Intelligence and the war in Bosnia 1992-1995: The role of the intelligence and security services–(UPDATED–3/14/2017)


CIA’s Muslim Brigades–Bosnia

Intelligence and the war in Bosnia 1992-1995:

The role of the intelligence and security services

Intelligence and the war in Bosnia 1992 – 1995 

NIOD Report on Srebrenica


Summary for the press

The following is an authorized summary of the conclusions from the epilogue of the main report Srebrenica, a ‘safe’ area – Reconstruction, background, consequences and analyses of the fall of a safe area. The numbers in the left-hand margin refer to the numbered main points in the epilogue.

(1) The mass slaughter of thousands of Bosnian Muslims in Srebrenica is a horrifying and probably the most violent excess to take place in the process of disintegration of Yugoslavia in the first half of the 1990. Great-Serbian ambition played a major role in this process. This places a relatively large share of the responsibility with the Serb political leaders, especially former president Milosevic. The very large degree of responsibility for the violent nature of the process of disintegration borne by the Serb leaders should not encourage turning a blind eye to the responsibility of other leaders for the resort to violence. All of the warring factions were guilty of gross violence. The reports and images of this violence met with repugnance throughout Europe and provoked a strong call for intervention by the international community, in which the Netherlands figured prominently.

(2) International interventions are rarely so preventive that they can be made before excesses occur. If those excesses do occur, the public debate on them often leads governments to intervene on moral and humanitarian grounds. More in-depth analysis of the background of trouble spots and measures based on such analysis rarely play a major role. The actual intervention is more often a matter of trial and error than of the coherent implementation of a well-conceived programme. As a result, many processes of intervention look like ‘muddling on’. This was the case in Yugoslavia as it disintegrated. From lack of a better alternative, international intervention right from the right down to the aftermath of the fall of Srebrenica was dominated by the ‘muddling on’ scenario.

(2) The promise made by UN general Morillon in 1993 to the people of Srebrenica that they were under the protection of the UN and would not be abandoned fits in with this picture of muddling on. He hereby produced a fait accompli, which the Security Council turned into the decision to label Srebrenica as a ‘safe area’, a new and undefined concept. Nothing was clear about it except that the dominant light option ruled out a mandate for a genuine military defence of the area or its population. The presence of UN troops was intended rather to be a warning by the international community not to attack (“to deter by presence”). The proclamation of the zone as a safe area created an illusion of security for the population.

(3) The Netherlands was in the forefront of the international call for some form of intervention. In 1993 a combination of humanitarian motivation and political ambitions led the Dutch government, on its own initiative and without prior conditions, to make an Air Battalion available for the UNPROFOR mission in Bosnia. The Netherlands could use this to show its worth, and Dutch prestige would be enhanced in the world. This took place amid wide political and media support and without a proper analysis of the far-reaching consequences beforehand. These were among the factors which led to the Netherlands being destined for Srebrenica, which had been turned down by other countries with arguments to back up their refusal.

(3) In its desire to achieve a consensus, the Dutch parliament closely followed cabinet policy. Parliament and the parliamentary commissions were regularly given detailed information, sometimes behind closed doors. This strengthened the cabinet, but at the same time it seriously undermined the critical and regulatory role of parliament. Criticisms were voiced in politics and in the media – some of them were forcefully expressed by the military – but critics ran the risk of being disqualified by the rest for their lack of moral fibre.

(3) The decision to become one of the main suppliers of troops for a peace mission moved many at the time. Dutch politics were dominated by the call to intervene on moral grounds. This humanitarian motivation, coupled with the ambition to improve Dutch credibility and prestige in the world, led the Netherlands to offer to dispatch the Air Brigade. By playing down the possible risks of the behaviour of the warring parties so much, a large circle of those involved in this policy, and in particular its advocates, took on a large responsibility for it.
In practice, Dutchbat was dispatched:

  • on a mission with a very unclear mandate
  • to a zone described as a ‘safe area’ although there was no clear definition of what that meant
  • to keep the peace where there was no peace
  • without obtaining in-depth information from the Canadian predecessors in the enclave (Canbat)
  • without adequate training for this specific task in those specific circumstances
  • virtually without military and political intelligence work to gauge the political and military intentions of the warring parties
  • with misplaced confidence in the readiness to deploy air strikes if problems arose, and
  • without any clear strategy for leaving.

