090505ED

Page 5783

1 WITNESS: EWA TABEAU

2 Examination by Ms. Carter:

3 MS. CARTER: Thank you, Your Honours.

4 Q. Dr. Tabeau, will you please state your name for the record.

5 A. I'm Ewa Tabeau.

6 Q. What is your area of expertise?

7 A. It is demography, and in particular demography of conflict.

8 Q. Before we begin the substance of your testimony I would like to

9 go through a few predicate matters. Can you please describe the field of

10 demography?

11 A. Demography is a research area that focuses on the population, on

12 the -- describes the population, changes in the population, explains the

13 trends and patterns in demographic phenomena. Among the phenomena there

14 are deaths, there are births, marriages, divorces, and migration. These

15 are the basic demography processes.

16 Q. And you noted for the Court that your specific area of expertise

17 is demography in relation to conflicts. Is that correct?

18 A. Yes, it's correct.

19 Q. And how does that -- how is that described?

20 A. Demography of conflict is a subarea of demography. The main

21 subject of demography of conflict is the demographic consequences of

22 conflict, but in -- within the area also other subjects are studied, like

23 demographic causes of conflict and the similar.

24 Q. Okay. We do have your resume through the time of your Galic

25 report. However, can you please the education and training that you have

Page 5784

1 developed regarding this expertise?

2 A. I studied statistics and econometrics at the Warsaw Economic

3 University in the 1970s and graduated with the highest grade in

4 statistics and econometrics with a Master degree. Later I engaged in a

5 Ph.D. programme in demography, which I completed successfully, in the

6 area of mathematical demography, also with the highest grade, also from

7 the same university, Economic University of Poland. I worked a number of

8 years at the same university where I taught statistics and demography to

9 undergraduate courses. Later I moved to the Netherlands and became a

10 researcher and later a senior researcher in the Netherlands

11 Interdisciplinary Demographic Institute, which the National Dutch

12 Institute for Demography. And finally, in the year 2000 I joined this

13 Tribunal, the Office of the Prosecutor where the became the head of the

14 Demographic Unit.

15 Q. During your time as an expert have you had the opportunity to

16 testify?

17 A. Yes, I testified several times -- I think 13 or 14 times before

18 the ICTY Trial Chambers.

19 Q. Okay. Can you name for us which cases you have testified in?

20 A. I testified in several cases related to Sarajevo. In the

21 General Galic case, Dragomir Milosevic case. I testified in

22 Slobodan Milosevic, also on Sarajevo but as well on population movements

23 internally displaced persons and refugees. I testified in Prlic which is

24 the Herceg-Bosnia conflict. I testified in Popovic which is Srebrenica.

25 I testified in Seselj and a number of other cases; Lukic and Lukic, for

Page 5785

1 instance, recently; and Vasiljevic, a number of years ago; in Simic et

2 al, which is Bosanska Posavina.

3 Q. During your time as an expert have you always testified for the

4 Prosecution or have you had the opportunity to testify for the Defence?

5 A. I testified one time for the Defence. It was in Lukic and Lukic

6 case, recently this year.

7 Q. Thank you. As you are aware -- I have been advised to please

8 slow down so --

9 A. I'm sorry.

10 Q. Actually it was for me.

11 As you were aware, the reports you compiled regarding the siege

12 of Sarajevo for the Galic and the Slobodan Milosevic cases with the

13 corresponding addendums have been admitted as P2325, through P2331.

14 In today's testimony, I want to focus on the killed and the

15 wounded in the Perisic indictment period, beginning on 23 August 1994.

16 I'd like to begin with the killed.

17 In P2325, your report from the Galic case, at Table A3.13 at

18 e-court page 69, as well as P2331, your report in the Slobodan Milosevic

19 case, at Tables A3 and A4, which is e-court pages 9, you give a monthly

20 break down of the killed soldiers and civilians.

21 These numbers and the totals given do not fully correspond with

22 each other during the overlapping period of September 1993 through

23 August 1994. Can you please explain to the Chamber why these numbers are

24 different?

25 JUDGE MOLOTO: That explanation for me doesn't -- is not helpful

Page 5786

1 unless I know exactly where we are Madam Carter. You just started

2 talking about the Galic and Slobodan Milosevic cases, and I'm not sure --

3 and their addendums. Can you refer us to what you are looking at so that

4 if we do have it -- is it in the report of the witness?

5 MS. CARTER: Certainly, Your Honour. The Galic report is P2325

6 and I refer to Table A3.13, and that's at page 69 of the report.

7 JUDGE MOLOTO: Well --

8 MS. CARTER: And it is currently on the screen. And, if possible

9 can both pages be brought up, meaning, P2331, at page 9, next to it.

10 JUDGE MOLOTO: Thank you so much, Madam Carter.

11 Now, what doesn't correspond.

12 THE WITNESS: If I may have one comment, the second table on the

13 right -- yes, that should be replaced by a different one. Yes. This is

14 what we need for our discussion.

15 JUDGE MOLOTO: Thank you so much.

16 THE WITNESS: We have two pages from two different expert

17 reports, both related to Sarajevo. On the left, Table A3.13, this data

18 comes from the Galic report. It is a simple time distribution of killed

19 persons, civilians and soldiers, and individuals with a non-status, shown

20 for the period from September 1992 until August 1995 -- 1994. It's a

21 monthly distribution, so the numbers in the table are absolute numbers of

22 killed persons by month in the mentioned period.

23 This period, September 1992 to August 1994, yes, the period of

24 the Galic report, which is different from the indictment period of this

25 case, and as well, of the period which was discussed in the Sarajevo

Page 5787

1 report that I made with my colleagues for the Slobodan Milosevic case.

2 So, the other page, on our screen, on the right-hand side, comes

3 from the Slobodan Milosevic report. The period of analysis is far

4 broader in this report than in the Galic report. In the

5 Slobodan Milosevic report, we were interested in the casualties of the

6 entire conflict period, starting in April 1992 until November 1995. In

7 the Table A3 which is made for soldiers and A4 which is for civilians the

8 period of analysis is even broader. It begins in January 1992 and ends

9 in December 1995.

10 There is even one more category, unknown month of death included

11 which is the first row in each of these tables.

12 The data shown in the tables A3 and 4 is of the same type as the

13 data shown for the Galic report. It is monthly distribution, monthly

14 distribution of killed persons separately shown here for soldiers and for

15 civilians. Of course, the numbers for civilians are our first interest,

16 so we will be focussing on table A4 from the right side of the screen,

17 and the first column numbers for civilians, from the other table, A3.13.

18 These numbers are not identical, at least we can compare in the

19 period that we can directly compare. There is an overlap between these

20 two periods, and the numbers can be for this period compared.

21 I also prepared a small extract of these data from these two

22 tables for the period relevant for the Perisic case. The period that is

23 12 months that are the same in the Galic report, in Slobodan Milosevic

24 report and in the indictment of Perisic case. This period begins in

25 September 1993, and lasts until August 1994. I have made a separate

Page 5788

1 table with exactly the same figures and I'm not sure that we can use this

2 table but if we may use this table, then I -- I would like to rather use

3 this because it is easier to see the differences in this smaller sample

4 of data.

5 MS. CARTER: Your Honour, we were provided with this extract as

6 of yesterday. I did provide it to Defence counsel at that time. It is

7 not on our 65 ter list, so I'm not certain if it would be appropriate to

8 use before us. It is also not on e-court. We can use the ELMO system if

9 the Court would like to review this item and with leave of Defence

10 counsel.

11 JUDGE MOLOTO: Mr. Lukic.

12 MR. LUKIC: [Interpretation] Your Honour, it is true that we

13 received this yesterday afternoon. After an information -- informative

14 conversation with Ms. Tabeau, we received several new documents or

15 excerpts from these new documents. I don't have any objection to us

16 seeing them now, but I would like to reserve the right to consult with my

17 client once these documents are translated into B/C/S, because right now

18 they're only in English, and I -- for those reasons I would suggest that

19 we only marked for identification so that I can give our position later

20 on.

21 What I have requested yesterday from Ms. Carter, and she

22 confirmed this, was whether these documents were based on the two earlier

23 reports exclusively and she confirmed that. So if that is -- if it is

24 clear, if these documents are clearer than the ones that we used in the

25 earlier cases, I don't object to them being used here but I would like to

Page 5789

1 be able to review them in detail, but I do agree they be used now and

2 marked for identification for those purposes.

3 JUDGE MOLOTO: You can use them and we mark them for

4 identification, Madam Carter.

5 MS. CARTER: Thank you, Your Honour. And we --

6 JUDGE MOLOTO: Madam Carter, you do confirm that they're based on

7 the earlier reports, exclusively.

8 MS. CARTER: That is the advice I was given by Dr. Tabeau. I can

9 confirm again.

10 Q. Dr. Tabeau, the graphs that you provided to me as of yesterday,

11 what is the provenance of those items?

12 A. These is a sample of data and results from the two reports that

13 we are discussing now: One being the Galic report on Sarajevo; and one

14 being Slobodan Milosevic report on Sarajevo.

15 Q. Okay.

16 MS. CARTER: And with the assistance of the usher, we do have the

17 report to be placed on the ELMO.

18 Q. Ms. Tabeau, on first glance I do note there are some markings,

19 specifically a straight line and a circle, next to the civilians on this

20 chart that is not present in the document that you provided to me or to

21 the Defence. Who made these markings?

22 A. I have made them just a second ago.

23 Q. Okay. Thank you. You were going to advise the Court as to

24 what -- what created the difference in the numbers of the civilians in

25 these two reports.

Page 5790

1 A. Now, first of all, I would like to say that there is a difference

2 in the number of killed persons, both civilians and soldiers, between

3 these two reports, and I'm speaking of the period that is overlapping

4 between Galic report and Slobodan Milosevic report and is part of the

5 indictment, Perisic indictment. This is a period, once again, of

6 12 months, beginning as shown in this table in September 1993 until

7 August 1994.

8 So these are 12 months that can be directly compared and for each

9 month, the number of civilians and of soldiers can be seen as in this

10 table next to each other.

11 Generally, the number of civilians killed within these 12 months

12 is higher in the Slobodan Milosevic report. It is 548, as opposed to the

13 lower number from Galic report, which is 370.

14 These are the two numbers I have just mentioned, 548 from

15 Slobodan Milosevic, and 370 from Galic report.

16 There is also a difference in the number of soldiers. The

17 difference is less significant. There are 499 killed soldiers reported

18 in the Galic report, and 472 in the Slobodan Milosevic report.

19 So what is the source of these differences? The reports, first

20 of all, were made based on different sources. The Galic report used only

21 one source, Sarajevo Household Survey, that was a special survey run

22 during the conflict by mid-1994 by the research institute in Sarajevo,

23 and we used this one source to compile all statistics in the Galic

24 report.

25 The source is very large. It was a Household Survey that was

Page 5791

1 meant to be complete to cover all households living in Sarajevo by

2 mid-1994. All together 85.000 households were interviewed in this

3 survey. The 85.000 households represented the population of

4 approximately 340.000 individuals. We used this source which we

5 computerised, not all of it, a part of it, by the OTP and established a

6 database, a database that included records of killings, woundings, and

7 natural deaths which we studied in the Galic report. This is a good

8 source but just one source.

9 And when it comes to sources when studying demographic

10 consequences of war, we must be aware of the fact that there is no source

11 which is complete and which is perfect, doesn't have any deficiencies so

12 is also the Household Survey. It was meant to be complete but most

13 certainly incomplete and there are certain deficiencies like

14 incompleteness in data that actually made us exclude certain records from

15 our analysis.

16 In order to improve this situation, in the Slobodan Milosevic

17 report that was made later we have applied a different approach. We

18 worked with a number of sources, big sources, including the Sarajevo

19 Household Survey, which were merged with each other, duplicates were

20 eliminated, overlap was eliminated and one master list of casualties was

21 established. This could only be done for the killed persons, because

22 sources covering injuries and wounding of the population are very

23 infrequent. Sarajevo Household Survey is just one of the few sources

24 that we have in our archive that could be used for this analysis.

25 So in Slobodan Milosevic we worked with a number of sources. It

Page 5792

1 was Household Survey Sarajevo. It was a statistical database on wartime

2 mortality, a big one, established at our request, at the OTP request by

3 the statistical authority of the Federation of Bosnia and Herzegovina.

4 In this database there are 20 -- approximately 25.000 war-related deaths

5 reported of both civilians and soldiers for the entire country, Bosnia

6 and Herzegovina, and entire conflict period, starting April 1992 until

7 December 1995.

8 Of course, only a small sample of this data is related to

9 Sarajevo. As far as I remember, about 15.000 records in this database

10 relate to Sarajevo. In addition to this databases, we use ICRC List of

11 Missing Persons --

12 JUDGE MOLOTO: Could you please slow down and --

13 THE WITNESS: I'm sorry.

14 JUDGE MOLOTO: It is very difficult for the court reporter to

15 catch up with you. I'm not quite sure because you're talking in English

16 I suspect the interpreters might also be having difficulties, those who

17 are interpreting into other languages. I don't hear any interpretation,

18 I hear you only, but I think -- will you please just slow down.

19 THE WITNESS: I'm sorry. Of course, I will.

20 I was saying, in Slobodan Milosevic report a number of sources

21 were used together, combined with each other and gleaned from

22 deficiencies. Among these sources there was Sarajevo Household Survey.

