■ See Korean Version on EAI Website

I. Introduction

As U.S.–China competition intensifies, the concepts of ‘entrapment’ and ‘abandonment’ are frequently invoked in domestic security discourse. When someone points to the risk of entrapment in connection with strengthening ROK–U.S.–Japan security cooperation or a contingency in the Taiwan Strait, the rejoinder that follows is typically that “if the contribution to the alliance is insufficient, there is also a risk of abandonment.” Such exchanges usually stagnate without yielding much, or end with the platitude that “a proper balance must be struck between the two.” Refutation is difficult, yet it is equally difficult to draw any meaningful policy implications from them.

Within this discursive framework, the North Korean threat is invariably invoked. The logic runs that responding to North Korea’s nuclear weapons makes alliance cohesion and the strengthening of extended deterrence unavoidable, and that being passive toward the alliance invites abandonment. The argument of this paper, however, is that the two risks do not coexist symmetrically. Whereas the threat originating from North Korea is a relatively well-specified risk that the ROK–U.S. alliance has ‘managed’ over a long period, entrapment in great-power conflict is a risk of an entirely different order of magnitude and lethality—one that could turn South Korea into the very primary target of a conflict.

Accordingly, this paper critiques the ‘mechanical-balance thesis,’ which treats the risks of entrapment and abandonment as equivalent. The relative magnitude of the two risks is not a constant but a variable, and which one predominates depends on the given structure and conjuncture. It follows that asserting a risk—whether of entrapment or of abandonment—requires specifying the concrete actors, structural circumstances, causal pathways, and time horizons involved. To claim, without furnishing any of this, that “the opposite risk also exists” is not, in fact, a balanced proposal but rather an exercise in futile spinning of wheels, or a mere deferral of judgment.

Entrapment is a risk that is already underway, one that arises from military and geopolitical causes alone. Abandonment sits on a different timeline: it emerges as a probable risk only after the structure of entrapment has matured. Overlooking this difference legitimizes a dangerous prescription—namely, the ‘preemptive-contribution thesis,’ which holds that “in order not to be abandoned, one must cooperate more and concede more right now.” This prescription, invoking a hard-to-specify possibility of abandonment as its justification, induces the acceptance of an entrapment risk that already exists in the present, and moreover, it actually hastens the arrival of an abandonment risk that could only otherwise materialize after entrapment.

II. Are Entrapment and Abandonment Symmetrical Risks?: A Theoretical Reexamination

Even within its own theoretical lineage, the mechanical-balance thesis is hard to substantiate. Snyder specified the conditions that determine which of the two risks—entrapment or abandonment—predominates:[1] the degree of dependence on the alliance, the explicitness of the commitment, the scope of shared interests, the availability of alternative partners, and the polarity of the system. In fact, no theory dealing with entrapment and abandonment holds that, because the two risks coexist, one must always maintain a balance between them. Rather, such theories are concerned with discerning which variable becomes more decisively activated under a given structure and conjuncture. Identifying the dilemma is the starting point of analysis, not its conclusion.

Victor Cha’s theory of quasi-alliance gives concrete form to this point in the context of ROK–U.S.–Japan relations.[2] According to Cha, the trajectory of ROK–Japan cooperation hinged on which risk the two countries came to regard as more dominant, depending on shifts in the U.S. commitment. In this theory, the weight of risk tilts to one side or the other depending on the conjuncture, and a state of equivalence between the two is never posited. Moreover, what determines this is not so much the domestic politics or bilateral relations of Korea and Japan as the variable of the United States. That security cooperation between the two countries has remained at an unprecedented level today—when the United States emphasizes, more than ever before, the strengthening of an alliance encompassing both Korea and Japan, notwithstanding changes in the domestic political landscape of each—vividly illustrates this structural determination.

The mechanical-balance thesis does not merely homogenize the differences in the pathways and timelines of the two risks; it also leads to a hazardous policy prescription—the preemptive-contribution thesis. This prescription fails to specify the pathway and mechanism of the very abandonment it seeks to forestall. It also carries an internal explanatory burden, in that it posits the United States simultaneously as an actor capable of abandoning an ally according to its own interests and as an actor bound by the contributions of its partner countries. While this will be discussed in detail later, it suffices to note here that what determines the likelihood of abandonment is not the level of South Korea’s contribution, but the degree to which the United States needs South Korea’s assets.

The entrapment risk that South Korea currently faces stems from a structural transformation that has already been unfolding gradually, over a considerable period, in the geopolitical and military domains as U.S.–China strategic competition proceeds. The risk of abandonment, by contrast, does not lie on the same pathway. If anything, the more intense U.S.–China competition becomes, the more the United States is compelled to place greater value on South Korea’s geopolitical utility. Furthermore, given the current state of U.S. manufacturing capacity, South Korea’s capabilities in semiconductors, shipbuilding, and batteries are not the kind of assets the United States can easily forgo. Conversely, the risk of abandonment becomes probable only after the United States has, at a minimum, transferred a substantial portion of South Korea’s capabilities into its own ‘endogenous’ capacity (that is, the ‘internalization of industrial capacity’). In other words, abandonment is not likely to be the symmetrical counterpart of entrapment—occupying the same pathway and the same point in time—but rather the eventual consequence of the entrapment structure itself.

How, then, do the mechanical-balance thesis and the preemptive-contribution thesis distort South Korea’s discursive space? The crux of the matter is that they treat the relationship between entrapment and abandonment—which major theorists such as Snyder presented only conditionally—as though it were an (unconditional) law. The linkage that “distancing oneself to avoid entrapment increases the risk of abandonment” holds only under specific conditions, and the fulcrum determining whether it holds is not South Korea’s own behavior but the structural condition of how much the United States needs South Korea’s assets. As long as U.S. need persists, any increase in the risk of abandonment is constrained even if South Korea tries to keep its distance; conversely, once that need disappears, it is more probable that there are limits to how much the risk of abandonment can be reduced no matter how closely South Korea tries to align itself.

The mechanical-balance thesis strips away precisely this conditionality, and the preemptive-contribution thesis compounds this by concluding that, if South Korea raises its level of contribution, the United States cannot abandon it. This is a dual flaw: the error of mistaking a conditional relationship for an unconditional law, and the leap of building policy atop that mistaken premise.

