²⁰⁸²

Wednesday, 13 March 2002

[Open session]

[The witness entered court]

[The accused entered court]

--- Upon commencing at 9.32 a.m.

JUDGE MAY: Yes, Mr. Milosevic.

WITNESS: BAJRAM BUCALIU [Resumed]

[Witness answered through interpreter] Cross-examined by Mr. Milosevic: [Continued]

Q. [Interpretation] We left off yesterday with this big -- well, big hump of paper and it says "Timetable" and a "Log." Do you consider that this log should have been kept in the Albanian language?

A. Yes. I think it should have been kept in the Albanian language because that was the case until before 1990s, as I said.

Q. First of all, that's not how it was. And the logbook is kept in Serbian for purposes of transport. What do you think at the railway station in England? Is it kept in English or what language? And in France in French, in Germany in German? What do you think?

JUDGE MAY: That's not for the witness. He's given his view about it.

MR. MILOSEVIC: [Interpretation]

Q. How, for example, could a controller coming from Belgrade, from the headquarters, how would he be able to control the work going on at the railway station where you were employed? How could he look through the traffic logbook if he didn't know the Albanian language, in a company ²⁰⁸³ which was a state-owned company and owned by the State of Serbia, in fact?

A. In such cases, either we translate it or the controller from Belgrade, if that was the case, he would have asked a colleague in our department.

Q. All right. I think that's quite sufficient by way of an explanation. I think any reasonable person understands what it means. You said that there was a board in two languages which had been taken down in Urosevac, which is not true. Throughout that time, it always said "Urosevac" and "Ferizaj," the two names. Am I right or am I not?

A. This is pure lie, because since when I worked from 1982 to 1989, it was first "Ferizaj" then "Urosevac." After 1990s, that signboard was removed and then it was written only "Urosevac."

Q. Yes. Only the same sign, it said "Urosevac" and "Ferizaj." But let's leave that for a moment.

At the entrance to your own village, on both sides there were signs both in Albanian and in Serbian, the name of your village; is that correct? There were name plaques in both languages; is that right?

A. Yes, that's right.

Q. Do you know who Njazi Asllani is from your village?

A. No. There isn't anyone by this name in my village.

Q. All right. He is the registrar in your office. Does that jog your memory perhaps?

A. Yes. There is a Njazi in -- I mean, that was -- used to work in ²⁰⁸⁴ my office, but he's not from our village. He worked there in that office.

Q. All right. But he was an Albanian; is that right? He was working in your village, an Albanian?

A. Yes, he was Albanian.

Q. Did he issue documents to you in the Albanian language or in the Serbian language?

A. I can't tell now, but when I needed to have a document, of course I got it in the Albanian language. Then I never had to have a document, to ask for one, so I don't know. But until 1989, all the documents we used to get in the Albanian language.

Q. I'm asking you about after 1989. Did he issue documents in the Albanian language to you or not?

A. I can't say anything now because, as I said, I never asked for a document after 1989, so I don't know what he did. But I don't think that they must have been in Albanian language. Probably they must have been in Serb.

Q. All right. But you know whether in those ten years Albanians received documents in the Albanian language or not. Surely you know that. Just say yes or no. I don't mind what your answer is; just give me an answer. That's the easiest possible thing to ascertain. Just say "Yes, they did," or "No, they didn't."

A. I can testify to what I saw and experienced in my workplace, where the Albanian language was not used. In relation to other documents that you are asking, I cannot testify anything, because I don't deem it necessary to say anything about that here. ²⁰⁸⁵

JUDGE MAY: We have now spent half an hour on this question of language, which is completely peripheral, Mr. Milosevic. Let us move on to another topic.

THE ACCUSED: [Interpretation] The question of language is a basic question, because untruths are being spoken here which claim that there was some sort of discrimination.

JUDGE MAY: In the view of the Trial Chamber, it is a peripheral issue. Now, move on.

THE ACCUSED: [Interpretation] The topic that I haven't exhausted and which was broached by the witness himself is the question of constitutional changes, so I asked what rights were curtailed with these constitutional amendments, to the Albanians and to him. He spoke about language. So was any right curtailed with the advent of these constitutional changes? What did he experience?

JUDGE MAY: You have explored that very fully over the last half hour. Now, you've exhausted the topic. Move on.

THE ACCUSED: [Interpretation] Does that mean that there is no other right which he says he lost with the constitutional changes except for the one referring to language?

JUDGE MAY: That is the one that he's referred to in his examination-in-chief and your questions must be directed towards it. Now, let's move on to another topic.

MR. MILOSEVIC: [Interpretation]

Q. You said that the Albanians who used the Albanian language were deprived of a percentage of their salary. I say that nobody was punished ²⁰⁸⁶ because of that in that way. Do you insist that they were punished? Do you maintain that?

A. Yes, very much so. We were docked certain shares of our salary. We don't have written statements to prove that, but many people have been punished for using their language, in this case, the Albanian.

Q. So that means when you received your salary, they would take away out of your hands that percentage and wouldn't give you a slip of paper saying that you were in fact being penalised; is that what you're saying?

A. You could see the payroll, in which it was reflected that so-and-so was docked his salary, but we haven't kept those papers, because more than a decade has passed. As far as I'm concerned, all my documents were burned, and this happened to many other Albanians, so it was impossible for us to have these papers to show here.

Q. I don't challenge that you can't show documentation, your documents here, but not because they were burned. How come you didn't have any documents, when there was no fire?

JUDGE MAY: He explained that. He said they were burned.

THE ACCUSED: [Interpretation] All right. If that's an explanation that you're satisfied with, then I'm satisfied with that too.

MR. MILOSEVIC: [Interpretation]

Q. In the written statement which you gave, you said that you were penalised, up to 30 per cent of your salary was deducted, and yesterday you said that your salary was docked up to 50 per cent of your salary. So when is it that you're speaking an untruth?

JUDGE MAY: That's not a proper question. ²⁰⁸⁷ It's said that in the statement you referred to 30 per cent, now you refer to 50 per cent. Can you assist us to which of those is right?

THE WITNESS: [Interpretation] No. I didn't say 30 or 50. I only say 15, 1-5 per cent, and up to six months. I said over a period of up to six months. This is what I said.

MR. MILOSEVIC: [Interpretation]

Q. It says here: "If, for example, an Albanian would write out a ticket for a passenger with the name of the station of departure and destination in the Albanian language, 30 per cent would be deducted from his salary." That is what it says in your written statement, what I read out. Yesterday you said 50 per cent, and that is what it says here on the computer. So one of the two is obviously not true, and I say that both those facts are not true, but you gave two completely different statements.

But never mind. We're not only talking about that. You said yesterday, as we're discussing the differences of this kind, you said yesterday that the Serbs set fire to four houses on that particular day, and it says so in exact terms here.

JUDGE MAY: Which day are we dealing with?

THE ACCUSED: [Interpretation] Let me just find it. Just a moment. Although I assume that you all have this. Just a moment. In the first witness statement, one, two, three, four, five, the sixth paragraph, at the beginning, it says:

"On the 5th of April, 1999, at around 9.00, the army, while ²⁰⁸⁸ leaving the village, set fire to two houses which it had previously taken over."

That is what it says in the statement. And yesterday your witness, or rather, the witness of this indictment, said that four houses were set on fire. Probably --

JUDGE MAY: Now, there is a divergence between your statement about the 5th of April when you refer to two houses being set on fire and your evidence yesterday when you said there were four houses. Now, can you clarify that, please?

THE WITNESS: [Interpretation] Yes, I can. I don't think it's very important, because that was only the beginning that I said. Then afterwards, many other houses were burned. It may happen that I gave the wrong date about the first four houses that were burned.

JUDGE MAY: Yes.

THE ACCUSED: [Interpretation] Well, I think that you have some of your own Rules and Regulations and obligations to take measures with respect to false testimony, which is quite obvious here.

JUDGE MAY: No, Mr. Milosevic. That is a comment. Now, if you've got any other questions of this witness, ask them.

THE ACCUSED: [Interpretation] Of course I have. I've just begun.

MR. MILOSEVIC: [Interpretation]

Q. I did not understand what you actually noticed with respect to the movement of trains, the passage of trains. Yesterday, when you were examined in chief, you said there were some sort of irregular train running, so an extraordinary, irregular train which was not on a regular ²⁰⁸⁹ line according to the timetable and log book, traffic logbook, and so it did not stop at Urosevac. And then you say, according to the timetable, it should have stopped there.

Now, my question for you is the following: How could it have -- should have stopped there, according to the timetable, if it wasn't a regular running train on the timetable itself? How do you explain that?

A. Of course I can explain. If Your Honours allow me, I need a little bit more time, if you think. Please tell me. You give me back the log and I will explain to you.

JUDGE MAY: Yes. Let the witness have the exhibit, the log.

JUDGE KWON: Let the witness have the original.

MS. ROMANO: Yes, I have.

JUDGE KWON: That's better.

THE REGISTRAR: The log is Prosecution Exhibit 63. It was already exhibited yesterday.

THE WITNESS: [Interpretation] We took the example of train number 37893. This train is a normal train, a regular train that passes through Ferizaj. When there is -- there are more passengers, they add a "3" number. This is, this train should run according to the regular train. That is, this train should normally have stopped in Ferizaj because that part, "7893," stops. So when we add the "3," it is almost the same train, and it should have stopped.

MR. MILOSEVIC: [Interpretation]

Q. But it did not stop. So how come? What happened?

A. It didn't stop because it was full with passengers, and then it ²⁰⁹⁰ left in the direction of Han i Elezit.

Q. So it didn't stop because there was no room in the train; is that right? It was full?

A. That's right. It was full.

Q. And who decides whether a train is going to stop or not; the stationmaster or the signalman, the dispatcher? Who makes that decision?

A. Usually the decision can be made by the most important authority for Fushe Kosove, that he should order that the train mustn't stop. But in extraordinary cases, even the dispatcher might do that.

Q. And in public traffic, urban traffic, when a bus is full, it doesn't stop at a bus-stop, for example. What seems to you to be out of the ordinary there?

A. But in the case of train circulation, there isn't such a rule. Please. The train has to stop, no matter if there is room or not. It is a rule. I think everybody should know that. Everybody has taken a train. This is the case everywhere, in the West countries and everywhere.

Q. All right. The point here is that the stop -- sorry, that the train didn't stop at your station; is that right? Am I to understand what you're saying like that? And you said that when there was no room in the train, when the train was full, then the stationmaster would call the bus company to help with the transport of the passengers. Was that standard practice?

A. Normally this was not a standard practice. We didn't have such instances before, but that happened that time when people were transported ^{2091 BLANK PAGE 2092} even by trucks, lorries, or by freight carriages, which was not -- which was a very unusual thing. My point was that unusual things kept occurring then.

Q. All right. You said that you heard from the priest about the killing of three persons.

A. Yes, that's right.

Q. In your first statement, you said you heard about it on the 14th. Yesterday, you said you heard about it on the 13th. But at all events, you did not see those bodies, and you did not see anybody actually being killed; is that right?

A. It's true that these three killings I didn't see myself, but the fact is that people were killed on the 13th at about 8.00 p.m., and I found out about that on the next day, on the 14th, at about 10.00. It is likewise true that when we did return to the village, we found new dead bodies. Not the three ones that I mentioned earlier but four other persons. They were thrown in a well or a ditch maybe. Then we found the remains of two elderly people who were burned in their homes, who were my neighbours. But the bodies of the three that were killed on the 13th of April, we have never found them to this day.

Q. How come in the written statement you gave you didn't mention these new bodies?

A. In my written statement, which I gave in April 1999, I have mentioned only three people who were killed and four who were taken ransom, hostage. But after our return from -- to the village, we found the bodies of these four people who were taken hostage. And then I can ²⁰⁹³ give you even the names of those two who were burned in their own home. And I couldn't give them in my first statement because this -- I found this out when I returned home.

Q. When you returned home. You mean after the war; is that right?

A. Yes. Yes, after the war.

Q. So that means that when you returned, you didn't find any fresh bodies that day, but you heard about it when you returned home after the war; is that right?

A. [No audible response]

Q. In your written statement, you emphasised, in passing but you did write it specifically, that you saw two truckloads of Serb families: men, women, children.

A. Yes, that's true. I didn't say they were full of people, but I said many -- I said I saw some Serb families getting off the train and coming to Ferizaj. This is the truth.

Q. It says in your statement that you said they were jam-packed, overfull, jam-packed with Serb families.

A. I don't know what is written there. I know what I said and what the situation was like in reality.

Q. Well, there's an old well-known proverb that when you don't tell the truth, you have to remember minutely what you said beforehand.

JUDGE MAY: Is that a question?

THE ACCUSED: [Interpretation] Yes. I'm asking whether he knows of that proverb, a proverb to that effect.

A. I don't know that old proverb, but that does not apply to me. ²⁰⁹⁴ That might apply to you, the accused.

MR. MILOSEVIC: [Interpretation]

Q. I'm asking you. I just wanted to know whether you had heard of the proverb. But anyway --

JUDGE MAY: You got the answer. If you ask questions like that, you must expect answers like that. They're not helpful. Now, let's move on.

MR. MILOSEVIC: [Interpretation]

Q. So you saw Albanians on the move and Serbs on the move, in trains in Urosevac, both Albanian families and Serb families. You claimed that Albanian families were fleeing out of fear from the Serbs. Do you have any assumption as to why the Serb families were fleeing?

A. I don't know when they went away. What time are you talking about here?

Q. I am talking about these Serb families that you talked about. I'm only talking about what you talked about in your written statement.

A. Those Serb families didn't leave. They came to Ferizaj, I said. They got off the train and remained in Ferizaj. They didn't leave Ferizaj. I can't tell you from where they came, but it was from the direction of Fushe Kosove maybe, whereas the Albanian families left out of terror exercised against them by the Serb police and army.

Q. Well, according to your testimony, there weren't any such measures in your village except that you had heard that three men had got killed. Is that right or is that not right?

A. Yesterday I was -- we passed very quickly over these events that ²⁰⁹⁵ happened in my village, and I was therefore unable to speak in greater length about what happened then in my village. But since I have the opportunity to answer to your questions, then I can explain to you, if Your Honours allow me, to tell you what real terrible things happened to us during all that time, and the accused can get convinced that it was -- whether it was a right decision to leave the village or not.

JUDGE MAY: Just deal with it briefly.

A. When the Serb Chetnik forces occupied our village, it was really horrible just to see, to look at them, Distinguished Honours, let alone to have anything to do with them. They were bloodthirsty people. You could see in their eyes their thirst for blood. They kept in their hands knives stained with blood. Only to look at them, it was terrible, let alone to have anything to do with them or to stay with them in the village. Then they started to loot and raid the village all over. First they took away all our vehicles of all the inhabitants of the village and they drove around the village with our own very vehicles, showing off. Then they started to rob us of our money. Every day they came and asked us to give them money, and you had to give them money, because they were armed, and we have given them money.

Then our sons, for nine or ten days, dug trenches for them, for those Chetnik forces, and they maltreated them every day. The accused would say that I was not there to see it, but my brother was there, and many other witnesses, who can testify to what these young people went through in the course of their labour for those forces. They were threatened every day that after they finished digging the trenches, they ²⁰⁹⁶ would have to kill them, because this is what they did in Bosnia too, they were telling them. You can imagine yourselves what position we were in at that time, when every day we heard threats that, "You are going to get killed tomorrow."

JUDGE MAY: Yes. Thank you.

MR. MILOSEVIC: [Interpretation]

Q. How come you said none of that yesterday or in any one of your written statements, what you've been saying just now?

JUDGE MAY: We need the statements. Can we have the copies, please. There's going to be cross-examination about it.

MS. ROMANO: Yes, Your Honour. There are two statements.

JUDGE MAY: In fairness to the witness, he wasn't asked about this yesterday, but you can ask him, Mr. Milosevic, about the statements. What's said, Mr. Bucaliu, is that there is no mention in your statements of this other maltreatment of villagers by the paramilitaries. Can you help us as to that?

THE WITNESS: [Interpretation] Yes. Distinguished Honours, I may have not said it in my statement, and I don't know how that has come about because I have said all these things even when I was giving the statement. Maybe they haven't written it down because then the statement would have become very lengthy. I don't know. But this is what I said even during the time I gave my statement. I don't know if they have taken down all of them.

MR. MILOSEVIC: [Interpretation]

Q. All right. This explanation is sufficient. Let's go back to ²⁰⁹⁷ yesterday, to what you said yesterday. You explained that they had forbidden you any kind of movement in the village except from 7.00 a.m. to 7.00 p.m. and that they told you that at no cost should you leave the village, and you said that you were very frightened. How did you leave, then? How did you leave the village when they had forbidden you to leave the village and when they did not allow you to move around from 7.00 a.m. to 7.00 p.m., only in that period?

A. I explained that. On the day that the Chetnik forces came, they showed us the rules of how we could go out from our homes. I said that they told us that we could leave the village only with their own special authorisation, and they had closed both exits to the village. Then a Chetnik called Milja, another called Musa, who were the most active in our village, they told us that, "You cannot leave the village because if the infantry troops of Clinton come to these lands, you will become our shield and you will fight together with us." That is, they wanted to use us as human shields or as cannon fodder, as we say. And this made us even more afraid, fearing that one day they would really mistreat us and use us for that purpose. But we stood there, nevertheless, until after all what we went through, until the first killings came. If the accused denies that, I don't know. He can do what he wishes.

Q. I am putting the following question: How did you leave the village if it is true that they had forbidden you to leave the village? How did you manage to leave the village, all of you together, as you had put it?

