Trial Proceedings Transcript Vol. 8
Public Court Documents
August 18, 1983
199 pages
Cite this item
-
Case Files, McCleskey Legal Records. Trial Proceedings Transcript Vol. 8, 1983. 9a6cfec3-5aa7-ef11-8a69-6045bdd6d628. LDF Archives, Thurgood Marshall Institute. https://ldfrecollection.org/archives/archives-search/archives-item/d0402e99-6943-42d2-a3a2-524d6cf52d3e/trial-proceedings-transcript-vol-8. Accessed August 14, 2025.
Copied!
1422
1 IN THE UNITED STATES DISTRICT COURT
FJ
FOR THE NORTHERN DISTRICT OF GEQRGIA
ATLANTA DIVISION U3
4 : -—
S WARREN MCCLESKEY )
}
)
PETITIONER, ) ATLANTA, OCEORGIA
7 )
~-VS~ ) AUGUST 18, 1983
: 8 )
| WALTER DB. ZANT, WARDEN, )
)
10 RESPONDENT. }
5 iv
12 VOLUME VIL)
13 TRANSCRIPT OF PROCEEDINGS
14 BEFORE THE HONORABLE J. OWEN FORRESTER, UNITED STATES DISTRICT
13 JUDGE.
17 APPEARANCES OF COUNSEL:
IZ FOR THE PLAINTIFF: JOHN CHARLES BOGER: TIMOTHY K. FORD
AND ROBERT H. STROUP.
20 FOR THE DEFENDANT: MARY BETH WESTMORELAND AND PAULA K,
SMITH.
21 |
|
20
JIM PUGH
23 OFFICIAL COURT REPORTER
ROOM 2367, 75 SPRING STREET,
24 ATLANTA, GEORGIA 30303&
11
}
L
] i
» a
r
o
t
a
m
e
r
inet. pr———. ————— — we — ~ " Trosnmam— ww ———
1422
1 HI TNESSES
2 DIRECT CROSS REDIRECT RECROSS
3 WITNESSES FOR THE PLAINTIFF:
4
S
e
.
a
9
10
11
13 WITNESSES FOR THE DEFENSE:
14 KATZ, JOSEPH LAUREN 1426 1571
hH3
5H
DOCUMENTARY EVIDENCE
MARKED
1357
1358
1361
1424
RECEIVED
133%
1339
1541
1423
1 | (ATLANTA, FULTON COUNTY, GEQRGIA3 AUGUST 13, 1983,
2 IN OPEN COURT.)
3 rts
4 THE COURT: WELL, MS. WESTMORELAND. WHAT LUCK DID YOU
3 HAVE"?
A 3 b MS. WESTMORELAND: YOUR HONOR. WE HAD SOME LUCK AND
7 MAMAGED TO RERUN, I THINK. THE APPROPRIATE TABLES THAT WERE
8 DESIGNATED BY MR. BOGER ON FRIDAY, WITH THE EXCEPTION OF TWO,
7 WHICH WE DETERMINED TO JUST OMIT FROM OUR PRESENTATION OF THE
10 CASE.
11 THE COURT: ALL RIGHT.
12 MS. WESTMORELAND: I BELIEVE MR. STROUP HAD SOMETHING
13 HE WANTED TO BRING TD THE COURY’S ATTENTION BEFORE WE BEGIN.
14 MR. STROUP: IF I MIGHT. YOUR HONOR, JUST FOR THE
13 RECORD, STATE THAT WITH RESPECT TO PETITIONER’S EXHIBIT LuW-1,
16 WHICH IS THE FIELD OPERATIONS MANUAL WHICH WAS INTRODUCED ON
17 TUESDAY, AT THE COURT’S DIRECTION I HAVE SIGNED THAT OUT
18 OVERNIGHT AND PULLED OUT THE PORTIONS THAT ARE NOT RELEVANT TO
@ 1? THIS PROCEEDING, AND THEREFORE. 1 WOULD LIKE TO OFFER AT THIS
20 TIME, A REDUCED VERSION OF LW-1., YOUR HONOR,
21 THE COURT: FINE. IF MS. WESTMORELAND THINKS YOU HAVE
2 REDACTED IT FAIRLY, THAT WILL RE NO PROBLEM.
23 MR. STROUP: 1 HAVE SHARED A COPY OF THE REDUCED
24 EXHIBIT.
235 MS. WESTMORELAND: I HAVEN'T LOOKED AT IT CLOSELY, YOUR
+ — er ————— ———" ————" — mr — tp
1424
KATZ —- DIRECT
; HONOR. BUT IF I HAVE PROBLEMS, I WILL ADD MY OWN PAGES AND
SUPPLEMENT IT AT A LATER TIME FBEFORE THE HEARING IS OVER. I
DON’T THINK THERE WILL BE ANY.
THE COURT: FINE.
.
3
4
5 MR. STROUP: FINE.
& THE COURT: AFTER THE HEARING ON THE RECORD CONCLUDED
7 THE OTHER DAY, M3. WESTMORELAND AND MR. BOGER WORKED TQ SEE IF
a THEY COULD MINIMIZE THE PROBLEMS THAT DEVELOPED. AND AS I
? UNDERSTAND IT. MR. BOGER‘S POSITION WAS THAT THE CHANGES WOULD
10 MAKE ONLY AN INFINESTIMAL AMOUNT OF DIFFERENCE IN SOME OF THE
11 TABLES. AND THEY IDENTIFIED FOUR TO SIX THAT» CHANGES IN THE
12 DATA BASE MIGHT MAKE SOME MORE DIFFERENCE. AND THATS WHAT MS.
13 WESTMORELAND WAS REFERRING TO WHEN SHE SAID SHE HAD RERUN SOME.
14 HAVE I CORRECTLY STATED YOUR POSITION, MR. BOGER?
15 MR. BOGER: THAT 3 CORRECT, YOUR HONOR.
14 THE COURT: ALL RIGHT. YESTERDAY MORNING, I MADE A
i7 STAB AT CREATING A LAWYERS MODEL AS I DISCUSSED WITH YOU. IT
1 IS BEING TYPED AND I WILL GIVE IT TO YOU TO SEE IF IT CAN BE
@ 19 RUN. I DON’T KNOW WHETHER IT IS COMPREHENSIBLE. MORE OR LESS
20 WHETHER IT CAN BE RUN. IM CURIOUS TO SEE WHAT HAPPENS.
21 80, I WILL HAVE THAT FOR YOU SOMETIME THIS MORNING.
<2 MR. BOGER: WE-’LL BE HAPPY TO DO THAT, YOUR HONOR.
23 THE COURT: G0 AHEAD, MS. WESTMORELAND.
24 a
23 JOSEPH LAUREN KATZ,
KATZ ~ DIRECT
{ |BEING PREVIOUSLY DULY SWORN, RESUMED THE WITNESS STAND AND
2 |TESTIFIED FURTHER AS FOLLOWS!
3 DIRECT EXAMINATION (CONT’D)
M5. WESTMORELAND: YOUR HONOR, RATHER THAN ATTEMPTING
TO DELETE TABLES FROM THE NOTEBOOK WE HAD PREVIOUSLY PREPARED, I
PREPARED THE ADDITIONAL TABLES AND RELABELED THOSE FOR THE
COURT, AND I WILL RETURN THESE TO THE CLERK AT THIS TIME, TO BE
UTILIZED DURING THE TESTIMONY.
C
B
E
+
BR
E
TR
E
SE
R
S
l
THE COURT: ALL RIGHT.
10 MS. WESTMORELAND: AND THESE ARE THE ADDITIONAL COPIES
11 FOR THE COURT OF THOSE PARTICULAR COPIES.
12 THE COURT: THOSE ARE MY COPIES. MRS. PACE.
13 i MS. WESTMORELAND: THERE SHOULD BE TWO SETS. ONE FOR
14 THE CLERK AND ONE FOR THE COURT.
15 THE COURT: FOR EXAMPLE. RESPONDENT S EXHIBIT 17. NOW
16 BECOMES 17A. AND YOU HAVE FOLLOWED THAT SAME DESIGNATION
17 THROUGHOUT?
1S MS. WESTMORELAND: THATS CORRECT, YOUR HONOR.
w 19 THE COURT: ALL RIGHT.
20 MS. WESTMORELAND: AND OUT OF THIS SIX THAT MR. BOGER
21 HAD INDICATED, WE WILL BE DELETING CERTAIN OF THOSE SPECIFIC
22 TABLES AND THEY HAVEN‘T BEEN IDENTIFIED OR ANYTHING OF THE SORT,
23 SO THEY WILL NOT BE UTILIZED AND CAN BE REMOVED FROM THE
=4 NOTEBOOK BY THE CLERK AT THE END OF THE HEARING. THAT WON'T BE
25 ANY PROBLEM.
KATZ - DIRECT
1 THE COURT: ALL RIGHT. GO AHEAD.
2 BY MS. WESTMORELAND:
3 QR. DOCTOR KATZ, I REMIND YOU, YOURE STILL UNDER OATH THIS
4 MORNING.
= IF WE COULD BACKSPACE JUST BRIEFLY. DOCTOR KATZ, TO
® 6 CATCH UP A LITTLE BIT ON WHAT WE“VE DONE IN THE PAST DAY.
’ .
CAN YOU EXPLAIN FOR US JUST VERY BRIEFLY WHAT DATA SETS
8 YOU HAVE UTILIZED IN PREPARING THE TABLES THAT WE“RE PRESENTED
@ TO THE COURT?
10 A. YES. I RECEIVED THREE DIFFERENT TAPES CR BOXES OF CARDS.
53) IN LATE JANUARY. I RECEIVED FOUR BOXES OF COMPUTER CARDS WHICH
12 CONTAINED THE INFORMATION FOR THE PROCEDURAL REFORM STUDY, AND A
13 MAGNETIC TAPE, WHICH CONTAINED THE DATA FOR THE CHARGING AND
14 SENTENCING STUDY.
13 THEN APPROXIMATELY JULY 30, I RECEIVED A SECOND TAPE
1& WHICH CONTAINED 3 FILES CONCERNED WITH THE CHARGING AND
17 SENTENCING STUDY.
13 FINALLY, ON AUGUST 8, I RECEIVED ANOTHER MAGNETIC TAPE
@ 19 WHICH CONTAINED 3 FILES FOR THE CHARGING AND SENTENCING STUDY
20 AND TWO FILES FOR THE PROCEDURAL REFORM STUDY.
2 @. AND IS IT THIS LATER TAPE THAT YOU UTILIZED YESTERDAY IN
2 RERUNNING CERTAIN TABLES?
24 Q. WOULD YOU REFER TO A DOCUMENT THAT HAS BEEN MARKED
<3 RESPONDENT “3 EXHIBIT 1747
KATZ - DIRECT
MS. WESTMORELAND: AND AT THIS TIME. I“LL NOTE FOR THE
po
ke
RECORD WE'RE WITHDRAWING RESPONDENTS EXHIBIT 17 WHICH I BELIEVE
WAS IDENTIFIED BUT NOT ADMITTED.
BY M3. WESTMORELAND:
Q. WOULD YOU IDENTIFY RESPONDENTS EXHIBIT 17A PLEASE?
MR. FORD: YOUR HONOR, I DON’T HAVE ANY DOCUMENTS
MARKED WITH EXHIBIT NUMBERS. SO PERHAPS IF WE COULD REFER TO
TABLES OR SOME OTHER WAY I COULD FOLLOW ALONG.
o H
E
L
E
BE
RR
Y
SE
RR
E
0
0
1
HE
E
A
MS. WESTMORELAND: CERTAINLY. WE DIDN‘T QUITE GET THE
pt
Q OPPORTUNITY.
Fr
y
ry
RESPONDENT’S EXHIBIT 17 WAS LABELED AS TABLE 1 AND THAT
12 IS THE TABLE THAT IS BEING WITHDRAWN.
13 BY MS. WESTMORELAND:
14 3. AND IF DOCTOR KATZ COULD STATE THE HEADING OF RESPONDENT 3
13 EXHIBIT 17A.
id A. “COUNTS OF THE NUMBER OF UNKNOWNS FOR VARIABLES IN THE
17 PROCEDURAL REFORM STUDY."
R= @. AND WOULD YOU JUST IDENTIFY THIS TABLE FOR US, PLEASE,
w $9. DOCTOR KATZ?
20 A. YES. 1 FREPARED THIS TABLE YESTERDAY, AFTER RUNNING SOME
21 COMPUTER RUNS UTILIZING THE DATA FROM THE PROCEDURAL REFORM
22 STUDY THAT I WAS GIVEN ON AUGUST 8.
23 I HAVE SELECTED THE 407 CASES OUT OF THE 818 CASES THAT
24 WERE GIVEN TO ME IN THOSE COMPUTER FILES THAT WERE EQUIVALENT TO
23 THE 4607 CASES THAT I WAS GIVEN EARLIER IN JANUARY.
KATZ - DIRECT
I ALSO LOOKED AT BOTH PROCEDURAL REFORM STUDY FILES
THAT I HAD GIVEN TO ME ON AUGUST 8, AND I GOT PRETTY MUCH THE
SAME RESULTS.
2. AND HAS THIS. WAS THIS TABLE THEN RUN FROM THE MOST RECENT
TAPE THAT YOU WERE PROVIDED?
A. YES. AND I COMPARED THE TWO DIFFERENT FILES THAT I HAD FOR
THE PROCEDURAL REFORM STUDY.
AS FAR AS I CAN TELL. THEY ARE EQUIVALENT, ALTHOUGH I
HAVEN“T BEEN ABLE TO DO INTENSE ANALYSIS TQ DETERMINE THAT FACT.
BUT I DON’T BELIEVE THAT IT WILL AFFECT THE NUMBER OF UNKNOWNS,
SINCE I CHECKED IT OVER WITH RESPECT TO BOTH THOSE FILES.
MS. WESTMORELAND: YOUR HONOR, AT THIS TIME, WE WOULD
SUBMIT INTO EVIDENCE AND ASK THE COURT TO ADMIT RESPONDENTS
EXHIBIT 117A.
THE COURT: MR. FORD?
MR. FORD: AS I UNDERSTAND IT, THIS IS THE SAME AS
TABLE 1, WHICH THE COURT HAS ALREADY ADMITTED, SO --
THE COURT: I HAVE NOT.
MR. FORD: YOU HAVE NOT ADMITTED IT?
MY ONLY QUESTION WOULD BE THEN ULTIMATE RELEVANCE OF
THE UNKNOWNS, YOUR HONOR, AND NOT ALL OF IT HAS BEEN TIED UP,
BUT LATER ON, BUT IF $0. I HAVE NO OBJECTION.
THE COURT: IT SHOWS SOMETHING ABOUT THE COMPLETENESS
OF THE DATA BASE, AND I WILL ADMIT IT.
MR. FORD: THANK YDU, YOUR HONOR.
1431
KATZ ~ DIRECT
fo
o MS. WESTMORELAND: THANK YOU, YOUR HONOR.
BY M3. WESTMORELAND: [3
@. DOCTOR KATZ, I BELIEVE AT THE PREVIOUS. ON, ON TUESDAY. WE
HAD ALS0 REFERRED TO A DOCUMENT WHICH AT THAT TIME WAS MARKED
RESFONDENT”S 18. WHICH WAS TABLE ROMAN NUMBER 2.
; AND I DON’T KNOW IF WE HAD SUBMITTED IT OR NOT. IF WE
HAVE, I“LL WITHDRAW THE EXHIBIT. :
WOULD YOU REFER TO WHAT YOU HAVE LABELED AS
RESPONDENTS 18A., AND GIVE THE HEADING FOR THAT TABLE, PLEASE?
A. YES. THE HEADING OF THE TABLE IS "COUNTS OF THE NUMBER OF
Fo
y
fe
BE
E
a
B
O
E
T
B
E
fo
>
UNKNOWNS FOR VARIABLES IN THE GEORGIA CHARGING AND SENTENCING
12 STUDY."
13 ©. AND COULD YOU IDENTIFY THAT TABLE FOR US. PLEASE?
14 A. YES. THIS IS ANOTHER TABLE I PREPARED YESTERDAY WHICH
15 COUNTS THE NUMBER OF MISSING VALUES INDICATED FOR THE VARIABLES
16 FROM THE QUESTIONNAIRE ITEMS FOR THE CHARGING AND SENTENCING
17 STUDY.
ia R. WAS THAT PREPARED FROM THE MOST RECENT TAFE YOU HAVE?
eo. 19 A. YES, FROM THE AUGUST & TAPE. >
20 MS. WESTMORELAND: YOUR HONOR, WE WOULD SUBMIT INTO
21 EVIDENCE RESPONDENT'S 18A, REFERRING TO UNKNOWNS IN THE CHARGING
22 AND SENTENCING STUDY DATA AT THIS TIME.
23 THE COURT: LET ME MAKE SURE I UNDERSTAND IT.
24 HOW MANY CASES WERE IN THE SAMPLE IN THE GEDRGIA
23 CHARGING AND SENTENCING STUDY?
1432
KATZ - DIRECT
THE WITNESS: THERE WERE 1082 CASES.
THE COURT: IS THIS A COUNT, LETS GO DOWN TO PLEA
BARGAIN. WHICH HAS HAD SOME DISCUSSION, WHICH IS LDF-40.
YOU SHOW 445 UNKNOWNS. DOES THAT MEAN THAT OUT OF 445,
EXCUSE ME. OUT OF ONE THOUSAND PLUS QUESTIONNAIRES. THE PLEA
BARGAIN FOIL WAS NOT FILLED IN IN 445, OR IS THAT EXPANDED TO
REPRESENT THE UNIVERSE?
THE WITNESS: I BELIEVE THIS PARTICULAR UNKNOWN
DESIGNATION IS THE RESULT OF A CODE THAT WAS REPRESENTED FOR
THAT QUESTION, WHICH I THINK WAS A ZERO CODE. AND THERE WAS NO
EXPLANATION AS TO WHAT THAT SIGNIFIED IN THE GUESTIONNAIRE.
THE COURT: I“M NOT SURE I ASKED THE QUESTION RIGHT.
WHAT I WANT TO KNOW, DOES THIS MEAN TO ME THAT.FROM THE
THOUSAND PLUS QUESTIONNAIRES. THERE WERE 445 WHERE THAT CODE WAS
UNKNOWN, REGARDLESS OF HOW IT WAS CODED?
THE WITNESS: YES, OF THE ~
THE COURT: ON THE BASIS OF THE UNIVERSE OF 23007
THE WITNESS: NO, I DO NOT HAVE INFORMATION ABOUT
SPECIFIC PLEA BARGAIN CONCERNING THE WHOLE UNIVERSE.
THE COURT: YOU DIDN’T USE THEIR WEIGHTING?
THE WITNESS: NO, I DID NOT.
THE COURT: TO EXPAND ON A STRATIFIED SAMPLE DR ANY
OTHER BASIS. THIS IS JUST A RAW NUMBER?
THE WITNESS: YES. YOUR HONOR.
THE COURT: TO SAY IT ANOTHER WAY, THAT WOULD MEAN THAT
1433
KATZ - VOIR DIRE
1 ABOUT FORTY-FIVE PERCENT OF THOSE RUESTIONNAIRES HAD NO
2 INFORMATION ON PLEA BARGAIN.
3 THE WITNESS: YES.
4 THE COURT: ALL RIGHT.
3 MS. WESTMORELAND: WE SUBMIT IT. YOUR HONOR, AND ASK
a & THAT IT BE ADMITTED AT THIS TIME.
7 THE COURT: ALL RIGHT, MR. FORD.
8 MR. FORD: COULD I ASK JUST A COUPLE VDIR DIRE
2 RUESTIONS, YOUR HONOR?
10 THE COURT: YOU MAY.
11 VOIR DIRE EXAMINATION
12 |BY MR. FORD:
12 |@. AGAIN ON THE LABEL, "UNKNOWNS," IS THIS WHAT CAME OFF THE
14 |TAPE. NOT THE QUESTIONNAIRES, IS THAT CORRECT, THESE NUMBERS?
15 |A. THESE CAME OFF THE TAPE. YES.
1&4 |G. AND YOU DON’T DIFFERENTIATE HERE BETWEEN THE FOILS THAT ARE
17 |CODED 1. 2. BLANK. "U", AND THE ONES THAT ARE CODED IN OTHER
13 |FASHIONS, IS THAT RIGHT?
Ww 19 |A. THAT'S CORRECT.
20 |@. AND WHERE IT INDICATES THAT SOME OF THE MATTERS ARE
21 |UNDETERMINED. IS THAT THE PLACES WHERE THERE ARE FOILS AND THERE
22 |WAS NO INFORMATION AS TO WHETHER THERE MIGHT HAVE BEEN MORE —-
23 |A. THERE WAS NO SPECIFIC ENTRY AS TO WHAT "UNKNOWN" REPRESENTS.
24 |THERE WAS NO "UNKNOWN" ENTRY THAT I COULD —-
23 THE COURT: COUNT.
A
R
1
EE
RR
R
+
1434
KATZ - DIRECT
THE WITNESS: COUNT. YEAH.
BY MR. FORD:
8. AND AM I CORRECT THOSE ARE BOTH IN THE FOIL SITUATION?
A. YES.
@. 1S THAT WHERE BOTH OF THOSE COME IN?
A. YES.
MR. FORD: AGAIN. YOUR HONOR. WITH THE RESERVATION THAT
OUR POSITION WILL BE ULTIMATELY THAT THIS IS NOT RELEVANT, I
HAVE NO OTHER OBJECTION AT THIS TIME.
THE COURT: WELL. I-M NOT IN POSITION TO ASSESS ITS
STATISTICAL SIGNIFICANCE, BUT I THINK IT IS RELEVANT AS
REFLECTING ON THE COMPLETENESS OF THE DATA BASE, SO I WILL ADMIT
7. :
M5. WESTMORELAND: THANK YOU, YOUR HONOR.
THE COURT: 1 MIGHT NOTE IN THAT RULING THAT SOME, I
DONT MEAN TO INFER THAT I THINK THAT ALL OF THESE OUGHT TO BE
JERO. SOME OF THESE ARE MEASURING THINGS, WELL, NO, EXCUSE ME,
THESE ARE UNKNOWNS, THESE ARE NOT PRESENT. ALL RIGHT, ILL
ADMIT IT.
BY MS. WESTMORELAND:
9. DOCTOR KATZ, DID YOU CONDUCT A SIMILAR TYPE OF ACCOUNTING,
COUNTING PROCEDURE IN THE CHARGING AND SENTENCING STUDY AND THE
PROCEDURAL REFORM STUDY FOR THE ITEMS THAT YOU PREVIOUSLY
DISCUSSED AS "OTHER" ITEMS, THE PROVISIONS IN THE QUESTIONNAIRE
FOR AN "OTHER" DESIGNATION?
1435
KATZ —- DIRECT
i A. YES, I DID.
2 @. COULD YOU REFER TO WHAT HAS BEEN LABELED RESPONDENTS
3 EXHIBIT 19. AND WHICH ALS0. 1 BELIEVE. HAS TABLES ROMAN NUMBER
2A AND TABLES ROMAN NUMBER 2B ON THAT EXHIBIT AND IDENTIFY THAT
FOR US. PLEASE?
