The Korean War and the Post-war Prisoner of War Regime, 1945–1956

This paper examines the framework for the treatment of prisoners of war that emerged after 1945. It focuses on one of the key elements of the post-war prisoner of war (POW) regime, the role of neutral bodies – state authorities acting as ‘protecting powers’ or humanitarian agencies such as the International Committee of the Red Cross – in supervising the implementation of the 1949 POW convention. It examines the importance of neutral supervision for the POW regime, and shows how the events of the Korean War affected the willingness of states to comply with their obligations under the new convention.

contaminated international dialogue and distracted discussions over the future of humanitarianism and international law. The fact that a fresh set of conventions was agreed in 1949, including, for the first time, one covering civilians, was clearly 'a noteworthy event in this day and age', as one participant put it, but the outcome had been confusing and contradictory. 1 The Western powers found themselves pressing for elements in the civilian code which they had only recently condemned at the Nuremburg trials. 2 More worryingly, the arrival of atomic weapons cast doubt over whether conceptions of humanity, discrimination, and proportionality were still relevant to the modern battlefield. Contemporary commentators such as J.M. Spaight were therefore left 'puzzled' by the 'failure of the powers [. . .] to do anything to regulate those methods of war which, if continued, will make the [conventions'] humanitarian provisions [. . .] read like hypocritical nonsense'. 3 How the 1949 conventions came into being, and the compromises and deals that were struck on the way, has been thoroughly explored elsewhere, and it serves no point to repeat the exercise here. 4 Instead, this paper will focus on one of the key elements of the post-war 'POW regime': namely the role of neutral bodies -state authorities acting as 'protecting powers' or humanitarian agencies such as the International Committee of the Red Cross (ICRC) -in supervising the implementation of the 1949 POW convention. The role of oversight mechanisms was by no means new in 1949. By the time of the Great War, all major belligerents had come to rely on neutral 'protecting powers' to supplement the work of civilian aid agencies and meet the needs of their nationals in enemy hands. 5 The success of these initiatives led to protecting powers being accorded specific responsibilities in the POW convention of 1929. Although protecting powers were excluded from the Red Cross convention (dealing with sick and wounded on the battlefield) and states disagreed over the balance between the rights of the neutral inspectors and the security interests of the detaining powers, all of the major former belligerents agreed that protecting powers were needed to hold belligerents to their obligations under the POW convention. 6 'Organs of control' were, therefore, a core element in the modern 7 For a broader study of this theme: Joan Beaumont, 'Protecting Prisoners of War, 1939War, -1995 in POW regime: attitudes towards them after 1945 promise to tell us a great deal about contemporary expectations. 7 This paper will examine the post-war debates over the type of organs of control required in the revised POW convention, and then assess how these faired in their first real test, the Korean War of 1950-3. Before doing so, it will address what lessons the international community drew from its experiences in the Second World War and how these shaped future debates.

I
The general collapse of accepted standards of behaviour -by all sides -after 1937 was inevitably mirrored in the experience of the protecting powers and humanitarian agencies. Contrary to expectations, it was not so much the expansion of the concept of military necessity that stymied the work of neutral inspectors as the political and ideological interests of the belligerents themselves. 8 Berlin's dismissal of Swedish efforts to protect Polish prisoners after 1939, on the grounds that Poland no longer existed as a state, was an early harbinger of a policy that was soon extended to most corners of Hitler's New Order. 9 An exception was made for Vichy France, where Berlin grudgingly allowed the Vichy minister, George Scapini, to represent the interests of the 1.8 million French prisoners remaining in Germany after the armistice. But this was prompted by political considerations -a desire to shore up support for the Vichy regime among French prisoners and workmen in the Reich -not humanitarian, and Scapini's authority fell short of that stipulated under the convention. 