaction that deniability is meant to produce (Fernández-Molina & Tsourapas, 2024; Bergaust & Sellevåg, 2024). Neither measure stops a determined coercer, and the model does not promise that the leverage can be reduced to zero, since a sufficiently committed sponsor against a sufficiently entrapped target will always generate some pressure. What the model offers is a map of where the target's effort is best spent, and its clear answer, that the effort belongs on entrapment, resilience, and deniability rather than on the flow, is the third hypothesis stated as a result.

CONCLUSION

The movement of people has joined the standard repertoire of coercion below the threshold of war, and the Belarus–Poland episode of 2021 showed both how effective the instrument can be and how poorly the existing, mainly descriptive accounts explain its workings. This article set out to provide the missing mechanism, and the model it proposes supports an ordered set of conclusions.

The first hypothesis, that leverage derives primarily from normative entrapment rather than from the magnitude of the flow, is supported by the model and by the cross-case comparison. The case with the largest overt mobilisation generated the least leverage, while the Belarus–Poland case, with a modest flow but a deeply entrapped target and a deniable channel, generated the most, which is intelligible only if the source of leverage is the target's normative bind rather than the number of arrivals (Greenhill, 2010; Ganty et al., 2024). The second hypothesis, that the sub-threshold and deniable character of the instrument is constitutive of its effectiveness, is likewise supported: deniability is the index component most sensitive to the coercer's conduct, and the collapse of the Turkish case's leverage upon Ankara's overtness shows that staying below the threshold and preserving ambiguity are not incidental conveniences but the conditions of the technique (İşleyen & Karadağ, 2023; Bergaust & Sellevåg, 2024).

The third hypothesis, that effective counter-coercion lowers the index by reducing entrapment and raising resilience rather than by hardening the border, follows from the structure of the model and is consistent with the case record. Border hardening attacks the weakest term while incurring the normative costs that constitute the coercion's yield, whereas lawful and legitimacy-preserving responses that close the gap between security and norms, together with the material and institutional resilience that absorbs the shock, attack the terms that actually drive the leverage (Grześkowiak, 2023; Korosteleva & Petrova, 2021; Ceccorulli, 2025). The European Union's turn to crisis instruments within the legal order can be read as an institutional attempt to lower entrapment, with the legal literature still divided over whether it succeeds.

The principal original contribution of this article is the Sub-Threshold Coercion model and its Coercive Leverage Index, and above all the concept of normative entrapment that names and locates the source of leverage the prior literature had only gestured toward. The model joins the coercion-theory, the migration-diplomacy, and the securitisation literatures into a single mechanism, it explains why a small engineered flow can coerce a great power, it predicts the conditions under which the instrument works better or worse, and it identifies where a target can intervene. That integrative and generative ambition is the proper aim of an original contribution, and it is the one pursued here.

The limitations are real and bound the claims. The index components are estimated ordinally and await an agreed operationalisation that would let the model be tested rather than illustrated; the case set is small and confined to the European periphery, so the asserted transferability to other regions remains to be shown; the target is modelled as unitary where it is in truth a contentious coalition; and the focal case is politically contested, so the intent and direction attributed to the coercer rest on the published analyses rather than on adjudicated fact. These boundaries define the agenda the model now sets: an operationalisation of the index components that survives measurement, a wider comparison that tests the mechanism beyond Europe, and a disaggregation of the target that treats its internal division as part of the leverage rather than as noise. Until that work is done, the safest reading is that the power of weaponised migration lies not in the people moved but in the values of those they are moved against, and that a target's best defence is to honour those values in a manner that denies the coercer the gap on which the instrument feeds.