A. YES. THIS IS TABLE 2A, "COUNTS OF THE NUMBER OF “OTHER”
ITEMS OF VARIABLES IN THE PROCEDURAL REFORM STUDY" WHICH I
PREPARED AND COUNTED THE NUMBER OF CASES IN WHICH AN "OTHER"
L
a
h
B
E
E
B
E
R
E
S
R
DESIGNATION WAS GIVEN AS ONE OF THE FOILS FOR THE PROCEDURAL
10 REFORM STUDY, AND COUNTED THE NUMBER OF CASES WHERE "OTHER" WAS
11 INDICATED AS AN ITEM FOR THE CHARGING AND SENTENCING STUDY.
12 RA. THAT 15 —-
13 A. S50. FOR EXAMPLE, FOR THE PROCEDURAL REFORM STUDY. I GIVE THE
14 RELEVANT QUESTION THAT’S BEING REFERRED TO IN THE LEFTHAND SIDE
135 COLUMN.
14 THEN THE SECOND COLUMN I GIVE THE ITEM NUMBER
17 REFRESENTING "OTHER" RESPONSE.
13 THEN I GIVE THE NUMBER OF CASES IN WHICH THAT "OTHER"
$ 19 RESPONSE WAS DESIGNATED.
20 ; AND THEN I GIVE A SHORT DESCRIPTION OF THE VARIABLES
21 THAT THE QUESTION IS TRYING TO RELATE TO.
ol Q. S0. DO THE COLUMNS UNDER "QUESTION", THEN, REFER TO SPECIFIC
23 QUESTION OR QUESTION NUMBERS OR RESPONSE NUMBERS IN THE
24 AUESTIONNAIRE ITSELF?
25 A. YES.
1436
KATZ - DIRECT
1 0. AND WHERE DID YOU OBTAIN THESE "OTHER" INFORMATION FROM.
FROM THE TAPE OR FROM OTHER ITEMS? [3
3 A. THIS PARTICULAR TABLE WAS DONE IN TERMS OF THE TAPE I
4 RECEIVED IN LATE JANUARY.
3 AND TABLE 2B. I ALSO COUNTED THE NUMBER OF "OTHER"
4 CASES FOR THE CHARGING AND SENTENCING STUDY, AND GENERALLY THERE
7 WERE TWO WAYS IN WHICH "OTHER." "OTHERS" COULD APPEAR. ONE AS A
8 PARTICULAR DISTINCT ITEM. AS IS INDICATED IN CERTAIN OF THESE
2 VARIABLES, AND ALSO "OTHER" CAN APPEAR AS AN OPTION IN ONE OF
10 THE TWO FOIL QUESTIONS,
12 AND SO ON THE LEFTHAND SIDE OF THE FIRST COLUMN I REFER
32 TO THE PARTICULAR ITEM DESIGNATION IN THE CHARGING AND
123 . |SENTENCING STUDY.
14 THEN I GIVE THE COUNT OF THE NUMBER OF CASES IN WHICH
13 "OTHER" WAS INDICATED.
16 THEN I GIVE A SHORT DESCRIPTION OF THE VARIABLE RELATED
17 TO THAT.
18 @. DID YOU MAKE ANY EXAMINATION OF THE DATA. THE DATA BASE AND
Es 19 THE INFORMATION YOU RECEIVED, TO DETERMINE IF VARIABLES HAD BEEN
20 DEFINED FOR THE SPECIFIC ITEMS?
21 A. YES. AND I COULD NOT SEE HOW ANY OF THIS INFORMATION WAS
22 UTILIZED, EXCEPT IN RARE CASES.
23 FOR THE MOST PART. THIS INFORMATION WAS NOT PUT DMN THE
24 COMPUTER VARIABLES FOR THE COMPUTER TAPE, AND AS FAR AS I CAN
25 TELL, WAS NOT UTILIZED IN ANY ANALYSIS.
1437
KATZ ~ DIRECT
1 THE COURT: LET‘S BE A LITTLE MORE PRECISE ABOUT THAT.
2 LOOKING DOWN AT THE CHARGING AND SENTENCING STUDY. MR.
3 BOGER, LET ME MAKE SURE I“M CLEAR ON SOMETHING.
4 THERE ARE VERY FEW OF PROFESSOR BALDUS” TABLES THAT
= RELY ON THE FIRST STUDY. IS THAT CORRECT OR INCORRECT?
& 6 MR. BOGER: THAT’S CORRECT, YOUR HONOR.
7 THE COURT: SO WERE PRIMARILY INTERESTED IN THE
a CHARGING AND SENTENCING STUDY AND SECONDARILY ON THE
? TRIANGULATION NOTION OF THE OTHER --
10 MR. BOGER: THATS RIGHT, YOUR HONOR.
11 THE COURT: IS THAT A FAIR STATEMENT OF YOUR --
12 MR. BOGER: I DO THINK THATS RIGHT. THE PLACE WHERE
13 THE PROCEDURAL REFORM STUDY HAS THE MOST RELEVANCE. OF COURSE,
14 IN THE CHARGING DECISIONS, THE DECISIONS THAT TAKE THE CASE ON
15 TO A PENALTY PHASE, AND THE DECISION OF THE JURY AT A PENALTY
14 PHASE.
17 AS TO THOSE, IT REPRESENTS THE COMPLETE UNIVERSE OF
18 CASES IN THE TIME PERIOD PERIOD THAT WE'RE LOOKING AT. AND
w 19 ALMOST ALL THE "OTHER" INPUT IS ON CHARGING AND SENTENCING --
20 THE COURT: ALL RIGHT. NOW LET ME ASK YOU A FEW
21 QUESTIONS.
22 I NOTICE THAT YOU HAVE A 139 "OTHER" RESPONSES IN THE
<3 VARIABLE "SPECIAL AGGRAVATING FEATURES OF OFFENSE."
24 IS IT YOUR TESTIMONY AS TO THAT ITEM THAT THAT “OTHER™
23 DATA WAS NOT USED IN SUBSEQUENT ANALYSIS.
1438
KATZ - DIRECT
THE WITNESS: THAT’S CORRECT. I DO NOT RECALL ANY
PARTICULAR SUBDIVISION OF THESE "OTHER" ITEMS IN TERMS OF
ADDITIONAL AGGRAVATING FEATURES THAT WERE DEFINED IN THE
COMPUTER CODES.
THE COURT: WELL, MY QUESTION IS, WAS THE DATA UTILIZED
AT ALL?
THE WITNESS: AS FAR AS I CAN RECOLLECT, NO.
THE COURT: WOULD THAT ALSO BE TRUE WITH VICTIM
MITIGATING CIRCUMSTANCES?
THE WITNESS: 1 BELIEVE THERE WAS OCCASIONALLY A
QUESTION DEALING WITH WHETHER OR NOT VICTIM MITIGATING
CIRCUMSTANCES WERE PRESENT OR NOT, WHERE THIS PARTICULAR ITEM
WAS INCLUDED WITH PERHAPS 13 OR 14 OTHER ITEMS.
BUT THERE WAS NO ATTEMPT TO BREAK DOWN THE INFORMATION
CONTAINED IN THIS RESPONSE, LDF-30&.
THE COURT: HOW ABOUT CONTEMPORANEOUS OFFENSE? DO YOU
KNOW IF THAT WAS USED?
THE WITNESS: I BELIEVE IT WAS USED IN CERTAIN WAYS,
BUT SOME OF THESE "OTHER" RESPONSES RELATED TO EITHER OTHER
FELONIES OR OTHER MISDEMEANORS, AND I BELIEVE THAT THAT
INFORMATION ‘WAS INCLUDED IN CERTAIN GENERAL VARIABLES RELATING
TO THOSE AREAS.
THE COURT: HOW ABOUT DEFENDANTS MOTIVE?
THE WITNESS: 1 DO NOT BELIEVE ANY OF THE DEFENDANT
MOTIVE “OTHER" ITEMS WERE UTILIZED.
oi
e
Lh
4a
0]
[8
]
gL
ER
A
B
R
10
11
BJ
(a
1439
KATZ - VOIR DIRE
| THE COURT: ALL RIGHT. .
MS. WESTMORELAND: YOUR HONOR, AT THIS TIME WE WOULD
LIKE TO SUBMIT INTO EVIDENCE RESPONDENTS EXHIBIT 19 AS
REFLECTING "OTHER" ITEMS, AND FOR THE PURPOSE OF ESTABLISHING
THAT THERE WERE CERTAIN ITEMS THAT WERE LISTED THAT WERE NOT
SPECIFICALLY CODED AS NEW VARIABLES, AND TO INDICATE POSSIBLE
MISSING INFORMATION, FOR THAT PURPOSE AND THAT RELEVANCE.
THE COURT: MR. FORD?
MR. FORD: COULD I ASK ONE QUESTION, YOUR HONOR?
THE COURT: YOU MAY.
VOIR DIRE EXAMINATION
BY MR. FORD:
Q. DOCTOR KATZ, DID YOU DO ANY ANALYSIS TO DETERMINE WHETHER
INCLUSION OF THIS MATERIAL WOULD MAKE ANY DIFFERENCE?
A. I DID NOT HAVE THE INFORMATION ABOUT THE "OTHER" ITEMS,
BECAUSE THIS WAS INFORMATION THAT ONE COULD GET FROM THE
AUESTIONNAIRES. ALL I HAD WAS THE DESIGNATION ON THE FOILS THAT
"OTHER" HAD BEEN INDICATED. I DID NOT HAVE THE QUESTIONNAIRES
WHICH WOULD ALLOW ME TO GO AND LOOK AT THE "OTHER" ITEMS TO SEE
IF THEY WOULD POSSIBLY AFFECT OR COULD BE USED.
THE ONLY EXPOSURE IVE HAD TO THAT WAS WHEN I WAS IN
SYRACUSE AND 1 LOOKED THROUGH THE QUESTIONNAIRES, AND I LOOKED
AT THESE "OTHER" ITEMS TO SEE WHAT WAS CONTAINED THERE.
HOWEVER, I DO NOT HAVE COPIES OF ALL THE QUESTIONNAIRES
AND I AM —-
1440
KATZ - VOIR DIRE
THE COURT: WELL, -—
@. I TAKE IT THE ANSWER IS NO?
A. I DON’T KNOW IF IT WOULD AFFECT THE ANALYSIS. I DON’T HAVE
THAT INFORMATION.
@. DID YOU DO ANYTHING TO DETERMINE WHETHER THERE-S ANY
SYSTEMATIC BIAS IN THE WAY THESE "OTHERS" ARE DISTRIBUTED
THROUGHOUT THE SAMPLE?
A. NO. I DIDN’T.
MR. FORD: I“M SORRY. YDUR HONOR, I°VE ASKED 3
QUESTIONS.
THE COURT: ILL TAKE IT OFF ON ANOTHER SERIES.
MR. FORD: RELEVANCE, YOUR HONOR.
THE COURT: WHAT WAS THAT?
MR. FORD: RELEVANCE IS OUR ONLY OBJECTION AT THIS
POINT, YOUR HONOR.
THE COURT: ALL RIGHT. ILL ADMIT IT.
DIRECT EXAMINATION (CONT I)
BY MS. WESTMORELAND:
@. DOCTOR KATZ, IN EVALUATING THE DATA BASE. THE DATA BASES
THAT YOU HAD AVAILABLE TO YOU, DID YOU DO ANY COMPARISONS
BETWEEN THE TWO STUDIES. THAT IS. THE PROCEDURAL REFORM STUDY
AND THE CHARGING AND SENTENCING STUDY?
A. YES. I DID.
THERE ARE 361 CASES WHICH HAVE THE SAME CASE NUMBERS
BETWEEN THE CHARGING AND SENTENCING STUDY AND THE PROCEDURAL
1441
KATZ - VOIR DIRE
1 REFORM STUDY.
2 AND IN TRYING TO EVALUATE THE CONSISTENCY OF THE DATA.
3 I LOOKED AT VARIABLES THAT SEEMED TO BE RELATED, CLOSELY RELATED
4 BETWEEN THOSE TWD STUDIES TO SEE HOW THEY WERE CODED BETWEEN THE
5 TWO STUDIES FOR THE SAME CASES.
b Q. AND HOW DID YOU MAKE THAT DETERMINATION, OR WHAT DID YOU DO
ve IN MAKING THAT EXAMINATION?
8 A. 1 MERGED THE TWO FILES SO THAT I COULD DIRECTLY COMPARE THE
9 3561 CASES. AND FOR A SAMPLE OF ABOUT 30 OR SO VARIABLES THAT
0 IVE LISTED IN THE TABLE, I --
11 Q. AND YOU MENTIONED TABLE, WOULD YOU REFER TO RESPONDENT'S
12 EXHIBIT 208, WHICH IS A DOCUMENT THAT IS NOT IN THE NOTEBOOK AT
13 THIS TIME, AND --—
14 A. YES. THIS IS A TABLE I PREPARED YESTERDAY.
15 Q. WHAT“S THE HEADING ON THAT TABLE?
16 A. “TABLE OF NON-MATCH COUNTS."
17 @. AND WAS THAT PREPARED FROM THE MOST RECENT TAPE AVAILABLE TO
Wo 19 A. YES, IT WAS.
20 QR. WOULD YOU EXPLAIN WHAT YOU DID IN PREPARING THIS TABLE?
21 A. IN PREPARING THIS TABLE FOR THE VARIABLES DESCRIBED, I, FOR
22 A PARTICULAR CASE. I DETERMINED WHETHER THE OUTCOME FOR THAT
23 VARIABLE WAS THE SAME. AND IF THE OUTCOME WAS NOT THE SAME, I
24 DESIGNATED THAT A NON-MATCH.
23 AND THEN FOR THE CASES, I THEN COUNTED THE NUMBER OF
1442
KATZ —- VOIR DIRE
1 NON~-MATCHES FOR THESE PARTICULAR VARIABLES.
ON THE LEFTHAND SIDE AFTER DESCRIPTION OF THE VARIABLE
THATS BEING REPRESENTED. THE SECOND COLUMN GIVES THE TOTAL
2
3
A NUMBER OF NON-MATCHES THAT WERE FOUND.
3 THE THIRD COLUMN GIVES THE PERCENT OF THE 341 CASES
& THAT THESE NON-MATCHES REPRESENT. >
7 THE FOURTH COLUMN GIVES THE PROCEDURAL REFORM STUDY
2 VARIABLE THAT I UTILIZED.
? AND THE FIFTH COLUMN GIVES THE CHARGING AND SENTENCING
10 VARIABLE THAT I UTILIZED TO COMPARE FOR THIS PARTICULAR
11 VARIABLE.
12 @. YOU INDICATED THAT YDU COMPARED TO SEE IF THE RESPONSES WERE
13 A NON-MATCH,.
14 WHAT PARTICULAR THINGS WERE YOU COMPARING? IF MY
15 RECOLLECTION IS CORRECT, THE PROCEDURAL REFORM STUDY MAD IN
146 CERTAIN CASES A YES OR NO TYPE DESIGNATION WHEREAS THE CHARGING
17 AND SENTENCING STUDY HAD A PROVISION FOR FOUR SPECIFIC
18 RESPONSES.
w» 19 HOW DID YOU MAKE THAT PARTICULAR COMPARISON?
20 A. 1 WAS PARTICULARLY CONCERNED WITH THE ULTIMATE WAY IN WHICH
21 THIS VARIABLE WOULD BE CLASSIFIED. AND IN THE CHARGING AND
22 SENTENCING STUDY. I, I CODED THINGS IN TERMS OF HOW PROFESSOR
23 BALDUS HAD ULTIMATELY CODED THIS VARIABLE. AND THAT IS USING
24 HIS CONVENTION OF DESIGNATING “U’S" AND BLANKS TO BE ZEROES. AND
23 13 AND 2-5 TO BE CODED AS {I FOR A PARTICULAR VARIABLE.
oo
S
at
e
GH
EE
Gu
t
lp
(0
RUE
144%
KATZ —- VOIR DIRE
IN THE PROCEDURAL REFORM STUDY VARIABLE. 1 USED THE,
SPECIFICALLY THE VARIABLE THAT IS INDICATED, IN THAT THESE ARE
FOR THE MOST PART» ZERO — 1 VARIABLE. ZERO INDICATING THAT THE
VARIABLE IS NOT PRESENT IN THE CASE, 1 INDICATING THAT IT IS
PRESENT.
SO I WAS MATCHING, OR TESTING MATCHES FOR WHAT
ULTIMATELY I BELIEVE WAS USED IN THE ANALYSIS BY PROFESSOR
BALDUS. | |
8. IN MAKING THESE COMPARISONS, HOW DID YOU DETERMINE WHICH
PARTICULAR VARIABLES OR WHICH PARTICULAR ITEMS TO COMPARE?
A. 1 TRIED TO PICK ITEMS THAT WERE ALMOST PRECISELY DEFINED THE
SAME IN THE TWO QUESTIONNAIRES, AND TO ELIMINATE ANY JUDGMENT ON
MY PART. I LISTED THE STATUTORY AGGRAVATING CIRCUMSTANCES. AND
THEN. FOR EXAMPLE, AN ITEM LIKE SLASHED THROAT, WOULD BE, THAT
OPTION WOULD BE THE SAME FOR THE CHARGING AND SENTENCING STUDY
AND THE PROCEDURAL REFORM STUDY QUESTIONNAIRE. THE INDICATION
WOULD BE SLASHED THROAT.
SO I COMPARED WHETHER A CASE HAD THAT ATTRIBUTE CODED
OR NOT. SO THE "12" INDICATES THERE WERE TWELVE CASES WHERE THE
CODINGS WERE NOT IDENTICAL OUT OF THE 3461 CASES.
A. IN MAKING THESE COMPARISONS, DID YOU MAKE ANY LEGAL JUDGMENT
OR ANY OTHER KIND OF JUDGMENT AS TO WHAT WOULD BE EQUIVALENT?
A. 1 TRIED TO RESTRICT THIS TO QUESTIONNAIRE ITEMS IN WHICH NO
LEGAL JUDGMENT WOULD BE REQUIRED. OTHER THAN PERHAPS A JUDGMENT,
ASSUMING THAT HOMICIDE, AT THE TIME OF THE HOMICIDE. IS THE SAME
1444
KATZ - VOIR DIRE
i AS THE TIME OF THE KILLING. THOSE KINDS OF WORD CHANGES.
2 I TRIED TO MAKE THESE ITEMS, THESE VARIABLES THAT IVE
3 LISTED, THE ITEMS THAT ARE THE SAME BETWEEN THE PROCEDURAL
REFORM STUDY AND THE CHARGING AND SENTENCING STUDY IN THE WAY
ITS DESIGNATED ON THE QUESTIONNAIRES.
MS. WESTMORELAND: YOUR HONOR. AT THIS TIME. WE WOULD
4
5
&
7 LIKE TO SUBMIT RESPONDENT’S EXHIBIT 20A, THE TABLE OF NON-MATCH
8 COUNTS. INTO EVIDENCE, FOR THE PURPOSE OF ILLUSTRATING THAT
9 THERE ARE AT LEAST APPARENT INCONSISTENCIES BETWEEN THE TWO
0 STUDIES.
i1 I DON’T BELIEVE DOCTOR KATZ IS INDICATING EITHER ONE IS
12 NECESSARILY RIGHT OR WRONG IN HIS JUDGMENT. HE'S JUST
13 INDICATING HE'S DONE A COMPUTER COUNT AND FOUND THESE
14 INCONSISTENCIES. AND FOR THAT PURPOSE WE WOULD LIKE TO ASK THAT
15 IT BE ADMITTED.
14 THE COURT: MR. FORD?
17 MR. FORD: COULD I HAVE A COUPLE QUESTIONS. YOUR HONOR?
1a THE COURT: ALL RIGHT.
@ 19 VOIR DIRE EXAMINATION
20 |BY MR. FORD:
21 Q. [DOCTOR KATZ» WHAT SOURCE DID YOU USE, THEN. TO DETERMINE :
25 |WHETHER OR NOT THE VARIABLES WERE EQUIVALENT BETWEEN THE TWO
23 |STUDIES?
2 A. THE QUESTIONNAIRE. THE DESCRIPTIONS ON THE QUESTIONNAIRES AS
23 TO THOSE VARIABLES.
1445
KATZ - VOIR DIRE
1 ©. IS THAT THE ONLY SOURCE YOU USED OR DID YOU USE ANY OTHER
2 REFERENCE?
3 A. SOME OF THE VARIABLES INVOLVED LOOKING AT THE COMPUTER
4 PROGRAM THAT WAS, THAT GENERATED THESE VARIABLES, AND I LOOKED
3 AT THAT TO MAKE SURE THAT EACH OF THE ITEMS THAT WERE BEING
& COMPARED WERE REPRESENTING THE SAME THING.
7 @. IS THAT ALL YOU USED?
a A. IN TERMS OF THE STATUTORY AGGRAVATING CIRCUMSTANCES. I JUST
¥ LISTED THOSE AS STATUTORY AGGRAVATING CIRCUMSTANCES. I KNOW
10 THAT THESE ARE NOT COMPARABLE IN TERMS OF HOW THEY ARE CODED,
11 COMPUTER CODED, IN TERMS OF THE INFORMATION THATS UTILIZED.
iz THESE ITEMS ARE JUST THERE FOR. FOR COMPARISON
13 PURPOSES.
14 A. IM JUST TRYING TO FIND OUT THE SOURCE OF THE DEFINITION OF
135 THESE PARTICULAR VARIABLES.
1&6 YOU SAID THE QUESTIONNAIRE, AND THATS ALL?
17 A. THE QUESTIONNAIRE AND HOW THESE VARIABLES WERE ULTIMATELY
13 CODED.
o 1% @. DID YOU HAVE AVAILABLE TO YOU THE SET OF INSTRUCTIONS FOR
20 CODERS THAT WERE PROVIDED WITH THE CHARGING AND SENTENCING
22 A. YES, I DO.
23 @. DO YDU RECALL IF YDU REFERRED TQ THAT FOR ANY SPECIFIC
23 INSTRUCTIONS WITH REGARD TO THE DEFINITION OF THESE TERMS FOR
23 PURPOSES OF THAT STUDY?
1444
KATZ — VOIR DIRE
A. 1 REFERRED TO THEM IN DETERMINING HOW EACH OF THESE
VARIABLES WERE DEFINED AND THEN LOOKING AS TO HOW THE "OTHER"
VARIABLE WAS DEFINED AND SEEING THAT THEY REPRESENTED THE SAME
ITEMS. | :
I DID LOOK AT THE COMPUTER CODES. AND I ALSO HAD A
SOURCE OF THE VARIABLE DIRECTORY FOR THE PROCEDURAL REFORM
STUDY. WHICH GIVES INDICATION OF HOW THESE ITEMS WERE CODED. SO
I WANTED TO MAKE SURE THESE ITEMS WERE REPRESENTED, WERE
REPRESENTING THE SAME THINGS.
@. WELL, PERHAPS THERE ARE SEVERAL DIFFERENT DOCUMENTS HERE
THAT MAY WAVE SOME RESEMBLANCE TO WHAT IM ASKING ABOUT.
YOU SAID YOU USED THE QUESTIONNAIRES AND THE COMPUTER
CODES AS YOUR REFERENCES ON THIS. IS THAT RIGHT?