10 Moscow's attempt to enlist Swedish protection after July 1941 was rebuffed, with disastrous consequences for POW mortality rates on both sides of the eastern front. 11  duration of the war. 12 Taking the war as a whole, some 70 per cent of prisoners lived beyond the reach of third-party protection. 13 Yet if, in global terms, the record was a bleak one, the fact remains that in certain circumstances even brutal states such as the Third Reich would moderate their behaviour out of a concern for its impact on neutral inspection reports. The majority of British protests lodged by American diplomats in Berlin before December 1941 received satisfactory replies. Washington even came close to convincing the two belligerents to hold face-to-face talks over POW issues in the spring of 1941. The Swiss, though lacking Washington's political muscle, capitalized on their position in representing both parties to the conflict, and helped head off several potential crises involving the treatment of POWs. 14 Swiss diplomats were generally held in high regard, despite grumbles over the wording of camp inspection reports, or Berne's occasional foray into areas deemed beyond its authority. The summary execution of the Allied 'great escapers' in early 1944 naturally underlined the fragility of the POW regime and the speed with which Hitler's regime could slip into barbarism. But the fact that this brutality was the exception rather than the rule had much to do with the presence of Swiss diplomats and ICRC delegates, holding the Nazi regime to its obligations under the POW convention, even in its final death throws. 15 The Second World war thus provided a mixed legacy for third-party involvement in POW affairs. Protecting powers, the principal 'organs of control', remained widely recognized as the institution of choice, wherever possible. Their greatest supporter was probably the UK government, whose long experience of US protection, dating back to the Boer War, had made it particularly partial to state-based protection. Even states denied direct formal assistance of protecting powers emerged from the war convinced of their utility. The one notable exception to this was the Soviet Union. Not only had Moscow fought the war entirely without the services of a protecting power, but it had refused to acknowledge Switzerland's claims to neutrality, on account of the real and perceived level of Swiss collaboration with the Nazi regime. Tentative Swiss attempts to re-establish diplomatic ties in late 1944 were publicly rebuffed, triggering the resignation of the Swiss foreign minister shortly afterwards; the following year Berne had to seek British help to secure the repatriation of its Berlin embassy staff from Soviet custody.
The war's legacy for humanitarian organizations such as the ICRC was even more problematic. Although the ICRC had been formally recognized in the 1929 POW convention, its role was limited to dispensing 'humanitarian services'. Consequently, almost every aspect of its work for POWs depended on it first securing prior agreement from the governments concerned. True, the ICRC's delegates had furnished valuable camp inspection reports, overseen the repatriation of sick and wounded POWs, and arranged for the delivery of relief parcels. It had also interceded in support of individuals -civilian internees and camp inmates -who lay beyond the reach of the protecting powers. Its efforts to bring succour to prisoners in the Far East probably exceeded those of the protecting powers. 16 But the war left deep scars on the ICRC's reputation. Its failure to speak out against the Holocaust, the intimacy of its relations with the Reich's leadership, its delegates' naivety in working with the Axis occupation forces in distributing relief supplies, and their embroilment in Germany's exploitation of the Katyn incident in 1943 left the institution open to charges of collaboration that were difficult to shrug off. 17 Geneva's standing with the Western governments was likewise marred by its outspoken criticism of Allied bombing of Axis cities, blockade of food supplies, and denial of POW status from surrendered enemy forces in 1945. 18 Finally, Geneva's success in providing relief parcels was slowly eclipsed, especially in American eyes, by the role assumed by the national Red Cross societies -which produced the majority of the parcels -and the work undertaken at the international level by the United Nations Relief and Rehabilitation Administration (UNRRA). 19 Such work appeared to signal a shift in the focus of international relief activity from Geneva to New York and into the hands of the national societies, and stymie the ICRC's hopes of staking a claim over this area of activity.