III. The U.S. Variable: Changes in the Historical Dynamics of ROK–U.S.–Japan Trilateral Cooperation

During the Cold War, the United States frequently called for ROK–U.S.–Japan trilateral cooperation. Yet in periods when the U.S. commitment and involvement were strengthening, Korea and Japan instead avoided trilateral cooperation. It was only when the United States scaled back its security commitment that Korea and Japan actually drew closer to one another.[3] For instance, the Nixon Doctrine and the partial withdrawal of U.S. Forces Korea in the 1970s, together with Carter’s attempted troop withdrawal, made the risk of abandonment the predominant concern in both countries, and at such times Korea and Japan leaned toward cooperation despite their mutual antipathy. Conversely, in periods when the U.S. commitment and involvement were solid, both countries—Japan in particular taking the lead—rejected calls for trilateral cooperation.

This strategic stance of Japan’s had a clear geopolitical background. In this period, Japan faced no imminent threat. The main force of the Soviet Union was deployed on the European front, and both North Korea and China were economically and militarily weak. Above all, as long as the Korean Peninsula functioned as a buffer zone in the U.S.–Soviet confrontation, Japan could stand clear of the front line, entrust its security to the United States, and devote itself to economic development. That Japan enjoyed its greatest economic boom in history while keeping defense spending below 1 percent of GDP was a product of precisely this structure.

Trilateral cooperation, which implied institutionalizing security cooperation with South Korea, was for Japan an excessive choice that raised risk without offering any benefit. This was because it meant not merely raising the probability of becoming entangled in a conflict on the Korean Peninsula, but also entrapping itself more deeply in a confrontational structure with the Soviet Union—a nuclear power and the United States’ true target. A similar pattern persisted in the early post–Cold War period as well. With the Soviet Union gone and China still poor, Japan had no incentive to respond even though the United States repeatedly demanded that it contribute more to regional security.

After the end of the Cold War, however, as China’s rise became visible, Japan’s calculus began to change. Through the 1995–1996 Taiwan Strait Crisis, Japan came to perceive China’s belligerence clearly, and the confrontation over the Diaoyu/Senkaku Islands turned into a concrete prospect of direct clashes with China over territorial and maritime sovereignty and hegemony in the East China Sea. In 2010, China’s GDP overtook Japan’s, and today the gap exceeds fourfold. Added to this was Japan's historical memory of its aggression against China compounded this, leading Tokyo to perceive itself as exposed to the Chinese threat.

Crucially, unlike the Soviet Union during the Cold War, China’s expansion no longer passes through the Korean Peninsula. China confronts the United States and Japan directly through its own maritime access routes in the South China Sea and the East China Sea. It is now Japan itself that is exposed to the front line of great-power conflict. The very logic that once led Japan to avoid trilateral cooperation during the Cold War has now been inverted into a logic demanding trilateral cooperation. In other words, by institutionalizing ROK–U.S.–Japan security cooperation so as to shift the initial target in a contingency from Japan to South Korea, the aim has become to reconfigure South Korea into a buffer zone (in a new context) for both the United States and Japan.

The implication is clear. The current effort to deepen trilateral cooperation is not, as in the 1970s during the Cold War, the product of growing Korean and Japanese fears of abandonment stemming from a weakening of U.S. involvement. At present, the United States has codified its security commitment at Camp David in August 2023 and, in accordance with this, is strengthening institutional and military cohesion through measures such as the establishment of a trilateral secretariat and the reinforcement of trilateral military cooperation. In other words, the dynamic now at work is not the resolution of Korean and Japanese anxieties over abandonment arising from a weakened U.S. involvement, but rather the deepening of an entrapment structure centered on a strengthened U.S. strategy toward China. This configuration has persisted despite the prolongation of wars in Europe and the Middle East, suggesting that the priority the United States places on its China strategy has not wavered even amid the dispersal of multiple fronts.

The shift in the United States’ primary threat perception is evident not only in ROK–Japan relations but also in the way it has handled the North Korea issue. When the Soviet Union, the principal adversary, disappeared with the end of the Cold War, the United States had to secure a new justification for stationing its forces in Korea, and the North Korean nuclear threat filled that role. Since North Korea revealed its intent to develop nuclear weapons in the early 1990s, the United States has managed this threat while gradually reorienting the ROK–U.S. alliance toward China. The 1999 demand for joint missile defense was the concrete opening signal of this, and through the Bush, Obama, and Trump administrations, the strengthening of trilateral cooperation was consistently pursued under the guise of responding to the North Korean threat. Even the policy of ‘strategic patience’ toward North Korea’s nuclear program, pursued within the broader framework of the Obama administration’s ‘pivot to Asia,’ served as a means of laying the groundwork for trilateral cooperation in earnest. By the time of the Trump administration, it became openly apparent that trilateral cooperation was aimed at China, and this has continued on a bipartisan basis through the Biden period and into Trump’s second term.

What deserves attention is that responding to North Korea’s nuclear weapons does not, in itself, require trilateral cooperation. South Korea’s GDP is dozens of times that of North Korea, and its defense budget alone exceeds the entire size of the North Korean economy. North Korea’s nuclear weapons are a problem to be addressed through the ROK–U.S. alliance’s extended deterrence or South Korea’s own capabilities, not a matter requiring a ROK–U.S.–Japan trilateral framework. That the North Korean threat has nonetheless been repeatedly invoked as the justification for ‘strengthening’ trilateral cooperation and extended deterrence suggests that, in essence, it has functioned as a mechanism binding South Korea to the United States’ China strategy and regional strategy, well beyond merely addressing North Korea. As will be discussed below, this point calls for reconsidering the conventional wisdom that reads concerns over extended deterrence as a sign of abandonment.