A. AND THE VARIABLE DIRECTORY FOR THE PROCEDURAL REFORM STUDY.
@. 1 DON’T THINK I KNOW WHAT A VARIABLE DIRECTORY IS. IS THAT
A LIST OF VARIABLES AND WHAT THEY MEAN IN PLAIN ENGLISH?
A. NO.
G9. OH?
A. IT’S A LIST OF THE VARIABLES ON THE LEFTHAND SIDE. ON THE
RIGHTHAND SIDE IN COMPUTEREZE IS A DESCRIPTION OF HOW THOSE
VARIABLES WERE GENERATED FROM THE QUESTIONNAIRES.
@. THOSE ARE THE 3 THINGS YOU USED, THEN?
THE COURT: YES OR NO.
THE WITNESS: 1 BELIEVE S0, YOUR HONOR.
THE CRURT: ALL RIGHT.
0
N
o
“g
1447
KATZ - DIRECT
BY MR. FORD:
2. THEN YOU DID NOT USE THE INSTRUCTIONS FOR CODING THE GEORGIA
PARDONS AND PAROLES BOARD QUESTIONNAIRES BACK IN 1981, THAT
DEFINED IN SOMEWHAT MORE DETAIL AS MR. GATES TESTIFIED ABOUT
LAST WEEK?
A. NO, I DIDN'T.
@. DO YOU RECALL WHETHER THERE WAS A SIMILAR SET OF
INSTRUCTIONS WITH REGARD TO THE DEFINITION OF THOSE TERMS IN THE
PROCEDURAL REFORM STUDY?
A. NO, I DONT.
THE COURT: THE QUESTIONS YOURE ASKING DON'T GO TO
ADMISSIBILITY. THEY GO TO WEIGHT. NOW YOU CAN CROSS-EXAMINE
AFTER EVERY DOCUMENT IF YOU WANT TO. BUT THAT LL BE LAST CHANCE
YOULL HAVE TO CROSS-EXAMINE ON THAT DOCUMENT.
MR. FORD: YOUR HONOR, I BELIEVE YOUR HONOR”S CORRECT.
I HAVE CROSSED THE LINE INTO CROSS-EXAMINATION.
I WOULD LIKE TO RESERVE MY CROSS-EXAMINATIUN.
I HAVE NO OBJECTION TO THIS DOCUMENT AT THIS POINT.
THE COURT: ALL RIGHT. IT WILL BE ADMITTED.
DIRECT EXAMINATION (CONT-D)
BY MS. WESTMORELAND:
R. DOCTOR KATZ, MOVING INTO, PAST THE DATA AND PAST THE
QUESTIONNAIRES THEMSELVES, DID YOU DO AN ANALYSIS OR DID YOU
EVALUATE THE APPROACHES UTILIZED BY PROFESSOR BALDUS IN
ANALYZING THE DATA IN THE PROCEDURAL REFORM STUDY?
1448
KATZ - DIRECT
1 A. YES. I DID.
- Q. AND WHAT APPROACH DID YOU TAKE IN YOUR EVALUATION OF THESE
3 ANALYSES?
4 A. 1 CONSIDERED. AFTER, AFTER READING HIS EARLY DOCUMENTS THAT
- WERE GIVEN TO ME, I HAD CONCLUDED THAT HIS HYPOTHESIS WAS A
& STATISTICAL HYPOTHESIS BASED ON A LEVEL OF AGGRAVATING AND
7 MITIGATING CIRCUMSTANCES. IN BLACK VICTIM CASES AND WHITE
2 VICTIM CASES.
7 TO RESTATE IT. I BELIEVE IT GOES. PROFESSOR BALDUS
10 BELIEVED THAT THE HIGHER LEVEL OF AGGRAVATING CIRCUMSTANCES AND
33 A LOWER LEVEL OF MITIGATING CIRCUMSTANCES ARE TOLERATED IN THE
12 GEORGIA CHARGING AND SENTENCING SYSTEM FOR BLACK VICTIM CASES,
13 RATHER THAN WHITE VICTIM CASES. BEFORE HIGHER SENTENCES ARE
14 IMPOSED.
1% @. AND WHERE DID YOU OBTAIN THAT PARTICULAR HYPOTHESIS?
146 A. THAT WAS STATED IN THE PRELIMINARY REPORT. SO, —-—
17 R. WAS THAT A DOCUMENT YOU RECEIVED IN -——
13 A. THAT WAS A DOCUMENT I RECEIVED IN NOVEMBER. 1782.
w 1% @. AND 50 DID YOU PROCEED. BASED ON THAT PARTICULAR HYFOTHESIS.
20 TO DO ANY ANALYSIS?
21 A. YES. I WANTED TO DIRECTLY TEST THAT HYPOTHESIS. AND THE
22 |FIRST THING I DID WAS DETERMINE WHAT PROFESSOR BALDUS OR TRIED
23 |T0D DETERMINE WHAT PROFESSOR BALDUS BELIEVED WERE AGGRAVATING AND
24 [MITIGATING CIRCUMSTANCES.
25 @. WOULD YOU REFER TO A DOCUMENT THAT”3 LABELED RESFONDENT”3
1449
KATZ — DIRECT
1 EXHIBIT 23. TABLE ROMAN NUMBER SIX, AND IDENTIFY THAT DOCUMENT.
PLEASE? NN
A. YES, THIS IS A — 3
4 THE COURT: WAIT JUST A SECOND. YOU’RE GETTING AHEAD OF
A MY ABSORPTION RATE.
MS. WESTMORELAND: I APOLOGIZE, YOUR HONOR.
wi
U
D
THE COURT: GIVE ME THAT HYPOTHESIS AGAIN, THE GEORGIA
0 SYSTEM TOLERATES HIGHER LEVELS ——
RL
THE WITNESS: OF AGGRAVATION AND LOWER LEVELS OF
10 MITIGATION IN BLACK VICTIM CASES. RATHER THAN WHITE VICTIM
11 CASES, BEFORE HIGHER SENTENCES ARE SOUGHT OR IMPOSED.
12 THE COURT: ALL RIGHT. GO AHEAD.
13 BY MS. WESTMORELAND:
14 Q. DOCTOR KATZ, REFERRING TO RESPONDENT‘S EXHIBIT 23, COULD YOU
13 IDENTIFY THAT, PLEASE?
16 A. YES. THAT'S A LIST OF VARIABLES IN WHICH I ATTEMPTED TO
17 CLASSIFY THEM ACCORDING TQ PROFESSOR BALDUSY CLASSIFICATION.
is UTILIZING INFORMATION I RECEIVED FROM QUESTIONNAIRES AND OTHER
w 19 |DOCUMENTS IN WHICH I PICKED OUT A LARGE SET OF VARIABLES THAT
20 WERE USED IN THE PROCEDURAL REFORM STUDY AND ATTEMPTED TO
21 CLASSIFY THEM APPROPRIATELY. ACCORDING TO PROFESSOR BALDUS”
22 CLASSIFICATION OF WHAT HE BELIEVES ARE AGGRAVATING AND
23 MITIGATING FACTORS.
24 RQ. AND WHAT PARTICULAR DOCUMENTS OR WHAT DID YOU REFER TO IN
25 MAKING YOUR DETERMINATIONS ON AGGRAVATING AND MITIGATING?
KATZ - DIRECT
1 A. THERE WERE SEVERAL DOCUMENTS THAT I REFERRED TO. THE MOST
2 IMPORTANT DOCUMENT WAS THE QUESTIONNAIRE ITSELF. IT DESIGNATED
3 EACH ITEM IN A CERTAIN CLASS. AND BASED ON THE CLASS THAT WAS
4 REPRESENTED, I TRIED TO DUPLICATE THAT DESIGNATION IN
3 CLASSIFYING THOSE VARIABLES.
& FURTHERMORE, THERE WERE METHODS USED IN DEF INING THE
7 VARIABLES THAT GAVE ME SOME INFORMATION AS TO WHETHER HE
3 BELIEVED THERE WERE AGGRAVATING OR MITIGATING
7 VARIABLES. FOR EXAMPLE. IF HE WOULD DEFINE A VARIABLE CALLED
0 AGGRAVATED MOTIVE, AND WOULD LIST A CERTAIN SET OF MOTIVES IN
ii WHICH, IF ANY OF THOSE MOTIVES WERE PRESENT, HE CONSIDERED THAT
12 TO BE AN AGGRAVATED MOTIVE, I WOULD CONCLUDE THAT HE MEANT FOR
13 THOSE CIRCUMSTANCES OR THOSE PARTICULAR MOTIVES TO REPRESENT
14 ACGRAVATED MOTIVES.
13 @. YOU ALSO LIST ATTRIBUTE VARIABLES AND TOTAL VARIABLES IN
16 THIS PARTICULAR TABLE.
17 WHAT ARE YOU REFERRING TO BY THOSE TWO DIFFERENT
18 DESIGNATIONS.
w 1 A. YES. AN ATTRIBUTE VARIABLE IS A VARIABLE IN WHICH EITHER
20 THAT VARIABLE HAS OCCURRED OR HAG NOT OCCURRED. WHERE IT CAN
21 TAKE ON EITHER A ZERO VALUE FOR NON-OCCURENCE., AND A 1 IF IT
22 OCCURS.
23 A TOTAL VARIABLE IS A VARIABLE THAT COUNTS THE NUMBER
24 OF A PARTICULAR FACTOR.
kd in Q. FOR WHAT PURPOSE DID YOU CONSTRUCT THIS PARTICULAR TABLE?
¢
3
10
KATZ - DIRECT
'A. THE PURPOSE WAS TO THEN AFTER DETERMINING AS BEST I COULD
WHAT THE CLASSIFICATIONS WERE FOR THE AGORAVATING AND MITIGATING
FACTORS AS I SUSPECTED PROFESSOR BALDUS HAD BELIEVED THEM TO BE.
I THEN WANTED TO UTILIZE THIS INFORMATION IN TESTING HIS
HYPOTHESIS. SINCE TO TEST HIS HYPOTHESIS I NEEDED TO HAVE SOME
INFORMATION AS TO WHAT ARE THE AGGRAVATING AND MITIGATING
CIRCUMSTANCES THAT HE BELIEVES THE GEORGIA CHARGING AND
SENTENCING SYSTEM TOLERATES MORE OF IN BLACK VICTIM CASES THAN
WHITE VICTIM CASES.
@. 50 IS THIS A TABLE YOU UTILIZED IN YOUR LATER ANALYSES?
A. YES.
MS. WESTMORELAND: YOUR HONOR, AT THIS TIME, 1 WOULD
LIKE TO SUBMIT RESPONDENT’S EXHIBIT 23 INTD EVIDENCE FOR THE
PURPOSE OF BASICALLY ILLUSTRATING WHAT DOCTOR KATZ DOES IN HIS
LATER ANALYSES AND TO SHOW THIS IS WHAT HE UTILIZED IN HIS LATER
ANALYSES.
MR. FORD: TABLE SIX?
MS. WESTMORELAND: YES.
MR. FORD! YOUR HONOR, IT“S OUR POSITION THAT THIS
ANALYSIS IS WELL BEYOND THE SCOPE OF DOCTOR KATZ. COMPETENCE AS
AN EXPERT. AND THAT ULTIMATELY, THEREFORE, IT WONT BE RELEVANT,
BUT CERTAINLY THIS LIST. I DON’T HAVE ANY PARTICULAR PROBLEMS
WITH. WE CERTAINLY DONT AGREE THAT THIS FOLLOWING ANALYSIS. AS
1 UNDERSTAND, IS ANYTHING HE’S COMPETENT TO TESTIFY TO. SO I
GUESS I OBJECT ON THAT GROUND.
1432
KATZ - DIRECT
i : THE COURT: WHY. WHAT I5 THERE THAT MAKES YOU HOLD THAT
2 CPINION?
3 MR. FORD: BASED ON THE INFORMATION WE RECEIVED. AS I
UNDERSTAND FROM DOCTOR KATZ‘ REPORT. OF THE, OF WHAT WERE, AS I
UNDERSTAND HE-5 SAYING HE IS TESTING A HYPOTHESIS AND MAKING A
4
3
a & NUMBER OF SOCIAL SCIENCE JUDGMENTS ALONG THE WAY AS TO WHAT
4 WOULD TEST THAT HYPOTHESIS AND WHAT WOULD BE THE SIGNIFICANCE OF
2 CERTAIN MATHEMATICAL MANIPULATIONS. CERTAINLY MATHEMATICAL
? MANIPULATIONS DOCTOR KATZ PERHAPS CAN DO. BUT THEY DON’T MEAN
10 ANYTHING AT ALL UNTIL YOU MAKE SOCIAL SCIENCE AND LEGAL
11 JUDGMENTS, AND WE DON-T BELIEVE HE IS COMPETENT TO MAKE.
12 THE COURT: I DON’T KNOW WHETHER YOU’RE RIGHT OR WRONG.
13 FROM MY POINT OF VIEW, WHAT I KNOW FROM THE CASE. CANT
14 DETERMINE IT.
15 : BUT IF HE IS DOING NQ MORE THAN TESTING THE JUDGMENTS
146 OF DOCTOR BALDUS, THAT CLEARLY IS NOT WITHOUT HIS REALM.
17 IF HE IS STARTING ANEW TO CREATE HYPOTHESIS. THAT SORT
1s OF THING, THAT MIGHT BE ANOTHER MATTER. BUT I UNDERSTOOD HIS
w i% TESTIMONY AT THIS POINT TO BE THAT HE WAS TAKING AS GIVEN
20 PROFESSOR BALDUS” JUDGMENT. THATS ALL I HAVE BEFORE ME AT THE
21 MOMENT AND IF THAT IS THE CASE, THEN I DON’T THINK IT”S EVEN
22 ARGUABLY OUTSIDE OF HIS EXPERTISE.
23 BUT IN ANY EVENT, SINCE YOU HAVE NO OBJECTION TO 23 IT
24 WILL BE ADMITTED.
23 BY MS. WESTMORELAND:
1453
KATZ - DIRECT
1
1 a. DOCTOR KATZ, I WOULD LIKE TO REFER YOU NOW TO. IF YOU WOULD
2 LOOK TO RESPONDENT“S EXHIBIT 2%, WHICH IS TABLE 8, AND IDENTIFY
THAT DOCUMENT. PLEASE?
A. YES. THIS IS A DOCUMENT WHICH I FREPARED IN WHICH I
COMPARED WHITE AND BLACK VICTIM CASES IN THE PROCEDURAL REFORM
STUDY IN TERMS OF ALL THE VARIABLES THAT WERE DEFINED IN TABLE
SIX.
LEFT-HAND COLUMN I HAVE THE VARIABLE.
H
E
E
T
R
E
N
B
E
A
J
TR
RA. BY TABLE &, YOU MEAN RESPONDENTS EXHIBIT 237
10 A. YES.
11 WHICH IS THE LIST OF THE VARIABLES.
12 ON THE LEFTHAND SIDE ARE THE VARIABLE NAMES, WHOSE
13 INTERPRETATION IS GIVEN IN TABLE 6.
14 THE SECOND COLUMN IS THE NUMBER OF WHITE VICTIM CASES
15 OUT OF THE TOTAL NUMBER OF WHITE VICTIM CASES THAT HAD THE
16 PARTICULAR ATTRIBUTE.
17 THE THIRD COLUMN, LABELED PERCENT, GIVES THE PERCENTAGE QF
is CASES, WHITE VICTIM CASES THAT HAD THE ATTRIBUTE, AGGRAVATED
w 19 BATTERY.
20 FOURTH COLUMN, LABELED BLACK VICTIM, THEN DOES A
SIMILAR COUNT FOR BLACK VICTIM CASES. AND QUT OF 247 BLACK ka
pee
s
BJ
IJ
VICTIM CASES 10 HAD AGGRAVATED BATTERY.
| []
THE FIFTH COLUMN GIVES THE PERCENT RELATED TO THAT, THE
PJ
45
BLACK VICTIM CASES.
J | TABLE, EXCUSE ME, COLUMN 6 THEN GIVES THE PERCENT
w
(2
——— St ——————— —— S——
1454
KATZ — DIRECT
DIFFERENCE CALCULATED BY TAKING THE PERCENT OF WHITE VICTIM
CASES MINUS THE PERCENT OF BLACK VICTIM CASES FOR THAT
PARTICULAR VARIABLE.
AND THEN THE LAST COLUMN IS THE "Z" VALUE.
fl. AND WHAT IS THE THE "Z" VALUE?
A. THE "IZ" VALUE IS A STANDARDIZED MEASURE BY WHICH ONE COULD
DETERMINE THE PROBABILITY THAT THE OBSERVED PERCENT DIFFERENCES
ARE DUE TO RANDOM CHANCE.
Q. AND WOULD YOU REFER THEN BACK TO RESPONDENT’S EXHIBIT 24.
TABLE 7. AND IDENTIFY THAT TABLE. PLEASE?
A. YES. IN MY ANALYSIS. I WILL BE USING THE "Z" VALUE
FREQUENTLY, AND IN THE HYPOTHESES THAT I WILL TEST. I WILL USE
THE MEASURE AT THE .0S LEVEL OF SIGNIFICANCE.
HOWEVER, IN RESPONDENTS EXHIBIT 24, I GIVE A LISY OF
POSSIBLE "I* VALUES, AND THE CORRESPONDING "P* VALUES. OR
SIGNIFICANCE LEVELS, THAT WOULD BE ASSOCIATED WITH THOSE "Zz"
VALLES.
IT-S DIFFICULT TO COMPUTE A PARTICULAR "P" VALUE, GIVEN
A HIGH “Z" VALUE, SINCE THE TABLES GENERALLY ONLY GIVE "ZI"
VALUES UP TO 3, SOMETIMES EVEN 4.
THIS ALLOWS ONE TO PUT THE "2" VALUE IN PERSPECTIVE.
SO A "I" VALUE THAT IS GREATER THAN 3, ONE CAN GET SOME
IDEA AS TO THE RELATIVE SIGNIFICANCE OF THAT VALUE.
NOW. I HAVE POSITIVE "ZI" VALUES IN RESPONDENTS EXHIBIT
24 AND NEGATIVE “ZI VALUES.
1455
KATZ - DIRECT
1 @. DID YOU MAKE THE CALCULATIONS INVOLVED IN THIS TABLE?
bJ
A. YES. THESE CALCULATIONS ARE APPROXIMATE. IT IS VERY
3 DIFFICULT TO EXACTLY DETERMINE WHAT THOSE PROBABILITIES ARE, BUT
4 THEY ARE APPROXIMATE.
@. DID YOU DO THIS BY SOME STANDARD FORMULA THAT'S UTILIZED BY a
5 STATISTICIANS OR HOW DID YOU COMPUTE THESE?
7 A. I COMPUTED THESE BY NOTING THAT BASED ON SOME MATHEMATICAL
a THEOREMS A CERTAIN COIN TOSSING EXPERIMENT CAN BE RELATED TO THE
? CALCULATION OF "“P“ VALUES FOR "ZI" DISTRIBUTION. AND GIVEN THAT
10 KNOWLEDGE I WAS ABLE TO APPROXIMATE WHAT THESE "P" VALUES ARE.
11 @. AND IS THIS WHAT YOU UTILIZED IN YOUR LATER ANALYSES. IN
$2 DETERMINING STATISTICAL SIGNIFICANCE?
12 A. YES. THE IMPORTANT "“Z" VALUES, HOWEVER. ARE THE "ZI" VALUES
14 OF PLUS 1.445, AND MINUS 1.4435.
is HA. AND WHY ARE THOSE IMPORTANT?
146 A. IF WE GET A "I" VALUE OF PLUS 1.4645, THAT WILL MEAN THAT AT
17 THE .0% LEVEL OF SIGNIFICANCE, WHITE VICTIM CASES HAVE MORE OF
18 THAT ATTRIBUTE THAN THE BLACK VICTIM CASES.
w 12? IF WE ACHIEVE THE "“Z" VALUE OF MINUS 1.6435 OR LESS,
20 THAT WILL MEAN THAT BLACK VICTIM CASES HAD MORE OF THAT WHITE
2 VICTIM CASES AT THE .0%5 LEVEL OF SIGNIFICANCE.
22 MS. WESTMORELAND: YOUR HONOR, MERELY FOR THE PURPOSES
23 OF ILLUSTRATING THE VALUES DOCTOR KATZ USER IN HIS LATER
24 ANALYSES, 1 WOULD SUBMIT RESPONDENTS EXHIBIT 24 AT THIS TIME.
— MR. FORD: NO QBJECTION, YOUR HONOR.
ett —t—— t— o——— t—
1456
KATZ —~ DIRECT
THE COURT: IT WILL BE ADMITTED.
BY MS. WESTMORELAND:
Q. DOCTOR KATZ, REFERRING BACK TO RESPONDENT S EXHIBIT 235. AND
THEN ALSO, IF YOU WOULD LOOK AT THE NEXT DOCUMENT LABELED
RESPONDENTS EXHIBIT NUMBER 26, WHICH IS TABLE 9.» WHAT ARE THE
DIFFERENCES IN THESE TWO TABLES AND WOULD YOU IDENTIFY FIRST
RESPONDENTS EXHIBIT 247
A. RESPONDENT'S EXHIBIT 26 IS TABLE 9, IS A COMPARISON OF WHITE
AND BLACK VICTIM CASES FOR THE PROCEDURAL REFORM STUDY IN WHICH
I SELECTED VARIABLES FROM THE PREVIOUS TABLE, RESPONDENT S
EXHIBIT 2%, THAT SHOW STATISTICAL SIGNIFICANCE AT THE .03 LEVEL
OF SIGNIFICANCE.
FOR THOSE VARIABLES. I THEN CATEGORIZED THEM AS TO
WHETHER THE "ZI" VALUES WERE POSITIVE. THAT MEANS GREATER THAM
1.44%, NEGATIVE, LESS THAN MINUS 1.643,
I THEN CATEGORIZED THEM IN TERMS OF WHETHER WHITE
VICTIM CASES SHOWED MORE OF THAT ATTRIBUTE OR BLACK VICTIM CASES
SHOWING MORE OF THAT ATTRIBUTE.
30 FOR EXAMPLE. THE VARIABLE ARMED ROBBERY, WE REFER
BACK TO THE RESPONDENTS EXHIBIT 25, IN THE ARMED ROBBERY CASE
THERE WERE FORTY PERCENT OF THE WHITE VICTIM CASES IN WHICH AN
ARMED ROBBERY WAS A CONTEMPORANEOUS OFFENSE AND THERE WERE 18.2
BLACK VICTIM CASES IN WHICH ARMED ROBBERY WAS A CONTEMPORANEOUS
OFFENSE.
GQ. 138.2 PERCENT?
[0
(8
b
7
8
2
10
1457
KATZ - DIRECT
A. PERCENT. YEA, THE DIFFERENCE WAS 21.8 PERCENT. AND THE "ZI"
VALUE ASSOCIATED WITH THAT, WAS 5.493. $0 THE INDICATION OF ARME]
ROBBERY 5.693, REFERS TO THE PREVIOUS TABLE WHERE ARMED ROBBERY —1
Q. YOU MEAN IN RESPONDENTS EXHIRIT 26, THAT INDICATION?
A. THAT INDICATION IS THAT WHITE VICTIM CASES HAVE A HIGHER
PROPORTION OF ARMED ROBBERY PRESENT THAN BLACK VICTIM CASES, AND
IT’S SIGNIFICANT AT THE .0T LEVEL OF SIGNIFICANCE.
