1 Thursday, 23 August 2007

2 [Open session]

3 [The accused entered court]

4 --- Upon commencing at 2.16 p.m.

5 JUDGE ANTONETTI: [Interpretation] Mr. Registrar, please call the

6 case.

7 THE REGISTRAR: This is case number IT-04-74-T, the Prosecutor

8 versus Prlic et al.

9 THE INTERPRETER: Microphone, please.

10 JUDGE ANTONETTI: [Interpretation] Thank you, Mr. Registrar.

11 Let me greet everyone in this courtroom, the Prosecution, the

12 Defence, the accused. Today we have a witness to hear. As I indicated in

13 a written statement, both the Defence and the Prosecution will have three

14 hours.

15 Furthermore, I would like to tell you that the Trial Chamber has

16 issued a number of rulings these past few days. Yesterday two major

17 rulings have been filed. One of them is related to the admission of the

18 written statement made by an accused. The second decision is related to

19 the 23 additional hours that we granted to the Prosecution in order for

20 them to complete the presentation of their case. I would like to invite

21 all of you to read these decisions related to these two topics.

22 Furthermore, I'd like to tell the parties that there is bank

23 holiday in October, the 24th of October. It's a Wednesday. Because of

24 this -- because we should be sitting on Thursday the 25th of October, but

25 I think it would be better for everyone if the Prosecution did not have

1 any witnesses on the 25th of October in order to give a break for

2 everyone, and even more so since at that time we will be in the final

3 stretch of the of Prosecution case. That's a time when the Defence will

4 have to get ready for the rest of the proceedings. Therefore, I'd like to

5 invite the Prosecution not to have a witness testify on the 25th of

6 October.

7 As you know, for reasons beyond our control we were not able to

8 sit two days this week, and the same will happen next week because next

9 week, on Wednesday the 29th, we won't be sitting because other trials are

10 taking place in the other courtrooms, in the courtrooms. Cases have to

11 take turns in the courtrooms, and as a result we won't be sitting next

12 Wednesday.

13 Furthermore, I will immediately issue two oral rulings. The Trial

14 Chamber would like to ask the Defence to give an oral response to the

15 Chamber's legal officer in relation to protective measures that were

16 requested on the 12th of June, 2007 for a witness referred to in an

17 application for admission of evidence pursuant to Article 92 bis A and B,

18 Rule 92 bis A and B, Ljubuski.

19 I would like to ask you by the end of the day to tell the

20 Chamber's legal officer what is your position with respect to this

21 application for evidence.

22 Second ruling: The Defence asked for additional time to respond

23 to three motions. On the 21st of August, 2007, the Chamber received a

24 joint motion by the Defence where the Defence asked to be granted

25 additional time to be granted until the 8th of October, 2007 to respond to

1 three motions of the Prosecution, the first one being application for

2 admission of evidence pursuant to -- to Rule 92 bis A and B for Dretelj,

3 Gabela, and others, dated 13th of August, 2007.

4 Second Prosecution motion, application for admission of

5 documentary evidence with respect to the Heliodrom, dated 15th of August,

6 2007.

7 Third motion: Application for admission of documentary evidence

8 with respect to Dretelj and Gabela, dated 21st of August, 2007.

9 The Trial Chamber would like to remind the party that it has

10 already granted additional time for several responses, namely by its oral

11 decision of the 12th of July, 2007. Considering that the three

12 above-mentioned motions that the Defence needs to respond to require a

13 great deal of work, the Trial Chamber believes that additional time should

14 be granted to the Defence.

15 Having said that, the Trial Chamber believes that two additional

16 weeks will be amply sufficient. As a result, pursuant to Rule 126 bis,

17 the Trial Chamber partially grants the Defence motion and grants the

18 Defence 14 additional days to file their responses to the three motions.

19 Let me add that pursuant to Rule 126 bis, the Defence normally has 14 days

20 to respond to a motion. We grant them two additional weeks. In other

21 words, the Defence has now one month to respond to each one of these

22 motions.

23 We are now going to have the witness brought in for the hearing to

24 be organised properly, we need the Prosecution to complete the

25 examination-in-chief today, and the cross-examination will take place next

1 time, because the witness will come back to be cross-examined next week.

2 This is the way we are going to proceed today.

3 Let me ask the usher to bring the witness in.

4 Oh, yes. Mr. Karnavas, you wanted to take the floor.

5 MR. KARNAVAS: Yes. Two points. First with respect to the -- the

6 decision that was recently made with respect to the statement. I'm not

7 going to comment on behalf of my colleagues because they represent their

8 own clients, but it's my understanding that the statement comes in against

9 all of the accused, they may wish to speak up on that, but we would be

10 moving for certification, and because it is in French and we want to be

11 precise in how we respond, we would ask some additional time, say for

12 instance five days after it is fully translated into English to file our

13 motion for certification for appeal of that particular decision.

14 That's the first, and you may decide what to do with that, but

15 primarily I want to speak with respect to something else that I find

16 rather regrettable to speak about and rather disturbing.

17 Last Friday there was a hearing where apparently -- you were

18 presiding over that where Mr. Seselj indicated that I had been engaged in

19 illegal activity, including smuggling. Now, I am told -- I saw a part of

20 the transcript. I'm told that there was no reaction from the Bench. In

21 other words, one could interpret that as being an acquiescence or

22 appeasement to what Mr. Seselj was saying. Now, I'm not interested in

23 responding to that (redacted)

24 (redacted). I do not intend to do that. But I do expect

25 that when such matters are being brought before a Presiding Judge in this

1 particular trial where it affects my client because it affects my ability

2 to advocate properly for the client, I expect the Trial Chamber either to

3 ask for clarification or do what Judge May did on many occasions in

4 Milosevic, and that is to turn the mic off.

5 There are rules and procedures and everyone including Mr. Seselj

6 has to comply with those rules and procedures. The Judge cannot allow

7 that individual, albeit he might have the backing of one of the permanent

8 members of the Security Council, Russia in particular, it's an open

9 secret -- or I should say not secret, but it's been widely held that

10 Russia is pressuring this Tribunal to give Seselj whatever he wishes,

11 which may be the reason why it appears that this institution is rather

12 bankrupt when it comes to finding ways of controlling that gentleman, but

13 I think that when his goal, as he stated, is to undermine the integrity of

14 this Tribunal, is attacking another lawyer who is representing someone

15 else, a Croat, one of his -- a nation which he, during his period in his

16 activities, in his admissions were against -- was against, that is a point

17 in time where I must say something has to be done.

18 Now, if, Your Honour, you would believe what Mr. Seselj said, and

19 there may be individuals out there that believe that, but if you do indeed

20 believe that, but then I have no choice but to ask for your immediate

21 disqualification effectively as of this moment, to step down and let

22 somebody else preside, because that means that I have been casted as a

23 criminal, and I have been judged as such, and as a result Mr. Prlic cannot

24 get a fair trial.

25 If on the other hand, which is what I believe the case to be, you

1 do not believe what Mr. Seselj said, then I think that Mr. Seselj must be

2 publicly reprimanded and henceforth in any future proceedings you, as long

3 as you preside in that case, need to control that gentleman. He is not

4 afforded the opportunity, every time he comes into court, to disparage the

5 Prosecution or Defence counsel or witnesses or individuals that have

6 nothing to do with his case. He may represent himself, but he must

7 conduct himself in accordance with all the rules and the procedures and

8 the code of conduct that all lawyers must comply with. He is not above

9 the law, and on behalf of my colleagues here who have been on the

10 receiving end and on behalf of the an association whom I represent as a

11 president, I say this institution has to protect the integrity of the

12 Defence counsel. We are not punching bags. And individuals such as

13 Seselj are not permitted to make gratuitous remarks whenever they feel

14 like, however they feel like, with impunity.

15 I regret to have to say this, but I must say on behalf of my

16 client, because I am concerned that because of these sorts of remarks that

17 went unanswered now I'm less of an advocate, less capable to put my case

18 forward on behalf of Mr. Prlic. I care not for my reputation. I care for

19 my client's rights, and my client has been affected by those remarks.

20 Thank you.

21 JUDGE ANTONETTI: [Interpretation] Thank you, Mr. Karnavas, for

22 these observations. Yes, indeed during the last Status Conference in the

23 Seselj case he made the statements you have mentioned. The Trial Chamber

24 will consider the matter and will consider your submissions, the

25 submissions you have made, and the Trial Chamber will consider whether

1 any -- anything should be done as a result of this.

2 Of course as a Pre-Trial Judge in that case, I'm in a position to

3 ask Mr. Seselj to supplement his previous statement. That's all I can say

4 at this stage. We'll now have the witness brought in.

5 Yes, Mr. Praljak?

6 THE ACCUSED PRALJAK: [Interpretation] I'd like to ask Your Honours

7 to allow half an hour longer for this witness, because it is a complicated

8 expert document, and if you could accord half an hour longer for the

9 examination.

10 JUDGE ANTONETTI: [Interpretation] We'll consider the matter once

11 the examination-in-chief has been completed.

12 We'll now have the witness brought in.

13 We need IC numbers to be given to a number of documents. We'll do

14 that before we bring the witness in.

15 Mr. Registrar.

16 THE INTERPRETER: Microphone, please.

17 THE REGISTRAR: Several parties have submitted lists of documents

18 to be tendered through Witness Antoon van der Grinten and the following IC

19 numbers have been allocated to them. The list submitted by the OTP shall

20 be given Exhibit number IC 635; the list submitted by 3D shall be given

21 Exhibit number 636; the list submitted by 5D shall be given Exhibit number

22 IC 637; the list submitted by 6D shall be given Exhibit number 638; the

23 list submitted by OTP in -- concerning the witness Larry Forbes shall be

24 given Exhibit number 639; the list submitted by 3D concerning the same

25 witness shall be given IC 640; and the list submitted by 5D concerning

1 Larry Forbes shall be given IC number 641. Thank you, Your Honours.

2 JUDGE ANTONETTI: [Interpretation] Fine. Thank you very much.

3 [The witness entered court]

4 WITNESS: EWA TABEAU

5 JUDGE ANTONETTI: [Interpretation] Madam, could I ask you to stand

6 up.

7 Could you give me your first name, last name, and date of birth.

8 THE WITNESS: My name is Ewa Tabeau. I'm born 26th of April,

9 1958.

10 JUDGE ANTONETTI: [Interpretation] What is your current occupation?

11 THE WITNESS: I am a demographer. I'm working at the Office of

12 the Prosecutor.

13 JUDGE ANTONETTI: [Interpretation] Madam, have you ever testified

14 in other cases in this Tribunal? If that's the case, in which trials have

15 you testified?

16 THE WITNESS: Yes, I did. I testified in this Tribunal at several

17 cases. I testified in cases such as Slobodan Milosevic,

18 Dragomir Milosevic. I have here a list of cases. In -- in Stakic,

19 indeed, and in other cases, several other cases.

20 JUDGE ANTONETTI: [Interpretation] Will you be testifying in other

21 cases in the near future?

22 THE WITNESS: Not that I am aware right now.

23 JUDGE ANTONETTI: [Interpretation] Fine. I'm going to ask you to

24 read the solemn declaration.

25 THE WITNESS: I solemnly declare that I will speak the truth, the

1 whole truth, and nothing but the truth.

2 JUDGE ANTONETTI: [Interpretation] Thank you very much. You may be

3 seated.

4 Madam, I'm going to give you a few explanations, but I will be

5 short because you're a member of the OTP so you're familiar with the way

6 we proceed here. Furthermore, you've already been examined and

7 cross-examined. As a result all the information I could give you, you

8 already know.

9 You will first answer questions put to you by Mr. Stringer. That

10 will take place today. And next week you will be answering to the

11 questions put to you by the Defence as part of the cross-examination, and

12 also to questions put to you by one of the accused, because I think one of

13 the accused wants to put questions to you. That's what I wanted to tell

14 you.

15 Mr. Stringer, you have the floor.

16 MR. STRINGER: Thank you, and good afternoon, Mr. President,

17 Your Honours, counsel, and everyone else in and around the courtroom.

18 As the Trial Chamber knows, we will be presenting three expert

19 reports prepared by Ms. Tabeau as well as some of her -- or together with

20 some of her colleagues in the Office of the Prosecutor Demographic Unit,

21 and the intent or the hope is that we'll be able to do that most

22 efficiently and hopefully in a way that's most clear for everyone by way

23 of a PowerPoint presentation that's been made by Ms. Tabeau and me working

24 together, and hopefully those have -- those slides have been distributed

25 to all counsel, Defence counsel, as well as the Judges and in the booths

1 to assist everyone's understanding of the testimony, so hopefully all of

2 you have those in front of you.

3 As well, we've loaded the PowerPoint presentation on our computer,

4 and I will ask that the technical people if they can put it on the video

5 screen so that I can follow along myself, and there it is.

6 Examination by Mr. Stringer:

7 Q. Ms. Tabeau, just starting with the first page of the presentation,

8 the slide, there's reference here to three reports that you'll be

9 addressing during the course of your testimony, and before we talk about

10 those I would like then to ask you just to give us a little bit of a

11 summary or a resume of your education, your educational background in the

12 field of demographics and then we'll talk a little bit more about some of

13 work that you've done in that field as well?

14 A. I have one question. Are we supposed to see the presentation

15 in -- on the screen as well or we don't?

16 Q. You should -- everyone should be seeing it. It may be --

17 A. I don't have it on the screen.

18 Q. Push the button there.

19 A. Okay. Could I have the presentation and the transcript at the

20 same time or is it possible or not? Sorry for this disruption. Okay.

21 Q. And for the benefit of all, at the top of this slide you'll see a

22 reference to the exhibit and the annex, and we'll be doing this throughout

23 the presentation, Your Honours. The slides will refer to specific parts

24 of the reports so that you can refer back to the reports if you wish.

25 A. Thank you very much.

1 MR. KOVACIC: [Interpretation] Your Honour, with your permission

2 I'd like to make an observation. I highly regard this approach by my

3 colleagues of the Prosecution trying with the PowerPoint tool to improve

4 the quality of the presentation and to assist those are listening, but we

5 cannot forget that the accused have the right to receive it in their own

6 language. I know that it will be simultaneously interpreted but it won't

7 be the same because a PowerPoint is in fact a precis, a summary of the

8 testimony, although it is a tool of course, but it is a summary of sorts,

9 and it would be a good idea if that were to be prepared in Croatian. I

10 don't know whether a Croatian version or copy exists or not.

11 JUDGE ANTONETTI: [Interpretation] Since this is a presentation and

12 here we have a page related to the education of the witness, we can see

13 that the witness studied in Warsaw. No need to understand English to --

14 to see what it's all about. And then the questions that will be put to

15 the witness will serve to understand what's on the screen.

16 Please proceed, Mr. Stringer.

17 MR. STRINGER: Thank you, Mr. President.

18 Q. Ms. Tabeau, could you give us a brief resume of your educational

19 and professional background in the field of demographics?

20 A. As I said I'm a demographer. I studied and graduated statistics

21 and econometrics. I come from Poland and studied in Warsaw at the Warsaw

22 School of Economics. In 1991 I obtained my Ph.D in mathematical

23 demography there. In 1981, I obtained my masters degree in econometrics

24 and statistics.

25 As part of my education I would also like to mention several

1 scientific internships that I completed. I spent some time in the 90s,

2 mid-90s is in the French National Demographic Institute. I also visited

3 the German National Demographic Institute, and most recently in July, I

4 had the opportunity to participate in a course given by epidemiologists on

5 public health in human emergency situation.

6 Q. Excuse me, Ms. Tabeau, if I could just ask you to slow down just a

7 little bit so that the interpreters can work more accurately.

8 A. Thank you. I'm sorry. At this course, I also had the opportunity

9 to lecture in relation to issues of measurement and analysis of conflict

10 deaths.

11 I now work at the Office of the Prosecutor. I am the project

12 leader of the Demographic Unit at the Office of the Prosecutor, and I have

13 been doing this since September 2000. Before I spent almost nine years in

14 the Dutch National Demographic Institute where I was conducting research

15 on mortality and public health issues, and longevity for countries such as

16 the Netherlands and other Western European countries, but also Central and

17 Eastern European countries.

18 Q. Okay.

19 A. Yes.

20 Q. Let me just jump in there and ask you --

21 JUDGE TRECHSEL: Excuse me for interrupting, but on the first page

22 the degrees the witness obtained are described as "with honours", I think,

23 and I would just like to know whether this would correspond to "cum laude"

24 in the traditional academic qualifications or whether it is a different

25 qualification.

1 THE WITNESS: At that time in the system we had in Poland "with

2 honours" had the meaning of obtaining the highest grade which at the time

3 was the 5. That was the best possible, and this is what I achieved in

4 both cases, but the cum laude as such was not explicitly used.

5 JUDGE TRECHSEL: Thank you very much.

6 MR. STRINGER:

7 Q. Now Ms. Tabeau, moving to the next slide that's on the screen

8 there are references to expert testimonies that you've given in other

9 cases here at the Tribunal as well as expert reports that have been

10 submitted in additional cases; is that correct?

11 A. Yes, it is correct. I already mentioned cases such as

12 Slobodan Milosevic, Dragomir Milosevic, General Galic, Stakic. I

13 testified twice in Stakic. I also testified in -- in the Simic et al.,

14 that is a case related to Bosanska Posavina. And in Lukic. There were

15 several more reports. I didn't have testimonies in relation to the other

16 reports, but the reports were used -- submitted, tendered into evidence

17 and used in the other cases as well, altogether more than 20 reports.

18 Q. And we will talk about this in some greater detail later, but in

19 general, can you tell us whether methods of analysis, sources of data that

20 you used to make the reports in this case are similar or correspond to

21 those that you've used in other cases in which you've appeared as an

22 expert?