IV. The Immediacy of the Entrapment Risk and the Latency of the Abandonment Risk

1. The Entrapment Risk in Progress

Entrapment is not a mere discursive assumption but a structural transformation actually underway. Developments since the end of the Cold War—particularly since China’s rise—bear this out. The U.S. National Defense Strategy (NDS), released in January 2026, sets out four priority tasks, which explicitly include the defense of the homeland and the Western Hemisphere, deterrence of China, an expanded burden-sharing by allies, and the mobilization of the defense industrial base.[4]

It stated that the United States would build a ‘robust denial defense’ along the First Island Chain in the Indo-Pacific. With respect to the Korean Peninsula, it emphasized that South Korea has the capacity to bear the ‘primary responsibility’ for deterring North Korea, under U.S. support that is ‘decisive but more limited.’ If this signifies a shift in the character of U.S. Forces Korea from deterrence of North Korea to deterrence of China, then the probability increases that the operation of forces stationed on the Korean Peninsula will become linked to contingencies outside the peninsula. Some in South Korea liken this to a ‘new Acheson Line’ and read it as a sign of abandonment, but this analogy warrants reexamination, since the Acheson Line itself, contrary to the conventional view, may be interpreted as part of a larger strategic design. As this issue requires separate discussion, it is not taken up here.

Statements by the command level have also confirmed this trend. General Brunson, Commander of U.S. Forces Korea, in his keynote address at the AUSA LANPAC Symposium in Hawaii in May 2025, likened South Korea to “a stationary aircraft carrier off the coast of China.” He stated that U.S. Forces Korea does not focus solely on repelling North Korea but also focuses on regional operations as part of the broader Indo-Pacific strategy. He further noted that, in order to overcome the ‘tyranny of distance,’ U.S. Forces Korea could play a significant role in supporting regional operations.[5] At the joint press conference following the 57th ROK–U.S. Security Consultative Meeting in November 2025, when asked whether U.S. Forces Korea could be used in operations in the Taiwan Strait or the East China Sea, U.S. Secretary of War Pete Hegseth responded to the effect that greater flexibility of U.S. Forces Korea to respond to other regional contingencies was necessary.[6]

This is no longer confined to the level of statements by senior officials. In 2026, there was even an incident in which U.S. fighter jets took off toward the Yellow Sea without prior coordination with South Korea and came to the brink of military friction with China. Forces launched from South Korean territory being committed to a third-party conflict irrespective of South Korea’s will, and South Korea being exposed to tension in the process, is the textbook definition of entrapment.

The institutionalization of ROK–U.S.–Japan trilateral security cooperation, advanced to an unprecedented degree by the previous South Korean administration, has also entrenched this structure. The Camp David commitment to consult, the Memorandum of Cooperation on the Trilateral Security Cooperation Framework signed in Tokyo in July 2024, and regularized combined exercises are all part of an effort to build institutions for trilateral cooperation that transcend changes in government.[7]

Institutionalization is not inherently an evil, and it may be pursued where necessary. Nevertheless, one cannot help but raise questions about the verification and consultation procedures for determining whether it truly reflects a judgment based on our own national interest. The elements described above shorten the causal pathway by which South Korea arrives at entrapment. Codification in strategic documents, statements by the command level, the actual diversion of forces, and institutional entrenchment have all already been observed. Each stage makes the next stage more probable. Moreover, this institutionalization raises the cost of exit. Once consultation procedures and combined plans are already in place, it becomes harder for South Korea to withdraw from a particular matter, and the room for entrapment in the extraterritorial operations demanded by the United States grows correspondingly larger.

The risk of entrapment is not confined to the military dimension. As will be discussed below, the United States seeks to compensate for the limits of its own defense industrial base with the manufacturing capacity of its allies, and within this configuration, if South Korea comes to serve exclusively as the United States’ ‘forward arsenal’ in a contingency, this too increases the probability that it will emerge as a priority target for China and Russia in the event of a U.S.–China conflict.

Costs can be imposed even prior to the stage of military conflict. China, which accounts for more than 20 percent of South Korea’s trade, holds the supply chains of key inputs such as core battery raw materials and rare earths, and the possibility that Russia will raise the level of its military-technology transfers to North Korea cannot be ruled out. Entrapment thus materializes not only as direct military attack but also in the form of the activation of economic and technological vulnerabilities.

In this way, the ‘risk of entrapment’ allows us to specify the structural circumstances, the actors, and the incentives, pathways, and timetable involved, all at once. We must now ask: can the ‘risk of abandonment’ likewise be specified in this way?

2. The Conditions of the Abandonment Risk: Geopolitical Value and Industrial Value

In the United States’ strategy toward China, South Korea’s value is broadly twofold. First, industrially, South Korea possesses capabilities that the United States has, in effect, lost—top-tier semiconductor production capacity, the capacity to build large commercial vessels and warships, and battery production capacity, among others. For example, it is openly acknowledged across the U.S. policy community that the assistance of South Korea’s shipbuilding industry is urgently needed to counter China’s maritime expansion. Naturally, the United States cannot discard an indispensable asset amid its strategic competition with China. Second, geopolitically, South Korea is a forward military base positioned right in front of China.

First, behind this industrial value lies a structural limitation in the U.S. defense industrial base. For example, amid the depletion of stockpiles of precision interceptor systems such as Patriot and THAAD due to support for Ukraine and involvement in the Middle East, a weakened manufacturing base means that a considerable lead time is required from the ordering to the delivery of a single missile. Even if the budget is increased, there is an absolute shortage of skilled labor to staff the manufacturing floor, compounded by dependence on China for the rare-earth supply chains essential to advanced weapons manufacturing. This is why diagnoses have emerged that a sense of crisis is prevalent within the U.S. Department of Defense over the prospect of losing a full-scale war not on the battlefield but due to ‘the limits of the supply chain.’ The NDS task of mobilizing the defense industrial base noted above, the recommendation to raise the defense budget to the mid-4-percent range of GDP, the mass production of low-cost weapons, and large-scale investment in nontraditional defense firms are all products of this sense of crisis.