(4) Dutchbat arrived in 1994 in an enclave with extremely complicated relations, and it was badly prepared for the actual situation there. Apart from two exploratory parties, they made little attempt to obtain information from the Canadians, their predecessors, about their experiences. The Canadian government in Ottawa was not asked for information either. The preparation and training were sufficient, generally speaking, for the military aspect, but were inadequate as regards the provision of information and insight into the situation of the population, its cultural background and experiences during the civil war. This meant that stereotypes and prejudices could already take hold during the training. All this did little to ease the relation between Dutchbat and the population. There were relatively frequent contacts during Dutchbat I, but they gradually grew less. That contributed to an introverted mentality and a reinforcement of negatively coloured stereotypes. Dutchbat was often negative about the population in the enclave, but there was no question of a deliberate anti-Muslim attitude. Contrary to what has been suggested, Dutchbat III was not conspicuous for a relatively high level of misbehaviour.

(4) Minister Voorhoeve, who as director of the Clingendael Institute had been a fervent advocate of intervention in the early 1990s, remarked in the summer of 1994 that it was an impossible assignment. The battalions often had to carry out their work in a spirit of frustration and lack of motivation. Especially in the case of Dutchbat III it became shaken and introverted. But that does not mean that it was dysfunctional. It performed its task, but that came nowhere near the desired effect. This is more the fault of the inadequate resources and the policy of the UN and UNPROFOR. Dutchbat grew less and less able to carry out its task.

(5) UNPROFOR was caught between two fires. The supposed demilitarisation in the enclave was virtually a dead letter. The Bosnian army (ABiH) followed a deliberate strategy of using limited military actions to tie up a relatively large part of the manpower of the Bosnian Serbian army (VRS) to prevent it from heading in full force for the main area around Sarajevo. This was also done from the Srebrenica enclave. ABiH troops had no qualms about breaking all the rules in skirmishes with the VRS. They provoked fire by the Bosnian Serbs and then sought cover with a Dutchbat unit which then ran the risk of being caught between two fires. On the other hand, the VRS blockade policy was a significant contributor to a frustrating and demotivating situation. As a result, the strength of Dutchbat III had been depleted by one third by the beginning of July 1995 and there was a serious shortage of supplies, from food to diesel oil for the vehicles. The last two battalions in particular became mentally and physically exhausted in the course of their mission.

(5) With hindsight there are no indications that the increased activity of the VRS in East Bosnia at the beginning of July 1995 was aimed at anything more than a reduction of the safe area Srebrenica and an interception of the main road to Zepa. The plan of campaign was drawn up on 2 July. The attack commenced on 6 July. It was so successful and so little resistance was offered that it was decided late in the evening of 9 July to press on and to see whether it was possible to take over the entire enclave.

From a military perspective Dutchbat had few grounds for mounting a counterattack on its own initiative because:

  • active defence of the enclave by military means was not in accordance with the mandate, the UN policy (the maintenance of impartiality) or the rules of engagement
  • the instruction (“to deter by presence”) was for military reaction to be above all reticent
  • the military balance of power was such that, without outside support, Dutchbat would have been defenceless in a serious confrontation
  • as a result of the ‘stranglehold strategy’ (the blockade policy of the VRS), Dutchbat III was no longer a fully operational battalion in terms of manpower, supplies or morale.
  • military means could only be deployed if the safety of the battalion was in danger and if it was the target of direct fire – the ‘smoking gun’ requirement – which the VRS deliberately avoided
  • the circumstances, such as the failure of air support to materialise and the death of the Dutchbat member Van Renssen through an action by a Bosnian Muslim, hardly encouraged the mood for a counterattack on its own initiative.

(5) The question of whether another battalion (in a different condition, or more heavily armed) would have acted differently is impossible to answer. In terms of political psychology, it was conceivable that, from fear of a negative reputation for the Bosnian Serbs, the commander of the Bosnian Serbian army (VRS), general Mladic, would have held back from armed resistance with the risk of victims on the UNPROFOR side. His decision to press on to Srebrenica was primarily motivated by the lack of any significant armed resistance. This indicates that such considerations did play a role with the VRS. Since action by Dutchbat on its own initiative contrary to the instructions was not an available option, the initiative for that would have had to come from one of the higher UN echelons. But those higher echelons were very reluctant to use close air support, let alone considering serious operations in the form of active fighting back by Dutchbat. Besides, the Bosnian Serbs deliberately avoided the ‘smoking gun’ by skirting the blocking positions and not giving UNPROFOR any direct pretext to initiate an armed counterattack on the ground. The attack on Srebrenica was characterised on a small scale by what characterised policy on a large scale: muddling on.