II
The task of revising the 1929 conventions, and negotiating a code for the protection of civilians, was carried out over a four-year period, culminating in a diplomatic conference in Geneva between 21 April and 12 August 1949. In between, the matter was debated at a number of Red supervisory mechanisms were agreed with little ado. 20 The priority given to official protecting powers was accepted early on. 'Common article 8', found in all of the 1949 conventions, stated that they 'shall be applied with the cooperation and under the scrutiny of the Protecting Powers whose duty it is to safeguard the interests of the Parties to the Conflict' (emphasis added). A consensus also emerged over the need to block recourse to 'Scapini' missions. 'Common article 10' prohibited the substitution of protecting powers with inferior agencies, based on 'special agreements between the Powers one of which is restricted, even temporarily, in its freedom to negotiate with the other Power or its allies by reason of military events, more particularly where the whole, or a substantial part, of the territory of the said Power is occupied'. So widespread was the belief in the importance of proper third-party assistance that from the earliest drafts, agreement was reached on establishing what might be called a 'hierarchy of protection'. At the top was the traditional protecting power, operating with the consent of all parties, and with responsibilities eventually itemized in 34 of the 143 articles of the 1949 POW convention. If this was not possible, for whatever reason, the detaining power was to request a second neutral state or 'an organization which offers all guarantees of impartiality and efficiency' to act as a 'substitute' protecting power. If all else failed, detaining powers were to 'request' or 'accept' the services of 'a humanitarian organization such as the ICRC' which would 'assume the humanitarian functions performed by protecting powers'. Though inelegantly expressed, the 'hierarchy of protection' offered a chance of tackling the problems uncovered in the last war. Never before had the safety net of prisoner protection and third-party surveillance of the conventions been so tightly woven. 21 This was, however, the limit of the post-war consensus, and divisions over some of the outstanding issues proved so intractable that they came close to jeopardizing the entire exercise. As in 1929, one of the central areas of debate concerned the scope of protecting powers' supervisory authority and the extent to which they could encroach on state prerogatives. In keeping with the attitude shown after 1941, Moscow was inclined to limit the scope of the protecting power mandates. An amendment tabled at the diplomatic conference called for an injunction against protecting powers 'infring[ing] the sovereignty of the State' or acting in 'opposition to State security or military requirements'. The Western powers, by contrast, whose servicemen had generally benefited from third-party protection, were reluctant to see neutral inspectors unduly impeded. 22 23 'Report on Geneva Conventions of 1949: Joint Committee', p. 5, NARA, RG59, entry 5210, box 9. 24 The additional clause was included in article 8 of the first and second conventions: text avail- When put to a vote, the majority decided that Moscow's concerns over state sovereignty were misplaced. After all, the protecting powers' authority was restricted to only those actions required to ensure that parties lived up to their commitments, and in ratifying the convention, states freely entered into accepting these commitments. 23 A compromise solution, sufficient to satisfy Soviet anxieties, was eventually agreed, whereby protecting powers were, at all times, to 'take account of the imperative necessities of security of the State' and avoid 'exceed[ing] their mission under the . . . Convention'. An additional clause was added to the sick and wounded conventions -though not the POW and civilian codes -allowing states to impose 'exceptional and temporary' restrictions on the protecting power if 'rendered necessary by imperative military necessities'. 24 Given the breadth of their duties, the possibility of protecting powers falling foul of these restrictions was perhaps slim. But the latitude left for detaining powers to define 'imperative necessity' naturally disappointed those who had hoped to have closed this loophole at the Stockholm conference the previous year. Still, the fact that some accommodation had to be found on this issue was hardly new: the balance between humanitarianism and military necessity has been a recurrent theme since the 1860s, and in 1949 the Soviet Union merely represented a position that had been held, twenty years earlier, by the Japanese and Romanians. It was, in this sense, a modern version of a familiar refrain. 25 What were new in the post-war debates were questions over the identity and scope of those institutions included in the convention's 'hierarchy of protection'. The first of these concerned the kind of institution deemed capable in extremis of acting in lieu of protecting powers. Discussion on this issue partly boiled down to a question of what role the ICRC should occupy as an 'organ of control'. At first sight, this might seem rather odd. After all, the ICRC had begun the codification of customary practice in 1864, and had been the principal architect and guardian of international humanitarian law ever since. It was the ICRC that had first called for a revision of the conventions in 1945 and shepherded successive drafts through subsequent meetings. In a conference of national Red Cross societies held in the summer of 1946, most delegations agreed that specific areas of competency for the ICRC had to be spelt out in Critics of the ICRC were, however, thick on the ground after 1945, and there was no certainty that it would be able to retain its position, as of right, in the new conventions. The institution had already found itself under pressure in 1945 when the Western powers resisted ICRC efforts to provide succour to Axis servicemen who had been re-categorized as 'surrendered enemy personnel' and denied protection under the POW convention. 