Two reports released successively by CSIS on July 6 and 7, 2026 encapsulate the way the United States views South Korea in this context. The July 6 report, in the course of assessing progress in shifting the U.S. defense industrial base onto a wartime footing, treats industrial cooperation with allies as one of the key tasks and identifies Hanwha Ocean’s acquisition of the Philadelphia shipyard and its $5 billion investment in converting warship-construction facilities as a representative case of allied industrial cooperation.[8] So acute is U.S. demand for South Korean shipbuilding’s mass-production capacity that President Trump has even mentioned the possibility of circumventing the domestic law (the Burns–Tollefson Amendment) that has barred the U.S. Navy from procuring foreign-built vessels. In short, at present the United States has no incentive to give up South Korea’s industrial capabilities; rather, the dominant incentive is to secure and make use of them.

Fundamentally, the United States’ East Asian security architecture was designed with the Korean Peninsula as a geopolitical constant rather than a variable. Because this is a matter of structure rather than policy preference, it is not something that changes with a change of administration. It is conventionally understood that the cornerstone of U.S. East Asian strategy is Japan, not South Korea, and indeed Japan’s industrial base and network of bases have long sustained U.S. military operations across the entire Western Pacific. Yet building this structure required the premise that Japan itself must not become the front line. A front-line state is required to maintain a large standing army and a correspondingly extensive mobilization system. Had Japan been directly exposed, throughout the Cold War following the Second World War, to the confrontation line between two nuclear powers, the Yoshida Doctrine would have been fundamentally impossible. Only when the Korean Peninsula performed the role of that front line could Japan’s transformation into a rear base, and the broader East Asian security architecture, come into being.

This is not a retrospective interpretation. It was a structure repeatedly confirmed, in practice, at the level of policymakers in both the United States and Japan during the Cold War. The structure of the Korean Peninsula as forward base and buffer zone was the most fundamental condition of ROK–U.S.–Japan security cooperation. Japan’s position had been consistent since the Cabinet of Yoshida Shigeru: it could not cooperate with the United States unless U.S. forces were stationed on the Korean Peninsula to guarantee security there. This rested on the recognition that Japan must not stand at the front line of the U.S.–Soviet confrontation. The U.S. calculus was no different. In order to bind Japan to its own camp for the sake of containing the Soviet Union, the United States had to place the Korean Peninsula, rather than Japan, as the buffer zone at the front line.

In short, South Korea’s buffer role was, for both the United States and Japan, the very premise for the formation of their security cooperation. Article 4 of the November 1969 Sato–Nixon Joint Statement likewise codified that the security of the Republic of Korea was essential to Japan’s own security—the so-called ‘Korea clause.’ That this clause emerged as part of the Okinawa reversion negotiations is no accident. Even as it returned administrative rights over Okinawa, the United States had to preserve operational discretion over the largest base in the Far East, and this required Japan’s assurance that it would not restrict base usage in the event of a Korean Peninsula contingency. Indeed, in August 1972 Nixon warned Tanaka that if Japan made base usage difficult, the United States would have no choice but to withdraw its forces from Korea. This is a scene in which the fact that the U.S. commitment to the Korean Peninsula and access to Japanese bases each conditioned the other is directly confirmed.

The difference between the geopolitical conditions of the Cold War and the present must also be read precisely. During the Cold War, the Soviet Union was a continental power centered on land forces, and in effect had no ice-free port usable year-round in the Pacific theater. As a result, for Soviet power to be projected, it had, as a matter of military strategy, to pass through the land axis of the Korean Peninsula. In other words, the Korean Peninsula was both a buffer zone and the chokepoint blocking the sole land axis of Soviet expansion.

At the time, the principal adversary was the Soviet Union for the United States and North Korea for South Korea, so the objects of the threat were not identical. Yet the direction and character of the threat were the same: both the Soviet Union and North Korea posed a threat coming from the north and centered on ground forces. The strategic vectors for U.S. containment of the Soviet Union and South Korean containment of North Korea fully aligned. The ROK and the U.S. were thus able to avoid, throughout the Cold War, the alignment problems that arise from a divergence in primary threat perception.

Today’s adversary, however, is not the Soviet Union but China, and China does not share the Soviet Union’s geopolitical conditions. China possesses the world’s largest fleet and a powerful air force, along with its own maritime access routes through the East China Sea and the South China Sea. Unlike the Soviet Union, then, passage through the Korean Peninsula is not essential for the projection of Chinese power beyond its borders. The direction and character of the threat, which had been relatively convergent between Korea and the United States during the Cold War, has fundamentally changed, and this has become the root cause of today’s alignment problem.

That said, the fact that threat perceptions have come to diverge fundamentally, and the question of whether the strategic value of the Korean Peninsula has changed, belong to different levels of analysis. The alignment problem arises at the level of what the United States and South Korea regard as the primary threat, whereas the strategic value of the Korean Peninsula is not determined by that alone.

In fact, the ‘indispensability’ of the Korean Peninsula has not disappeared; only its context has changed. If, during the Cold War, its value lay in geographic denial—blocking the Soviet Union’s southward advance—today it is, on the one hand, a hub for the capabilities the United States mobilizes against the threat posed by China, and, on the other, an indispensable and vital asset for Japan as well, which must also reckon with the possibility of direct conflict with China.

Viewed in this context, U.S. abandonment of South Korea would not be a decision to give up South Korea alone; it would be, in a chain reaction, tantamount to a decision to give up Japan as well, since Japan cannot stand alone against China. In 2025, China’s defense expenditure was estimated by SIPRI at approximately $336 billion, compared with $62.2 billion for Japan—a gap of more than fivefold. The cause of this gap is not a matter of will but of economic scale: even if Japan doubled its spending ratio, the gap would remain large. To this must be added China’s nuclear forces, Japan’s non-nuclear status, and demographic structure, among other factors.

Consequently, if the United States were to abandon South Korea, only two probable options would remain for Japan: departure from the U.S.-led order, or independent nuclear armament. The former could take the form of an adjustment of relations with China or of equidistant neutrality; either way, it would collapse the U.S. position in the Western Pacific. The latter would unravel the nonproliferation order the United States has maintained for seventy years. Either way, the resulting cost would far outweigh whatever savings might be gained by abandoning South Korea. The cost of abandonment does not remain confined to a single object but spreads across the entire system.