BIBLIOGRAPHY

Abbassi, N., & Triculescu, R.-M. (2025). The Crisis and Force Majeure Regulation through the lens of synchronization: Institutionalizing the EU migration crisis response. European Politics and Society, 27(2), 273–285. https://doi.org/10.1080/23745118.2025.2576701

Álvarez-Miranda, B., & Brey, E. (2025). Reframing coercive engineered migration theory: Ceuta and the Western Sahara. Mediterranean Politics, 30(2), 316–337. https://doi.org/10.1080/13629395.2023.2293417

Bartasevičius, V., Fuks, V., & Nakrošis, V. (2025). Lithuania's response to the crisis of irregular migration in 2021–2024. Politologija, 119(3), 44–84. https://doi.org/10.15388/polit.2025.119.2

Bello, V. (2022). The spiralling of the securitisation of migration in the EU: From the management of a 'crisis' to a governance of human mobility? Journal of Ethnic and Migration Studies, 48(6), 1327–1344. https://doi.org/10.1080/1369183X.2020.1851464

Bergaust, J. C., & Sellevåg, S. R. (2024). Improved conceptualising of hybrid interference below the threshold of armed conflict. European Security, 33(2), 169–195. https://doi.org/10.1080/09662839.2023.2267478

Braghiroli, S., & Hagelin, S. (2024). More than neighbours: Assessing crisis narratives of Nordic-Baltic elites in the context of the European Union's eastern border. New Perspectives, 32(2), 113–144. https://doi.org/10.1177/2336825X241230833

Ceccorulli, M. (2025). Geopolitics and migration: The European Union's misstep. European Politics and Society, 1–12. Advance online publication. https://doi.org/10.1080/23745118.2025.2583071

European Commission. (2021). Proposal for a Regulation of the European Parliament and of the Council addressing situations of instrumentalisation in the field of migration and asylum (COM(2021) 890 final). https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:52021PC0890

European Union. (2024). Regulation (EU) 2024/1359 of the European Parliament and of the Council of 14 May 2024 addressing situations of crisis and force majeure in the field of migration and asylum. Official Journal of the European Union, L 2024/1359. https://eur-lex.europa.eu/eli/reg/2024/1359/oj

Farrell, H., & Newman, A. L. (2019). Weaponized interdependence: How global economic networks shape state coercion. International Security, 44(1), 42–79. https://doi.org/10.1162/isec_a_00351

Fernández-Molina, I., & Tsourapas, G. (2024). Understanding migration power in international studies. International Affairs, 100(6), 2461–2479. https://doi.org/10.1093/ia/iiae230

FitzGerald, D. S. (2020). Remote control of migration: Theorising territoriality, shared coercion, and deterrence. Journal of Ethnic and Migration Studies, 46(1), 4–22. https://doi.org/10.1080/1369183X.2020.1680115

Freier, L. F., Micinski, N. R., & Tsourapas, G. (2021). Refugee commodification: The diffusion of refugee rent-seeking in the Global South. Third World Quarterly, 42(11), 2747–2766. https://doi.org/10.1080/01436597.2021.1956891

Gabrielli, L., & Garcés Mascareñas, B. (2025). The framing of migration as a hybrid military threat: Insights from Ceuta's 2021 'crisis'. Migration Studies, 13(1), mnaf007. https://doi.org/10.1093/migration/mnaf007

Ganty, S., Ancite-Jepifánova, A., & Kochenov, D. V. (2024). EU lawlessness law at the EU-Belarusian border: Torture and dehumanisation excused by 'instrumentalisation'. Hague Journal on the Rule of Law, 16(3), 739–774. https://doi.org/10.1007/s40803-024-00237-0

Graban, A. (2024). Instrumentalisation of fear and securitisation of the 'Eastern Borders Route': The case of the Poland–Belarus 'border crisis'. European Security, 33(2), 236–260. https://doi.org/10.1080/09662839.2023.2287499

Greenhill, K. M. (2010). Weapons of mass migration: Forced displacement, coercion, and foreign policy. Cornell University Press.