(6) The rapid fall of Srebrenica and the little resistance that was put up have led to questions and allegations about possible secret deals. The most important of these concerns the UN commander-in-chief general Janvier. It has been alleged that he and VRS general Mladic made a secret agreement on 4 June 1995 that no more air strikes (like those at Pale) would be forthcoming, not even in the future, in return for the release of mainly French hostages. This hypothesis does not stand up to criticism. Even without a deal, Mladic could still understand that the general conviction on the UN side was that air strikes were ruled out as long as there were hostages or the possibility of new ones. Attacks from the air had become extremely hazardous as long as the ground forces of UNPROFOR were still in the enclaves. That was why Janvier had already asked the Security Council to abandon the eastern enclaves to facilitate a stronger action.

(7) The expectation on the part of Dutchbat and the Sector North East in Tuzla that help would come from outside on the morning of 11 July in the form of massive air strikes was misguided. The UNPROFOR command had completely ruled out air strikes, but was also extremely reticent about lighter support from the air in the form of close air support. It hereby crushed the Dutchbat illusion and the enclave became an easy target for the VRS. On the Dutch side, the potential for the deployment of air support was overestimated. It was expected that “robust” action would be taken if necessary. That had been an important argument in the political and military decision-making right from the start. The UN was very reluctant to deploy the air force because it wanted to remain impartial in the conflict and to limit its task to that of peace-keeping, not peace-enforcement. The hostage crisis after the Pale air strikes had further increased doubts about the usefulness of massive air support.

(6) Fear for the fate of a group of members of Dutchbat held hostage elsewhere whose numbers had grown to 55, as well as for that of the refugees and members of Dutchbat in the enclave, who could also be regarded more or less as hostages, led Minister of Defence Voorhoeve to make an urgent appeal to the highest UN commander in the region, Akashi, to call off the requested air support on 11 July. This appeal had no effect whatsoever because the decision in question had already been taken. Akashi later falsely used Voorhoeve’s request in order to get out of having to say that the UN itself had not planned any further action. He appealed to Voorhoeve’s request by saying that it left him no alternative.

(8) The attack on the enclave, and in particular its complete collapse, came as a total surprise to Dutchbat and to all of the other parties involved. This can be explained from the lateness of the decision of the Bosnian Serbs to attack and eventually take over the whole enclave, but it was also due to the weak intelligence position of the UN and the lack of sufficient capacities and the right resources to collect and analyse intelligence. The fall of Srebrenica was thereby partly a failure of military intelligence. The Netherlands also played a part in this. The Cabinet, the Ministry of Defence and the Dutch Parliament adopted an anti-intelligence attitude. The Military Intelligence Service in particular did not receive sufficient extra resources to collect additional intelligence, and this service was not involved enough in the decision-making on Srebrenica. As a result, far fewer observational data were obtained than was technically possible. The United States had the strongest intelligence position in Bosnia. The Netherlands could have benefited from this, but lack of interest and the negative attitude of the military and political leadership stood in the way. The army top repeatedly turned down an offer by the CIA to smuggle a number of so-called Comint cases into the enclave with equipment to tap the communications of the ABiH and the VRS in the region. Defence thereby missed the opportunity to strengthen its own information position in the field of intelligence in return. The result was a weak Dutch information position in the field of intelligence. Tapped messages would at least have given Dutchbat ‘ears’, and would have put the army in a strong position vis-à-vis the US intelligence and security services to obtain additional intelligence from US sources.

(10) The tragic nadir of the fall of Srebrenica was the mass killing of thousands of Muslim men by Bosnian Serbian units. A large number of the men were members of the Bosnian Muslim army (ABiH) who had attempted to break out of the enclave to Tuzla with some of the male population during the night of 11 July. The decision to break out and thus to give up further resistance was taken entirely outside the UN and UNPROFOR. The flight to Tuzla and the mass executions which followed took place entirely out of view of Dutchbat. Suggestions that the Muslims were killed “under the eyes of Dutchbat” are unfounded in relation to this mass slaughter.