A. I have explained the way how we managed to leave. I don't want to ²⁰⁹⁸ repeat. If they saw us or not, if they wanted to prevent us or not, I can't say, but we went, as I said, through that valley. Maybe they were not very close, positioned very close to the place from where we left. Maybe they didn't see us leave, because the emptying of the village lasted only 30 minutes.

Q. All right. That assumption that they did not see you is probably probable. Is it your claim that you did have paramilitaries in the village?

JUDGE MAY: Yes, he said that.

THE ACCUSED: [Interpretation] All right.

MR. MILOSEVIC: [Interpretation]

Q. Who is this Mile person that you've mentioned as somebody I cannot make out? Who is this person? What's his role? Who is this man anyway, this man you've been calling Mile? In all these events, that is.

A. Mile and Musa were two rabid Chetniks and I know that they were their names. This is all I know. But I also know our neighbour Milan Rudovic was always seen in their headquarters. He was in uniform and kept going to the headquarters of the Chetniks. And he may know better who Mile and Musa were. We couldn't know more than that, than the pseudonyms they used.

Q. And who knows better who Mile and Musa were? I did not quite understand what you've been saying.

A. The person who may give you better information is my neighbour Milan Rudovic, who was wearing uniform, Serb uniform, and who used to stay with them all the time at their headquarters. ²⁰⁹⁹

Q. Did you believe these claims? Did you believe that they wanted to use you as a human shield?

A. When we saw what people they were and how they were behaving with us, we believed everything they said, everything bad that they said, with the exception of anything good.

Q. But according to what you said, nobody stopped you, nobody directed you in a certain way, nobody made you go in a certain direction. So is that correct?

JUDGE MAY: It's right. Just say yes.

THE WITNESS: [Interpretation] Please, I need to explain. I just can't say no or yes.

JUDGE MAY: Mr. Bucaliu, we've heard your evidence on the point. It's right, isn't it - you haven't said otherwise - that nobody directed you to go to a particular place when you left the village?

THE WITNESS: [Interpretation] Unfortunately, nobody told us where to go, but they threatened us at --

JUDGE MAY: No. We'll get on more quickly if you just answer the questions shortly.

THE ACCUSED: [Interpretation] Can we go on?

JUDGE MAY: Yes.

MR. MILOSEVIC: [Interpretation]

Q. You worked in Slovenia as a signalman, and you came to Urosevac in the late 1970s, and then you continued to work as a signalman; is that correct?

A. Yes. I did my school in Slovenia and worked there for one and a ²¹⁰⁰ half years. Then in 1982, I started to work in Ferizaj train station.

Q. In 1987, you caused an incident when you let a fast train go by and it hit a freight train. There were many injured persons, and you were suspended from work for six months. And then instead of being dismissed altogether, you were returned to work but you got a job as a cashier in the commodities department, and you remained in that position before the war and during the war; is that correct?

A. Yes. It's correct that there was an accident in 1987. There were accidents, but this happens everywhere. It's nothing extraordinary that happened. I went even to -- I was taken to police stations and taken to a detention cell, but after that, I was never punished. And that is true that I continued my work for a time as inspector of goods or commodities as then as a cashier.

Q. And you stayed there all the way up to April or, as it says here, May 1999?

A. I worked there until 24th -- sorry, until -- yes, on the 24th of March in my workplace, but in the station, I reported until the 14th of April.

Q. And do you know the names of Mahmut Shabani, Muhail Rexhepi, Medi Hoxha [phoen], and Yseni Ysen who worked with you?

A. Yes. They were my colleagues. I know them.

Q. Precisely at that time these four men reported to Ilic, the trainmaster, and informed him that they had orders from the KLA to go, with the explanation that they would return earlier in that way. However, as opposed to them, I here have a note that you did not report to the ^{2101 BLANK PAGE 2102} trainmaster as they did.

JUDGE MAY: There are two questions there, two questions there. Do you know about the four men it's alleged having orders from the KLA to go?

THE WITNESS: [Interpretation] This is ridiculous, distinguished Honours. I can't say anything to testify that -- to that because it is equally true in their case, like it was in my case. That is, we were ordered not to work any more there. It's not true at all what the accused is alleging here.

MR. MILOSEVIC: [Interpretation]

Q. So it is your claim that you did not get orders from the KLA to leave the workplace?

A. No. There wasn't any KLA in that part where I lived, and I've had never any contacts then with KLA, and I never got any orders from them.

Q. All right. And do you know Ramadan Muharemi who worked there as well?

A. Yes. He was also another colleague.

Q. He also reported to the trainmaster, and he said that due to KLA pressure, he had to leave. He had to sell his house in the centre of Urosevac and to move, because he was not an Albanian, he was a Roma. He did that indeed, and he moved to Subotica where he's been living until the present day, and he can confirm this. Do you know about that?

JUDGE MAY: What is the question? Do you know anything about that?

THE WITNESS: [Interpretation] I know Ramadan Muharemi. He was a ²¹⁰³ cleaner in our station and he then worked there. He was in Ferizaj after the war for a long time. Then he why left his house -- it was three or four months after the war. He stayed in the Ferizaj station for only three or four months. I don't know where he is now. If he said that he has left the village for this reason that you are claiming, I don't know about that.

MR. MILOSEVIC: [Interpretation]

Q. Is it correct that you held meetings at the warehouse of the train station, and when salaries were handed out, you collected money for the KLA? Yes or no.

A. No, that's not true.

Q. And is it correct that you were involved in smuggling foreign exchange, and that is to say, to and from Slovenia, relying on your old connections?

A. That's not true either. The accused is trying to catch at a straw because he's seeing that he's drowning. So he is saying things here which I don't think matters for this Court either.

JUDGE MAY: Now, Mr. Bucaliu, it's for us to decide what matters. Yes, Mr. Milosevic.

MR. MILOSEVIC: [Interpretation]

Q. Do you know a person by the name of Ymer Neziri?

A. Ymer Neziri, yes. He's my neighbour.

Q. Together with Ymer Neziri, you participated in the demonstrations of 1981.

JUDGE MAY: No. We've been back to 1981 before. We're going to ²¹⁰⁴ move on.

THE ACCUSED: [Interpretation] I'm not dealing with 1981 in relation to the demonstrations. I want to ask him whether he remembers that together they were breaking windowpanes on the department store called Napredak. That is the department store from Djakovica and Urosevac. And they also took goods from that store, and there is a document to prove that.

MR. MILOSEVIC: [Interpretation]

Q. Do you recall that?

JUDGE MAY: You are making a series of allegations against this witness which you will have to produce some evidence to support. At the moment, they are pure allegation.

We will allow the witness to answer this one and shortly. It's alleged that you were breaking windows in a demonstration in 1981. Were you?

THE WITNESS: [Interpretation] I said that at that time I was in Slovenia, in the school. You can say what you want.

JUDGE MAY: Yes. That's the answer. He was Slovenia at the time. Now, let's move on to something more recent and more relevant to the indictment.

THE ACCUSED: [Interpretation] Everything is related to the indictment, everything that has to do with the witness, and the validity of his statements, the reliability and authenticity of them.

JUDGE MAY: I'm not sure we're going to be helped, even if he was breaking windows in 1981, in deciding whether he's telling the truth about ²¹⁰⁵ events in 1999.

THE ACCUSED: [Interpretation] I'm talking about credibility, please.

JUDGE MAY: Yes, I'm talking about credibility too. It's a question what is relevant to credibility.

MR. MILOSEVIC: [Interpretation]

Q. All right. The first cousin of witness Bajram Bucaliu, who is also called Bajram, do you remember when he was caught in 1987 at the Bulgarian border and 300 kilogrammes of drugs were found in his van?

JUDGE MAY: What is the relevance of what his cousin was doing, even if his cousin was doing it? It's totally irrelevant.

THE ACCUSED: [Interpretation] It is a well-known fact that the entire family was involved in smuggling tobacco, weapons, and drugs.

JUDGE MAY: Very well. You can put that to him.

THE WITNESS: [Interpretation] These are genuine lies. That's not true at all, that my entire family, or even members of it, were ever engaged in illicit activities or trade.

MR. MILOSEVIC: [Interpretation]

Q. Is it correct that your father was in prison in Nis because of smuggling?

JUDGE MAY: No. I think this is enough. Attempts to smear witnesses do not help the Court. They don't help anybody. If you've got an allegation to make involving the witness's credibility, then you can put it, as you have done, but to start attacking his own family is not relevant. ²¹⁰⁶

THE ACCUSED: [Interpretation] I was just asking questions which relate exclusively to the facts and not to any kind of value judgement, but let's move on from that topic to another area.

MR. MILOSEVIC: [Interpretation]

Q. As the witness explained the reasons which, according to him, were the reason for the departure of citizens from his village to Macedonia, I have the following question: In the neighbouring villages, that is to say, in that area, the neighbouring villages of Talin -- Talinovac [phoen], Nekodim, the village of Varos, Tankosic, Gadnje, Trebno, Plesina, Mirosavlje -- I have been asked by the interpreters to speak more slowly. So the villages of Talinovac [phoen], Nekodim, Varos, Tankosic, Gadnje Trebno, Plesina, Mirosavlje, all of mixed composition, and the village of Rajovica, where the inhabitants are purely Albanian, the inhabitants, throughout the time, remained in their villages; is that true? Is it true that the villagers in all these surrounding villages did not move but stayed in their villages throughout? Is that right?

A. I don't know about all these villages you enumerated. I knew about the neighbouring villages which were all emptied of people. I said Kamoja [phoen], Sore [phoen], Bibaj, and Varosh. These four villages, they were evicted from Albanians [sic].

Q. NATO bombed a large storehouse area belonging to your cousin, where your father and brother were employed; is that correct or not?

A. NATO had struck that object after we had left the village, because in that place, military forces were deployed, which were there even when we were there, but the bombing came after we left the village. And not ²¹⁰⁷ only that facility was bombed, but all other facilities. Two tanks were burned and a truck was burned from this bombing.

Q. Do you know why a warrant was issued for [indiscernible] Bucaliu, your first cousin?

A. I am Bajram Bucaliu. I don't know what was issued. I don't understand that.

Q. Have you got a first cousin whose name is Brahim Bucaliu, that is to say, your uncle's son on your father's side?

A. Yes, I have a cousin called Brahim, and there's another one called Brahim too.

Q. I'm referring to the person who has a house between the railway station and bus station. Have you been able to identify him now?

A. Yes, I can identify him, yes.

Q. Do you know why a warrant was issued?

JUDGE MAY: Do you know -- first of all, do you know if a warrant was issued?

THE WITNESS: [Interpretation] I don't know anything about a warrant. How could I know anything about that? I don't know anything about that, no.

MR. MILOSEVIC: [Interpretation]

Q. All right. I won't pursue this and go on with my questions relating to that.

Does your family own a gas station on the Urosevac-Pristina road?

A. Yes. Not my direct family, but my relatives or cousins do have a gas station. Two cousins, in fact. ²¹⁰⁸

Q. And do you work at that gas station yourself?

A. Yes, I work there now at the gas station.

Q. Is that a gas station that was built by your cousin Nazmija [phoen]?

A. Yes. Muhamet, his son. It's his son, Muhamet, who built the gas station.

Q. Was the gas station built on the property of a Serb called Milivoj Injic [phoen], on his land?

A. That's not true. He built it on his own property, on property that he had bought with his own money.

Q. Do you remember the bombing when Bajram Reka's house was hit? And otherwise, next to the house of Bajram Reka, you have a field that you own, so I assume that you know about that.

A. Around the house of Bajram, there's no land, but I know that the house of Bajram Reka was used as a centre for the Chetniks. Bajram Reka and all his family and all the neighbours were thrown out of their houses. And it was there that the Chetnik staff had their centre. Tanks didn't destroy the house. They were near the building, but the house itself was not. It was the tanks which were burned. The house itself was not touched. And a truck was destroyed from the bombing.

Q. And do you know whether anybody fell victim to the bombing? Were there any casualties?

A. I don't know about that. No Albanians, at least, from our village. As to the forces, the military forces, I don't know. I couldn't tell you. ²¹⁰⁹

Q. You claim that none of the Albanians at the time were either wounded or injured or killed, that there were no casualties. Is that what you're saying?

A. That's true. We weren't there -- the houses weren't bombed. That happened after we left. Everything took place after we left.

Q. All right. But when you returned, do you know that all the Serb houses were set fire to except the house belonging to Dragutin Veljkovic [phoen]?

A. Yes, they were burnt. They were all burnt down, both Serb and Albanian houses, yes.

Q. Did you go, just before the aggression, to the village of Biba at Osman [phoen] and Nabi Biba [phoen] to attend meetings there?

A. I don't know what time you're talking about, but I was an activist of the LDK, and it's possible that I was there.

Q. And the meetings that you held, were they in connection with the KLA activities?

A. No. We had nothing to do with the KLA. We didn't even know anything about the KLA at the time, and we had nothing to do with their activity, with such activities.

Q. All right. Let's go back to the end of your statement, what you said at the end, and your departure for Macedonia. Is it true that in the meantime you went back from Macedonia, back to Urosevac, for short periods, at certain intervals, before you finally went back after the war?

A. Well, the first day we tried -- on the 15th of April, we tried to cross the border, but we couldn't. When we tried to get to Macedonia on ²¹¹⁰ train and were returned to Ferizaj, that's the only time. That's not true what you're saying.

Q. I'm asking you: Did you go back now and again? I don't claim that you did; I'm asking you.

A. No, that's not true. I didn't return. I was in Stankovec for about a month, which the accused said didn't exist, but I was there.

JUDGE MAY: Deal with it shortly, if you would. The answer is you didn't go back.

THE WITNESS: [Interpretation] No, I didn't, because --

JUDGE MAY: It doesn't matter about the reason. Yes, Mr. Milosevic.

MR. MILOSEVIC: [Interpretation]

Q. Did anybody take away your documents when you went to Macedonia?

A. No, they didn't take away our documents.

Q. Did anybody direct you to go to Macedonia, tell you to go?

A. They didn't tell us to go, but we went to Macedonia. We were accompanied by the police, by the police and the army. They didn't say to go to Macedonia, but they accompanied us, escorted us to the border.

Q. I understood that they drew your attention to the fact that you shouldn't leave the railway line because it was -- that you shouldn't go along the railway line because it was mined and somebody might be killed. Not to go outside the lines.

A. Yes, that's what they said. We were afraid. That's what they said, at any rate. I don't know if it was true. So we kept on the rail track. ^{2111 BLANK PAGE 2112}

Q. Well, it was undoubtedly true. But do you consider that they helped you in that way or that they were doing something evil towards you by telling you not to leave the railway tracks and not to lose your lives by doing so?

A. I don't know what they were thinking, whether they were trying to save our lives or not. I told you what happened.

Q. Well, I'm asking you how you -- what your understanding was of this gesture on their part. How did you understand it?

A. I don't know what I understood. They just said, "Keep on the train tracks." And I know they said that. What they thought, what their reason was, I don't know. We didn't think about that.

THE ACCUSED: [Interpretation] I have no more questions.

JUDGE MAY: Yes, Mr. Tapuskovic.

MR. TAPUSKOVIC: [Interpretation] Thank you, Your Honours. As Slobodan Milosevic today brought up the statements himself and presented them to Witness Bucaliu, it will be easier for me to focus on matters which he did not touch upon when asking the witness questions. So let me start off with the list, the list that was brought in and presented as an exhibit.

Questioned by Mr. Tapuskovic:

Q. [Interpretation] Mr. Bucaliu, on the 24th of March, it says that you were not working, as you said. Is that correct?

A. On the 24th of March, I was working.

Q. That's why I'm asking you. So on the 24th of March, you were still working. Now, on this list, do you see that half the column has not ²¹¹³ been filled in and that on that day particularly not half the trains passed by. At least, there were half as many trains as there were the previous day, because four, six, seven trains passed by on that particular day, whereas the previous day many more trains passed by. How do you explain that?

A. On the 24th, trains were moving, but on the 25th, there were no trains.

Q. How do you explain that, that after the bombing started, the NATO bombing started, not only did the number of trains not increase, not only was there more in circulation, but on the 24th and 25th of March, on one day there were no trains at all, and on the 24th of March there were half as many trains as usual. How do you explain that? And you said that the number of trains increased.

A. It's possible the 24th, 25th, up to the 29th, it's possible that there were no trains. But from the 29th on, there are no -- lots of notes.

Q. As far as I can see here, and I'm looking at it, I have the 29th in front of me and that's where I'll end, the 29th, and then it goes on to the 1st of April, there were trains, and you said, depending on what the last figure was, whether it was an odd number or an even number, that that meant that the train was moving in one direction or in another direction. Now, when the last digit is 0, what does that mean?

A. The 0 is an even number. It's an even number.

Q. Now, if you look at the list dated the 29th of March or any other after that date, you will see at that there are about seven, seven trains ²¹¹⁴ moving in one direction and seven trains moving in another direction. So the number of trains was equal going to Macedonia and from Macedonia.

A. But that's normal. Every train that goes to Macedonia has to return. They can't all stay in Macedonia.

Q. All I'm saying is that there were no upsets in the timetable, that everything took place regularly, that trains moved regularly in both directions, as stipulated in the timetable. Is that correct?

JUDGE MAY: So we can understand it, this is -- just a moment. So you're putting that up until the 29th of March or on the 29th of March, the timetable was working normally.

MR. TAPUSKOVIC: [Interpretation] And onwards. It was working normally afterwards too. Trains moved in a regular fashion after that date, going to and from.

JUDGE MAY: The witness's evidence was there were more trains on the 3rd of April. That's the date that he dealt with in his evidence. What is suggested is that you're wrong about that and there were -- the trains were running normally. What is your answer to that suggestion?

Less the witness answer.