23 A. The methods, sources and approaches that we've been using in all

24 these cases are similar. Worth mentioning is perhaps the approach we've

25 been using for the quantification of the numbers of internally displaced

1 persons and refugees. This approach has been used many years now, and it

2 was presented in several cases. We originally prepared these statistics

3 for use in the Slobodan Milosevic case where we presented an expert report

4 with the same type of statistics.

5 Q. Okay. Now, I've moved to the next slide which has a reference to

6 some of your more latest projects, and there are references here to

7 publications that have appeared in the scholarly journals within your

8 profession, and those publications relate in part at least to the work

9 that you've been doing here at the Tribunal, and I simply wanted to ask

10 you whether in your own field whether the work that you've been doing here

11 has been received favourably within the field of the demography that

12 you're engaged in.

13 A. These titles that are shown right now on the screen are a few

14 latest achievements. They are all related to the quantitative analysis of

15 demographic consequences of conflict, mainly these papers are related to

16 the casualties, to the killed persons in relation to conflict, as

17 consequence of conflict. I certainly believe that several of these

18 publications had a lot of impact in the field, but also on public opinion.

19 We, at some point, published an article together with Jakub Bijak in which

20 we discussed the number of war casualties in the conflict in Bosnia and

21 Herzegovina, and I believe this particular paper certainly had a lot of

22 impact in terms of both the method, the approach we applied, the sources

23 we applied, but also the result that is obtained and presented in this

24 paper.

25 Another publication is a book which I edited with Helge Brunborg

1 and Henrik Urdal. Helge Brunborg used to work at OTP demographic before I

2 worked here and Henrik Urdal also spent some time working on this project.

3 We know each other from the past and the book is a collection of

4 articles that were presented at a seminar organised by -- by a working

5 group of which we are all members, and this working group is called, right

6 now, the scientific panel of -- on demography of conflict. I believe it

7 is an important working group that is a leading group of researchers in

8 demography in this area.

9 Q. Okay. Now, these are the three reports now appearing on the

10 screen, the first of which is a report that you and your colleagues made

11 in respect of ethnic composition, internally displaced persons and

12 refugees from eight municipalities of Herceg-Bosna, and after we talk

13 about that, we'll move on then to additional reports or studies that you

14 did that related to siege of East Mostar, killed persons, and wounded

15 persons during the course of those months in 1993 in East Mostar.

16 I can skip over this slide because I think we've already touched

17 on this. The fact that at least with respect to -- well, in addition, I

18 should say, to the study, the work that's been done on wounded and killed

19 in a siege situation, the methods that you're going to be talking about in

20 respect of your demographic studies of refugees and IDPs from these

21 municipalities, those same methods and sources have been used by you in

22 these additional cases; is that correct?

23 A. Yes, this is correct. The results come from one project made for

24 the case of Slobodan Milosevic and statistics are all consistent

25 throughout all these cases. So are the sources.

1 Q. Now focusing on the first of the studies that relates to IDPs and

2 refugees, just keying off this slide Ms. Tabeau in general can you tell us

3 what was the subject or objectives in performing this specific study?

4 A. Yes. This is a study of population movements, of migration that

5 we should see as forced migration, both internal and external, and in the

6 context of the war in Bosnia and Herzegovina, it's important to note that

7 population migration is the major consequence of the conflict as the size

8 of this migration, during and at the end of the conflict, was enormous.

9 More than 50 per cent of the original census pre-conflict population was

10 displaced externally or internally. So measurement of migration as a

11 consequence of conflict is a very important issue. At the same time, it

12 is very hard to find sources that could be used for this kind of work.

13 Sources of migration especially from the conflict period are not possible

14 to obtain. There is no statistical data, individual data that would

15 include personal information about the migrants that could be used for

16 this kind of study. This is why a different type of approach has to be

17 applied to measure the migration.

18 We used the approach which, in order to quantify the numbers of

19 migrants in our approach, we compare the population at the outbreak of the

20 conflict with the population at the end of the conflict, and we compare

21 place of residence of this population whether it remained the same or

22 changed during this period.

23 Q. Okay. Now, on this slide there's a reference under the first

24 point to eight municipalities and could you tell us please just what was

25 the geographical, if you will, universe in which you were operating for

1 purposes of this particular study?

2 A. Yes. We used as a study area the area of eight municipalities in

3 Bosnia and Herzegovina. The municipalities are listed on the slide. It

4 was Capljina, Gornji Vakuf, Jablanica, Ljubuski, Mostar, Prozor, Stolac,

5 and Vares. It is not that these eight municipalities we considered to be

6 the Herceg-Bosna area. This is our study area which belongs to a larger

7 area called Herceg-Bosna, but for this particular report we called these

8 eight municipalities the Herceg-Bosna area, but these two issues shouldn't

9 be confused.

10 Q. Okay. And then we'll talk in greater detail about the sources.

11 There were three general sources; is that correct, for this study the

12 population census from 1991 and then OSCE voter registers from 1997, 1998,

13 and then thirdly registration of IDPs and refugees that was begun by UNHCR

14 and then continued by the government of Bosnia-Herzegovina?

15 A. Yes. These are the three sources that we used. It is very

16 obvious right now that the sources are dated not necessarily as the

17 indictment period is defined. We deal here with a broader period starting

18 in March 1991, which gives us a measurement of the population at the

19 outbreak of the conflict and at the other end we have the voters

20 registration from 1997, 1998, and one additional source that is dated as

21 of the year 2000.

22 Well, we have good reasons that we actually used these sources,

23 and -- well, first reason was that sources that could be covering exactly

24 the indictment period don't exist. There is nothing that could be used

25 for this kind of analysis as presented here. So the best we can get is on

1 one hand the 1991 census and on the other hand is the voters registration

2 1997, 1998, and this is what we use.

3 Q. Okay. And on that again we're going to discuss these in greater

4 detail but when you say these are the best you can get, given that that is

5 the best, are these sources and the data contained in these sources

6 nonetheless, did they provide you with a sufficient basis to do this study

7 under the rules and practices that would be acceptable within the field of

8 statistics and demographics?

9 A. Well, the answer is certainly yes, although I must here note that

10 sources that are used in human emergency situations, and conflict is

11 certainly part of this human emergencies, perhaps the most important part,

12 sources used in analysis of human emergency situations are not necessarily

13 of the same quality and the same standard as sources used in, for

14 instance, official statistics. There are very good reasons for this but

15 in this particular case we shouldn't be complaining at all, because the

16 population census is the best source of information about the population

17 in any country. It is a complete population survey covering many aspects

18 of the population and including a lot of individual details, including

19 names and date of birth and father's name and place of birth and place of

20 residence and ethnicity and religion and socio-economic variables and

21 marital status and the children ever born and professional area is

22 covered, et cetera, et cetera. So the census is an excellent source.

23 Well, the beginning for our approach is excellent.

24 On the other hand, we used voters registration. Voters

25 registration is not a source that would be used by professional

1 statisticians and official statistics. Nevertheless for our purposes for

2 comparison with the census, this source is fairly satisfactory and can be

3 used in the analysis as those presented in these reports. Not to mention

4 the importance and value high value of the last source, the registration,

5 official registration of internally displaced persons and refugees. This

6 is also a very good source developed by -- very professionally and up to a

7 high standard.

8 Q. Okay. Now there's a reference here to methods and I think I'm

9 going to skip that now because we're going to address it in greater

10 details as we get closer to your findings.

11 Just to give some attention to the geographical area that you've

12 described, we're looking at a map now and this will contain shaded

13 municipalities. Are these the shaded municipalities here are the

14 municipalities that form the basis of your study?

15 A. Yes. These are the eight municipalities.

16 Q. Okay. And at this point I think it's necessary for us to address

17 the issue of the boundaries of the municipalities and how they changed

18 from this time in 1991 to 1997 and 1998, which is sort of the back end of

19 your study after the Dayton agreement, and so I'm going to move from this

20 slide to the next one, and perhaps you could just describe for us what

21 we're seeing in this next slide here with these greater number of

22 municipalities.

23 A. Yes. First of all what is marked with colour on this map is

24 exactly the same what we have just seen on the previous map. What is

25 different is the number of municipalities shown, and in several cases the

1 municipal borders are different as well.

2 On the previous map we saw pre-conflict municipalities. There

3 were 109 of them in Bosnia. This is that map, yes. And on the next map

4 we see many more municipalities, and in addition to this a red line

5 splitting the country into two political entities. This part is -- is the

6 part of Republika Srpska located at the eastern border with Serbia and in

7 the north of Bosnia and Herzegovina and the rest of the country is the

8 Federation of Bosnia and Herzegovina.

9 Q. Okay. Let me --

10 A. And our study area is mainly located in the Federation of Bosnia

11 and Herzegovina.

12 Q. Okay, and on that point because it's one that affects the counting

13 that comes later, we went -- is it correct to say, then, that we're going

14 from eight pre-conflict or pre-Dayton municipalities to 14 post-Dayton

15 municipalities within what had been the original eight and that the only

16 two affected municipalities are Mostar, which was divided in, as part of

17 the Dayton agreement, into a number of smaller municipalities, and then

18 also Stolac which was divided into two municipalities, one of which

19 remained or is part of the federation and then the eastern half being --

20 falling within the Republika Srpska?

21 A. Well, it is basically the case only that the eight pre-war

22 municipalities became not 14 but 16 post-war municipalities. Fourteen of

23 them are in the federation area and two of them are in the area of

24 Republika Srpska. And the two is one part of Stolac, Stolac-Berkovici it

25 is called, and the other one which is composed of three small pieces is

1 the Mostar-Srpski Mostar. And the rest, the 14 municipalities to the left

2 of the Dayton line are in the federation area. So we will be basically

3 speaking -- I will be basically speaking about changes in the 14

4 municipalities.

5 Q. Okay. So we're going to be talking about eight municipalities,

6 pre-Dayton, that became 14 municipalities post-Dayton?

7 A. Yes, may I have one remark. So the major change introduced by the

8 Dayton peace agreement is that a number of municipalities were split

9 between the federation and the Republika Srpska and population movements

10 in the split municipalities were related to the fact that Bosnian Serbs

11 were moving from the federal area to the RS area and mainly Bosnian

12 Muslims were moving from the RS area to the area of the federation. There

13 were also some other changes. Some municipalities within the federation,

14 like Vlasenica for instance, were also divided in two and some other

15 municipalities, but right now these are less important.

16 Q. Okay. So that's the geographical area and how it changes or the

17 boundaries change pre-conflict and post conflict. Now you've already

18 discussed the census, and perhaps using this slide you could very

19 briefly - because I do want to keep going rather swiftly through this

20 part - tell us more about the census, what kind of data it provided, the

21 quality of the data that you found in the census?

22 A. Well, I said the census is the most complete population survey.

23 The principle of the census is to measure the population at a given moment

24 of time. It is for statistical authority probably the most important and

25 organisationally the most difficult and most challenging project. The

1 census is conducted once per 10 years in most countries. It's a

2 population census but in the case here of Bosnia and Herzegovina not only

3 population but also households, dwellings, and agriculture were part of

4 the census.

5 Census is conducted -- census is conducted by using a

6 questionnaire. A questionnaire is always prepared very carefully. It

7 is -- in the former Yugoslavia it was a standardised questionnaire that

8 was largely identical in all republics. The republics had the right to

9 add some questions. Not all republics used this opportunity, but the core

10 information was identical for every republic. The questionnaire included

11 questions as those I mentioned earlier related to personal information, to

12 place of birth, place of residence, et cetera. I don't want to repeat.

13 The questionnaire is extensive and gives a very detailed picture of every

14 individual registered in the census.

15 The information is collected or was collected, we are speaking of

16 1991, in face-to-face interviews. The group of interviewers employed in

17 the census was very large. These were several thousand of interviewers

18 who were going and talking to people and collected this information. Of

19 course the interviewers were prepared to do so because they were offered

20 intensive and extensive training on how to do that. There were also

21 guidelines printed in advance and distributed among interviewers and

22 instructors, municipal instructors and republican instructors in advance,

23 that data would be collected consistently, that the reporting would be

24 uniform across the country.

25 Well, the preparation of the census and the conduct of censuses is

1 a long talk and I don't want to go into details of that right now.

2 Q. And it's largely found in annex B of your report --

3 A. Yes.

4 Q. -- isn't it?

5 A. That's right.

6 Q. Okay, so the result being and let me then make a couple points and

7 ask you to comment on them, the result was that data was obtained as to

8 4.4 million individuals in the entire country, Bosnia-Herzegovina in 1991.

9 You make reference to some of the deficiencies in the data, in the

10 census, and there is a distinction between agriculture, dwelling and

11 population data that I'd like you to address, and then finally briefly if

12 you can just describe for us how you dealt with deficiencies that you

13 found in the data.

14 A. Yes. First my comment on the size. Yes, indeed, approximately

15 4.4 million records related to individuals residing in Bosnia as of 31st

16 of March, 1991, was collected in the census.

17 Regarding the deficiencies, indeed there are deficiencies, some

18 deficiencies, this the census. Here I want to say there is no survey and

19 no census anywhere else that is error free. This is impossible. Every

20 survey conducted, whether large or small, includes errors. So it is not

21 that the Bosnian census or census in Bosnia and Herzegovina is

22 exceptional. It has certain deficiencies but the deficiencies can be

23 dealt with in ways that prevent producing biased results.

24 Well, the major deficiency from the point of view of our work

25 comparing individual data from before the conflict with from after the

1 conflict, what is most important is the personal information about

2 individuals.

3 Personal information includes names on the first place. Names

4 contain spelling mistakes, and the spelling mistakes are related to the

5 fact that the census material wasn't computerised. By using data entry it

6 was scanned and therefore spelling mistakes related to the optical

7 character recognition method used in the scanning.

8 Q. Just to be clear on that that's when a computer programme tries to

9 read writing and then to assign an electronic or a digital character to

10 that handwriting; is that correct?

11 A. That's right.

12 Q. Okay.

13 A. Yeah. Well, this is one deficiency. There are also other

14 problems with the census. For instance, duplicates is an issue, although

15 duplicates were checked and eliminated by the statistical authority in

16 Bosnia and Herzegovina. Not all duplicates were eliminated. We

17 additionally checked whether we could identify more and we did. These

18 are -- these were not big numbers, marginal numbers.

19 Another deficiency is that the usual checking of the survey

20 material and census in particular is not only the checking of the

21 completeness of the census and elimination of the duplicates from the

22 census. One more step is the so-called logical control, that is,

23 relationship between different data items reported for one the same person

24 in the census questionnaire. The logical control in the 1991 census for

25 Bosnia and Herzegovina was completed for the population part of the census

1 but it wasn't completed for the agriculture and dwelling. So actually

2 from our point of view, this is not a big issue because what we use is the

3 population data.

4 Q. You didn't rely on the agriculture or dwelling data?

5 A. No. No. No.

6 Q. Okay. So the population data that you contained was completed or

7 was --

8 A. Yes.

9 Q. Now, again, very briefly how did you handle these deficiencies?

10 What did you do to overcome them?

11 A. Well, we invested a lot of time and effort in correcting spelling

12 mistakes in the names. For this purpose we -- we had run three different

13 projects. We, first of all, used computer programmes that checked the

14 spelling of the names and identified impossible combinations of letters,

15 impossible in the Bosnian language, and names with these combinations were

16 checked by native B/C/S speakers who corrected these impossible

17 combinations by replacing them with possible combinations.

18 Another method was that we compiled a table, a data table with

19 names, complete names, first names, and surnames, and native B/C/S

20 speakers were checking these names and proposing viable names for the

21 names that didn't exist in the B/C/S language.

22 And finally in the third project we used a household method to

23 correct the names. Well, household information is available in the census

24 so we were able to check within every household the names reported. And

25 if within households correct names were reported, then all other incorrect

1 names were replaced by the correct names.

2 Q. And just so we're clear because we've talked about how this data

3 has been used in other trials, what you're talking about here for the 1991

4 census was that done for just the eight municipalities in this report or

5 has it been done for the entire country?

6 A. No, it was done for the entire country, for the entire census, and

7 it took us several years to improve the names, to correct the spelling

8 mistakes.

9 Q. Okay.

10 A. There were several persons working on this project, B/C/S native

11 speakers, who were studying and correcting the results from our computer

12 programmes. There were people who were -- who wrote these programmes,

13 people who had the necessary IT background in order to make this kind of

14 work happen, yep.

15 Q. All right. Ms. Tabeau, I've moved to the next slide because this

16 is sort of the point that I'd like for you to address because you've

17 identified deficiencies in the data or the census and I'd like for you to

18 give us your overall impression of the quality or reliability of the data

19 for purposes of your study?

20 A. As I said there were some deficiencies in the census but many of

21 them have been improved, corrected. Not only spelling mistakes but also

22 some other data problems like inconsistent -- inconsistent date of birth

23 or some other issues of this kind.

24 I believe that data problems in the census do not discredit this

25 source as an excellent source of information about the pre-conflict

1 population in Bosnia and Herzegovina, and I don't think it is only my

2 opinion. Well, the statistical authorities in Bosnia and Herzegovina

3 published several bulletins and they also published this data in

4 statistical yearbooks. Authority, statistical authority of Croatia, also

5 published the census data for Bosnia and Herzegovina. So I believe the

6 fact of publishing these results is -- is obvious, right? The publication

7 would never take place if the quality of the source and deficiencies would

8 be unacceptable.

9 Q. All right. And for purposes of your report and your testimony in

10 this case, can the census be reliably used in producing the statistics and

11 the figures that you've provided in your report in this case?