(10) The decision on mass executions was most probably made after 11 July when it became clear that the breakout led by the 28th division prevented the handling of affairs that had been planned. No written order has been found. The outbreak was a complete surprise, and it came at a very bad time for the VRS. Along with the already existing hatred, eagerness for revenge and the wish for ethnic cleansing, it was one of the factors that led the Bosnian Serbs to settle accounts harshly with the Muslim population of the enclave. This turned into an organised mass slaughter. It is unlikely that is was planned long in advance with this specific form and scope. It is more plausible to suppose that the Bosnian Serbians had counted on a surrender of the ABiH troops and a deportation of the population from the enclave after ‘screening for war criminals’ and transfer of the troops to prisoner-of-war camps.

(10) It was precisely in the treatment of the many prisoners that the Bosnian Serbs lost control of themselves. In some places the Muslims were slaughtered like beasts. They included men who had been separated from the women outside the Dutch compound in Potocari at the transfer of the population to Tuzla. That mass murder was the intention is also clear from the fact that, after the prisoners had been taken, no measures were taken to see whether there were ‘war criminals’ among them, no prisoner-of-war facilities were found, no food or drink were organised, identity cards were destroyed, and no distinction was made between combatants and non-combatants.

(10) There are a number of indications that it was a central command from the General Staff of the VRS. There are none pointing to political or military liaison with Belgrado. The involvement of the then president Karadzic (Republik Srpska) is unclear. In any case, the main responsibility for the mass slaughter lay with the military. Mladic’s central role was unmistakable and beyond doubt. He was a dominant presence during these days and was clearly in command. That does not alter the responsibility of others in leading positions in the VRS, the Drina Corps and in the special troops and security services.

(11) After the fall of the enclave, several tens of thousands of the population went to Potocari, some of them to the local Dutch compound. That mass of people stayed there for a few days in terrible and chaotic conditions and suffered from the heat and the lack of food and water. There was a risk of a humanitarian catastrophe and in view of the hygienic conditions the fear of epidemics was justified. There was no doubt about where the priority lay. Evacuation or deportation of the population was the top priority. When all is said and done, that was also what all of the parties wanted. The refugees were also very keen to leave for a variety of reasons, not least fear. There was no way of offering the refugees in Potocari armed protection. This was ruled out not only by the lack of military balance between Dutchbat and the VRS, but also because the presence of an enormous crowd of refugees packed together made effective action impossible. It was essentially a hostage situation in which any violent resistance would have provoked a bloodbath. There was no doubt about the main priority. The top priority was the evacuation of the population, and in the last analysis that was what all of the parties wanted.

(11) When buses appeared, there was a rush to get a place as quickly as possible. Many had wanted to leave earlier, and now there was very heavy congestion. That was one of the reasons for Dutchbat to perform a regulatory function in this connection. Another was that the refugees were in this way – at least not directly – at the mercy of VRS troops. It was also clear to Dutchbat that the Muslims had only one choice: to leave as soon as possible. It was the natural task of Dutchbat to supervise that and to collaborate with it, even though in the given circumstances it was tantamount to collaborating with ethnic cleansing.

(11) Deputy Dutchbat commander Franken stated later that he had recognised the danger of excesses. He had not expected mass murder, nor had he expected the VRS to abide by the rules, and he realised that in the worse scenario there might be some killing. From fear of panic and a direct catastrophe, he was silent about these fears and assisted with the evacuation in full awareness that the fate of the men was uncertain and that they, as he put it, “might end up in the most appalling circumstances”. His choice was basically to avoid a humanitarian disaster that would affect everyone. That choice had far-reaching consequences for the men among those present. The battalion command realised that the fate of the men among them, who were separated from the women and children, was uncertain, but could not have suspected that this would lead to the mass slaughter of these and many other men who fell into the hands of the Bosnian Serbs during their flight to Tuzla.

(11) Franken’s dilemmas included that of whether or not to admit persons to the compound or to place them on the list of local personnel to whom they did not actually belong. There was a general awareness in UNPROFOR about how strict the VRS checks were and of how important the necessary written permits were in this respect. Franken felt that there was too large a risk that, during the supposedly inevitable inspection, the VRS might come across someone who had no papers or untrustworthy ones. That could endanger the lives of the others. For this reason he also refused to put the brother of an interpreter (Hassan Nuhanovic) on the list of people allowed to leave with Dutchbat. His father was allowed to stay because he had taken part in the discussion between Karremans and general Mladic, but he decided to stay with his family, which cost him his life. In at least one case (the electrician Mustafic) someone who was entitled to be included on the list was refused – a communication failure with fatal consequences.