23 There was the federal authority, statistical authority database on

24 wartime mortality in Bosnia. There was ICRC list of missing persons for

25 Bosnia and Herzegovina. There a database compiled by an NGO in Bosnia,

Page 5793

1 Muslims Against Genocide. Later this database evolved and became the

2 Bosnian Book of Death, which is quite known in Bosnia as the largest

3 existing database.

4 MS. CARTER: If I may briefly interrupt. From looking at the

5 court reporter, I think we're still having difficulty with speed. If you

6 it slow down a bit further.

7 THE WITNESS: Right. I'm very sorry.

8 So Muslims Against Genocide is an NGO in Bosnia and Herzegovina

9 which at some point was collecting information about victims of conflict,

10 of civilians and soldiers, and they combined this information in a

11 database on war-related casualties and established a database called

12 Muslims Against Genocide database which we used in this project.

13 So all together, a number of large sources. So sources is the

14 major difference between the statistics compiled in Galic and

15 Slobodan Milosevic report. But not the only difference.

16 JUDGE MOLOTO: I just want to ask you a question.

17 Do I understand you to be saying, in fact, that your report is

18 the result of you getting information from data that was built up by

19 other people and other organisations. It's not entirely your own work

20 product.

21 THE WITNESS: Any report I made is based on sources made by

22 others. So is the Galic report so is the Slobodan Milosevic report. It

23 is -- we are speaking of large numbers of victims of this war, and large

24 number of people worked down there in Bosnia and Herzegovina to put

25 together record of information about these victims.

Page 5794

1 JUDGE MOLOTO: I'm not asking for an explanation.

2 THE WITNESS: Yes.

3 JUDGE MOLOTO: I'm just asking whether that is the fact.

4 THE WITNESS: Yes, it's the fact.

5 JUDGE MOLOTO: Thank you.

6 THE WITNESS: Yes.

7 JUDGE MOLOTO: I'm sure Madam Carter will get to the point but I

8 can't wait to hear you tell me about information relating to deaths of

9 civilians and soldiers, and wounded of civilians and soldiers that

10 relates to the period 26 August 1993 to the end of the war.

11 THE WITNESS: I am not sure that I understood the question. Is

12 the question about --

13 JUDGE MOLOTO: The question is: I would like to hear what you

14 extrapolate from these reports. I understand that these reports were not

15 drafted for this case. They were drafted for Galic and

16 Slobodan Milosevic.

17 Now, to the extent that they may cover the period covered by the

18 Perisic case, I'm expecting that you're going extrapolate from them and

19 tell us what the figures are relating to the indictment period in the

20 Perisic case. That's what I'm interested in.

21 THE WITNESS: Yes, I understand. I suggest we will look at first

22 page from this set of paper that was just introduced this morning. I

23 have included a little table on the first page and this is a table that

24 summarises the number of casualties for the entire indictment period of

25 the Perisic case. That would be -- I could not include the data from the

Page 5795

1 26th of August but the first month next after it is September 1993, until

2 the end of November 1995.

3 So the overall number of victims is 1.782 and this includes both

4 civilians and soldiers. These are killed persons. The number of killed

5 civilians is 949 and soldiers 833. These numbers come from the

6 Slobodan Milosevic report that I started explaining in terms of sources.

7 This one report actually can be used to extract the relevant numbers for

8 the Perisic case.

9 So that's the number, that's the major information that can be

10 extracted from Slobodan Milosevic report.

11 JUDGE MOLOTO: Do you have any similar information regarding the

12 wounded?

13 THE WITNESS: Unfortunately, there is no information about

14 wounding in the Slobodan Milosevic report. However, this can be

15 extrapolated using the information from Galic report. Based on these

16 12 months overlapping between these two reports, it is possible to

17 estimate the relationship between killed and wounded civilians and the

18 relationship is some kind of ratio 1:3.2, that means per one civilian

19 3.2 -- per one killed civilian 3.2 civilians were wounded. So this ratio

20 can be applied and the general number of wounded persons can be obtained

21 for the Perisic period as well. Applying the numbers that we see on the

22 screen. That would be a roughly three times more wounded persons --

23 wounded civilians compared with the killed civilians.

24 JUDGE MOLOTO: Yes. But I did think that a little earlier you

25 showed us a table that you referred to as absolute numbers but not

Page 5796

1 estimates.

2 THE WITNESS: Yes. The table is absolute numbers and these are

3 counts, that means numbers that can be documented by records, individual

4 records of information about persons. So these are not estimates. And

5 these are minimum numbers, the lowest numbers we have to think of. When

6 it comes to estimation, there is some estimated numbers in the Galic

7 report as well. This ratio I mentioned is not an estimate. It is a

8 ratio that is obtained from the count, from the real data. So the ratio

9 is not an estimate it is the quantification of a relationship --

10 JUDGE MOLOTO: From absolute figures?

11 THE WITNESS: Yes, from absolute figures.

12 JUDGE MOLOTO: Okay.

13 THE WITNESS: But as we know, in statistics once a measure is

14 estimated from the actual data, there can be an extrapolation and the

15 same measure can be applied to a bigger population for which we don't

16 know the same thing. That means I'm speaking here about wounded persons.

17 If there are, of course, good reasons to apply these measures and I think

18 there are good reasons to apply the measure of wounded persons to overall

19 killed number of civilians and obtain an estimated number of civilians

20 for the Perisic period.

21 JUDGE MOLOTO: Yes, Madam Carter.

22 MS. CARTER: Thank you, Your Honour.

23 For clarification of the record, if we can re-call P2325 at

24 page 69. That's Table A3.13. It is already up on the e-court. Thank

25 you.

Page 5797

1 I believe that Your Honour was referring to the chart to the

2 left-hand side and you indicated that we did have some raw data in

3 relation to the wounded civilians -- I'm sorry. I apologise. Actually

4 this is actually for the killed civilians. The wounded civilians would

5 say be A4.13.

6 JUDGE MOLOTO: [Microphone not activated] ... or A3.13.

7 MS. CARTER: It's Table A4.13, document ID page 02194824.

8 THE WITNESS: Excuse me? May I have your help, please? Can you

9 help me, please?

10 Thank you.

11 MS. CARTER:

12 Q. Dr. Tabeau, what we now see before us at Table A4.13 purports to

13 the monthly number of persons wounded within the front lines of Sarajevo

14 and you've compiled the same monthly breakdown. Did you use the same

15 method that you were describing to the Court previously in order to

16 derive these numbers?

17 A. Yes. These the are the same -- these are also counts of wounded

18 persons. The table comes from the Galic report, so from this table we

19 unfortunately cannot have a figure that would be covering the indictment

20 period of the Perisic case. From this table we can only take the data

21 for one year, beginning in September 1993 until August 1994.

22 Q. Okay. When you indicated that there was a ratio of 3.2, 1:3.2,

23 how did you arrive at that figure?

24 A. This figure was obtained using data on civilians from the Galic

25 report, only from the 12 months that overlap with the Perisic indictment

Page 5798

1 period, which is the first year basically of the indictment period, and

2 the overall total of killed civilians was compared with the overall total

3 of wounded persons. It is basically dividing the wounded by the killed.

4 Q. However, you previously indicated that based on that one year,

5 you can extrapolate forward past August 1994; is that correct?

6 A. Yes. It is possible. That -- what can be -- this can be done,

7 indeed, yes.

8 Q. Okay. Is it a normal demographic tool in order to complete these

9 extrapolations?

10 A. Demographic extrapolation can be done in many ways, extrapolating

11 this kind of ratio is one way of doing this. There are more

12 sophisticated tools that can be used for this kind of extrapolation. In

13 Dragomir Milosevic report, which unfortunately is not discussed today,

14 another type of extrapolation was presented from which the monthly totals

15 of wounded persons for the entire siege period were obtained by

16 comparison of two sources reporting the -- on the wounding.

17 So a statistical model, more complex model than a ratio was used

18 in this kind of extrapolation.

19 Q. When using that secondary type of operation did you obtain

20 similar numbers to the numbers that you're testifying here today,

21 specifically 3.2 times per every 1 killed?

22 A. Well, I don't know for this particular period because I don't

23 have it at hand and I haven't made it. But generally, this model-based

24 extrapolation is even more sophisticated and detailed. That means not

25 only the relationship, the general relationship as the ratio 1:3.2 is

Page 5799

1 contained in this extrapolation but also the monthly distribution or even

2 weekly distribution can be produced using the model.

3 So generally I would think the answer is yes, these kind of

4 ratios are preserved because these are the most general measures, but the

5 more sophisticated the model, the more sophisticated the output of the

6 model.

7 Q. But as to a bottom line, as you're sitting here today with the

8 information that you have available, specifically the -- I'm sorry,

9 specifically the Galic report and the Slobodan Milosevic report, can you

10 testify before this Court today that we would be able to use a 3.2 ratio

11 in order to determine the wounded post-August 1994?

12 A. Yes, I would think yes. Yes.

13 Q. Okay. Thank you.

14 JUDGE MOLOTO: Yes, Mr. Lukic.

15 MR. LUKIC: [Interpretation] I didn't want to interrupt,

16 Your Honours, but now that I saw on the screen a technical problem with

17 the translation of this document, this is an official translation made by

18 CLSS which was used already in the Galic case. I'll read it out what it

19 says in B/C/S. It says: A 413, number of persons killed monthly within

20 the front lines in Sarajevo.

21 Obviously this is meant to be a translation of the English

22 version. Unfortunately, in preparing for the examination, I used the

23 English version. I'll try to make use of this fact that I established

24 all the figures are the same, but the heading of the table is completely

25 different.

Page 5800

1 JUDGE MOLOTO: Yes, Madam Carter.

2 MS. CARTER: My B/C/S is terribly poor. I would not be able to

3 verify or to deny that that fact is accurate. We would ask if Defence

4 counsel could provide for us if there other corrections that need to be

5 made, we can certainly ask for this to be retranslated. But the report

6 was originally drafted in the English and that would be the controlling

7 language of it. But, certainly, we would want to have an accurate record

8 in the B/C/S.

9 JUDGE MOLOTO: Well, your colleague has indicated therefore that

10 this -- the B/C/S indicates this is the dead [Realtime transcript read in

11 error "date"] instead of the wounded, if you can take that up with your

12 translators, please.

13 MS. CARTER: Certainly, Your Honour. I will do that at the first

14 opportunity.

15 JUDGE MOLOTO: Thank you.

16 MS. CARTER:

17 Q. Ms. Tabeau, given the questioning of the Court as well as the

18 answers that you have provided to me, I have no further questions of you

19 at this time.

20 MS. CARTER: And I pass the witness.

21 JUDGE MOLOTO: Can I just correct the transcript. I don't think

22 this may be corrected later. I said it indicates the dead instead of the

23 wounded, not the date of the wounded. Okay?

24 Mr. Lukic.

25 MR. LUKIC: If you can put this ELMO ...

Page 5801

1 Cross-examination by Mr. Lukic:

2 Q. [Interpretation] Good morning, Ms. Tabeau.

3 A. Good morning.

4 MS. CARTER: I do apologise for interrupting Mr. Lukic. Given

5 the fact that we are now going back to the item on the ELMO, I would ask

6 that it be tendered as an IC exhibit and marked for identification

7 awaiting translation.

8 MR. LUKIC: [Interpretation] I agree, Your Honours.

9 JUDGE MOLOTO: What is an IC exhibit, Madam Carter?

10 MS. CARTER: A court exhibit Your Honour. An in-court exhibit.

11 I apologise, I may use a term that's not ...

12 JUDGE MOLOTO: Okay. The document -- now, this is the document

13 that is on the screen here? According to my e-court, I have got this one

14 that has got 5090 at the end, which has got -- which is mistranslated,

15 but then there was the previous one. What are you actually wanting

16 admitted, Madam Carter?

17 MS. CARTER: Your Honour, we're asking for the charts that are on

18 the ELMO not the item that's on the e-court. I believe it is being

19 provided to you now.

20 JUDGE MOLOTO: The graphs. Okay. Okay. Fine. These documents

21 are admitted into evidence. May it please be given an exhibit number and

22 marked for identification.

23 THE REGISTRAR: Your Honours, that will be Exhibit P2332, marked

24 for identification.

25 JUDGE MOLOTO: Okay.

Page 5802

1 MR. LUKIC: [Interpretation] Let me just check with Ms. Carter.

2 These tables that were now marked for identification do not include the

3 additional informative report that we got, the interview made with

4 Ms. Tabeau. You are just tendering the tables, right, and the

5 informative conversation that took place on the 4th of May, 2009, is not

6 part of it, or is it.