12 A. I certainly believe it can be used in this kind of study.

13 Q. Okay. Now, the next slide is -- we're going to talk about the

14 census in relation to the eight municipalities of Herceg-Bosna that

15 you've -- you've described, and with this slide could you just tell us

16 what the numbers are. What are the census figures in respect of these

17 eight pre-Dayton municipalities?

18 A. Yes. We speak of -- of the population in eight municipalities,

19 eight -- in the eight Herceg-Bosna municipalities in 1991 being 281.366

20 persons. This is the entire population in these municipalities including

21 the children, as well as those at older ages.

22 Q. Okay. Now let me stop you there. That's the population of all

23 persons who were living in these municipalities at the time the census was

24 taken in 1991; correct?

25 A. That's right.

1 Q. Okay. Now the next point has a smaller number of those

2 individuals as people who were born before 1980, and I'm going to ask you

3 about this a little more later, but is that -- does that have to do with,

4 then, the voting age of the people who were actually people who were old

5 enough to vote in 1997, 1998, when you were then examining the other data,

6 the OSCE voter registration?

7 A. Yes. The age those born before 1980 of course has to do with the

8 age of being eligible to vote. Well, for the 1997-8 we could only use the

9 voters registers. That means only a group of the population. So in order

10 to compare two things that are comparable we had to adjust the census

11 population, and from the census we had to study only those born before a

12 certain year, 1980 in this case.

13 Q. Okay. So then to the third point. This 231.610 figure, this is

14 sort of the starting point of your -- of your study group for purposes of

15 your report then?

16 A. Yes. This is the study population which is coming from these

17 eight municipalities shown right now on the map. This is the population

18 born before 1980.

19 Q. Okay. Now, let's talk about the second source in a little more

20 detail, the voters registers from OSCE. This provides us with the ending

21 point or the sort of the --

22 JUDGE ANTONETTI: [Interpretation] Just a moment, please. I am

23 trying to follow the answers given by the witness, but I would like to

24 make sure that I understand something.

25 Madam, when you talk, we are under the impression that the entire

1 census that was carried out that you were the one carrying out the census.

2 I am slightly confused, because in your sources you talk about the 1991

3 census, the statistics of the former Yugoslavia, and they had taken 4.4

4 million people in order to carry out the census.

5 The charts that are in your document, these charts are your own

6 charts, aren't they, or are those charts that belong to the official

7 statistics office for the year of 1991.

8 THE WITNESS: Yes. This is a very good question. Well, I want to

9 stress that the 1991 census was conducted by the statistical authority of

10 the former Socialist Republic of Bosnia and Herzegovina. So this is a

11 source that comes from Bosnia and Herzegovina, was compiled there by the

12 statistical office in Sarajevo.

13 We do have a copy of the census data. We not only have aggregate

14 statistics from the census as those published by the statistical

15 authorities at several occasions. We do have individual census records

16 for every individual reported in the census. So we are able to use this

17 individual data to compile our own statistics.

18 When having individual data, we are able to make a selection or

19 take a group from the census and analyse this group. This is what we did

20 in this project. We took a group of the census records. The group was

21 the individuals born before 1980, and we studied this group and compared

22 this group with information that we had about the same people from other

23 sources.

24 So in fact we can summarise this as follows: We do use an

25 official source, that is the census, but because of having access to

1 individual census records we are in the position to compile any statistics

2 that we need for this or other projects. So in my reports -- reports,

3 charts, maps, data tables, even though the source is mentioned, the 1991

4 population census for Bosnia and Herzegovina, these statistics were

5 compiled in my unit, in the demographic unit here at the OTP.

6 JUDGE ANTONETTI: [Interpretation] Very well. Thank you very much.

7 I understand better. Thank you. You may proceed.

8 MR. STRINGER: Thank you, Mr. President.

9 Q. Just to clarify or to follow up on the president's question, you

10 were not -- you were not personally involved in the gathering of the

11 census data in 1991, were you?

12 A. No, I was not.

13 Q. And you received all of the census data from the authorities of

14 Bosnia-Herzegovina?

15 A. Yes, that's right.

16 Q. And then -- and then you've given us the number of 4.4 million

17 people. That's roughly or approximately the population of

18 Bosnia-Herzegovina as of 1991 according to the census?

19 A. Yes, that's right.

20 Q. Okay.

21 A. As of 31st of March, 1991.

22 Q. Okay.

23 A. Approximately 4.4.

24 Q. And then bringing that number of 4.4 million down to the eight

25 municipalities that are the subject of this report, how many people of

1 those -- from 4.4 million we're going to come down to a number that's

2 going to be the beginning point of your survey or your report, and what's

3 that number?

4 A. For me, it is the number 231.610.

5 Q. Okay.

6 A. This is the study population from the population census for Bosnia

7 and Herzegovina.

8 Q. And again, that's the population of -- your study population from

9 these eight municipalities?

10 A. Moreover it is not the entire population; only those born before

11 1980.

12 Q. That's right. And the reason you've excluded people born before

13 1980 relates then to this next source; isn't that right?

14 A. That's right. It is related to the voters registers from the 1997

15 municipal election and 1998 parliamentary elections in Bosnia and

16 Herzegovina.

17 Q. So what did you use this source for?

18 A. This source is used in our project to compare with the population

19 census. We used this source in order to study whether or not place of

20 residence of individuals listed in the '91 census changed when observed in

21 the period of '97, '98.

22 Q. Okay. Let me just stop you right there because I just want to --

23 so we're clear. This is a source that you used to study whether or not a

24 person's place of residence changed from 1991 to this time, 1997, 1998?

25 A. That's right.

1 Q. Okay. And this OSCE voters registration, did this cover people

2 who were living in Bosnia-Herzegovina as well as people who were living in

3 neighbouring countries in 1997, 1998?

4 A. It is a register of all voters eligible to vote. It covers people

5 in Bosnia and Herzegovina and people from Bosnia and Herzegovina but

6 living at the time of elections in other countries, in neighbouring

7 countries, in the area of the former Yugoslavia, but also in other

8 countries, in Western European countries and even other countries. So it

9 is the basis for the establishing of the voters registers, especially the

10 register of 1997, was actually the population reported in the 1991

11 population census. The lists of voters eligible to vote were made based

12 on the names reported in the census.

13 Q. Okay. So --

14 A. And even technically speaking, electronic files from the census

15 were used for the voters registration in 1997-8.

16 Q. So OSCE actually used the 1991 census data to facilitate making

17 these voter lists in 1997 and 1998?

18 A. Yes, that's right.

19 Q. Okay.

20 A. Of course, it was possible to add new names if a person was not

21 reported in the census, and there were a few cases like that.

22 Q. Okay.

23 A. But, basically, it is the 1991 census population later used in the

24 electoral lists.

25 Q. Okay. So now on this slide you make reference to these two lists

1 which were then merged into a single list which you used, and I'm going to

2 move to the next slide then and just ask you to talk about how then you

3 used this merged list of voters registration to come down to a number that

4 you used again to track the people who had moved or who had stayed in

5 those eight municipalities from 1991.

6 A. The voters registration -- the combined voters registration we

7 combined these voters registers because it was important for us to include

8 in our post-conflict sources as many names of survivors as possible. We

9 merged these two sources by taking the 1997 register as the primary source

10 and expanded this source by adding records of new voters registered in

11 1998. So it is basically the population of 1997 that is reported in the

12 merged sources here in the merged voters registers. Well, the voters

13 registers do not contain all the information from the census. They only

14 contain personal information of individuals, but the personal information,

15 this is what we actually need.

16 Q. Right.

17 A. As this together with the municipality where the voters register

18 gives us the information that is needed to study in this project.

19 Q. Okay. I'm going to move to the next slide and ask you then to

20 just bring it then down to the level of these eight municipalities, the

21 information that turns out to be the most useful to you from the voters

22 registration, how that relates then to the 1991 census.

23 A. Well, the first observation is that the size of the merged

24 1997-'98 voters register is approximately 2.700.000 individuals. This is

25 of course much less than 4.4 million individuals registered in the census.

1 But first of all this is because of the age groups covered by the voters

2 register. Voters register is just a smaller group. But secondly, not all

3 individuals eligible to vote registered to vote and participated in the

4 elections. So there is a group that is not in the register at all. So

5 this is another reason that we don't see it here.

6 So, for 1997-8, we don't speak of a complete population of record

7 of -- of the complete population of Bosnia and Herzegovina. We have a

8 huge sample of this population registered.

9 For the eight municipalities, this means quite some -- quite

10 different numbers. Registered voters who registered in the eight

11 Herceg-Bosna municipalities accounted to 145.061 voters. This is the

12 first statistic. This is what we have for the eight Herceg-Bosna

13 municipalities.

14 Now, when we want to show the voters who registered in 1997 and 8,

15 and at the same time were reported in the population census in the same

16 eight municipalities, this group is slightly smaller. It is 142.204

17 voters.

18 And this is a very important group, very important group. This is

19 the group that will be -- is studied in our report. Once again I would

20 like to repeat. These 142.000, approximately, voters are individuals who

21 registered to vote in the 1997-1998 elections, but at the same time they

22 were reported in the eight Herceg-Bosna municipalities in the 1991 census.

23 Q. Okay. So that what we've got now are 142.204 people, and you know

24 where each of those persons are in 1997, 1998, and you also know where

25 each of those people were in 1991; is that correct?

1 A. Yes, it is.

2 Q. Okay. So then is your study then looking at the differences in

3 the composition of the people, where they lived in 1991 versus where they

4 were found or where they were living in 1997, 1998?

5 A. Yes. This is actually the basic principle of our approach. We

6 have a starting population, pre-conflict population that is population

7 census for eight municipalities.

8 Q. Okay.

9 A. And we traced this population in other sources and of the

10 approximately 231.000 in 1991 we were able to find in the voters register

11 142.000, approximately, individuals. These are same individuals in both

12 sources. We have the names of these individuals, other details of these

13 individuals. We have place of residence of these individuals in 1991 and

14 also we know for these individuals where they registered to vote whether

15 in their home municipalities or in different places.

16 Q. Okay, so that you know if someone who lived in Jablanica, for

17 instance, someone living in Jablanica in 1991, for these 142.000 or so,

18 you're able to tell where that person was living when he registered to

19 vote in 1997 whether he was in Jablanica or in Stolac or in Germany?

20 A. Yes. Yes, we are able to say this.

21 Q. Okay. Now, we'll talk more about that matching process, but let's

22 move to the third source because, and I don't want to spend too much time

23 on this, because I know in the next slide it's an indication that this is

24 a sort of subsidiary source, but if you could describe this displaced

25 persons and refugees database and what use you made of it?

1 A. As you said, it is a subsidiary resource. We looked at it for

2 contextual purposes once we produced our statistics on displaced persons

3 and refugees. In this case, displaced persons in particular.

4 We wanted to check whether sources like the registration, official

5 registration, of internal displacement showed similar statistics or

6 similar geographic patterns. So our purpose was just contextual. We

7 didn't use this sort for producing any statistics in our report.

8 It is an important source, though, I must admit. It reports

9 internal displacement and refugees in Bosnia and Herzegovina. It contains

10 people -- records of people who acquired the legal status of an

11 internal -- internally displaced person or refugee. These records are

12 important in Bosnia and Herzegovina, because they offer a basis for

13 applications for compensations for the property lost. So this is a very

14 good source, basically speaking, only that it is reporting the data as of

15 the year 2000, which is even more distant from the conflict period than

16 what we used in our study.

17 Q. Okay. Now, I moved to the next slide in which you indicate that

18 this was a subsidiary source, it served as a reference for the OSCE

19 statistics only on IDPs. I think with that actually we can move to the

20 next slide.

21 You made reference to this marching of names from 1997, 1998 to

22 the names from the census data from 1991. Tell us about how the extent of

23 matching, the level of matching that you were able to accomplish as

24 between these two sources.

25 A. Perhaps it's important to explain a little bit what the matching

1 stands for. Matching and what I earlier said, tracing records of a group

2 of individuals in other sources is actually one the same thing. The idea

3 of matching is checking whether or not a given record from a given source

4 is also reported in another source. The method of matching is a

5 well-recognised standard in social sciences, and especially in

6 Scandinavian countries this method has been used for many years with a

7 great degree of success.

8 In the developed countries like the Scandinavian countries,

9 matching is done easily on the basis of individual identification numbers

10 reported for persons in sources through comparing these numbers. These

11 are just numeric characteristics, it's easy to conclude records the same,

12 into different sources or in more sources.

13 Matching does not necessarily have to be conducted exclusively on

14 the basis of numeric characteristics of individuals reported in the

15 records. Matching can be done also on the basis of non-numeric

16 characteristics like names, other personal information reported, including

17 date of birth, place of birth, place of residence, et cetera, et cetera.

18 Q. Could I just ask you there the matching process or the method that

19 you used in this study, is it a process or analysis that is accepted and

20 used more generally or would it be an accepted practice in the field of

21 statistics and demography?

22 A. It certainly is but it is much more than statistics and social

23 sciences or demography only. Matching has been used -- used by

24 institutions like life insurances, health insurances, tax offices.

25 Statistical authorities of countries use matching to increase the quantity

1 of information about individuals. Well, it's important to understand that

2 in many countries we don't have the population census as such at all. All

3 the information about the population comes, first of all, from population

4 register, which is matched or other term that is often used, linked with

5 additional sources of information.

6 Depending on the needs in developed countries, we can produce any

7 statistics on the population, and it is only a beginning to have the

8 population register. Having the register and individual identification

9 numbers, it is a common practice nowadays to increase the information

10 about individuals in any required way.

11 Q. Now, one the point here in the middle of the page, you indicate

12 that 80 per cent of those in the voter register were matched with

13 individuals found in the 1991 census. Is that a high or low matching

14 level for your needs?

15 MR. KARNAVAS: Well, as compared to what, Your Honour? As

16 compared to what, this 80 per cent? I think the question needs to be

17 rephrased. And I would go back also to the previous -- answer to the

18 previous question. Matching methods that she used, the particular ones,

19 are these the ones accepted within the scientific community, not in

20 general whether matching is, but when -- when -- the question -- the

21 question 80 per cent for her needs, what does that mean?

22 MR. STRINGER: I can clarify that question, Mr. President.

23 JUDGE ANTONETTI: [Interpretation] Very well. Those are usually

24 questions that should be put during the cross-examination, but, yes, you

25 may ask for some precision, Mr. Stringer.

1 MR. STRINGER:

2 Q. Ms. Tabeau, you heard --

3 A. Yes.

4 Q. -- counsel's comment. Would you like to address it?

5 A. Yes. First of all, it is 80 per cent of records in the voters

6 register that have been matched with the census. Well, whether it is a

7 big matching -- high matching rate or not, ideally, we would like to see

8 100 per cent matching rate. This is, of course, the purpose of any

9 matching. Well, whether it is possible to achieve 100 per cent of matches

10 in a certain source, I believe, yes, it is. If there is a good system of

11 individual, unique identification numbers, then, yes, it is. Of course

12 there will be always a small group of records that are deficient in some

13 way, just human errors or something like that that cannot be matched, but

14 we can achieve a matching rate in principle close to 100 per cent.

15 Q. Okay, let me --

16 A. In excellent sources. In excellent sources, not necessarily in

17 sources used in human emergency situations.

18 Q. Okay. Just so this is clear, are you saying that 80 per cent of

19 the names found on the combined voter register were individually matched

20 to a name appearing in the 1991 census?

21 A. 80 per cent of the names found on the combined voters register

22 were matched with the census.

23 Q. Okay. And that number of matched names, then, is that the 142.204

24 people you've described?

25 A. That's right.

1 Q. Okay.

2 A. Well, regarding the method of matching, I was just about to speak

3 about it. In the former Yugoslavia there was a system of individual

4 identification numbers available. These -- the so-called jedinski

5 [phoen], well, individual identification JMBG numbers were introduced

6 around the year 1980 in the former Yugoslavia, and for all those born

7 after this year, 1980, were received -- were issued at birth. For the

8 older population these numbers were issued subsequently at different

9 occasions but there was a deadline, at which, I think in the second half

10 of the'80s, at which the population was to acquire these numbers. In the

11 population census, the JMBG number is available. It is not 100 per cent

12 correct and not 100 per cent available for all individuals. But the JMBG

13 was the basis for our matching of the census with the voters

14 registration. Basis, but we also used other criteria, additional

15 criteria.

16 JUDGE TRECHSEL: I would like to come back to question by

17 Mr. Springer which has prompted Mr. Karnavas to his feet.

18 80 per cent. We now know what 100 per cent would be, but if you

19 compare your study with comparable studies, is it something to be

20 expected, is it better than expected, is it lower than expected. To what

21 extent does it affect the reliability of the results?

22 THE WITNESS: It is a very good matching rate. We have been

23 matching sources since 1998 in this unit, and we've been matching all

24 kinds of sources, always with the population census, and the 80 per cent

25 matching rate is a very good matching rate. Only in a few more cases,

1 perhaps one, two, we obtained a matching rate higher than 80 per cent.

2 JUDGE TRECHSEL: Thank you.

3 MR. KARNAVAS: Well, Your Honour, your question wasn't answered.

4 I mean, compared to what? With her studies? But the way I read your

5 question was on outside matters, not just within the OTP. 80 per cent may

6 suffice for her, because she's a member of the OTP, paid by the OTP,

7 employed by the OTP but in comparison to outside is that an acceptable

8 standard of 20 per cent variance, plus or minus. I think -- that's what I

9 thought you wanted the answer to and I don't think she answered that

10 question.

11 JUDGE ANTONETTI: [Interpretation] Let me supplement the question.

12 Witness, you stated that you made a comparison to the level of 80 per cent

13 of the registers vote. This total of 80 per cent leaves 20 per cent we

14 don't know anything about. Is that approach, is that work method

15 compatible with a standard applied the world over in your field, and

16 furthermore, when you have an 80 per cent matching rate, can you deem the

17 result to be reliable in terms of international standards? And I'm not

18 talking about the standards applied at the OTP. I believe that you're

19 perfectly capable of answering this question, which is quite

20 straightforward.