27 But its most outspoken critic was the Soviet government. Moscow's opposition was so trenchant that it absented itself from all meetings before 1949 on the grounds that the ICRC's lamentable wartime conduct disqualified it from hosting such events. Its denunciation of the ICRC three days before the opening of the Red Cross conference in Stockholm was particularly caustic and prompted the organizers to reschedule the first day's proceedings to allow the ICRC president, Paul Ruegger, to respond to Soviet charges. Ruegger's speech and the three-volume report on the ICRC's wartime operations published for the occasion were well received, but few present could ignore the fact that Soviet opposition to the committee was deep-seated and likely to shape its subsequent view of the conventions. 28 In reality, Soviet criticism of the committee was important not so much in affecting the outcome of any particular vote -the Eastern bloc vote was too small to muster a majority -but rather in influencing behaviour of other delegations. Geoffrey Best has argued that French delegations were particularly sensitive to Soviet wishes, in large part because of the waxing electoral fortunes of the communist party at home. He suggests that the French proposal for a 'high international committee' to oversee the conventions was essentially driven by the desire to keep Moscow happy and provide an alternative mechanism to the ICRC. 29 The 'high international committee' comprising 30 distinguished lawyers, luminaries, and international 'elder statesmen' was to be convened at the start of a conflict to 'supervise the application and ensure respect for the convention'. In fact, the Cahen-Salvador proposal, named after the lead French delegate, reflected a trend in French thinking that extended back before the war -the French had proposed a similar body at the 1929 conference -and was designed to deal with situations in which there were no neutral states to act as protecting powers, and no neutral Switzerland to provide a base for the ICRC. First raised in 1947, the proposal for a high international committee had particular traction at the diplomatic conference where 'extreme pressure' from the French had secured the presidency of one of the commissions for Cahen-Salvador. 30  Cahen-Salvador's proposal was, however, only one of a number of initiatives that emerged over the period that threatened to undercut the ICRC's standing in the Red Cross movement and, by extension, its place in the new conventions. By the eve of the Geneva Conference, the ICRC faced challenges from two different directions. On the one hand, the Eastern bloc societies, suspicious of the committee's Western, 'bourgeois' leanings, insisted that its membership be 'internationalized' through the introduction of members from across both sides of the Iron Curtain. On the other hand, some Western societies were anxious to bridge the gap between the international committee and the umbrella organization for the national societies, the League of Red Cross Societies. The Swedish, American, French, and Belgian societies all lent their voice to calls for a 'standing commission', chaired by the charismatic vice president of the Swedish Red Cross, Count Folke Bernadotte, to sit between the ICRC and the League, and coordinate their respective activities. 32 In both cases, the ICRC was confronted with a challenge to its leadership of the Red Cross family and its authority as the principal supervisory institution for the movement in times of war.
Over time, the danger of a fissure opening up along Cold War lines gradually helped rein in some of the Western societies' enthusiasm for reform. But while pressure on the ICRC ebbed, lingering doubts over the suitability of its statutes, form, and membership inevitably coloured thinking over the wisdom of leaving the ICRC as the principal default 'organ of control' in the revised conventions. 33 The committee's very 'Swissness', hitherto a guarantee for its neutrality and independence, was much less valued in a world governed by a United Nations charter that cast doubt over the viability of neutrality in international politics, or one teetering on the brink of another global war. 34  outspoken criticism of the committee before the Stockholm conference inevitably threw such concerns into sharp relief. As Harold Starr, head of the American Red Cross delegation in 1948 bluntly put it: 'as the ICRC was no longer persona grata with the USSR and its neighbours, it was not opportune, if one desires to have these states adhere to the new conventions, to allow the ICRC to figure too prominently'. 