7 MS. CARTER: That's correct. We're only tendering the graphs.

8 It is a six-page item beginning with a coloured graph entitled

9 "S. Milosevic Report: Time, trend, and killings."

10 MR. LUKIC: [Interpretation]

11 Q. Ms. Tabeau, good morning. My name is Novak Lukic and I represent

12 the Defence of Mr. Perisic. I will be examining you.

13 A number of years ago, we met in this very courtroom, in the

14 Bosanski Samac case. You may not remember but I certainly do.

15 I will ask you an initial question. I believe it has to do with

16 the methodology. You told this Court that in many cases before the

17 Tribunal you acted as expert mostly for the OTP. You still work for the

18 OTP, don't you?

19 A. Yes, I do.

20 Q. Did you ever have the chance, did you ever read the decisions of

21 Tribunal in the cases where you were summoned as an expert?

22 A. I occasionally do. I read the decision, for instance, related to

23 admitting my reports in this case.

24 Q. I'm asking you this because both you and everybody else here

25 knows that it is the business of an expert to help the Court to arrive at

Page 5803

1 certain conclusions, and that was your role in the cases where you

2 testified, isn't it?

3 A. Yes, it was.

4 Q. Were you interested in whether afterward your report did indeed

5 help the Court to -- to decide in possible future cases that were similar

6 to the one in which you were testifying?

7 A. Well, I think, yes, generally, yes, in any case to which a report

8 is relevant, most certainly.

9 Q. And I suppose that you read the decision of the Tribunal in the

10 Galic case, where your report was partly taken into consideration and

11 partly was not.

12 A. If you are speaking of the decision in this case, I -- I don't --

13 you need to clarify.

14 Q. The verdict of the Tribunal in the Galic case where you were --

15 testified as an expert, and a part of the verdict reference is made to

16 your report but other parts of your report were not taken into

17 consideration. Did you actually read the first instance decision in the

18 Galic case?

19 A. If you refer to the Judgement, I did read part of it. But it's

20 never the case that a report is accepted as in a whole. There will be

21 results that are accepted and will be deferred to in a Judgement, and

22 there will be parts that are not referred to in a Judgement.

23 Q. My question was: If some parts, some comments of the Tribunal,

24 with regard to your report were not accepted, would you be -- would you

25 be inclined to correct some -- those parts of your report by looking for

Page 5804

1 other sources or trying to apply different calculation methods? Was the

2 position of the Trial Chamber some sort of green light for you to try and

3 find something different?

4 A. I think, sir, that decisions in the Judgement are justified by

5 numerous factors. It is not only my report that is taken into account

6 but several other factors, like witness testimonies, additional reports,

7 et cetera, et cetera.

8 What you are saying is if there were errors in my report, whether

9 I would be inclined to correct them. Is that so?

10 Q. Well, perhaps trying to find another method, another way of

11 arriving at certain conclusions that could assist the Trial Chamber,

12 because the Trial Chamber may decide that Ms. Tabeau's report was not

13 useful to us in certain parts.

14 Well, would that induce to you change something in your -- in

15 your method of work, because there are many similar cases before this

16 Tribunal?

17 A. Sir, in every next report I make I try to improve things. That

18 is the normal way of doing things for me. I am -- I have been learning,

19 as all of us here in this Tribunal, and I am always trying, in every

20 report, and I do my best to present a good, reliable picture and good

21 statistics. But if you are aware of errors in my Sarajevo report, then

22 please tell me, because, so far, I don't recall that errors were pointed

23 to me by Trial Chambers, which I would need to correct in my next report.

24 Q. All right. Let's not dwell upon that any longer because that is

25 for the Court to decide anyway.

Page 5805

1 A general question: You now -- you now fully stand by your

2 report, both the Galic reports and the Milosevic reports. You stand by

3 the findings and the facts contained in those reports, don't you?

4 A. Yes, I do. But for the Slobodan Milosevic, there is a more

5 recent update, which is the Dragomir Milosevic report, not used in this

6 case.

7 Q. I will focus my whole examination on these two reports that are

8 considered here. So the report from the Galic case and the

9 Slobodan Milosevic case.

10 Do you fully stand by the findings in these two reports? That's

11 my question.

12 A. Yes, I do.

13 Q. I'll ask you about your sources now.

14 I suppose that you agree with me when I say that for any

15 expertise, irrespective of the area of the expertise, it is of paramount

16 importance that the sources of information be of high quality.

17 A. Yes, it is.

18 Q. The quality of that information presupposes that in your

19 judgement as a researcher they be of the kind that you, as an expert, may

20 trust them. Isn't that so?

21 A. Yes, it is.

22 Q. And in your reports, in a certain way, you even assist the

23 quality of the information obtained from various sources that you took

24 into consideration and graded them.

25 A. The assessment of sources is an important step in our work.

Page 5806

1 Every source is assessed and various detailed -- in a very detailed way.

2 There are much more information and results from assessment than what is

3 reported in our expert reports because the assessment process is a

4 lengthy procedure and has to do with studying the data items in the

5 database; how deficient they are; how many have empty cells, that means

6 lacking values; how many show logical inconsistencies; how -- what is the

7 coverage every source; what are the biases of every source; whether

8 ethnic groups are well represented or underrepresented or

9 overrepresented. So assessment procedure is a thorough lengthy process.

10 And it is true that there this is an hierarchy of sources that we

11 used in the Galic report which certainly has to do with the reliability

12 and deficiencies of sources, but at the same time as far as I remember,

13 there is a statement in the report that all sources used were of

14 acceptable quality and were relatively reliable.

15 I'm sorry, Slobodan Milosevic report, not Galic report.

16 Q. That's correct. You said as much in the introductory part of the

17 Slobodan Milosevic case that all sources have a certain quality and you

18 rank them hierarchically when you couple them to certain events, and I

19 believe that -- this had to do with the army of BiH?

20 A. I disagree, it has not to do with the army. The hierarchy of

21 sources was applied to the sources that were used for the master list of

22 victims and not to army records because -- and well, hierarchy of sources

23 was a very practical issue related to the fact that when overlapping

24 records between the sources were found, there was a need to decide which

25 record from which source will be taken first and which ones will be

Page 5807

1 considered to be duplicates.

2 So that was the meaning of the hierarchy. And the lists of

3 military records were not used as part of the master list. They were

4 used for a totally different purpose. They were used for monitoring of

5 the status, military status of victims.

6 Q. Actually it was my error. You said that the report was the

7 report of the federal statistics bureau and that in your report that

8 report had the priority in the hierarchy and not the army report;

9 correct?

10 A. The database established by the federal statistical authority was

11 indeed the first priority in our duplicate checks. It is a database that

12 is established by professional statisticians, based on a standardised

13 death notification for every death case. So it has to have a better

14 quality than any other source. All other sources were not compiled by

15 professionals -- professionals, including even the ICRC which is

16 considered to provide highly reliable lists of missing persons. But,

17 still, it is not a statistical authority. They are not professionals.

18 Q. In order to draw a conclusion that a certain piece of information

19 is of good quality, I assume that it is of paramount importance that you

20 can compare this information and perhaps find it -- confirmation for it

21 in some other source that this is something that would actually raise or

22 attribute to assessing this as a high quality source of information --

23 piece of information; correct?

24 A. Well, we cross-references sources all the time, sir. It is a

25 continuous process. This is the major part of the work of my unit,

Page 5808

1 cross-referencing sources through individual matching. We have the

2 population census, as you remember from Simic case, the population census

3 is the core of our population database. We link sources with the

4 population census, compare how information about individuals is reported.

5 Q. Perhaps I wasn't precise enough, but I did not get the answer to

6 my question.

7 The quality of a piece of information, or, rather, a source is

8 better if you can cross-reference the information, if you can compare it

9 with some other information from some other source, rather than if you

10 only have one source of information; correct?

11 A. The quality is as it is. The quality of every source is a

12 feature of this source, right? It has nothing to do with

13 cross-referencing which can help the understanding of a given source in a

14 better way.

15 So what you're trying to say that using multiple sources is

16 better than using one source. I hope this is what you are saying. Is

17 that so, am I right or not?

18 Q. Well, here is what I want to ask you, and it has to do with what

19 I have read in your reports.

20 Your assessments -- your conclusions regarding the number of

21 killed individuals in the period that you surveyed, you had several

22 sources for this case. In particular for the Milosevic report, you used

23 five different sources that dealt with a number of individuals killed.

24 One of those sources, for instance, was the source the Federal Bureau of

25 Statistics. Now the quality of the information contained for you was

Page 5809

1 considered to be good, if you could actually find confirmation, or if you

2 could actually compare and collate them with information that you found

3 in other sources, information relating to the same facts.

4 A. No. I think you misunderstood the procedure. It is -- the

5 federal statistical authority database is a very good database. But this

6 doesn't mean that all other sources are completely to be neglected. That

7 would be very wrong to understand this, you know, like this. Using more

8 sources is better because the picture is more complete. But using one

9 large source, as the Sarajevo Household Survey, which is huge, really,

10 believe me, it's unbelievably large, there is nothing wrong in using one

11 source and produce minimum numbers from this source.

12 So I believe that numbers in both these reports, Galic report and

13 Slobodan Milosevic report, are very good, and believe me, I have compared

14 my sources so many times, from the Galic report, from Slobodan Milosevic

15 report, from Dragomir Milosevic report, and you would be surprised to see

16 the consistency in trends and patterns obtained from each of these

17 reports.

18 There is a new report in which this comparison is presented. We

19 don't have it here but I know what I am saying. There is nothing wrong

20 in using one source, a single source, like Sarajevo Household Survey, to

21 produce these kinds of statistics. There is nothing wrong to work with

22 merged sources, five of them, seven of them, because the picture is even

23 more complete. The problem with sources on war victims is that they are

24 all incomplete, whether you like it or not. That is how it is.

25 So you either have to work with sources that are showing a good

Page 5810

1 coverage and you know they are unbiased, or you have to work with

2 multiple sources.

3 Q. I fully agree with you, and in that sense, you actually compared

4 sources and assessed their quality.

5 Now my question now is this: Your conclusion and your findings

6 when talking about the wounded individuals, for them, you had -- you used

7 only one source and this is the Sarajevo Household Survey for 1994.

8 Now when these -- when that source is concerned, the information

9 contained therein, you did not compare them with any other sources.

10 A. Oh, yes, I did. Not in these two reports, but in the

11 Dragomir Milosevic report there is a wonderful comparison with hospital

12 records from major hospitals in Sarajevo, a sample of more than 3.000

13 records from Sarajevo hospitals related to only civilians and only

14 victims of shelling/sniping and other violent deaths. And there is a

15 wonderful consistency in the trends between wounded persons reported in

16 Household Survey Sarajevo with the Sarajevo hospital records, a

17 consistency that could be used to estimate the overall trend in the

18 wounded persons, both hospitalised and non-hospitalised, not only for the

19 period related to Dragomir Milosevic but to the period of the entire

20 siege.

21 Q. You know that that is not the subject matter of your testimony in

22 this -- before this Trial Chamber.

23 Now, I said that I will focus all my questions on your findings

24 in the Galic and Slobodan Milosevic cases, which are part of the record,

25 and now I will go back to the same question.

Page 5811

1 In the case of the wounded in the Galic report, because there is

2 no mention of wounded individuals in the Milosevic case, so in the Galic

3 report, did you use any other report other than the Sarajevo Household

4 Survey for 1994? Just answer my question with yes or no. It is simple.

5 A. In the Galic report, it is only wounded reported in Household

6 Survey Sarajevo.

7 Q. And that source, for the purposes of that report, you were unable

8 to compare it with any other source in respect of those facts; correct?

9 A. In the Galic report, that's no comparison with any other source

10 on wounded persons. Just the Household Survey Sarajevo.

11 Q. When you were preparing your report and working on the data, did

12 you, and I mean your entire team that worked on it, did you look for

13 sources yourself or were they offered to you by the Prosecutor?

14 A. No. We looked for the sources ourselves.

15 Q. Did you try to find certain sources and were unable to get to

16 them, or were prevented from doing so by any authorities?

17 A. Not that I recall. The source we wanted for the Galic report,

18 Sarajevo Household Survey, was offered to us by the owners of the source.

19 Q. Now, I would like to ask you about all the sources from both

20 reports.

21 In view of the fact that most of those sources other than the

22 ICRC and the BiH -- the Serbian Army report and the HVO report, all the

23 other sources are sources whose provenance are the federal authorities,

24 the authorities of the federal Bosnia and Herzegovina, in other words,

25 one of the sides in the conflict.

Page 5812

1 So did you have certain reservations in respect of the

2 authenticity and actual correctness of the information contained therein?

3 A. Well, not really. Why statistical authority would be believed to

4 be biased. Later, I also collected the same database from the RS

5 statistical authority and used these two databases together. So when it

6 comes to Sarajevo, the sources used in the Slobodan Milosevic report were

7 those that were covering the -- these episode of the conflict the best

8 possible way, so that was the major criterion of selecting our sources.