Grześkowiak, M. (2023). The 'Guardian of the Treaties' is no more? The European Commission and the 2021 humanitarian crisis on the Poland–Belarus border. Refugee Survey Quarterly, 42(1), 81–102. https://doi.org/10.1093/rsq/hdac025

Hagelin, S. (2024). Framing entangled borders in the Baltic states. Journal of Borderlands Studies, 39(6), 1101–1120. https://doi.org/10.1080/08865655.2023.2276475

İşleyen, B. (2021). Technology and territorial change in conflict settings: Migration control in the Aegean Sea. International Studies Quarterly, 65(4), 1087–1096. https://doi.org/10.1093/isq/sqab046

İşleyen, B., & Karadağ, S. (2023). Engineered migration at the Greek–Turkish border: A spectacle of violence and humanitarian space. Security Dialogue, 54(5), 475–492. https://doi.org/10.1177/09670106231194911

Kinacioglu, M. (2023). Militarized governance of migration in the Mediterranean. International Affairs, 99(6), 2423–2441. https://doi.org/10.1093/ia/iiad232

Koinova, M. (2025a). Governing transit and irregular migration: Informality and formal policies. Journal of Ethnic and Migration Studies, 51(2), 423–444. https://doi.org/10.1080/1369183X.2024.2371204

Koinova, M. (2025b). Power and informality in the polycentric governing of transit and irregular migration on the EU's eastern border with Belarus. Journal of Ethnic and Migration Studies, 51(2), 501–525. https://doi.org/10.1080/1369183X.2024.2371214

Korosteleva, E., & Petrova, I. (2021). Community resilience in Belarus and the EU response. JCMS: Journal of Common Market Studies, 59(S1), 124–136. https://doi.org/10.1111/jcms.13248

Léonard, S., & Kaunert, C. (2022). The securitisation of migration in the European Union: Frontex and its evolving security practices. Journal of Ethnic and Migration Studies, 48(6), 1417–1429. https://doi.org/10.1080/1369183X.2020.1851469

Liszkowska, D. (2024). Securitisation of migration in the European Union during the EU-Belarus border crisis (2021–2022). Politeja, 21(1/88), 41–56. https://doi.org/10.12797/Politeja.21.2024.88.3.03

Mészáros, E. L., & Țoca, C. V. (2023). The EU's resilience and the management of hybrid threats coming from the Eastern neighbourhood: Belarus and the deliberate facilitation of irregular immigration. Eastern Journal of European Studies, 14(1), 5–30. https://doi.org/10.47743/ejes- 2023-0101

Micinski, N. R. (2023). Threats, deportability and aid: The politics of refugee rentier states and regional stability. Security Dialogue, 54(6), 529–547. https://doi.org/10.1177/09670106211027464

Straczuk, J. (2024). 'Hybrid war', military humanitarianism, and epistemic friction: Framing illegalised migration on the Polish-Belarusian border. Journal of Ethnic and Migration Studies, 50(16), 4234–4252. https://doi.org/10.1080/1369183X.2023.2293637

Tolay, J. (2022). Interrogating and broadening the emerging narrative on migration diplomacy: A critical assessment. Millennium: Journal of International Studies, 51(1), 354–375. https://doi.org/10.1177/03058298221139589

Tsourapas, G. (2019). The Syrian refugee crisis and foreign policy decision-making in Jordan, Lebanon, and Turkey. Journal of Global Security Studies, 4(4), 464–481. https://doi.org/10.1093/jogss/ogz016

Tsourapas, G., & Zartaloudis, S. (2022). Leveraging the European refugee crisis: Forced displacement and bargaining in Greece's bailout negotiations. JCMS: Journal of Common Market Studies, 60(2), 245–263. https://doi.org/10.1111/jcms.13211

Yeliseyeu, A. (2024). Belarus's coercive engineered migration case of 2021–2022: Categorisation of state media narratives. Studia Migracyjne – Przegląd Polonijny, 49(3), 79–99. https://doi.org/10.4467/25444972SMPP.23.030.19145