Counting and Miscounting Palestinian Refugees – Part I

Counting and Miscounting Palestinian Refugees – Part I

by

Howard Adelman

Counting is one of the most basic things we learn to do. No sooner does a child learn to recognize mom and dad and to understand he or she has only one of each, than that child learns to count siblings and then cousins. ‘Well the new baby means you now have two brothers.’ ‘Three!‘ ‘No, not three brothers. Anna is your sister. You have two brothers and one sister.’ ‘No, momma has four children, not three. You have to count yourself. I have three boys and one girl.’ And so it continues.

As we grow, we learn to aggregate and differentiate. And we also learn our first lessons in lying and cheating. ‘I told you that you could have three M&Ms, not five.’ ‘But two were broken. They don’t count.’ And when we are mature, we learn to become sloppy in our counting and even deliberately mislead for entertainment or supposedly humanitarian and political purposes. In 1982, just after the Israelis invaded Lebanon, OXFAM Britain published full page ads that claimed 600,000 had been made homeless in southern Lebanon by the Israeli invasion. Israel responded and issued a press release insisting that only 27,000 had been made homeless. The Centre for Refugee Studies (CRS) at York University, then called the Refugee Studies Project, earned its initial international reputation in part by proving its thesis that, even in the midst of conflict, reasonably accurate counts of refugees can be carried out. Based on this and other research, CRS was dubbed a Centre of Excellence and received research funding of $5,600,000 over five years.

The accurate number made homeless by the invasion in South Lebanon was 40,000. That figure was subsequently used by all sides in the conflict. The credibility was earned in the following ways:

·         Demonstrating that the 600,000 figure originated in an ICRC report that 600,000 had been affected, not made homeless, by the invasion

·         That the research on the ground by a highly reputable Israeli academic erred by 10,000 as a result of one unintended addition error and by 3,000 because several pockets of refugees had been missed

·         That the twelve counts on the ground carried out by Palestinian elementary teachers, municipal authorities, the Society for Engineers in Lebanon, and nine other institutions in Lebanon, including the International Red Cross, though varied somewhat, could all be reconciled if the same definition of a refugee was used (not Palestinian homes in camps but Palestinian homes occupied by ethnic Palestinians – many homes had been rented out to itinerant workers in Lebanon from South Asia primarily –that were destroyed by the invasion), the same territory covered and corrections plus allowances taken into account for errors easily made when counting in such difficult circumstances

·         By showing that the International Committee of the Red Cross count, which ICRC only shamefacedly owned up to because the method of counting lacked any rigour, was simply based on its years of experience using only a simple rule of thumb rather than actual counting; ironically, that estimate turned out to be the most accurate – simply count the number of kitchen kits distributed and multiply by three.

One might think that, unlike Lebanon where refugees were simply scattered among the local population, counts of refugees in camps would be easy. The people are all supposedly contained in a given area. When, as part of one study, the Centre for Refugee Studies at York University went to verify the numbers of Somali and a few other types of refugees in Dadaab Camp in Northern Kenya, well before the camp numbers were enormously increased by the most recent continuing conflicts in Southern Somalia, on census day conducted by the UNHCR, over 100 buses and many other vehicles, mostly from Nairobi, had arrived to allow refugees to return so they could be included in the count and ensure their families had adequate rations, which then consisted of only 1,600 calories a day per adult. We could find no accurate method to differentiate between local Northern Kenyan ethnic Somali citizens who had joined the camp population to get cards entitling them to rations, but estimates were arrived at through small samples of very defined areas. From reviews of such counts, we began to adopt a rule of thumb for camps administered and controlled by UNHCR. Discount published census figures by 10% to gain a more accurate estimate of the number of actual refugees living in a camp.

Sometimes miscounting can have horrible consequences. We learned that when the actual control of camps was in the hands of militants, and even worse, defeated soldiers, then the numbers were inflated far more because the militants used ration cards to acquire foodstuffs and sell them on the black market. The monies were used to purchase arms as well as to pay themselves. Then we used as a guide a 25% inflation figure.