9 If at the time of the Slobodan Milosevic report, of making of Slobodan

10 Milosevic report, I already had the RS mortality database, I would most

11 certainly have taken it and included it as part of my analysis.

12 Q. All right. Now just a question that has more to do with your

13 impressions but I believe that you may answer that question, being an

14 expert witness here.

15 Now, during your communication in preparing the reports when you

16 communicated with various institutions, and I mean the Bosnia and

17 Herzegovina institutions who provided the information, was it your

18 impression that they had a vested interest in your producing the report

19 and did they ask you to provide them the report once you have completed

20 it?

21 A. Well, indeed, I shouldn't be speaking of my impressions. But, as

22 a matter of fact, it is part of the discussion of sources. You know, I

23 said sources are biased and bias often depends on who makes the sources.

24 As a matter of fact, in every conflict every party in the conflict is

25 interested in showing their victims were the most significant ones. So

Page 5813

1 that is what happens all the time. But this is my role as a

2 statistician, first of all, to recognise the bias, and second, to be able

3 to deal with the bias. And this is what I have been doing. This is what

4 I have been doing in the reports like Slobodan Milosevic.

5 First of all, bias is a very essential issue when it comes to

6 ethnic cleansing, and in the cases of Sarajevo we don't speak about

7 ethnic cleansing. We are speaking of civilians being killed perhaps

8 indiscriminately, so this is a different issue so what really matters are

9 the numbers. The numbers are the crucial issue and from this point of

10 view, in every source all ethnic groups are represented. Every source I

11 used in Slobodan Milosevic case.

12 Sarajevo was multi-ethnic community, as you know, and the

13 Sarajevo part that was under siege was falling under the authority of the

14 federation statistical authority, right? So they were registering deaths

15 of every victims disregarding the ethnicity because is that the role of

16 the statistical authority in every country in every time. So there is no

17 reason to believe that they were selective and they were skipping one

18 ethnic group because they didn't like it. I have no reasons to believe

19 that anything like this happened.

20 Q. Is your position on that the same with regard to the Sarajevo

21 Household Survey of 1994, or when this survey is in question, did you

22 have certain greater reservations before you reached your conclusions?

23 A. Sarajevo Household Survey, again, was meant to cover the entire

24 population that lived within the front lines, confrontation lines in

25 Sarajevo. I believe that possibly because the Bosniaks ran the survey,

Page 5814

1 it was easier for them to get access to the Bosniak families than to

2 minority groups there. But as a matter of fact, as I said, it is not the

3 ethnic composition of victims that is studied in my reports as the most

4 important issue. It is the numbers of victims.

5 It might be so that Bosniak victims are reported more

6 systematically and more completely than other ethnic groups but the other

7 ethnic groups are reported as well in the survey.

8 MR. LUKIC: [Interpretation] Your Honour, perhaps this is a good

9 time for the break.

10 JUDGE MOLOTO: We'll take a break and come back at quarter to

11 11.00.

12 Court adjourned.

13 --- Recess taken at 10.18 a.m.

14 --- On resuming at 10.47 a.m.

15 JUDGE MOLOTO: Yes, Mr. Lukic.

16 MR. LUKIC: [Interpretation]

17 Q. Ms. Tabeau, I would like to continue our discussion regarding

18 sources.

19 Now, the census of 1991, we can agree, was a high-quality source

20 or basis on which you could draw comparisons; correct?

21 A. Yes.

22 Q. Were you able to see the federal law, federal code on the census,

23 which was promulgated in 1990, where instructions for conducting the

24 census of 1991 were provided? I believe this was prepared by the then

25 federal statistical bureau.

Page 5815

1 A. Yes.

2 Q. Would you agree with me that this census of the population,

3 because of the -- the situation, as it was, the political situation and

4 the sensitivity of that moment, but regardless of that, that in fact the

5 census was very thoroughly and well prepared?

6 A. Yes, it was prepared very well.

7 Q. The census-takers, the individuals who were involved in this,

8 were individuals who had previously been trained - they were especially

9 instructed - in how to conduct the survey?

10 A. Yes. There was a training for the interviewers and for the

11 instructors, and a lot of other activities around the census. The

12 preparations took about ten years, as for every census.

13 Q. And the handbooks or the material that the interviewers received

14 in the course of the training provided for certain definitions of certain

15 terms that were to be explored in the census?

16 A. Yes. But I want to note we are speaking about the most important

17 survey, complete survey of the population in every country. This

18 population survey, the census, is the most essential source of the

19 information about the population in every country as it is complete, it

20 is huge, there is an army of people involved in the preparation and

21 conduct of the census, and after all, it is the major task of the

22 statistical authority in every country to prepare and conduct the census.

23 And, again, I want to make it clear, to avoid any unnecessary

24 discussion, that demographic of war and sources used in the study of

25 conflict and demographic consequences of conflict shouldn't be measure

Page 5816

1 the same way as official statistics. Official statistics and statistics

2 produced in the area of demography of conflict cannot be directly

3 compared because these two situation, like, you know, the time when the

4 census was prepared and conducted and the time when sources on victims of

5 war are compiled, these two situations cannot be compared, because they

6 are totally uncomparable.

7 Q. All right. So it is your position that they cannot be compared

8 but as a demographer when you assess the quality of a certain source, you

9 have to consider certain standards that in your assessment would be valid

10 when assessing the quality of the source. So we're not talking just

11 about the census but also about war demographics. When you assess a

12 certain source, a source that was relevant for your find information this

13 particular case, for instance, the Household Survey Sarajevo, you used

14 certain standards in assessing the quality of information contained

15 therein and you compare them to the standards -- the world standards, as

16 it were, in compiling such a service; correct?

17 A. You are speaking, sir, of a standard, a standard of assessment of

18 a source. Standards in well established statistical service like the

19 census cannot be taken and applied in war demographics, in sources for

20 war demographics. That is, first, impossible, and second, it would be

21 wrong. One important task of war demographics is first to identify the

22 sources that can be used in a study of casualties of a conflict, to

23 provide a thorough assessment of these sources, to identify deficiencies

24 and biases of sources, and having done all this, to deal with the sources

25 accordingly.

Page 5817

1 So that is my position. So it is not about the standard,

2 worldwide standard of official statistics. It is about the area of war

3 demographics that requires a different approach to sources dealing with

4 sources and analysis. It is exactly the same as you would have said

5 Western countries demography and third-world countries demography is the

6 same; it is not. Believe me, non-Western countries demography is

7 different than Western countries demography, as are war demographics and

8 official statistics. So you can't measure everything with the same

9 standard. That is impossible, and wrong, simply.

10 Q. Well, I don't want to polemicize here the theory of statistics,

11 but if we have in mind the standards used in Western countries and you

12 compare them to third-country standards, the quality of the statistics

13 is, at the same time, compared with the standards that exist in other

14 countries. Is that correct? Well, just answer me with yes or no. We

15 don't have to elaborate on this.

16 A. I can't answer yes and no because you are speaking about

17 standards in western and third-world countries. In third-world country

18 there are no sources as they are in Western countries. In many African

19 countries, for instance, there is no census at all, and still there is a

20 need to have some information about the population, analyse this

21 information and come up with certain meaningful outcomes. So we are

22 speaking of the same situation. We have war demographics that need to be

23 compiled for an area affected by war. That cannot be done using the

24 regular statistical sources and worldwide statistical standards --

25 Q. Well, let's narrow that down, let's narrow down our topic under

Page 5818

1 discussion here.

2 You said that you were aware there was law on the census that

3 regulated the census of 1991. Do you remember, and I will tell that you

4 it was Article 21 of the law, of the code, that regulated or provided for

5 sanctions for -- for sanctions for providing false information, so anyone

6 who would provide false information to an interviewer would be punished.

7 Yes or no, do you remember that there was such a provision?

8 A. I don't exactly know whether it was 21 Article or something else,

9 but most certainly people were, first of all, obliged to participate in

10 the census, everybody, and everybody by law was obligated to provide the

11 true information.

12 Q. And under the law, every individual was to provide information on

13 themselves. So the -- the basic sentiments of citizens had to be

14 provided by that citizen himself; for instance, the ethnicity or the

15 religious affiliations and so on.

16 A. That's right.

17 Q. And now you're telling us that these two criteria, these two

18 forms that are significant in providing information to an interviewer

19 could not be compared to the surveys or censuses from the time of war.

20 The 1994 census provided for one member of the household providing

21 information for all other members of the household to the interviewer.

22 Is that correct?

23 A. Yes, that's correct. However, we didn't use this information in

24 our analysis in Galic. As you might remember, 81.6 per cent of the

25 records of the Sarajevo Household Survey, the part that we computerised

Page 5819

1 for the Galic report which was about 40.000 records, 37.022, 37.022

2 records, 81.6 per cent of those records were matched with census, 1991

3 population census. Having done the matching, sir, you can transfer

4 information between sources. So for the records matched you could use

5 the ethnicity reported in the census and this is what we have done. So

6 there is no bias in ethnicity in our report because it is taken from the

7 census, and there is a small group of the individuals that are not

8 matched for which we don't have ethnicity and in our report it says

9 ethnicity unknown.

10 Q. You will agree with me, and I believe that you also said so

11 during the examination-in-chief, that the many purpose of your report in

12 the Galic case was to establish the number of civilians who were killed

13 or wounded. That was your task, the status -- the victim status was the

14 focus of your work in both reports. Correct?

15 A. Yes.

16 Q. In the Sarajevo Household Survey of 1994, in respect of the

17 wounded, you, the individuals who compiled this report, there was a --

18 there was a question relating to the status of a member of the household,

19 whether they were killed or wounded. When I say "status," I mean either

20 this could refer or apply to either a civilian or a military person.

21 Correct?

22 A. Yes.

23 Q. The person giving that information to the interviewer was not

24 obliged to prove the information given to the interviewer by any document

25 when it comes to injury or death, or other status. Whatever was said to

Page 5820

1 the interviewer was entered into the form without any proof or evidence.

2 A. Yes. There wasn't -- there was no requirement to prove the facts

3 of death, injury, dates, marital status.

4 Q. Let us now take a look at this questionnaire. It is part of this

5 document P2325. It's English page 56, and in B/C/S it's page 55.

6 Let us just briefly analyse what kind of information was given or

7 was available.

8 MR. LUKIC: [Interpretation] If we could move one page back on the

9 left side. Or, rather, no, leave it this way because the legend is both

10 in B/C/S and in English.

11 Q. Let us start with the following: This is the questionnaire used

12 by the interviewers in the Sarajevo Household Survey in 1994, isn't it?

13 A. Yes, it is.

14 Q. It is in two languages, B/C/S and English.

15 As far as I remember from your report, you only processed part of

16 the information from that questionnaire; is that correct?

17 A. Yes, it is.

18 Q. You processed the information from questions 1 and 2. Isn't it?

19 A. Yes. Yes, among others.

20 Q. Yes, yes, we'll get to that. I find number 2 interesting,

21 because it provides information about whether that family household lives

22 at the same address where it lived before the war or if, in the meantime,

23 it had moved to another of the six Sarajevo municipalities. Is that

24 correct?

25 A. Yes, it is.

Page 5821

1 Q. I think - and correct me if I'm wrong - that did you not analyse

2 the answers under 4 in your report?

3 A. No, I didn't.

4 Q. That is a question about the members of the family household who

5 had fled or were displaced from Sarajevo or who remained in the

6 territory, as it says, controlled by the aggressor, so that during the

7 population census in 1991, were members of the family household and lived

8 there and, at the time when this survey was made, for some reason, were

9 no longer to be found at that place. Is that correct?

10 A. I'm not sure what you are saying. Well, it is a question that is

11 related to listing the family members of a given household and giving

12 information about whether or not they moved from their original

13 pre-census location.

14 So in this sense, that is correct.

15 Q. Then you separately processed information which I presume were

16 important to you, the answers under 5, 6, and 12. Did I miss anything?

17 5, 6, and 12, and that is information about casualties and those who died

18 a natural death, including their status. Isn't that right?

19 A. That's right.

20 Q. At the time when you were drafting the Galic report, this was an

21 important database for you, regarding the status of the victim, wasn't

22 it?

23 A. Yes, it was.

24 Q. As far as I remember, the only other source which also mentioned

25 the status - namely, civilian or soldier - was the report of the burial

Page 5822

1 authority. I believe they were the only ones that had that information.

2 A. No, you are wrong. There was also the ABiH list of fallen

3 soldiers and other military personnel. That was used in the Galic

4 report.

5 Q. Could you please tell me where in the Galic case you relied on

6 the ABiH report for me to be able to find it?

7 A. In the executive summary in the most important table, Table 1, in

8 the Galic report, which is page 3.

9 Q. Okay. We'll get to that table later on.

10 A. We have -- we have two definitions of status listed in this

11 table. The first one is the status, marital status as reported in the

12 household survey. And the second one is called status external

13 definition. It is page 3, English version of the Galic report.

14 Q. For the purpose of precision, that is at Table 1 in the Galic

15 report at page 3, right?