@. AND IS THAT THE REPRESENTATION MADE IN BOTH OF THESE TABLES,
AS YOUR CALCULATIONS?
A. YES. IT IS.
FOR THE TOTAL VARIABLES, HOWEVER. I UTILIZED A SLIGHTLY
DIFFERENT APPROACH. SINCE THESE WERE NOT ATTRIBUTE VARIABLES, I
COMPARED MEAN DIFFERENCES.
THE COURT: WAIT JUST A MINUTE, NOW. WHAT --— IF I GO
OVER TO ABOUT THE THIRD PAGE, DO I PICK UP THE FIRST TOTAL
VARIABLES?
THE WITNESS: YES. YOUR HONOR.
THE COURT: QCKAY.
THE WITNESS: AS AN EXAMPLE, UNDER THE CATEGORY,
SPECIAL AGGRAVATING FEATURES OF THE OFFENSE, 1 HAVE THE VARIABLE
AGGCIRX, WHICH COUNTS THE NUMBER OF AGGRAVATING CIRCUMSTANCES
PRESENT AT THE OFFENSE OUT OF A LARGE LIST OF POSSIBLE
AGGRAVATING CIRCUMSTANCES. AND THE MEAN NUMBER OVERALL. THE
WHITE VICTIM CASES WAS 2.831. THE AGGRAVATING -- EXCUSE ME,
YOUR HONOR.
k]
1458
KATZ - DIRECT
THE COURT: IM NOT WITH YOU IN TERMS OF WHERE YOU ARE.
WHAT PAGE ARE YOU ON?
THE WITNESS: THIRD PAGE OF RESPONDENT 3S EXHIBIT Z5-
YOUR HONCR.
THE COURT: ALL RIGHT. ABOUT WHERE ON THE PAGE?
THE WITNESS: THE MIDDLE OF THE PAGE.
THE COURTS ALL RIGHT. HEADING?
THE WITNESS: SPECIAL AGGRAVATING FEATURES OF THE
OFFENSE.
THE COURT: OKAY. I SEE IT.
THE WITNESS: AGGCIRX IS THE VARIABLE NAME.
THE MEAN NUMBER OF AGGRAVATING CIRCUMSTANCES OVERALL
WHITE VICTIM CASES. IS 2.831, AND THE MEAN NUMBER OF AGGRAVATING
CIRCUMSTANCES OVERALL BLACK VICTIM CASES, IS 1.964.
AND THE MEAN DIFFERENCE IS CALCULATED AND THE "ZI" VALUE
ASSOCIATED WITH THIS MEAN DIFFERENCE IS PRESENTED.
BY MS. WESTMORELAND:
@. AND THEN DID YOU MAKE THE SAME DETERMINATIONS FOR THOSE
|PARTICULAR VARIABLES IN ACCUMULATING THE INFORMATION ON TABLE 97
A. EXCUSE ME, WOULD YOU REPEAT THE QUESTION, PLEASE?
@. YES. WE TALKED ABOUT THE ATTRIBUTE VARIABLES. WITH REGARD
TO THE TOTAL VARIABLES, DID YOU ALSO TAKE THE SIGNIFICANT “Z"
VALUES AND MOVE THEM TO TABLE 9 AS WELL?
Ah. YES. 1 DID.
@. DID YOU MAKE ANY KIND OF DETERMINATION OR JUDGMENTS IN
1457
KATZ - DIRECT
i CREATING TABLE 9 FROM. I“M SORRY, RESPONDENT-S EXHIBIT 26, FROM
< RESPONDENT “S$ EXHIBIT 257
2 A. NO. I SIMPLY SELECTED THOSE VARIABLES THAT WERE SIGNIFICANT
4 AT THE .05 LEVEL OF SIGNIFICANCE, AND THE CLASSIFICATIONS HAD
3 ALREADY BEEN PROVIDED FOR IN TABLE & AND I SIMPLY REPRESENTED
® & THEM IN TABLE 9.
7 2. AND IN MAKING THIS REPRESENTATION IN TABLE 9, WHAT IS SHOWN?
8 THE COURT: NOW. WAIT A MINUTE, WHAT IS TABLE ?7
g MS. WESTMORELAND: I“M SORRY, YOUR HONOR. RESPONDENTS
10 EXHIBIT 26.
11 THE COURT: IS THAT WHAT WE-VE BEEN LDOKING AT?
12 MS. WESTMORELAND: IT’S THE SECOND OF THE TWO TABLES,
13 YOUR HONOR.
14 THE WITNESS: IT SHOWS THERE ARE A LOT OF VARIABLES,
15 AGGRAVATING VARIABLES IN WHICH WHITE VICTIM CASES HAVE A HIGHER
1&6 PROPORTION OF THAT VARIABLE THAN BLACK VICTIM CASES.
17 AND THERE ARE A CREAT DEAL OF VARIABLES IN WHICH,
1a PARDON ME, MITIGATING FACTORS, IN WHICH BLACK VICTIM CASES HAVE
w 17 A HIGHER PROPORTION OR A HIGHER AVERAGE NUMBER OF THOSE
20 MITIGATING FACTORS. THAN WHITE VICTIM CASES.
=! @. AND DID YOU YOU UTILIZE THESE TWO TABLES IN TESTING THE
22 PREVIOUS HYPOTHESIS YOU OBTAINED FROM PROFESSOR BALDUS” EARLIER
23 WORK?
24 A. YES, I DID.
Pb Q. AND STATISTICALLY. DID YOU FIND THAT THIS TABLE SUPPORTED OR
14460
KATZ - DIRECT
DID NOT SUPPORT THAT HYPOTHESIS?
MR. FORD: 1 OBJECT TO THAT, YOUR HONOR, AT THIS POINT
WE“RE JUMPING AHEAD. IT SEEMS TQ ME. TO AN ULTIMATE CONCLUSION
THAT CLEARLY THERES NO FOUNDATION FOR AT THIS TIME.
THE COURT: CAN YOU BE MORE SPECIFIC IN YOUR VIEW OF
WHY THIS FOUNDATION IS LACKING?
MR. FORD: YOUR HONOR, AGAIN WHAT THESE NUMBERS. AS I
UNDERSTAND WHAT THESE ARE ULTIMATELY GOING TO BE USED FOR. TO
TEST A HYPOTHESIS, THE HYPOTHESIS IS ONE, STATED BY DOCTOR KATZ.
IS ONE THAT HE HAS DERIVED, 1 BELIEVE COUNSELS WORD IS OBTAINED
FROM FROFESSOR BALDUS.
WE BELIEVE HE HAS STATED THE WRONG HYPOTHESIS. WE
BELIEVE HE“S NOT COMPETENT TO STATE THE HYPOTHESIS, THEREFORE.
WE THINK IT’S IRRELEVANT.
THE COURT: IT IS IRRELEVANT ONLY IF HE HAS NOT
ACCURATELY STATED A HYPOTHESIS OF PROFESSOR BALDUS.
HE HAS TESTIFIED THAT HE OBTAINED A HYPOTHESIS FROM
DOCTOR EBALDUS” REPORT, WHICH IS THE EXTENT OF THE EVIDENCE I
HAVE ON THE SUBJECT.
ILL BE GLAD TO HEAR YOUR ARGUMENT AS TO WHY THAT S NOT
MR. FORD: WELL, YOUR HONOR. PERHAPS I. I DON’T KNOW IF
|THE COURT WANTS TO HEAR FROM ME ON WHERE I“M GOING OR PERHAPS IF
I COULD ASK DOCTOR KATZ.
THE COURT: YOU CAN VOIR DIRE HIM.
1461
KATZ ~- VOIR DIRE
1 VOIR DIRE EXAMINATION
BY MR. FORD: Pr
3 @. DOCTOR KATZ. IN ORDER TQ REACH YOUR ULTIMATE CONCLUSIONS
4 ON THIS SUBJECT WITH THIS DATA, WAS IT NOT NECESSARY FOR YOU TO
35 SET A SECOND LEVEL OF HYPOTHESIS. WHICH YOU ULTIMATELY TESTED
8 A WITH REGARD TO THE NUMBER OF AGGRAVATING CIRCUMSTANCES IN WHITE-
7 BLACK VICTIM CASES AND THE SIGNIFICANCE OF THAT REGARDING
a8 DISCRIMINATION?
P A. WHAT WOULD THAT HYPOTHESIS BE THAT YOURE REFERRING TO?
10 @. IS NOT THE HYPOTHESIS ON THIS SUBJECT THAT YOU ULTIMATELY
11 TEST WITH THIS DATA THE HYPOTHESIS THAT IF HIGHER LEVELS OF
12 AGGRAVATION ARE TOLERATED FOR DEFENDANTS IN BLACK VICTIM CASES
13 AS COMPARED TQ DEFENDANTS IN WHITE VICTIM CASES BEFORE A HIGHER
14 SENTENCE IS SOUGHT OR IMPOSED, THEN MANY DEFENDANTS OF
15 AGGRAVATED BLACK VICTIM CASES WILL HAVE RECEIVED LESSER
1& SENTENCES THAN DEATH. IS THAT NOT THE HYPOTHESIS YOU ULTIMATELY
17 TEST WITH THIS DATA?
is A. THE HYPOTHESIS I ULTIMATELY TEST IS. ARE HIGHER LEVELS OF
W 1? AGGRAVATION AND MITIGATION PRESENT IN BLACK VICTIM CASES AS
20 COMPARED TO WHITE VICTIM CASES.
21 A. AND ANYWHERE IN PROFESSOR BALDUS REPORT DOES IT SAY THAT IF
22 HIGHER LEVELS OF AGGRAVATION ARE PRESENT IN WHITE VICTIM CASES,
23 THAT NECESSARILY MEANS ANYTHING ONE WAY OR THE OTHER WITH REGARD
24 TO DISCRIMINATION?
29 A. NO.
wd
KATZ ~- VOIR DIRE
@. THAT’S A CONCLUSION THAT YOU REACH YOURSELF, IS THAT RIGHT?
MS. WESTMORELAND: YOUR HONOR, I DONT BELIEVE DOCTOR
KATZ HAS TESTIFIED ABOUT ANY CONCLUSIONS AT THIS TIME.
BY MR. FORD:
@. ANY CONCLUSION ON THAT WOULD BE SOMETHING YOU REACHED
YOURSELF?
THE COURT: WHAT HE-S TRYING TO FIND OUT. MS.
WESTMORELAND, IS WHERE HE GOT HIS HYPOTHESIS HE STATED IN
RESPONSE TO YOUR QUESTION EARLIER.
MS. WESTMORELANDz I BELIEVE HIS TESTIMONY WAS THAT HE
OBTAINED IT FROM PROFESSOR BALDUS” EARLY PRELIMINARY REPORT.
THE COURT: IS THAT STATED IN HAEC VERBA IN HIS EARLY
REPORT?
MR. FORD: THERE“S TWO HYPOTHESES. YOUR HONOR.
THE COURT: LET’S START OUT WITH THE ONE HE STATED FROM
THE STAND. IS THAT STATED ESSENTIALLY AS THE WITNESS STATED IT
IN THE NOVEMBER OR THE REPORT HE RECEIVED IN NOVEMBER OF “82
FROM PROFESSOR BALDUS., YES OR NO.
MR. FORD: I DON‘T KNOW OF A REFERENCE THAT STATED
PRECISELY THOSE TERMS. I CAN‘T REPRESENT TO THE COURT THAT
THATS THE ~—
THE COURT: HOW ABOUT IT. DOCTOR KATZ?
THE WITNESS: YES, YOUR HONOR. I DON‘T, I MIGHT BE
ABLE TO EVEN QUOTE THE REFERENCE. |
MR. FORD: I“M INFORMED BY PROFESSOR WOODWORTH THAT
14463
KATZ - VOIR DIRE
1 THERE IS A STATEMENT SIMILAR TO THE, SUBSTANTIALLY SIMILAR TO
2 THIS STATEMENT THAT’S COME FROM THE STAND AT THIS POINT IN THE
REPORT.
MY QUESTION IS WITH REGARD TD THE SECOND LEVEL OF
3
4
3 HYPOTHESIS THAT TESTS THAT, YOUR HONOR, AND I THINK THAT’S WHERE
& WE HAVE TO GET --
7 THE COURT: ALL RIGHT.
8 HAVE YOU GOT IT WRITTEN DOWN. AND IS YOUR HANDWRITING
@ LEGIBLE OR IS IT TYPED?
10 MR. FORD: I HAVE IT WRITTEN DOWN. YOUR HONOR, AND I
1} HAVE IT TYPED, AND ITS LEGIBLE AND IT CONTAINS AN ERROR.
12 THE COURT: WELL, CORRECT THE ERROR AND HAND IT TQ THE
13 WITNESS, AND THEN LETS LET HIM DECIPHER. ABSORB IT FOR A MOMENT
14 AND THEN WELL GO FROM THERE.
1% MR. FORD: SHOULD THIS BE MARKED, YOUR HONOR? I WAS
1& ULTIMATELY PLANNING ON USING IT.
17 THE COURT: I DON’T REALLY CARE. WHATEVER YOURE
13 COMFORTABLE WITH.
ol 1¢ MR. FORD: WE‘VE GOT A LOT OF PIECES OF PAPER AROUND
20 HERE, YOUR HONOR. I THINK A NUMBER WOULD BE APPROPRIATE.
23 THE COURT: ALL RIGHT. HAS IT BEEN MARKED?
ne MR. FORD: IT’S BEEN MARKED AS JK-2, YOUR HONOR, AND I
3 HAVE A COPY FOR THE CQURT AND A COFY FOR MS. WESTMORELAND AND A
24 COPY FOR DOCTOR KATZ OF THE MARKED EXHIBIT ITSELF.
235 VOIR DIRE EXAMINATION
KATZ - VOIR DIRE
BY MR. FORD:
@. IS THAT NOT THE HYPOTHESIS?
THE COURT: WAIT A MINUTE. IF HE UNDERSTANDS IT, HE'S
QUICKER THAN 1 AM, AND HE MAY BE. LET ME STUDY IT A MINUTE.
THE COURT: DOCTOR KATZ?
BY MR. FORD: .
@. IS THAT NOT A CORRECT PARAPHRASE AND INDEED AN EXACT QUOTE,
EXCEPT THE PORTION IN PARENTHESIS, OF THE HYPOTHESIS THAT YOU
SET OUT TO TEST ON PAGE 157 OF YOUR REPORT?
I‘M SORRY, EXACT QUOTE WITH CERTAIN PORTIONS DELETED,
BUT ISN“T THAT WHAT YOU TESTIFIED, DOCTOR KATZ?
A. THIS OUTLINES THE METHODOLOGY THAT I EMPLOYED TO TEST
PROFESSOR BALDUS” HYPOTHESIS.
@. WITH REGARD TO THAT STATEMENT, THEN IT IS AN ACCURATE
STATEMENT OF THE HYPOTHESIS. THE SECOND LEVEL HYPOTHESIS YOU SET
OUT IN YOUR REPORT. IS THAT FAIR?
A. IT’S A STATEMENT OF THE METHODOLOGY IM USING TO TEST
PROFESSOR BALDUS” HYPOTHESIS.
@. IS IT A STATEMENT THAT ANYBODY ELSE STATES OTHER THAN YOU,
TO YOUR KNOWLEDGE?
THE COURT: DID YOU GET IT FROM OUT OF PROFESSOR
BALDUS” --—
THE WITNESS: NO.
THE COURT: OR, IS THIS A PROFESSOR OR A DOCTOR?
DOCTOR WONDWORTH,
“w
i
i
e
A
R
o
d
38
9
10
14465
KATZ - VOIR DIRE
OR IS IT ORIGINAL WITH YOURSELF?
THE WITNESS: THE METHODOLOGY I DEVELOPED?
THE COURT: THE STATEMENT.
THE WITNESS: YES, THAT S A PARAPHRASE OF THE STATEMENT
I MADE. YES.
THE COURT: NOW WE HAVE THAT ESTABLISHED.
BY MR. FORD:
(= THIS IS THE PROPOSITION THAT ULTIMATELY YOU TEST OR YOU
PURPORT TO TEST WITH YOUR STATISTICS, IS THAT NOT RIGHT?
A. NO, THIS IS THE METHODOLOGY THAT I°VE USED TO TEST PROFESSOR
BALDUS” HYPOTHESIS.
@. IN ORDER TO TEST PROFESSOR, YOU SAY IT’S PROFESSOR BALDUS”
|HYPOTHESIS. YOU IN FACT, IN YOUR REPORT GO THROUGH THE
REASONING OF SETTING OUT THIS STATEMENT. IS THAT NOT RIGHT?
A. THATS SETTING OUT THE METHODOLOGY 1M EMPLOYING IN TESTING
THE HYPOTHESIS.
THE COURT: IS THIS AN INTERMEDIATE STEP OR IS IT A
FINAL STEP IN A FINAL TEST OF THE HYPOTHESIS?
THE WITNESS: GIVEN THE HYPOTHESIS THAT I OUTLINED,
THIS IS A METHODOLOGY THAT I HAVE DEVELOPED TO TEST THAT
HYPOTHESIS.
| THE COURT: IS IT INTERMEDIATE OR FINAL? BASED ON THE
RESULTS OF THIS ANALYSIS, DO YOU HAVE AN OPINION, OR DO YOU GO
TO ANOTHER STEP?
THE WITNESS: THIS IS, THIS WOULD BE THE FINAL
ka
14646
KATZ - VOIR DIRE
METHODOLOGY.
THE COURT: OKAY. NOW I THINK I UNDERSTAND WHAT WE
HAVE.
WHY IS IT OBJECTIONABLE?.
MR. FORD: WELL, YOUR HONOR. MAY 1 ASK ONE MORE
QUESTION?
THE COURT: SURELY.
BY MR. FORD:
QR. THE WORD "MANY", IS THAT A TERM OF STATISTICAL ART?
DOCTOR KATZ?
A. I GUESS IT CAN BE. IT DEPENDS ON THE CIRCUMSTANCES. I
IMAGINE, AS TO WHAT KIND OF PRECISION YOU CAN ATTACH TO A
STATEMENT OR, S0 IT”S, IT’S NOT AS PRECISE AS OTHER WORDS. BUT
SOME CIRCUMSTANCES, THAT IS THE BEST INFORMATION. THE BEST
CONCLUSIONS YOU CAN COME TO.
MR. FORD: YOUR HONOR, IT-S OUR POSITION THAT THIS IS
THE CRITICAL POINT AT WHICH DOCTOR KATZ” ANALYSIS TURNS AWAY
FROM HIS AREA OF EXPERTISE AND THE CRITICAL POINT AS TO WHICH IT
BECOMES TOTALLY ERRONEOUS IN THE ANALYSIS OF RACIAL
DISCRIMINATION IN THIS CASE.
HE HAS TESTIFIED THAT THIS IS A PROPOSITION THAT HE AND
HE ALONE CONSTRUCTED AND IT IS NOT A STATISTICAL PROPOSITION. IT
IS A LOGICAL PROPOSITION THAT IS BASED ON ASSUMPTIONS WITH
REGARD TO THE MEANING OF DISCRIMINATION. AND THE MANNER OF
TESTING A SOCIAL SCIENCE HYPOTHESIS THAT ARE NOT WITHIN THE
14467
KATZ - DIRECT
REALM OF THIS WITNESS‘ EXPERTISE, AND THAT'S THE GROUNDS FOR MY
oo
k
DBJECT ION.
THE COURT: I AM NOT PERSUADED ON THE FIRST PART OF
YOUR ARGUMENT, I.E., THAT HES DEPARTED FROM HIS EXPERTISE.
I REMAIN OPEN ON THE SECOND PART. MR. FORD.
FOR THE MOMENT. I THINK I WILL NOT SUSTAIN YOUR
OBJECTION.
LIKE MUCH ELSE, I“VE SIMPLY GOT TO SEE WHERE WERE
3
©
~~
¢
UG
$
O
P
N
GOING, AND SINCE WE“RE TRYING TO THIS NONJURY, SOME CF YOUR
10 OBJECTIONS MAY END UP GOING TD WEIGHT, RATHER THAN EXPERTISE.
11 BUT I ASK YOU TO BEAR WITH ME WHILE I FIND WHAT IS TO COME.
12 MR. FORD: SO THAT I WON’T BE INTERRUPTING, MAY I HAVE
13 A GO, CONTINUING OBJECTION ALONG THIS LINE?
14 THE COURT: YOU BEST NOTE YOUR OBJECTION BECAUSE IT IS
15 SUCH A FINE LINE THAT AT SOME POINT YOU MAY FEEL WE HAVE A
16 GREATER DEVIATION FROM THE MEAN THAN OTHERS, AND I WOULD MOST
17 SPECIALLY APPRECIATE YOUR POINTING THAT OUT TO ME.
13 : MR. FORD: ALL RIGHT, YOUR HONOR.
ee 19 THE COURT: BUT I THINK YOU BEST OBJECT ALONG.
20 |BY MS. WESTMORELAND:
21 §. DOCTOR KATZ, REFERRING AGAIN TO RESPONDENT'S EXHIBIT,
22 RESPONDENTS EXHIBITS 25 AND 26, WHAT DID THESE PARTICULAR
23 TABLES REFLECT IN LIGHT OF YOUR ANALYSIS?
24 A. THESE TABLES REFLECT THAT THERE ARE MANY VARIABLES THAT HAVE
25 BEEN PREVIAUSLY DEFINED IN TABLE &, HIM AGGRAVATING VARIABLES.
14463
KATZ - DIRECT
THAT WHITE VICTIM CASES HAVE A HIGHER PROPORTION OF THESE
ATTRIBUTES, AND BLACK VICTIM CASES HAVE A HIGHER PROPORTION OF
THE MITIGATING ATTRIBUTES.
I UTILIZED THIS TABLE IN TRYING TO DETERMINE HOW THIS
WOULD AFFECT THE TEST THAT PROFESSOR BALDUS HAD CONDUCTED ON THE
|DATA FOR THE PROCEDURAL REFORM STUDY AND I CONCLUDED THAT --
MR. FORD: YOUR HONOR, OBJECTION, BEYOND THE SCOPE OF
THE QUESTION, I BELIEVE. E |
THE COURT: YEAH. IT IS. BEYOND WHERE I WANT TO BE.
LET ME. IT WOULD BE MOST USEFUL IF YOU WOULD REFER TO
THE DOCUMENTS BY EXHIBIT NUMBERS, TO BEGIN WITH, -- ALL RIGHT,
ASK YOUR QUESTION.
MS. WESTMORELAND: YOUR HONOR, I THINK MY QUESTION WAS
BASICALLY WHAT DOCTOR KATZ HAD CONCLUDED FROM AN EXAMINATION OF
THE STATISTICAL STUDIES HE HAD DONE IN RESPONDENT'S EXHIBITS 25
AND 24.
THE COURT: CONCLUDED ABOUT WHAT?
MS. WESTMORELAND: IN REGARD TO THE HYPOTHESIS THAT HE
WAS TESTING.
THE COURT: THE VALIDITY OF IT?
MS. WESTMORELAND: WHETHER THESE TABLES SUPPORTED THAT
HYPOTHESIS, OR INDICATED ANYTHING TO HIM IN RELATION TO THAT
HYPOTHESIS.