21 THE WITNESS: Yes, sir. Well, thank you for the question. I must

22 say I am perhaps alone here at the OTP in doing this work. I have myself

23 at my unit, of course, for these kind of issues, but I'm lucky to have

24 been in touch with several other experts working in the field, including

25 great people like Helge Brunborg, Patrick Ball, who are also experts and

1 presented their reports here. I also worked with other experts, recently

2 with Mr. Philip Verwimp. We had a joint project which I presented earlier

3 today in which we made an assessment of a very exceptional source on

4 casualties of the Bosnian war.

5 I don't think that there is a problem here with this matching rate

6 80 per cent. I certainly can tell you that this is a generally very good

7 matching rate for this kind of work and for these sources.

8 Of course, there are methods to increase the matching rate, but

9 we've been conservative with matching sources because it's better to

10 undermatch sources than to overmatch.

11 JUDGE ANTONETTI: [Interpretation] You're not answering my

12 question. What I would like to know is the following: Is this 80 per

13 cent rate compatible with methods used by other experts? You've named a

14 number of experts. Is it compatible with the methods used in other

15 countries where the same methodology is applied and where it is thought

16 that a rate of 80 per cent, a figure of 80 per cent is a reliable figure?

17 THE WITNESS: Well, we speak of research into consequences of

18 conflict, and I don't think it is -- that it is -- we are speaking of a

19 situation that is not a huge research field in the sense that I could

20 present you here with 10 comparable studies to the study that I'm

21 discussing here. This work is exceptional, the work I'm presenting here.

22 It's exceptional and people follow our example, other people working in

23 the field.

24 So what I'm trying to say, the most of the existing experience

25 comes from this office, and I can compare the results in this project I

1 obtained in very many other projects, and at the same time I know this

2 method has been used by several other experts I know personally, with whom

3 I even worked, and they use this approach the way we do it, and we are all

4 aware of problems when applying this approach, and we apply similar

5 solutions to these problems. And as I said, it's possible to increase the

6 matching rate by using statistical techniques, but then we run the risk

7 that wrong records are mapped. I believe that there is a limit when a

8 match can be declared a true match and this limit cannot be crossed

9 because the risk of making an error when declaring two wrong matches a

10 good match is too high. It's better not to increase the matching rate by

11 statistical method and computer algorithms but better stick to results

12 that are very reliable and at the same time offer a very good matching

13 rate, 80 per cent.

14 MR. KARNAVAS: Again, Mr. President, the employee of the Office of

15 the Prosecution has refused to answer the question concretely and

16 directly.

17 MR. STRINGER: Well, Mr. President, I --

18 JUDGE ANTONETTI: [Interpretation] If I understand correctly,

19 Witness, to my question that was extremely specific you responded that the

20 work you conducted is rather -- is unique, because the consequences of a

21 conflict on a given population is a unique type of work that to your

22 knowledge only the OTP was able to conduct in relation to the former

23 Yugoslavia. As a result of this, you are not in a position to compare

24 your work could work accomplished on other conflicts because according to

25 you the work you have done is unique.

1 Is it really what you're telling us?

2 THE WITNESS: Well, I think what is really unique is the scale of

3 the analysis. I don't think that I could give you examples, other

4 examples of other research groups who would conduct the matching of the

5 census data with another large source like the voters register. I cannot

6 give you these kind of examples. So the scale is absolutely unique. But

7 not the method as such. I can give more examples of the matching done

8 between sources of smaller scale.

9 JUDGE ANTONETTI: [Interpretation] We need to take a break now.

10 We'll resume at 10 past 4.00.

11 --- Recess taken at 3.53 p.m.

12 [The witness stood down]

13 --- On resuming at 4.14 p.m.

14 JUDGE ANTONETTI: [Interpretation] Fine. The hearing is resumed.

15 Before we bring the witness in, let me read a short oral decision

16 related to the admissibility of exhibits related to witness who testified

17 on the 15th of March, 2007. The Trial Chamber has decided to admit

18 Exhibit P 09882 under seal, as well as P 02732. These exhibits had been

19 submitted by the Prosecution through list IC 00492. They have been

20 admitted since they have a certain probative value and certain relevance.

21 The Trial Chamber most that the Defence has not requested the

22 admission of any exhibits.

23 Let's have the witness brought in.

24 MR. MURPHY: Your Honour, while the witness is being brought in,

25 there's no objection to her hearing any part of this, I would like to make

1 an oral application in the interests of saving time that the Trial Chamber

2 dispense with any further evidence from this witness. There is a basic

3 requirement in the case of an expert witness that there should be at least

4 some degree of scientific comparison and validity to the scientific

5 methodology which is being used --

6 MR. STRINGER: Excuse me, counsel. Mr. President, do believe that

7 this is an argument that should take place out of the hearing of the

8 witness. Sorry for the interruption.

9 MR. MURPHY: I'm in the Chamber's hands. I have no objection to

10 the witness hearing it.

11 JUDGE ANTONETTI: [Interpretation] Mr. Stringer will proceed with

12 the examination-in-chief, and he will deal again with the methods used for

13 comparison purposes. That will shed some light on the matter.

14 Having said that, Mr. Murphy, this witness has already testified

15 many times here in this Tribunal.

16 Mr. Stringer, you have the floor.

17 MR. STRINGER: Thank you, Mr. President.

18 Q. Witness, before the break you were getting questions from me and

19 then subsequently counsel and then -- and then the Trial Chamber itself

20 about the 80 per cent matching issue. What I want to do is to take it one

21 step beyond that and maybe that will shed light on the 80 per cent issue.

22 And we're looking at the same slide again as we were before, and I'm

23 looking at the second-to-last point in which you say as a -- the total

24 number of linked or matched individuals for the eight municipalities came

25 to 142.204. You see that?

1 A. Yes.

2 Q. Now, my question is whether that number of individuals, that

3 number of cases, 142.204, whether that provided you with a large enough

4 sample --

5 A. Yes, of course.

6 Q. -- in which to conduct your studies and particularly whether it's

7 a large enough sample within the standards that would prevail in your

8 profession.

9 A. Certainly it is. It is a sample which is approximately 46 per

10 cent of the census population from these eight municipalities, and it's a

11 large sample and as such it can be safely used in the analysis.

12 I want to comment on this 80 per cent as well if I may. Well, the

13 80 per cent should be seen in the broader perspective of our work, of our

14 project. It wasn't our goal to achieve complete numbers of displaced

15 persons and refugees because that is impossible. Through matching, what

16 could be achieved and has been achieved in this project is the minimum

17 number of the displaced persons and refugees. So if the matching rate

18 could have been 100 per cent, our statistics on the displacement could

19 have been better, higher numbers. They are based on 80 per cent matching

20 rate and they are minimum numbers, but they are coming from a very good

21 sample. In addition to them, to the minimum numbers, we produce more

22 complete numbers based on the statistical estimation, and I think we will

23 discuss these numbers later today.

24 JUDGE ANTONETTI: [Interpretation] Witness, I want things to be

25 very clear. There's no problem as far as I'm concerned, but apparently

1 it's not the case for everyone, but at paragraph 5 on the screen you're

2 saying that the vast majority of individuals included in the voters

3 register were matched with individuals found in the 1991 population

4 census.

5 When I read that sentence on the screen in English, it leads me to

6 think that you and your associates went through the voters registers and

7 you took note of the fact that amongst the people who had voted 80 per

8 cent were to be found in the 1991 population census, and the 20 per cent

9 difference might be either new voters or people who had not been

10 registered as part of the 1991 population census.

11 Is that the way you assess this 80 per cent figure?

12 THE WITNESS: The 80 per cent the matching rate which relates to

13 the registered voters. 80 per cent of the registered voters have been

14 matched with the census. The registered voters is not a complete

15 population and it has nothing to do with the 20 per cent unmatched

16 records. Those voters who are not in the register simply are not there

17 because they didn't register, and there is another group which is not in

18 the voters register. This is population younger than 18 years. So it is

19 per se a sample, right, of the 1997, 1998 population, the voters register.

20 JUDGE ANTONETTI: [Interpretation] Mr. Murphy, have you understood?

21 Is it clear now?

22 MR. MURPHY: Your Honour, I think -- I think that is clear. At

23 least we can clear it up during cross-examination. I'm trying to save

24 time. The other point I wanted to make, if I may, is perhaps Mr.

25 Stringer -- Your Honour, may I just add this: That Mr. Stringer could

1 assist also explaining the relevance of this evidence given that the

2 comparison runs from a date in 1991 a date before anything significant

3 happened in our case to a point between three and four years after the

4 indictment period.

5 MR. STRINGER: We will be addressing all of that, Mr. President.

6 If I could just ask perhaps to -- to go back into the presentation and

7 then if -- if points like that haven't been addressed to everyone's

8 satisfaction --

9 JUDGE ANTONETTI: [Interpretation] Yes, please proceed.

10 MR. STRINGER:

11 Q. Ms. Tabeau, in order to move things on a little more quickly

12 because I want to get -- we haven't really gotten into the numbers yet and

13 I think that's obviously of greatest interest to everyone, so I'm going to

14 sort of move quickly through the next couple of slides. The information

15 in them is already found in your reports which everyone has.

16 With those 142.000 individuals now that you've got both a

17 beginning point and a ending point in terms of their place of residence,

18 you're then going to assign an ethnicity to those individuals, and on this

19 slide you indicate that essentially the ethnic grouping that you applied

20 is the one that that individual -- each of those 142.000 individuals

21 reported for themselves in the 1991 census; is that correct?

22 A. Yes, it is correct. They reported this in an open-ended question.

23 They had no restrictions, no lists to -- to choose from. They reported

24 what they thought they were at the time of the census.

25 Q. Okay. So you didn't change that.

1 A. No.

2 Q. All right. And then that resulted in four principal groupings

3 that you used in your report.

4 A. Yes. We -- actually, the reported ethnicities are much more than

5 just three major groups, Bosnian Croats, Bosnian Muslims, and Bosnian

6 Serbs. There are also other groups reported including Yugoslavs and mixed

7 ethnicities. Well, we used the original reporting to regroup the

8 individuals. We separated those who reported themselves explicitly as

9 Muslims, Croats and Serbs and took them as reported, and the rest was

10 assigned a new name, "Others."

11 JUDGE PRANDLER: Mr. Stringer.

12 MR. STRINGER: Yes, Your Honour.

13 JUDGE PRANDLER: I'm sorry to interrupt you. I would like to ask

14 only one question from the witness about the second paragraph in what we

15 have now before us as far as the methods are concerned, and in paragraph

16 two it says that "in the original 1991 census forms, the citizens of

17 Bosnia and Herzegovina mentioned several hundreds of ethnic categories.

18 We regrouped these categories into four major clusters," which is clear.

19 So for me frankly, it's a bit strange when you mention here several

20 hundreds of the ethnic categories. To my knowledge, both in Yugoslavia

21 and in neighbouring countries, I do not believe that we have several

22 hundreds of ethnic categories either in Yugoslavia or Hungary or in

23 Austria or Bulgaria, et cetera. So I wonder if you may clarify this issue

24 Ms. Tabeau.

25 THE WITNESS: Yes, thank you very much. I actually paid

1 particular attention to this particular formulation this morning. I was

2 checking this in the census 1991 and exactly 91 ethnic categories were

3 reported. So, sorry for the several hundred, but 91 is certainly more

4 than just three plus one major groups.

5 Well, why that many? Why 91? Categories like Poles, French,

6 German, Norwegian, Australian, whatever, and small regional groups also

7 reported themselves separately are there in the original report. Mixed

8 marriages, Serbo-Croat, for instance, is there as well in several

9 combinations. So that's the reason why.

10 JUDGE PRANDLER: Thank you.

11 JUDGE ANTONETTI: [Interpretation] Let me follow up this question

12 and my question might help us to save time.

13 As for the categories, in the tables you've prepared we see the

14 following categories: Croats, Muslims, Serbs, and at page 25, we see

15 Yugoslavs, others and unknown. So in actual fact we have five categories.

16 Further on we find tables where you've included the Yugoslavs in --

17 under "Unknown" and "Others," but at page 25 of your report on Mostar, we

18 see that almost 10 per cent of the population has registered -- have

19 registered themselves as being Yugoslavs, whereas in the other

20 municipalities the rates are much lower. In Mostar, we have something

21 rather unusual because these people register themselves as Yugoslavs and

22 not as Croats, Serbs, or Muslims.

23 In the other tables did you not take that into account or did you

24 believe that it had no impact at all, this particular fact?

25 THE WITNESS: If I may have a question about which report,

1 Your Honour, we're talking. I -- the first report? I -- yes.

2 JUDGE ANTONETTI: [Interpretation] Yes, the first report. Yes.

3 It's the first report. Page 25 of the first report. We find a table --

4 THE WITNESS: Yes, I see it. I see this table, and this table is

5 taken from publication. The source for this table is mentioned under the

6 table. So it doesn't come from our work.

7 In our project, we didn't separate the category "Yugoslavs", and

8 we didn't look at this category separately. This category is included in

9 the "Others" in all analysis we presented in our report. And we did it

10 the same in the first report as also in the other reports, the reports on

11 the killed and wounded persons. We looked at three major groups, Bosnian

12 Croats, Bosnian Muslims, Bosnian Serbs, and the rest was just taken

13 together. But this table is for reference purposes taken as originally

14 published. This is why.

15 JUDGE ANTONETTI: [Interpretation] Fine. At least that's clear

16 now. Thank you very much.

17 You may proceed.

18 MR. STRINGER: Thank you, Mr. President.

19 Q. So that was assigning ethnicity for purposes of your report. And

20 the next question, again I'm hoping to be brief, is just for you to tell

21 us how then you determined or classified an individual within this 142.000

22 group as an IDP, that is, an internally displaced person, or a refugee.

23 A. We used statistical definition of a displaced person or a refugee,

24 which is not necessarily -- which is not the same as legal definition used

25 by the UNHCR. In our definition, a person was declared a displaced person

1 or a refugee if the place of residence in 1991 was different than the

2 place of registration to vote in 1997-8. The place of registration to

3 vote in 1997 and '8 was used as a proxy for the place of residence in this

4 period. All those individuals who reported different place of residence

5 in 1991 versus 1997-8, different places of residence in these two periods

6 of time, were declared as -- well, declared by us in this study as

7 displaced persons or refugees. Displaced person is a person who stayed in

8 Bosnia but moved to a different municipality. A refugee is no more in

9 Bosnia by 1997-8, who left the country and was reported as a voter in a

10 different country.

11 Q. Okay. Now, you have also --

12 JUDGE TRECHSEL: I'm sorry. Would a person who was born in

13 another country but then registered in 1997, 1998 in Bosnia, would that

14 also be a refugee, or did you not include such persons?

15 THE WITNESS: Well, the basis for this comparison is the

16 population census of 1991. The population census of 1991. So we compared

17 the places only for the persons enumerated in the census. So if the

18 person was not reported in the census in 1991 but still was reported in

19 the voters register, it is not in our study. This person is not in our

20 study, cannot be in our study because we can't match the person with the

21 census.

22 JUDGE TRECHSEL: So it's a one-way definition in a way.

23 THE WITNESS: Yes.

24 JUDGE TRECHSEL: And persons who do not register but stay in the

25 country would nevertheless have figured as refugees in your books.

1 THE WITNESS: If they would never register but stay at the same

2 place of residence would be not shown in the minimum numbers.

3 JUDGE TRECHSEL: They would be regarded as refugees.

4 THE WITNESS: No, they wouldn't be shown in my statistics.

5 JUDGE TRECHSEL: How do you how did you make a difference between

6 these and the refugees, because you will not have figures of refugees

7 properly, will you?

8 THE WITNESS: Perhaps I misunderstood the question. If you could

9 perhaps repeat the question, please.

10 JUDGE TRECHSEL: The starting point is you define refugees as

11 persons who in 1991 lived somewhere in Bosnia-Herzegovina, and in 1997,

12 1998 had the residence abroad.

13 THE WITNESS: That's right.

14 JUDGE TRECHSEL: Now, you had, of course, no means, I suppose, but

15 perhaps I'm wrong, to trace persons that in 1997, 1998 were residing --

16 were --

17 THE INTERPRETER: Microphone, please.

18 JUDGE TRECHSEL: Somebody -- were residing outside of Bosnia and

19 Herzegovina, which then would lead me to the conclusion that you cannot

20 distinguish between people who have stayed in the country but for some

21 reason or other did not choose to register and persons who were de facto

22 refugees who had, in fact, left Bosnia and Herzegovina; is that correct?

23 THE WITNESS: First if -- the registered voters, it is not only

24 people in Bosnia. It is also people in other countries. So the register

25 includes names of those voters who registered to vote outside Bosnia as

1 well, most of them from the neighbouring countries in the region of the

2 former Yugoslavia. But in my project my starting point is the original

3 pre-conflict population. Original pre-conflict population. I'm

4 interested in this group, and I want to find out what happened to this

5 group. If the person did not register, I am unable to say what happened

6 to this person, and this person is not matched with the census and the

7 person is out from my statistics, is not included.

8 JUDGE TRECHSEL: Right. And my mistake has been that you do know

9 about those who left the country, because they register and they indicate

10 that they are living outside the country in 1997, 1998.

11 THE WITNESS: Well, if they are living outside country, they are

12 in the voters register if they registered to vote.