THE WITNESS: [Interpretation] I can only tell you that the trains up to -- the trains were running regularly. They were going in both directions. And after the beginning of the bombings, there were more trains, and they were exceptional trains running an exceptional -- an exceptional schedule. There were also regular trains, normal trains, but most of them then were irregular trains. That's what I said yesterday. I ²¹¹⁵ said there were regular and non-regular trains.

JUDGE KWON: Just a second. Because you are looking at the date of 29th of March, I don't see any trains in this log between 25th of March to 28th of March. Am I right? Could you check up this matter?

THE WITNESS: [Interpretation] Yes, you're right. There were no trains from the 25th to the 28th.

JUDGE KWON: Thank you.

MR. TAPUSKOVIC: [Interpretation] That's what I wanted. I wanted to reach the 3rd of April, in fact, and I have it here in my hand.

Q. If 0 is an even number, as you yourself say, then seven trains left in one direction towards Macedonia and seven trains returned from Macedonia on that day in April. That's what I wanted to hear. I wanted to hear the witness's explanation of that. Nothing changed.

A. No, it changed. I said on the 3rd of April, I told you how they passed through -- how the trains passed through Ferizaj. They didn't stop there except for the train 37897, which stopped in Ferizaj, and the other train that returned from Elez Hani. Aside from that, all the other trains just passed through Ferizaj.

Q. No. My question wasn't that. Yes, I know. You explained that. You explained that. All I'm doing is noting that you said that nothing changed. There were no irregular trains. Seven regular trains went in one direction, and according to the timetable, they returned from the opposite direction as the timetable states.

JUDGE MAY: His whole evidence is that the trains were irregular. That's the whole point of his evidence. ²¹¹⁶

MR. TAPUSKOVIC: [Interpretation] That's why I'm asking, Mr. May -- Judge May. That's what I'm asking, because --

JUDGE MAY: Why are you challenging it? What is the basis of your challenge to this evidence which he's given?

MR. TAPUSKOVIC: [Interpretation] I'm not challenging it. All I'm doing is saying that I think the documents show that for the most part everything ran regularly, that a certain number of trains run in one direction, a certain number of trains ran in the opposite direction.

JUDGE MAY: [Previous translation continues]... more quickly. We'll ask the witness.

Were the trains running regularly that day or not, according to the document which you have in front of you?

THE WITNESS: [Interpretation] If we're talking about this date -- there's a lot of dates, but let's take this one here. According to the list, there is no regular train. They are all irregular trains. According to the numbers we have here, you can see that they are irregular trains.

MR. TAPUSKOVIC: [Interpretation] Thank you. I won't pursue the matter. I think there is a logic that has to be applied, but I'll leave it to the Trial Chamber to assess.

Q. What I wanted to ask in connection with everything that was said here with respect to the events at the railway station, I won't go back to the language problem, the language issues, but in your previous statement, that is to say, in your statement dated the 16th of -- no, the 29th of August, 2001, you said that the money fines were prescribed generally ²¹¹⁷ because for the train tickets, the name of "Urosevac," the place, was written in the Albanian language, "Ferizaj," as the ultimate destination; is that right?

A. I said there were times when it was -- it was questioned why was it written "Han i Elezit" on the ticket. A friend of mine was punished for that because he wrote it "Han i Elezit."

Q. I wanted to ask you something else in fact. You work at the railways, in the railways, and you've been working there for many years. You know the problems that existed. When you leave Kosovo, it's not a large, a great distance. When you leave Kosovo -- the Albanians travelled a great deal. They went to Bosnia and Montenegro, passed through Serbia on their way to third countries, and they used that main road a lot through Croatia. And you know that very often there were problems on the railway lines, at railway stations, that when you write "Ferizaj" most people, not in Kosovo but further on in Yugoslavia, which existed until 1990, that there were problems because they couldn't tell where the people had -- where the point of departure was for these passengers, that they were even asked to leave the trains because of that. Is that true? Did you know about that problem?

A. I don't know anything about that. There may have been certain problems. I think the inhabitants of Yugoslavia itself didn't have any problems whereas people abroad, from abroad, might have had a problem with that.

Q. Let me now go on to the statement. Slobodan Milosevic has focused on the statement. But I should just like the witness to explain one ²¹¹⁸ point. He said on the 2nd of April, when the army arrived, when the army came to your village, Witness, and you explained that there was an officer, that you had a conversation with him, but in your statement, you told the investigators that, "The officer had said that they were the regular army and that we should not flee." Was that it? Is that what he told you, that you shouldn't flee because they were the regular army? Is that what he said to you and is that what you've said in the statement?

JUDGE MAY: Mr. Tapuskovic, so that we can follow your examination, will you tell us where this is; which statement and where we can find it.

MR. TAPUSKOVIC: [Interpretation] The statement is dated the 24th and 25th of April. It is his first statement given in 1999. And on page 1, paragraph 4 is what I'm referring to. In the English version, the one you have, that is --

JUDGE MAY: Yes, we have it.

MR. TAPUSKOVIC: [Interpretation]

Q. Is that correct, that he said, "We are the regular army, and you needn't flee"?

A. Yes, that's true. That's what he said. But later, we found out that that was not the case at all. He did say that. He said, "There's no reason to leave. We are the army. Don't be afraid of us." That's what he said.

Q. But you also said over the past two days that you left but that nobody told you to leave, that nobody demanded that you leave.

A. Yes. We stayed in the village for several days after that. ²¹¹⁹

Q. And then you decided to leave the village?

A. Yes, on the 14th of April.

Q. Thank you. Now we come to your second statement, dated the 16th of June, and the English version -- in the English version, it is on page 4. Page 4. And the date is the 15th of April, the date that is mentioned. And you say, "That first morning, the 15th of April, 1999, once again at the railway station I saw the police." Is that right? Is that what you said? Is that right?

A. Yes, that's right.

Q. Next you go on to say, "They were looking at what was going on and tried to introduce a certain amount of law and order." Is that what you said?

A. I didn't think they were -- say that they were trying to introduce law and order. I didn't see anything particular. I was just there.

Q. That's what it says in your statement.

THE INTERPRETER: The interpreters apologise, but they haven't got a copy of the statement.

MR. TAPUSKOVIC: [Interpretation]

Q. All right, but what is important is the following. You say: "At the station, there were about ten policemen. I didn't notice whether they were preventing people or directing people in which direction to go."

Is that correct?

A. They didn't tell anyone where to go. They -- people were leaving. It didn't matter where they went. ²¹²⁰

Q. Thank you. Two paragraphs lower down -- so they didn't address you in any way or make any suggestions to you, did they?

A. Not at the station in Ferizaj, no.

Q. Thank you. Now, two paragraphs further down, you state the following --

JUDGE MAY: I'm going to interrupt you. It's now a minute past 11.00. Are you going to be many minutes longer?

MR. TAPUSKOVIC: [Interpretation] Two more minutes. Two more minutes.

JUDGE MAY: Very well.

MR. TAPUSKOVIC: [Interpretation]

Q. Two paragraphs lower down, you state the following: Vule, you mentioned Vule. Who is Vule?

A. Vule was the station head at Han i Elezit.

Q. In your statement, you said: "Vule offered to see my family and myself off to the border, to escort me, together with the policemen, but I didn't take up his offer." So he offered to transfer -- transport you and your family, but you didn't accept the offer because it wouldn't have been honourable for you to leave your fellow villagers; is that right?

A. Yes, that's true, because the station head at Elez Han, after the first day when we couldn't get across the border into Macedonia, offered to help me and my family, but because I was with all the other people who were in the train, I didn't want to accept the offer to cross the border just with my family. And then the next day, we tried and got across.

Q. Thank you. ^{2121 BLANK PAGE 2122}

MR. TAPUSKOVIC: [Interpretation] Judge May, with the Court's indulgence, I wanted to clear up one more matter, and it is to be found on paragraph 3 of the second statement, on page 1. After the break, may I be allowed to clarify one more point, after the break, in that second statement of his?

JUDGE MAY: Yes. We'll adjourn now for half an hour.

--- Recess taken at 11.05 a.m.

--- On resuming at 11.35 a.m.

JUDGE MAY: Yes, Mr. Tapuskovic.

MR. TAPUSKOVIC: [Interpretation] Your Honours, I just want to draw your attention to the third paragraph in the second statement, please.

Q. Mr. Bajram, earlier on you said that all the way up to 1998, you did not support the KLA.

A. Yes, that's right, because at the beginning the KLA was something which I knew nothing about, so how could I respect it?

Q. Also you said then: "I thought I could not do anything against Serbia, and I knew that it would be hard to achieve anything without some help from outside the country."

A. Yes, this is what I thought then, because having to deal with such a major military force like Serbia was, I would normally think that we would suffer great consequences.

Q. You said that, "When the bombing started, I changed my position." What did you mean by that?

A. It may coincide with the bombing, but when I saw what one ²¹²³ man -- one human being can do to another human being, from what I saw happen in my village, it was but normal and clear to all of us that if you -- that if we sat cross-legged, we would be killed just the same. Then I had to change my mind.

Q. Before I put another question to you, let me ask you the following: Once, as you said, you helped a KLA soldier when the unit withdrew due to the presence of strong Serbian forces. In which way did you help him?

A. Yes. There was a great offensive of the Serb forces against the KLA positions in the Sharre Mountains, and the KLA had withdrawn. In so doing, the son of a colleague of mine happened to be left behind in the village, and my colleague asked me to help him, to transfer him from -- to take him to another place.

Q. How big was this unit that was retreating? How many men were there in that unit?

A. I have no information about their number. I can't tell you.

Q. All right. Do you know whether before the bombing and after the bombing there were clashes between the KLA and the army and the police of Serbia?

A. I know about that after I read in the information media. Yes, there were clashes.

Q. I'm just asking you the following. That's why I've been asking you all these other questions. Due to these clashes between the KLA - I mean before the bombing and after the bombing - did that lead to the withdrawal of the population, either Serb or Albania, from these areas ²¹²⁴ where the fighting was taking place?

A. That is not true. It's not that it was because of the NATO bombing that we left the village, either we Albanians or Serbs. The Albanians left because of the tragedy perpetrated against them by the Serbs, while the Serbs stayed there and didn't leave. I can testify to this in several ways.

I may tell you -- I may cite only an example when I was in town and the alarm went off about -- for the NATO airstrike, and I could see the Albanians, that none of them left the town because they thought that NATO doesn't shoot at the civilians, doesn't strike at the civilians, and that was the case, while the Serbs left. They tried to find shelter in cellars, in other secure places.

Q. You said in your statement here - I don't want to show it to you yet again - that you and all your people gathered together during the bombing in one house. While the bombing lasted, that is.

A. Yes. We got together in a house. Not because of the bombing but because we felt safer in the face of the Chetnik forces who were in the village. Because if it were for the bombing, there were no place we could go because there wasn't any shelter for us. But the fact is that we felt so secure, so safe that only the fact that when bombing started on the 24th of March it was a great joy for us because it was there that we saw our salvation.

Q. There were power cuts often from the 24th onwards. That's the way you had put it as well.

A. Yes. The power was cut at about 8.00 p.m. and came -- was ²¹²⁵ released in the morning.

Q. Let me ask you one more thing. Did this have anything to do with the bombing?

A. I think that this was so to make it easier for the Chetnik military forces to do what they intended in the night, to circulate freely without being seen. And this is what happened, because the people, the three people, were killed precisely when the power was cut, Emin Zeka and his two sons.

Q. This happened every evening? The first five or six days, the way you had put it.

A. During the time I was in the village, this is what happened. The power was cut off every night.

MR. TAPUSKOVIC: [Interpretation] Thank you.

JUDGE MAY: Yes. Any re-examination?

MS. ROMANO: Yes, Your Honour, two areas just. Re-examined by Ms. Romano:

Q. Mr. Bucaliu, on cross-examination you were asked about the 2nd of April when the army came to your village and they told you that they were the regular army and that you should not flee. But you continued saying that it changed after. What changed?

A. Yes. That situation changed because that army, which described itself as a regular army, started to burn our homes, to raid and loot. So the very army that told us not to be afraid started to do all these things. So they didn't suit action to words.

Q. Mr. Bucaliu, did you see them burning the houses, and did you see ²¹²⁶ them looting?

A. Yes. I saw them burning the houses. As to the looting, I can give my own example. They took away my car, my tractor, and so.

Q. Thank you. Also on cross-examination, you answered that you decided to leave on the 14th of April. Did you want to leave?

A. I don't know if there is anyone in the world to wish to leave his own property, home, everything at the mercy of such Chetniks and to leave his home and his fatherland. If there is someone to think that, to leave his home and go to an unknown destination, I don't think that this can happen. But for us, it was important to save our lives. We didn't feel any pains at leaving behind all our properties. The only concern for us was to save our lives.

Q. So was there any particular moment or particular fact that made you decide that you were leaving at that moment?

A. Yes. As I said during my testimony, apart from various acts of maltreatment, looting and raiding, they started to kill people. That did it, when we heard that people were starting to get killed. Then we decided to leave, because we were afraid that the same thing might happen to us. And this is shown by the people who were left behind, Fehmi Hoxha and his wife, my first cousin, who we found only their remains, when we returned, in their burned house.

Q. Mr. Bucaliu, that applies also to the other people that left with you; do you know if that was the same -- they would have the same fear as you had?

A. Yes. Yes. It was normal for all of them to feel the same thing. ²¹²⁷ When we start to hear that people began to be killed, nobody had anything to say. It was chaos. It was confusion prevailing among all of us. We didn't know where to go. It was quite spontaneous when we started to flee the village on -- through the road that I always -- already explained to you.

Q. Thank you, Mr. Bucaliu.

MS. ROMANO: No more further questions. Questioned by the Court:

JUDGE ROBINSON: Mr. Bucaliu, so I can be clear about this question of the regularity of the movement of the trains, I'd like the usher to give you the logbook and for you to look at each day from the 24th of March until the 3rd of April. And I'd like you to comment on the question of the movement of the trains on each day; how many trains there were, whether the trains were full, matters of that kind which are relevant to the question of the regularity of the movement of the trains. So each day from the 24th of March.

A. As I said earlier --

JUDGE KWON: Could you please put the log on the ELMO, please.

JUDGE ROBINSON: Are you now looking at the 24th?

A. This is the 24th March.

JUDGE ROBINSON: How many trains were there that day?

A. On the 24th of March, there were seven trains running: three in one direction, apparently four in the opposite direction.

JUDGE ROBINSON: And what is your evidence - remind me - as to the average number of trains each day? ²¹²⁸

A. The average number was six: three in one direction, three in the opposite. That was the regular schedule. Six trains over 24 hours.

JUDGE ROBINSON: So on the 24th, we had three and four. And were the trains full, are you able to say, on the 24th?

A. On the 24th, I don't think they were full, because I was at work, and the trains had only two carriages each and were not full. But after the 24th of March --

JUDGE ROBINSON: The next day now, the 25th, and give us the same information on the 25th.

A. On the 25th of March -- there are no notes until the 29th of March. If I may continue from 29th onwards --

JUDGE ROBINSON: Until the 29th, there's no record in your logbook of the movements of the trains?

A. No, there isn't any.

JUDGE ROBINSON: Why is that?

A. I can't be sure why that is, but probably the trains didn't run. I can't remember precisely this particular time, but this is what can be inferred from the log.

JUDGE ROBINSON: Thank you. Let's move to the 29th, then, the next day when you have information. How many trains were there on the 29th?

A. On the 29th, there was only one train. It has come up to Ferizaj station and has returned, from what I see in the log.

JUDGE ROBINSON: Any indication as to how many passengers were in that train? ²¹²⁹

A. There isn't any indication as to the number of passengers.

JUDGE ROBINSON: Let's move to the 30th.

A. On the 30th, there were two trains, I see here. On the 31st, there were more trains.

JUDGE ROBINSON: No. Just -- we haven't finished with the 30th. Two trains in each direction or just one direction?

A. One in the direction of Ferizaj, one in the direction of Fushe Kosove; opposite.

JUDGE ROBINSON: Yes. And the 31st?

A. On the 31st, there were eight trains; four in one direction and four in the opposite direction.

JUDGE ROBINSON: And any indication as to the capacity, the number of passengers?

A. No. Unfortunately, no, I don't see any records here about the number of passengers. As you can see, there are records only about the number of trains that passed through. As to the number of passengers, this is taken down at the departure station.

JUDGE ROBINSON: Let's move to the 1st.

JUDGE KWON: Just a second. In terms of a capacity problem, I'd like to clarify this matter. You testified that the reason the train did not make a stop at Ferizaj station, you explained the reason as being due to the fact that the train was full; is it right?

A. Yes, that is right. When I was there, I saw such trains passing by and not stopping, and they were full of travellers.

JUDGE KWON: According to this log, not only the trains with odd ²¹³⁰ numbers, but those with even numbers didn't make a stop at Ferizaj station. Does that mean that trains from the border were also full?

A. In very rare cases they have been full, as the case was with our train when we returned back from the border to Ferizaj station. But in other cases, the trains that went were full; when they returned, they were empty.

JUDGE KWON: But also some empty trains did not make a stop at Ferizaj station; am I right?

A. Yes. When they returned from Han i Elezit to Ferizaj, they were empty. Sometimes they did not stop, from what I see in the log.

JUDGE KWON: Thank you.

JUDGE ROBINSON: Let's continue. I think we were at the 1st of March -- 1st of April, rather, 1st of April. How many trains were there on the 1st of April?

A. On the 1st of April, there were seven trains.

JUDGE ROBINSON: Give us the breakdown as to the direction. How many in the direction of the border, going to the border?

A. Four went to the border direction, three returned from Han i Elezit.

JUDGE ROBINSON: And is there an indication as to the capacity?