This argument carries two implications. First, from the U.S. perspective, South Korea’s value is not merely bilateral but systemic. The preemptive-contribution thesis, confined largely to the bilateral dimension, asks what South Korea must give the United States in order to prevent abandonment. Yet the fundamental value of the Korean Peninsula derives, above all, from its geopolitical position within the regional security architecture. Second, this geopolitical value is not singular. Whereas the value of denial during the Cold War was physically fixed, the value of alliance cohesion today depends to a considerable degree on capability, perception, and the institutionalization built upon them.

This does not, however, justify the claim that ‘abandonment’ is a realistic risk. Even if the value of cohesion depends in part on perception, what ultimately underpins that perception is the position the Korean Peninsula occupies within the security architecture and its own capabilities. It is worth noting, though, that this value is not, relatively speaking, given automatically the way Cold War geography was; the perception of capability and the mutual trust founded upon it are matters that require active management. At the same time, if this perception is not grounded in structure and fact, and if institutionalization is built upon such an ungrounded perception, that institutionalization can itself become a source of risk. This is precisely the reason the ‘preemptive-contribution thesis’ cannot be sustained. As will be discussed below, the path to safeguarding South Korea’s value lies, if anything, closer to the opposite—namely, in the precise management of the relationship.

Beyond its non-transferable positional value, South Korea today also possesses substantial capability-based value. The problem is that industrial capability can be transferred, and it is precisely this transferability that gives rise to a genuine abandonment problem, since current U.S. policy fundamentally seeks both to bind South Korea and to transfer what South Korea possesses to itself. Moreover, this policy trend has shown consistency, not confined to a single administration. Through the 2022 CHIPS and Science Act and the Inflation Reduction Act, the Biden administration drew South Korea’s advanced manufacturing capacity into the United States by conditioning subsidies and market access. Under a structure in which failing to produce in North America meant exclusion from subsidies while receiving U.S. support constrained production in China, Samsung Electronics’ Taylor foundry and the large-scale U.S. investments of the three South Korean battery makers were, in effect, coerced. The Trump administration merely changed the form of pressure. In place of institutionalized conditions, it foregrounded the blunt threat of tariffs, extracting a $350 billion investment package and shipbuilding cooperation in the 2025 tariff negotiations. The means differed only in degree of sophistication—all were coercive, and the direction was singular. In terms of the transfer of South Korean capability to the United States, the two administrations were continuous with one another.

The CSIS report of July 7 shows how the discourse justifying this transfer is constructed. The report characterizes South Korea’s record-high 2025 export performance ($709.7 billion) as concealing more than it reveals, and, citing the sharp decline in the trade surplus with China (from $99.7 billion in 2018 to $19.6 billion in 2025) and a demographic structure marked by a total fertility rate of 0.75, warns that a country that remains within its existing export model will increasingly be marginalized and reduced to a substitutable entity. At the same time, while taking the form of acknowledging South Korea’s strengths across four domains—AI, semiconductors, and telecommunications; batteries and precision manufacturing; robotics and machine tools; and critical minerals and energy—it presses South Korea to fold these capabilities into a technology and security ecosystem led by the United States. This structure, in which a diagnosis of crisis and a prescription of integration are presented as a single package, suggests that the logic of the preemptive-contribution thesis is being formulated first within U.S. discourse, even ahead of domestic discourse in South Korea.

The target is clear: to build manufacturing capacity inside the United States using South Korean capital, technology, and labor. In 2026, this was followed by the enforcement of the special act on investment in the United States, the establishment of the ROK–U.S. Strategic Investment Corporation, and the opening of a ROK–U.S. Shipbuilding Cooperation Center in Washington. Joint research and development, technology exchange, and workforce training were formally specified as the content of this cooperation. In the semiconductor sector, large-scale investments by Samsung Electronics and SK Hynix are replicating advanced production capacity inside the United States.

Structurally, this can be described as a process in which the very elements that constituted South Korea’s ‘irreplaceability’—production facilities, process knowledge, and skilled labor—are being relocated to the United States with a time lag. The cooperative language of technology exchange and workforce training, Viewed from another perspective, the cooperative language signifies the transfer of tacit knowledge accumulated over decades. The problem is that, once this process is complete, semiconductors will be produced and warships built inside the United States even without South Korea.

Conceptualizing this process as ‘reshoring’ distorts its essence, because it is not a process of American firms bringing back production they had once sent abroad. Samsung Electronics’ investment in the United States is not the return of an American firm but a new greenfield investment by a Korean one. Hanwha’s acquisition of the Philly Shipyard is not the recovery of an asset but the acquisition of foreign capital. Above all, what is being transferred includes something the United States has never possessed—namely, process knowledge that South Korea accumulated independently. One cannot ‘bring back’ what one never had. Nor, for that matter, is ‘onshoring’ an appropriate term for the United States either; the term is misleading insofar as it highlights only the benefits to the United States while obscuring the corresponding losses for South Korea.

The current phase is, in effect, a process of ‘internalization’ of capability, whereby the United States relocates the manufacturing function it has relied on from an external supplier—South Korea—into its own territory and technological base, converting it into its own capacity. The completion of internalization means the disappearance of dependence on the external supplier, and that supplier’s obsolescence. Internalization originally refers to the process by which a firm absorbs an external function within its own boundaries, but what is at stake in this context is not ownership but control. Even facilities within the United States owned by Korean firms function, in effect, as assets under U.S. jurisdiction, through export controls, extraterritorial jurisdiction, and the conditioning of subsidies. The separation of ownership from location, and the effective control exercised by the state where the assets are located, is precisely what characterizes this internalization.

This is precisely the point this paper wishes to underscore decisively: the material condition of the abandonment risk lies not in the fact that a factory has appeared in the United States, but in the relational change by which South Korea’s irreplaceability is being eroded. The domestic expression ‘industrial hollowing-out’ captures the result but misses the cause. Hollowing-out ordinarily refers to a spontaneous outflow resulting from a loss of cost competitiveness. What is at issue here, however, is that this is the product of coercive negotiations that the United States has waged against an ally in order to maintain its hegemony and compete with China.