13 JUDGE TRECHSEL: Yes. Okay. Thank you.

14 MR. KARNAVAS: Your Honour, just a follow-up to your question

15 though, does that mean that the expert knows exactly when they moved and

16 the reasons? So in other words if somebody in 1995, post-Dayton, moved to

17 Scandinavia legally for reasons whatever, you know, has nothing to do

18 with -- is that a refugee? Because -- is there a distinction? I think

19 that was part of your question.

20 JUDGE TRECHSEL: No --

21 MR. KARNAVAS: Well, it's related to it at least.

22 JUDGE TRECHSEL: It is -- it is your interest in the question,

23 because with the -- those who changed their place of residence within

24 Bosnia and Herzegovina, there is no indication whether they left because

25 they wanted to stay with their mother-in-law or whatever. So in this

1 respect it's the same for both categories, and that's why I didn't ask the

2 question.

3 MR. STRINGER: Thank you, Mr. --

4 JUDGE ANTONETTI: [Interpretation] Please proceed, Mr. Stringer.

5 MR. STRINGER: Thank you, Mr. President.

6 Q. I'm going to move to the next slide, because you've made reference

7 now a couple of times to a minimum number, and this is something that

8 appears in actually all three of your reports, what's called an absolute

9 minimum number and what you call a more complete minimum number, and it's

10 important but I'm also watching the clock.

11 The absolute minimum numbers are the concrete 142.000 names that

12 were matched between the two sources; is that correct?

13 A. Well, no. This 142.000 of names are the matches.

14 Q. Yes.

15 A. Out of them some stayed in home municipalities. Some of them

16 moved out and lived elsewhere by 1997-8. So these 142 are all. 142.000,

17 of course.

18 Q. Okay. So people who were part of that initial 142.000 who were

19 living elsewhere in 1997, 1998, they've been classified as refugee or IDP,

20 and you can actually give us a list of the names of those people; correct?

21 A. That's right.

22 Q. And those are your absolute minimum number individuals.

23 A. Yes.

24 Q. Okay.

25 A. That's right.

1 Q. Okay. Now, the bottom two points here then refer to more complete

2 absolute minimum numbers, and this is based on estimation method. And if

3 you could describe for us what the more complete numbers are and describe

4 the estimation method that you used to arrive at that more complete

5 minimum number.

6 A. Yes. Knowing that the absolute minimum numbers are incomplete and

7 knowing the reasons why they are incomplete, we made a very strong attempt

8 to compensate for the incompleteness, and we tried to come up with better

9 statistics.

10 In statistics, sampling methods are often used to produce values

11 of the unknown characteristics of the entire populations. Also, in this

12 project we applied sampling approach, a sample-based method in order to

13 produce more complete numbers.

14 From the matches and statistics of -- on minimum numbers of IDPs

15 and refugees, we were able to estimate for every ethnic group in every

16 municipality the proportion of the displaced and refugees in the entire

17 voters population of this ethnic group in this municipality. The

18 proportion is a sample-based measure that can be used in relation to the

19 population from the 1991 census and the proportion can be used to obtain

20 more complete minimum numbers of displaced persons and refugees. It is a

21 simple statistical approach. Very old, a well-established standard used

22 all the time in many areas in science, in research, and also in practical

23 life.

24 Q. Okay. And much of what you've said then is already -- is

25 contained on the next slide which I've moved to. You describe in greater

1 detail --

2 MR. STRINGER: Mr. President and Your Honours, all the mathematics

3 are found in Annex C2 and I do not propose to go into the mathematics of

4 it. It's all there for everyone to use as they wish.

5 Q. Ms. Tabeau, I do want to take you to the very last point on this

6 slide which relates to confidence intervals and we'll be seeing those and

7 so again briefly a description of what a confidence interval is.

8 A. It's a concept that is used to provide a measure of uncertainty of

9 a statistical estimate. Take, as an example, the proportion P of IDPs and

10 refugees in a given ethnic group in a certain municipality.

11 Q. Can I just interrupt you there. I'm looking at the transcript

12 that says -- you indicated that it's a concept that is used to provide a

13 measure of uncertainty is what it says and I don't know whether you

14 intended to say uncertainty or certainty in terms of a statistical

15 estimate?

16 A. Yes, these two terms are related. Of course, we want our estimate

17 to be as certain as possible, as good as possible. That means the

18 uncertainty of estimation must be as low as possible. So these are two

19 related concepts. Uncertainty tell us at the same time how good the

20 estimate is. Simply speaking, uncertainty tells us how large the error of

21 the estimation is.

22 When -- going back to the example of the proportion P of the IDPs

23 and refugees of a given ethnicity in a given municipality based on the

24 voters registration, this proportion is a sample-based measure of the

25 unknown, generally unknown proportion of all IDPs and refugees in the

1 entire population that was exposed to risk. So the sample-based

2 proportion comes from the voters register which is our sample, but if we

3 would take another sample, I wouldn't know what sample, but if we would

4 take another source and estimate another proportion based on another

5 sample, then most likely we would end with a slightly different estimate

6 of this proportion. This is how the sampling methods work.

7 So in order to have an idea of how large the error is of our point

8 estimate of proportion, we come up with the concept of confidence interval

9 which shows the lower end and upper end and between those two ends, lower

10 and upper, our estimates based on different samples would fall in.

11 If the interval is narrow, the quality of the estimation is good.

12 The error we make is made -- is small. We present in our report

13 confidence intervals for 5 per cent error or 95 per cent confidence

14 intervals. That means with 95 per cent we can say that our point estimate

15 of proportion of IDPs and refugees of a given ethnicity in a given

16 municipality will always fall within this interval.

17 Q. Okay. Two questions then from that. The estimation method, the

18 formulas, the principles that you used in moving from absolute minimum

19 numbers to the more complete, are those methods and principles generally

20 accepted and used in your field?

21 A. Of course. This is a standard. This is in textbooks for

22 statistics. Students have to learn how to do this and why.

23 Q. And the next question then is whether in this case in your study

24 whether the confidence intervals that you determined are -- are acceptable

25 within your field.

1 A. Well, these, very narrow intervals which give us 95 per cent of

2 confidence that our estimate of -- of P, of the proportion P, is covered

3 by this interval. So I can only say it is an excellent result.

4 Q. Okay. Now, we're talking about more complete minimum numbers, and

5 I'd like for you to briefly tell us why you're not able to say that these

6 are complete numbers. And I've moved on to the next slide.

7 A. Yes. There are three reasons. First reason is that people

8 younger than 18 years at the time of elections are not in our study. They

9 are excluded.

10 Second reason is that the voters population did not -- the voters

11 population, the matched population, only is related to voters who

12 registered to vote. So those who didn't register could not be matched,

13 and they are out.

14 Q. Okay. So that's -- you say you matched 80 per cent of the voters

15 registration. So these are the 20 per cent who were excluded because they

16 were not matched.

17 A. Yes. They were not matched. And there is another group of voters

18 who did not register, so they could not be matched as well, and here we

19 can only present our guesstimate of how large this group is. There are no

20 statistics, no numbers, no sources that could be used for better

21 assessment of this group of people.

22 Q. Okay. So now that we've --

23 JUDGE TRECHSEL: Just a small question. What about people who

24 died between 1991 and 1997, 1998.

25 THE WITNESS: They are excluded as well. They could not be

1 included. But at the same time we are unable to mark all that in the

2 census, because we don't have these names. I will discuss later today the

3 report on the killed persons, and then I will explain the difficulties of

4 measuring and collecting even information about the deaths.

5 JUDGE ANTONETTI: [Interpretation] Madam, let's not waste any time

6 on statistical data or math. We know and we can understand the figures.

7 Those who have studied mathematics will be able to add up by themselves,

8 but what we see on page 98, we can see that according to what you're

9 telling us, the rate would be 95 per cent of almost certainty and that

10 there would be a rate of 5 per cent left. So all these calculations bring

11 us to the following conclusions, that the figures that you will give us

12 later, you gave us those figures based on 95 per cent of certainty.

13 Should I understand you this way?

14 THE WITNESS: Your Honour, we're talking about the estimated more

15 complete minimum numbers of IDPs and refugees. And indeed these figures

16 tell us with 95 per cent of confidence, of certainty, that these will be

17 the right numbers.

18 JUDGE ANTONETTI: [Interpretation] Please pursue, Mr. Stringer.

19 MR. STRINGER: Thank you, Mr. President.

20 Q. Now, Ms. Tabeau, we're actually going to get into the findings,

21 and the next slide -- the next couple of slides are going to relate to the

22 composition, the ethnic composition of the municipalities and how the

23 ethnic composition of the municipalities changed as from 1991, based on

24 census, and then based on where those 142.000 people were living when they

25 registered in 1997, 1998.

1 So if you could just perhaps go through these -- these points and

2 give us this information in terms of general overview, differences

3 pre-conflict and post-conflict?

4 A. Well, generally speaking, we will be looking at 14 post-Dayton

5 municipalities within the area of the Federation of Bosnia and

6 Herzegovina. Fourteen municipalities in total.

7 Before the conflict, in 1991, eight of these municipalities had a

8 relatively majority of one ethnic group and six had an absolute majority.

9 What is an absolute majority? What's the relative majority? Absolute

10 majority is when one ethnic group is at least 50 per cent of the

11 population in a given municipality. Fifty per cent or more is absolute

12 majority.

13 Relative majority is when the largest group is smaller than 50 per

14 cent but the difference between the largest and the second largest is at

15 least 5 per cent. That is the relative majority. All other cases would

16 be considered cases of mixed ethnic composition.

17 So in the case of eight Herceg-Bosna municipality, pre-war

18 municipalities and 14 post-war municipalities we speak of eight

19 municipalities in 1991 with a relative majority and six municipalities

20 with absolute majority.

21 Q. Witness, while you're talking I'm going to move to the next slide

22 which has the table that I think might assist everyone in following, then,

23 what you're saying.

24 A. Yes. It is a list here in this table of all municipalities. It

25 is not only federation but also Republika Srpska. If we will be looking

1 at the municipalities marked or labelled in this table as the Federation,

2 and if we would be looking at light colours, perhaps here it is a good

3 moment to explain that I used green for the Bosnian Muslims, blue for the

4 Bosnian Croats, and red for the Bosnian Serbs. If these are dark colours,

5 they mean absolute majorities of these groups. The same colours but light

6 would symbolise the relative majorities in the municipalities.

7 For instance, if we take Vares as an example, in 1991, the colour

8 is light blue. This means it was a municipality with a relative majority

9 of Croats. The percentage, 40.6, is the size of this population in this

10 municipality. After the conflict, in 1997, 1998, the colour changed. It

11 is not blue any more, it is green, dark green. That means in this

12 municipality we observe an absolute majority of the Bosnian Muslims. With

13 the size of the Muslims 63.3 per cent.

14 So if we now look at the light colours in 1991, then it is eight

15 municipalities that are marked with light colours, disregarding whether

16 blue or green or whatever else. There are eight such municipalities in

17 the federal part of the area.

18 If we look at 1997-8, there are no light colours at all. All

19 municipalities are showing absolute majorities. It is when we look at the

20 federation either an absolute majority of the Bosnian Croats or absolute

21 majority of the Bosnian Muslims. And the numbers in 1997 are 7 and 7. It

22 is 7 absolute majorities of Muslims, 7 absolute majorities of Croats.

23 So the conclusion is that in this period, from 1991 to 1997, 1998,

24 the population in this area had become divided very strongly along the

25 ethnic lines.

1 Q. All right. So then, Ms. Tabeau, for all of these municipalities,

2 and these are now with the post-Dayton boundaries in order to -- so that

3 you can move from one to the next in terms of pre- and post-conflict --

4 well, let me do this. I want to use the ELMO, with the assistance of the

5 usher if I could, because I think it may be useful to show everyone then

6 how this information in this table, this information on the screen,

7 relates to the annexes that are found in your report, because what you're

8 going to be able to find in these annexes then is specific numbers in

9 terms of absolute minimum numbers of -- and how those numbers then changed

10 for each of these municipalities, and it may be useful just to show how

11 these tables fit in with that.

12 Well, I don't know how useful that's going to be because the one

13 I'm looking at is not very well in focus. I don't know whether they can

14 improve that or whether each of us can go straight to our own copy of the

15 report. Are they able to bring that up at all?

16 MR. KARNAVAS: Perhaps in part of this exercise, the gentleman

17 could also get from the witness how these figures of 1991 in the Mostar

18 municipalities, how they were arrived given that this was one municipality

19 and now we have several municipalities. How did those percentage arrive?

20 Based on what? I think that's relevant information for the

21 cross-examination that should be elicited on direct.

22 MR. STRINGER: Well, counsel can ask the question, Mr. President.

23 I'm in the middle of something that I consider to be important.

24 Q. I don't know that we're going to be able to use the ELMO for this.

25 THE INTERPRETER: Microphone, please.

1 JUDGE ANTONETTI: [Interpretation] I believed that Mr. Karnavas

2 raised a good point. I actually wanted to intervene myself.

3 Madam, let's go back to the coloured chart, the previous one. Not

4 this one but the previous one that we just saw where we had Mostar in

5 various different parts. So I'm referring to the previous chart. So not

6 the one -- yes.

7 JUDGE ANTONETTI: [Interpretation] Yes, the chart, indeed, the

8 table. This table, ma'am, could you very briefly give us some comments on

9 the changes in Mostar? I take it those are various areas of Mostar,

10 Mostar Zapad, Mostar Centre. What was the evolution? How did it change?

11 Can you briefly tell us?

12 THE WITNESS: I mentioned earlier that the administrative division

13 of Bosnia and Herzegovina into municipalities changed with the Dayton

14 peace agreement. Instead of 109 pre-war municipalities there were many

15 more established. Mostar is one of the municipalities that had been

16 split. Mostar had been split into eight municipalities. The precise

17 territorial division was not available until approximately mid-1997. At

18 mid-1997, we were able to acquire from the statistical authority in Bosnia

19 and Herzegovina a detailed bridging system of -- that allowed us to

20 recalculate the census data according to the new post-Dayton division of

21 Mostar.

22 A bridging system is, basically speaking, a big table in which

23 there are two types of territorial units, coded, of course. One is a

24 municipality, and another one is a settlement. A settlement is used in

25 statistics, official statistics usually, as the basic statistical unit.

1 It's a small area. Settlements can be grouped in different ways. We can

2 group settlements in order to obtain the pre-war municipalities, post-war

3 municipalities, whatever, other larger areas required for whatever

4 purpose.

5 This is basically how statistics are compiled. We need

6 information, a code, for every record in a source related to a place of a

7 residence and for us this small -- smallest area used in the aggregation

8 of data is the settlement. There are also other systems like census

9 enumeration areas, other systems. We worked with settlements as it is

10 normally done in official statistics.

11 JUDGE ANTONETTI: [Interpretation] Very well. I'm not talking

12 about the method, however. I'm just talking about the results. If I look

13 at Mostar, Mostar is divided in eight parts. I can conclude, according to

14 your table, that Mostar Jug went from 89 to 76. So it means that the

15 Croats have increased. The number of Croats had increased. And that

16 increase of Croats is also visible in Mostar Jugozapad, 47.9 versus 78.2.

17 And when we look at Mostar Zapad, 41.6 and 75.2. However, in the other

18 areas, it's an increase of Muslims that we note and in increase of Serbs

19 and for Mostar Srpski, for instance. That's what we see. Is this what we

20 can conclude upon reading your table?

21 THE WITNESS: Yes. This is exactly what is in there.

22 JUDGE ANTONETTI: [Interpretation] Very well. That's all I needed

23 you to say. Thank you. Please proceed, Mr. Stringer.

24 MR. STRINGER: Thank you, Mr. President.

25 Q. So I would like to try to bring this back to one of the tables and

1 maybe rather than the ELMO if I could direct you to page 43 of your report

2 Ms. Tabeau. This is annex 2(A), and all of this information by

3 municipality and by ethnic group is contained in annex 2(A) in the series

4 of tables that are found there. So page 43 has the table 2M for Muslim in

5 Annex 2. And just looking at the first municipality there, Capljina,

6 perhaps you could -- let me ask you a couple questions about that. Again

7 my intention is to hopefully bring us back to this table that we've

8 been -- that we've been looking at. I'm sorry, I should have directed you

9 to the table on page 38, Annex A1, because we're in Annex A1, not Annex

10 A2.

11 A. That's right.

12 Q. My apologies. Okay. Now, are you able to indicate for us in this

13 table how -- well, the relationship between this -- this annex 1M and the

14 table which is the PowerPoint slide? And you can pick any of the

15 municipalities you wish.

16 A. Yes, I suggest to pick Gornji Vakuf, which is the next one after

17 Vares. Gornji Vakuf. If we refer now to the table, in the first part of

18 the table which is related to the 1991 population, in the third column for

19 Gornji Vakuf we see the number 55.4 per cent. That's the number in the

20 table.

21 Q. That's the percentage of Muslims living in Gornji Vakuf in 1991.

22 A. That's right.

23 Q. Okay.

24 A. So if we move within the same table to the second panel of the

25 table with three columns again, this panel is titled 1997 sample

1 population, in the third column in this panel we read the number 61.5.

2 61.5. And this is what we see for Gornji Vakuf in this table.

3 So for Gornji Vakuf, the ethnic composition was characterised in

4 both years, 1991 and 1997-8, by absolute majority of Muslims. This is how

5 to read this table.

6 Q. Okay. Let me just ask you to look at Stolac while we're on this

7 one.

8 A. Mm-hmm.

9 Q. And again this table, 1M, will only have the numbers and the

10 percentages for Muslim populations.

11 A. Mm-hmm.

12 Q. Table 2C would have the information that would correspond to the

13 Croat populations in these -- in these areas, but for Stolac, anyway,

14 again we have a pre-conflict Muslim population of 48 per cent.