(11) A number of Muslim men had already been killed by the Bosnian Serbs in Potocari in a local act of revenge. Their number is estimated at between 100 and 400, with a wide margin. The battalion command received limited information about this, but what was reported was alarming enough: two sightings of nine or ten corpses and indications that assaults were taking place during interrogations in a house near the compound. Much more happened in Potocari than what the members of Dutchbat saw. But not all of what they saw was reported in those days. Communication was a complete failure at the time. One reason for this is the great stress to which the members of Dutchbat were exposed, their narrowing of focus, and their shutting themselves off from the world around them. In some cases too, concern about their own survival in this hell will have meant more to them than the fate of the Muslim men who had made things so difficult for Dutchbat.

(11) Be that as it may, communication on 12 and 13 July, when these killings of men taken from the crowd were carried out, failed completely. This failure of the internal humanitarian reporting raises the question of why the battalion command itself did not play a more active role. During the week after the evacuation, when the situation had quietened down, Commander Karremans and his deputy Franken did not attempt to issue a call or make inquiries to investigate the serious violation of human rights either. Afterwards they are themselves surprised by this failure to assume responsibility. By the way, they share this responsibility with officers at higher levels who – earlier than Karremans and Franken – had also received the alarming reports from survivors in Tuzla.

(12) The battalion command, particularly commander Karremans, have been strongly criticised in public, partly due to their very unflattering way of presenting themselves in the media. If we look further than this public image, it can be stated that, under the responsibility of that battalion command, most of the military and humanitarian tasks were properly performed until the hectic days of July. When the VRS opened its attack on the enclave, however, Dutchbat was no longer able to determine the course of events, though it should be borne in mind that military defence of the enclave was explicitly not a part of the task or instruction and that passive action had been enjoined. Nevertheless, there is room for criticism of a number of separate aspects, including the nature and action of the battalion command.

(13) Reactions to the reports from refugees in Tuzla about massive violations of human rights were varied. Minister Pronk (Development Cooperation) first used the word “genocide” without any qualification in a NOVA broadcast on 18 July. Ministers Van Mierlo (Foreign Affairs) and Voorhoeve (Defence), on the other hand, expressed their concern but usually in cautious terms, partly motivated by concern for the vulnerability of Dutchbat as long as it was still stationed in the former enclave. The commander of the Dutch army, general Couzy, suggested on the basis of Dutchbat reports that it was not as bad as some made out. He barely countenanced the possibility, and certainly any (indirect) complicity by Dutchbat, from uncertainty about what had actually gone on in the enclave and from a desire to protect the image of the battalion and the army. Yet on the basis of everything that he had so vaguely and incompletely learned, he could have also concluded that that was precisely a reason for alertness and further investigation.

(13) The press conference by Defence in Zagreb marked a fundamental change of direction in the media. Until then the picture had been one of Dutchbat doing its work under difficult conditions. Journalists well understood that, as long as Dutchbat was still in the enclave, reticence was called for because of the risk of being taken hostage. The first press contacts of commander Karremans after the departure from the enclave damaged the image of both Dutchbat and Defence. His attempts to characterise the relation of the warring factions (“no good guys, no bad guys”) and his value judgements on general Mladic backfired completely in the light of what many journalists now regarded as the highly probable and serious killings by the Bosnian Serbs. Given the shocking excesses that had taken place, the media were somewhat ashamed of the compliant role they had played so far, and there was now a tendency to go to the opposite extreme. The situation was aggravated by understandable and in some cases well-founded complaints about the lack of information by Defence and the disruption of personal relations between Defence and journalists as a result of the failure to fulfil commitments.

(13) The negative image that emerged at that time was also fuelled by footage of a party after the arrival of Dutchbat in Zagreb. Shots of members of boisterous Dutchbat drinking beer have become the symbol of the alleged indifference of the Netherlands in general and of Dutchbat in particular to the fate of the Muslims killed in the mass executions. This footage, which has often been shown on television, was taken from a short excerpt from a much longer internal video, the full version of which was deposited in a public film archive by Defence in the autumn of 1995. This selection of images suggests that Dutchbat was partying in full awareness of the mass murders under the eyes of the press. All three suggestions are incorrect. The nature and scale of the mass killings were not yet known, yet alone fully realised; the partying was the spontaneous release of emotions after a very moving memorial service for their dead colleagues Raviv van Renssen and Jeffrey Broere. A few dozen men took part, a small minority of the battalion.