35 In line with this thinking, Starr tabled a series of proposals at Stockholm which effectively airbrushed the committee out of the revised conventions, and replaced it with a 'competent international body', a term that was sufficiently vague to leave open the possibility of a variety of ad hoc institutions, whether drawn from the international Red Cross movement or not. The American amendments not only covered prisoners' external relations and the 'organs of control', but included the ICRC's role in transportation and relief activities and its place in convening the 'central POW information agency'. It took all the tact and back-room diplomacy the ICRC delegates could muster to overturn the American initiative. Part of the problem lay in the Americans' conflation of two different levels in the 'hierarchy of protection': namely the 'substitute protecting powers' (neutral states that stepped in when the original protecting powers were unable to function) and the 'quasi protecting powers' ('a humanitarian organization such as the ICRC', which would 'assume the humanitarian functions performed by protecting powers' where statebased protection was not possible). Naturally, any thought that the ICRC might act as a 'substitute' protecting power, rather than a 'quasi' (humanitarian) one entailed an accretion of powers to Geneva that would elevate, rather than diminish, its status in the international Red Cross movement. The fact the ICRC itself was averse to shouldering political responsibilities under the convention -lest they interfere with its traditional humanitarian activities -had little appreciable effect on American attitudes. According to ICRC observers, it was only the 'very dignified' intervention of the South African Red Cross delegate that swung the debate in the ICRC's favour. Even then, the ICRC only remained in the Stockholm draft as an 'organ of control' by a vote of 10 to 8. 36 Starr's anxiety over the danger of naming the ICRC as a 'quasi' protecting power was not entirely without foundation. As we will see, Moscow did nothing to help the ICRC in North Korea after June 1950; thirty years later, it barred the ICRC from operating inside Afghanistan. In 1949, however, Soviet interests lay elsewhere, and much to Geneva's relief, Moscow chose not to contest the ICRC's position in the conventions. Instead, and rather unexpectedly, it turned its attention to the issue of the 'substitute' protecting power. The crux of the Soviet position lay in their opposition to the idea that the right to choose a substitute lay with the detaining power, as this opened the possibility of interned Soviet civilians and servicemen being 'protected' by a state not of Moscow's choosing. Soviet concerns were principally directed towards the fate of their own personnel, but the fact that German and Japanese POWs were still detained in Soviet camps from the last war was hardly lost on the other delegates. 37 Moscow's opposition to the 'substitute protecting power' clause gave rise to one of only two 'reservations' tabled at the signing ceremony, and was to have profound implications for Western confidence in the POW convention over the succeeding decades. The Soviets refused to 'recognize the validity of requests by the Detaining Power to a neutral State or to a humanitarian organization, to undertake the functions performed by a Protecting Power, unless the consent of the Government of the country of which the protected person [POWs] are nationals has been obtained'. 38 When set alongside Moscow's other reservation -its insistence on the right to try prisoners for war crimes (article 85) -Soviet objections threatened to frustrate the humanitarian ambitions of the conventions. Instead of taking the convention's criteria of 'impartiality' and 'humanitarianism' as good faith guarantees to be lived up to, the Soviets sought to secure compliance through the arbitrary power to withhold consent. In reality, it meant that in a Soviet-US conflict, Washington would be obliged to employ a protecting power or humanitarian organization in order to meet its legal obligations, even though Moscow, under its reservation, could withhold recognition. The Soviets, for their part, could insist on drawing on the services of a communist aid society or a satellite state which might, as a US State Department memorandum put it, 'purposely have been kept apart from the hostilities to qualify as a neutral, to "protect" the interests of US POWs and detainees in Russian hands'. 39 Western efforts to deal with the Soviet challenge were unsuccessful. Any objection to the Soviet reservations ran the danger of either delaying Soviet ratification or provoking the Soviets into tabling objections to the various reservations made by the Western powers to articles in the civilian convention. For all the progress made in bolstering the supervisory regime, the new 'hierarchy of protection' unveiled in 1949 was therefore noticeably weaker than many scholars initially assumed. 41 True, the ICRC's status had been elevated: its right to intervene on behalf of POWs was affirmed in 13 separate articles, and in explicitly defining its status as 'an impartial humanitarian body' (articles 3 and 9), impartiality was acknowledged as a 'constitutive quality' of the ICRC (despite its remaining a Swiss institution) rather than merely a principle observed in the course of its operations. 42 But the defeat of Cahen-Salvador's and Starr's proposals scarcely amounted to a ringing endorsement of the Geneva committee. Indeed, the back-room tussles only underscored the lack of confidence in the institution as currently conceived. Equally troubling was the position of the protecting power. Though their remit had been widened and clarified, the Soviet reservation cast doubt over whether this vital element in the POW regime could remain insulated from the corrosive effects of Cold War politics.