15 A. Yes, which is a relative majority, a relative majority of Bosnian

16 Muslims.

17 Q. Okay. And then moving across to the post-conflict population of

18 Muslims in Stolac.

19 A. We would need to take another table from the report. At this

20 moment we were using the table 1M, the letter M indicates it's the table

21 for the Muslims, and we have to look now for the table 1C.

22 Q. Okay, but before we do that, I just want to stay with this in

23 terms of staying with the Muslim population there in Stolac. Moving to

24 the far right column, then, does this indicate that the Muslim population,

25 percentages of Muslims living in Stolac after the conflict had decreased

1 by greater than 99 per cent?

2 A. Yes, this is right.

3 Q. Okay.

4 A. The last -- very last column to the right tells us the percentage

5 change in the size of this group in this municipality, and it is more than

6 99 decrease, 99 per cent decrease.

7 Q. Okay. Okay. So that is hopefully offered to assist the Trial

8 Chamber and the parties in linking up the information and the table on the

9 PowerPoint slide to the individual tables that are contained in Annex 1.

10 And we've included -- as well, Witness, I think just perhaps we could --

11 JUDGE ANTONETTI: [Interpretation] Madam, for Stolac, in 1991

12 Stolac-Berkovici, there was 714 Muslims, and in 1997 there was only 1

13 left, is that right?

14 THE WITNESS: Yes. This is what is in the table.

15 JUDGE ANTONETTI: [Interpretation] Very well. Thank you. Please

16 proceed.

17 MR. STRINGER:

18 Q. And just to clarify on this now, with these numbers, these numbers

19 of individuals that are referred to, these people, who figure in this

20 table, are these the 142.000 that were the absolute minimum number of

21 matches or does this include some level of estimation?

22 A. Well, it doesn't include any estimation. It is the statistics

23 that relate to the sample of voters that we had in our data. There is no

24 estimation included here.

25 Q. Okay. All right. Now, if you were to turn to page 39 then, which

1 you wanted to do, and look at the numbers in respect of the ethnic Croats

2 in these municipalities, what do you find in respect of Stolac?

3 A. Well, for Stolac, for 1991, the percentage of the Croats was 35.4,

4 and it increased in 1997-8 to 95.5 per cent.

5 Q. Okay. So that figure 95.5, that's an absolute majority that's

6 reflected, then, on the table that is found in the PowerPoint slide, this

7 one.

8 A. Yes. Yes, that's right. That's exactly the same figure.

9 Q. Okay. So then would this indicate that the percentage of Croats

10 living in Stolac, the federation part of Stolac, increased by 169 per cent

11 as between 1991 and 1997, 1998?

12 A. Yes. This is what is here in the table.

13 JUDGE ANTONETTI: [Interpretation] Witness, one should always be

14 aware of figures and percentages. When you look at Stolac, FBH, number of

15 Croats, in 1991 I see 4.663 Croats. Do you agree with me? In 1997,

16 however, people who were registered on voting list 4.959 people. Is that

17 right?

18 THE WITNESS: Yes, that's right. That's the effect of population

19 movement.

20 JUDGE ANTONETTI: [Interpretation] Very well. But in reality this

21 movement of population concerns 600 people; is that right? 600 people out

22 of 4.000 people, because there are 600 additional Croats. Is that right?

23 So the increase is 600 people with respect to 4.000 people. And when you

24 say there is an increase of 169.5, or 600 out of 4.000, this does not add

25 up to 165.

1 MR. STRINGER: Mr. President, if I could clarify that, because I

2 had some difficulty with this one as well.

3 Q. Ms. Tabeau, pre-conflict in the federation part of Stolac can we

4 say that 35.4 per cent of the population in Stolac, this part of Stolac,

5 was Croat?

6 A. Yes, we can.

7 Q. Okay. And that there were 4.363 Croats according to the census

8 there?

9 A. Yes, that's right.

10 Q. Okay. And then after the conflict, first of all, it would appear

11 that the -- the total population of this part of Stolac diminished by

12 greater than half over the course of the conflict because originally there

13 had been 12.000?

14 A. That's right. But this is the only a sample in 1997 but that is

15 the case.

16 Q. These are the absolute minimum numbers?

17 A. That's right, yes.

18 Q. So post-conflict we have 5.192 people living there, of which 95.5

19 per cent are Croats.

20 A. That's right.

21 Q. So the figure then to the right the 169.5 per cent, that reflects

22 the change in the percentages, not the changes in the number of people.

23 So moving from 35.4 per cent to 95.5 per cent represents 169 per cent

24 increase in the percentages. So that's --

25 A. That's very correct. That's very correct. This is how it was

1 obtained.

2 Q. Okay.

3 A. But basically this percentage just resembles the change that we

4 see in absolute numbers in some sense as well.

5 Q. Okay.

6 JUDGE ANTONETTI: [Interpretation] Madam, this is what I notice:

7 There were 4.300 Croats in 1991. 4.300 Croats; is that right? Now in

8 1997, we find 4.959 Croats. So there are 600 Croats more with respect to

9 the ones who were there in 1991. Those additional 600 Croats, if I

10 compare it to the 4.300 who were there originally, I can notice in

11 relative figures an increase of Croats of 15 per cent. We agree on this,

12 do we?

13 THE WITNESS: You -- Your Honour is right, only that we are not

14 allowed methodologically, not allowed to work with absolute numbers. This

15 is wrong, because on one hand, we have 1991 population census, complete

16 population and on the other hand we have a sample of this population. So

17 absolute numbers should not be compared. They may not be impaired. What

18 may be compared are percentages. The percentages are, statistically

19 speaking, reliable measures of the ethnic composition and we may compare

20 them. It's basically the same thing we did only we did it with

21 percentages not with absolute numbers as Your Honour did.

22 JUDGE TRECHSEL: Remaining for a moment with absolute figures,

23 it's important to know that of 5.905 Muslims in this period some 5.887

24 left in the time, because only 18 remained and this is -- this is the big

25 change in absolute numbers. Here we have the drain really.

1 THE WITNESS: This is how it is.

2 MR. STRINGER: Yes --

3 THE WITNESS: But still we measure this in terms of terms of

4 percentages because that's a good way to do it.

5 MR. STRINGER: And that was going to be the next point,

6 Your Honour, actually is that really it's useful to move from one table to

7 the other because, obviously, you can only put so much information on one

8 piece of paper, and so that really I guess the point would be that we

9 would be offering these as a unitary picture that really can't be viewed

10 in isolation in only one table. It really requires some assimilation of

11 all of the information.

12 I'll move to the next couple of slides, and it will hopefully

13 illustrate graphically what we've been talking about only with respect to

14 all the eight municipalities, and these slides are in your packages so you

15 can put them together or side by side if you wish?

16 Q. Ms. Tabeau, this is the configuration of the ethnic composition

17 for the pre-conflict; is that correct?

18 A. Yes, this is 1991.

19 Q. Okay. And again, with relative and absolute majorities indicated

20 by the colour.

21 A. Exactly the same system of colours is used here as in the table.

22 Q. And then this is now the slide that relates to the post-conflict

23 configuration; is that correct?

24 A. Yes, it is correct.

25 Q. Okay. And now what we see are all of the municipalities in

1 absolute majority.

2 A. Yes, that's true.

3 Q. And so that one can get some idea of comparing these two maps

4 which direction the populations moved; is that correct?

5 A. Yes. It's -- it's good to pay attention to two municipalities,

6 Vares and Stolac, Vares and Stolac. It's light blue, light blue in 1991,

7 and dark green in 1997-8. And Stolac, light green in 1991, dark blue in

8 1997-8. So this is not only a change from relative to absolute majority.

9 It is a change from one ethnicity to the other, and it happens that these

10 two groups were engaged in the conflict in Herceg-Bosna.

11 Q. That's a glimpse of the ethnic composition of the municipalities

12 and, Mr. President, what I propose to do now is to move on to examine more

13 closely the ethnic composition of these refugees and IDPs, the 142.000

14 that we've been discussing and conclusions that can be taken from -- from

15 those.

16 Again, bringing it back to the matching process, can you tell us

17 again number -- how many people are we looking at now, those who didn't

18 live in their 1991 place of residence when you captured them in the voter

19 records in 1997 and 1998?

20 A. So the minimum number, absolute minimum number of internally

21 displaced persons and refugees who at 1997-8 did not live at their pre-war

22 residence is 61.487 in the eight Herceg-Bosna municipalities. It's the

23 absolute minimum number. This is people whose names can be documented

24 from the results of the matching, 61.487 individuals.

25 And in addition to this number of which we know it's severely

1 incomplete, we present on the same slide and in the report the second

2 number, which is the more complete, still minimum number obtained based on

3 sample-based estimation, more complete absolute minimum number is 101.107.

4 Q. Excuse me, Ms. Tabeau, I don't mean to cut you off.

5 MR. STRINGER: Mr. President, there is some discussion taking

6 place in the courtroom that is becoming a distraction to me, and I believe

7 to the witness as well, and I wonder if counsel could try to refrain. I

8 know that it's difficult sometimes, but we're trying to do a direct

9 examination, and it's quite noisy.

10 JUDGE ANTONETTI: [Interpretation] I don't know. I looked at the

11 counsels and they were hanging at your every word. I did not hear them

12 talk amongst themselves, but maybe there is some interferences. We are

13 hanging to your lips, all of us.

14 MR. STRINGER: Thank you, Mr. President. I'll choose my words

15 very carefully then.

16 Q. Okay. Ms. Tabeau, you were then describing the more complete

17 minimum number which has been arrived at through the estimation of process

18 and the statistical analysis that you described earlier; is that correct?

19 A. Yes, it is correct.

20 Q. Okay. All right. So then for the following numbers we're going

21 to be presenting, Mr. President, both the absolute minimums based on the

22 known 142.000 people that is followed, and then the more complete number

23 that is based on the estimation methods that have been described. And

24 this is getting into a series of tables and figures that are found in

25 annex 2 of Ms. Tabeau's report; is that correct?

1 A. Yes, it is.

2 Q. And it may be that this is a table that we can use on the ELMO

3 with greater success if I can find one. But I can't find one at the

4 moment, so we'll continue looking, and I'll just direct everyone to the

5 slide that is in front of each of you. Here it is. Thank you. Okay.

6 It's on the ELMO, and it may be useful, Ms. Tabeau, you may want to use

7 the pen and point to the numbers as you talk about them. Tell us what

8 this table is about. Now that we're talking about the ethnic composition

9 of the people and not the municipalities?

10 A. Yes. This table contains some more statistics that are related to

11 the area of eight Herceg-Bosna municipalities studies -- studied. The

12 first familiar number here in this table is 231 - approximately -

13 thousand. This is the population from the 1991 census. This is the

14 population born before 1980 that we studied in this project.

15 The second familiar number is the 142.000 approximately. This is

16 a record from the voters registration. These are records that had been

17 matched with the census, so the 142.000 are at the same time here as part

18 of this population. This is complete population. This is our sample.

19 Based on this sample, we were able to identify approximately

20 61.000 of internally displaced persons and refugees. The rest of this

21 group were persons who stayed after the conflict at the time of elections

22 at their pre-war place of residence, but 61.000 did not. 61.000. This is

23 the minimum number.

24 Q. So the 61.487, again this is a documented number of persons who

25 were not living at their 1991 place of residence?

1 A. That's right.

2 Q. Okay. And again moving across to the estimated number, what would

3 that be?

4 A. It is 101.107.

5 Q. Okay. All right. So if I could direct you to the third column,

6 which is the observed minimum number of IDPs and refugees, let's look at

7 the composition of those 61.487. First of all -- well, can you just move

8 down that column and tell us what it means.

9 A. This column tells us the ethnic composition of the IDPs and

10 refugees. It shows us the size of every ethnic group of IDPs and

11 refugees, and the first number given for Muslims here, I skip the first

12 row which is for all known Croats jointly, first number, the 26.663, is

13 the number of IDPs and refugees of Muslim ethnicity who didn't stay at

14 their pre-war residence in 1997-8.

15 Q. Then moving across, Muslims account for 43.4 per cent of all of

16 these people who don't live any more at their 1991 place of residence.

17 A. Yes, that's right. That's right. This is the largest group, 43.4

18 per cent. The second largest is the Croats with 25.6 per cent, then the

19 Serbs 23.8 and others jointly at 7.3 per cent.

20 Q. And what you've demon the final column is to apply your estimates

21 to what you call the more complete minimum number?

22 JUDGE ANTONETTI: [Interpretation] Mr. Praljak, yes.

23 THE ACCUSED PRALJAK: [Interpretation] I just have a question. Why

24 aren't relevant percentages being used here in relation to the population

25 but, rather, absolute figures? If the refugees were to be compared with

1 the previous population, then these percentages wouldn't be valid, so we

2 are having double standards here.

3 JUDGE ANTONETTI: [Interpretation] Witness, indeed. Why did you

4 not use the same method.

5 THE WITNESS: I believe I did use the same method. I first of all

6 present both, absolute numbers and percentages. I made it clear that the

7 1991 population is a complete population, and that 1997-8 is a sample.

8 For my sample, which is very big and solid, statistically speaking, I am

9 allowed to calculate ethnic composition, the same as I did for the 1991

10 census population. There is nothing wrong in this method absolutely. I

11 see no methodological bias in this table.

12 JUDGE ANTONETTI: [Interpretation] Very well. Thank you. You may

13 proceed.

14 MR. STRINGER:

15 Q. And just one last question about this then. In terms of absolute

16 minimum numbers of documented persons as well as the estimated more

17 complete, what is the percentage of non-Croats who no longer live at their

18 1991 residence?

19 A. Well, it is according to the observed minimum 74.4 per cent, and

20 according to the estimated minimum it is approximately 75 per cent,

21 three-quarters.

22 Q. Now, I would like to direct you to your table found at page 43 of

23 the --

24 JUDGE ANTONETTI: [Interpretation] Yes, Mr. Prlic?

25 THE ACCUSED PRLIC: I just need one clarification because I'm not

1 able to understand. This is not the essence of -- now we are dealing with

2 the 61.000 refugees and displaced persons. Before we dealt with 118.792

3 those who remain in municipalities. So the total number is 179.000. But

4 we are dealing all the time with 142.000. You told just before on page

5 75, line 19 and 20, that out of 142.000 we should reduce for this 61.000

6 to have those who remain in the their municipalities. So it should be

7 less than 81.000 but before we all -- we have all percentages, all data

8 based on this 118.000. I really don't understand. This is just the issue

9 of clarification of numbers; I'm not speaking about methodology or

10 anything. Because I'm not able to analyse that without that.

11 JUDGE ANTONETTI: [Interpretation] Yes, madam.

12 THE WITNESS: Yes. I want to explain again that in this table we

13 show first census population complete, then the sample incomplete. The

14 sample size is 142.000 approximately. And then this total can be further

15 distributed into two groups. One group is IDPs and refugees. They are

16 mentioned in the table here. There are 61.000 approximately mentioned in

17 the table. The other component is not included in the table, and this is

18 perhaps what you are missing.

19 THE ACCUSED PRLIC: [Interpretation], no, I'm not missing anything,

20 but we are dealing with page 39.

21 THE WITNESS: 39 of the report --

22 THE INTERPRETER: Please slow down for the record.

23 THE ACCUSED PRLIC: [Interpretation] [Previous translation

24 continues] ... 118.

25 THE INTERPRETER: Please do not overlap, thank you.

1 THE INTERPRETER: We cannot follow. It's going too fast and

2 overlapping.

3 MR. KARNAVAS: I wish if the OTP witness an employee would

4 allow --

5 JUDGE TRECHSEL: Mr. Karnavas, it is not necessary at every

6 occasion to polemically refer to the employment situation of the witness.

7 It is pure polemics.

8 MR. KARNAVAS: Well, she is an employee. Very well. If this

9 expert witness would allow my client to make his record and not be

10 interrupted because he's asking for clarification which is obviously

11 needed, given -- given that she's mixing percentages and figures.

12 JUDGE ANTONETTI: [Interpretation] Madam, Mr. Prlic asked you to

13 give him some precisions regarding the table that we have on the screen

14 or, rather, in 1991 population census and 1997 register of voters, and we

15 have another figure, 142.000 people, then we have 61.487 people, and then

16 101.107 people. So would you be able to give an answer to Mr. Prlic

17 regarding his -- these questions -- or to these questions?

18 THE WITNESS: I can. The reference was made to table in Annex A1.

19 This is the table 39, and this is table 1C, as I understand. And in this

20 table, this is again I repeat table 1C, page 39 of the report. I think

21 the question is related to this table and not to the table that is on the

22 screen or although these two are, of course, related to each other.

23 The 1997 population in table 1C, page 39, is 118.792. And this is

24 a very good point you made. We have measured population in 1997 in two

25 different ways. One is the so-called "in" population. Someone the

1 so-called "from" population. If you will refer to the report where the

2 table 2 is included, then you will see that in the report you will see two

3 totals for 1997-8, the in population and from population. So the 118.000

4 approximately is included in the table in the report. For simplicity, we

5 skipped it here.

6 But what is this 118.000? This is the registered voters in --

7 observed in the eight Herceg-Bosna municipalities, in, disregarding,

8 disregarding where did they come from. We didn't pay attention at this

9 moment where did they come from. It is 1997-8 in population. But next to

10 this there is another population. There is this from population, and this

11 from population is very different, because from population is closely

12 related to the census population. It is a subgroup from the census

13 population that we traced in the voters register, and this is this

14 142.000. Part of them are in as part of the in population, but part of

15 them are elsewhere, in other municipalities.

16 So the displacement, if you think about it, is not necessarily

17 within the Herceg-Bosna only. The displacement is much broader than that.