For abandonment along this pathway to become a realistic risk for South Korea, at least three conditions must be met from the U.S. perspective. First, a substitute for the industrial value must be secured. Second, the regional strategy must be reconfigured to the point of giving up the positional value, or that positional value must be securable by some other means. Third, the cost of ‘maintaining’ South Korea, taken as a whole, must come to exceed its residual value.

At present, none of these conditions has been met. Yet the first condition is being satisfied progressively as the internalization of U.S. industry advances. Once the domestic production base matures and the transfer of technology and personnel is complete, the industrial incentive to retain South Korea disappears, and the U.S. calculus converges on a purely geopolitical cost-benefit analysis. As shown above, however, the second condition is systemic in character and unlikely to be met within a short period.

The fulfillment of this first condition is already proceeding in an observable form. In the semiconductor sector, even as the United States depends on South Korean HBM (high-bandwidth memory), it is intensively nurturing its own firm, Micron, and once the next-generation HBM production base linked to Japanese government subsidies becomes fully operational, the irreplaceability of Korean memory will be eroded to that same extent. The structure of exports to the United States is also suggestive. There are analyses indicating that a substantial portion of the recent increase in exports to the United States consists of materials and equipment for building factories on American soil, which means that the current export boom may, paradoxically, be a process in which South Korea is building, at its own expense, the very U.S.-based production base that will one day replace its exports of finished goods. Add to this the transfer of smart automation processes and robotics, and a path opens for U.S. domestic production to become self-sufficient without any large-scale movement of skilled personnel. In short, internalization is not a hypothesis but a plainly ongoing trend.

Taken together, the probable phase awaiting South Korea after the internalization of U.S. industry is not the risk of total abandonment but partial instrumentalization. In other words, once industrial value disappears, what remains for the United States is not an ally to be protected but merely a position of value to be secured and utilized. U.S. forces would remain stationed, but their mission would no longer be the defense of South Korea. South Korea would become not a protected party but a base to be exploited. This is, in fact, more difficult to deal with than abandonment itself. That the United States pursues the internalization of industrial capability while simultaneously shifting the responsibility for deterring North Korea onto South Korea and strengthening strategic flexibility points precisely in this direction: retaining geopolitical access and absorbing our capabilities, all while passing on the strategic costs and burdens.

For South Korea, then, the risk of abandonment is not simply the symmetrical counterpart of entrapment but is closer to a consequence of the entrapment structure itself. If entrapment is a present problem, abandonment is a conditional future, one more likely to materialize in the form of ‘instrumentalization’ than as total abandonment. This can lead to a far more devastating outcome than simple abandonment, and policy grounded in the preemptive-contribution thesis itself creates or accelerates precisely this risk condition.

3. The Dual Vulnerability of the Preemptive-Contribution Thesis

The foregoing reconstruction warns of the risk posed by the ‘mechanical-balance thesis’ and the ‘preemptive-contribution thesis’ regarding entrapment and abandonment, both widely circulating in domestic discourse. The ‘preemptive-contribution thesis’ expresses concern over the risk of abandonment while failing to specify what that risk actually is. Beyond identifying the United States as the actor, it offers no concrete account of why, through what pathway, when, or to what degree. Preparation for an unspecified risk cannot have its effectiveness assessed. Because there is no criterion for judging what must be done, and how much, to be sufficient, this argument can always slide into yet another leap of a different kind—namely, the claim that it is ‘still not enough.’

Let us examine in more detail the more fundamental problem foreshadowed in Section II—namely, the internal incoherence of this argument. The preemptive-contribution thesis posits the United States in two contradictory ways at once. On the one hand, the United States is an actor capable of abandoning an ally according to its calculation of interests; otherwise, there would be no reason to worry about abandonment at all. On the other hand, the United States is simultaneously an actor bound by the degree of an ally’s contribution. Yet this ignores the fact that, for an actor moving purely by interest, contributions already received can become sunk costs. Above all, the United States does not consider its bilateral relationship with South Korea in isolation. If the United States has, for some reason, entered a stage of considering abandonment, it may weigh not what South Korea has already given but, within a larger framework, only what more it might be able to give going forward.

More fundamentally still, there are two further problems. First is the unfalsifiability of this claim: the preventive effect of contribution cannot be disproven, since one can always say that abandonment did not occur because of the contribution, or, if abandonment does occur, that the contribution was insufficient. Second, while this unscientific mode of reasoning is itself a problem, an even more serious problem lies in what actually constitutes the ‘effective contribution’ the preemptive-contribution thesis calls for. Most of this content consists of components deeply bound up with entrapment—cooperation in extraterritorial operations, the deepening of interoperability, and the like—since the very definition of strengthening interoperability is to maximize effectiveness against a common enemy. In the end, this amounts to arguing that a specific present entrapment must be accepted in order to forestall an unspecified future abandonment. And the (partial) agreement to strategic flexibility, or the strengthening of the combined system, that such an argument demands, is hard to retract even once the risk of entrapment visibly arrives. In the worst case, the price South Korea would pay is becoming a party to a great-power conflict.

The prescription of the preemptive-contribution thesis does not operate solely in the geopolitical realm. The demand to contribute reliably whenever the United States needs it was also the logic used to justify the $350 billion investment package. But if abandonment comes after ‘internalization,’ this very contribution betrays its own purpose: the more deeply South Korea contributes and the more it transfers, the more it is, in effect, building its own replacement on American soil at its own expense—rather than proving its own irreplaceability.

As a result, the preemptive-contribution thesis produces losses in two directions at once. Geopolitically, it leads to the acceptance of a specific present risk on the pretext of an unspecified future risk. Industrially, it increases the probability of hastening the very completion of the risk conditions that the ‘payment’ was meant to prevent. This prescription amounts, not to a balance between the two risks, but to something closer to their simultaneous maximization.

V. Anticipated Objections

Counterarguments to this view are certainly conceivable. The first is the objection that concerns over the credibility of extended deterrence are themselves evidence that the risk of abandonment is real—a genuine concern that has, in fact, continuously driven discussions of South Korea’s own nuclear armament.