For an example of the abuse of refugee figures for both humanitarian and political purposes, the Hutus who fled to Eastern Congo then called Zaire provide an excellent example. Ignore for a moment the large numbers of Rwandan refugees that fled to Tanzania (500,000) and Burundi (200,000). In the Eastern Congo to which the defeated army in Rwanda had fled in the summer of 1994 along with hundreds of thousands of additional Hutu refugees following the genocide of 800,000 Tutsi and moderate Hutu in Rwanda, the reported number of refugees in camps in the Eastern Congo (Zaire) mostly controlled by the ex-FAR (Forces Armée Rwandaises) as well as the interahamwe and impuzamugambi militias, was said to be 1,400,000 out of a total Hutu refugee population of 2.1 million. The definition of refugees excluded soldiers and militants estimated at 50,000. If Hutu civilians who participated in the killing are counted among the genocidaires as well as members of ex-FAR families who accompanied the soldiers into exile, then these along with the soldiers and militia members totaled about 200,000.

Were the over 100,000 Hutu civilians who had participated in the massive slaughter of innocents to be counted as civilians or as militants? What about the members of the families of the soldiers? After all, they were civilians so they should definitely be counted as refugees.

When the camps were broken up by an invasion of Zaire in 1996 by Kagame’s forces, an estimated 800,000 Hutu refugees returned from Zaire to Rwanda. 600,000 Hutu were reported as having fled deeper into Zaire. These included members of the ex-FAR (Forces Armée Rwandaises), their families, the interahamwe and impuzamugambi and other civilians who participated in the killing. The number 600,000 was obtained by subtracting the 800,000 who returned from the estimated number of 1,400,000 in the Zaire camps. But if those numbers were inflated by at least 25% given the military control over the camps, then the figure of the number that fled was inflated by 350,000. Though up to 50,000 of those who fled were estimated to have been killed by Rwandan armed forces under Kagame, including a large number who were civilians, the charge that a second genocide had been committed because the 350,000 phantom refugees were “missing” and presumed dead. So a figure began to be distributed that Kagame’s forces had killed 400,000 to 600,000 in a second genocide rather than the estimated up to 50,000 who had died.

The actual numbers in Zaire were as follows:

Estimated number of refugees (not soldiers or militia members) in camps 1,400,000

Less inflation factors of 25% (-350,000)                                                        1,050,000

Numbers who Hutu refugees who returned to Rwanda from Zaire                 800,000

Numbers of civilians, including civilian genocidaires   100,000

plus family members of militants                        100,000

others                                                                       50,000

Total of Civilians who fled deeper into the Congo                                         250,000

Actual Total Number of Hutu Refugees in Zaire                                     1,050,000

This figure of 600,000 refugees who fled deeper into Zaire included 350,000 phantom refugees. The large number was used to accuse the Rwandan government of committing another genocide of 400,000 to 600,000 Hutu refugees. Thus are inflated imaginary figures used to conduct propaganda wars.

What has this to do with counting Palestinian refugees?

In the next two days, the answer.

The BBC Report: Genocide Denial – A

The Rwandan Genocide Revisited:
Part IV: The BBC Report – Genocide Denial – A

by

Howard Adelman

The BBC video is available at http://vimeo.com/107867605.

In a rare joint sitting of Rwanda’s two houses of parliament this past Thursday (23 October 2014), the legislators condemned the BBC for broadcasting the documentary accusing Kagame of both being a major cause of the Rwanda genocide and for being responsible for downing President Juvenal Habyarimana’s aircraft on 6 April 1994 when he was returning from the Tanzanian peace talks with the Tutsi RPF rebels. There, Habyarimana had finally agreed to implement the Arusha Peace Accords.

Parliament banned the BBC from Rwanda and accused the documentary makers of genocide denial, a charge which BBC, as expected, denied. Interestingly, Paul Kagame in an interview, at Chatham House after the BBC program was broadcast, did not at first appear to call for censorship of the BBC for its broadcast. Kagame said, “The BBC can say whatever they want to say. They don’t have to say or do, whatever they do or say, because that is right. They say or do whatever they say and do because they can.” Though to some he appeared to be defending BBC’s right to broadcast, he was actually asserting that what BBC broadcast was not right and the BBC only broadcast what it did because it could get away with it – with the implication that such behaviuor would not have been permitted in Rwanda.

While I am very critical of the documentary and see it as a very well-made piece of schlock with some very interesting interview segments, the producer and director are not guilty of genocide denial. But they come very close. A genuine case can be made that they crossed the line. It is not because of the accusations leveled against Paul Kagame. It is because of the way the program played with numbers.

Those killed in the genocide are generally believed to have been mostly members of the minority ethnic Tutsi group, though the slaughter included the killing of moderate Hutus as well as random killings for an assortment of reasons. The documentary, The Untold Story, interviews and endorses the views of a pair of US academics, Dr. Allan Stam, currently Dean at the Batten School of Leadership and Public Policy at the University of Virginia, and Dr. Christian Davenport of the University of Michigan. (See their web site: http://genodynamics.weebly.com/) Stam was previously Director and a Professor at the Gerald R. Ford School of Public Policy at the University of Michigan. Both are reputable and widely recognized as scholars.