A. No, there isn't any indication as to the capacity. Unfortunately, no such records are ever kept.

JUDGE ROBINSON: 2nd of April.

A. On the 2nd of April, there were 13 trains.

JUDGE ROBINSON: And give us the breakdown as to the direction. ^{2131 BLANK PAGE 2132}

A. Seven in the direction of Han i Elezit and six in the opposite direction, from the border to Fushe Kosove.

JUDGE MAY: Now, can you tell by looking at the log whether these were regular or irregular trains?

A. On the basis of the number of the trains I see here, all these trains were irregular trains, I mean not normal trains, and from the log, not regular -- not a single regular train has run. All have been irregular trains.

JUDGE MAY: Did they stop at Ferizaj or not?

A. On the 2nd of April, only train with number 14[as interpreted] stopped. It came to Ferizaj at 6.08 and left at 15 [as interpreted]. The other trains didn't stop.

JUDGE MAY: And how can you tell that they didn't stop?

A. I can tell from this arrow that you see here, which shows that it has only passed through the station. The train that stops, you see a mark the time when it has arrived, how long it has stopped, and when it has left the station. So this train 4935 [as interpreted] has come at 6.08 and has left at 7.15. And you can see the same with other trains have not stopped at Ferizaj, and we come at another train here that has stopped, and so on.

JUDGE MAY: Yes. Thank you.

JUDGE ROBINSON: And is the situation the same as to the capacity; you are not able to say whether the trains were full on that day, the 2nd?

A. I didn't keep a diary for every day, but most of the time that I was there, and I saw with my own eyes, the trains were full of people. ²¹³³

JUDGE ROBINSON: Okay. On the 3rd, the last day now, the 3rd of April, how many trains were there?

A. On the 3rd of April, there were 16 trains.

JUDGE ROBINSON: How many in the direction of the border?

A. Seven towards the border.

JUDGE KWON: Isn't it 14? Could you count it again?

A. Yes, 14. You're right.

JUDGE ROBINSON: Did they stop at Ferizaj?

A. Only the train number 37897 stopped at Ferizaj, and the other one from the border stopped at 1456, 1458. The other trains didn't stop at Ferizaj at all. There, at the bottom, you can see trains stopping at Ferizaj, et cetera.

JUDGE ROBINSON: Were they regular trains?

A. No, none of them were regular trains.

JUDGE MAY: Did the trains normally stop at Ferizaj?

A. The Ferizaj station is one of the largest stations, and all major trains, all international trains, always stopped in Ferizaj. All passengers trains stopped, because it's one of the biggest stations in Kosova.

JUDGE MAY: Mr. Bucaliu, that concludes your evidence.

MR. TAPUSKOVIC: [Interpretation] One thing was not done, Your Honours. These two statements that were made by the witness to the investigator, they did not receive a number.

JUDGE MAY: Yes. They will be exhibited.

THE REGISTRAR: Witness statement dated 24, 25 April 1999, will be ²¹³⁴ Prosecution Exhibit 64, and the next one Prosecution Exhibit 65.

JUDGE MAY: Yes. Now, what is it, Mr. Milosevic?

THE ACCUSED: [Interpretation] I assume that I have the right to an additional question with respect to what has been said so far after my cross-examination, the re-examination. Do I have that right or not?

JUDGE MAY: No. No, you can't have another question. The only question you could ask would arise from the questions which the Judges ask. If you want to ask anything about the train timetable, you could.

THE ACCUSED: [Interpretation] Only on the basis of what the Judges asked; is that right?

JUDGE MAY: Yes, about the train timetable.

THE ACCUSED: [Interpretation] About what Judge Robinson asked.

JUDGE MAY: Yes, you can ask about.

THE ACCUSED: [Interpretation] We just heard the witness answer that on the 25th, 26th 27th, 28th there were no trains and that on the 29th there was just one train which arrived at Urosevac and returned, went back.

That means that for practically five days there was no communication and traffic towards Djeneral Jankovic. And then the next five days are taken, of which the first two days was regular traffic and in the following three days, it was doubled. Now, if we make an average for the first five days in which there was no traffic and the next five days in which traffic was increased, can we note there that, on an average, there was a regular number of trains running over a period of ten days. Because if there are no trains five days, then it is quite normal ²¹³⁵ that over the next five days, traffic is increased. Can we draw that observation? Can we take note of that? That is my question. Is that what we are able to conclude?

JUDGE MAY: Mr. Milosevic, that is a comment. The witness has given his evidence about the number of trains and you can address us on it and on the average.

Now, Mr. Bucaliu, thank you for coming to the Tribunal to give your evidence. You are free to go.

THE ACCUSED: [Interpretation] I have --

JUDGE MAY: No more questions. No, Mr. Tapuskovic. The witness has been here long enough and answered enough questions. Yes. You're free to go.

Let the original of that document be given to the Court. The witness can go.

[The witness withdrew]

[Trial Chamber confers]

JUDGE MAY: Yes, Mr. Nice.

MR. NICE: Patrick Ball, please. The Chamber has his report. I don't think it yet has his curriculum vitae. I'm not sure about that. It has a supplement to his report, or an addendum.

Your Honour, I'm proposing to take the report shortly but to explain it in terms that are comprehensible to non-experts, and I hope that's an acceptable course.

MR. TAPUSKOVIC: [Interpretation] Judge May.

JUDGE MAY: There is a way to address a Court, Mr. Tapuskovic, ²¹³⁶ which you should bear in mind. Yes. Let the Senior --

MR. TAPUSKOVIC: [Interpretation] I apologise.

JUDGE MAY: Let the Senior Legal Officer come up.

[Trial Chamber and Senior Legal Officer confer]

JUDGE MAY: Yes.

MR. TAPUSKOVIC: [Interpretation] Judge, I was on my feet. I didn't think you saw me on my feet.

Before Patrick Ball comes in, as the amicus curiae, we decided to set forth certain matters in respect of that. So my I please have several minutes during which I can set forward some of our positions with respect to --

JUDGE MAY: Just keep the witness out for a moment, please. Yes?

MR. TAPUSKOVIC: [Interpretation] What I want to say is the following: We first of all have to clear up the status that the next report entitled Killings and Refugee Flow in Kosovo, March to June 1999, and it is dated the 3rd of January.

Patrick Ball has the status of an expert witness. Now, during disclosure and all the addendums, the Prosecution refers to Rule 94 bis. That means that it acts according to Rule 94 bis, and the report plus the additions may be considered as expert evidence.

I agree with the President of the Chamber, presented two days ago, that the report can be read and should be read by at least two members of the amicus curiae. I read it on time. I have my assistant, and I have become well-acquainted with the report when it was disclosed, since the ²¹³⁷ very beginning. However, that is not what I want to bring up here, whether the amici have read the report or not or which language. But what we're discussing here is the accused and his guaranteed minimum rights, according to Article 21 of the Statute, and one of these rights is to provide the accused enough time to prepare for his defence. For the first time, the report of Mr. Ball, in the English language, was served on the 18th of January, 2001, to both the accused and the amicus curiae. This disclosure of the material and the report, with some exceptions, was performed simultaneously. We took that. Then we received another copy of that same report in English and became acquainted with it. In its lists, the Prosecution, with respect to the order of witnesses dated the 7th of February and the 8th of February, indicated to us that the translation would be disclosed on the 18th of February 2002, but it was disclosed three weeks late, exactly three weeks late. The Rule is, Rule 94 bis and the amended Rule of the 20th of December, 2001, when on the 6th of March when we were given an addition, provides for the fact that expert witness reports should be disclosed within a time limit prescribed by the Trial Chamber or by the Pre-Trial Judge.

This Trial Chamber, on the 11th of January, issued an order confirming the oral orders given from the pre-Status Conference on the 9th of January, and it was decision number 4 or order number 4 which states the following, and I'm going to read it out word-for-word: "Written statements by all witnesses who are to testify in person must be disclosed to the accused in a language which he understands before ²¹³⁸ the beginning of trial. The witness whose report has not been disclosed in this way can be called exclusively with permission by the trial -- pre-trial Chamber."

That is what the order states. After a request made by the Prosecution after the 18th of January, the Trial Chamber changed its order of the 11th of January. Therefore, after that, at the request of the Prosecution on the 18th of January, the Chamber changed its order dated the 11th of January, extended the time limit for disclosure, and in point 3 of that order, of the order of your Trial Chamber, it states that the accused and the amicus curiae should make maximum time of the time disposal and prepare for cross-examination of these witnesses, although the Trial Chamber, at the trial, will take into consideration all demands and requests for deferring cross-examination because of insufficient time for preparation. That was your own decision.

And the report of expert Ball is very complex. It requires knowledge of the material in hand and methodology. As the amicus curiae and the obligations entrusted to us, I wanted to draw your attention to these circumstances to ensure a fair trial for the accused, and we therefore consider the Trial Chamber should assess whether disclosure was conducted in keeping with the Rules and Regulations and whether it is in keeping with the time limits prescribed. If the Trial Chamber decides that disclosure in the language understood by the accused was done within the stipulated time limit, we are of the opinion that the position of the accused set out on the 11th of March towards the end of the day at trial, when he said he didn't have ²¹³⁹ enough time to prepare for this expert witness, namely Mr. Patrick Ball, that there was reason enough to defer cross-examination as you yourself provided for in your order dated the 21st of January, 2002. Thank you. And I apologise, Judge May, for the misunderstanding a moment ago. I really was on my feet because I didn't think you had seen me standing.

Those, then, are our reasons and our position. Thank you.

JUDGE MAY: Yes, Mr. Nice.

MR. NICE: The report was disclosed in accordance with the Rules in English on several occasions. It was on 18th of January first, on the 6th of March second. I can deal with the addendums and appendices and matters of that sort which have different dates. The B/C/S was forecast as being available at an earlier date than in the event CLSS were able to provide, but it was certainly provided at the first available opportunity. The material has been available for the accused to read by himself or his associates, and in our respectful submission, the evidence should be taken not just in chief with deferred cross-examination but in full now.

What I propose to do is to go through, as I've already suggested, the report in a way that will make its concepts understandable by the lay observer, reader, or listener, and it may be that any question of deferment could be considered at the end of the examination-in-chief.

[Trial Chamber confers]

JUDGE MAY: We'll hear the evidence in chief.

MR. NICE: The witness can come in, please. ²¹⁴⁰ It may be the Chamber, before the contribution from Mr. Tapuskovic, was checking on the documents it had before it, which is the report itself, and then there's the addendum, which is a three-page document, I think, and the curriculum vitae, which I think you don't have.

JUDGE MAY: Yes. Let the witness take the declaration.

WITNESS: PATRICK BALL

THE WITNESS: I solemnly declare that I will speak the truth, the whole truth, and nothing but the truth.

JUDGE MAY: Thank you. If you'd like to take a seat. Examined by Mr. Nice:

Q. Full name, please.

A. My name is Patrick Ball.

MR. NICE: Your Honour, I propose, rather than going slavishly through a curriculum vitae, with the Court's leave to produce as an exhibit his curriculum vitae, simply to ask him a couple of questions about his background, if that would be acceptable.

JUDGE MAY: Yes.

MR. NICE: Curriculum vitae, then.

THE REGISTRAR: Prosecution Exhibit 66.

MR. NICE:

Q. Dr. Ball, you, I think, have university degrees, ultimately a Ph.D. in Sociology.

A. Yes.

Q. Sociology, including matters of statistics? ^{2141 BLANK PAGE 2142}

A. Yes.

Q. Have you, in the course of passing those various exams, which are set out in the curriculum vitae, used statistics?

A. Yes, quite a bit.

Q. And for the last how many years have you been full-time engaged in working as a statistician?

A. Full time since 1997, I think all of my work has been quantitative and analytical. Quite a bit before that, but not full time.

Q. Have you, for some years, focused on human rights events and the application of disciplines of sociology and, in particular, statistics, to those events?

A. Yes, since 1991.

Q. And so if we look - and this comes on the fourth and fifth pages, I think, of your curriculum vitae, fifth and sixth - working in reverse order, have you worked on projects in Sri Lanka; Kosovo; Guatemala; South Africa, for the Truth and Reconciliation Commission; Haiti; Guatemala again; Ethiopia, two projects, I think; and El Salvador?

A. Yes.

Q. Dr. Ball, in the course of working for human rights events as a statistician, did you develop something of a relationship with the organisation Human Rights Watch?

A. Yes, I did.

Q. And did you find yourself lecturing them or addressing courses in some way from time to time?

A. I give talks at Human Rights Watch offices in New York and ²¹⁴³ Washington, yes.

Q. And in the course of your lectures to them, did you focus on something, perhaps a bit of an obsession or a bee in your bonnet?

A. I often spoke about the difficulty of understanding a massive phenomenon with purely qualitative methods; that is, if the human rights event being studied has more people involved than you can easily interview, then perhaps statistical methods would assist an understanding of.

Q. And does that explain how you first became involved in the Kosovo project --

A. Yes.

Q. -- at least for the preparation of this report?

A. Yes.

Q. Did Human Rights Watch get in touch with you in April of 1999?

A. First in March of 1999.

Q. March of 1999.

A. Yes.

Q. Inviting you to do what and to go where?

A. They asked me to assist them in the field, perhaps to design my own project, not a Human Rights Watch project, but they saw the need for a project like the ones I had done in the other countries you had listed. So they invited me to accompany them to Albania to design such a project.

Q. I'm going to remember, Dr. Ball, to leave a gap between your answer and my question, because as we are speaking the same language, there's the risk that we will overtax the interpreters otherwise. ²¹⁴⁴

A. Understood.

Q. Did you go to Kosovo, and in particular, to a border point in the south-west?

A. I went to Albania, actually, and --

Q. Sorry, yes, to Albania.

A. Yes. And then to a border point in northern Albania.

Q. Which is on the south-west of Kosovo?

A. Yes.

Q. When there, did you notice something about the method or methods of recording events that was going on?

A. Yes. There were two methods by which refugees were being counted as they crossed the border from Kosovo into Albania.

Q. Two different methods, being what?

A. The first were registries prepared by the Albanian border guards. They had developed a form on which they attempted to register the name, or the surname, the year of birth, the place of residence or location, and the quantity of each party or group that crossed the border from Kosovo into Albania. That was the first method.

Q. The second?

A. The second were teams of observers from UNHCR and the OSCE, or former OSCE mission in Kosovo, who maintained head counts with mechanical hand-held counters that they reported to UNHCR, I believe in Kukes, the Albanian town some distance from the border, a few kilometres, and then those were recorded in daily press reports in Tirana.

Q. To a statistician, is there any significance or value in matters ²¹⁴⁵ being recorded not in one way but in two or more?

A. Yes. Having two measurements of the same thing is enormously valuable because we can compare the two and determine how they might differ, and then if they differ, which might be more important, reliable, or sound.

Q. And I'll take the story comparatively swiftly and then get to the report. Did you return to America to seek initial funding for a project and did you obtain funding from, I think, the Ford Foundation?

A. Yes. I returned to America --

Q. If it's yes, yes will do.

A. Yes.

Q. I'll deal with how you obtained some of the records, in any event, when you were back at the border crossing shortly, but I just want to conclude the question of funding. Were there earlier publications by you or by you and others in relation to the work you did, publicly available?

A. Earlier than the current report.

Q. Yes, earlier than this report.

A. Yes. There were two.

Q. Were they funded by the Ford Foundation or were they funded by the second funder, to whom we'll turn?

A. The first report, entitled "Policy or Panic?" was published in March of 2000 and was funded by the Ford Foundation and the John D. and Catherine T. MacArthur Foundation and the Institute for Civil Society, all American private foundations.

The second report, published in September of 2000, entitled ²¹⁴⁶ "Political Killings in Kosovo," was published with the American Bar Association and Central and East European Law Initiative, and it was funded by funds provided to ABA/CEELI by the US government.

Q. And is that the funding that leads to the report that's before this Court --

A. Yes.

Q. -- today?

A. Yes.

Q. Did you make any presentations to the United States funders?

A. Yes, I did.

Q. One or more than one?

A. I made one presentation in July 2001.

Q. Did you have any further contact with the United States funders other than that?

A. No.

Q. Let's go back, then, to your return to the border once you had started the process of obtaining funding for the project, and you can deal with this very shortly, I think. Did there come a time when it was not possible to go with safety to the border point, where you believe records to have been?

A. Yes. When I returned to the border in early June 1999, the border post had been fired upon a few days previously. A journalist had been seriously hurt and the border guards decided to withdraw several hundred metres into Albania, to a point where they believed they would be safe.

Q. To cut this long story short, I think you believe the records to ²¹⁴⁷ be in the border post, or some of them.

A. Yes.

Q. Eventually you went to the border post yourself.

A. Yes, I did, personally.

Q. With an accompanying guard. Did you find some records?

A. I did.

Q. Which type of records did you find? Did you find the first type or the second type that resulted from the clickers?

A. I found the border guards' registry forms.

Q. And did you gather those up --

A. Yes.

Q. -- bring them back, and use them as part of your source material?

A. Yes.

Q. Before we now move to the report itself, at that early stage, was there any divergence between the total numbers of people moving that you can tell us about at that post?

A. Divergence --?

Q. In numbers, at that early stage.

A. After we had obtained the records and keyed them into a database, we looked at the number of people reported each day crossing the border and compared the numbers reported by UNHCR to the numbers reported by the border guards.

Q. So this is right at the beginning, raw material?

A. Yes.

Q. Did the figures differ, and if so, what was each? ²¹⁴⁸

A. They differed on some days, and all of the days on which they differed were days of very heavy flow over the border. And in every difference, the registries by the border guards were fewer; they were smaller numbers than the UNHCR figures.