Before responding to this objection, however, the character of extended deterrence must first be made clear. The entire authority over nuclear weapons—from planning and operation through to final use—belongs, under constitutional law, to the exclusive national command authority of the nuclear power. Even if an allied country participates in a consultative body and receives some shared information, that authority itself is not thereby divided. In the end, extended deterrence is structurally not a sharing of authority but a matter of ‘expectation,’ and the credibility of that expectation always carries an inherent unverifiability. The rhetoric of ‘strengthening the expectation relationship’ is therefore, logically, an exceedingly awkward construction.

Because anxiety over an unverifiable expectation does not resolve itself, it can, in practice, easily be exploited as a pretext for endless binding. This unverifiability, as will be seen below, defines the very character of the discourse surrounding extended deterrence. Acknowledging this inherent limitation is the starting point for the discussion that follows.

On this principle, the conventional view—that concerns over the credibility of extended deterrence demonstrate the reality of the abandonment risk—should, if anything, be inverted. Contrary to the conventional wisdom that treats the credibility problem of extended deterrence as evidence of abandonment, it is closer to yet another channel through which the United States binds South Korea. The more South Korea fears North Korea’s nuclear weapons, the deeper its dependence on the U.S. nuclear umbrella becomes, and that dependence in turn functions as a mechanism binding South Korea to U.S. regional strategy, or as leverage for extracting concessions.

As noted above, moreover, the North Korean nuclear threat that serves as the justification for this umbrella is not, in itself, purely exogenous. In the process of reorienting the ROK–U.S. alliance toward China since the end of the Cold War, the United States has ‘managed’ the North Korean threat while also using it as a justification for its China strategy and regional strategy. South Korea’s anxiety over extended deterrence is therefore not simply a ‘harbinger of abandonment’; rather, its character as a strategic mechanism binding South Korea within the entrapment structure is more essential. In other words, unless it reaches an extreme, the anxiety arising from extended deterrence functions not as a signal of abandonment but as a useful driver that efficiently deepens South Korea’s voluntary entrapment.

Compounding this is another problem: the lack of specification. The worst-case scenario contemplated by extended deterrence—North Korea’s use of nuclear weapons against the South—can only reasonably be posited in an extreme situation where the survival of the regime itself is at stake. Nevertheless, South Korea’s preemptive demands have concentrated not on a concrete judgment about the probability of such use, but on measures that can simply be labeled as ‘strengthening’ extended deterrence—the creation of consultative bodies, the regular deployment of strategic assets, and the like. Yet how much these measures actually ‘strengthen’ the credibility of extended deterrence remains unclear in itself; nowhere is there any basis for the claim that the frequency of consultative meetings or the visibility of asset deployments is directly proportional to the effectiveness of deterrence. In the end, this is a structure in which demands are repeated without any criterion for what must be obtained, or to what degree, for extended deterrence to be ‘strengthened’ to a sufficiently credible level. This absence of a criterion works to the advantage of the United States, since, in the absence of a criterion, the United States can present a bill unbounded in category or scale as the price for it.

To be sure, if the United States were actually to withdraw the nuclear umbrella, that would be unambiguous abandonment. That the provision of extended deterrence functions as a binding mechanism does not mean that its withdrawal would not be abandonment. In the current phase, however, the movements surrounding extended deterrence are not withdrawal but a surface-level ‘strengthening,’ and institutionalization at various levels under that pretext—the Washington Declaration being a clear example. In other words, the trend currently observed is not a sign of the umbrella being withdrawn, but a direction in which the institutional binding of various forms of entrapment deepens, mediated by measures such as nuclear consultative bodies and strategic asset deployments whose effectiveness is unverified—and unverifiable. The objection that cites concerns over extended deterrence as grounds for abandonment therefore overlooks, or evades, precisely how those very concerns substantively reinforce entrapment.

That said, discussing the credibility of extended deterrence as a policy problem must be distinguished from the rhetorical move of offsetting the discussion of entrapment by declaring that “since extended deterrence is uncertain, there is also a risk of abandonment.” This distinction, however, does not hold because the former satisfies the discipline this paper demands—specifying the actor, pathway, and timetable. It is, rather, the opposite. As shown above, the authority to use nuclear weapons is, in principle, never shared with an ally, and its credibility belongs, from the outset, to a domain of expectation that can be neither verified nor specified. In this respect, extended deterrence is an exceptional topic to which the discipline this paper demands cannot, in principle, be applied.

The problem lies precisely here: because specification is impossible, disproof is likewise impossible, and because disproof is impossible, extended deterrence can be invoked indefinitely as a generic term for the abandonment risk. The moment we classify extended deterrence as a matter belonging to the dimension of abandonment, we come to mistake an inherent conceptual limitation for a legitimate concern about a logic of binding that the United States itself has designed.

The second is the objection that reads the 2026 NDS’s transfer of primary responsibility for deterring North Korea onto South Korea as a sign of abandonment. It is more accurate, however, to read this as a redistribution of burden, since its purpose is not to abandon South Korea but to convert U.S. forces to missions directed at China. This is less abandonment than part of a reconfiguration of the entrapment structure, and even this fact—which appears to be a strong basis for the abandonment thesis—is, upon closer examination, written in the grammar of entrapment. In particular, there is the claim that the statements on reducing U.S. Forces Korea and the pressure over defense-cost sharing that emerged under the Trump administration amount to a risk of abandonment. That pressure, however, was not a herald of alliance dissolution but a negotiating lever: increased defense-cost sharing and the $350 billion investment package were recovered as its price. The outcome of the pressure that invoked abandonment was not departure but greater binding and deeper internalization.

VI. Conclusion

1. Disciplining the Discourse and Managing the Entrapment Risk

For a claim of risk—whether entrapment or abandonment—to be recognized as a contestable argument, it must specify the actor, the incentive, the pathway, and the timetable. A claim that merely invokes the opposite risk without specifying the actor and pathway is not argumentation but an evasion of argumentation. The reflexive, mechanical balancing of discourse gives rise, in itself, to a further risk: it conceals the actual imbalance in reality and leads to an uncritical treatment of an ongoing structural transformation. Mechanical balance does not mean neutrality. A rhetoric of balance deployed atop a tilted reality is merely a reflection of that tilt.