THE COURT: ALL RIGHT, DOCTOR KATZ, WHAT'S YOUR ANSWER?
THE WITNESS: THESE TABLES ARE VERY EARLY IN THE ACTUAL
14469
KATZ - DIRECT
1 TESTING OF PROFESSOR BALDUS< HYPOTHESIS. SO 1 DID NOT COME TO
2 ANY CONCLUSIONS OR, BASED ON HIS HYPOTHESIS, BASED ON THESE
3 PARTICULAR TABLES.
4 BY MS. WESTMORELAND:
5 @. WERE THESE TABLES UTILIZED TQ DEVELOP YOUR LATER ANALYSES?
[3
] A. YES. THEY WERE.
7 MS. WESTMORELAND: YOUR HONOR. ON THAT BASIS,
& |UTILIZATION TO TEST LATER. TO DEVELOP LATER ANALYSES AND AS A
9 FOUNDATION FOR DOCTOR KATZ” ANALYSES IN GENERAL, I WOULD SUBMIT
10 |RESPONDENT’S EXHIBIT 25 AND RESPONDENTS EXHIBIT 26 AT THIS
11 TIME.
12 THE COURT: AS THE COURT HAS DONE IN DOCTOR BALDUS‘ I
13 WILL ADMIT IT TO SHOW PROCESS AND FOUNDATION DATA. BUT NOT AS
14 |EVIDENCE OF THE VALIDITY OR INVALIDITY OF THE PROPOSITION.
15 |EITHER AS STATED BY DOCTOR KATZ, OR AS STATED BY PROFESSOR
16 |BALDUS.
17 BY MS. WESTMORELAND?
13 |@. DOCTOR KATZ. AFTER YOU DID THIS INITIAL PRELIMINARY
- 19 |PREPARATION OF TABLES, DID YOU PROCEED TO EXAMINE THE PARTICULAR
20 TECHNIQUES EMPLOYED BY PROFESSOR BALDUS?
21 A. YES. 1 DID.
22 |@. AND HOW DID YOU TEST THOSE TECHNIQUES, OR WHAT TECHNIQUES
23 DID YOU EXAMINE?
24 A. THE OVERALL APPROACH THAT PROFESSOR BALDUS HAS EMPLOYED IN
<3 BOTH STUDIES IS AN ATTEMPT, GENERALLY, TO MATCH SIMILAR CASES IN
EE NESSES.
0H
o
~
O
a
1470
KATZ - DIRECT
TERMS OF AGGRAVATING AND MITIGATING CIRCUMSTANCES. AND THEN
COMPARE THE DEATH SENTENCE RATES BETWEEN WHITE VICTIM AND BLACK
VICTIM CASES.
RESPONDENTS EXHIBIT 26 SUGGESTS THERE ARE A LOT OF
VARIABLES THAT PERHAPS NEED TO BE CONSIDERED BEFORE TWO
SUB-POPULATIONS OF WHITE VICTIM BLACK VICTIM CASES ARE PROPERLY
MATCHED.
AN EXAMPLE OF THE TYPE MATCHING PROCEDURES THAT WERE
EMPLOYED BY PROFESSOR BALDUS IS INDICATED IN RESPONDENT'S
EXHIBIT 27.
Q. AND WOULD YOU IDENTIFY RESPONDENTS EXHIBIT 27, TABLE 107
A. YES. RESPONDENT‘S EXHIBIT 27. TABLE 10, IS A TABLE THAT I
CONSTRUCTED TO UPDATE AN EXPERIMENT THAT WAS PERFORMED BY
PROFESSOR BALDUS EARLIER.
THIS IS A TABLE CONTROLLING FOR EACH STATUTORY
AGGRAVATING FACTOR.
IN THE FIRST COLUMN, I LIST THE STATUTORY AGGRAVATING
FACTORS.
THE SECOND COLUMN, I GIVE THE DEATH SENTENCE RATE FOR
WHITE VICTIM CASES. SO OUT OF 26 WHITE VICTIM CASES IN WHICH A
PRIOR CAPITAL RECORD. Bi, WAS INDICATED, 13 RECEIVED DEATH
SENTENCES.
THE NUMBER TQ THE LEFT OF THAT IS THE PERCENT.
THE THIRD COLUMN IS THE DEATH SENTENCE RATE FOR BLACK
VICTIM CASES, S80 OUT OF 15 BLACK VICTIM CASES IN WHICH El WAS
wi
Q
e
C
e
W
N
8
?
10
1471
KATZ — DIRECT
INDICATED. ONE RECEIVED A DEATH SENTENCE. AND NEXT TO THAT. ON
THE LEFTHAND SIDE. IS THE PERCENT REPRESENTATIVE FOR THAT.
COLUMN 4 GIVES THE PERCENT DIFFERENCE.
AND COLUMN S GIVES THE "Z" VALUE FOR THIS PERCENT
DIFFERENCE.
RA. AND DID YOU UTILIZE THIS TABLE IN ANY FASHION IN ANY
SUBSEQUENT ANALYSES OR DID YOU DO ANY SUBSERUENT ANALYSES BASED
ON THIS TABLE?
A. YES. ACCORDING TO THIS TABLE. THERE ARE MANY SIGNIFICANT
DIFFERENCES INDICATED WHEN WHITE AND BLACK VICTIM CASES ARE
MATCHED IN TERMS OF EACH STATUTORY AGCORAVATING FACTOR. AND MY
CONSIDERATION WAS WHETHER THE ACTUAL SUB-POPULATIONS THAT WERE
BEING MATCHED WERE SIMILAR WITH REGARD TO ALL AGGRAVATING AND
MITIGATING FACTORS, AND IF IN FACT THEY WERE COMPARABLE.
TO ANALYZE THAT. 1 SELECTED FOR EACH OF THE STATUTORY
AGGRAVATING FACTORS IN WHICH THE SIGNIFICANCE WAS INDICATED, AND
FOR THOSE SUB-POPULATION OF CASES I COMPARED THE FROPORTIONS FOR
EACH OF THE VARIABLES THAT I DEFINED EARLIER IN RESPONDENTS
FXHIRIT
@. 237
A. 23, LIST OF VARIABLES. TO SEE HOW THESE FACTORS WOULD RELATE
RETWEEN WHITE AND BLACK VICTIM CASES.
“L* FOR THOSE CASES WHICH HAD THE PARTICULAR
AGGRAVATING CIRCUMSTANCE.
R. IS THIS A SIMILAR TYPE OF ANALYSIS TO THAT WHICH WE-/VE JUST
a J
Fa
1&6
pp —— lr er Se fortrtint
1472
KATZ -~ DIRECT
EXAMINED IN RESPONDENTS EXHIBIT 23579
MR. FORD: I OBJECT TO THAT AS LEADING, YOUR HONOR.
THE COURT: IT“S DIRECTORY. ILL OQVERRULE YOUR
OBJECTION.
THE WITNESS: YES, RESPONDENTS EXHIBIT 28 THROUGH -——
BY MS. WESTMORELAND:
@. WE’LL GO THROUGH AND IDENTIFY THEM IN A MOMENT. BUT WOULD
youl TELL US IF IT7°5 A SIMILAR TYFE ANALYSIS?
A. YES. CORRELATIONS WERE USED TO COMPUTE THESE "ZI" VALUES.
RATHER THAN PERCENT DIFFERENCES. BUT THE MEANING OF CORRELATION
1S SIMILAR, IN THE CONTEXT OF ZERO ~ 1 VARIABLES.
MS. WESTMORELAND: YOUR HONOR. BEFORE WE PROCEED INTO
LATER TABLES. I WOULD LIKE TO SUBMIT RESPONDENT” S EXHIBIT 27 TO
SHOW PROCESS, AGAIN, AS A STEP IN THE PROCEDURE THAT DOCTOR KATZ
UTILIZED IN TESTING TECHNIQUES BY PROFESSOR BALDUS.
THE COURT: FOR THAT PURPOSE. ITLL BE ADMITTED.
BY MS. WESTMORELAND:
Q. ALL RIGHT, DOCTOR KATZ, YOU WERE BEGINNING TO REFER TO LATER
TABLES.
COULD YOU IDENTIFY RESPONDENTS EXHIBIT 287
A. YES. RESPONDENT’S EXHIBIT 28 IS A TABLE I PREPARED.
@. TABLE 11, I“M SORRY. PROCEED?
A. IN WHICH I CHOSE ALL THE CASES REPRESENTED FROM RESPONDENTS
EXHIBIT 27 FOR THE STATUTORY AGGRAVATING FACTOR. B2, IN WHICH
THERE WERE 194 WHITE VICTIM CASES AND 70 BLACK VICTIM CASES.
~N
um
B
W
Le
@
10
1473
KATZ —- DIRECT
THE COURT: SEVEN?
THE WITNESS: EXCUSE ME, YOUR HONOR. SEVENTY BLACK
VICTIM CASES.
AND I COMPARED THEM IN TERMS OF ALL THE VARIABLES
DEFINED EARLIER IN RESPONDENTS EXHIBIT 23.
TABLE 11, RESPONDENT'S EXHIBIT 28, LISTS THE VARIABLES
THAT SHOW STATISTICAL SIGNIFICANCE AT THE .035 LEVEL.
BY MS. WESTMORELAND:
fl. AND HOW DID YOU DETERMINE THE STATISTICAL SIGNIFICANCE IN
THIS PARTICULAR TABLE?
A. AGAIN, IF THE "Z" VALUE WAS GREATER THAN 1.643, THAT WOULD
INDICATE THAT WHITE VICTIM CASES HAD A HIGHER PROPORTION OF THAT
ATTRIBUTE THAN BLACK VICTIM CASES. HELD ON ALL THOSE CASES IN
WHICH STATUTORY AGGRAVATING CIRCUMSTANCE B2 WAS PRESENT.
IF A “Z" VALLE WAS MINUS 1.445 OR LESS, THAT WOULD
INDICATE THAT BLACK VICTIM CASES HAD A HIGHER PROPORTION OF THAT
ATTRIBUTE THAM WHITE VICTIM CASES, HELD FOR THOSE --
THE COURT: AT THE .05 LEVEL.
THE WITNESS: AT THE .05 LEVEL, YES, YOUR HONOR.
BY MS. WESTMORELAND:
2. WOULD YOU, THEN, JUST LOOK AT RESPONDENT S EXHIBIT 2%, TABLE
12, AND IDENTIFY THAT EXHIBIT. PLEASE?
THE COURT: NOW, WAIT A MINUTE. WHAT I3 287 WHY I3 IT
DIFFERENT FROM ANYTHING ELSE I“VE SEEN?
MS. WESTMORELAND: YOUR HONOR, THIS TABLE IS A
1474
KATZ —- DIRECT
1 REFLECTION OF AN EXAMINATION OF THE CASES IN WHICH STATUTORY
2 AGGRAVATING CIRCUMSTANCE B2 OCCURRED.
3 THE COURT: THIS IS A LIST OF THE AGGRAVATING AND
3 MITIGATING CIRCUMSTANCES THAT APPEARED IN ALL B2 CASES?
5 THE WITNESS: THIS IS A LIST OF, ALL B2 CASES WERE
Ri IA COMPARED FOR WHITE AND BLACK VICTIM CASES FOR ALL THE VARIABLES
7 |LISTED IN RESPONDENT'S EXHIBIT 23.
a THESE ARE THE VARIABLES SUBSEGUENTLY THAT WERE SHOWN TO
9 BE SIGNIFICANT AT THE .0S LEVEL FROM THAT EARLIER LIST OF
10 VARIABLES IN RESPONDENT’S EXHIBIT 23. SO THESE ARE A LIST OF
11 VARIABLES SIGNIFICANT AT THE .0S LEVEL OF SIGNIFICANCE.
2 THE COURT: DID YOU RECOMPUTE THE “Z" VALUE WITHIN THE
13 SUBSET OF B2 CASES OR ARE THESE THE "ZI" VALUES FROM 237
14 THE WITNESS: THESE ARE THE "ZI" VALUES RECOMPUTED
15 TWENTY SUBSET OF B2 CASES.
16 BY MS. WESTMORELAND:
17 Q. AND DOCTOR KATZ, DID YOU DO A SIMILAR TYPE OF COMPARISON FOR
13 ANY OTHER STATUTORY AGGRAVATING FACTORS?
Ww 19 THE COURT: WAIT A MINUTE. THE BOTTOM LINE IN THIS
20 TABLE IS THAT GENERALLY IN B2 CASES THE WHITE VICTIM CASES ARE
21 MORE AGGRAVATED AND THE BLACK VICTIM CASES ARE MORE MITIGATING,
22 PORT OF WHICH I BELIEVE I“VE HEARD BEFORE FROM PROFESSOR BALDUS.
23 15 THAT RIGHT. DOCTOR KATZ, IS THAT ABOUT ALL YOU COULD CONCLUDE
24 FROM THAT TABLE?
Pin THE WITNESS: YES, YOUR HONOR.
Pa
TH
RE
E
a
n
a
I
T
T
pe
o
11
13
14773
KATZ - DIRECT
THE COURT: OKAY. LETS HAVE OUR MORNING RECESS.
WE“LL BE IN RECESS UNTIL TEN MINUTES AFTER ELEVEN.
(RECESS TAKEN.)
THE COURT: ON THE BREAK, I GAVE DOCTOR WOODWORTH A
COPY OF SOMETHING I HAVE PUT TOGETHER WHICH I ENTITLED “LAWYERS
MODEL" AND I°M ABOUT TO GIVE YOU A COPY OF IT.
THIS IS A FIRST EFFORT BY ME TO LOOK AT THE
QUESTIONNAIRE WITH SOMETHING OF A FROSECUTOR’S VIEW AND PERHAPS
SOMETHING OF A DEFENSE LAWYER’S VIEW, AND ATTEMPT FROM THE DATA
ON THE QUESTIONNAIRE TO DEFINE A SUBSET OF CASES WHICH SOMEONE
IN THE CRIMINAL JUSTICE SYSTEM WITH EXPERIENCE MIGHT EXPECT
WOULD HAVE A BETTER THAN AVERAGE CHANCE OF AT LEAST BEING DEATH
ELIGIBLE.
I DON’T CLAIM THAT IT IS PERFECT, BUT I HAVE DONE IT
ONCE, AND DOUBLE-CHECKED IT.
MY QUESTION TO DOCTOR WOODWORTH WAS, OR MY REQUEST OF
HIM WAS TO LOOK AT IT AND TELL ME IF CONSIDERING HIS DATA AND
HIS PROGRAMS. IF HE COULD RUN IT. TO SHARE IT WITH PROFESSOR
BALDUS TO SEE IF PROFESSOR BALIUS HAD ANY FEELING THAT
CONCEPTUALLY IT WAS IN SOME FASHION FLAWED. AND OBVIOUSLY THE
SAME THING QUESTIONS GO TO YOU.
IF THERE ARE NO COMPUTER LIMITATIONS ON RUNNING IT. AND
IF SOME GLARING MAJOR CONCEPTUAL PROBLEM ISNT RUN. I HAD ASKED
DOCTOR WOODWORTH TO RUN IT.
| a
ad
CSNY
+)
J
RE
T
e
BE
E
BR
T
T
l
BR
1476
KATZ - DIRECT
HIS RESPONSE WAS THAT THE DATA BASE AT SYRACUSE MIGHT
HAVE TO BE USED AND IT MIGHT BE MONDAY BEFORE WE COULD GET IT
DONE.
THE, IT 13 A 3~PAGE PAGE DOCUMENT, AND I-LL HAVE ONE OF
THEM MARKED BY THE CLERK AS C-1.
I ASKED HIM TO RUN IT IF POSSIBLE. AS A FUNCTION, RUN
IT TWICE MAYBE 3 TIMES, THE LAST DIVISION BEING AS FUNCTION OF
A. STATUS OF DEFENSE COUNSEL AS BETWEEN PRIVATE APPOINTED. OR
APPOINTED STATUS UNKNOWN ON THE ONE HAND, VERSUS COUNSEL COMING
TO THE CASE FROM SOME CIRCUMSTANCE WHICH MIGHT INDICATE A HIGHER
LEVEL OF EXPERTISE THAN SIMPLY BEING A PRIVATE APPOINTED COUNSEL
FOR THE COURT KNOWS OF THE PROCLIVITIES OF THE SUPERIOR COURTS
IN THIS STATE TO APPOINT VERY YOUNG LAWYERS IN CRIMINAL CASES,
WHEN THEY“RE MAKING APPOINTMENTS FROM THE PRIVATE BAR.
I THOUGHT IT INTERESTING TO SEE IF THAT HAD ANYTHING TO
DO WITH THE DUTCOME OR HANDLING WITHIN THE SYSTEM.
I TOLD DOCTOR WENTWORTH THAT CONCEPTUALLY WHAT I WAS
DOING WAS RUNNING A CROSS TABULATION TO WINNOW DOWN THE CASES. I
BELIEVE THAT’S HIS TERM, TO THE SUBSET AND TO USE WHATEVER
ANALYTICAL TOOLS USED TO ATTEMPT TO MEASURE RACE EFFECTS.
HERE ARE TWO COPIES ARE FOR YOU, ONE FOR YOU AND ONE
FOR DOCTOR KATZ.
HERE“S ONE FOR THE CLERK TO MARK C-1. WE WON"T LEAVE
IT IN THE RECORD IF NOTHING COMES OF IT, BUT JUST SO THE RECORD
WILL HAVE IT. ILL PUT IT THERE FOR THE MOMENT.
1477
KATZ - DIRECT
1 G0 AHEAD.
3 JOSEPH LAUREN KATZ:
3 BEING PREVIOUSLY DULY SWORN, RESUMED THE WITNESS STAND AND
3 TESTIFIED FURTHER AS FOLLOWS:
e 4 DIRECT EXAMINATION (CONT D)
7 M5. WESTMORELAND: YOUR HONOR, ALONG THOSE SAME LINES,
3 YOU HAD INDICATED WE MIGHT SEEK TO DO A SIMILAR TYPE OF THING
9 WITH THE PROSECUTOR AND WE“RE MAKING INQUIRIES INTO THAT
10 PARTICULAR AREA AT THIS POINT. IT HAS NOT BEEN DONE. BUT -——
i1 THE COURT: I TOLD MR. BOGER. OF COURSE, HE COULD DO
12 THE SAME THING. GQ AHEAD.
13 BY MS. WESTMORELAND:
14 Mm. DOCTOR KATZ, 1 BELIEVE WE HAVE PREVIOUSLY REFERRED TQ AND
13 IDENTIFIED RESPONDENT’S EXHIBIT 28 WHICH DEALT WITH STATUTORY
1&4 AGGRAVAT ING CIRCUMSTANCE, B2. AND I THINK I HAD ASKED OR WAS
17 GETTING READY TO ASK YOU IF YOU HAD DONE A SIMILAR ANALYSIS
12 REGARDING ANY OF THE OTHER STATUTORY AGGRAVATING CIRCUMSTANCES
ho. 19 LISTED IN RESPONDENT'S EXHIBIT 27.
20 A. YES. FROM RESPONDENT’S EXHIBIT 27. I CHOSE ALL THESE
23 STATUTORY AGGRAVATING CIRCUMSTANCES THAT SHOWED A SIGNIFICANT
22 "zu VALUE AT THE .05 LEVEL OF SIGNIFICANCE. AND I ARRIVED AT A
23 SIMILAR ANALYSIS FOR THOSE SUB-POPULATIONS GENERATED.
z4 R. DID YOU PRESENT. FREPARE TABLES?
pad A. YES. 1 PREPARED TABLES THAT COMPARED WHITE AND BLACK VICTIM
1473
KATZ - DIRECT
fy
CASES. HELD FOR STATUTORY AGGRAVATING FACTOR Bil.
@. COULD YOU RESPOND TO RESPONDENT'S EXHIBIT 29, PLEASE, WHICH
IS TABLE 12, AND IDENTIFY THAT DOCUMENT?
A. YES, THESE ARE THE RESULTS OF THAT EXPERIMENT. THIS TABLE
LISTS ALL THE VARIABLES OUT OF THE TOTAL NUMBER OF VARIABLES
REPRESENTED IN RESPONDENT’S EXHIBIT 23, THAT SHOWED A
~
a
A
»
L
N
SIGNIFICANCE. STATISTICAL SIGNIFICANCE AT THE .0S LEVEL.
z @. AND THEN WOULD YOU REFER TO RESPONDENTS EXHIBIT 30, WHICH
9 15 TABLE 13 AND IDENTIFY THAT DOCUMENT, PLEASE?
10 A. IN RESPONDENT’S EXHIBIT 30. ALL CASES IN WHICH B3 OCCURRED
11 WERE COMPARED IN TERMS OF RACE OF THE VICTIM AND THIS TABLE
12 LISTS THOSE VARIABLES OUT OF THE ORIGINAL VARIABLES IN
13 RESPONDENTS EXHIBIT 23. WHICH SHOWED A SIGNIFICANT PROPORTION
14 OR MEAM DIFFERENCE AT THE .0S LEVEL OF SIGNIFICANCE.
is THE COURT: B37
16 THE WITNESS: B3. YES. YOUR HONOR.
17 EY M3. WESTMORELAND:
13 QR. THEN WOULD YOU REFER TO RESPONDENTS EXHIBIT 31, TABLE 14,
Ww 15 AND IDENTIFY THAT?
20 A. YES. A SIMILAR ANALYSIS WAS PERFORMED FOR ALL CASES IN
21 WHICH B4 WAS INDICATED TO HAVE OCCURRED. AND THIS LISTS THE
22 VARIABLES THAT ARE SIGNIFICANT AT THE .0S LEVEL FOR THOSE CASES,
23 IN COMPARING WHITE AND BLACK VICTIM CASES.
24 @. THEN IF YOU WOULD REFER TO RESPONDENT-S EXHIBIT 32. TABLE
23 1S, AND IDENTIFY THAT DOCUMENT. FLEASE.
~
4
p
W
8
?
10
1479
KATZ ~ DIRECT
A. IN RESPONDENTS EXHIBIT 32, I GIVE A LIST OF VARIABLES THAT
ARE SIGNIFICANT AT THE .0S5 LEVEL THAT COMPARE RACE OF VICTIM,
THAT COMPARE THESE VARIABLES IN TERMS OF THE RACE OF THE VICTIM
FOR ALL CASES IN WHICH B7 OCCUR,
Q. WOULD YOU LOOK AT RESPONDENT'S EXHIBIT 33. TABLE 14. AND
IDENTIFY THAT DOCUMENT, PLEASE?
A. IN RESPONDENTS EXHIBIT 33, 1 COMPARED ALL VARIABLES FROM
RESPONDENT’S 23 IN TERMS OF ALL CASES IN WHICH BY OCCURRED.
COMPARING WHITE AND BLACK VICTIM CASES. AND I LIST THOSE
VARIABLES THAT ARE SIGNIFICANT AT THE .0S LEVEL.
0. AND THEN WOULD YOU LOOK FINALLY AT RESPONDENTS EXHIBIT 34,
TABLE 17 AND IDENTIFY THAT DOCUMENT?