That does not mean that all their work is excellent, though some certainly has been highly regarded. I met Allan Stam at the International Studies Association meeting in 2004 when he was awarded the Karl Deutsch Award for his scholarly contributions to that date on the origins of war and on democracies at war. The body of work did not include his research with Davenport on the Rwandan genocide, though his work had begun before that date and the year before he had received a National Science Foundation Grant to undertake research on “Mass Killing and Oases of Humanity: Understanding Rwandan Genocide and Resistance”. Thus far, none of the results have been published in either an academic journal or as a book, though publication in the near future has been promised. Given the sensation aroused by the BBC documentary, such a book could garner high sales as a trade rather than an academic publication.

Though not yet published, the data gathered and a number of relevant references, including criticisms of their work for the BBC and on the BBC documentary, have been included on their web site. They offer three reasons for their failure to publish after 14 years of work on the topic: 1) Their attempt to be as accurate as possible; 2) criticisms they encountered surprised them and set them back; 3) the fact that, nevertheless, they have posted their research online and in a transparent way. They should be congratulated for the latter; it makes it much easier to offer a critique. They include a very useful list of a large number of commentaries on the BBC broadcast. They do not offer the explanation offered by Filip Reyntjens (who is also used in the BBC documentary) that the research was too inadequate to be worthy of scholarly publication.

As far as the killers, Stam and Davenport reiterate that the murderers “were a group of extremist Hutu members of the Rwandan Armed Forces [the FAR], the Presidential Guard, national police, the ‘Zero Network death squads’ as well as affiliated militias: the Interahamwe and Impuzamuga.” Because I and most other scholars contend that the militias were primarily responsible for most of the actual killing, I use the simpler designation, the FAR and the Interahamwe militias whereas they refer to FAR+. The killers targeted Tutsi to exterminate them.

In the BBC documentary, and in their research, the controversy is about those killed not about the killers. The pair claimed that most of those killed were not Tutsi but Hutu. “(O)ur best estimate of who died during the 1994 massacre was, really, an educated guess based on an estimate of the number of Tutsi in the country at the outset of the war and the number who survived the war. Using a simple method —subtracting the survivors from the number of Tutsi residents at the outset of the violence — we arrived at an estimated total of somewhere between 300,000 and 500,000 Tutsi victims. If we believe the estimate of close to 1 million total civilian deaths in the war and genocide, we are then left with between 500,000 and 700,000 Hutu deaths, and a best guess that the majority of victims were in fact Hutu, not Tutsi.”

If the hundreds of thousands of Hutu allegedly slaughtered by Kagame’s Rwandan Patriotic Front (RPF) which won the civil war both in Rwanda and, after the war, in the Congo, the implication is that the Akazu, the party intent on the extermination of the Tutsi, were but a pinprick in their slaughter of Tutsis compared to the massacres of Hutu committed by both the extremist Hutu Akazu and the predominantly Tutsi RPF. As Stam and Davenport claim. “there was clearly a genocidal campaign, directed to some degree by the Hutu government, resulting directly on the deaths of some 100,000 or more Tutsi.” (my italics)
The number of Tutsi was first said to be 300,000 to 500,000, then reduced to 300,000, and then to 200,000. Then the number of those deliberately killed by the Akazu was reduced to a guess of approximately 100,000. The last figure was offered without explanation, without an examination of the mass graves, without an examination of all the evidence that countered such a claim. This may not be genocide denial, but it is on the same plain as those who whittle the number of Jews killed in the Holocaust down to 1,000,000 to 1,200,000.
This is when the scholarship could be accused of crossing the line into denial. Over 460 graves with more than 20 in each grave does not support a thesis that 90% of those killed were not products of genocide. Stam and Davenport insist that the correct comparison is not the Holocaust but the civil wars in England, Greece, China and Russia where civil order broke down and ethnic-based violence took place combined with random rather than intentional slaughter. However, unlike those civil wars, and characteristic of Rwandan society in general, and in spite of the presence of extremism, civil order did not break down, even in the refugee camps.