Because the two risks operate on different timescales and in different domains, the response to each must also be separated. Managing entrapment is a present-tense, geopolitical, and military task. Above all, South Korea must hold firmly to the principle that it cannot become the front line of U.S.–China conflict. As a corollary, prior consent and consultation regarding the exercise of strategic flexibility by U.S. Forces Korea must be made institutionally mandatory.

The issue of troop reductions needs to be addressed head-on here. Conventional wisdom has read any reduction in U.S. Forces Korea as a sign of abandonment and has made preventing it the top priority of security policy. Yet, as shown above, the character of today’s U.S. Forces Korea is changing. As long as its mission is being redefined from deterring North Korea to deterring China, maintaining troop levels means, in effect, maintaining the entrapment structure. Under this condition, a reduction could be an occasion for mitigating entrapment rather than a sign of abandonment. Accordingly, if the United States persists in extraterritorial deployment, it may be preferable for South Korea to accept some reduction of U.S. Forces Korea in order to lower the risk of being dragged into the middle of a great-power conflict. The assumption that a reduction automatically equals a defense gap holds only if U.S. Forces Korea remains dedicated to defending against North Korea. Conversely, the equation that reduction equals mitigated entrapment can likewise hold only once the mission composition of the remaining forces is specified, so efforts toward that specification must proceed in parallel.

The basis of this judgment lies in the asymmetry between the two risks. The threat from North Korea is a risk of relatively specified scale and character that the ROK–U.S. alliance has managed over a long period. Entrapment in great-power conflict, by contrast, carries a high probability of turning South Korea into the very first target of a dispute between great powers, and the damage could reach a level from which recovery cannot be assured. Taking on an irrecoverable risk for the sake of a manageable one is not a rational trade. In this sense, the risk of entrapment in great-power conflict—at least in terms of its potential scale and lethality—is a matter that should take priority over the North Korean threat. In the same vein, activities in sensitive airspace such as the Yellow Sea should be subject to mandatory information-sharing from the planning stage onward. In short, what is at issue is not the depth of the commitment but the conditions of the commitment.

2. Preparing for the Abandonment Risk and Policy Implications

Criticizing the abuse of the ‘risk of abandonment’ does not mean denying the possibility of abandonment. Rather, it means that its precise pathway must be identified. Preparing for abandonment is a task for the future, and its substance lies in managing the pace of the United States’ industrial ‘internalization.’ If abandonment comes on condition that internalization is complete, then genuine preparation does not mean ‘giving more right now’ but rather protecting South Korea’s ‘irreplaceability.’ South Korea must make effective the mechanisms for recovering principal and sharing returns on its investments in the United States, and must manage the pace and scope of technology and personnel transfer. The core upstream capabilities of design and process must, without exception, remain domestic. Efforts must also proceed in parallel to move into next-generation capabilities the United States cannot replace, thereby renewing irreplaceability itself. South Korea’s safety and prosperity lie in a direction diametrically opposed to a prescription that invokes the ‘risk of abandonment’ to preemptively demand what is, in effect, unconditional contribution.

The worst combination arises from conflating two tracks: abandoning the management of entrapment out of anxiety over abandonment, while accelerating internalization within the entrapment structure in an attempt to prove one’s ‘trustworthiness’ as an ally—thereby, in fact, enlarging the very possibility of the abandonment risk. Regrettably, this is the point that the ‘mechanical-balance thesis’ and the ‘preemptive-contribution thesis’ are, in practice, likely to arrive at. The word ‘balance’ carries a positive connotation in itself. But balance is something one states after reading the scale, not something one declares without even looking at what has been placed upon it. What is needed now is not a balance between entrapment and abandonment, but concrete explanatory power regarding each of the two concepts in its own right. ■ 

[1] Glenn H. Snyder, “The Security Dilemma in Alliance Politics,” World Politics, vol. 36, no. 4 (1984), pp. 461–495.

[2] Victor D. Cha, Alignment Despite Antagonism: The United States-Korea-Japan Security Triangle (Stanford, CA: Stanford University Press, 1999).

[3] Jaewoo Jun, “Underlying Dynamics for US-Japan-South Korea Trilateral Security Cooperation,” ISPI Commentary, Istituto per gli Studi di Politica Internazionale, 8 April 2024, https://www.ispionline.it/en/publication/underlying-dynamics-for-us-japan-south-korea-trilateral-security-cooperation-169126.

[4] U.S. Department of War, 2026 National Defense Strategy (Washington, D.C.: Department of War, 23 January 2026), https://media.defense.gov/2026/Jan/23/2003864773/-1/-1/0/2026-NATIONAL-DEFENSE-STRATEGY.PDF.

[5] Gen. Xavier T. Brunson, Commander, U.S. Forces Korea, Keynote Address at the Association of the U.S. Army (AUSA) Land Forces Pacific (LANPAC) Symposium, Honolulu, Hawaii, 15 May 2025 (local time).

[6] Joint press conference by Minister of National Defense Ahn Gyu-back and U.S. Secretary of War Pete Hegseth following the 57th ROK–U.S. Security Consultative Meeting (SCM), Ministry of National Defense, Yongsan, Seoul, 4 November 2025. Related coverage: “Hegseth: ‘Actively Supporting South Korea’s Construction of Nuclear Submarines’ … Strengthening ROK–U.S. Defense Industry Cooperation,” Hankook Ilbo, 4 November 2025.

[7] “The Spirit of Camp David: Joint Statement of Japan, the Republic of Korea, and the United States” and “Commitment to Consult,” Camp David, 18 August 2023; “Memorandum of Cooperation on the Trilateral Security Cooperation Framework,” Tokyo, 28 July 2024.

[8] Jerry McGinn and A.J. Dilts, Is the Industrial Base on a Wartime Footing? A Progress Report (Washington, D.C.: Center for Strategic and International Studies, 6 July 2026), https://www.csis.org/analysis/industrial-base-wartime-footing-progress-report.

■ Jaewoo JUN is Research Fellow at the Korea Institute for Defense Analyses (KIDA).

■ Translated and edited by Sangjun LEE, EAI Research Associate
    For inquiries: 02 2277 1683 (ext. 211) | [email protected]