(13) The public condemnation of the role of Karremans was also affected by video footage, in this case recordings made by the VRS during his talks with general Mladic. It has to be stated that he was the victim of manipulation which led to his being filmed in a way that was fatal for his image. Someone put a glass in his hand in a situation in which he had just managed to meet a number of Dutchbat members who were being held hostage. Mladic’s attitude left Karremans no chance at all, but where there were opportunities he seems to have said the minimum and not to have made any promises or concessions which were out of order in the given circumstances.

(13) From Zagreb on a highly critical view of what had happened predominated. The situation did not improve when it was discovered that the supply of information by Defence left a lot to be desired. Suspicion arose that an unjust attempt was being made to preserve the reputation of Dutchbat untarnished. Minister Voorhoeve was constantly under fire, was uninformed of the facts, and therefore called for a broad, comprehensive inquiry, a continuation of the debriefing on a larger scale. He counted on the loyalty, support and correct political sense of the army. However, the army top had different priorities, such as preserving the image of Dutchbat and of the army, as a result of which the minister was informed relatively late, often inadequately, and on a few occasions was not even informed at all. The debriefing report was inadequate. The army could clearly put its stamp on it. Voorhoeve and his aides knew that. Nevertheless, he considered that there was no alternative but to praise the report, because he had used it to keep questioners at bay for weeks. Immediate recognition that the report was inadequate would be a disowning of the army, and he did not want that. Voorhoeve had become the prisoner of the debriefing report.

(14) The limited debriefing policy of the army top bounced back on the defence organisation like a boomerang. It was virtually paralysed, especially in the summer of 1998, by the constant Srebrenica exposures in the media. The new Minister of Defence, De Grave, felt the effects of this immediately after assuming office. He decided to act more forcefully: he ordered an inquiry to see whether Srebrenica had been hushed up. The Van Kemenade report found no evidence for that, but his investigation was not in the first instance intended to discover the truth; in the last resort it was about a problem of political administration that required an acute solution. Van Kemenade therefore did not conclude, as the present report does, that there was not only incapacity but also a deliberate attempt by the army top, contrary to the wishes of the Minister, to limit the flow of information and, where possible, to avoid sensitive issues. Van Kemenade failed to recognise the evidence for that; it should have led him, at least, to be more suspicious of his early conclusion that there had been no hushing up.

(15) After July 1995 the tragedy of Srebrenica has continued to have a profound effect, first of all for the survivors and next of kin of the population. For most of them, the events have wrecked their lives. In particular, many women from Srebrenica still suffer every day from what they have been through. Many of them had already had a series of exhausting and horrifying experiences before the setting up of the safe area. The attack, deportation and mass killings came after two years in a besieged fortress, in which the VRS used its ‘stranglehold strategy’ in an attempt to create an impossible situation in humanitarian terms. The process of digging up and identifying the corpses has still not been completed yet. In many respects the next of kin feel abandoned by the rest of the world, in Bosnia and outside it.

(15) The members of Dutchbat have also been deeply affected by their time in the enclave. Many of them have long-term psychological problems arising from their experiences in Srebrenica, and in some cases they are still serious. Many of them were not impressed by the counselling and aftercare they received. Their reception in the Netherlands, in an atmosphere of public debate in which Dutchbat was often presented in a very negative light, certainly did not help them to cope with their problems in a healthy and balanced way on the home front. That atmosphere left little room for understanding what it had ‘really’ been like according to the Dutchbat members. They did not recognise themselves in the image that dominated in the media, of deep black (mainly Bosnian) Serbs and lily-white Muslims. Most of the members of Dutchbat had difficulty in accepting that picture. The world that they had known during their stay in the enclave had been different.

(15) The mass killing had a strong shock effect in UN and NATO and contributed to a U-turn in US policy. In July 1995 a line was drawn at the menace to Gorazde, where British troops were stationed. The UN’s lack of flexibility because of the need to maintain impartiality diminished and there was an increased readiness to deploy the air force. The military were given wider powers to deploy it, and the ‘smoking gun’ condition was dropped. The shelling of the Markale market in Sarajevo in August 1995 triggered tough air strikes by the UN and NATO against the Bosnian Serbian air defence. That united action – “Deliberate Force” – led to Dayton November 1995, where the reluctant parties were nevertheless forced to sign an agreement. The Netherlands played no role at all at this stage. It was even banned from the conference table.