16 A. Right.

17 Q. All right. Thank you.

18 A. So --

19 Q. We'll get back to that later. But I haven't seen -- oh, yes.

20 Can you tell me where you referred that ABiH source? Is it

21 marked ABiH/VRS?

22 A. This is the lists of the fallen soldiers, the military lists of

23 fallen soldiers and other military personnel that were provided to us, to

24 the Demographic Unit and for the OTP by the Defence ministries, the ABiH

25 at least by the federal Ministry of Defence and the VRS by the RS

Page 5823

1 Ministry of Defence. So we used these lists in addition to the status

2 reported in the Household Survey, as a more objective definition of the

3 military status, we believed, namely, that the status that was

4 self-reported which was not documented by any formal documents might be

5 an over-reporting of soldiers, and this actual fact is a rather known

6 fact in sever other sources. It is not only this source. People tend to

7 over-report soldiers.

8 So for the distinction between civilian victims and military

9 victims it is really important to have well-reported soldiers on the

10 first place.

11 Q. I agree fully that is a very important topic, but I really fail

12 to understand. Maybe it is my fault. When I was reading your report in

13 the Galic case, when I read the content of the report, I cannot see in

14 the content or the report itself that you refer, as you do in the

15 Milosevic case, to a BiH source. I can see you referring to the

16 Household Survey and the database of the Bakije undertaking company, and

17 you analyse those.

18 So my question is: Did you in the Galic report, rely on the --

19 the report of the Ministry of Defence? Because I haven't seen that

20 anywhere. You mentioned the table, and say that the table considers

21 information from those sources but did you analyse them anywhere? Do you

22 mention that?

23 A. I just started explaining where in the report you can find

24 references to ABiH list, and Table 1 is just one place.

25 If you go to page 27 of the English version of the report, there

Page 5824

1 is a paragraph, that is the second paragraph for -- from the bottom --

2 from the top, in which I explain how we used the military lists.

3 Page 27, English version of the report. It's the page where the Table 5

4 is included and look at the second paragraph and actually the first

5 paragraph from the top is related as well.

6 In this paragraph we explain how the external definition of

7 status was made and from this paragraph you see that we used the records

8 of fallen soldiers through the matching with the census and Household

9 Survey, and we used the records matched for marking them as soldiers, and

10 it was for about 80 per cent of matched records. For the remaining

11 20 records that were not matched we used the self-reported status. So it

12 is an integration of two definitions. For 80 per cent of records it is

13 the external status from the military lists, and for the remaining

14 20 records are matched records. It is as reported in the Household

15 Survey Sarajevo, but because it was only 20 per cent of the records for

16 which we used the Sarajevo Household Survey definition, and for the 80,

17 it is the external source used for this definition this item expresses,

18 as a matter of fact, the external definition as obtained from applying

19 the military lists.

20 Q. I must ask a layman's question to be able to understand.

21 In any case, generally speaking, in the Household Survey you

22 established that there were no more combatants than civilians who came to

23 harm, based on that information. Or let us take a look at the tables.

24 This will simplify things.

25 Let me ask you this: When you refer to the Ministry of Defence

Page 5825

1 that gave you information about killed combatants, that's the same source

2 that you used in the Milosevic case for the members of the army BiH, or

3 until -- up until 1995 or are we talking about two different sources

4 there?

5 A. It's the same source.

6 Q. I'll proceed along my way of questioning the status of civilian

7 or combatant; that, I believe, is important to all of us here in this

8 courtroom.

9 But let me get back to the data from the Household Survey. Did

10 the authors of that survey, from 1994, inform you that all households

11 were registered in the territory of the parts of the six Sarajevo

12 municipalities?

13 A. Yes, they said all to which they had access were registered.

14 Q. About 85.000 households were registered there; isn't that right?

15 A. That's right.

16 Q. But neither they nor you were able to tell whether all households

17 that, at the time, were present in the parts of those six municipalities

18 were indeed registered, or do you suppose that, indeed, all households

19 were registered?

20 A. Well, I think marginal numbers could have been skipped. But it

21 doesn't matter really. Because 85.000 households is a really huge number

22 if you compare this number from the pre-war data from the census and

23 publications on the households then you see that it is almost all. So it

24 really doesn't matter whether the few you are thinking of were left out

25 or not. That is one thing.

Page 5826

1 Another thing is the numbers we presented in the Galic report are

2 presented as minimum numbers, minimum numbers, at least numbers. So we

3 don't claim it is a complete number that we have there covering all the

4 victims. It is a minimum number. We are aware of the fact that some

5 victims are left out and we even made an estimate to account for those

6 that are not listed in our sources. To address this fact explicitly.

7 So I think all in all what we presented in the Galic report is

8 first coming from a large source, a source that was well prepared, this

9 regarding the difficult circumstances at that time, a source that was

10 collected -- data was collected through interviewers that were trained.

11 It is not that they didn't have any training at all. The structures used

12 in collecting the data were the "mjesna zajednica," the local communities

13 that had a very easy access to the population, so this is all together a

14 very good source and in addition to this, it is large. And we are saying

15 in addition to this, even though it is large and well done, we still

16 believe it is incomplete and we come up with minimum numbers, only

17 minimum numbers, conservative numbers.

18 Q. I would prefer a shorter answer so that we might be able to

19 finish today. I understand that you want to be as precise as possible.

20 But you used this source to statistically establish the

21 population that lived in that territory at that time; is that correct?

22 A. Yes. We tried to estimate the size of this population.

23 Q. You established it by multiplying 85.000 by 4, as the estimated

24 number of members of a household at that time and that territory, isn't

25 it?

Page 5827

1 A. That's right.

2 Q. And you said yourself that in your report, statistically for the

3 territories of all the municipalities that could be processed before the

4 war, that the overall numbers were raised due to the migrations of

5 population during the period of time in question; isn't that correct?

6 A. Yes, that's correct. I think the average household size was

7 lower that for -- according to the pre-war data. It was 3 point

8 something, as far as I remember.

9 Q. You said a minute ago you did not include question 4 that gave

10 information about the population that had left the territory, in order to

11 get more precise data about the persons who had left Sarajevo in --

12 during the period in question. You did not consider that, or did you?

13 A. Yes and no, I must say. I did not include the data from this

14 questionnaire, but in addition to the main data I used for the Galic

15 report, we at some point collected a pilot sample of the data which was

16 several thousands of questionnaires from two municipalities, I think one

17 of them was Stari Grad and the other one was Centar. And these

18 questionnaires, these pilot questionnaires that we collected were all

19 questionnaires available for these areas and these questionnaires

20 contained data on all demographic events, including the displaced

21 population.

22 So not only killed and wounded and natural deaths were there, but

23 also births, live births and still births, also deaths during the Second

24 World War, and the displaced population living in Sarajevo at the time of

25 the survey. Using the pilot we estimated the displaced population in

Page 5828

1 these two locations within Sarajevo and the number is in the Galic

2 report, and it was about 86 of original population living in these two

3 locations versus 14 per cent of the displaced population in these

4 municipalities.

5 So we had good ideas about the displacement present within the

6 Sarajevo.

7 Q. That pilot project you analysed encompasses two municipalities,

8 that is about 4.000 households, I believe, and that is .5 per cent of the

9 number of households established, isn't it?

10 A. That is correct.

11 Q. And according to you, as a demographer, that was a sufficient

12 parameter for you to be able to draw conclusion about the number of

13 persons who had left the territories of those municipalities; correct?

14 A. I have not drawn any conclusions based on this number. I am

15 mentioning this number to you in response to your question. Most

16 importantly, what's the importance, what the meaning -- what's the

17 meaning of the displaced persons? The report is about the number of

18 casualties within the confrontation lines. And if you are leading with

19 these questions to the calculation of rates, then, again, it is

20 irrelevant because rates are calculated only for the population

21 originating from these municipalities and not for the displaced

22 population.

23 Q. Ms. Tabeau, reading your findings, I conclude that the -- your --

24 the assumed number of persons living there, about 300.000, is the basis

25 that you -- for several analyses of yours. Your assumption of the number

Page 5829

1 of inhabitants of that area is very important, isn't it?

2 A. Yes, it is.

3 Q. Why didn't you simply add up the -- the -- the members of the

4 households in these 85.000 households from the questionnaire? I can't

5 see that this is such a huge task. Why didn't you proceed in this simple

6 fashion?

7 A. I think are you wrong about that it is a simple task. It is a

8 huge task. It is a huge task. We are speaking of entering information

9 for at least 300.000 individuals. 300.000 individuals. Do you realise

10 that? In order to have this information that are you addressing right

11 now, you would need to --

12 Q. No --

13 A. -- enter all records of both non-displaced and displaced

14 population from this survey, and every questionnaire, and there are

15 85.000 questionnaires, contains a number of names listed as household

16 members.

17 Q. I must interrupt you. Regarding this information, I believe that

18 the names and personal information are not important. Only numbers are

19 important. And you are saying that merely adding up numbers from 85.000

20 questionnaires is more difficult than making assumptions the way you did.

21 Is that what you are saying?

22 A. No, sir, this is not what I'm saying. This is -- what I'm saying

23 is that you have no idea about how statistics are made, how much it takes

24 from the moment a survey --

25 Q. I must interrupt you.

Page 5830

1 A. [Previous translation continues] ...

2 Q. I really am soliciting a simple answer. I'm saying that up until

3 the total population, if we rely on the household survey from 1994 as a

4 high-quality source, in a simple fashion, that would take some time.

5 However, could have made -- could have accurately established the

6 population present there, instead of making assumptions.

7 You did not simply count the persons that were -- that could be

8 found in the sources.

9 A. I didn't count the displaced persons and the majority of the

10 survey population because I didn't need them. In order to come up with

11 numbers of casualties, of killed persons and wounded persons, I needed to

12 computerise the records of killed and wounded persons and this is what I

13 did, and it really doesn't matter whether a person originally lived in

14 Sarajevo in 1991 or came to Sarajevo from a different part of Bosnia, as

15 long as the person is listed as a killed or wounded person.

16 Q. But the figure of 340.000 at which you arrived based on your

17 assumption was used as a basis for your statistical estimate, wasn't it?

18 A. For the estimation of rates, and if you would think a little bit,

19 the 340 [sic] I estimated from the average size of four for a household

20 is more than if you would have used 3 as the average household. And if

21 you think a little bit how the rates are calculated and that the

22 population is actually in the denominator, so the larger size is better

23 for calculation of the conservative rates, that is lower rates, lower,

24 not higher but relatively lower, lower.

25 So what I did I took the conservative approach not to

Page 5831

1 overestimate the rates, which is good for the Defence, actually.

2 Q. I am logically fully aware of what you have said but I don't want

3 to go into the issue of what is good for me or not.

4 I just want to know whether it was statistically correct if you

5 have the relevant data to simply count the persons. You didn't do it.

6 You think that this method of counting people was not -- was not

7 necessary for you, that it was sufficient for you to make an estimate.

8 Is that correct?

9 A. Well, there were two things. It wasn't absolutely necessary,

10 that is one. And, second, it was extremely resource consuming, and I'm

11 not the person to decide about the resources, sir.

12 Q. All right. That's an answer.

13 I'll ask you a few questions about another source. With regard

14 to the Milosevic case, that's the ICRC report that you have mentioned,

15 and you consider it a high-quality report, and I understood this to be a

16 report about missing persons which, according to the conclusions of the

17 ICRC, can be used as accounting for missing people, so missing persons

18 that were to be found in their database?

19 A. Yes, these are missing persons, not killed, explicitly killed or

20 wounded. These are missing person who are believed to be dead and who

21 died in violent circumstances.

22 Q. And from that source, from that list of missing persons of the

23 ICRC, you matched, as far as I understood, 570 persons who were

24 inhabitants of Sarajevo, according to the 1991 census. Did I understand

25 that correctly? Which means that ...

Page 5832

1 A. I think in order to get the right answer, you, sir, have to go to

2 page 8 of the Slobodan Milosevic report and look at Table A2.

3 Q. Mm-hm.

4 A. In this table - do you have it? - there is a category "Unique

5 ICRC." The Table shows how many records entered per source. So there is

6 a category "Unique ICRC" and it says that 70 records were taken from the

7 ICRC alone.

8 And the number you mentioned, the 570 records have been indeed

9 matched with the census and come from the ICRC, but it is not that all of

10 them are unique ICRC records taken for the Slobodan Milosevic report.

11 The -- another 500 records out of 570 overlapped with other sources,

12 predominantly with sources like Household Survey Sarajevo or Muslims

13 Against Genocide, and were accepted on the basis of the other sources for

14 the analysis.

15 So it is at the same time a proof that the 570 missing persons

16 matched with the census have been confirmed for as many as 500 as deaths

17 in other sources.

18 Q. That's exactly what I assumed, and my question was, or is, did

19 the Prosecutor inform the ICRC about these 500 individuals that they had

20 on their list as missing persons that they had actually been killed based

21 on other sources. I mean, maybe has nothing to do with our proceedings

22 here, but it seems to me if that's how you matched them, that this is

23 something that would you have done.

