Srebrenica Dutch Liability: The 10 Percent That Does Not Mean What It Says (Cold Files 020)

2 days ago

9 min read

Cold Files 020. The Netherlands spent twenty-eight years investigating and adjudicating its own conduct at Srebrenica. Project Leaflet graphic.

In July 2019 the Dutch Supreme Court put a number on the Netherlands and Srebrenica, and almost every headline that followed got it wrong. The figure was ten percent. It was reported, in wire copy and broadsheets alike, as a finding that Dutch peacekeepers bore ten percent of the blame for the massacre, or that there had been a ten percent chance they could have stopped it. It was neither. It was an estimate of a probability, in a private tort action, about one afternoon inside one compound, converted into a fraction of a money judgment. The gap between what that number is and what it was read to mean is the subject of this file.

Part of the Cold Files series.

The Report That Declined to Judge

The Dutch government commissioned its own history of the catastrophe in the autumn of 1996. The job went to the Netherlands Institute for War Documentation, and NIOD took five and a half years over it. The main report alone ran to 3,393 pages, and with its sub-studies the whole came to some six thousand six hundred.

By any measure of state self-scrutiny this is an enormous object. The editors of the standard scholarly volume on the Srebrenica investigations, Isabelle Delpla, Xavier Bougarel and Jean-Louis Fournel, note that Britain and the United States, both of which might equally have been called into question, produced no investigations or reports. France's National Assembly did report, on 22 November 2001, five months before NIOD, but it reported as a mission d'information rather than a commission of inquiry, without the power to compel a witness or put one on oath. The United Nations examined itself too, and its 1999 report to the General Assembly was the most unsparing of any of them.

What NIOD produced was a reconstruction of extraordinary density that declined, in terms, to say who was at fault. The epilogue written by the institute's director, J. C. H. Blom, states that the report does not aim to supply political conclusions or judgments, and that this is a task for the political and public arena. The finding was the record. The verdict was somebody else's problem.

The report was not gentle about the mission. The safe area, it said, was a new and undefined concept, and declaring the enclave one created for its population an illusion of safety. It found that Dutch politicians who had pushed hardest for intervention had taken on a heavy responsibility by understating the risk. On the handover at Potocari, NIOD's authorised summary records that helping supervise the evacuation was Dutchbat's duty even though, in the circumstances, that meant assisting in ethnic cleansing.

That sentence describes the act and declines the judgment, which was the instrument's design rather than its limit.

The Inquiry That Blamed No Politician

The Binnenhof in The Hague seen across the Hofvijver, the parliamentary complex where the Srebrenica inquiry sat in 2002. Photograph by Txllxt TxllxT, CC BY-SA 4.0, via Wikimedia Commons.

NIOD handed the report to the cabinet on 10 April 2002. Six days later, on 16 April, Prime Minister Wim Kok told the Tweede Kamer that he had offered the Queen the resignation of the entire cabinet. He accepted political co-responsibility and rejected guilt.

The resignation is usually described as the report bringing down a government, and that is literally true and practically thin. A general election was already fixed for 15 May, twenty-nine days later. Kok had already announced he would not stand again. There had been no parliamentary defeat; the Dutch reference work parlement.com records flatly that no conflict with the chamber underlay it. The Queen asked the ministers to stay on as caretakers until the election, and they did. The cabinet resigned without losing a day of governing.

Parliament then reached for the one instrument with real teeth. On 25 April 2002 the Tweede Kamer instituted a full parliamentary inquiry under the chairmanship of Bert Bakker. Unlike NIOD it could compel attendance and administer oaths, and it used both: fifty confidential preliminary interviews, and forty-one public hearings under oath across ten hearing days between 11 November and 6 December 2002.

Its report, Missie zonder vrede, appeared on 27 January 2003 and ran to 467 pages. It went further than NIOD had been willing to go. The Netherlands, it found, bore a responsibility entirely its own, and The Hague could not offload Dutchbat's failures onto the international political context. It found no worst-case planning for the enclave's fall, an unclear mandate, and hesitancy by the French UN commander Bernard Janvier that cost Dutchbat its air support. At Potocari, it found, Dutchbat made no real attempt to accompany the Muslim men to the last.

Then it stopped. The committee concluded that it was not realistic to hold any individual politician accountable, persuaded the chamber of that, and the matter closed. Two inquiries, six and a half years between commissioning and conclusion, and not one Dutch politician held to account.

The Only Defendant a Court Could Reach

The International Criminal Tribunal for the former Yugoslavia in The Hague, which tried the Srebrenica crimes while the Dutch state was being sued across the same city. Photograph by the ICTY, CC BY 2.0.

What the inquiries had not done, the survivors tried to make the courts do. In June 2007 the Mothers of Srebrenica foundation and a group of relatives issued proceedings against two defendants: the Netherlands, and the United Nations.

The second never answered. The UN declined to appear, and the Dutch courts held its immunity absolute at every level: the district court in July 2008, the appeal court in March 2010, and the Hoge Raad on 13 April 2012. The claimants took it to Strasbourg, where on 11 June 2013 the European Court of Human Rights declared the application inadmissible. Immunity reflecting generally recognised rules of international law, the Court held, could not in principle be regarded as a disproportionate restriction on access to a court, and the prohibition of genocide being a peremptory norm did not displace civil immunity. The absence of another forum did not change the answer.