18 People can move within Herceg-Bosna from one municipality to another,

19 but also from Herceg-Bosna municipalities to any municipality in Bosnia,

20 or to other countries.

21 So there is a very considerable difference between population

22 called "in" and population called "from". Population called "from" is

23 essentially for the assessment of displacement, but population called

24 "from" cannot be used for assessment of ethnic composition, of attempting

25 the ethnic composition.

1 MR. KARNAVAS: Your Honour, I object to the word "displacement"

2 unless she can prove that there is actual displacement. When somebody

3 moves from one place to another it doesn't necessarily mean that they're

4 being displaced. It could be voluntarily. I think that was the essence

5 of Judge Trechsel's earlier question to which I sought some clarification.

6 So I just think that -- And I don't wish to engage in polemics, but

7 let's stay with statistics, mathematics but not political or legal terms

8 such as "displacement."

9 JUDGE ANTONETTI: [Interpretation] Madam, while you were talking I

10 was thinking exactly along the same lines as Mr. Karnavas. You're talking

11 about the movement, movement of a population or a displaced people. Maybe

12 there is a Croat who, for instance, lives in an X municipality and for

13 some reason this particular person decides to go to another municipality,

14 let's call it municipality Y. So this person just changes places but that

15 person can also change whether he will be registered or not in the voters

16 list, but this person is not displaced in the sense that he was displaced

17 because of the conflict. Maybe in your figures you can find people who

18 for various reasons they changed the place of residence because they

19 bought another apartment, they bought another house and they decided to

20 move.

21 So when you use the term "displacement" in your mind, who do you

22 cover with that term? You talk about displaced people, but who are these

23 people exactly?

24 THE WITNESS: I simply use this term for people who change their

25 residence between 1995 and 1997-8. There is no negative connotations.

1 There is nothing. I made a distinction earlier today. I said I use

2 statistical definition of displacement; I don't use legal definition.

3 There is no notion of the fear of being persecuted. There is nothing like

4 that in the way I'm using this term.

5 JUDGE ANTONETTI: [Interpretation] Very well. I believe it's time

6 to take a break. Mr. Stringer, how much time do you need? How much more

7 time do you need?

8 MR. STRINGER: I need all of my three hours, Mr. President. I'm

9 not sure where I -- I'm not sure where I stand at this moment.

10 JUDGE ANTONETTI: [Interpretation] Well, you must have taken and

11 I'm going to tell you almost two hours. We will actually count it for you

12 if you wish.

13 You've used 110 minutes so far, so one hour and 50 minutes.

14 MR. STRINGER: That's consistent with our count. So --

15 JUDGE ANTONETTI: [Interpretation] Very well. So you may not be

16 able to finish today, I suppose.

17 Let's take a short 15-minute break because I would like to finish

18 at five to 7.00. If you are not finished, we will continue next time.

19 Since Mrs. Tabeau is always here anyway, I do not believe that it will be

20 difficult for her to come back to -- at another date. But since you

21 haven't talked about the injured and the dead people, I think you wanted

22 to talk about that.

23 THE INTERPRETER: The killed and the wounded, correction.

24 MR. STRINGER: I think it's still possible to get through this

25 today, Mr. President. I'm halfway through the slides, and we already know

1 a lot about the estimation methods and things that we won't have to cover

2 again when we talk about the other studies. So it's possible. I'm

3 certainly going to do my best to try and finish the direct today if I can

4 at all.

5 JUDGE ANTONETTI: [Interpretation] Very well. So we will take a

6 short 15-minute break.

7 --- Recess taken at 5.41 p.m.

8 --- On resuming at 5.58 p.m.

9 JUDGE ANTONETTI: [Interpretation] Mr. Stringer, you may proceed.

10 Sorry, Mr. Prlic. I had not seen you.

11 THE ACCUSED PRLIC: I didn't get the answer on my question. If

12 you are satisfied, that's fine, but I think nobody is able to follow

13 further direct examination without clarification. We heard many times

14 during the afternoon that the total number of matched applied for voting

15 was 142.214. Again, many times we mentioned and we made all those

16 calculation with colours, with the Vares municipalities with 118.792

17 registered in those eight municipalities. So the differences -- the

18 difference between those two data is 23.412. So those data should apply

19 to those registered for voting out of those eight municipalities or

20 abroad.

21 Now we receive information that the minimum number of those apply

22 for voting in 1997, 1998 abroad or in different municipality is 61.487.

23 If we, very simple, add those 61.000 to 118 registered in those

24 municipalities and we have all those percentages, based on that, we are

25 going to have the total number of registered for voting 180.289, those who

1 matches with the data from census. So the difference between the first

2 information, 142.000 and this one is almost 40.000. So all calculation

3 based on those differences are not credible, and I think without

4 clarification of that it is impossible to continue.

5 JUDGE ANTONETTI: [Interpretation] Witness, Mr. Prlic has just made

6 a demonstration for our benefit. He took into account the 140.000, and he

7 looked at the figure of 118.792. He tells us, based on what I don't know,

8 but he probably will be able to establish it, he says that 61.487 persons

9 were registered abroad to vote. In other words, the total number of

10 voters was 180.299 and not 140.000 something. In other words, according

11 to Mr. Prlic, there is a difference of 40.000 people, and according to him

12 this could have an impact on your statistics, on your figures on your

13 conclusions what do you have to respond to his submissions?

14 THE WITNESS: Well, making these kind of calculations is a

15 dangerous exercise. This shouldn't be done, simply speaking. These two

16 populations cannot be compared in the way you did. The population in the

17 118.000 is the population coming from the voters register, the voters who

18 registered in the eight municipalities. Some of them are originally from

19 1991 census population and some are not. Some are newcomers who arrived

20 in these municipalities between 1991 and 1997-8 for whatever reason. So

21 this is included in 118.000. So there is an overlap, an overlap between

22 this population, 118.000, and the 142.000 and the census population, but

23 this is only an overlap, not one-to-one correspondence.

24 So what I'm trying to say, the from population, 142.000, as a

25 whole can be traced back in the census, the population, census population,

1 from these eight municipalities, which cannot be said about the population

2 in 118.000.

3 MR. STRINGER:

4 Q. Ms. Tabeau, let me just ask you a couple of questions to follow

5 up.

6 THE ACCUSED PRLIC: If you are satisfied, I am satisfied. I'm

7 completely fine with it.

8 THE WITNESS: Yes.

9 THE ACCUSED PRLIC: But, madam, you made calculation based on

10 that, for Stolac and for other municipalities showing the real data. Very

11 clearly, you told the total registration which matches this 142.000.

12 Those registration consisted of registration in municipalities, plus out

13 of municipalities in municipalities 118.000 registered, "yes," or "no", in

14 eight municipalities?

15 THE WITNESS: Well --

16 MR. STRINGER: Mr. President, this is for cross-examination.

17 THE ACCUSED PRLIC: You are not able to continue.

18 JUDGE ANTONETTI: [Interpretation] Wait a minute. Mr. Prlic very

19 rarely takes the floor here, and when he decides to take the floor he

20 always does it in a very sensible manner.

21 According to him there is a major problem here, so let us listen

22 to his questions. Of course, the questions might be asked during

23 cross-examination, but if we can save time, let's go for it. So I'll ask

24 the witness to answer these technical questions. Amongst other things a

25 question you have not answered yet, the number of people registered

1 abroad, 61.487, does this have an impact on your conclusions? Because it

2 seems to me that this has not been taken into account. So I'd like to ask

3 you to answer Mr. Prlic's questions very specifically.

4 Mr. Prlic, can you please repeat your question.

5 THE ACCUSED PRLIC: So madam, you told that 118.000 eligible

6 voters registered for voting in those eight municipalities. I'm not

7 making any distinction between them, are they from 1991 or not. So this

8 is the result.

9 THE WITNESS: Yes. 118.000 registered in the eight

10 municipalities, yes.

11 THE ACCUSED PRLIC: Yes.

12 THE WITNESS: Disregarding where they come from.

13 THE ACCUSED PRLIC: The second number to get this 142.000 is

14 registered abroad.

15 THE WITNESS: Not abroad.

16 THE ACCUSED PRLIC: Abroad or in other municipalities.

17 THE WITNESS: Displaced and refugees.

18 THE ACCUSED PRLIC: I am speaking for the third time, and I'm

19 speaking about refugees and displaced persons, I am using your definition.

20 So those registered out of these eight municipalities in

21 Bosnia-Herzegovina or abroad; is it correct?

22 THE WITNESS: What number did you mention?

23 THE ACCUSED PRLIC: [Interpretation] You mentioned 61.000.

24 THE WITNESS: Yes.

25 THE ACCUSED PRLIC: [Interpretation] But in accordance with this

1 those information we have 118.000 registered in those municipalities plus

2 23.000 is 142. I don't know what is the root of this 38.000. This is

3 crucial for understanding. Maybe I'm wrong but --

4 THE WITNESS: I will read your numbers first on the screen.

5 MR. STRINGER: Mr. President, could I ask two questions of the

6 witness in the hope that I can clear this up?

7 MR. KARNAVAS: I believe a question has been posed to the witness.

8 The witness should be entitled to answer it.

9 MR. STRINGER: It's my direct examination. That's my response,

10 Mr. President.

11 MR. KARNAVAS: I believe, Your Honour, that you've given leave to

12 Mr. Prlic to ask questions and I believe there is no need to assist this

13 witness.

14 THE WITNESS: Well, the 61.000 if I --

15 JUDGE ANTONETTI: [Interpretation] Please answer, madam, because

16 Mr. Prlic has raised an objection so we need to deal with this. Please

17 answer the question put to you by Mr. Prlic. You've asked to be able to

18 have a look at the screen, then of course Mr. Stringer will be able later

19 to ask additional questions. Please give us your answer, madam.

20 THE WITNESS: Regarding the 61.000, these are people who did not

21 register at the pre-war place of residence. 61.487, these are the IDPs

22 and refugees. So these are -- this is the minimum number of IDPs and

23 refugees. This is what is in the table number 2 which I still have on the

24 screen. I don't understand the question perhaps.

25 JUDGE ANTONETTI: [Interpretation] Witness, this figure of 61.487

1 corresponds apparently to persons who are registered to vote abroad, but

2 amongst these 61.487 individuals there might be people, Muslims working in

3 Berlin, for example, and who were registered to vote abroad, but it does

4 not necessarily mean that we have to do with refugees or IDPs.

5 THE WITNESS: But what's the question? I still don't understand

6 the question. What question am I supposed to answer?

7 I, again, repeat the 61.486 is the number of internally displaced

8 persons and refugees at whatever location they are in 1997-8. Partly they

9 are in 8 Herceg-Bosna municipalities, partly they are or not in other

10 municipalities in Bosnia-Herzegovina, partly in other countries. And it

11 has very little to do with the 118 population registered in the eight

12 municipalities. Very little to do because I suppose some of the 61.000,

13 some of the 61.000, are still in the eight Herceg-Bosna municipalities,

14 only different municipalities than in 1991, but not all. It is not that

15 simple that you can add up these numbers and compare them, because these

16 groups overlap simply with each other. It's not that simple.

17 JUDGE TRECHSEL: If I may take your term "overlap." Is it so that

18 part of those 61.000 are included in the 118.000.

19 THE WITNESS: Very correct. This is how it is. Part, but not

20 all. I can answer this. If you are interested which part, I can answer

21 that. I can run queries and compare this, but it is not right now I can

22 give you this number.

23 Some are in 118.000, some of 61.000. Certain not all, 100 per

24 cent sure.

25 JUDGE TRECHSEL: In other terms, the figures, the quantities that

1 you mention can be visualised as partly overlapping circles.

2 THE WITNESS: Excellent, yes. This is how it is. This is a

3 complex statistical analysis that is made on the basis of individual data.

4 Moreover, several concepts are used that are usually used in the

5 analysis of migration. There is a municipality of departure. There is a

6 municipality of arrival or destination. We have a study area and a

7 broader universe that is Bosnia and Herzegovina. So these are complex

8 population movements, and it is very hard to measure them and to express

9 them in a simple way. We think we did it in a simple way, but adding

10 sources like this cannot be done.

11 JUDGE ANTONETTI: [Interpretation] In other words, what you're

12 saying in response to Mr. Prlic is that these 61.487 persons are included

13 in the 118.000, partly included, and that we should not add them up as

14 Mr. Prlic had done. That's your answer.

15 THE WITNESS: That's my answer. That's my answer.

16 JUDGE ANTONETTI: [Interpretation] Mr. Stringer, you may proceed.

17 MR. STRINGER: Thank you, Mr. President.

18 Q. Witness, I've put on the ELMO there during the break a table or

19 two from Annex 2, because again to bring it back to where I'm at, which is

20 the composition of the 62.487, that is, the ethnic composition, the

21 profile of those people who were not there any more, they were there in

22 1991, they were gone elsewhere, maybe a different municipality in

23 Herceg-Bosna, maybe a different municipality of Bosnia-Herzegovina, maybe

24 a different country, but I want to talk about the ethnic composition of

25 those 61.487 people who weren't there any more in 1997, 1998. They were

1 not at the place they'd been living in in 1991. Okay?

2 Now, the ELMO is table 2M. This relates to the ethnic composition

3 of -- of Muslims as within those 61.000; is that correct?

4 A. Yes, it is.

5 Q. Okay. And again just to give us some general numbers, moving

6 across the top line again we have the 61.000 who are -- you're -- you've

7 called IDPs or refugees because they're not living at their 1991 residence

8 any more.

9 A. Yes, that's right.

10 Q. Okay. And you've got numbers for Muslims there then; right?

11 A. Yes.

12 Q. Okay. Now, there's a 49, 49 per cent. What does that relate to?

13 A. This is the proportion of Muslims, IDPs, and refugees of Muslim

14 ethnicity, in the total of IDPs and refugees in 1997-8, that is, 26.663 as

15 related to 61.487, and that's the proportion, 43.4 per cent.

16 Q. Okay. So 43.4 per cent of all the people who had moved were

17 Muslims.

18 A. Yes.

19 Q. Now moving to the left one column there's the 49 --

20 A. Well, excuse me, no. 43.4 is the fraction of Muslims within all

21 IDPs and refugees.

22 Q. Yes. So you've got Muslims, Croats, Serbs.

23 A. Yes.

24 Q. And others, all of whom fall within the 61.000.

25 A. Yes.

1 Q. Muslim account for 43 --

2 A. -- 3 per cent.

3 Q. 43.4 per cent of all of those people?

4 A. That's right.

5 Q. Okay, and moving to the left. The 49, what does the 49 per cent

6 relate to?

7 A. This is the proportion or the fraction or per cent of Muslims,

8 IDPs and refugees of Muslim ethnicity in the population of Muslims

9 identified in the voters register 1997-8, which says how many of them in

10 the sample are displaced in 1997-8.

11 Q. So now that's looking only at all the displaced Muslims.

12 A. Right.

13 Q. And what you're finding or saying is that 49 per cent of all the

14 Muslims who lived in these municipalities were living somewhere else in

15 1997, 1998?

16 A. Yes. Yes, based on the sample voters register 1997-8.

17 Q. Okay. Taking it down to the municipality level, let's talk about

18 Capljina there is number 1?

19 A. Yes. The proportion is 96 per cent.

20 Q. Okay. 96 per cent of all the people who left were Muslims? I'm

21 sorry, no.

22 A. No, no. It is the Muslims identified in the voters register, all

23 of them, and 96 per cent are IDPs and refugees.

24 Q. Okay. So 96 per cent of --

25 A. It is the -- oh, sorry.

1 Q. 96 per cent of the Muslims who lived in Capljina before the

2 conflict weren't there in 1997, 1998?

3 A. No, not really. 96 per cent is that based on the sample of voters

4 we estimated the proportion of displaced and refugees of Muslim ethnicity.

5 It is 96 per cent of them are no more at their place of residence.

6 Almost everybody had become displaced.

7 Q. Okay?

8 A. Within this group. We are looking just now at one group.

9 Q. The Muslim group?

10 A. The Muslim group.

11 Q. And then moving one come up to the right --

12 A. Right.

13 Q. -- this is a 65 per cent figure. What does that signify?

14 A. 65 per cent, it is the proportion of Muslims among all IDPs and

15 refugees. So this is two different measures, very different measures.

16 Q. Okay.

17 A. We have an ethnic composition of all IDPs and refugees. Ethnic

18 composition tells us the ethnic structure within the IDPs and refugees.

19 This is the last column. For Capljina there are 65 per cent of Muslim

20 IDPs and refugees.

21 Q. So for all the people, Croats, Serb --

22 A. Yes.

23 Q. -- whatever who left Capljina --

24 A. Yes.

25 Q. -- 65 per cent of all of those were Muslims?

1 A. That's right, that's right.

2 Q. Okay. And then I don't know if you have it, the next page of your

3 report is page 44. This one relates to the Croat component of all of the

4 IDPs and refugees; is that correct?

5 A. Yes, it is.

6 Q. Okay. Okay. So then the numbers and the percentages here again

7 would be the same only as applied to Croats.

8 A. Yes, that's right.

9 Q. Okay. And then in your report you have additional tables for the

10 Serbs and for the others and for --

11 A. Yeah.

12 Q. And then you've actually got at page 46 a table that captures that

13 information for all the groups in a summary fashion. Do you have -- do

14 you see that?

15 A. Yes. Yes.

16 Q. So on page 46 of your report then just to connect the numbers,

17 again if you look at the second group which are the Muslim group on this.

18 A. Yes.

19 Q. I see the 49 per cent as being the percentage of Muslim -- the

20 Muslim population that moved?

21 A. Mm-hmm.

22 Q. And then the 43.4 being the Muslim component of all of the

23 displaced.