24 A. Sir, I don't want to speak about these issues because this is not

25 part of this trial, but we do indeed have an ongoing cooperation with the

Page 5833

1 ICRC and that should clarify your question.

2 Q. I agree. I agree there is no need to go any further into this

3 but this is something that just occurred to me while I was preparing for

4 this cross, considering it useful information.

5 But let's move on to something else. I would like to discuss now

6 the procedure for identifying those killed, and now, let's go to your

7 report in the Milosevic case, the Slobodan Milosevic case.

8 If I understood this correctly, what you did is you matched

9 information from several different sources, and once you matched that

10 information, you actually produced new information or, rather,

11 conclusions that then you incorporated into your report. Is that

12 correct?

13 A. I didn't produce any new information. I produced a new list of

14 victims and the information in this list, which I call a master list, or

15 integrated list, is exactly the same as in the underlying sources.

16 Q. And now we will discuss absolute figures, not assessments.

17 So you've, after you have matched the information, you have

18 learned the names, the first and last names, of individuals who were

19 killed. That was one type of information that you matched. Correct?

20 A. Yes. Names were part of the records matched.

21 Q. In some sources you also had a very useful bit of information

22 called the pin number for citizens which was very useful for matching;

23 correct?

24 A. You are speaking of JMBG, right? The pin number.

25 Q. Yes, that's correct.

Page 5834

1 A. [Previous translation continues] ... in some sources, yes.

2 Predominantly for a number of records in the military lists. But in a

3 majority of sources in Slobodan Milosevic, we didn't have JMBG reported.

4 Q. You then received information as to the place of death of that

5 individual, and you established that all these individuals who were the

6 subject of your survey had been killed in those areas of these six

7 municipalities under the BH Army control; correct?

8 A. Yes. The place of death was part of the record from mortality

9 sources and was taken for the master list.

10 Q. The next type of information was the time of death; correct?

11 A. Yes.

12 Q. Then you also had information as to the cause of death or injury;

13 correct?

14 A. Yes.

15 Q. And finally, perhaps I omitted something, but I am just

16 mentioning here the things that are important to me, the next item was

17 the status of that individual, whether it was a combatant or a civilian.

18 A. Yes.

19 Q. The status as category is corroborated by the following sources.

20 All the three sources of the three armies: The BH army, the VRS and the

21 HVO, the information provided from those sources. Correct?

22 A. Yes.

23 Q. Then, as we saw earlier, you also used, for this category, the

24 Sarajevo Household Survey of 1994, which also included this; correct?

25 A. Yes.

Page 5835

1 Q. And also the Bakije Funeral Home lists, they were -- the

2 information from that were also contained in the source. Correct?

3 A. Yes. And I think federal authority also included for a few

4 records, reporting of the military status.

5 Q. Do you mean the Federal Bureau of Statistics or the federal

6 Ministry of Defence?

7 A. Federal Bureau of Statistics. But in order to be sure about it,

8 I would have to double-check this. But it wasn't in any case a

9 systematic reporting of the military status in this source.

10 Q. You've -- have also pointed out in your report, and I found this

11 information acceptable but I just want to verify it with you, the lists

12 that you received from the BH army on the soldiers killed, none of those

13 lists contained the information as to the place of their death; correct?

14 A. Yes, that's right.

15 Q. I assume that you -- that the most reliable source for you

16 regarding this issue of status, whether the person was a combatant or a

17 civilian, was -- or were the military sources; and, for the moment, I

18 would like to discuss the BH army source.

19 A. Yes.

20 Q. Is it -- am I correct if I claim that if you found in another

21 source that a person was a combatant but this did not match with the

22 status as provided for in the lists of the BH army, that you would then

23 change that status to civilian, and I'm -- I refer here to those

24 individuals who have been -- who were killed.

25 A. Sir, one clarification. The military lists were not used to

Page 5836

1 create the master list as such, because place of death was not reported.

2 Military lists were used for monitoring the status of the individuals

3 that entered the master list based on the other sources used in the

4 project.

5 Whatever record was matched from the master list with military

6 lists, and there were 80 per cent of matched, the military status was

7 assigned as military because the match was established. For a small

8 number of records we couldn't do that because these records were not

9 matched. For these records the status was taken as reported in the

10 source from which the record originated.

11 So it is again a merge here. Basically it is mainly the military

12 status as reported in the ABiH lists, and VRS lists as well, we used

13 those as well, and for a small number of unmatched record it is the

14 status as reported in the original sources.

15 So these are important facts, and we should have a clear picture

16 of them.

17 Q. Well, I have to put a very basic question.

18 If you found based on the Household Survey of 1994 that an

19 individual was listed as a combatant and you didn't find that same person

20 on the BH Army list, what group did you put this individual under in your

21 report in the Milosevic case? When I mean what group, I mean whether you

22 grouped them as -- you put them under civilian or combatant.

23 A. As a combatant. Because he is - or she - reported as a combatant

24 in the original source.

25 Q. Very well. Thank you. You must know, Ms. Tabeau, that a

Page 5837

1 combatant in an armed conflict is not -- does not imply only a member of

2 an army, a soldier of an army. I assume you know that there is this

3 category of combatant.

4 A. Yes, I'm aware. But as I explained in the Galic report that if a

5 person is reported as a soldier or combatant or whatever way, we just

6 take it as a person who possibly could have died as a combatant. That

7 means in a combat. But we, as a matter of fact, don't know the specific

8 circumstances for death of all people listed as soldiers in any source.

9 So in order to decide whether these people died in combat, we would have

10 to go through these victims on case-by-case basis and study circumstances

11 and then decide. And this is obviously impossible, because the numbers

12 are so large.

13 So what we can do, we can take the status as reported, and we can

14 assume it's a proxy for the circumstances of death, but this is by no

15 means a specific confirmation of the fact that they died in combat.

16 JUDGE MOLOTO: In fact, that fact would have to be verified by

17 the person who compiled the source document?

18 THE WITNESS: Even this person probably is --

19 JUDGE MOLOTO: If at all.

20 THE WITNESS: If at all, yes. That's my answer.

21 MR. LUKIC: [Interpretation]

22 Q. Are you aware that members of paramilitary units and militias in

23 an armed conflict are not considered civilians? Do you know that this is

24 a category under the Geneva Conventions?

25 A. Well, I'm not a lawyer but I can imagine that they are not

Page 5838

1 considered, yes. It is not my area, so I wouldn't like to make

2 statements about these kinds of legal issues.

3 Q. Yes, yes, of course. Did you by any chance request that the

4 police of the BH provide you are lists of their members killed during the

5 conflict?

6 A. I think the police members are reported in the databases provided

7 by the Ministries of Defence as one of the categories. I didn't explain,

8 but these two databases, or three, we have three databases, one for ABiH,

9 one VRS, and one HVO, these three databases include not only soldiers but

10 also some other categories. They include police members who died during

11 the conflict. They include even the personnel of the ministries of the

12 Defence and associated organisations, and even persons who were part of

13 the production section -- sector providing goods and services for the

14 army.

15 So these are very broad lists and, of course, a large number of

16 individuals included in these lists died as soldiers in -- most likely in

17 combat situations but by no means all of them should and can be seen like

18 this. So police, we didn't request these records separately but police

19 records are part of the databases that we used in this report.

20 Q. Let us now take a look at your first report, and we've already

21 mentioned this topic earlier but let's now discuss it more thoroughly.

22 MR. LUKIC: [Interpretation] Can we please see document P2325,

23 Table 1.

24 Q. This is from your Galic report.

25 JUDGE MOLOTO: Just to avoid any confusion, this P2325, that's a

Page 5839

1 number for the Galic case. It's not an exhibit number for this trial.

2 THE WITNESS: Yes.

3 JUDGE MOLOTO: Okay. Thank you.

4 [Trial Chamber and registrar confer]

5 MR. LUKIC: [Interpretation] I will have to intervene.

6 JUDGE MOLOTO: It is -- it is an exhibit in this case, I'm told.

7 Thank you.

8 MR. LUKIC: [Interpretation]

9 Q. So this is a table of statistical overview. It's called "An

10 Overview of Summary Statistics from the Household Survey Sarajevo 1994."

11 And as regards the individuals killed you established that

12 1.399 civilians were killed, and 2.381 soldiers; correct?

13 A. Yes. These are minimum numbers.

14 Q. Based on that source?

15 A. Based on that source and related to the indictment period only.

16 Galic indictment period.

17 Q. That's right. When you matched this with other sources, this

18 period, you established that the number of civilians was somewhat higher,

19 the number of soldiers, combatants, somewhat lower, but that the number

20 of soldiers was still greater, far greater than the number of civilians

21 killed. Correct?

22 A. Well, we can take a look these additional pages which were used

23 at the beginning of my testimony today. I studied these numbers --

24 Q. No, I have to ask you based on my prepared questions. Please

25 just answer my questions. I'm just referring to what I have prepared

Page 5840

1 for.

2 So do you agree with me that the number of combatants under both

3 sources was greater, the number of combatants killed was greater than the

4 number of civilians killed in both these sources?

5 So please answer with yes or no.

6 JUDGE MOLOTO: You're perfectly entitled to ask questions

7 according to your preparation. I think the witness is entitled to give

8 the answers according to her preparation.

9 MR. LUKIC: [Interpretation] I agree. I agree, absolutely.

10 JUDGE MOLOTO: So she should be able to go to the documents that

11 will give her the answer that you asked for.

12 MR. LUKIC: [Interpretation] Yes, Your Honour. I will follow your

13 instructions.

14 JUDGE MOLOTO: Sorry, if you still remember the question, please

15 answer it.

16 THE WITNESS: Yes. First of all, the number of civilians in the

17 Slobodan Milosevic report for the period of the Perisic indictment, the

18 number of civilians is higher than the number of soldiers killed. I'm

19 speaking of killed persons.

20 But now we can go back to the Galic report and we can pay

21 attention to these 12 months that can be compared between Galic report

22 and Perisic -- and Slobodan Milosevic report, and we can take a look at

23 the table, if you allow me to do so. And in these cases, in Galic the

24 number of civilians will be lower than the number of soldiers, and in the

25 Slobodan Milosevic the number of civilians will be higher than the number

Page 5841

1 of soldiers.

2 So there will be an opposite picture from these two reports. We

3 can take a look at the Table --

4 MR. LUKIC: [Interpretation]

5 Q. I'm aware of that and that's exactly what I wanted to ask you

6 about. That is what I would like to us talk about here.

7 So let's just talk about the same period of time, just the period

8 of time that you reported in the Galic report, and your findings in the

9 Milosevic case. And perhaps we can proceed in the following manner.

10 MR. LUKIC: [Interpretation] Let us look -- just a moment, please.

11 Could we now please see the Galic table, A3.13. That will be on

12 page -- just a moment, please.

13 Table A3.13, Your Honours, I wonder if you have a copy of that

14 report before you. This is in Annex 3, Table 13.

15 JUDGE MOLOTO: Yes, Madam Carter is on her feet.

16 MS. CARTER: Just for the assistance of the Chamber and Defence

17 counsel, that would be page 69 in the e-court.

18 MR. LUKIC: [Interpretation] Thank you very much.

19 Q. I'm not going to ask you to make a calculation here. But please

20 take a look at 1993 and the number of soldiers, combatants killed.

21 That's in the second column.

22 According to your sources, the sources that you used in the Galic

23 case, in absolute figures, this is the total number of combatants killed

24 in the period from January through December 1993, and I calculated that

25 that figure is 1.389 or thereabout, the total for this. If you have a

Page 5842

1 calculator and if you wish to add it up yourself, I would like to compare

2 this to the same category that you described in the Milosevic case, and I

3 would like to compare this by year.

4 So, according to this table, am I correct that the total number

5 of combatants killed is what we see in column 3, for the period

6 January through December 1993, and we -- if we add them all up, we will

7 get the total number of combatants killed in the Galic case. Correct?

8 A. In 1993, killed in 1993, combatants, yes.

9 Q. That's right.

10 A. Soldiers. I would prefer to say soldiers. Is it okay?

11 Q. Well, I am using the B/C/S term.

12 You're right, actually. I wanted to check. It is soldiers.

13 That's how it is stated here.

14 Now, I would like to compare the totals that we have here with

15 your table in the Milosevic case, Table A3, where you mentioned the same

16 category, the soldiers except here you use the term "combatant." And in

17 absolute figures for the same territory, the same period you used the

18 figure 1.150 persons. Correct?

19 A. Sir, I don't have these numbers calculated right now, except for

20 this little table that I made for you, actually, to simplify your task.

21 And the total in that table, you're speaking about the same 12 months, I

22 guess, because that is the only period that is the same in these two

23 reports.

24 So what is your total, 1.150 and this is for 1993, entire 1993?

25 Q. Well, my total -- correct, that's the period. But when I added

Page 5843

1 up all the numbers in the A3.13 figure without the unknown, all the

2 combatants killed in the period January through December 1993, I came up

3 with the number 1.389. Maybe I'm mistaken but I don't think so. But

4 that number is higher than the number that you have come up with. Even

5 if you look at the breakdown by month, you can see that that number is

6 higher per month in the table that -- in the Milosevic case. Correct?