III
Debate over the robustness of the 1949 conventions' oversight mechanisms was overtaken by the outbreak of fighting in Korea in June 1950. Although the 1949 conventions were not in force -only Switzerland had ratified them by this date, and many states suspended discussions for the duration of hostilities 43 -the warring parties all proclaimed their readiness to adhere to the 'humanitarian principles' underpinning the conventions: the US State Department in a letter to the ICRC on 3 July, and Pyongyang 10 days later, on 13 July. Today the war is principally remembered for the politicization of the POW issue, with prisoners subjected to political indoctrination and scripted into propaganda films and broadcasts. 44 Korean forces on both sides routinely ill-treated their captives, and while the UN forces were frequently shocked by the behaviour of their allies, they too had to resort to tanks and tear gas to restore order among rioting communist prisoners at the UN detention facilities at Koje-do. 45 Most famously, progress in the armistice talks, which began in early July 1951, was repeatedly stalled over the issue of prisoners of war, and the desire of the Western governments to secure agreement on the principle of 'voluntary repatriation' at the close of hostilities, as a 'substitute for victory' over communist forces on the battlefield. 46 The war was also, however, an important test for the new norms of POW protection. As will be seen, the test was failed in nearly every respect, but the practices that emerged in their wake were to have a profound impact on the shape of international humanitarianism for the next half-century. Interestingly, neither side moved to employ a protecting power at the outset of hostilities. While Pyongyang's attitude towards the issue can be deduced from its subsequent refusal to agree to external interference over its POW policies, the West's hesitation initially stemmed from a determination to avoid any action likely to bestow political legitimacy on the North Korean regime. As the UN intervention in Korea was deemed to be a 'police action', not a war, the West did not consider the new conventions legally binding and therefore limited their observance to only 'such detailed provisions as [were] appropriate to the situation'. 47 The protecting power provisions were noticeably excluded from these provisions. 48 The arrival of Chinese 'volunteers' in Korea in November 1950, and the headlong retreat of UN forces that ensued, prompted a rethink in Western capitals over the question of POWs. The options were, however, hardly appealing. As no Western state had recognized the North Korean regime, the coalition found itself facing precisely the situation envisaged in Moscow's reservation to the 1949 conventions. With only Eastern bloc embassies in Pyongyang, the choice of protecting power by early 1951 came down to either the Soviet Union or Czechoslovakia, neither of which was particularly attractive to the Western governments. 49 Action was suspended over the first half of the year in the hope that ICRC delegates would gain entry into North Korea. It was only in the autumn, with the front stabilized and armistice talks opened at Kaesong, that the UN secretary general approached the Chinese and North Korean governments, asking them to appoint a substitute protecting power or accept the services of the ICRC as a 'quasi' protecting power. 50 Western governments were rightly, as it turned out, sceptical over the chances of success, and backed the initiative primarily to strengthen their negotiating position at Kaesong. 51 Statements by Chou En-lai the following July, hinting at Peking's willingness to recognize the 1949 conventions and the 1925 Geneva gas protocol, prompted the West to appeal for Soviet assistance to sway opinion in Peking and Pyongyang, but once again the initiative came to nought. Chou En-lai's trumpeting of the 1925 protocol was too convenient for communist propaganda over the West's alleged use of chemical weapons in Korea to be taken seriously, while his flattering references to Soviet reservations in the 1949 conventions appeared to suggest that he was intent on trying UN airmen for war crimes, while denying the ICRC access to the conflict as a substitute protecting power. In both cases it was difficult to see how the interests of Western POWs could be advanced by Peking's sudden affection for the humanitarian codes. 52 An equally instrumentalist approach can be detected in attitudes towards the ICRC. Geneva offered its services to both parties at the start of hostilities, but was never formally invited to act as a 'quasi' protecting power. Consequently, just as in the Second World War, any initiatives on behalf of the prisoners depended on securing the consent of the belligerent governments. In the North, this was not forthcoming. Of the 200 communications sent by Geneva to Pyongyang between June 1950 and December 1951, only two elicited any reply. Paul Ruegger's visit to Peking in the spring of 1951 provided a brief glimmer of hope, but by the summer it was clear that the Chinese Red Cross would not press the ICRC's case or arrange for the distribution of medical and relief supplies. Similar approaches via Moscow proved equally forlorn. Peking denied responsibility for the Chinese 'volunteers' and remained unmoved by Geneva's release of names of communist prisoners held by the UN. 53 The ICRC was thus forced to watch on impassively as communist 'peace movements' and their Western sympathizers were given free access to 'report' on the communist camps, while notification of the identity of individual inmates filtered back to the West through staged radio broadcasts by 're-educated' prisoners. 