Q. And the two different figures were?

A. The total figures that there were 272.000 people registered by the border guards, and four hundred, four thousand [sic] people registered in the UNHCR daily counts reported in their daily press reports.

Q. So the clickers reported more than the handwritten records?

A. That's correct.

MR. NICE: Your Honour, I'm going to ask that the report that has already been made available should become an exhibit, and I'm proposing in due course that we might use a version of the report which I've got here, page by page laid on the overhead projector so that those listening and watching can understand the points we're making. But first of all, may the report become an exhibit, Exhibit 67.

JUDGE MAY: Yes.

MR. NICE: I'm afraid not every page is numbered, because the first thing I want to look at, very briefly, is the Scholarly Review Panel, which comes two pages after the last numbered page, page 69.

THE REGISTRAR: Prosecution Exhibit 67.

MR. NICE: And if the usher would be so good, what we can do with our loose-leafed version is make available to the usher - it might be easier that way - pages to which I'm referring, and then he can display them. It's two pages after 69, the Scholarly Review Panel. If that could ²¹⁴⁹ just go on the ELMO, please, the overhead projector.

Q. You'll see that on the screen in front of you. Did you, in accordance with -- well, you must tell us if it's in accordance with practice. Did you submit drafts of the report that's been produced to a scholarly review panel?

A. Yes, I did.

Q. Is that standard practice for reports of this general kind or not?

A. This is a bit more rigorous than reports submitted to non-expert groups. Things that are published in professional journals are reviewed according to a standard and formal process. Since we were submitting this to the Office of the Prosecutor, we had to create a panel of reviewers who could give this equally rigorous review.

Q. Did they review it on two occasions?

A. They did.

Q. At each review, did you receive advice as to further steps to be taken, further elements to be introduced into the report?

A. That's correct, yes.

Q. And we see listed various names: Dr. Helge Brunborg, Dr. Ronald Lee, and Dr. Francoise Seillier-Moiseiwitsch, the first three. Do any of those occupy particularly senior positions in the world of statistics?

A. I consider that all three of them occupy senior positions. Dr. Brunborg formerly worked for the Office of the Prosecutor here and is very well regarded in international demography, in addition to his --

Q. Well, that's enough for these purposes. Dr. Ronald Lee?

A. Was formerly President of the Population Association of America, ²¹⁵⁰ the Professional Association of Demographers in the United States.

Q. And then Dr. Francoise Seillier-Moiseiwitsch?

A. She is the chair of the Human Rights Committee of the American Statistical Association.

Q. Well, that's all I need to ask. Others can ask about other names if necessary. Just this: At any stage in the review process, did you either reject the advice you were given by the panel or did the panel ever challenge your conclusions as unacceptable or in any doubt?

A. No.

Q. Is that no to both?

A. That's no to both. That's correct.

Q. Thank you. Now let's go to the report.

JUDGE ROBINSON: Mr. Nice.

MR. NICE: Yes.

JUDGE ROBINSON: Earlier the witness said that since he was submitting the report to the Office of the Prosecutor, I wanted to find out whether the report was commissioned by the Office of the Prosecutor or whether it was done independently and then submitted.

MR. NICE:

Q. Can you give us the history of that? You've already told us that your earlier two reports or the two earlier reports with which you were associated were published publicly. Just explain how it was that you were involved with the Office of the Prosecutor.

A. After I completed each of the previous two reports, before they were made public, I came to The Hague and gave brief presentations to ^{2151 BLANK PAGE 2152} people here at the Tribunal, covering the basic findings of the report. On that basis, I was asked by people in the Office of the Prosecutor if this evidence would be sufficient to go -- to bring before the Court, and I answered that, in order to bring it before the Court, it should be done with additional data and in an even more rigorous fashion. They assisted me, negotiating with the Organisation for Security and Cooperation in Europe for a third source of information, and they provided me, the Office of the Prosecutor provided me with a database of exhumation records based on the fieldwork of the exhumation teams. At no time did the Office of the Prosecutor suggest findings or suggest what I should conclude. They did ask me to do this with the highest scientific standards, and they emphasised the importance of that.

MR. NICE: We're going to see some of these matters reflected at two places in the evidence in the course of the morning or the early afternoon.

Q. Before we turn to looking at particular pages of the report and in order to have exhibits in good order, can we produce one exhibit that's outside the report, which is an example of the record you took from the border crossing point, which I think was Morina, wasn't it?

A. Yes.

THE REGISTRAR: Prosecution Exhibit 68.

MR. NICE: Thank you. Which I think has a translation with it. If the original can go onto the overhead projector. And the -- if the original could go onto the overhead projector - thank you very much - briefly so those viewing can see it. And then having been viewed, if the ²¹⁵³ original can go -- well, no. We can now look at the translation, putting that on the overhead projector in English.

Q. Can you just tell us, please, Dr. Ball, what this particular page is.

A. This is one of the 690 pages that I recovered from the border post at Morina.

Q. And it shows, in respect of individuals, their name, their address - that is the address from which they had come - their year of birth, and what does the last column help us with? Do you know?

A. That's the total number of people in their group that crossed the border with them. Only the name of the head of household or head of the party was registered.

Q. So that name number one from Suhareke, 27 people in that party, is that what you're saying? It says "Tool type" I see on the translation, but --

A. I don't think it's "Tool type." Yes, that's what it means.

Q. Thank you. I think it will be helpful if we go first of all to a first graph which we can find on page 5 of the report, and then we'll come back a few pages to fill in a few blanks, but let's go to page 5 first of all.

MR. NICE: We have both text and the graph, but I'd like the unit, if it can, to focus on the graph, please. Thank you.

Q. This is one of many graphs that we're going to be looking at. This is -- or several, not many. This is an "Estimated total refugee flow over time," and in the vertical axis it's "People leaving their homes," ²¹⁵⁴ and we can see the dates concerned moving along the bottom. Yes. Now, this particular graph, was this a graph simply drawn from the raw material that you were able to collect? We'll discover what sources of raw material you had in due course, but yes or no, was this a graph that was provided from raw material?

A. No.

Q. All right. What was it, then?

A. The raw material consisted of the border records we discussed earlier, supplemented by the UNHCR reports and a number of other smaller sources.

Q. Yes. Am I or am I not using the word "raw material" correctly? You had other sources.

A. Yes.

Q. We'll identify them in due course. Having various sources of material, you then read from the material whether someone's left their home in the period 24th of March, 25th of March, 26th of March, and so on.

A. Uh-huh.

Q. You total up the figures from all sources, making such allowances for duplication as you have to, and then you've got this particular graph.

A. We also corrected for the time it takes people to travel to the border --

Q. Yes.

A. -- from their homes. ²¹⁵⁵

Q. We'll come to all those details in a minute. Yes.

A. With that clarification, yes.

Q. Right. To a statistician and perhaps to anybody, is there a characteristic in this particular graph?

A. Yes.

Q. Namely?

A. There seem to be three quite distinct phases or points at which refugee flow is high. First, the flow rises to high points during late March and early April. Then this declines to a relatively much lower point. It rises again to a peak in mid-April, subsequently declining to a low point in later April. Then toward the end of April and early May, it rises in a series of small peaks, finally declining toward near zero for the remainder of the period.

Q. So three peaks, as it were, and two troughs?

A. That's correct.

Q. We're going to see something similar in other graphs a little later, but immediately what does that shape suggest when contrasted with, for example, the possible assertion that there was a constant cause causing people to become refugees in the way suggested by the line?

A. This pattern would not be characteristic of a constant cause. The stop-and-start pattern of the peaks and troughs suggests more strongly a cause that itself started and stopped.

Q. Well, again we'll break this down in due course into a little bit more detail, but did you from some stage then in the project, once you became aware of these peaks and troughs, consider three possibilities as ²¹⁵⁶ causes or possible causes for peaks and indeed troughs?

A. Having seen this graph and others with this characteristic pattern of peaks and troughs, we concluded that there must be some cause. Accordingly, we considered three possible causes that might account for this peak and trough pattern.

Q. Namely?

A. Those were, first, that KLA activity, in particular clashes with Serb forces, could be a cause; second, that NATO airstrikes might have been a cause; and third, that Yugoslav force activity might have been a cause.

Q. At the end of the exercise, when we reach your conclusions, in general - just yes or no - is it possible to identify causes or is it more the scope of the statistician in this sort of work to consider what causes may be excluded?

A. The methods we used test hypotheses. Each of the three causes I proposed, we treated as a hypothesis. The methods allow us to reject hypotheses or to find that data are consistent with a hypothesis; however, the matters do not allow us to prove a hypothesis.

Q. With that overview of what you found yourself engaged in and the scope of the statistician's skill, range of his potential skill as applied in this case, can we now go back, please, to page 3. There are one or two things I wanted to clarify in the report, and it will be helpful to have those pages on the overhead projector so that people can follow what I'm asking about.

And "Data and analysis," paragraph 1.2, you set out, as we already ²¹⁵⁷ know, that you went to Morina, which is the one border point in the south-west of Kosovo, north of Albania.

A. Yes.

Q. Could one border point be sufficient? What's your comment on that?

A. This border point could be sufficient because, according to the available data, I believe that more than half of all Kosovar Albanians who left Kosovo did so through this point. In the report mentioned earlier that looks specifically at refugee flow, we found that all the other evidence available from other border points in Albania as well as the border points in Macedonia and data from Bosnia, Kosovars who went to Bosnia, suggest that their movement patterns and departure patterns were substantially similar to the patterns found among refugees who entered Albania.

So, yes, I believe these data are sufficient for these conclusions.

Q. The second bullet point of the same paragraph turns from the material you had in respect of refugee flow to the data you had in respect of killings, because as we're going to discover, you were looking at simultaneously -- not simultaneously. In the course of your work, you were looking both at refugee flow and at killings. Set out there and just run through for us the sources of information you had to tell you when, where, and whatever else you knew about killings.

A. We learned about killings from interviews conducted by the ²¹⁵⁸ American Bar Association, Central and East European Law Initiative, and their partners, interviews that they conducted, as well as interviews conducted by Human Rights Watch and interviews conducted by the Organisation for Security and Cooperation in Europe. We --

Q. So that - just pausing - are three, as it were, narrative records.

A. That's correct.

Q. Right. And then you had a fourth record or fourth source.

A. Yes. And those were the exhumation records of the exhumations conducted by international teams on behalf of the Tribunal.

Q. So you did not have and therefore did not rely on witness statements of the ICTY, OTP?

A. No, I did not.

Q. You did have records in relation to the bodies found and material in relation to that?

A. That's right.

Q. Human Rights Watch -- I beg your pardon. OSCE's records are, of course -- well, you better tell us if they are. Are those records that are also reflected in the two OSCE reports, one on the suffering of the Kosovar Albanians and one on the Serbs --

A. I believe so.

Q. -- the KLA?

A. Yes.

Q. And as we can see, in all, "The statistical analysis of killings aggregates information from more than 15.000 interviews and exhumation ²¹⁵⁹ reports."

A. Yes.

Q. All right. Can we go now just to clarify one or two things on page 4. At the bottom of page, I think, "Statistical analysis of refugee flow."

We've already looked at the graph because I wanted to provide the Court with an overview.

MR. NICE: The top of that paragraph. Well -- thank you very much. That's -- no. Paragraph 3. Thank you. That's fine.

Q. You set out in this part of your report a description of the departure of ethnic Albanians from Kosovo from late March to May of 1999, saying that approximately 95 per cent of the Kosovo Albanian refugees who did so entered between the 24th of March and the 19th of May. And there's a reference. It says: "(Ball 2000, p.5)." Is that a reference to one of the two earlier reports published publicly that you've already spoken about?

A. Yes.

Q. Was the work of those reports, although you yourself judged them not to be sufficient for presenting to Court in themselves, was the work of those earlier reports incorporated in the report that you have presented to this Court?

A. Yes.

Q. The 95 per cent, is that an estimate?

A. Yes, it is.

Q. Is it one by which you stand? ²¹⁶⁰

A. Yes, it is.

Q. In what way? Just tell us about it.

A. I used the border records supplemented by the UNHCR records and the much smaller head-count records done by the Albanian government's emergency management group in May, and I divided the refugees who flowed in during this period from those who flowed in either before or after, according to those records, and calculated a simple proportion.

Q. Footnote 6 we may focus on. I think deal -- it says at the end of it --

MR. NICE: That's it, thank you very much.

Q. Says this: "The three phases..." and we've already looked at the diagram. We'll look at it once more in its first depiction. But it says this:

"The three phases reflect patterns of refugees departing their homes, not the patterns of refugees crossing the border. On any given day, slightly more than half of the refugees who crossed the border had left their homes earlier that same day. However, the remaining refugees crossing the border that day had been in transit for longer times. The transit delay was accounted for in the analysis of the data." Well, you've already dealt with that in general. How did you calculate the transit delay for your statistical purposes?

A. I had three surveys that had been conducted in refugee camps and other refugee housing locations in Albania and Macedonia during March through June of 1999. In those surveys, refugees reported the sequence of events that took them from their homes, their home departure, to the ^{2161 BLANK PAGE 2162} camps. From that, I was able to calculate the transit time for refugees and then look at that time for refugees who left different places and different times and compare the process of transit in a statistical sense over those different -- over refugees from different times -- different places and who had left at different times.

Q. Now, if we go back to page 5 on which I have one supplementary question and following which, with the Chamber's leave, I would seek a one-minute private session to deal with the ordering of witnesses today and tomorrow, looking at page 5 again, and just to remind us, the graph at the top is the graph from the material that you had available on refugee flow, adjusted for the time that people left their homes before they reached the border; yes?

A. That's correct.

Q. If we can go down the page to the last paragraph, you say this in the last sentence: "The mass exodus of Kosovar Albanians on this scale and in this pattern could only have been driven by a common cause." We've seen three peaks. Common cause or two or three different causes? Can you help us?

A. By "common cause" here, my co-authors and I refer to the -- that the refugees who departed would have had a cause common with respect to each other. At this point, we, without additional data, I don't think we could distinguish between whether there was one cause for all three peaks or three different causes. Additional data would allow us to refine that.

Q. Or I suppose, indeed, two causes and one cause, or two causes, rather. ²¹⁶³

A. Any combination, sure.

Q. Turning to page 6, on which I have a few questions --

MR. NICE: I don't know if that would be a convenient moment.

JUDGE MAY: If there's another matter you want to raise, we'll adjourn now.

MR. NICE: Yes.

JUDGE MAY: Dr. Ball, you are giving evidence. Don't speak to anybody about it during the adjournment or any other adjournments there may be until it's over, and that does include members of the Prosecution team. Could you be back, please, at half past 2.00. Yes. We'll go into private session.

MR. NICE: Can Dr. Ball withdraw? It has nothing to do with him, I hasten to say.

[The witness stands down]

[Private session] (redacted)

(redacted) (redacted)

(redacted) (redacted)

(redacted) (redacted)

(redacted) (redacted)

(redacted) ²¹⁶⁴ Page 2164 - redacted - private session.

--- Luncheon recess taken at 1.05 p.m. ²¹⁶⁵

--- On resuming at 2.33 p.m.

[Open Session]

JUDGE MAY: Yes, Mr. Nice.

MR. NICE: Page 6 of the report. We'll be looking at all of this. This has just been laid on the overhead projector, and I'd ask the booth to focus on relevant parts as we require them. This is one of about two or three parts of the report which need some explanation, for they are, to a degree, technical.

Q. If we look at the body of the report first, at the end of the first paragraph, we see a reference to 10.000 victims - thank you very much - reported by name. This process identified 4.400 unique individuals. In a couple of sentences, explain how 10.000 reduces to 4.400.

A. Many of the people who were reported to the data-gathering projects that I described earlier are reported by more than one witness, so a particular individual may be reported two or three times. One individual in particular, you'll remember, was reported 27 distinct times to one of the four projects. So we removed the individual - I mean the duplicate references - we identified them and removed them, and then we identified the references which were in common among more than one project. Someone may have been reported to the OSCE but also identified in an exhumation. After all of those reductions are complete, we are left with 4.400 unique individuals.

Q. Let's now go to the next paragraph, which begins in this way: "Second, because the victims identified in the data sources were ²¹⁶⁶ not the only victims of killings, it was necessary to estimate the number of undocumented victims to determine the overall estimate of total number killed. This figure, 10.356 ..."

I pause there straight away. Ten thousand, three hundred and fifty-six, the estimated number, is rather close to the 10.000 figure from which the reduced figure was itself derived. Is there significance in the similarity figures or is that entirely coincidental?

A. Entirely coincidental.

Q. So you're now going from your 4.400 reduced figure to calculate another figure, and as you say in the paragraph, "generated by means of a widely used demographic statistical technique known as multiple systems estimation." In terms that the layperson can understand, please explain.

A. Multiple systems estimation uses the overlap among multiple projects to determine how many people fall outside the documentation of any of those projects. So, for example, there were 4.400 unique individuals reported by name to one or more of the four projects. There are some additional numbers of people who were not reported by name to any project. Our effort is to estimate that.

Q. Now, because of the constraints of time and because this is a technical matter, I am probably going to leave it there, although I know there are illustrative diagrams you can draw to help us, and I may ask you to come back to one of them. But am I right in suggesting this: Because you had four sources of information - not just one, not just two, not just three, but four - differently prepared, you're able to engage in this multiple systems estimation to reach an approximate, an estimated global ²¹⁶⁷ figure?

A. Yes. The method can be done with only two systems. With the addition of each additional system, we gain the ability to more deeply test our assumptions. Four systems is a quite thorough test.

Q. Now, you, of course, proceeded in your report on the base of 10.356; is that correct?

A. Yes, it is.

Q. But it happens - and this is shown in the footnotes to this page - it happens that you had available to you, and we can see it at footnote 10, I think, some other independent studies that had been prepared in entirely different ways; is that right?