7 A. Yes. This is why I insist to show you this table for the

8 12 overlapping months for Perisic in which we can see these numbers per

9 month. Per month. And you are right, the number of soldiers will be

10 higher in Slobodan Milosevic report, generally. That is what you are

11 saying, and I agree with that.

12 I would like to add that the numbers for the Galic report are the

13 numbers from the self-reported status not from the external definition.

14 This is self-reported in Galic. So we have two different definitions

15 used for data shown in these two reports.

16 Q. You told me a minute ago that where the status was given -- the

17 status given was different, you accepted the status that was

18 self-reported.

19 Now, why did you -- the status that you -- that was used in the

20 household review -- survey, you -- in other words, let me ask you this:

21 You used -- in the Milosevic case you used the Household Survey of 1994.

22 Correct?

23 A. Right. But I can answer your question already --

24 Q. Did you include all those individuals who were mentioned in that

25 survey as combatants, did you use those figures in the Milosevic case?

Page 5844

1 Yes or no?

2 A. For the remaining unmatched 20 per cent of records.

3 Sir, there are two definitions in Galic, two definitions of

4 status in Galic. All statistics in all tables in the Galic report are

5 presented based on self-reported status taken from the Sarajevo Household

6 Survey. The other definition is shown in the summary table, Table 1, and

7 there is another table, 3, and perhaps more tables only in terms of

8 summary statistics to address the issue of over-reporting the military

9 status by the respondents of the Sarajevo Household Survey which, of

10 course, has impact on the numbers of civilians. But you are comparing

11 two things that are not directly comparable because these are two

12 different definitions of status used in these two reports.

13 Q. Well, we are drifting away from each other now.

14 Now my question was very simple, Ms. Tabeau. Please try to

15 answer in simple terms because I'm a lawyer, and I, unfortunately, cannot

16 fully follow your expert report here.

17 The status of combatant or the status of soldier was set forth in

18 the Galic report for a certain period of time, for the same location,

19 where you established that the greater number of combatants were killed

20 in that period than civilians, than the number of civilians killed in

21 that period. Correct?

22 A. The status in the Galic report is self-reported by the

23 respondents of the Sarajevo Household Survey. So there will be -- if you

24 compare this with the external definition of status, there will be more

25 soldiers, according to the internal definition.

Page 5845

1 Q. That was my question. According to the source that you used in

2 the Galic case, there was a greater number of soldiers killed than

3 civilians for the reported period; correct?

4 A. Compared to the Slobodan Milosevic, yes.

5 JUDGE MOLOTO: Maybe on that note, then we can take up from there

6 later, don't you think?

7 We will come back at half past 12.00.

8 Court adjourned.

9 --- Recess taken at 12.02 p.m.

10 --- On resuming at 12.31 p.m.

11 JUDGE MOLOTO: Yes, Mr. Lukic.

12 MR. LUKIC: Thank you, Your Honour.

13 [Interpretation] Could we please get the second report of

14 Ms. Tabeau back on the screen. That is Exhibit P2325. That is

15 Table A3.13. I said the second report from the Galic case. Maybe I

16 misquoted it. I'm sorry, the first report from the Galic case, P2325.

17 Q. We'll return to this table for a little while, Ms. Tabeau.

18 We're now looking at the period from September 1992 through

19 August 1994 in the Galic case report.

20 Do you agree with me that when I say that in the period in

21 question, in all months, except for October 1992, more soldiers were

22 killed than civilians, according to this report?

23 A. Well, generally, this is the case. But there will be more cases

24 like October 1992. But generally --

25 Q. Yes, there are some. October 1992 and now I see February 1994.

Page 5846

1 But in the other months, and in some months, significantly fewer

2 civilians than soldiers were killed in these municipalities, or parts of

3 municipalities.

4 A. Generally, this would be the expectation that fewer civilians

5 than soldiers were killed.

6 Q. Let us now look at your report P2331, that is the Milosevic case

7 report, Tables A3 and A4.

8 Let us first agree that we're talking about the same territory;

9 is that correct?

10 A. Not necessarily, because the territories were, in the Galic

11 report, within the confrontation line, which covered large parts of the

12 six municipalities, and in the Slobodan Milosevic report, these were

13 six -- the same municipalities -- municipalities taken as a whole.

14 So there is a slight difference in the territories.

15 Q. Uh-huh. This is news to me. So you're saying that in the

16 Slobodan Milosevic case, you analysed the territory within the boundaries

17 of the municipalities that were covered by the 1991 census. That is to

18 say, a larger territory than that in the other case.

19 A. Yes, slightly larger. We studied the population from the census,

20 1991 population, for both territories. Using the bridging system between

21 [B/C/S spoken] from the Galic report and the municipalities reported in

22 the census, we were able to establish that the population within the

23 confrontation lines was approximately 80 per cent of the census

24 population. I'm not speaking of geographic territories because that is

25 irrelevant. What is really relevant is the population size in these

Page 5847

1 territories.

2 Q. I agree. But let us make something clear: The population that

3 you examined that lived in the territory in the Milosevic case and the

4 population in the Galic case is 340.000 persons. That is the number.

5 And that's the number that is based on the Household Survey in 1994.

6 A. It is an estimate of this population.

7 Q. All right. That means that we're talking about the same number

8 of persons as estimated in both reports.

9 A. I'm now confused. Sorry, you -- please rephrase the question.

10 Q. In the Galic case, you found out and expressed the opinion that

11 the population was 340.000, based on the 85.000 households that were the

12 source of your work from the Household Survey carried out in 1994, and

13 you rely on the same population size in the Milosevic case report, as the

14 population you are examining.

15 A. Well, I don't recall that I mentioned the population size. I

16 mentioned six municipalities in the Slobodan Milosevic report, and the

17 population size in these municipalities is not 340.000.

18 JUDGE MOLOTO: What is the population size in those

19 municipalities?

20 THE WITNESS: In -- according to the 1991 population census --

21 JUDGE MOLOTO: No. According to the Milosevic report.

22 THE WITNESS: It is unclear because the population changed since

23 after the 1991 population census and we have the estimate, 340.000

24 individuals who stayed within the confrontation lines and it is unclear

25 how many stayed outside the confrontation lines. And because of the

Page 5848

1 freedom of movement outside the confrontation lines, it is expected that

2 large parts of the population outside the confrontation lines moved out

3 and separated as much as they could, went far away from the conflict,

4 which is not the case within the population --

5 JUDGE MOLOTO: Thank you so much --

6 THE WITNESS: -- within the population lines.

7 JUDGE MOLOTO: -- for that answer.

8 MR. LUKIC: [Interpretation]

9 Q. Let us now take a look at your report for the Milosevic case, the

10 following table, Tables 2 and 3.

11 These tables show the number of combatants and civilians that

12 came to harm. If we look at the table for the Galic case, for -- which

13 covered the period from September 1992 till August 1994, we saw that in

14 all months there is a greater number of killed combatants than civilians

15 except for two months.

16 Isn't that correct? And if we look at the same months, the same

17 period in these two tables, for each month, if we compare from

18 September 1992, in the same period, we can see that there are more

19 civilians listed here than combatants. Correct?

20 A. Well, I think -- I agreed already that the number of soldiers

21 will be greater in the Slobodan Milosevic report. Sorry. I'm now

22 confused. Let me think about it. Will be lower. Will be lower than in

23 the Galic report, generally speaking.

24 Q. Correct.

25 A. In all periods. In all periods, yes. So if you have a

Page 5849

1 particular period in mind, it is most likely correct what you are saying.

2 Q. Yes. I was comparing the two periods that are comparable from

3 two reports.

4 So the same period in two reports reflects different status

5 categories of the persons killed in the same territory. Isn't that so?

6 You have said as much. In the Milosevic case, there were more civilians

7 killed in the same period than in the Galic case.

8 A. Yes.

9 Q. Although you said a minute ago that where the source of your

10 information stated that a person was a combatant or a soldier,

11 respectively, that you stuck to that categorisation. Isn't that correct?

12 A. In the Slobodan Milosevic, it is the military lists, as the main

13 source for the military status. In the Slobodan Milosevic report.

14 Does that answer your question?

15 Q. Which means that, in the Slobodan Milosevic case, your

16 Household Survey from 1994 were persons were characterized as soldiers,

17 you did not count as soldiers unless they were mentioned in the ABiH

18 report. I'm speaking about the report for 1995. Instead, you reduced

19 the number of soldiers as compared to the number of civilians.

20 A. No. We had a long discussion about how this was done in

21 Slobodan Milosevic report. For the matched records we took the military

22 status as reported in the military lists, for the matched. For the

23 matched individuals.

24 For the remaining, it was as reported in the original source.

25 However, it is not that all household records were taken and used as such

Page 5850

1 in the Slobodan Milosevic report, because there was a hierarchy of

2 sources, so most records entered from the federal statistical authority

3 database.

4 Q. [Microphone not activated]

5 A. [Previous translation continues] ... compare these sources as if

6 they could correspond one to one to each other. They don't. They are --

7 they don't. On one hand we have a report in which one source was used,

8 Household Survey, and in the other report we have a number of source,

9 large sources used and also organised hierarchically, and there will be

10 differences in reporting in the sources whether you wanted it or not.

11 Our way of dealing the differences in reporting of the military

12 status in the sources was by using one uniform, consistent definition of

13 the military status as reported in the military lists. So this is what

14 we did, and this is a different definition of the status than the

15 definition used in the Household Survey. In addition to this, the number

16 of sources which was bigger, which is another source for differences in

17 the outcoming statistics and there is a slight difference in that

18 territory so there is enough differences not to compare these sources

19 directly.

20 Q. Madam, a simple question: Did you, in the Slobodan Milosevic

21 case, discount sources that stated a person as a soldier --

22 JUDGE MOLOTO: [Previous translation continues] ... Madam Carter.

23 MS. CARTER: Your Honour, this has been asked an answered

24 repeatedly.

25 JUDGE MOLOTO: Mr. Lukic.

Page 5851

1 MR. LUKIC: [Interpretation] I may have asked the question several

2 times, but I don't believe I got an answer.

3 JUDGE MOLOTO: You have got an answer, sir.

4 MR. LUKIC: [Interpretation] All right.

5 JUDGE MOLOTO: The witness has said to you that if the source

6 categorised the person as a soldier, she kept it as a soldier.

7 MR. LUKIC: [Interpretation] Yes.

8 JUDGE MOLOTO: [Previous translation continues] ... she's told you

9 that the reason for the difference between the Galic list and the

10 Milosevic list are the following and I may not remember them all. One,

11 in the Galic list, these were self-reported people who tended to

12 over-increase the number of soldiers; two, in the Galic -- in the -- in

13 the Milosevic list, they used soldiers -- army, military lists, not

14 self-reported people; then she said in the Galic report, they went

15 municipality by municipality for the six municipalities. In the

16 Milosevic report they went for the whole municipalities as one single

17 whole, which was slightly bigger than -- than the municipalities in the

18 Galic case.

19 So -- and that's why she is saying -- I see you shaking your

20 head. But if you want, we can look it up in the record.

21 THE WITNESS: Yeah.

22 MR. LUKIC: [Interpretation] No, I fully agree with your

23 interpretation of the witness's words. But I did not get an answer that

24 would be satisfactory for my analysis. But obviously I cannot get an

25 answer that would be of use to the Trial Chamber, but that is obviously

Page 5852

1 my problem.

2 JUDGE MOLOTO: Yeah.

3 Yes, Madam Carter.

4 MS. CARTER: Respectively to Defence counsel, I point him to

5 page 70, it's beginning at line 11, where she specifically indicates that

6 if it was a matched person from the household list to the military lists,

7 she used the military list. For all of the remainder, she kept the

8 self-reported status. She has been very clear on this issue.

9 MR. LUKIC: [Interpretation] All right.

10 THE WITNESS: May I add something?

11 JUDGE MOLOTO: Yes, ma'am. Did I interpret you correctly?

12 THE WITNESS: Very, very correctly, Your Honour. Very, very

13 well. Thank you.

14 I just want to remark that perhaps will be helpful when reading

15 these reports. It is less about directly comparing the statistics from

16 these two reports. It is more about, you know, one report being

17 complimentary to each other. I believe that the Galic report is an

18 important report because it shows unquestionable minimum number of

19 victims and for these victims shows a number of very important

20 distributions, including cause of death distributions in terms of

21 shelling and sniping. It shows sex distribution, age distribution, how

22 many women, how many elderly were killed. It shows municipality, saying

23 where the activities were the most intense.

24 So this is an important starting point. And then we have the

25 Slobodan Milosevic report which is extremely general and gives only

Page 5853

1 overall numbers, but if you read these numbers together with the specific

2 details from the Galic report, then the picture becomes very complete, in

3 my view. There might be a degree of extrapolation which is allowed

4 between these two reports but there shouldn't be a direct comparison of

5 every single statistic between these two reports.