54 South of the 38th parallel, the situation was different. The fact that Geneva had nothing to show for its efforts in the North did not materially affect the scope or effectiveness of its advocacy and inspection work on behalf of POWs in the South. In truth, despite repeatedly complaining about the inconvenience and unreality of applying the POW convention unilaterally, the UN command could do little but accept the ICRC's presence. To do otherwise would damage its international standing, add fuel to allegations over its use of germ warfare agents, and banish any chance of delivering aid to its prisoners in communist hands. Indeed, the military authorities came to value the ICRC's expertise, and occasionally appealed to its delegates to assist in areas that were not explicitly 'humanitarian'. 55 The lack of any sustained effort to adopt or adapt the convention's 'hierarchy of protection' should prompt us to question the depth of Western commitment to these mechanisms. Clearly the existence of Moscow's reservations -and Chou En-lai's apparent sympathy for them -cast doubts in Western minds over the whole question of POW protection. The politicization of the POW issue during the war inevitably fanned Western suspicions over communist intentions, and strengthened the conviction that the 1949 conventions were 'entirely inadequate for and inapplicable to' a conflict with the communist bloc. 56 The re-emergence of neutrality in the early 1950s did, it is true, raise hopes that 'neutral' and 'impartial' protecting powers might yet survive the onset of the Cold War. The four-power Neutral Nations Supervisory Commission for Korea, though framing the European neutrals along ideological lines, implied that predictions over the imminent demise of neutrality were premature. Moscow's flirtation with a 'neutral' Germany in the early 1950s, the signing of the Austrian State treaty in 1955, and the gradual emergence of 'Third World' neutralism under Indian leadership inevitably strengthened this belief, and goes some way to explaining the West's renewed confidence in the Geneva Conventions by the late 1950s.
Yet, the real lesson taken from the Korean War was the belief that, under Cold War conditions, it paid to operate outside the protecting power framework. This conclusion obviously jarred with the concerns expressed by the UN coalition over the fate of its men in communist hands. But there is little doubt that the experience of confronting communism in the developing world convinced many in the UN command that there was some advantage in curtailing the work of the conventions' 'organs of control'. The rarefied academic discussions over whether the Korean War was a unique case under international law soon gave way to the realization that the conflict was ultimately a harbinger of things to come. The benefit of withholding political recognition from one's adversarywhether in the form of a puppet communist regime, or an anti-colonial insurgency -had obvious appeal.
The same was true for the informal character of the ICRC's mandate. Unlike the 1929 codes, protecting power arrangements were common to all four of the 1949 conventions. The implications of this were not initially appreciated by the UN command. In August 1950 General MacArthur confidently announced that his forces would be guided by all four conventions, but the conduct of fighting, not least the appalling treatment of civilians by South Korean troops, soon convinced him to limit his responsibility to the POW convention alone, and leave Seoul to answer questions on the civilian, maritime and Red Cross conventions. As nothing was done to advertise this change, it was not until the ICRC raised the issue in February 1951 that the US government became aware of the situation, and not until the middle of that year that the UN command confirmed that it was 'not extend[ing] recognition of the ICRC with respect to all four conventions because of what its delegates might discover and report'. 57 Despite efforts to have the UN command live up to its public pronouncements, the Red Cross, maritime, and civilian conventions were never applied to the Korean conflict. Indeed, the UN command attempted to walk back from its commitment to the POW convention, by trying to downgrade the status of ICRC delegates -from 'accredited' to 'invited' -and insisting that their camp reports were submitted for comment first, before being forwarded to the UN headquarters in New York. 58 The ICRC took this in its stride, and chose not to make an issue out of the UN's difficulties at Koje-do or elsewhere. America's 'inexperience in managing POW camps', Paul Ruegger wryly remarked in conversation with the British foreign secretary in January 1953, 'sometimes led them astray', but there was little doubt that the UN command was 'doing [its] best to carry out the various Geneva conventions'. 59 As for the committee's part in supervising the application of the conventions, the events in Korea clearly had a chastening effect on Swiss thinking. By mid-1951 the committee was increasingly warning its delegates to avoid encroaching into areas that properly lay within the jurisdiction of the protecting power. 60 By that winter, as the conflict moved from the battlefield to the court of world public opinion, the committee issued a note on 'alleged violations of international law' in which it sought to clarify its remit. While admitting to exercising 'a certain degree of supervision' over the conventions, it insisted that 'its essential task' lay in carrying on 'the humanitarian work entrusted to it by the Conventions and devolving on it under its Statutes and those of the International Red Cross. Instead of passing judgement, the Red Cross must bring help. Before theorizing about principles, it must translate them into action.' 61 Such statements drew approving nods in Western capitals.