A. RESPONDENTS EXHIBIT 34, I COMPARE ALL THE VARIABLES THAT
WERE INITIALLY DEFINED IN RESPONDENTS EXHIBIT 23 FOR ALL CASES
IN WHICH B10 DCCURRED AND I LIST THOSE VARIABLES AT THE, THAT
ARE SIGNIFICANT AT THE .0% LEVEL.
Q. AND THEN. ARE ALL OF THESE, THESE LATTER TABLES,
RESPONDENTS EXHIBITS 30 THROUGH 34, THE SAME TYPE OF TEST AS
PREVIOUSLY DONE AND TALKED ABOUT IN RESPONDENT'S EXHIBIT 297
A. YES.
0. AND IN OBSERVING THE TABLES. WHAT OBSERVATIONS DID YOU MAKE
FROM THESE PARTICULAR TABLES?
A. AFTER CONTROLLING FOR A PARTICULAR STATUTORY AGGRAVATING
CIRCUMSTANCE, WHITE VICTIM CASES HAD A NUMBER OF AGGRAVATING
CIRCUMSTANCES IN WHICH THEY HAD, AT A HIGHER PROPORTION OF THAT
KATZ - DIRECT
l
e
AGGRAVATING VARIABLE THAN BLACK VICTIM CASES: AND BLACK VICTIM
CASES TENDED TO HAVE MORE MITIGATING FACTORS THAN WHITE VICTIM
CASES AT THE LEVEL OF SIGNIFICANCE.
MS. WESTMORELAND: YOUR HONOR. AT THIS TIME I WOULD
SUBMIT FOR EVIDENCE RESPONDENT’S EXHIBITS 28, 29, 30. 31. 32, 33
AND 24, REFLECTING THE ANALYSIS BASED ON THE STATUTORY
AGGRAVATING FACTORS OF THE PRESENCE OF EACH INDIVIDUAL STATUTORY
AGGRAVATING FACTOR.
MR. FORD: SAME, OBJECTION, YOUR HONOR, ON RELEVANCE.
po
is
Iw
S
E
E
BR
T
R
E
B
E
E
BE
E
T
E
B
R
Sa
THE COURT: ALL RIGHT. THEY’LL BE ADMITTED. THAT'S 29 +—
33 THAT 3 28, AND 2% THROUGH 347?
i2 MS. WESTMORELAND: YES, YOUR HONOR, THAT’S CORRECT.
13 BY MS. WESTMORELAND:
14 A. DOCTOR KATZ, DID YOU, AFTER MAKING THIS PARTICULAR ANALYSIS,
13 HAVE A TECHNIQUE UTILIZED BY PROFESSOR BALDUS IN HIS EARLY
1& WORKS, DID YOU MAKE AN ANALYSIS OF ANY OTHER TECHNIQUES EMPLOYED
17 BY PROFESSOR BALDUS?
13 A. YES, I DID. I DID AN ANALYSIS OF THE COMPARISONS OF WHITE
. 12 AND BLACK VICTIM CASES, CONTROLLING FOR THE NUMBER OF STATUTORY
20 AGGRAVATING CIRCUMSTANCES PRESENT IN THE CASE.
21 @. AND WOULD YOU REFER TO RESPONDENTS EXHIBIT 35, TABLE 18 AND
22 IDENTIFY THAT DOCUMENT. PLEASE?
<3 A. YES. RESPONDENTS EXHIBIT 35 IS AN UPDATE OF AN EARLIER
24 TABLE THAT WAS DONE BY PROFESSOR BALDUS IN THE PRELIMINARY
25 REFORT, IN WHICH I LIST A NUMBER OF STATUTORY AGGRAVATING
1421
KATZ - DIRECT
1 CIRCUMSTANCES THAT ARE BEING HELD FOR IN THE LEFT-HAND COLUMN.
< SECOND COLUMN GIVES THE DEATH SENTENCING RATES FOR
3 WHITE VICTIMS IN A SIMILAR FASHION AS THE EARLIER TABLE DEALING
4 WITH STATUTORY AGGRAVATING CIRCUMSTANCE COUNTS.
S THE THIRD COLUMN GIVES THE DEATH SENTENCING RATES FOR
& BLACK VICTIM CASES.
7 THE FOURTH COLUMN GIVES THE PERCENT DIFFERENCE.
a AND THE FIFTH COLUMN GIVES THE ASSOCIATED "ZI" VALUE
? ATTACHED TO THAT PERCENT DIFFERENCE.
10 RQ. IS THIS TABLE BASED ON DATA FROM THE PROCEDURAL REFORM
11 STUDY?
12 A. YES.
13 @. AND ONCE YOU OBTAINED THE INFORMATION IN THIS TABLE. WHAT
14 DID YOU DO WITH THIS INFORMATION?
15 A. LOOKING AT THE "Z" VALUES, I FOUND TWO INSTANCES IN WHICH
16 SIGNIFICANT DIFFERENCES WERE INDICATED AT THE .05 LEVEL oF
17 SIGNIFICANCE. THESE CASES ARE WHEN 3 STATUTORY AGGRAVATING
18 CIRCUMSTANCES ARE PRESENT AND FOUR STATUTORY AGGRAVATING
1% CIRCUMSTANCES ARE PRESENT.
<0 MS. WESTMORELAND: YOUR HONOR, BEFORE WE PROCEED TO THE
21 LATER TABLES. I WOULD LIKE TO INTRODUCE RESPONDENT’S EXHIBIT 33
22 AT THIS TIME FOR THE PURPOSES OF SHOWING THE PROCESS. AND THE
23 BASIS FOR THE SUBSEQUENT ANALYSIS.
24 MR. FORD: WE HAVE NO OBJECTION TO THIS ONE, YOUR
23 HONOR.
1482
KATZ - DIRECT
1 THE COURT: JUDGE HILL NOW OF THE ELEVENTH CIRCUIT
= MAKES AN ELOQUENT TALK ON THE JUST BECAUSE YOU CAN OBJECT.
THAT DOESN‘T ALWAYS MEAN YOU SHOULD. I THINK YOU HAVE JUST
DEMONSTRATED THAT HYPOTHESIS VERY WELL.
MS. WESTMORELAND: I WAS NOT SURPRISED.
THE COURT: IT WILL BE ADMITTED.
BY MS. WESTMORELAND:
R. DOCTOR KATZ, YOU WERE INDICATING THAT YOU HAD DONE
R
E
+
E
Y
B
R
A
S
E
SUBSEQUENT ANALYSIS BASED ON RESPONDENTS EXHIBIT 35.
10 WOULD YOU LOOK AT RESPONDENT'S EXHIBIT 3& AND IDENTIFY
11 THAT DOCUMENT, PLEASE?
12 A. YES. RESPONDENTS —-
13 G. TABLE 19. EXCUSE ME?
14 A. RESPONDENT’S EXHIBIT 26, COMPARES CASES IN WHICH EXACTLY 3
135 STATUTORY AGGRAVATING CIRCUMSTANCES CCCURRED IN TERMS OF ALL THE
16 VARIABLES EARLIER DEFINED IN RESPONDENT-S EXHIBIT 23. AND LIST
17 THOSE VARIABLES THAT ARE SIGNIFICANT AT THE .03 LEVEL OF
13 SIGNIFICANCE.
& 1% @. IS THIS A SIMILAR TYPE OF ANALYSIS TO THAT WE JUST DISCUSSED
20 WITH REGARD TO THE PRESENCE OF EACH INDIVIDUAL STATUTORY
i AGGRAVATING FACTOR?
<2 A. YES, 1T 1%.
23 GB. AND DID YOU DO ANY ANALYSIS CONCERNING THE PRESENCE OF FOUR
24 STATUTORY AGGRAVATING CIRCUMSTANCES?
25 A. YE3. 1 DID.
SN
BE
ES
¢
SE
E
1
B
o
n
B
R
E
EE
KATZ - DIRECT
0. AND WOULD YOU RESPOND TO RESPONDENTS EXHIBIT 37. PLEASE.
A. RESPONDENT’S EXHIBIT NUMBER 37 ARE THE RESULTS OF MY
ANALYSIS IN COMPARING CASES IN WHICH EXACTLY FOUR STATUTORY
AGGRAVATING CIRCUMSTANCES OCCURRED.
G. AND THAT’S WHAT I BELIEVE HAS TABLE 20 ON THAT DOCUMENT?
A. YES.
@. AND WHAT OBSERVATIONS DID YOU MAKE FROM THESE PARTICULAR
TABLES? ;
A. AFTER HOLDING FOR THREE OR FOUR STATUTORY AGGRAVATING
CIRCUMSTANCES. WHITE VICTIM CASES STILL HAD SOME ADDITIONAL
AGGRAVATING CIRCUMSTANCES IN WHICH THEY HAD A HIGHER PROPORTION
OF, AS COMPARED TO BLACK VICTIM CASES: AND BLACK VICTIM CASES
STILL HAD A HIGHER PROPORTION OF MITIGATING CIRCUMSTANCES AS
COMPARED TO WHITE VICTIM CASES AT THE .0S LEVEL OF SIGNIFICANCE.
9. AND BOTH OF THESE TABLES ARE INDICATIONS OF THE “Z" VALUES
WHICH ARE SIGNIFICANT?
2. IS THAT WHAT YOUZRE INDICATING?
MS. WESTMORELAND: YOUR HONOR, AT THIS TIME. WE WOULD
SUBMIT RESPONDENTS EXHIBITS 34 AND 37.
MR. FORD: SAME RELEVANCE OBJECTION. YOUR HONOR.
THE COURT: ALL RIGHT, THEY‘LL BE ADMITTED.
R. DOCTOR KATZ, IN EXAMINING, AND IN DOING YOUR ANALYSES ON
PROFESSOR BALDUS” WORK, DID YOU LOOK AT THE INDEX METHOD
1434
KATZ - DIRECT
UTILIZED? Po
ni
A. YES, I PID. k2
3 @. COULD YOU PLEASE DESCRIBE OR TELL US WHAT THE PURPOSE OF AN
4 INDEX METHOD IN GENERAL IS?
3 A. THE PURPOSE OF AN INDEX METHOD IS TO DEVISE SOME STANDARD BY
A é WHICH OBSERVATIONS OF CASES WILL BE JUDGED FOR THE PURPOSE OF
? RANKING THESE CASES APPROPRIATELY.
a GENERALLY AN INDEX IS AN ARTIFICIAL MEASURE OF SOME
@ PROCESS IN WHICH NO EASILY OBTAINED MEASURE IS AVAILABLE TO. TO
10 ALLOW FOR THAT PROCESS OR TO MEASURE THAT PROCESS.
11 @. AND WHAT DOES THE INDEX METHOD DETERMINE? WHAT IS
12 DEVELOPED BY THESE INDEXES?
13 A. THE PURPOSE OF THE INDEX IS TO RANK DIFFERENT CASES, OR
14 DIFFERENT OBSERVATIONS SQ THAT ONE COULD CONCLUDE THAT FERHAPS
15 ONE CASE WAS, HAD MORE OF THAT ATTRIBUTE OR MORE OF THAT MEASURE
14 [THAN ANOTHER CASE.
17 @. WHAT KIND OF OUTCOMES ARE DETERMINED BY THE INDEX METHOD?
18 A. THE INDEX METHOD IS TRYING TO PROVIDE AN ARTIFICIAL RANKING.
w 17 IN THIS CASE, THE INDEXES WERE USED FOR, AN INDEX
20 TRYING TO CONTROL FOR AGGRAVATING AND MITIGATING CIRCUMSTANCES.
21 50 THE QUTCOMES WOULD BE NUMBERS THAT PURPORT TD REPRESENT THE
22 DEGREE OR LEVEL OF AGGRAVATION AND MITIGATION IN EACH CASE FUR
23 THE PURPOSE OF RANKING THESE CASES ACCORDING TO THOSE NUMBERS.
<4 THE COURT: I UNDERSTAND THE CONCEPT, BUT IT IE NOT
== CLEAR TO ME AS TO WHAT INDICES OF PROFESSOR BALDUS” DOES THIS
KATZ —- DIRECT
1 ADDRESS.
2 BY MS. WESTMORELAND:
3 @. DOCTOR KATZ, COULD YOU ADDRESS -—- ARE YOU -—-
THE COURT: OR WHICH.
BY MS. WESTMORELAND?
Q. DOCTOR KATZ, ARE YOU ADDRESSING PARTICULAR INDICES AT THIS
POINT OR THE INDEX METHOD IN GENERAL?
A. AT THIS POINT THE INDEX METHOD IN GENERAL.
SUE
«SR
Y
J
i
EE
©. AND HAVE YOU ADDRESSED ANY SPECIFIC INDICES AND ANALYZED ANY
10 OF THOSE PARTICULAR INDICES?
11 A. YES.
12 | MS. WESTMORELAND: I THINK, YOUR HONOR, AT THIS POINT
13 THE PURPOSE IS MERELY A GENERAL DISCUSSION OF THE INDEX METHOD.
14 BY MS. WESTMORELAND:
13 DN. DOES THE INDEX METHOD UTILIZE A PREDICTED QUTCOME OR AN
146 ACTUAL OUTCOME OR WHAT TYPE OF VALUE IN THAT REGARD?
i7 A. THERE ARE MANY DIFFERENT WAYS ONE CAN CONSTRUCT AN INDEX.
18 FROFESSOR BALDUS HAS USED REGRESSION ANALYSIS TO DEVELOP HIS
Ww 1? INDICES.
20 WHAT HE. WHAT 1 BELIEVE HE HAS DONE IS HE HAS RUN
21 REGRESSION ANALYSES AND USED A PREDICTED OUTCOME WHICH WERE THE
22 RESULT OF THE REGRESSION ANALYSIS TO FORM HIS INDEX FOR
23 AGGRAVATION AND MITIGATION.
24 @. WHAT WOULD BE THE DIFFERENCE IN A PREDICTED OUTCOME AND AN
23 ACTUAL OUTCOME IN THAT CONTEXT?
1436
KATZ - DIRECT
1 A. THE ACTUAL OUTCOME IS THE OUTCOME THAT HAS BEEN OBSERVED FOR
2 THE PARTICULAR VARIABLE.
3 THE REGRESSION PROVIDES A MODEL FOR THAT ACTUAL
OUTCOME. AND FROM THAT MODEL. WILL GIVE A PREDICTED OUTCOME,
BASED ON THE INDEPENDENT VARIABLES THAT IT HAS TO WORK WITH.
@. WOULD YOU REFER TO RESPONDENTS EXHIBIT 39, WHICH IS TABLE
4
5
&
7 21A. DOCTOR KATZ. AND IDENTIFY THAT DOCUMENT. FLEASE?
8 A. YES, RESPONDENT'S EXHIBIT NUMEER 29 GIVES A SERIES OF
@ REGRESSIONS THAT I RAN FOR THE FIRST ONE HUNDRED CASES IN THE
0 PROCEDURAL REFORM STUDY.
11 THE COURT: HUH?
12 THE WITNESS: FOR THE FIRST ONE HUNDRED CASES IN THE
13 PROCEDURAL REFORM STUDY.
14 BY MS. WESTMORELAND:
15 Q. ARE THESE REGRESSIONS THAT YOU DEVELOPED OR HOW WERE THESE
14 REGRESSIONS OBTAINED AND WHAT WAS THE PURPOSE OF DOING THIS?
17 A. THE PURPOSE ~~ |
18 THE COURT: EXCUSE ME. WHAT TABLE ARE YOU ON?
ww 19 | MS. WESTMORELAND: TABLE 21A, YOUR HONOR, RESPONDENTS
20 EXHIBIT 29. IT SHOULD BE IN THE NOTEBOOK.
2% THE COURT: ALL RIGHT.
22 THE WITNESS: THE PURPOSE OF THESE REGRESSIONS IS TO
22 SHOW HOW THE INDEX METHOD CAN BE SHAPED TO GIVE DIFFERENT
24 RANKINGS FOR, FOR DIFFERENT CASES. THESE REGRESZIONS ARE ONLY
<3 FOR THE PURPOSE OF DEMONSTRATION RATHER THAN AS A MEASURE OF AN
1487
KATZ ~ DIRECT
1 ACTUAL AGGRAVATION-MITIGATION INDEX.
2 ®R. IS THIS -- GO AHEAD. I‘M SORRY.
3 A. AND I LIMITED IT TO A HUNDRED CASES 50 THAT THE NUMBER OF
OBSERVATIONS THAT I WOULD HAVE TO TABULATE IN THE TABLES WOULD
BE LIMITED TO ONLY A HUNDRED AND NOT BURDEN ME AND OTHERS WITH
ALL 407 OBSERVATIONS.
80 FOR PURPOSES OF ILLUSTRATION, SHOWING HOW THE INDEX
0
~N
h
b
METHOD WOULD RESPOND TO PROFESSOR BALDUS® USE OF PREDICTED
7 OUTCOMES.
10 @. AND THEN IN RELATION TO THIS. AND WHAT YOU HAVE DONE WITH
11 THIS PARTICULAR TABLE, COULD YOU EXPLAIN HOW THE REGRESSION FITS
12 INTC THE INDEX METHOD?
13 A. YES. IF WE LOOK AT THE REGRESSION UNDER THE TITLE
14 "REGRESSION 1" AS AN EXAMPLE. THIS IS A REGRESSION WHICH RESULTS
13 FROM USING THE VARIABLE X431C, WHICH IS A VARIABLE FOR WHETHER
14 OR NOT A DEATH SENTENCE WAS IMPOSED. AS A DEPENDENT VARIABLE.
17 AND THE USE OF SIX INDEPENDENT VARIABLES REPRESENTING AGGRAVATED
1= BATTERY, AGGRAVATED MOTIVE. AND SO FORTH.
w 12 NOW, REGRESSION ANALYSIS TRIES TO EXPLAIN OR MODEL THE
20 DEPENDENT VARIABLE. IN THIS CASE, BEING THE DEATH SENTENCE
23 QUTCOME, BY THE INDEPENDENT VARIABLES THAT ITS GIVEN. AND IT
22 DOES THIS BY TRYING TO MAKE THE PREDICTED QUTCOMES, WHICH ARE
23 THE OUTCOMES THAT WOULD BE OBTAINED FOR EACH OF THE INDIVIDUAL
23 ONE HUNDRED CASES AFTER APPLYING THESE WEIGHTS IN TERMS OF WHAT
23 FEATURES OF THE OFFENSE WERE PRESENT IN THOSE CASES. AND SUMMING
14a3
KATZ ~ DIRECT
i UP THESE WEIGHTS AND TRYING TO HAVE THESE WEIGHTS AS CLOSE AS
2 POSSIBLE TO THE ACTUAL OUTCOME.
3 S0 FOR EXAMPLE. IF WE HAD A DEATH SENTENCE CASE, THE
4 VALUE FOR X481C WOULD BE 1, AND IF IN THAT DEATH SENTENCE CASE
ba’ WAS PRESENT AN AGGRAVATED MOTIVE AND BURGLARY, WHICH IS
& B-U-R-G-A-R~5, AS THE TWO VARIABLES. THEN THE PREDICTED OUTCOME
7 FOR THAT CASE WOULD BE .103 THE CONSTANT TERM. NOTHING, NO
8 |WEIGHT WOULD BE ADDED IN TERMS OF THE AGGRAVATED BATTERY
? VARIABLE, A-G-G-B-A-T~-T SINCE IT DID NOT QCCUR IN THIS
10 PARTICULAR CASE, THEN .3% WOULD BE ADDED FOR THE AGGRAVATED
3 MOTIVE VARIABLE, A-G-G-M-0-T, AND .38 WOULD BE ADDED FOR THE
is BURGLARIES VARIABLE, B-U-R-G-A-R-3.
13 : THE OTHER THREE VARIABLES WOULD HAVE ZERO VALUES AND SO
14 THOSE WITNESSES WOULDN’T BE COUNTED. S00 THE RESULTANT PREDICTED
13 OUTCOME FOR THIS PARTICULAR CASE, WOULD BE THE SUM OF .10 FLUS
1% .3%, PLUS .38, WHICH I BELIEVE IS .83.
17 NOW, THE CRITERION BY WHICH THESE WEIGHTS ARE ASSIGNED
19 IS BASED ON THE CRITERION THAT QVERALL OBSERVATIONS CHOOSE THE
w» 19 WEIGHTS OR CO-EFFICIENTS THAT MINIMIZE THE SUM OF THE
20 DIFFERENCES BETWEEN THE PREDICTED OUTCOMES AND THE ACTUAL
21 QUTCOMES. SOUARED.
22 THE REASON THE SQUARING PROCEDURE IS EMPLOYED IS TO. IS
23 THE SAME REASON THAT-S USED IN THE COMPUTATION OF VARIANCE, 30
24 THAT WE'RE MINIMIZING THE SUM OF POSITIVE VALUES AND WE DON’T
itn] HAVE POSITIVE AND NEGATIVE VALUES CANCELLING OUT.
4]
3
20
1489
KATZ —- DIRECT
2. IN THIS PARTICULAR EXPERIMENT THAT YOUVE CONDUCTED. WHAT IS
THE SIGNIFICANCE OF THE LATER REGRESSIONS THAT YOU‘VE LISTED OR
PROVIDED?
A. THE LATER REGRESSIONS ARE REGRESSIONS WHERE ADDITIONAL
VARIABLES WERE ADDED TO THE SET OF INDEPENDENT VARIABLES, AND
FOR THE PURPOSE OF RAISING THE R-SRUARED VALUE.
@. 1 NOTICE YOU UTILIZE R-SQUARE IN EACH OF THESE REGRESSIONS,
AND WE TALKED ABOUT R-SGUARE IN SOME OF THE OTHER WITNESSES’
TESTIMONY. |
COULD YOU GIVE A BRIEF EXPLANATION OF R-SQUARE AND ITS
SIGNIFICANCE?
A. THE R-SGUARED IS A MEASURE OF HOW CLOSE THE PREDICTED
OUTCOMES REFLECT THE ACTUAL OUTCOMES. AND ITS ESSENTIALLY THE
PROPORTION OF VARIATION OF, AS EXPLAINED BY THE REGRESSION MODEL
IN ITS PREDICTED OUTCOME, IN TERMS OF THE ACTUAL OUTCOME, IN
TERMS OF THE VARIATION OF THE ACTUAL OUTCOME.
THE PURPOSE OF REGRESSION IS TO TRY AND MAKE THE
PREDICTED OUTCOMES BE AS CLOSE AS POSSIBLE TO THE ACTUAL
OUTCOME. IF ONE IS SUCCESSFUL IN DOING THAT, ONE MIGHT SAY THEY
HAVE A REGRESSION EQUATION THAT MODELS THE SYSTEM. AND HOPEFULLY
HAS VARIABLES AND CO-EFFICIENTS THAT REPRESENT THE RELATIVE
IMPORTANCE OF THESE VARIABLES TO THE SYSTEM.
THE R-SQUARED IS A MEASURE OF HOW SUCCESSFUL ONE HAS
BEEN IN HAVING THE PREDICTED OUTCOMES BECOME VERY CLOSE TO THE
ACTUAL OUTCOME.
1420
KATZ ~ DIRECT
1 THE HIGHER THE R~SQUARED. THE CLOSER THE PREDICTED
2 QUTCOMES ARE TO THE ACTUAL OUTCOME OVER ALL THE CASES THAT ARE
BEING CONSIDERED.
®. LET ME BACKSPACE JUST A MOMENT.