24 A. Very correct.

25 Q. Okay.

1 A. Yes.

2 Q. Okay. Now, Mr. President, for whatever assistance it may provide,

3 there is a map that attempts to capture this information as well, and it's

4 one to be careful with.

5 Ms. Tabeau, is this a map that tells us the ethnic composition of

6 the people who had left from these municipalities?

7 A. Yes. Ethnic majority of these people.

8 Q. So, again, say pointing to Vares because it's easy and right up

9 there by itself, this would indicate that Croats were the greatest number

10 of people to leave Vares?

11 A. Yes.

12 Q. And something less than 50 per cent of the Croats left?

13 A. Yes.

14 Q. Okay. And then -- sorry. Prozor, then, again indicating here

15 that greater than 50 per cent of the Muslims left Prozor.

16 A. Yes.

17 Q. Okay.

18 A. This was the largest group to leave Prozor.

19 Q. Okay. That was the largest group.

20 MR. STRINGER: And anyway, so the colours are self-explanatory,

21 Mr. President, but this is just an attempt to show graphically the

22 composition of those who left these municipalities.

23 THE WITNESS: Yes.

24 MR. STRINGER:

25 Q. Now, the next slide, Ms. Tabeau is one that gets a little bit into

1 the issues, some of the issues that have been raised in some of the

2 questions, and it relates to the composition of -- we've been talking

3 about the people who left; correct?

4 A. That's right.

5 Q. All right. Now, is there a different population that figures in

6 your report, that is, those who stayed or who were found in these

7 municipalities as of 1997, 1998?

8 A. Yes. This population is called non-DPs in the table that we are

9 now seeing on the screen.

10 Q. Tell us about these non-DPs. Do these match with the ones you've

11 been able to track from the 1991 census?

12 A. Yes, this is all the same group matched with the people with the

13 census 142.000 approximately, and as I have been repeating, one part of

14 this group are people who had become displaced or refugees, and another

15 group is those who simply stayed at their home municipalities. This table

16 shows the specific distribution into those who stayed, those who left, and

17 those who left are separated in two groups, IDPs and refugees.

18 Q. So that of the 142.000 who were matched initially, the non-DPs,

19 then, are people who we know were there both in 1991 and who were still

20 there as of 1997, 1998?

21 A. Yes, that's right.

22 Q. Okay. So that total number would be 80.717 persons?

23 A. Yes, that's right. Approximately 57 per cent.

24 Q. Okay you mentioned overlap and actually, I think it was

25 Judge Trechsel also. Is there overlap between this number of people still

1 present in 1997, 1998, and those 118.000 who were put in these

2 municipalities by the voters register?

3 A. Yeah. Yeah, thank you for the question. Of course it is 80.000

4 that are included in 118.000 in the population called "in". This is the

5 component. This is the 80.000 people who are as well included in the

6 population, 1991 population in of 118.000. That's the overlap of the two.

7 Q. Okay. Now, by the way, are these -- these are absolute minimum

8 numbers based on distinct matching.

9 A. That's right.

10 Q. Okay. No estimation.

11 A. No.

12 Q. All right. Moving down to the percentage then, percentages

13 according to your findings, what's the largest group of -- of persons who

14 stayed?

15 A. Well, it is the Croats, which is almost 75 per cent. It is 74.8

16 per cent of them stayed at the pre-war residence.

17 Q. Okay. And then 50 per cent of the Muslims stayed?

18 A. Yes, that's right.

19 Q. And again just to be clear, this table does not indicate movement

20 within the eight municipalities?

21 A. Well, partly perhaps it does, but generally it is a distinction

22 between those who stayed and those who moved.

23 Q. Okay. I am going to -- well, just to summarise then this next

24 slide, are these -- are these -- would you summarise the findings in

25 respect of Muslims, Croats, and non-Croats in the way indicated on this

1 slide?

2 A. It is actually repeating some statistics from the tables that we

3 have just seen. First statement says 43.2 per cent of all non-Croats, all

4 non-Croats who lived in 1991 in Herceg-Bosna, in the eight Herceg-Bosna

5 municipalities, were still displaced in 1997-8, and for Muslims this

6 proportion was 49 per cent and for Croats 25.2 per cent.

7 Q. Okay. Now, I've got a series of slides now that relate to

8 applying those percentages to the absolute minimum numbers and yielding

9 more complete absolute minimum numbers, but in view of the time and the

10 explanations that have already been given by you about the difference

11 between those two figures, I think I'm going to skip that and just go

12 straight to then the overall conclusions about this report so that we can

13 then move on to other reports.

14 So here's the slide. Can you just summarise for us the findings

15 as indicated in this report on refugees and IDPs?

16 A. Well, the first observation is that the scale of population

17 movements, especially the size of internal and external migration are

18 extraordinarily high and cannot be seen as consequences of usual

19 demographic or socio-economic forces. There had to be other factors that

20 caused this migration, this size of migration, within such a short period

21 of time.

22 Secondly, the Herceg-Bosna population in the eight municipalities

23 became divided very strongly along ethnic lines after the conflict.

24 Thirdly, 32 -- 43.2, sorry, of the pre-conflict population was not

25 living in their 1991 places of residence after the conflict. This is a

1 very high percentage.

2 Q. This is a percentage of non-Croats?

3 A. This is the percentage of non-Croats. And 49 per cent of the

4 pre-conflict population of Muslims was not living at their pre-war

5 residences. 49, almost 50 per cent of this ethnic group.

6 So to conclude, I would like to stress that certainly every ethnic

7 group suffered in this conflict and was affected by migration. That in

8 most cases was a forced migration. There is no doubt about it.

9 MR. KARNAVAS: Your Honour, I'm going to object at this point

10 unless there's a foundation. She is not a political scientist. She's

11 here as an expert on statistics and demographics. Now she's drawing legal

12 conclusions. This is well beyond the scope of her expertise.

13 MR. STRINGER: Mr. President, I think I could clear that up with a

14 question. I'm happy to let the witness address that -- that point,

15 because it's in the report, and so if I could just propose to do that.

16 MR. KARNAVAS: It don't mean because it's in the report that it

17 should be in the report. That's the whole -- that's the whole point of

18 this objection, that she is not here -- you know, it's bad enough that

19 she's used technical terms, terms of art, displaced persons, refugees,

20 into the report in order to give it the added umph, speak of polemics, but

21 it seems to me that now she's opining as to what happened in the field,

22 and if that is the case, Your Honours, I will definitely go into certain

23 issues that at least in the Simic case she indicated she was unqualified

24 to testify about.

25 JUDGE ANTONETTI: [Interpretation] The Chamber will take that into

1 account in the light of what you said, because apparently in the Simic

2 case the witness said that she was not able to deal with these matters. I

3 don't know.

4 Mr. Stringer, there are conclusions in that report. You wanted

5 the witness to explain where these figures came from or --

6 JUDGE TRECHSEL: We are here again at the spot where we have been

7 twice today, namely the explanation for movements, and I have indicated

8 that there are movements all the time. Everywhere people change places.

9 And you ventured to give a reason here when you speak of violent forces.

10 I think that is indeed something which you cannot say as a -- as a result

11 of your figures. But what you can say is whether normal, general figures

12 of population movement are known, whether there are norms, as to what

13 percentage of a given population moves within a time lapse of six to seven

14 years.

15 MR. KARNAVAS: Your Honour, with all due respect and I understand

16 that you're entitled to an answer to that particular question.

17 JUDGE TRECHSEL: Yes.

18 MR. KARNAVAS: But if I may, with all due respect --

19 JUDGE TRECHSEL: I think I am entitled to the answer as you

20 correctly state and you speak afterwards, I would suggest.

21 MR. KARNAVAS: Very well. I would like the record to reflect

22 there is no foundation for that, for the basis of that question, but very

23 well, I will.

24 THE WITNESS: Thank you for the question, Your Honours. Well, my

25 main -- first main conclusion is related to the scale of this migration.

1 I clearly say that scale is extraordinary. It is such large scale of

2 migration that it cannot be seen as to be explained by the usual factors.

3 Well, in case of migration in normal situation when there is no conflict

4 we think of economic migration, educational migration, other types of

5 migration, but it never comes to 40 -- more than 40 of the population that

6 goes away within a short period of seven, eight years. That is not

7 possible, simply speaking.

8 So the first reason is the scale.

9 Secondly, I did indeed not study determinants of this movement in

10 this particular report but I am aware of additional materials related to

11 conflict in Herceg-Bosna. I had been studying reports by international

12 observers. I had been studying books written by journalists who were also

13 present in the conflict area during the conflict as observers, war

14 reporters. I saw videos showing scenes from the conflict period, and I

15 think there is very little doubt that there were violent sources active in

16 this period.

17 JUDGE ANTONETTI: [No interpretation]

18 THE INTERPRETER: Testing. Can you hear the interpretation into

19 English?

20 JUDGE ANTONETTI: [Interpretation] Okay. You can hear me now.

21 Let's take a demographer. Let's take an expert demographer. When

22 highlighting statistics, figures, that's what you do in your conclusions.

23 You say that about 43.2 per cent of the non-Croat populations in eight

24 Herceg-Bosna municipalities was -- was not on -- or in the location where

25 they resided before when they resided in 1991, and you say that on the

1 basis of figures from 1997, 1998. When as a demographer you reach that

2 conclusion, can you and should you give the reasons for these changes, the

3 causes?

4 THE WITNESS: No, I cannot. I have no sources, and I didn't use

5 any sources to quantify the causes of these movements.

6 JUDGE ANTONETTI: [Interpretation] I am putting that question to

7 you because some demographers, for example, studied migrations in cities

8 or towns that were partly reconstructed, and you might reach conclusions

9 on population transfers. Therefore, thanks to a demographic study you may

10 say that this phenomenon is due to the reconstruction of the society.

11 Let's take Beijing in China because of the preparations for the Olympic

12 Games populations are moving around and a demographer might say that the

13 cause for this phenomenon in Beijing is the Olympic Games. But when we

14 are talking about the work you conducted, you carried out, are you in a

15 position to come to any conclusions regarding the causes of these

16 phenomenon?

17 A. When this comes to quantification and use of the actual data on

18 determinants on factors that caused these movements I cannot. But I am

19 aware of sources, serious sources that indicate that the movements were

20 not voluntary. For instance, the database that I also included in this

21 study, internally displaced persons and refugees, the registration of

22 internally displaced persons and refugees in this report, maintained by

23 the government of Bosnia and Herzegovina and which was started by the

24 UNHCR. This is a very serious source. It contains 600 - almost -

25 thousand records of internally displaced persons and refugees as of the

1 year 2.000. This is the huge number for the entire Bosnia, of course.

2 And there were many more. There are many more in the earlier

3 years and this is a legal database. This is a legal database. This means

4 that these people who are registered there have the legal status of

5 internally displaced persons and refugees. So that is one good reason I

6 can use for my statement about the forced migration.

7 Second reason is I am aware of transit visa database that was

8 established as a registration system in Croatia by the Croatian

9 authorities for people who were transferred from Bosnia through Croatia to

10 other countries.

11 MR. KARNAVAS: Your Honour, now we're going beyond the scope of

12 the -- of the expert report. Now, I don't object to her testifying in

13 these matters. If that is the case, we're going to need an additional day

14 or two for this particular witness because now you're allowing this

15 witness to go out of the scope.

16 Now, in the Simic case, page 10805, she is asked a question about

17 certain international agreements at which point a member of the

18 Prosecution stands up and says: "I object to that question. This witness

19 is not a political expert." Goes on and then says: "She is a

20 demographer, and it is unfair on her to ask her questions unless her

21 expertise in the area has been established."

22 Now, her expertise into these other matters has not been

23 established. We do know she has written an article where she does

24 indicate after now saying on the record under oath that she is not

25 qualified to talk about these issues because on page 10806 she says: "I

1 know many things from different areas but I really don't feel qualified to

2 say anything about these types of issues."

3 Here she says in one of the articles which she's going to be

4 questioned on it's on page 188 of an article she wrote, "War-related death

5 1992-1995," it's referenced by the Prosecution, she says on page 188 she

6 says: "Bosnian Serbs and later also Bosnian Croats fought (often through

7 ethnic cleansing and terror campaigns) to take and control territories

8 that otherwise would be subject to the rule of Bosnian Muslims from

9 Sarajevo. The Muslims fought for these territories as they believe they

10 did not have much choice."

11 Now, if that is not a political statement coming from someone who

12 claims not to be an expert in politics but a demographer, a science that

13 deals with -- that deals with logic, that deals with statistics, that

14 deals with mathematics. She's not here as a politician. She is not here

15 as a political expert. She is not here as a politician. She is here as a

16 demographer and that's why Mr. President and that's why Judge Trechsel I

17 stress over and over again you may call it polemics that's why I stress

18 that she is an employee of the Office of the Prosecutor, making a fairly

19 good living as an employee of the United Nations.

20 MR. STRINGER: Objection to that last statement. It's

21 inappropriate.

22 MR. KARNAVAS: I mention that because -- I mention that because

23 this part, this part of her testimony is clearly beyond the scope and is

24 part of the Prosecution's theory of the case. That's why.

25 JUDGE ANTONETTI: [Interpretation] Very well. Mr. Stringer, you

1 will have the floor in a while but I believe that Judge Trechsel wanted to

2 say something.

3 JUDGE TRECHSEL: I ask for the floor internally before

4 Mr. Karnavas rose. I am hundred per cent in agreement with what you said,

5 Mr. Karnavas.

6 And I want to tell the witness that you were called as an expert

7 in demography. That is in dealing with figures, relating figures and

8 that's it. What you have added is your personal opinion as a citizen

9 which you entitled to have but which we cannot hear as evidence because

10 they are not part of your scientific knowledge. Therefore, the Defence

11 counsel was right on this.

12 What I wanted to ask you is whether you have, whether your science

13 has any figures, figures, on what a "normal" migration is in a given area

14 and over a period of six, seven years.

15 THE WITNESS: Not that I have these kind of figures at hand right

16 now, but I believe figures on normal migration in the region of the former

17 Yugoslavia do exist, and for reference purposes, I certainly can look for

18 these figures and I can provide this Chamber with these figures for

19 comparison.

20 JUDGE ANTONETTI: [Interpretation] Mr. Stringer, I will give you

21 the floor in a few seconds, but the best would be to conclude your

22 questions on this portion and next week you may start asking questions

23 about this other topic. So you may perhaps finish today on this topic and

24 then you may continue next week.

25 MR. STRINGER:

1 Q. Actually, I think the only question I've got on this and it goes

2 to something Judge Trechsel asked about earlier an I -- if I could direct

3 you to your report, Ms. Tabeau, page 10, the bottom paragraph, which I

4 think gets to some of what has been raised.

5 In this paragraph you say that it needs to be noted that internal

6 migration in the former socialist countries such as Yugoslavia, and in

7 particular Bosnia-Herzegovina, was limited in the years until 1991. And

8 then you make your reference to analysis of the difference in places, but

9 you talk about pre-conflict internal migration as being negligible, prior

10 to the conflict that is, and you make reference to the usual causes of

11 internal migration, labour market, housing, et cetera, those did not

12 operate during the conflict.

13 So just keying off that paragraph and, you know, bringing it back

14 to the point one of -- of the graphic which is this: The point being that

15 the facts were incomparable with those of usual demographic or

16 socio-economic factors. Now, as a demographer perhaps you can address the

17 issue of, as Judge Trechsel mentioned earlier, what would have been normal

18 migration patterns in this country or in any country as compared to the

19 migration patterns that you've seen in this study?

20 A. Well, as I said, I will look for the figures that are relevant to

21 this issue, and I will bring the figures. Let's talk about figures.

22 Q. But just in general terms, can you address what would be normal

23 factors and --

24 A. Yes. Oh, factors.

25 Q. -- as you mention in your report here?

1 A. Well, as you mentioned already, this is work migration; this is

2 housing; this is education. These are the normal causes that operate in

3 normal situations where there are no other factors. All people want to

4 have good education, a good house, and a better life, and these are the

5 major factors that drive people to move from one place to another.

6 Q. Okay. Now, based on your work here, based on your work for nine

7 years in the Dutch government here as a demographer, do those normal

8 factors, economics, housing, et cetera, do those sorts of factors result

9 in the scale of population movement that one finds in your report here?

10 A. Certainly not. And in official statistics the conflict-related

11 migration is largest component of all migratory movements. It's not

12 accidental that we have had a lot of discussions of the issues of refugees

13 in the Western European countries recently. There were several large

14 waves of refugees arriving in Western Europe from countries affected by

15 conflict, Bosnia and Herzegovina being just one of them. So this group of

16 migrants is the most serious issue because of the size and rapidity of the

17 arrivals.

18 So we certainly speak here of not the usual migration. It is

19 different type of migration. There is no doubt about it.

20 Q. Okay.

21 MR. STRINGER: Mr. President, that's all I have on this particular

22 report.

23 JUDGE ANTONETTI: [Interpretation] Regarding the oral request made

24 by Mr. Karnavas as to the additional five days, the Chamber will grant a

25 request to Mr. Karnavas. You will get the delay that you requested.

1 Madam, you will come back next week. We will continue, and we

2 will talk about the second report, and we will start with Mr. Stringer and

3 then we'll begin the cross-examination, and you don't have time you will

4 also have to come back another time, but because you are here in this

5 building I don't suppose that this will be a problem for you.

6 We have another witness for two days next week, so the first part

7 will be examination-in-chief, cross-examination, and for this witness --

8 by the Defence.

9 This is all I wanted to tell you. Thank you very much, and we

10 will see each other next Monday.

11 --- Whereupon the hearing adjourned at 6.49 p.m.,

12 to be reconvened on Monday, the 27th day

13 of August, 2007, at 2.15 p.m.

14

15

16

17

18

19

20

21

22

23

24

25