6 MR. LUKIC: [Interpretation]

7 Q. All right, I agree. Just please explain to me where such a

8 difference in categorisation came about, that in one report the number of

9 civilians killed is much greater than soldiers and the other, it's the

10 other -- it is vice versa. So what could have caused such greatly

11 differing results of your research? Because for the same period in two

12 reports you give very different data. What is the reason for that?

13 JUDGE MOLOTO: Madam Carter.

14 MS. CARTER: Your Honour, I would again object, asked and answer.

15 As the Court has laid out the reasons why, I'm certain that Defence

16 counsel should know as well.

17 JUDGE MOLOTO: I was trying to explain some of the reasons, but I

18 think you talked to the witness.

19 MR. LUKIC: [Interpretation] I have a problem, Your Honours,

20 because I don't understand. The witness explained that where the

21 territories differ. That's my problem. I thought that the territorial

22 and -- framework and the time framework were the same, but I understand

23 that -- that the categorisation differs, and Ms. Tabeau says that two

24 reports can be understood as being complimentary but they give a very

25 different picture of the events.

Page 5854

1 However, if it clear to you, I can proceed.

2 JUDGE MOLOTO: It is not clear to me but what is clear to me are

3 her explanations, and I was still trying to tell you the differences that

4 she has told us the whole of this morning. One of them was that -- one

5 of them was that -- in addition to what I have told you already, in the

6 Milosevic list there was a multiplicity of sources. In the Galic list,

7 is the HSS list source. In the Galic list, she mentioned the periods, I

8 think, starting from 1992 or somewhere, but then she's -- this was

9 in-chief. But in the Milosevic list, it was for the entire period of the

10 war.

11 Now, these are all factors that one can expect would result in

12 slight differences, in differences. To what extent they affect the

13 differences I do not know. But, you know -- but then, of course, she has

14 just explained to you now that the Galic list categorised according to

15 age, sex --

16 MR. LUKIC: Status.

17 JUDGE MOLOTO: -- whereas the other one is just general. It

18 doesn't give children, women, men, elderly. So ...

19 MR. LUKIC: [Interpretation] All right.

20 JUDGE MOLOTO: [Previous translation continues] ... I have heard

21 so far today.

22 MR. LUKIC: [Interpretation] All right.

23 Q. Ms. Tabeau, in your second report, let us take a look at Table 1A

24 and at Table 1B. That's at page -- it's at the very beginning.

25 MR. LUKIC: [Interpretation] In B/C/S, it's page 2, and in English

Page 5855

1 it's also page 2.

2 Q. These are your key conclusions.

3 Let me check if I understood this correctly. In Table 1A you

4 speak about the absolute numbers of those killed. From 1992 through

5 1995, you state that 3.686 soldiers were killed, and in excess of 3.000

6 civilians, as direct consequences of the war.

7 Now, these are absolutely numbers that are matched?

8 A. I think the number of civilians is 4.015, not 3.000.

9 Q. Yes. Yes, indeed. 4.015 civilians and 3.686 soldiers. That's

10 what the table states. And these are the figures that you matched from

11 different sources?

12 A. That's right.

13 Q. And then in Table 1B, you work with estimated categories by

14 raising the -- the figures by 24 per cent [as interpreted], as the

15 estimated number of casualties, both civilian and military.

16 A. That's right.

17 Q. This percentage of 23 per cent [as interpreted], you applied to

18 the population of all six municipalities, didn't you?

19 A. To the victims from these territories. So, yes, yes, to the

20 victims, to the numbers of victims from these territories.

21 Q. All right.

22 MR. LUKIC: [Interpretation] I've done with this table.

23 Q. So that we agree that the first table contains absolute figures

24 and the second table contains estimates, and it represents, as you said

25 in your conclusion, a -- that this is a -- an inference and it is

Page 5856

1 described how you arrived that inference?

2 A. Yes. One is observed, absolute, minimum numbers; one is

3 estimated. That's right. Table 1B is estimated.

4 Q. Mm-hm.

5 JUDGE MOLOTO: If I may ask a question.

6 If 1B is then estimate and 1A is absolute, why do you call 1B

7 corrected?

8 THE WITNESS: Because that is of the method of estimation, it's

9 just a matter of fact a correction for the -- it doesn't have a specific

10 meaning. It is an unfortunate term. It should be called estimated. I'm

11 sorry.

12 JUDGE MOLOTO: It's all right.

13 MR. LUKIC: [Interpretation]

14 Q. A general question that follows from what you have shown us today

15 when you analysed the period of the indictment against Mr. Perisic.

16 If we look chronologically at the number of victims, that number

17 globally speaking has decreased from 1992. It significantly decreased in

18 1994, and in the three spring months of 1995 it rose but then it dropped

19 again.

20 But generally, my question is: Your survey shows that the number

21 of victims, regardless of their status, drastically decreased from 1992

22 when the conflict started until the end the war?

23 A. I would disagree. I have a very good of chart of this in the set

24 of the materials distributed today in the beginning of this testimony.

25 There is a time trend in killings, on the first page, covering the entire

Page 5857

1 conflict period and we can take a look at this chart on the ELMO. And

2 see all --

3 Q. If we can put it on the ELMO, I would like to put Table 1 on the

4 ELMO from your report.

5 MR. LUKIC: [Interpretation] Oh, I'm sorry.

6 Q. This first table, or Figure 1, shows in red the oscillation of

7 mortality in civilians and the other, the mortality of combatants.

8 What we can observe is that there is a peak in the first months

9 of 1992 until June, and after that it varies but mostly falls. And the

10 green line represents the period covered by the indictment against

11 Mr. Perisic. If we look at that period until the end of 1995 do you

12 agree with me that this period, globally speaking -- or, rather, that the

13 number of victims globally speaking in that period is significantly below

14 that until August 1995?

15 A. Generally speaking, yes, most certainly. However, there are some

16 local peaks and can you see them in the October 1993 --

17 Q. Yes, we can perhaps take a look at the table to see that.

18 Maybe the easiest way, because that is what I found when I was

19 preparing for this cross, is to use Table -- you agree with me that each

20 of these peaks, in the period when Mr. Perisic was there, that this

21 highest peak, as it were, was always lower or about the same as the peaks

22 in the previous periods. Correct?

23 A. Yes, it's correct.

24 Q. On this table, we can see the period in the spring of 1995,

25 between May and August 1995, there is an increased number of victims,

Page 5858

1 both in the civilian and the soldiers' category, and I think that you

2 stressed this in your report, that there was significant surge as

3 compared to the earlier period, a surge in the number of victims, and

4 that can be seen on the Tables A3 and A4.

5 A. Yes.

6 Q. I will now ask you, although maybe this is not part of your

7 expertise, but are you aware that at this time a major BH offensive

8 started, it started in May and went on through May and June, up until

9 August 1995 when the NATO air-strikes began?

10 A. Yes, it's possible.

11 Q. And this can lead us to the conclusion that the victims,

12 regardless of the category, there is a correlation between the victims

13 and the conflict going on in -- on the battlefields at the time.

14 A. Sir, I didn't study the factors behind these trends, so, of

15 course, there must be a correlation, there must be an association between

16 combat activities and the number of victims. There is no doubt about it.

17 But I didn't study the causes and especially the military actions of both

18 armies, so I don't think I can answer this.

19 Q. Very well, fair enough. I would just now like to take a look at

20 one more analysis where the wounded are mentioned, that's in the first

21 report, because that is not the subject of the second report.

22 MR. LUKIC: [Interpretation] That is table -- in Annex 4, Table 9.

23 I believe that is on page 91, but I'm not sure if that's the same page in

24 e-court. Perhaps page 88. I think that's the right page.

25 THE WITNESS: It's 90, I think.

Page 5859

1 MR. LUKIC: [Interpretation] I would just like to know the pages.

2 So page -- that's the first report. Table A4.9, page ...

3 JUDGE MOLOTO: A4.9 is at page 91.

4 MR. LUKIC: [Interpretation]

5 Q. We discussed a bit earlier, Ms. Tabeau, that the only source for

6 the wounded, and you analysed this only in the Galic report, was the

7 report -- or, rather, information that you received from the survey or

8 census of -- the Household Survey which is the only source that actually

9 also provides information on the wounded; correct?

10 A. Yes.

11 Q. Just to make sure that I have correctly understood this category,

12 what you've said here - and that's interesting - you've established that

13 in this period under review, 178 women soldiers were wounded, 2 of whom,

14 as far as I can see, were wounded -- were injured, rather, unrelated to

15 the conflict, but 176 women soldiers were actually wounded in the course

16 of the fighting, correct, during combat, in this period that we are

17 reviewing, which is 1992 through 1994.

18 A. Number of women is 178. 178. And they are reported as soldiers,

19 wounded as soldiers. It is the self-reported status.

20 Q. Yes, yes. Yes. That's uncontested because this is the source

21 that you had, the -- in the Household Survey.

22 Now, in this same report, there is a total of 7.594 wounded men,

23 and I can also see in the penultimate column that there are also injuries

24 and accidents that's unrelated to the conflict. But all the other

25 reports had to -- all the other reported woundings were in correlation or

Page 5860

1 connected to the fighting, to the conflict. Correct?

2 A. Well, I think there are 36 cases for men of woundings in accident

3 and other cases, including self-inflicted injuries, but honestly,

4 actually it is a very difficult task to decide whether an accident was

5 war related or not. In the Sarajevo Household Survey, many accidents

6 were reported like gas explosions, accidental falls from damaged stairs,

7 accidental falls on street, traffic accidents related to the lack of

8 light regulation of the traffic, and -- et cetera.

9 So personally I believe these accidents were war-related. At the

10 same time I recognise that nobody killed these persons by shooting, by

11 shelling or sniping. But I did include these cases of accidents among

12 the war-related cases.

13 Q. Yes. And what you have just said, I believe you have also

14 explained this during your testimony in Galic, but that is only an

15 assumption, and what we are discussing here are the figures which was

16 your task to provide certain information based on those figures and not

17 to determine whether the wounding was inflicted out of combat or as

18 collateral damage, because that would be just speculating and that was

19 not your task. Correct?

20 A. Well, you are right. I didn't try even to distinguish between

21 collateral and other. However, you must remember that in the Sarajevo

22 Household Survey killed persons are reported separately from natural

23 deaths. All these cases of accidents reported in this table and wounded

24 persons are reported separately from natural deaths, so I believe that --

25 I have done it right. I didn't mix up, for instance, for killed persons,

Page 5861

1 causes of death of natural deaths with causes of death of killed persons,

2 and for wounded causes of death were reported -- causes of wounding were

3 reported separately.

4 So I have presented the survey data as they were reported. I

5 think I didn't cross any borders here.

6 Q. Very well. Now, I will present some of my conclusions based on

7 what we have heard today or what I have read in your reports.

8 First of all, you consider and you stand by both of these

9 reports, and there's nothing in those reports that you feel should be

10 altered in any way. As far as you're concerned, they're completely

11 accurate and properly done, the way that you have worked on these

12 projects.

13 A. Yes.

14 Q. And your primary task was to try, based on various sources, to

15 determine the number of civilian casualties in the territory under the

16 control of the BH army. When I say "casualties," I mean both dead and

17 injured, or wounded.

18 A. Well, my task was to quantify the numbers of killed and wounded

19 persons, and -- yes. Yes. On the first place. And, of course, show any

20 other details related to these basic numbers.

21 Q. When determining the status of those individuals, you used

22 various sources, and all those sources confirmed -- or, rather, you used

23 all of those sources for both of your reports to determine the status of

24 those individuals -- or, rather, whether they are a civilian or a

25 combatant soldier?

Page 5862

1 A. So one source, the one from Galic, included the reporting,

2 self-reporting of the status. I used this as reported. And in the other

3 report, in Slobodan Milosevic, I did use the military lists for the

4 determination of the military status. On the first place, yes.

5 Q. In your Milosevic report, you also used the self-reporting, the

6 Household Survey as a source for your report, for your research.

7 A. For a number of unmatched records -- small number of unmatched

8 records the self-reported status from -- well, whatever status was

9 reported in the sources was accepted for the unmatched records. So it is

10 a match. But generally for a majority of records it is as reported in

11 the military lists.

12 Q. And now I will again go back to what I did not quite understand.

13 In the Milosevic case, did you use the Sarajevo Household Survey

14 self-reporting in -- for in -- in the category status of individual. Yes

15 or no? And I apologise if I'm not perfectly clear again, but did you use

16 that report in the Milosevic case in the unmatched cases, the cases that

17 were not matched with other sources?

18 A. Yes.

19 Q. And then you just used that status, that is, self-reported, there

20 as the status in your reports?

21 A. I don't know what to say really, because your problem is that you

22 don't really understand how the -- the micro analysis of data is done.

23 So it is possible that you assign a certain value to 80 per cent of your

24 records and this value would come from the military lists, and there will

25 be 20 per cent of remaining records and you can assign to these records