WHAT IS THE DIFFERENCE BETWEEN AN ADJUSTED R-SGQUARED AN
3
4
S
a & UNADJUSTED R-SQUARED?
7 A. A UNADJUSTED R-SQUARED DOES NOT TAKE INTO ACCOUNT THE NUMBER
a OF VARIABLES THAT ARE USED IN THE SET OF INDEPENDENT VARIABLES.
? IT’S. THE UNADJUSTED R-SGUARED IS NOT TRULY A MEASURE OF THE
10 VARIANCE, THE EXPLANATION OF THE VARIANCE. THAT’S WHAT THE
11 ADJUSTED R-SQUARED DOES. BUT THE UNADJUSTED R~SQUARED IS
12 GENERALLY UTILIZED BECAUSE WE HAVE A CONVENIENT TEST TO TEST THE
13 SIGNIFICANCE oF THAT R-SQUARED VALLE.
14 GENERALLY THESE VALUES ARE APPROXIMATELY THE SAME. IF
15 YOU HAVE A LARGE NUMBER OF VARIABLES IN YOUR MODEL AS COMPARED
14 - |TO THE NUMBER OF OBSERVATIONS, THEN THE ADJUSTED R-SGUARED COULD
17 BE SIGNIFICANTLY LOWER THAN THE UNADJUSTED R~SQUARED VALUE.
18 @. WHY MIGHT THAT OCCUR?
w ie A. BECAUSE THE ADJUSTED R-SQUARED VALUE IS TRULY A RELATIVE
20 PROPORTION OF VARIANCE RATHER THAN VARIATION OF SUM OF SQUARES.
i, FOR MOST MODELS, IF THE NUMBER OF INDEPENDENT VARIABLES
22 COMPARED TO THE TOTAL NUMBER OF OBSERVATIONS IS A VERY HIGH
23 . |PERCENTAGE. THE UNADJUSTED R-SQUARED AND THE ADJUSTED R-SQUARED
24 WILL HAVE COMPARABLE VALUES.
25 Q. I BELIEVE YOU WERE DISCUSSING THE SPECIFIC REGRESSIONS THAT
1491
KATZ - DIRECT
ha HAD LISTED IN YOUR EXFERIMENT, AND THE PURFQSE OF THESE
2 ADDITIONAL REGRESSIONS.
3 WHAT IS THE PURPOSE OF REGRESSIONS 2 THROUGH 5 AND WHAT
4 DO THEY SHOW IN REGARD TO REGRESSIONS IN GENERAL?
3 A. PROFESSOR BALDUS HAS EMPLOYED HIS INDEX METHOD BY UTILIZING
s & THE PREDICTED OUTCOMES FOR A PARTICULAR REGRESSION.
7 1 BELIEVE THE STATISTICAL SIGNIFICANCE OR MEANING OF
a THIS INDEX HAS TO BE PUT IN TERMS OF HOW THE RANKINGS THAT
2 RESULT FROM PREDICTED OUTCOMES CAN BE AFFECTED BY A SET OF
10 INDEPENDENT VARIABLES THAT’S USED IN THE REGRESSION MODEL.
11 FURTHERMORE, I ALSO POINT QUT THAT AS THE R~-SQUARED
32 VALUE INCREASES, WHAT RESULTS ARE INDEX VALUES THAT RANK DEATH
13 SENTENCE CASES CLOSE TO 1 AND LIFE SENTENCE CASES WHICH HAVE
14 ZERO VALUES FOR X481C CLOSE TO ZERO.
13 @. AND THEN DO YOU, WOULD YOU LOOK AT RESPONDENTS EXHIBIT 40
1& AND IDENTIFY THAT DOCUMENT FLEASE?Y
17 A. YES. THIS IS A TABLE I PREPARED IN WHICH I COMPARED THE
18 PREDICTED QUTCOMES FOR EACH OF THE REGRESSIONS INDICATED ON
w 12 RESPONDENTS EXHIBIT NUMBER 39.
20 THE FIRST CATEGORY OF CASES ARE THE LIFE SENTENCE
21 CASES. NOW I SELECTED THE FIRST ONE HUNDRED CASES FROM THE
22 PROCEDURAL REFORM STUDY AND LISTED ON THE LEFTHAND SIDE UNDER
23 "CASE" THE CASE NUMBER.
24 THEN I LIST THE ACTUAL SENTENCE CUTCOME AND FOR LIFE
<3 SENTENCE CASES THEY ALL SHOULD BE ZERO.
1472
KATZ = DIRECT
AND THEN I LIST THE PREDICTED OUTCOME THAT WOULD BE
GENERATED BY REGRESSION 1 AND REGRESSION 2. AND REGRESSION 3.
AND REGRESSION 4 AND REGRESSION 35.
@. AND WHAT, I‘M SORRY, GO AHEAD.
A. THE SECOND PAGE CONTAINS A CONTINUATION FOR THE LIFE
SENTENCE CASES, WITH THE APPROPRIATE INDEX FREDICTED QUTCOME
DEMONSTRATED.
AND THEN THE THIRD PAGE IS THE PREDICTED QUTCOMES FOR
THE DEATH SENTENCE CASES FOR EACH OF THE FIVE REGRESSIONS.
@. COULD YOU EXPLAIN THE SIGNIFICANCE OF THESE TABLES IN LIGHT
OF THE EXPERIMENT YOU-RE DISCUSSING THAT YOU RAN?
A. WHEN PROFESSOR BALDUS RAN HIS REGRESSION ANALYSIS AND BUILT
HIS INDEXES, HE NOTED THAT THERE WERE MANY CASES THAT HAD
PREDICTED VALUES SOMEWHERE IN THE MIDDLE.
2. BY IN THE MIDDLE, WHAT DO YOU MEAN?
A. MID-RANGE CASES, CASES WHERE THERE WERE DEATH SENTENCE CASES
THAT WERE CLOSE TO .5 OR LOWER: LIFE SENTENCE CASES THAT WERE
CLOSE TO .5 OR LARGER. AND HE INTERPRETED THIS TO MEAN THAT THE
SYSTEM HAD A DIFFICULT TIME DISCRIMINATING BETWEEN THESE
CASES AS TO WHETHER A DEATH SENTENCE SHOULD BE APPLIED OR NOT.
fl. AND HOW DOES YOUR EXPERIMENT RELATE TO THAT?
A. WELL, FROM MY ANALYSIS. IT APPEARS THAT THIS, THIS MID-RANGE
OF CASES CAN BE ELIMINATED AS LONG AS ONE HAS A REGRESSION MODEL
IN WHICH THE R-SOUARED IS SUFFICIENTLY HIGH.
GQ. COULD YOU GIVE AN EXAMPLE BY LOOKING AT RESPONDENTS EXHIBIT
[3
[A
1493
KATZ - DIRECT
40 WHAT YOURE REFERRING TQ?
A. YES. IF WE LOOK AT SAY REGRESSION 1, IN THIS CASE.
ACCORDING TO RESPONDENTS EXHIBIT NUMBER 39, IT —-
THE COURT: WHAT PAGE ARE YOU LOOKING AT?
THE WITNESS: TABLE 221A, YOUR HONOR.
THE COURT: PAGE? RESPONDENT’S EXHIBIT 407
THE WITNESS: YES. AND RESPONDENTS EXHIBIT 3%.
BY MS. WESTMORELAND:
Ql. DKAY, WHICH PARTICULAR PAGE OF RESPONDENTS EXHIBIT 40 ARE
YOU REFERRING TO, DOCTOR KATZ? |
A. THE FIRST PAGE OF RESPONDENT’S EXHIBIT 4Q.
IF WE LOOK AT THE PREDICTED OUTCOMES FOR THESE CASES,
UTILIZING REGRESSION 1, WHICH R-SQUARED WAS .26, WE SEE A WIDE
VARIATION OF THE POSSIBLE INDEX VALUES.
AND LOOKING AT PAGE 2, DOWN THE FIRST COLUMN, THERES
AGAIN A WIDE RANGE OF VALUES. SOME OF THESE VALUES EVEN ACHIEVE
A PREDICTED QUTCOME OF .70 AS IN THE CASE OF CASE NUMBER 066 AT
THE BOTTOM. |
RA. BY A WIDE RANGE OF VALUES, COULD YOU GIVE US AN EXAMPLE IN
THE TABLE WHAT YOU’RE REFERRING TO?
A. WELL, SOME OF THESE PREDICTED OUTCOMES HAVE VALUES AS LOW AS
-.23 AND OTHER CASES HAVE PREDICTED OUTCOMES THAT ARE AS HIGH AS
«70.
@. WOULD YOU TAKE A SPECIFIC CASE AND EXPLAIN WHAT YOURE
REFERRING TO. PLEASE?
1494
KATZ — DIRECT
A. ON THE SECOND PAGE OF RESPONDENTS EXHIBIT 40, CASE NUMBER
Dil, HAS A PREDICTED OUTCOME OF -.23.
Q. THATS WITH WHICH REGRESSION?
A. THAT IS FOR REGRESSION NUMBER 1.
@. QUT OF RESPONDENT’S EXHIBIT NUMBER 397?
A. YES. WHEREAS CASE 064 HAS A PREDICTED OUTCOME OF .70. IF
WE LOOK AT PAGE 3 OF RESPONDENT’3 EXHIBIT 40, WHERE WE LOOK AT
THE INDEX VALUES OR PREDICTED OUTCOMES FOR THE DEATH SENTENCE
CASES. WE AGAIN SEE A WIDE RANGE OF VALUES. THERE ARE CASES
SUCH AS CASE NUMBER Z15 IN WHICH A PREDICTED OUTCOME OF .100 IS
INDICATED, AND THERE ARE CASES IN WHICH A VERY LARGE VALUE
1S INDICATED, SUCH AS CASE NUMBER D18, WITH A .90 VALUE.
NOW, THERE ARE SOME LIFE SENTENCE CASES WHICH HAVE
HIGHER INDEX VALUES THAN DEATH SENTENCE CASES, AND THERE ARE
SOME DEATH SENTENCE CASES THAT HAVE LOWER, THAT HAVE LOWER INDEX
VALUES THAN LIFE SENTENCE CASES.
THE POINT IS THAT THERE I3 A MIDDLE RANGE OF CASES
WHERE LIFE CASES AND DEATH SENTENCE CASES ARE MIXED UP IN TERMS
OF THE PREDICTED OUTCOME INDEX VALUES.
THIS HAS BEEN —-
THE COURT: ON REGRESSION 1 OR ACROSS THE BOARD?
THE WITNESS: ON REGRESSION 1, YOUR HONOR.
BY MS. WESTMORELAND?
Q. AND HOW DOES THAT. DOES THAT CHANGE IN ANY WAY IN THE
SUBSEGUENT REGRESSIONS THAT YOU UTILIZED IN YOUR EXPERIMENT?
1493
KATZ - DIRECT
1 ‘A. YES. 1 THEN RAN REGRESSION 2, WHERE I ADDED ADDITIONAL
rl
VARIABLES TO THE SET OF INDEPENDENT VARIABLES AND ACHIEVED AND
R-SQUARED OF .59 AS INDICATED ON RESPONDENTS EXHIBIT NUMBER 3%.
THEN I LISTED THE PREDICTED CUTCOMES FOR INDEX VALUES
THAT WOULD BE ASSOCIATED WITH THE RESULTS FROM REGRESSION 2.
AND WHAT I OBSERVED IS THAT THE AMOUNT OF DIFFERENCES
BETWEEN THE NUMBERS ARE LESS.
@. BETWEEN WHICH NUMBERS?
A. BETWEEN THE INDEX NUMBERS FOR PREDICTED OUTCOME NUMBERS FOR
OQ
N
o
A
B
W
po
y THE LIFE SENTENCE NUMBERS. THEY TEND TO BE CLOSER TO THE ZERO
11 VALUE, WHICH IS THEIR. THE SENTENCING OUTCOME WERE TRYING TO
12 PREDICT. ALTHOUGH THERE ARE SOME CASES THAT DO HAVE LARGE INDEX
13 VALUES.
14 THEN IF WE LOOK AT THE DEATH SENTENCE CASES ON PAGE 3
135 OF RESPONDENTS EXHIBIT 40, WE FIND THAT THE INDEX VALUES FOR
16 THESE CASES TEND TO BE LARGER. ALTHOUGH THERE ARE STILL SOME
17 CASES WHICH HAVE LOW INDEX VALUES AS COMPARED TO THE LIFE
13 SENTENCE CASES.
os 19 SO EVEN IN THIS REGRESSION. WE WOULD HAVE A SET OF
20 CASES IN THE MIDDLE IN WHICH WE WOULD HAVE DEATH SENTENCE CASES
21 HAVING LOWER INDEX VALUES THAN LIFE SENTENCE CASES.
22 80 THIS REGRESSION HAS NOT COMPLETELY SEPARATED QUT THE
23 LIFE AND DEATH SENTENCE CASES.
2% IF WE MOVE ON TO REGRESSION NUMBER 3. FROM RESPONDENT 3
2X3 EXHIBIT NUMBER 29, I“VE ADDED ADDITIONAL INDEPENDENT VARIABLES.
H
W
4]
1496
KATZ —- DIRECT
THE COURT: 1 HAVE THE POINT. I FIGURED QUT WHAT
HAPPENS AS YOU GO ACROSS,
LET ME ASK, WHAT I3 INCLUDED IN REGRESSION 1 AND WHAT
IS INCLUDED IN REGRESSION S? ANYBODY ABLE TO TELL ME? HAVE
YOU. LET ME ASK THIS.
LOOKING AT 39, YOU’VE GOT A NUMBER OF INDEPENDENT
VARIABLES, SUCH AS AGGRAVATED BATTERY, AGGRAVATED MOTIVE,
BURGARS, WHICH 1 SUPPOSE MEANS BURGLARY3 OR BURGLARY-ARSON.
THE WITNESS: BURGLARY OR ARSON, YOUR HONOR.
THE COURT: OKAY. I DON’T KNOW WHAT THE NEXT ONE IS.
ANYHOW. USING THE SAME INDEPENDENT VARIABLES ON EXHIBIT 40 FOR
COLUMN 1 FOR ALL THE CASES?
THE WITNESS: YES.
THE COURT: WHAT HAVE YOU. HAVE YOU PUT IN YOUR MODEL,
REGRESSION 57? DO YOU KNOW WHAT THOSE MEAN?
THE WITNESS: YES. YOUR HONOR.
THE COURT: WHAT IS "BRIGHT?"
THE WITNESS: THE DEFENDANT HAD AN I.@. OF ONE HUNDRED
OR GREATER.
THE COURT: ALL RIGHT. AND THAT 3S BURGLARY-ARSON.
WHAT IS "DEFEM?"
THE WITNESS: THE DEFENDANT WAS FEMALE, YOUR HONOR.
THE COURT: "DULL?"
THE WITNESS: I BELIEVE THAT“S THE DEFENDANT HAD AN
I.G. BETWEEN, OR LESS THAN 45, YOUR HONOR,
1497
KATZ - DIRECT
1 THE COURT: "KIDNAP" MEANS THAT ——
THE WITNESS: THE KIDNAPPING WAS A CONTEMPORANEOUS )
OFFENSE TO THE HOMICIDE. 3
+ THE COURT: AND "MULSCH" IS MULTI-SHOOTING., I BELIEVE.
n THE WITNESS: MULTIPLE SHOTS TO THE HEAD.
THE COURT: AND THE NEXT ONE, "MUTIL?"
THE WITNESS: “MUTIL" IS MUTILATION.
THE COURT: MUTILATION. WHAT DOES "NOARREST" MEAN?
W
o
o
o
on
THE WITNESS: 1 BELIEVE THE MOTIVE FOR THE KILLING WAS
10 TD RESIST ARREST OR NOT TO BE ARRESTED.
i1 THE COURT: WHAT IS “NOKILL"?
12 THE WITNESS: DEFENDANT WAS NOT THE TRIGGER MAN.
13 THE COURT: AND "NONFCRM"?
14 THE WITNESS: THE MOTIVE FOR THE CRIME, FOR THE
13 KILLING WAS A NONPROFIT CRIME MOTIVE.
16 THE COURT: NEXT ONE?
17 THE WITNESS: NO VIOLENT PERSONAL CRIMES.
18 THE COURT: OLD VICTIM?
w 19 THE WITNESS: THE VICTIM WAS &5 OR OLDER.
20 THE COURT: STATUTORY AGGRAVATING CIRCUMSTANCES?
21 THE WITNESS: YES, SIR, YOUR HONOR.
22 THE COURT: RAPE. SERIQUS CRIME.
23 THE WITNESS: YES. YOUR HONOR. TOTAL NUMBER OF SERIOUS
24 CRIMES.
23 THE COURT: CONNECTED WITH, OR PRIOR RECORD.
1493
KATZ - DIRECT
1 THE WITNESS: PRIOR RECORD, I BELIEVE.
2 THE COURT: WHAT IS THE "STMIT2"? STATUTORY AND
2 |MITIGATING?
a THE WITNESS: I BELIEVE THAT'S WHAT IT IS, YOUR HONOR.
5 THE COURT: TWO VICTIMS. THE VIOLENCE WAS UNNECESSARY :
64 |TO CARRY OUT THE FELONY.
7 THE WITNESS: YES. YOUR HONOR.
a THE COURT: IT WAS AN URBAN CASE,
= THE WITNESS: YES. YOUR HONOR.
10 THE COURT: THE VICTIM WAS A CRIMINAL.
11 THE WITNESS: OR CRIMINAL RECORD.
12 THE COURT: WHAT IS “VICDEFS?"
13 THE WITNESS: THE VICTIM WAS DEFENSELESS.
14 THE COURT: WHAT’S THAT NEXT ONE?
15 THE WITNESS: VICTIM WAS A POLICEMAN OR FIREMAN.
16 THE COURT! WHAT IS “VICPLEA," PLED FOR MERCY?
17 THE WITNESS: YES. YOUR HONOR.
18 THE COURT: THE NEXT ONE?
w 19 THE WITNESS: 1 BELIEVE THAT’S THE NUMBER OF VICTIM
20 MITIGATING CIRCUMSTANCES -- SOMETHING TO DD WITH VICTIM
21 MITIGATING CIRCUMSTANCE. I BELIEVE THE NUMBER OF THEM.
22 THE COURT: WHAT’S THE NEXT ONE?
23 THE WITNESS: WHETHER OR NOT A VIOLENT PERSONAL CRIME
2% HAD BEEN IN HIS PRIOR RECORD, THE DEFENDANT’S FRICR RECORD.
23 THE COURT: AND THE NEXT ONE?
[o
N
bh
fe
y
W
B
a
ES
,
i
M
R
SR
S
T
1499
KATZ - DIRECT
THE WITNESS: THE NUMBER OF VIOLENT PERSONAL CRIMES.
THE COURT: IN THE VICTIM’S PRIOR RECORD?
THE WITNESS: YES, SIR. EXCUSE ME, IN THE DEFENDANT’S
PRIOR RECORD.
THE COURT: IT SAYS THE PER CRIME?
THE WITNESS: I BELIEVE THAT’S VIOLENT.
THE COURT: OKAY. THE NEXT ONE? |
THE WITNESS: THAT DISTINGUISHES WHETHER OR NOT A
PRECIPITATING EVENT OCCURRED AT THE TIME OF THE HOMICIDE.
THE COURT: AND PREMEDITATED. I GUESS.
THE WITNESS: THE KILLING WAS PREMEDITATED.
THE COURT: ALL RIGHT.
BY MS. WESTMORELAND:
Q. DOCTOR KATZ, ~—
MS. WESTMORELAND: I-‘M SORRY, YOUR HONOR.
THE COURT: GO AHEAD.
BY MS. WESTMORELAND:
a. DOCTOR KATZ,» DID YOU UTILIZE ANY SPECIFIC METHOD IN CHOOSING
THE FACTORS TO ADD INTO THESE REGRESSIONS?
A. NO. 1 SIMPLY WANTED TO DEMONSTRATE THAT AS THE R-SQUARED
INCREASES, THE ACTUAL. THE PREDICTED OUTCOMES WILL BECOME CLOSER
AND CLOSER TO THE ACTUAL OUTCOMES.
AND THE PURPOSE OF THIS WAS TO SHOW THAT AT ANY STAGE,
WHAT IS HAPPENING WITH THE REGRESSION IN TERMS OF THE
INDEPENDENT VARIABLES IT HAS AVAILABLE TQ IT. IS THAT IT IS
ww)
a
1300
KATZ - DIRECT
TRYING TO WEIGHT THE VARIABLES OR ASSIGN CO-EFFICIENTS TO THE
VARIABLES SO THAT THE PREDICTED OUTCOMES FOR THE LIFE SENTENCE
CASES WILL HAVE ZERO VALUES AND THE PREDICTED QUTCOMES FOR THE
DEATH SENTENCE CASES WILL HAVE ONE VALUE. REGARDLESS OF THE
INDEPENDENT VARIABLES THAT IT HAS TO WORK WITH.
MS. WESTMORELAND: YOUR HONOR, AT THIS TIME, I WOULD
LIKE TO SUBMIT RESPONDENTS EXHIBITS 3% AND 40 TO ILLUSTRATE
DOCTOR KATZ TESTIMONY AS AN EXPERIMENT HE HAS CONDUCTED AND FOR
THAT PURPOSE ONLY.
MR. FORD: CERTAINLY AS AN EXPERIMENT. WE WOULD OBJECT.
YOUR HONOR. THIS IS AS I UNDERSTAND IT, IS NOT THE WHOLE DATA
SET. THIS IS NOT ON THE DATA IN THIS CASE, THIS IS A HAND
PICKED SAMPLE.
IM STILL NOT CLEAR AS TO HOW DOCTOR KATZ CAME UP WITH
THESE VARIABLES, WHETHER HE THREW AT A DART BOARD, LET THE
COMPUTER CHOOSE THEM OR WHAT. BUT IT DOESN’T SEEM TO ME. THIS
IS A PURELY ILLUSTRATIVE THING TO SAY SOMETHING ABOUT REGRESSION
EQUATIONS IN GENERAL. AND HOW THEY WORK ON SMALL, MANAGEABLE
DATA SETS. 1 DON’T KNOW THAT THIS IS AN EXPERIMENT THAT HAS
ANYTHING TO DO WITH THIS CASE. I CERTAINLY OBJECT ON THIS
FOUNDATION.
MS. WESTMORELAND: I BELIEVE I ASKED THAT IT BE
INTRODUCED TO ILLUSTRATE DOCTOR KATZ” TESTIMONY IN THIS REGARD.
AND PERHAPS "HIS EXPERIMENT" IS NOT A VERY GOOD CHOICE OF
WORDS.
KATZ —- DIRECT
THE COURT: WELL, EXPERIMENT IN THE INTERRELATIONSHIP
OF NUMBER VARIABLES AND R-SGUARE, YES: EXPERIMENT TO DEMONSTRATE
ANYTHING ABOUT THE ENTIRE DATA SET. NO.
MS. WESTMORELAND: THATS NOT THE PURPOSE. YOUR HONOR.
IT’S TO ILLUSTRATE THE TESTIMONY CONCERNING R-SQUAREDS AND
REGRESSIONS.
: THE COURT: FOR THAT PURPOSE, IT“5 ADMITTED.
MS. WESTMORELAND: THANK YOU, YOUR HONOR.