A. That's correct, yes.

Q. And so do these operate not as part of your database but as a check, if you like --

A. Yes.

Q. -- something to cause you to look again if they are wildly different? In this case, if we look at the reports, PHR report of ABA/AAAS, what figure did that reach by a different system?

A. The Physicians for Human Rights -- Human Rights report, cited as "PHR," approximated 9.600 deaths.

Q. And then if we go to the other one, which was Spiegel and Salama?

A. Yes. They are epidemiologists at the Centre for Disease Control.

Q. What figure did they reach? Although I think they included a year that you didn't include.

A. That's right. Their estimate includes 1998, which is excluded ²¹⁶⁸ from our analysis. They reached a figure of approximately 12.000.

Q. So more than you, but they've got another year.

A. Plus the months from January to mid-March, which are not in -- of 1999, which are not in our study.

Q. So thus you reached your figure of 10.356. And if we then take that figure and look at the top of the page to the two graphs, please.

MR. NICE: That's all right. I think the booth will do it for us. Thank you very much.

Q. We now see at the top the same graph of "People leaving their homes" prepared in the way you've already described, but we now have underneath it a "People killed" graph. And is this prepared on the basis of raw material in the loose way I used or is this prepared on the basis of the figures as derived from multiple systems estimation?

A. This is from multiple systems estimation.

Q. What do we notice about the similarity or dissimilarity of the shape of this graph in comparison with the earlier graph?

A. They are very similar in several respects.

MR. NICE: Sorry again to the interpreters. I've not been leaving a gap.

Q. My mistake, Dr. Ball. I will try and do so. Your observations about looking for a cause or separate causes for peaks, would that apply similarly for the "People killed" graph?

A. Yes, the same reasoning would apply.

Q. Page 7 of the report, please.

MR. NICE: Text first, for the booth, and then the map. ²¹⁶⁹

Q. The text still relates to the graph you've been showing, and it makes one further explanation that we should have in mind. "When the estimated number of people killed is considered over time, using the same two-day intervals employed with the refugee flow data..." and so on. Now, I don't think I got you to deal with the two-day interval. Perhaps you could now.

A. As I described earlier, the data from the Albanian border guards was transformed into data on people leaving their home by using the transit time. The transit time comes from surveys. As I described, several hundred surveys from my own project plus those from two other projects.

When we look at the survey data day by day, there are some days with no data, and that's because we -- a survey is conducted with only a small sample. So on a day with relatively few refugees crossing the border, there are relatively few survey datum about it. As a result, there can be a lot of zeros if we look over time. Zeros may be real information or they may simply be points in the data that are insufficiently covered. By summing days together into two-day periods, we stabilise those estimates.

I made this decision to do two-day intervals in the migration study, however, it proved useful throughout the subsequent study and the current study so we have maintained it all through for its property of giving us an additional stability, by which I mean specifically reducing the number of zero points in survey information.

Q. Thank you. Identifying intervals of whatever item or event ²¹⁷⁰ material, is that a common and necessary technique in statistics?

A. Very much. The language that statisticians use is that of aggregating or adding things to bigger units, or disaggregating, breaking things into smaller units. It's a standard technique that every data analyst must employ to find the place where there's enough data to say something meaningful.

Q. Thank you. Let's move on to the next topic, which is first revealed by the map at the top of this page.

Having prepared your first graph of "People killed," did you decide to break Kosovo down into four regions, north, east, south, and west, as shown on this map?

A. Yes, I did.

Q. The divisions between parts of the territory have no relationship with political, geographical, or any other divisions. They're simply for statistical purposes.

A. That's correct. That's correct.

Q. Can we go over now briefly to page 8. And at the top of the page we see "Refugee flow and killings by geographic location." "In addition to examining when refugee flow and killings happened, it is important to study where the events occurred. An analysis of the locations where the refugee flow originated, and the killings occurred, shows widespread patterns consistent with acts of violence associated with displacements."

Those last few words are, I think, the conclusions of the following passage. Would that be correct? ^{2171 BLANK PAGE 2172}

A. That's correct.

Q. Can we then go over -- we're going to come back to footnote 11, but can we now go to pages 9 and 10.

On 9, Figure 4 at the top of the page has again the same "People leaving their homes" graph, but the bottom one now shows what?

A. The bottom graph is the number of people killed. Both of the curves shown in Figure 4 are for people leaving their homes or people killed in the northern region.

Q. Yes. My mistake. The top map is not the familiar one. It's now unique to the northern region.

A. That's correct.

Q. It's been extracted. Thank you. And the bottom one, again prepared in the different way, still on the basis of the technique you told us about, is again just for the northern region.

A. The bottom one is for the southern region.

Q. Sorry. The bottom of the two graphs at the top is for the -- at the top you've got two graphs, the top one being for "People leaving their homes" and the bottom of the two for "People killed" in the northern region. Do we here see similarity with early graphs or dissimilarity?

A. The similarity we see in the two graphs of Figure 4 is with each other.

Q. Yes.

A. Here what's interesting to me in particular, I think, is the quite distinct peak in both graphs during phase 2. ²¹⁷³

Q. The bottom figure is for the southern region. Again, it contains two graphs. Are they similar, one to the other, in statistics terms?

A. These curves are quite similar to each other. They both show almost no activity in phase 2. Rather, in the southern region, nearly all the activity is concentrated in phase 1.

Q. We see -- sorry. I'll just pause. We see on the "People leaving their homes" graph a little free-floating point with discussion in the text. I'll come back to that. If we go now to page 10 --

JUDGE ROBINSON: Mr. Nice, before you go to page 10, can I just ask the witness to return to the peaks in phase 2 of the northern region, and he highlighted the significance of those peaks. But the peak at the top, that is, for people leaving their homes, is actually higher than the peak at the bottom in respect of people killed. Is there -- does that indicate anything or ...?

THE WITNESS: [Interpretation] The curves are not identical, as Your Honour has observed. What I think is most significant about these two curves is the existence of a very substantial peak during phase 2. That is what we will contrast with other regions in a moment.

JUDGE ROBINSON: I see.

MR. NICE:

Q. Reducing to non-technical terminology, correct me if I've got it wrong: Is it the fact and general quantum of a peak or trough that is significant rather than the precise scale of the peak? And if you don't accept that, don't adopt it simply because I've suggested it.

A. The general pattern of high points and low points is important. ²¹⁷⁴ The more closely they track each other, the more strongly a statistician would argue that they correlate.

Q. Right. We ought -- perhaps, since we're looking at figures 4 and 5, still on page 9, is it important for us just to have in mind that the scale -- the vertical scale in each graph is in fact different because you've adjusted the graphs to the maximum number of either people leaving their homes or people killed to make best use of the space?

A. That's right. More than -- approximately 40 times more people left their homes and are represented in these graphs than were killed, so the scales must, by necessity, be different.

Q. And indeed, as between the northern and southern region, the graph at the top, people-leaving graph is 7.900, the bottom people-leaving graph is 24.000?

A. That's right.

Q. And similar differences, even more substantial in proportionate terms, I think, as seen on page 10, which shows the eastern region at the top and the western region at the bottom. By now we can probably start to work things out for ourselves. In each case, the people-leaving and people-killed graphs, do they have similarities to each other?

A. Yes.

Q. However, is the western region graph similar or dissimilar from the eastern?

A. The west is dissimilar from the east.

Q. Coming back to page 9, does it have any points of similarity with the southern? ²¹⁷⁵

A. Yes. The western and southern patterns are quite similar, as distinguished by absence of significant peak during phase 2.

Q. Going back from page 9, and in particular, we'll go back to page 8, because I drew your attention to the free-standing point that is said to require discussion in the text. On the footnote at page 8, do you observe that there is an anomalous point in the southern region in late May? It's at the footnote on that page, right down at the bottom.

A. Yes, we do observe that, my co-authors and I.

Q. And you say this in the text of the footnote: "This estimate of more than 200 people killed in one two-day period results from fewer than 20 documented killings." And then you say this:

"Appropriately, this point also has a high level of error associated with it, as shown in Appendix 2, Figure 12. As is clear in that figure, most points have modest errors which do not weaken the interpretation of the pattern. This point, however, has a sufficiently wide margin of error that the point may not be significantly different from zero."

Two or three things from that: If statisticians, or even Their Honours, want to follow the error or the margin of error calculations, they may do so at Appendix 2. I think it's page 53. We don't need to turn to it. But looking at any one of the graphs - page 9 is open before me - do the figures represented on these graphs have margins of error associated with them?

A. Yes. Every point in the people-killed graphs has a margin of ²¹⁷⁶ error.

Q. Do you assess those margins of error to ensure that they are not at such a level as to invalidate the figure that you're representing?

A. Yes, we do.

Q. The free-standing point dealt with in the footnote and to be seen on Figure 5 is one which you eliminate because of the scale of the error, the margin of error, but as to the other figures, have you allowed for the possibilities of margins of error, nevertheless standing by the graph that you've drawn?

A. Yes, we have.

Q. Thank you. In that case, may we turn to page 11.

JUDGE KWON: Excuse me.

MR. NICE: Just one minute.

JUDGE KWON: I'm not sure. You might have mentioned this, but regarding the date in the refugee flow data, do they refer to the date when the refugees entered the border or those when they left their villages?

THE WITNESS: We refer to the latter, the dates they left their villages.

JUDGE KWON: Thank you.

MR. NICE:

Q. Page 11. We start with a text, which actually begins on the previous page, 8. Don't worry about that. The text explains how you're exploring the hypotheses you've already described; is that correct?

A. That's correct. ²¹⁷⁷

Q. So we can see at the beginning of the text: "However, if airstrikes and KLA activity do not precede the peaks in the number of killings ... then the causal relationship should be ... rejected."

Well, let's just have from you your approach to connection or non-connection of either KLA activity or, in due course, NATO activity, with refugee movement or killings. What was your approach?

A. Our idea is that if something is to cause something else, then the cause must precede the effect. Accordingly, we looked at the peak level of killings and refugee flow for each municipality. There are 29. We then considered whether, during the two-day period of the peak or in the preceding two-day period, there was some KLA activity. If we found any KLA activity whatsoever in the period of the peak or in the period immediately preceding the peak, we counted that in the first row. However, if we found KLA activity either not at all in that municipality or following the peak, we counted it in the second row. Finally, if KLA activity preceded the peak of the killings or refugee flow by more than four days, we counted that as an inconclusive pattern.

Q. Right. Well, before we look at the table that sets out your findings, it's necessary to know, first for the KLA and then for NATO, what your sources of material were on the basis of which you then built the table. Let's deal with KLA first. What sources of material for KLA activity did you have and rely on?

A. We began, my co-authors and I, began by examining open sources that were placed on Internet by the government of Yugoslavia, and as well ²¹⁷⁸ as Serbian non-governmental organisations. These documented casualties suffered by Serb authorities in Kosovo as a result of KLA attacks, and the sources documented exchanges of fire, battles, between KLA forces and Yugoslav government or police forces.

Q. Was that your only source or did you have other sources for KLA?

A. We supplemented those sources with additional sources provided to us by the Office of the Prosecutor. The open sources in total constituted 55 per cent of our source material, and the Tribunal sources constituted 45 per cent of our sources.

Q. Going back, as I said I would when His Honour Judge Robinson asked you this morning, to the topic of your instructions by the Office of the Prosecutor, two things about the material - three things - about the material provided by the Office of the Prosecutor: One, was it statements or was it reduced to a statistical or summarised form?

A. It was reduced to a highly summarised form.

Q. Second, how did that material compare in its general shape and effect with the 55 per cent of open-source material that you had obtained from Serb sources?

A. In the statistical terms, they looked approximately the same.

Q. Third, at the time that that material was provided to you, had the Office of the Prosecutor any advance knowledge or knowledge of the findings to which your report may have been moving?

A. When I met with the Office of the Prosecutor's representatives in early November, we had none of the results on killing that are presented here complete. However, the migration report had been in the public ²¹⁷⁹ domain for a year and a half or so, so that was available to anyone who wanted to look at it.

Q. Back to the text: In the next paragraph, beginning "To test the conclusions drawn by this method, we used another statistical method to consider the joint correlations of ..." I'll come back to that perhaps later. But -- yes, I'll come back to that later. Let's go to the chart at the top of the page, then, please, if we can. Reminding ourselves that "preceded or coincided with peak" means really the four-day period that ends in the peak; is that right?

A. That is correct.

Q. Of the 29 municipalities, there was killing preceded by KLA activity in 11; killing followed by KLA activity in 12; and killing with inconclusive findings as to KLA activity - either none or outside those parameters - in 6.

A. Yes, that's correct.

Q. And the refugee flow shows respectively 10 for Preceded or coincided, 11 for Followed, and 8 for Inconclusive. The percentages in each case being set out. What does that tell you?

A. If we hypothesise that KLA activity directly or indirectly caused killing or refugee flow, we would expect that KLA attacks and activity would be very frequently present at the peak of the hypothesised cause. I believe that this table shows that there is no KLA activity at all preceding the peak in more municipalities than we see the hypothesised coincidence. We interpret this to mean that this tends to reject the hypothesis that the KLA was the cause of killings and refugee flow. ²¹⁸⁰

Q. May we now move to the possible significance of NATO, page 12. What were your sources of information for NATO attacks on a geographical basis so as to explain or not to explain killings and refugee flow?

A. The data on NATO airstrikes came exclusively from Yugoslav press and government sources that were publicly available in the period approximately June to August 1999.

Q. You restricted yourself to that source for what reason?

A. There were two reasons: One, because my two phone calls to the US Department of Defence went unanswered and unreturned, so I did not get data from the US side; but second, it seemed to me the best possible test of this hypothesis to use the data from Yugoslavia. At the time, the hypothesis I was focusing on was the relationship between NATO airstrikes and refugee flow. So the claim that NATO airstrikes had motivated in some way refugee flow was in part advanced by authorities in the Yugoslav government. It seemed to me the best test of that hypothesis to use data from Yugoslavia.

Q. The table at the top of page 11, in a form similar to the previous table, shows thus that there were NATO airstrikes preceding refugee flow in three of the 29 municipalities. There were -- did I say that correctly?

A. I think it's preceding in killings in three.

Q. In killings, I should have said, in three of the municipalities. There were airstrikes preceding refugee flow in nine of the municipalities. The remaining figures, 20 and 6, 13 and 7 being municipalities where there was either such strikes or inconclusive ^{2181 BLANK PAGE 2182} material. Sorry, such strikes following or inconclusive material.

A. That's correct.

Q. Now, what does that tell you about the possible significance of NATO airstrikes in the killings or the refugee flow?

A. The pattern that in 20 municipalities NATO airstrikes followed the -- the first NATO airstrike followed the peak of killings makes NATO airstrikes a highly implausible cause of the pattern of killings. After all, the peak occurred and passed before the first airstrike in 20 municipalities. Similarly for refugee flow, NATO airstrikes are an implausible motivator, directly or indirectly, for refugee flow since, in 13 municipalities, the peak came and went without the first airstrike having occurred.

JUDGE KWON: Excuse me once again. Whether the NATO bombing preceded or not was decided municipality by municipality?

THE WITNESS: [Interpretation] That's correct.

JUDGE KWON: So if a nearby village was bombed yesterday and the villagers left on today and it was bombed tomorrow, so that was -- you mean that it is not -- the NATO bombing was not preceded in that case?

THE WITNESS: [Interpretation] If the nearby village was across a municipal border --

JUDGE KWON: Yes.

THE WITNESS: [Interpretation] -- that would be correct. We consider that in the next section.

JUDGE KWON: Yes. Thank you.

MR. NICE: ²¹⁸³

Q. Can we come back briefly to page 11, because I said there was something that needed explanation, it being a term of art, that we should have explained on the record for those of us who will be reading the report in months to come.

In the middle of the page you set out this: "To test the conclusions drawn by this method, we used another statistical method..." Are we now moving to a different tool from the one we've already looked at?

A. Yes.

Q. "... to consider the joint correlations of KLA and NATO activity with refugee flow and killing patterns. The point is to use the second statistical technique..." and then these words, "... to control for the correlation of KLA activity...."

We're going to see the word, or we may see the words "control for" elsewhere in your report. It's a term of art. What does it mean?

A. In this analysis, we are examining first the pattern of killings over time and space and then the pattern of refugee flow over time and space. Our question is: What statistical relationship do the patterns of NATO airstrike and KLA activity have with the patterns of killing and refugee flow? The technique employed is to use the statistical relationship of KLA activity and NATO airstrikes to make an estimate of each point and then remove the effect of that estimate from the pattern, from the pattern of killing or migration. That removal is what we mean by "control for."

Q. Before we leave this page, and I should have dealt with this ²¹⁸⁴ earlier, but I'd just like you to explain footnote 12, which actually relates to the earlier -- the earlier examination reflected in the two tables, the KLA and the NATO tables. You say that, "... for this analysis, we used only the number of reported killings, not the estimated total number."

Just explain that, please.

A. In order to determine when the peak occurred for killings in Figure 8 and 9, we took the simple number of victims reported in each municipality over time. That is, we used only the 4.400 people whose names we can list in order to conduct the analysis in figures 8 and 9. The reason was that, as I discussed earlier in my description of the two-day periods, when we break the data down to 29 municipalities, it is too thin. There is not enough data. There are too many zeros to make the estimates.

Q. Thank you. And I remind the -- it may help if we just remind the learned Judges that your initial process was to go from the figure of 10.000-odd or down to 4.400 and then to expand it, in the method you've described, to 10.356. What you're saying is, on this occasion, you couldn't do that second step?