While, then, the West might sympathize, philosophically, with both the principles and ambitions of the new conventions, the practical experience of applying them within the new Cold War context made them appear much less attractive. Far better to pare down the scope of the 'organs of control' and rely on the traditional humanitarian functions of the ICRC than expose Western conduct to the scrutiny of a protecting power, drawing on the full authority of the 1949 conventions. This was just as true for confrontations with the Soviet bloc -where captured enemy personnel would be regarded 'more as political refugees than as prisoners of war', 62 and where their adversaries questioned the whole premise of 'substitute' or 'quasi' protecting powers -as for conflicts with insurgent movements in the European colonial empires or their postcolonial client states.
Though clearly troubled by the increasing politicization of prisoner treatment in Korea, Western governments could afford themselves a guarded sense of optimism over the shape of the POW regime. The major sticking points during the armistice negotiations -the right of prisoners to seek asylum and the right of parties to use unified command structures in dealing with POW affairs -were both resolved in the West's favour. 63 Moreover, North Korea's challenge to the Geneva 'norms' had, at least in part, been overcome. The high death rates among UN prisoners were shocking, but no worse than those experienced by Allied forces in the Far East and Pacific theatres during the last war. Political indoctrination, which had led 21 Americans and 1 Briton to remain in China after the end of hostilities, clearly rattled Western militaries. But while an attentive protecting power might, conceivably, have helped mitigate the threat, the West could take comfort from the fact that the matter could be tackled through improved training in 'resistance to interrogation' and 'conduct after capture'. In all events, the majority of states that joined the UN coalition felt confident enough to ratify the POW convention within a few years of the end of hostilities.
This should not, however, lead us to belittle the impact of the Korean War on the development of the POW regime. The apparent self-evident importance of establishing a robust system of oversight for the new conventions did not long survive the conflict. Britain, France, and Egypt employed protecting powers during the Suez crisis in 1956, but on only three occasions since then have states lived up to their treaty obligations and accredited neutral states as protecting powers, and on each occasion 'exceptional' circumstances were deemed to apply. 64 It is ironic that just as the 'organs of control' in international humanitarian law reached their apogee, the international community chose to dispense with their services. The tireless work of the ICRC has ensured that we have not seen the complete return to barbarism that many had feared. But in insisting that it would 'bring help', not 'pass judgement', in its work, the ICRC was not merely reciting its founding mantra; it was, instead, retreating from a tradition of advocacy and debate that it had propagated during the Great War, but had increasingly discarded since the early 1940s. No doubt, in the politically charged world of the Cold War, this shift was wise, but it has left the conventions largely under-supervised and allowed belligerents to ignore their treaty obligations without fear of serious censure. Just as in Korea, 'reciprocity' has had little appreciable effect on belligerent behaviour. Indeed, one of the key characteristics of armed conflict since 1945 has been the denial of reciprocity, whether in the 'bush fire' wars of decolonization, the interstate conflicts along the political fault lines in South Asia or the Gulf, or the more recent episodes in the 'global war on terror', with tragic results for those detained by their enemies. 65 Moreover, it was not merely in the contraction of the humanitarian space where the absence of a protecting power was felt. The protecting powers were equipped and empowered to hold belligerents to account across a range of diplomatic and humanitarian concerns. It is surely telling that in deliberating over what was learnt from the Korean War, the ICRC was firmly of the view that had protecting powers been present, political tensions between the belligerents would have been eased, and the armistice negotiations brought to a speedier conclusion. It was for this reason that the committee spoke out in favour of protecting powers in international discussions over the course of the 1950s and 1960s. 66 As one senior member of the committee, Frédéric Siordet, remarked in July 1954, 'if protecting powers are not able to accomplish their tasks in good faith, then the entire application of the conventions is compromised'. 67 The history of the last sixty years sadly bears out his depressing prognosis.