The authors make numerous errors in their account, some of which I will enumerate at greater length tomorrow and the next day. As a single example of error, they are just incorrect that Romeo Dallaire claimed that the plan to exterminate the Tutsis had been hatched in 1988. No source is cited for such a claim. For good reason. Dallaire claimed that the genocide was planned two years before the actual mass slaughters occurred in 1994 and not two years before the invasion began in 1990. Trial runs of mass killings averaging 300 Tutsi took place over that two-year period. There is no evidence presented that the planning started six years before or that Dallaire made such a claim.
While Stam and Davenport may be stars in the quantitative world of current American political science, they are terribly inaccurate in their specific citations. Further, for those who claim that their research is “evidence-based”, where is the evidence that approximately 100,000 Tutsi were killed for genocidal reasons. That is not even a best or an educated guess. It is speculation based on whimsy. I cannot find a reasonable basis for such a calculation after several reviews of their documentation on their site.
Even Filip Reyntjens, who defended the BBC program in which he appeared and criticized its critics, wrote, “I do not need to dwell on the second claim considered untenable by the signatories (in a letter written by many experts criticizing the program). I agree with them that the figures provided by Professors Stam and Davenport on Tutsi and Hutu killed in 1994 do not appear to be based on solid research. At least the data they have published (not in a scientific journal or book, but merely on their website http://genodynamics.weebly.com) are insufficient to support their claim, which flirts with genocide minimisation or denial.”

However, let’s first examine the case they make. I will save the systematic rebuttal for two subsequent blogs. Stam and Davenport followed the lead of Alan Kuperman (now at the University of Texas in Austin who first gave his provocative paper as a young scholar at the 1998 or 1999 International Studies Association conference). Kuperman made a name for himself by attacking the position that the intervention of a peace force could have stopped the genocide. He did so by tracking the battles and showing that Tutsis were mostly slaughtered just before the RPF made significant advances. (See Alan Kuperman (2000) “Rwanda in Retrospect,” Foreign Affairs, Jan/Feb.) As he wrote, “even a massive Western intervention could have saved only a quarter (around 125,000) of the Tutsi lives lost in the massacres.”

Note that he had concluded that 500,000 Tutsi had been killed in the genocide. But, as he argued, “Intervention may be better late than never, but it requires military effort on a scale and for a length of time that will make most developed countries’ military staffs and politicians blanch.” (See Alan Kuperman’s book published in 2001, Genocide in Rwanda; another scholar, Jeffrey Herbst in his essay “The Unanswered Question: Attempting to Explain the Rwandan Genocide,” used the same figure of 500,000 as the number of Tutsi killed by the genocidaires.)

Countering the claim that 5,000 U.N. troops could have halted the Rwandan genocide in its tracks is one thing. Alison Des Forges vigorously attacked Kuperman for his thesis in a subsequent issue of Foreign Affairs. She also challenged Kuperman’s claim that Clinton not only did not know of the genocide until two weeks later, but that he could not have known, a nonsensical claim as Des Forges made very clear. Such a mistake is common among academics. However, the Stam/Davenport claim that only about 100,000 Tutsi were targeted and killed for genocidal reasons is another order of error altogether.
The two American scholars do not use the usual tactic that the numbers slaughtered totaled far less. Even Alison Des Forges, one of the very early and extremely reputable scholars who documented the genocide, had insisted that the total number was 540,000 of whom an estimated 500,000 were Tutsi. However, at a meeting in Geneva attended by a number of scholars on the genocide, we reviewed the various figures and the evidence for them. We agreed at that time to use a figure of 800,000 as the most reasonable estimate of those killed. We did not distinguish between Tutsi and Hutu killed, but presumed from all our studies that the vast majority were Tutsi, the product of mixed marriages or moderate Hutu. The American scholars insisted that the overall number killed was slightly over a million, not 800,000. Further, and this is the controversial part, the vast majority were Hutu. Only 200,000 (at most 300,000, though at another point they said at most 500,000) of the one million death toll in Rwanda had been Tutsi and only half of that 200,000 had been killed in genocidal motivated slaughters.

The vast majority of Stam and Davenport’s statistical work was used to aggregate the total of one million and not disaggregate that total. The Rwandan government, as well as myself, agree with Stam and Davenport with the revision of the total upward. Subsequent to the Geneva meeting, Astri Suhrke and I reviewed our figures. Our estimates were based on mass grave sites and sample reviews of three sites and a very detailed counting at one site, the technical school in Butare where all those gathered at the school to flee the massacres were gathered, then slaughtered over a 3-4 day period. Over 17,000 killed were buried in one mass grave which had been dug by a contractor three weeks before the genocide commenced on 6 April 1994. The bodies were unearthed from that one mass grave not long before we visited the site and were laid out on the benches and tables of the school.