A. That is correct.

Q. In general -- and I'm sorry to have done this slightly out of order, but there it is. In general, how conservative were the figures you were relying on in every stage of your report?

A. We made every attempt to be as conservative as possible. In this sense, I mean conservative as we made decisions that would all tend to ²¹⁸⁵ reduce the total estimate and to reduce the estimates at each space-and-time point. The point of that was to present a statistical case which would be as favourable as possible to the hypotheses which we ultimately have rejected here, but we gave that the strongest test we could.

The most obvious sense of conservatism that I can give as an example is that whenever we compared two people who may potentially -- two names which potentially might refer to the same person, we assumed that the names did refer to the same person unless we were able to demonstrate positively that they did not refer to the same person. Again, that tends to reduce the total lists and to reduce the overall estimates.

Q. After that digression, then I can return to the second method, the controlling-for method. Can we go to page 13 of a 15, 16-page report? We're approaching its conclusion. Here is the third and, I think, last part of the report where we'll have to have some guidance on statistical concepts from you.

If we look at the graph at the top of the page, we can see a new word featuring, underneath "People killed" and another line, called "Residuals." People can consume technical ideas in different ways, sometimes by reference to the matter in hand, sometimes by reference to generalities. Can you give us, in a sentence or so, first of all, how Residuals operates in the most general terms? And then we'll look at the facts of this case.

A. Referring to the discussion of controlling for, if we think about the top line in Figure 10, the dark line which -- ²¹⁸⁶

Q. You're now dealing with specifically on the facts of this case. I was wondering if you could just help us conceptually with what a residual is --

A. Right.

Q. -- and then we'll come to the facts of the case.

A. When we wish to examine the effect of one variable upon another, we control for, as I discussed, the hypothesised cause and then we can look at the resulting pattern in the hypothesised effect. If we deduct that quantity created or accounted for by the cause, the remaining value from the original is called the residual, that which remains. The pattern of the residual can tell us a great deal about the structure of the explanation, the internal relationship of these two variables.

Q. Now let's, having described it in conceptual terms, look at Figure 10, the facts of this case. What do these two lines, "People killed" and "Residuals" show?

A. The bottom line, the Residuals, describes the net effect or that which remains after we have controlled for KLA activity and NATO airstrikes. What is striking about the residual pattern is that it largely follows the original pattern. They follow quite closely.

Q. We'll come to the significance of that in a minute, because I think it may help the learned Judges if we break it down step-by-step. The top line is the line we've already seen before, for example, on page 6.

A. Yes.

Q. What you're doing in reaching the residual is deducting in some ²¹⁸⁷ statistical way that part of that line which is explained by NATO and KLA activity; correct?

A. That is -- that's broadly correct, yes.

Q. Were it the case that NATO/KLA activity explained every killing, what would be left? What would be the Residual line?

A. The Residual line, if the explanation were stronger, the Residual line would be merely a roughly flat line, squiggling around the horizontal axis.

Q. Because you will have explained by the hypothesis everything on the Residual line?

A. A statistical correlation would explain the original pattern, yes.

Q. So here you remove what may be explained by KLA or NATO activity and you're left with a line of shape similar to the original line.

A. Yes.

Q. What does that tell you as a statistician?

A. It tells us that the explanation has not changed the pattern very much. The explanation provided by the statistical relationships of NATO airstrike activity and KLA activity does not explain the pattern of killing over time.

Q. Page 14, please. This figure, the figure at the top of the page, Figure 11, is the same exercise conducted for refugee flow; correct?

A. That is correct.

Q. Same thing: If it was wholly explained by KLA and NATO, it would be a flat line hovering around -- or a flat-ish line hovering around the ²¹⁸⁸ north?

A. Correct.

Q. What do you say about the shape that we see of the Residual line?

A. The Residual line does track the original pattern, with peaks and troughs at roughly the same points.

Q. As strong as in the killings or not?

A. Less strong. Less strong.

Q. We see here in footnote 17 that you suggest in the middle of the page in the text that it has -- let's just read that. Shall we go down in the text a little bit? The results, you say, are shown --

MR. NICE: If the booth could take us down in the text a bit. Thank you very much.

Q. You say, in Figure 10 -- you set that out: "The same analysis can be performed for the refugee flow. The results are shown in Figure 11. As with killings, the pattern of the refugee flow, controlling for the correlations with the NATO and KLA patterns, is strongly similar to the original pattern. However, the statistical measures suggest that the KLA activity (but not NATO airstrikes) has a weak but noticeable relationship with the refugee flow pattern."

Just explain that again.

A. I think -- well, when a statistician looks at Figure 11, he or she will ask: There does seem to be some explanation here. Which factors might be responsible for that explanation? So she or he would want to conduct a more detailed analysis, looking at the individual factors - KLA ²¹⁸⁹ activity and NATO activity - and consider their individual contributions to the explanation. We also, my co-authors and I, also felt that this warranted a more detailed analysis over both region and time, which we conducted, and that is referred to in the footnote.

Q. And we should have in mind, without referring back to it, that the figures of refugee migration had shown different patterns regionally; in particular, the southern and the western differing from the northern and the east?

A. That is correct.

Q. So if we can go, quite briefly, in the annexes of the report, to pages 59, 60, and 61. We can look at 59 very briefly.

A. I believe you mean 58.

Q. Oh, yes, 58 as well maybe. Yes, 58. 58, I think, requires some technical explanation. Can you give us that, please. What is it, first of all? It's headed "Regression Coefficients" and "Response Variables," and then we have a number of titles: "Killings Over Time and Region," and so on. Can you explain it, please.

A. These are the coefficients, the actual statistical coefficients that are the model used for the residual analysis we just discussed. Each of the coefficients describes the relationship, all else being equal, between a particular variable and the response variable, that which we are trying to explain.

So, for example, going to the top of the third column in the "KLA (kill)" variable under "Refugee Flow Over Time," the interpretation of the number -634.9 is that, all else being equal, in a given region, at a ²¹⁹⁰ certain point in time, the presence of each additional Serb casualty caused by the KLA would result in 635 fewer refugees leaving that municipality - or excuse me, excuse me - that region in that time, in that period.

Q. Thank you. We're going to be able to understand this chart ourselves probably with but one more or, at the most, two more examples, but it's got one deficiency in it at the moment, by typographical error, and that is that there should be one more asterisk - we'll explain the asterisk - and the remaining asterisk should be in the figure immediately below that, 2.728.6; should it not?

A. That's correct. There should be an asterisk there.

Q. And the asterisks on page 58 - and it's either a single asterisk, as that last entry, or a double or a triple asterisk - what do they represent, or what do they signify, rather?

A. The asterisks describe the relationship between the coefficient listed in the table and the error associated with that coefficient.

Q. There can be cases where the error swamps the statistic so the statistic becomes of no value?

A. That is correct. In fact, those statistics without asterisks should not be interpreted. They are -- the error is too great.

Q. And in the most approximate terms, where the correcting error figure, in brackets, is half or more than the numeric figure above it, plus or minus, then the figure is of no value?

A. It's quite close to half. It's actually -- but it's 1.96. But yes, in terms, you're correct. ^{2191 BLANK PAGE 2192}

Q. The smaller the error figure is, as a proportion of the numeric figure above it, the larger the number of stars, the stronger the value of the conclusion?

A. Broadly, yes. Broadly, yes.

Q. So let us then do one more example to see if we understand this. And "KLA (kill)," "KLA (battle)," self-explanatory, as is "NATO," but what about "Lag-KLA (battle)"? Can you explain that as a title for us?

A. Yes. "Lag" is the statistician's term for what happened in the previous period. So "Lag-KLA (battle)" is the effect of a battle that occurred in the previous two-day period to the migration or killing we are explaining. So we want to consider in this model not only what has happened in the current two-day period, but also what happened in the previous two-day period in order to explain the pattern of killings and refugee flow.

Q. And so to check that we understand how the table may be read by the Chamber, or indeed by other statisticians who may look at the document, if we stay with "Lag-KLA (battle)" and then, at the top, "Killings Over Time and Region," so it's the figure 11.6, with 11.9 in brackets -- well, that won't do, because it hasn't got an asterisk to go against it, so we won't rely on that. "Lag-KLA (battle)" and then "Refugee Flow Over Time and Region," 2.138.4, with three asterisks, that interprets as --?

A. All else being equal, and here I mean all the KLA killings and battles, NATO activity being equal, the effect of each additional KLA battle in a region is to predict 2.138.4 more refugees departing their ²¹⁹³ homes on the next two-day period, in the following two-day period. So it's in the next two-day period.

Q. When you're referring to a battle, you're referring to a battle between the KLA and --?

A. And Serb forces.

Q. Page 59, please, briefly. Simply, the charts at the top within the appendix, this is the same scheme of killings charted, with the residuals underneath, but now broken down by region. And taking it very shortly, do we see a close correlation of residuals and original chart in both north, east, south, and west?

A. Yes. The relationship is close in all four regions.

Q. The same conclusion; namely, that KLA or NATO activity does nothing to explain the pattern of killings.

A. Not quite nothing, but it does very little.

Q. Very little. Thank you. Page 60, please. When we look at refugee flow with the same approach, broken down by regions north, east, south, and west, to the eye, the residuals track less closely than with killings, although they track in the south and west, it may be, more closely than the north and east. But what would your interpretation be?

A. Following that, in the north and east, until the phase 2 peak in mid-April of refugee flow in both regions, the patterns, residual pattern, does not track very closely the original pattern. Then in that mid-April peak, they converge closely, then follow each other roughly through the end, with some divergences, some divergences in the less significant periods toward the end when the magnitudes are very small. In the south ²¹⁹⁴ and west, the residual pattern actually tracks quite closely in the south, and closely, but not as closely, in the west, where there are a few smaller divergences.

JUDGE MAY: Judge Robinson.

JUDGE ROBINSON: Dr. Ball, I wanted to have clarified a few matters. In answer to me earlier, you said that when you first presented, I think, your first two reports to the OTP, you indicated that you would have to make them more rigorous for presentation in Court. I'd like you to tell me what you did to make it more rigorous. I heard you mention one particular matter which perhaps could have had that result; that is, you said wherever there was a conflict, then you took the course which was favourable to the hypothesis that you eventually rejected. If that is one, then you can say so and tell me what are the other methods that you used to make the presentation more rigorous.

THE WITNESS: That is indeed one, the one that Your Honour listed. Let me highlight three others quite briefly. The first is that, relative to the first two reports, we sought, my co-authors and I sought much more extensive peer review, as we described at the beginning of the presentation, with a substantial committee of statisticians from Europe and the United States, each of whom had the opportunity to review two preliminary drafts of the report. In earlier work, I had been less formal and worked with a number of advisors who read drafts and made comments to me, but not in this much more formal process that we instituted for this report.

The second additional rigour that we applied has to do with the ²¹⁹⁵ matching process. Recall that the data source for killings are lists of names of people who have been described as having been killed in witness reports, and to do the estimation that we describe here, from 4.400, we have to determine which killings are in multiple reports. I have to look at the list from Human Rights Watch, several hundred names, and determine which of those names are also in the other lists. This is hard because there are a number of factors which complicate making exact comparisons. To assure that that was as reliable and rigorous a process as we could, every intersystem comparison, as I just described, between the ABA/CEELI data, the exhumation information data, the HRW data - Human Rights Watch data - and the OSCE data, every one of those comparisons was done by two or more different coders, different staff people, independently. So I -- every single one of those matches was done by two people, and many of them were done by three, some by four, and some by five. Those people went through an extensive training process, as described in Appendix 1.

For purposes of your question, I think the most important note was that we were able to measure, and thus quantify, the reliability of the matching process, reliability here in a formal, statistical sense, which is that if we did it again, we would reach the same conclusions. That's what I mean by "reliability." The agreement between the raters was 94 per cent. Ninety-four per cent of the decisions were reached the same way by all the coders, all the staff people who looked at the names to make decisions about whether they should be matched. All the disagreements that were found in the remaining 6 per cent were reviewed by a supervisor ²¹⁹⁶ and a final decision reached.

So that's the second of the additional kinds of rigour we brought to this process that I had not done in earlier projects. It's very difficult. There's one more.

JUDGE ROBINSON: Yes. Go ahead.

THE WITNESS: The final one is that the statistical calculations that we've just been discussing, the estimation procedure and all this analysis, was done twice by two different people independently, on different computers, using different software, in fact, using completely different operating systems. We started with the same data, we established the mathematical procedure we were going to use, and then she, one of the co-authors here, and I independently wrote all of the software to do the estimation. And we kept doing it until our answers agreed to many different -- to many digits. We made sure that completely different estimation procedures, completely different software, different computers, different analysts, reached the same conclusions. Again, that's -- they're expensive, but we felt that it was warranted in this case, given the gravity of the questions. So those are the three.

JUDGE ROBINSON: Thank you. Notwithstanding the initial rigour and the additional rigour that you applied, are you able to put your conclusions any higher than what you said in relation to Figure 8 on page 11, that it tends to reject a particular hypothesis? Are you able to put it any higher than that, or that's the highest that you can put it?

THE WITNESS: Yes, Your Honour, I can put it higher than that. And I think what I would say is at the point of Figure 8, the strongest ²¹⁹⁷ conclusion, if I were to stop there, the strongest conclusion I would want to make is that it tends to reject. As we continue through these additional analyses which I've just described, we can make substantially stronger conclusions.

JUDGE ROBINSON: I see. Okay. The next question I wanted to ask you was whether this statistical method that you have developed, whether it has been used in court proceedings before, in judicial proceedings.

THE WITNESS: Let me separate my answer into three small parts, if I may. The procedure used to describe the migration process is composed of a statistical tool kit of many different methods, each one of which would be easily recognisable to any professional statistician or demographer. I don't know if it is -- if those methods have been combined in this way before, but certainly the component methods would be widely used in any statistical presentation, including those that would be given before judicial proceedings. I do not have specific knowledge. On the second method, which is the multiple systems estimation, which uses the -- which estimates the total number of people killed, again I don't have specific knowledge about judicial use; however, this method is widely used to adjust censuses, which, as Your Honours are probably aware, is the basis for legislative apportionment, and so tends to have very high stakes.

So it is a closely investigated scientific method. It is well known. And since censuses are always partial - they never count everybody - this is the method to which demographers turn to to adjust ²¹⁹⁸ censuses.

In the third answer, I can answer yes, although I can't provide examples. In the regression analysis, in the residual analysis, this is among the most common statistical techniques used anywhere. I mean, any time social scientists, statisticians, or demographers want to examine the relationship of several or many variables over a large --

JUDGE ROBINSON: I'm talking about court proceedings in particular.

THE WITNESS: [Interpretation] I believe it must have been, sir, Your Honour. I don't know of a specific case. I'm not a lawyer, so I'm unfamiliar with the details.

JUDGE ROBINSON: The next matter that you can help me with is this: I can understand the utility of your method in analysing broad patterns and broad trends, but in this case, we are concerned with specific allegations that relate to particular geographical areas. There's an allegation that there was a movement of people from A to B. This was caused by a particular phenomenon. There is evidence that, prior to the movement, there was KLA activity, there was NATO activity, and there was Serb activity. The task which the Trial Chamber has is to determine what was the cause in relation to the particular charges in the indictment.

I'm not sure I understand how your statistical method, which I think is very useful for broad patterns of conduct, is particularly helpful in relation to that specific issue which is particular and specific to a particular area and in relation to which there may or may ²¹⁹⁹ not be evidence.

THE WITNESS: [Interpretation] I think that when I reach the conclusions of the -- my direct presentation, it may shed some light on your question about the causes which you've proposed and which I've adopted as hypotheses here, as I will find two of those causes to be statistically implausible, given the available evidence.

JUDGE ROBINSON: Where, for example, the NATO bombing took place after the movement. I mean, I would have thought the conclusion that the NATO bombing was not the cause to be quite an unremarkable conclusion. But the real issue is where we have evidence of the two or three phenomena I mentioned taking place before and where there is a conflict, and that is what I think we need help with, and I'm trying to see how your exercise helps us with that particular issue. But I understand you'll be coming to that.

MR. NICE:

Q. We have yet to reach the conclusions, but in light of the questions you've been asked, just a couple of matters to pick up on. First of all, just dealing with peer review. Can we find some examples in the report? I think one may be at page 61 of additional things that you did as a result of peer review suggestion. Am I right about page 61?

A. Yes. Yes.

Q. I wasn't proposing to go into it in detail for the reason you'll give in a second, but "Estimated killings over time with alternative date assignments," what do these mean, "alternative date assignments"? ²²⁰⁰

A. Much of our argument turns on whether killings in this case happened exactly at the time that we estimate that they occurred, and so a reasonable question might be: What if the witnesses got the dates wrong? What would happen to your analysis? So our response to that quite reasonable question is to look at each date reported by the witnesses and move them as far earlier and later as is remotely plausible. And I can describe how that was done in detail if there are additional questions. But we shifted the dates as early and as late as were possible or we thought were plausible for each date.

Q. This was a peer review proposal?

A. It was a question that was asked, simply: What would happen to your analysis if the date reporting were imprecise?

Q. And -- sorry. I mustn't go too fast. And the charts or graphs we see, did they lead to any different conclusion of any kind?

A. I do not think so. I believe they are substantially the same, given the radical adjustment that we made to test that.

Q. The second point arising from His Honour's questions, or one of them, in particular multiple systems estimation and its use in court, in fact, multiple systems estimation was not used for the first and, in a sense, most readily accessible conclusions in figures 8 and 9, was it?

A. That's correct.

Q. It was used in the others.

A. It was used in the others.

Q. But now let's come to --

JUDGE MAY: Before you do, it's now five to four. Plainly we're ^{2201 BLANK PAGE 2202} not going to get to another witness tonight. Four o'clock is the time for a break. Are you going to be finished shortly?