This is the hinge of the whole story and it is routinely skipped. The United Nations created the safe area, held command of UNPROFOR, and ran the mandate. It could not be sued in any national court. The Netherlands did not become the principal defendant because it was the principal wrongdoer. It became the defendant because it was the only one a court could reach.

A criminal route was tried too, against three Dutchbat servicemen, and it failed at every stage: the prosecutor refused in March 2013, the military chamber at Arnhem-Leeuwarden dismissed the complaint in April 2015, and Strasbourg declared that application inadmissible in September 2016.

Full, Then Thirty, Then Ten

The Peace Palace in The Hague, seat of the International Court of Justice, which found Serbia in breach of the duty to prevent genocide at Srebrenica in 2007. Photograph by Yeu Ninje, public domain, via Wikimedia Commons.

The civil claim against the Netherlands had already produced one landmark. Hasan Nuhanovic, a Dutchbat interpreter, and the family of the electrician Rizo Mustafic sued over three men sent out of the compound: Mustafic, Nuhanovic's brother Muhamed, and his father Ibro, who had been allowed to stay and left anyway with his son. The district court dismissed the case in September 2008 on the ground that the conduct was attributable only to the UN. The Hague Court of Appeal reversed that on 5 July 2011, holding that the test was effective control and that the Dutch state had possessed the power to prevent the expulsions. The Hoge Raad affirmed on 6 September 2013. It remains, as far as the published record shows, the first time a troop-contributing state was held liable for the conduct of its own peacekeepers.

The group claim went differently, and the way it moved is the point.

On 16 July 2014 the Hague District Court found the state liable in respect of the men who had been inside the Dutchbat compound at Potocari on 13 July 1995, a group it described as more than three hundred. Dutchbat, it held, should not have participated in their deportation, and it was sufficiently certain that they would have survived had Dutchbat not cooperated. No percentage was applied. Liability was full.

On 27 June 2017 the Court of Appeal took that certainty away. It held it too uncertain that the men would have lived had they stayed, which on ordinary causation would have sunk the claim entirely. It rescued the claim by recasting it as the loss of a chance, estimating at thirty percent the chance the men would have escaped execution had Dutchbat offered them the choice to remain, and holding the state liable, in its own words, for thirty percent of the losses suffered by the relatives.

On 19 July 2019 the Hoge Raad cut the estimate to ten percent. The men's prospects, it reasoned, were very bleak even inside the compound; the Bosnian Serbs would probably have found them during inspections and pressed to remove them; effective international support was unlikely. But it could not be completely ruled out that Dutchbat, had it held out, would have deterred an attack. The chance was, in the court's phrase, small but not negligible. The Supreme Court also struck out the appeal court's declaration that forming the sluice at Potocari had been wrongful, and dismissed their other grounds.

So the number fell from full liability to thirty percent to ten, and at no point did it measure what the headlines said it measured.

What Ten Percent Does Not Say

Gravestones at the Srebrenica-Potocari memorial cemetery in Bosnia and Herzegovina. Photograph by MichaelBueker, CC BY-SA 3.0, via Wikimedia Commons.

Ten percent is the Supreme Court's estimate of a probability of survival, transposed under Dutch loss-of-a-chance doctrine into liability for ten percent of one group of families' civil damages. It is not ten percent of the genocide, not a finding that the Netherlands was one-tenth at fault and the Bosnian Serbs nine-tenths, and not a probability that Dutchbat could have prevented the massacre.

There is an arithmetic available here and this file is going to refuse it. Around eight thousand men and boys were killed. The courts' group was about three hundred and fifty. It is tempting to multiply the liability fraction by the headcount fraction and present the product as the measure of what the reckoning reached. That would be a category error of exactly the kind this series exists to dismantle: a precise-looking figure made to carry a meaning its source does not support. Nothing in any judgment authorises it.

The honest version is harder and better. The percentage tracks evidentiary difficulty about one afternoon, not the depth of Dutch culpability. Zane Ratniece, writing in the Utrecht Journal, calls it a reasoned guess and notes that the court gave no convincing reason for ten rather than twenty or forty, and that the parties were not heard on the figure. Tom Dannenbaum and Kristen Boon both fault the judgment, though for its attribution test rather than for the percentage.

What the state did pay is a separate record and a clearer one. The settlement scheme run by the Wortmann commission opened in June 2021, offering fifteen thousand euros to a widow and ten thousand for a father, son or brother. Its final report, tabled on 22 November 2024, recorded 8,135 applications, 2,374 of them granted, filed by the relatives of 612 victims, for roughly twenty-five million euros in all. The scheme's eligible class was the courts' class. Assessing claims one by one, the commission established 612 victims where the courts had estimated about three hundred and fifty. Four months before that scheme opened, the cabinet announced a one-off five thousand euros for each surviving member of Dutchbat III.

The apologies divide the same way. On 18 June 2022 Prime Minister Mark Rutte apologised to the women and men of Dutchbat III for the conditions of their deployment, their reception on return, and the lack of support. Three weeks later, at Potocari on 11 July, Defence Minister Kajsa Ollongren told the families that the international community had failed to protect the people of Srebrenica, and that for that the Netherlands offered its deepest apologies. One apology was for what was done to Dutch soldiers. The other was for what the world failed to do. In January 2020 the relatives took the ten percent to Strasbourg. No decision has been published.

Twenty-eight years of institutional reckoning produced a historical record that declined to judge, an inquiry that blamed the state and spared its politicians, and a tort action against the only defendant a court could reach. Each instrument answered the question it was given. The one number that escaped into public memory is the one that measures least.

Sources consulted