The most searing experience I had in conducting our study of the involvement of non-Rwandans in the killing was confirming the count. Because the bodies had been packed together so densely, they still had most of their flesh. The smell was overwhelming. A half-crazed woman wandered around the site, but at that time it was only managed by one person. An observer could even see how some were killed, especially women; long wooden spears had been stuck up their vaginas to penetrate their hearts.

I, for one, audited the official count of the bodies unearthed in that one grave by counting the bodies in four different school buildings. The official figures provided were accurate. When we audited two other killing sites, we came to concur with the government figure that over a million had been killed and began to use the figure of 800,000 to 1,000,000 as the number that were slaughtered. Fearing accusations of exaggeration, given the agreement we had made in Geneva, we did not simply use the one million figure. Further, we had never properly audited the total. But I was convinced that the revised government figure of over a million was the accurate one. The two Americans scholars agree. So I am pleased that Stam and Davenport concur that the total figure of those killed was over one million.

However, I am not pleased or at all persuaded by their disaggregation of that figure. Stam and Davenport claim that, “Essentially, research on the topic of Rwandan casualties has relied upon eye-witness testimony via survey, census and interviews.” What about the data from the mass grave counts? Even in the documentary, Jane Corbin uses the figure of 15,000 killed in the massacres at the school Murambi though official figures are much higher with estimates of 40,000 to 65,000. Wikipedia uses the highest figure. “On April 16, 1994, some 65,000 Tutsis ran to the school. After the victims were told to gather there, water was cut off and no food was available, so that the people were too weak to resist. After defending themselves for a few days using stones, the Tutsi were overrun on April 21. The French soldiers disappeared and the school was attacked by Hutu Interahamwe militiamen. Some 45,000 Tutsi were murdered at the school, and almost all of those who managed to escape were killed the next day when they tried to hide in a nearby church.”

It is not very important that Stam and Davenport ignore our scholarship since Astri and I were far more focused on the role of external actors. However, American anthropologists specializing on Rwanda, numerous historians and political scientists, particularly European ones, who are recognized authorities on Rwanda, are never cited and not even listed in their bibliography. But even if one simply adds up a few mass graves, the total easily far exceed a figure of 100,000 Tutsi killed through genocidal actions. Adding them all up leads to a figure of at least 500,000-700,000 Tutsi killed.

The Americans argued that large numbers of Hutu killed had been murdered for six different reasons. First, many members of the killing militias were non-locals who could not distinguish in many cases between a Hutu and a Tutsi, even if in his own district he could identify who was either a Tutsi, a Hutu, or, for that matter, a Twa, the tiny minority descended from the original inhabitants of the area. Secondly, the cover of war was used systematically using the mass killing to settle political, economic and personal scores. “(T)here was a large degree of random political violence taking place or what is referred to as ‘wilding’ (Fujii; Hatzfield). In this situation, ordinary/non-government affiliated Rwandans (both Hutu and Tutsi) squared off against other Rwandans (both Hutu and Tutsi) in an attempt to exact revenge for personal wrongs, financial gain or collective hatred – some ethnic, some political and some idiosyncratic in nature.” Third, many Hutu were killed who tried to protect or hide neighbours. In one case that I personally investigated, the body of a Tutsi killed by the militias had lain in the middle of the road for several days. A neighbour of the individual killed eventually went out to cover the body with an old coat. A member of the Interahamwe militia spotted him. That Hutu, appalled at the indignity of his neighbour left to rot in the middle of the road, was instantly killed. Fourth, many Hutu were simply killed in the fog of war. Fifth, moderate Hutus who were charged with supporting the RPF (and hence Tutsi rule according to the extremist ideology) were deliberately killed. Finally, the American scholars charged the RPF with killing large numbers of Hutu after they conquered a territory to impose their rule. This does not even include the number of Hutu killed in the Congo in the war that broke out there in 1996. Those estimates vary from 20,000 to 60,000.

I do not think that anyone disagrees with the classification of the reasons for those killed, but simply the distribution of the total numbers among those classes. The distribution Stam and Davenport arrived at just does not fit with what we observed, what we counted and what the vast majority of scholars have concluded who have studied the genocide. Their failure to question their own calculations adequately, their consistent insistence that is blind to their own fundamental errors, their seemingly unshakable belief that, through their quantitative analysis, they were correcting the errors of previous scholars, ended up producing results that were derided as genocide denial. Though that derision may be undeserved, it is understandable given the falsifications of the record resulting from their quantitative studies, about the disaggregation of those totals arrived at through inference, illogic and very questionable assumptions.