MR. NICE: I would think I'll be finished shortly because I only have the conclusion page of the witness to go to apart from one or two other references.

THE ACCUSED: [Interpretation] I have a comment to make.

JUDGE MAY: Well, Mr. Milosevic, make it shortly.

THE ACCUSED: [Interpretation] An objection, in fact. Very briefly. As you're speaking about the next witness, I did not ask for any postponement of the cross-examination. I have heard that that was something that one of the friends of the Court asked for. But as you know, the friends of Court do not represent me. And a moment ago when I showed a whole lot of papers, I did so, I showed them as being an example of the improper manner and prejudicial manner and not as being my request for anything, for any kind of understanding, because there's no understanding for me at all, and especially no understanding with respect to my possible consultations with the advisors and consultants that the other side mentioned, because I cannot contact my advisors other than by telephone. So please bear that in mind.

JUDGE MAY: Let us see what you're saying. Are you saying you want to cross-examine the witness now?

THE ACCUSED: [Interpretation] I said that I did not ask for any postponement of cross-examination, which implies that I shall continue, once this witness finishes his testimony, with my right to examine him in cross-examination. That is, I suppose, something which is the normal ²²⁰³ order of the day and nothing that I make a special request of or for.

JUDGE MAY: Very well. We'll adjourn now for five minutes for a break and then we'll go on with the rest of the evidence.

--- Break taken at 3.55 p.m.

--- On resuming at 4.01 p.m.

JUDGE MAY: Yes, Mr. Nice.

MR. NICE: Thank you.

Q. If we can now go to the conclusions on page 15.

JUDGE ROBINSON: Mr. Nice, just to explain, the questions that I asked, I had intended to ask them after your examination-in-chief. So I thought you had finished, because it's not my style to interrupt examination-in-chief.

MR. NICE: I'm grateful for your explaining that. No, no. I diverted to the appendices because they were referred to in the body of the text.

Q. Page 15, then, please. The format of it, we can see you've already been dealing with the two hypotheses. First of all, KLA, NATO, and then with the Yugoslav forces, but --

MR. NICE: A little up the page, please, if we may. The other way. Thank you very much. A bit further. Thank you very much.

Q. Dr. Ball, first of all, on the analysis so far described using the different techniques and samplings that you've referred to, what was your conclusion as to the possible involvement of NATO and or the KLA in either the refugee movement or the killings?

A. First, we rejected the hypothesis that NATO involvement caused ²²⁰⁴ either killing or the migration. In both of the methods, NATO showed inadequate correlations to be considered a plausible cause. Therefore, we rejected the hypothesis that NATO was a cause of either killing or migration patterns.

Second, we reject the hypothesis that KLA activity caused the patterns, the statistical patterns, of killing. Thus KLA activity is not a plausible cause for more than a small fraction of the killing, a cause directly or indirectly of more than a small fraction of the killings. The fourth sub-piece of it is a bit more complex, as we were discussing earlier. There does seem to be some correlation at certain times and certain regions between KLA activity and refugee flow.

Q. And by "KLA activity," as you've already told me, you were referring to KLA/Serb conflict?

A. That's right. Serb casualties caused by the KLA and exchanges of fire between KLA forces and Yugoslav forces.

Q. Let's now turn to the third hypothesis which we can see on the page, headed "Yugoslav forces."

A. Uh-huh.

Q. Turning to that third hypothesis, you say in your report that: "Yugoslav forces organised and implemented a systematic campaign of violence resulting in killings and refugee flow: the statistical analysis of correlations cannot prove that the Yugoslav forces were the external influence responsible for the observed patterns. However, the findings of this study are consistent with the hypothesis that action by Yugoslav forces was the cause of the killings and refugee flow." ²²⁰⁵ Just pausing there and knowing that there's something in the following paragraph, does that remain your opinion?

A. Yes.

Q. And of course, as well as -- I mustn't go too fast, particularly at this time of the day.

As well as peaks, there are troughs in both the refugee flow and in the killings, and do you find one circumstantial link that may assist?

A. Yes, we did.

Q. Which would be the best chart for us to look at to see this circumstantial link, or will any of them do?

A. Any of them will do. Perhaps Figure 2 is the clearest. I believe from page -- it might be 6. I'm not sure.

MR. NICE: I wonder if you can just lay that back on the ELMO.

Q. Page 6, the earliest and, in one way, the simplest, but the others to similar effect. Yes, please go on.

A. On the evening of April 6th, Yugoslav official sources announced a unilateral ceasefire by Yugoslav forces to honour Orthodox Easter, which that year fell on April 11th, Sunday. The 6th of April was a Tuesday, and the announcement was made that evening. On the following day, what is very interesting is that using these widely differing statistical techniques and completely different data, we find a consistent and drastic decline in both refugee movement and people being killed. That decline is exactly coincidental in time with the Yugoslav government declaration of a ceasefire. Both people leaving their homes and people killed declined drastically on the 7th and remained at low levels for several days. ²²⁰⁶ I don't know if the Yugoslav government rescinded the ceasefire, but it seems that it -- that hostilities began again sometime during the weekend, approximately at the time of Orthodox Easter. Both people leaving their homes and people killed increase again at that point. That circumstantial link, which is, I think, quite dramatic in light of the very high levels of killing and refugee flow a few days prior to the ceasefire announcement, leads my co-authors and I to confirm - or, excuse me - to state that the evidence that we have found, the statistical evidence, is consistent with the hypothesis that Yugoslav forces are the cause of the observed patterns.

Q. Thank you. In addition to requests from your peer review to consider certain matters and to add matters to your report, did you receive a request or instructions from the Office of the Prosecutor, to which you responded by the short two-and-a-half, three-page supplement, with annexes, that's been served on the Court?

A. Yes. We were -- several requests were made of us in February this year.

Q. I'm going to break them down, now that we are familiar with what you were doing, to very simple proposition and very simple conclusions, you being in a position to turn any questioner to the workings it asked. Were you asked to do some further breakdowns and to check results by different broken-down approaches?

A. Yes, we were.

Q. Explain just in a sentence or so.

A. We were asked to consider the relationship between refugee flow ²²⁰⁷ and killing directly, those two compared to each other, using the same method that we looked at in Figures 8 and 9, with the different percentages. We called that the peaks-versus-presence method.

Q. Yes.

A. We conducted the analysis, and the hypothesis there would be that if the two series are closely related, then we would find the presence of killings before the peak refugee flow for each municipality, there would be a high level of coincidence there. Similarly, we would expect that the peak in killings would coincide with or occur just before a very high - or, excuse me - a recorded refugee flow. We found that in both cases, the number of municipalities which satisfy the hypothesis was quite high, 69 and 83 per cent respectively.

Q. The second exercise you were asked to perform?

A. We were also to consider the independent effects of the NATO and KLA activity on killing and migration.

Q. So if we want to look at a chart or a map and say that you reconsidered the same topic but now with one factor on its own, which would be the best ones to look at? Figures 10 and 11 in the original?

A. Figures 10 and 11 would be the ones we took apart in the subsequent figures.

Q. So Figures 10 and 11 deal with KLA and NATO collectively. You then dealt with them separately?

A. That's right.

Q. Are the conclusions the same or different?

A. The conclusions are the same, that bombing has very little effect ²²⁰⁸ on either pattern -- NATO airstrikes, bombing, has very little effect on either pattern. KLA activity has no effect on killing but seems to have a relationship with refugee flow in the northern and eastern regions.

Q. And my last question, which is just a matter of detail, but it may help the Chamber. If you can just look at Appendix 3, the additional sources on KLA and NATO activity, please. It's page -- just the page before page 64. It's an unnumbered page, so it's really 63, but it hasn't got a number on it. Two pages of this, I think, that page and the following page.

To give us an idea of the rigour of your report, this is simply your sources or additional sources, as you explain, for KLA and NATO activity. You list, starting with the Albanian Human Rights Group and running on that page down to Kosovapress, and on the following page you can see all the names and bodies to which you referred. On the following page, we pick it up at Kosovar Media and run down to V.I.P. Daily News Report.

Are these all bodies that you turned to or scoured or searched for details of NATO and KLA activity in order to run your hypothesis?

A. Yes. The NATO activity was limited to the Yugoslav sources, a subset of the ones listed here, but all of these contributed elements to the sources we used for the KLA measures.

Q. Thank you, Dr. Ball. If you wait there, you may be asked some further questions.

JUDGE MAY: Yes, Mr. Milosevic.

THE ACCUSED: [Interpretation] I wish, first and foremost, to ²²⁰⁹ finish with my objection, because you interrupted me. I consider this to be an elementary minimum of decency --

JUDGE MAY: What is the objection?

THE ACCUSED: [Interpretation] I said why and for what reason I illustrated the extreme incorrectness and isolation and every possible way of depriving me of any equality before this illegal institution of yours. I asked for a minimum of decency at least for this list of witnesses that the opposing party is bringing here, that I should be given this list at least a week in advance, and I expect to be given in the future a list of the right order of witnesses one week in advance. As for this witness, I shall start with the cross-examination immediately. Cross-examined by Mr. Milosevic:

Q. [Interpretation] In order to put the first question I have, I shall first quote what it says in these papers, because there is an explanation here of the organisation that the witness works for, that is to say, the American Association for the Advancement of Science, et cetera.

This is what it says here: [As interpreted] "The material contained here are the opinions of the author and editors and cannot be considered to be the view of the American Association for the Advancement of Science...," in parentheses, "... (The original American Association for the Advancement of Science, AAAS)...." And further on, "... or the Central and East European Law Initiative."

So they cannot be construed to be the view of either one ²²¹⁰ institution or the other.

And also, it says here: "The ABA/CEELI, the American Bar Association, Central and East European Law Initiative, or any of the contributing organisations. The AAAS Committee on Scientific Freedom and Responsibility ... in accordance with its mandated and Association policy, supports publication of this report as a scientific contribution to human rights. The interpretations and conclusions are those of the authors and do not purport to represent the views of the AAAS Board, Council, the CSFR, or the members of the Association."

So they do not represent the views of any one of them. "Likewise, the views expressed herein have not been approved by the House of Delegates or the Board of Governors of the ABA and, accordingly should not be construed as representing the policy of the ABA. "Nothing contained in this publication is to be considered as the rendering of legal advice for specific cases, and readers are responsible for obtaining such advice from their own legal counsel. This publication and any forms and agreements herein are intended for educational and informational purposes only."

That's what it says here. Signed by the American Association for the Advancement of Science.

Since this is what it says here, my first question is the following: What it says here, does that mean that your work cannot be used for purposes of the indictment?

A. This report is not legal advice. I am not a lawyer. I am, ²²¹¹ however, a scientist, and I believe that I can offer some scientific insight into these patterns. That has been my effort.

Q. You have said that this is statistical research.

A. I'm sorry, is there a question?

Q. Yes, that is a question. This is statistical research.

A. Yes.

Q. So this is not sociological research.

A. Something may be simultaneously sociological and statistical. "Statistical" refers to the methodological approach taken.

Q. You spoke exclusively about statistics and statistics experts. There are no elements of sociological research in here. Yes or no.

A. The question confuses me. I think that sociological research in my country and as practised at the university where I took my doctorate is highly quantitative in nature and this would be recognised, I believe, by any American or probably most European sociologists as sociology. "Statistical," as I said earlier, refers to the methods that were used.

Q. These are not methods; these are methodological procedures that you apply. There is a major difference between a method and methodological approach. Do you know how to explain the difference between a method, a sociological method, for example, and a methodological procedure that you applied here?

JUDGE MAY: What is the point of this question?

THE ACCUSED: [Interpretation] Well, what is the point of explaining that this statistical analysis of incorrect data is a sociological research? The witness mentioned the application of a ²²¹² method. He mentioned that all sociologists accept, espouse, quantitative methods. Quantitative methods have their limited reaches, and not everything can be explained by quantitative methods, because if everything could be explained statistically, then there would be no need for sociology to exist, as the most general of sciences about society. Therefore, I ask the witness, who, as we have heard, has a doctorate in sociology, to define briefly the difference between sociological method and methodological procedure applied herein.

JUDGE MAY: This is an irrelevant question. Ask him about his evidence.

THE ACCUSED: [Interpretation] Mr. May, you cannot restrict me in my questions, especially --

JUDGE MAY: I can in what is relevant. Perhaps Dr. Ball can help us.

You've heard what's being put. Is there any point that you think you can answer in relation to it?

THE WITNESS: Not really.

JUDGE MAY: Thank you.

MR. MILOSEVIC: [Interpretation]

Q. That answer is sufficient for me as well. As a doctor of sociology, you cannot answer this question about the difference --

JUDGE MAY: [Previous translation continues]...

THE ACCUSED: [Interpretation] All right. I'm not going to make any comments.

MR. MILOSEVIC: [Interpretation] ²²¹³

Q. So in addition to the fact that we have such an explanation of the American association, your answer to the question whether this explanation precludes the possibility of using your work for the needs and purposes of the indictment is that you are not a lawyer. You didn't say yes or no.

A. I'm not a lawyer, and therefore I do not believe I'm competent to speak to whether this is -- this answers your question or not. That's my answer. I'm a statistician. I've offered statistical and sociological analysis.

Q. Well, that's precisely what I have been asserting, that you are a statistician, and that this is not a sociological analysis but a statistical approach.

My next question: Were you asked to look at the movements of Serb refugees during the NATO aggression?

A. No.

Q. You said that this work of yours was financed by US government resources.

A. In part.

Q. In greater part or smaller part? Predominantly or negligibly so?

A. The prior study to this was financed to my partners with US government funds, and this report was financed through funds to my partners with US government funds.

Q. Since you were not asked to study the flows of Serb refugees from Kosovo and Metohija, was it perhaps in your interest to compare in your statistical analysis, to do this at your own initiative?

A. No. ²²¹⁴

Q. Have you ever heard of any figure related to the number of Serb refugees who left the area of Kosovo and Metohija during the bombing?

A. I saw a paper presented at an international conference by a Serb statistician in September 2000 on this topic.

Q. And do you know that this figure is about 100.000 Serbs who fled due to the intensive bombing of Kosovo?

A. I do not know that. I do not recall the specifics from his presentation.

Q. And is it your assessment that for the purposes of statistically analysing the data you have analysed here, it would have been useful for you to have taken into account these flows as well, related to Serbs and the non-Albanian population of Kosovo in general, and compared it to these flows?

A. This study looks at refugee flow and killing against Kosovar Albanians. As such, it does not contain analysis of other groups.

Q. But you know very well that you are dealing with a complex issue related to a very complex conflict that was taking place in Kosovo and Metohija. Do you think that it is a considerable simplification or one-sided approach to have this boiled down to a statistical representation of figures pertaining to Albanian refugees only?

A. I'm sorry. There's a lot of parts to that question. I'm not sure I understood. Could you rephrase it for me? I'm sorry.

Q. In Kosovo, was this a complex conflict, a conflict that cannot be reduced only to one of its factors, that is to say, the statistical study of Albanian refugee flows? ²²¹⁵

A. I'm not a political analyst, so I'm not -- I don't think I can answer whether this is complex. I have done a study of killings and refugee flows.

Q. Can studies be made about refugee flows if there are no elementary ideas about the historical, political, military, economic, social - and let me stop there - and any other context in which this analysis is sought to be made?

A. This study examines the statistical patterns in light of three hypotheses, which I believe are the most important hypotheses that have been advanced to explain them. Neither I nor my co-authors are competent to make any additional analysis beyond the conclusions in this report.

Q. And who put these three hypotheses forth, the ones that you just mentioned?

A. We took the hypotheses from political debates which we have heard expressed by different actors. Most locally to our work today, I have heard two of the hypotheses advanced by the accused in this case, and the other is in some sense advanced by my highly non-technical understanding of the indictment. So I believe they are relevant to our discussion today.

Q. Are you referring to some of my hypotheses? Did I understand you correctly?

A. Yes.

Q. Then you misunderstood them, but we are going to get to that later.

When presenting your C.V., when your C.V. was being presented, we ²²¹⁶ heard that you had dealt in such statistics, in such statistical research, in a considerable number of countries over the past few years.

A. Yes.

Q. Is it possible to state that all the mentioned countries had crises, conflicts, and that the American administration was involved in all of these crises and conflicts?

A. Yes, all the countries had some sort of crisis or conflict. I don't think the involvement of the US government is clear in all of them. In some, it might be.

Q. And where was it not involved out of those that you mentioned?

JUDGE MAY: What's the relevance of this, Mr. Milosevic, in relation to his evidence?

THE ACCUSED: [Interpretation] Mr. May, the relevance is in the fact that statistics that are well-known among the greatest statisticians of the world that can prove anything by statistics [as interpreted], and this is done to serve the purposes of the American politics aimed at enslavement. Chemistry is being misused for making chemical bombs, atomic energy for making atomic bombs --

JUDGE MAY: [Previous translation continues]... the witness. In the various -- just a moment. In the various studies that you were carrying out, were they in any way to serve the purposes of American politics?

THE WITNESS: I have never had, knowingly, contact with a representative of the US government in any of my work, in the sense that they would advance US interests through my work. However, I think there's ²²¹⁷ a misconception, which may be helpful here, since my work in both El Salvador and Guatemala was actually quite critical of US interest and of US policy at the time. So this may help clarify the question.

JUDGE MAY: We'll adjourn now. It's past half past 4.00. Would you be back tomorrow, please, half past 9.00, Dr. Ball.

--- Whereupon the hearing adjourned at 4.35 p.m., to be reconvened on Thursday, the 14th day of March 2002, at 9.30 a.m.