BY MS. WESTMORELAND:
A. DOCTOR KATZ, HOW DO REGRESSION EQUATIONS REACT TO ERRORS IN
THE PREDICTOR VARIABLES THAT MAY APPEAR?
A. WELL, THE REGRESSION EQUATIONS TREAT THE INFORMATION IT HAS
AS ONLY, IGNORES THE ERRORS AND JUST GOES BY THE RULE OF TRYINO
TO MINIMIZE THE SUM OF THE SQUARE OF THE DIFFERENCES BETWEEN THE
ACTUAL AND PREDICTED OUTCOMES.
30 IF THERE ARE ERRORS THAT ARE INTRODUCED INTO THE SET
OF VARIABLES, THE REGRESSION THAT RESULTS FROM THAT ANALYSIS CAN
BE FAULTY AND CAN BE SUBJECT TO A GREAT DEAL OF ERROR.
Ql. WHAT WOULD BE GENERALLY ACCEPTED STATISTICAL PROCEDURE IN
UTILIZING REGRESSIONS INSOFAR AS TREATING UNKNOWN VALUES?
A. I BELIEVE THERE IS A PROBLEM IN USING REGRESSION IN THE KIND
OF ANALYSIS PERFORMED BY PROFESSOR BALDUS BECAUSE THERE ARE A
GREAT DEAL OF UNKNOWNS.
THE UNKNOWN VALUES WERE CODED TO HAVE ZERO VALUES, AND
THE USUAL STATISTICAL METHOD OF DEALING WITH UNKNOWNS I5 TO OMIT
KATZ ~ DIRECT
1 THOSE OBSERVATIONS IN THE PERFORMING OF THE REGRESSION ANALYSIS.
2 IT IS HOPED THAT WHEN REGRESSION ANALYSIS IS RUN THAT
3 THERE ARE VERY FEW INDEPENDENT OR DEPENDENT VARIABLES THAT WOULD
HAVE MISSING VALUES.
SINCE IN THE DATA SET THAT PROFESSOR BALDUS HAD
4
%
iB & AVAILABLE TO HIM THERE WERE A CREAT DEAL OF MISSING VALUES. AND
7 HIS REGRESSION HAD BEEN PERFORMED WHERE THOSE VALUES HAD BEEN
3 INDICATED AS MISSING, USING THE STANDARD CONVENTION OF OMITTING
Q OBSERVATIONS IN WHICH EVEN ONE MISSING VALUE IS INDICATED FOR AN
10 INDEPENDENT OR DEPENDENT VARIABLE, THERE WOULD HAVE BEEN VERY
11 FEW OBSERVATIONS LEFT IN WHICH THE ANALYSIS COULD HAVE BEEN
12 CONDUCTED WITH.
13 THE COURT: ARE YOU ABOUT TO CHANGE SUBJECTS?
14 MS. WESTMORELAND: YES, YOUR HONOR.
13 THE COURT: LET ME ASK YOU A COUPLE QUESTIONS ABOUT
16 THAT PIECE OF TESTIMONY.
12 BY THE COURT:
18 @. DO YOU UNDERSTAND, AS DO I, THAT A CODING OF "U" MEANT THAT
w 19 THERE WAS SOME EVIDENCE IN THE FILE TO INDICATE THAT THAT
20 VARIABLE MIGHT BE IMPLICATED IN THE CASE, BUT THE CODER WAS
23 UNABLE TO MAKE A CONCLUSIVE DETERMINATION AND SO HE PUT A "U"
22 THERE. IT IS NOT LIKE NOT EXISTENT,. IN QTHER WORDS?
23 A. YES, YOUR HONOR. THAT S MY UNDERSTANDING.
24 @. 50 IN GIVING THE TESTIMONY, THAT IS THE UNDERSTANDING THAT
<3 YOU ARE OPERATING ON?
1503
KATZ - DIRECT
A. YES, YOUR HONOR.
@. WE THINK OF —-— ALL RIGHT. NOW, ARE YOU SAYING THAT WHERE
THERE IS A “U" IN MAKING A COUNT, IF YOU WILL, THAT THE FACT
THAT THERE IS ONE “U“, SHOULD THROW OUT THAT ONE CASE. OR THE
FACT THAT THERE IS ONE "U" SHOULD THROW OUT THAT VARIABLE?
A. THE FACT THERE IS THAT ONE "U" SHOULD THROW OUT THAT CASE
FROM CONSIDERATION IN THE ANALYSIS. |
Q. SO THAT AFTER YOU HAVE THROWN QUT THE "U“S" THEN You ARE
COMPARING POSITIVE RESPONSES WITH AN ABSENCE OF THAT VARIABLE?
A. YES.
@. WHERE THE RESEARCHER OR THE CODER HAS PUT DOWN A “U",
INDICATING THE POSSIBILITY THAT THE DATA IS PRESENT. IS THERE,
IN YOUR KNOWLEDGE, A STATISTICAL CONVENTION WHICH WOULD ADDRESS
THE PROPRIETY OF CODING THAT AS ZERO. AS ABSENT. AS OPPOSED TO
PRESENT?
A. NOT THAT I-“VE BEEN ABLE TQ FIND.
I HAVE FOUND --
RA. WHAT IM SAYING IS, YOU COULD ARGUE THAT THERE IS A BARE
INFERENCE THAT IT IS PRESENT, OTHERWISE HE WOULDN‘T HAVE PUT
DOWN ANYTHING ABOUT IT.
A. I SUSPECT ALSO THAT PERHAPS THERE WAS JUST NO INDICATION
ABOUT THAT VARIABLE CONTAINED IN THE FILE.
@. WELL, THAT”S NOT MY UNDERSTANDING. AND THAT S WHY I ASKED
YOU THIS QUESTION. MY UNDERSTANDING IS IF THERE WAS NO
INFORMATION ABOUT THAT VARIABLE, HE PUT DOWN A ZERO. BUT IF
1304
KATZ ~ DIRECT
THERE WAS SOME INFORMATION THAT COULD SUGGEST THE VARIABLE WAS
PRESENT. HE PUT DOWN A "U",
A. WELL. THEN, I’M NOT CERTAIN THAT THAT’S MY UNDERSTANDING OF
HOW THESE THINGS WERE CODED. I“M NOT SURE EXACTLY HOW THEY
DECIDED WHETHER TO CODE IT A "U", OR “NOT PRESENT" IF THERE WAS
NO INDICATION ABOUT THAT PARTICULAR ITEM IN THE FILE.
Q. AS I STATED ON SEVERAL OTCASIONS: MY RECOLLECTION OF MR.
GATES” TESTIMONY, AND THE EXAMPLE THAT WE WERE CONSIDERING,
LEADS ME TO THE CONCLUSION, AT THIS POINT, IF MY MEMORY IS RIGHT
THAT THEY WERE USING THE “U" WHEN THERE WAS A POSSIBILITY THAT
THAT. THAT YOU COULD. THAT THERE WAS A POSSIBILITY THAT THAT
VARIABLE WAS OPERATIVE IN THAT CASE.
OPERATING ON MY ASSUMPTION. RATHER THAN YOURS, DOES
THAT CHANGE YOUR TESTIMONY?
A. IN TERMS OF THE, HOW TO DEAL WITH THE MISSING VALUES?
@. UH HUH?
A. NO.
THE COURT: ALL RIGHT.
BY MS. WESTMORELAND:
Q. DOCTOR KATZ, MOVING FROM REGRESSIONS AT THIS TIME. DID YOU
EXAMINE THE CROSS TABULATION PROCEDURE UTILIZED BY PROFESSOR
BALDUS?
Ia. YES, 1 DID,
Q. AND COULD YOU JUST TELL US STATISTICALLY WHAT ARE GENERALLY
PROBLEMS THAT MAY ARISE IN UTILIZING A CROSS TABULATION
1503
KATZ - DIRECT
TECHNIQUE?
A. THE CROSS TABULATION TECHNIQUE IS AN. IS AN ATTEMPT TO
MATCH CASES.
THE UNFORTUNATE PART IS THAT IT’S A VERY DATA
RAVENOUS TECHNIGUE IN THAT AS YOU MATCH CASES ON PARTICULAR
FACTORS, THE NUMBER OF CELLS IN WHICH THESE PARTICULAR CASES ARE
CLASSIFIED IN, RISE AT A VERY QUICK RATE.
AND WITH ONLY &07 CASES IN THE PROCEDURAL REFORM STUDY,
OR EVEN THE 1082 CASES IN THE CHARGING AND SENTENCING STUDY. ONE
CAN VERY QUICKLY RUN OUT OF A SUBSTANTIAL NUMBER OF CASES WITHIN
|EACH CELL TO DO MEANINGFUL STATISTICAL COMPARISONS.
@. DID YOU MAKE ANY CONCLUSIONS BASED ON THE USE OF THE CROSS
TABULATION METHOD IN THIS PARTICULAR INSTANCE?
A. 1 FELT THAT CROSS TABULATION METHODS. ALTHOUGH A 500D.
EFFECTIVE TOOL TO USE, WAS NOT GOING TO SUCCEED IN THIS KIND OF
ANALYSIS, BECAUSE OF THE PROBLEM OF JUST &07 CASES.
FURTHERMORE, THERES AN ADDITIONAL PROBLEM IN TRYING TO
CLASSIFY CASES WHEN THERE ARE UNKNOWN VALUES, AS WHETHER HAVING
A PARTICULAR ATTRIBUTE OR NOT.
S80 IT MADE IT DIFFICULT TO UTILIZE THAT TECHNIGUE.
Gl. WHAT DID YOU CONCLUDE IN YOUR ANALYSES IN REGARD TO THE USE
OF MULTIVARIATE ANALYSIS WITH THE DATA FROM THE PROCEDURAL
REFORM STUDY?
A. THE USE OF MULTIVARIATE ANALYSIS, ESPECIALLY MULTIPLE
REGRESSION, AS A TOOL TO MEASURE EFFECTS, REQUIRES THEORETICALLY
1306
KATZ - DIRECT
1 COMPLETE. CORRECT DATA TO BE UTILIZED. IT’S A VERY IMPORTANT
2 AND USEFUL TECHNIQUE IN FERRETING OUT RELATIONSHIPS IF ONE HAS
3 COMPLETE, CORRECT DATA.
OTHERWISE, THE RESULTS ARE, OR CAN BE FAULTY AND FILLED
WITH ERRORS AND NOT RELIABLE.
THE SAME KINDS OF PROBLEMS WOULD EXIST WITH OTHER
MULTIVARIATE TECHNIQUES IN THAT IN TRYING TO DETERMINE
RELATIONSHIPS BETWEEN MANY VARIABLES. THE ASSUMPTION IN THE
Lo
B
E
E
Y
ME
1
S
R
¢
B
E
THEORY OF THOSE TESTS IS THAT ONE HAS COMPLETE, CORRECT DATA.
10 |WHICH WE DID NOT HAVE AT THIS PRESENT TIME.
11 |@. DID YOU DO ANY SUBSEQUENT ANALYSIS ON THE DATA IN THE
12 |PROCEDURAL REFORM STUDY?
12 |A. YES, I DID.
14 2. WOULD YOU REFER TO RESPONDENT-S EXHIBIT 41, WHICH IS TABLE
1S |22, AND IDENTIFY THAT DOCUMENT, PLEASE?
16 |A. YES. IN THIS TABLE, I COMPARE LIFE AND DEATH SENTENCE CASES
17 |FOR ALL THE CASES IN THE CHARGING AND SENTENCING STUDY, BROKEN
18 |DOWN BY ALL THE VARIABLES THAT WERE DEFINED IN RESPONDENT'S
ww 19 |EXHIBIT 23.
20 |@. IS THIS THE, I’M SORRY, DID YOU SAY CHARGING AND SENTENCING
zi |OR PROCEDURAL?
22 |A. EXCUSE ME. THE PROCEDURAL REFORM STUDY.
23 |Q. AND WHAT DOES THIS TABLE REPRESENT? IS THIS TABLE ANY
24 |DIFFERENT FROM THE PREVIOUS TABLES WE HAVE SEEN?
pri] A. YES, THIS TABLE IS NOT COMPARING RACE OF THE VICTIM, BUT
1307
KATZ - DIRECT
1 COMPARING SENTENCE, EITHER A DEATH SENTENCE OR A LIFE SENTENCE.
@. AND COULD YOU JUST BRIEFLY EXPLAIN THE. WHAT YOU HAVE DONE
USING THIS TABLE, OR WHAT THIS TABLE REPRESENTS?
A. YES. USING THE VARIABLES DEFINED IN RESPONDENTS EXHIBIT
23. THE FIRST COLUMN LISTS THOSE VARIABLES.
THE SECOND COLUMN INDICATES QUT OF THOSE CASES IN WHICH
4
A DEATH SENTENCE WAS IMPOSED. THE PROPORTION THAT HAD EACH
©
NN
6
RB
2
W
N
PARTICULAR VARIABLE. THE THIRD COLUMN GIVES THE PERCENT
“ RELATIVE TO THE DEATH SENTENCE CASES THAT HAD THE PARTICULAR
10 VARIABLE.
11 THE FOURTH COLUMN DOES A SIMILAR COUNT FOR LIFE
12 SENTENCE CASES.
13 COLUMN 5 GIVES THE RESULTING PERCENT FOR THE LIFE
14 SENTENCE CASES.
15 COLUMN & TABULATES THE PERCEN, THE PERCENT DIFFERENCE,
146 TAKING THE PERCENT OF THE VARIABLE INDICATED IN DEATH SENTENCE
17 CASES MINUS THE PERCENT INDICATED OF THAT VARIABLE IN LIFE
1&8 SENTENCE CASES.
Ww 12 AND COLUMN 7 CALCULATES THE "I® VALUE, RELATIVE TO THAT
20 VARIABLE.
21 @. DID YOU DO ANY ANALYSIS BASED DN THIS TABLE IN YOUR
2 CALCULATIONS IN THIS TABLE?
23 A. 1 BELIEVE I JUST SET OUT THIS TABLE. AND DIDN’T DO ANY
24 FURTHER ANALYSIS, DEALING WITH THIS PARTICULAR ASPECT.
23 G. DID YOU MAKE ANY BREAKDOWNS CONCERNING RACE OF VICTIM
1308
KATZ - DIRECT
DURING, IN THE VARIOUS SENTENCING OUTCOMES?
A. YES, I DID.
R. WOULD YOU REFER TO RESPONDENT'S EXHIBIT 42, AND IDENTIFY
THAT DOCUMENT. PLEASE, WHICH IS TABLE 247
A. RESPONDENTS EXHIBIT 42 IS A TABLE THAT COMPARES WHITE AND
BLACK VICTIM CASES WHOSE DEFENDANTS RECEIVED LIFE SENTENCES FOR
ALL THE VARIABLES DEFINED IN RESPONDENTS EXHIBIT 23.
THE MAKEUP OF THE TABLE IS SIMILAR TO EARLIER TABLES IN
WHICH THE VARIABLES DEFINED IN COLUMN 1 OR THE APPROPRIATE
VARIABLE IS INDICATED IN COLUMN 1.
COLUMN 2 GIVES THE COUNTS FOR THE WHITE VICTIM CASES IN
WHICH A LIFE SENTENCE WAS IMPOSED, THAT HAS THE PARTICULAR
VARIABLE THAT IS BEING CONSIDERED.
COLUMN 3 GIVES THE PERCENT OF WHITE VICTIM CASES THAT
HAD THAT PARTICULAR VARIABLE.
COLUMN 4 DOES THE CORRESPONDING COUNTS FOR BLACK VICTIM
CASES.
COLUMN = GIVES THE PERCENT AGAIN FOR BLACK VICTIM
CAZES.
COLUMN & REPRESENTS THE PERCENT DIFFERENCE BETWEEN
WHITE AND BLACK VICTIM CASES.
AND COLUMN 7 GIVES THE "ZI" VALUE ASSOCIATED WITH THAT
VARIABLE. |
R. AND DID YOU DO A SEPARATE TABLE LISTING THE SIGNIFICANT “Z*
VALUES FROM THIS TABLE?
p-
lo
S
EL
BE
E
BR
U
R
Ta
TE
p
n
fon
t
po
KATZ - DIRECT
Bis YES, I DID.
a. WOULD YDU IDENTIFY RESPONDENT S EXHIBIT 43. PLEASE?
A. YES. IN RESPONDENTS EXHIBIT 43 1 SELECTED THOSE VARIABLES
THAT -=-
QR. I“M SORRY, THAT’S TABLE 237
A. =—— FROM RESPONDENT’S EXHIBIT 42, THAT SHOWED STATISTICAL
SIGNIFICANCE AT THE .0S5 LEVEL OF SIGNIFICANCE. AND THEY ARE
TABULATED IN RESPONDENTS EXHIBIT 43.
@. WHAT OBSERVATION DID YOU MAKE FROM THESE TWO TABLES?
A. THAT IN THE COMPARISON OF WHITE AND BLACK VICTIM CASES THAT
RECEIVED LIFE SENTENCES. THAT THERE ARE MANY MORE AGGRAVATING
CIRCUMSTANCES THAT. IN WHICH WHITE VICTIM CASES HAVE A HIGHER
PROPORTION OF THAT AGGRAVATING CIRCUMSTANCE THAN BLACK VICTIM
CASES. AND THERE ARE MANY MITIGATING FACTORS IN WHICH BLACK
VICTIM CASES HAVE MORE OF THAT MITIGATING FACTOR THAN WHITE
VICTIM CASES.
BY THE COURT:
RP. WHAT ARE THESE. LIFE CASES, DEATH CASES?
A. THESE ARE COMPARING LIFE SENTENCE CASES.
Q. WHAT DOES 42 SHOW?
A. 42 IS THE COUNTS, AND THE PERCENTS. THE ACTUAL PERCENTS FOR
EACH OF THE VARIABLES FOR LIFE SENTENCE CASES.
@. FOR LIFE SENTENCE CASES?
A. YES, YOUR HONOR.
RESPONDENTS EXHIBIT 43 THEN SELECTS FROM RESPONDENTS
1510
KATZ - DIRECT
1 EXHIBIT 42 THOSE VARIABLES THAT ARE SIGNIFICANT AT .0%5 LEVEL.
2 BY MS. WESTMORELAND:
3 @. DOCTOR KATZ, DID YOU DO ANY SIMILAR TYPE OF BREAKDOWN BASED
4 ON THE LIFE SENTENCE CASES WITH NO PENALTY TRIALTY
3 A. YES» I DID.
3 & R. AND WOULD YOU IDENTIFY RESPONDENTS EXHIBIT 44. PLEASE?
£ A. IN RESPONDENT’S EXHIBIT 44 I PREPARED A SIMILAR TABLE TO
WO
THAT PREPARED FOR RESPONDENTS EXHIBIT 42, EXCEPT IN THIS TABLE
4 I AM COMPARING WHITE VICTIM CASES WHOSE DEFENDANTS RECEIVED LIFE
10 SENTENCES AND NO PENALTY TRIAL WAS HELD.
7% THE COURT: EXCUSE ME, 437
12 THE WITNESS: 44, YOUR HONOR.
12 MS. WESTMORELAND: 44, YOUR HONOR.
14 | MR. FORD: TABLE 277%
13 MS. WESTMORELAND: TABLE 24 15 RESPONDENTS EXHIBIT 44.
164 |I“M SORRY.
17 |BY MS. WESTMORELAND:
18 |@. AND THEN WOULD YOU IDENTIFY RESPONDENTS EXHIBIT 43, WHICH
Ww 19 |IS TABLE 27?
20 |A. YES. RESPONDENT’S EXHIBIT 45, 1 SELECTED THOSE VARIABLES
21 |FROM RESPONDENT'S EXHIBIT 44 THAT SHOWED SIGNIFICANCE AT THE .0S
72 |LEVEL, AND CLASSIFIED THEM APPROPRIATELY.
23 |@. AND WHAT OBSERVATION DID YOU MAKE FROM RESPONDENTS EXHIBITS
24 |44 AND 457
en’ A. 1 OBSERVED IN RESPONDENTS EXHIBIT 43 THAT WHITE VICTIM
i S———— ———— . or T pp— ” a ra
KATZ - VOIR DIRE
1 CASES HAVE MORE OF THE MANY AGGRAVATING CIRCUMSTANCES AS
2 COMPARED TO BLACK VICTIM CASES WHEN THESE CASES ARE COMPARED
3 BASED ON LIFE SENTENCES THAT DID NOT ADVANCE TO PENALTY TRIAL.
4 AND THAT BLACK VICTIM CASES TEND TO HAVE MORE OF THE
be’ MITIGATING FACTORS THAN WHITE VICTIM CASES IN LIFE SENTENCE
® & CASES THAT DID NOT ADVANCE TO PENALTY ‘TRIAL.
.
MS. WESTMORELAND: YOUR HONOR. AT THIS TIME. I WOULD
LIKE TO SUBMIT RESPONDENT’S EXHIBITS 41, 42, 43, 44 AND 4S, I w
? BELIEVE.
10 MR. FORD: YOUR HONOR, COULD I VOIR DIRE, PLEASE?
11 THE COURT: UH HUH, VOIR DIRE EXAMINATION
12 BY MR. FORD:
13 @. DOCTOR KATZ. AT THIS POINT YOU JUST DROPPER OUT THE DEATH
14 PENALTY CASES, IS THAT RIGHT?
15 A. 1 DO NOT HAVE A TABLE FOR THE DEATH PENALTY CASES AT THIS
146 POINT, YES.
17 RN. WHAT YOU DISCOVERED HERE. AS I UNDERSTAND IT. THAT IN THE
18 LIFE SENTENCE CATEGORY, THE WHITE VICTIM CASES ARE MORE
Wi 1? AGGRAVATED THAN THE BLACK VICTIM CASES, IS THAT RIGHT?
23 0. DO YOU KNOW IF THAT’S TRUE IN THE DEATH SENTENCE CATEGORY
22 ALSO?
23 A. I1°VE RUN THAT EXPERIMENT, AND I GET SIMILAR RESULTS AS THE
2 TABLES REPRESENTED IN LIFE SENTENCE. AND LIFE SENTENCE NO
29 PENALTY TRIAL.
1512
KATZ - VOIR DIRE
Q. YOU JUST HAVEN’T INCLUDED THOSE TABLES HERE. IS THAT RIGHT?
A. THAT'S CORRECT.
@. YOU HAVEN‘T INCLUDED THEM IN YOUR ANALYSIS, HAVE YOU?
THE COURT: IS THIS CROSS-EXAMINATION OR DOES IT GO TO
ADMISSIBILITY?
MR. FORD: WELL YOUR HONOR, I HAVE ONE MORE QUESTION, I
THINK WILL GET TO THE POINT OF AT LEAST WHAT I UNDERSTAND YOUR
HONORS RULING TO BE ON ADMISSIBILITY.
THE COURT: GO AHEAD.
BY MR. FORD:
Q. DID YOU EVER [0 A COMPARISON OF SENTENCING OUTCOMES BY RACE,
WHERE YOU CONTROLLED BY THIS LEVEL OF AGORAVATION YOU'RE
DISCOVERING HERE?
A. DO YOU MEAN IN THE FORMS THAT PROFESSOR BALDUS, IN THE
EXPERIMENTS SIMILAR TO THE ONES PERFORMED BY PROFESSOR BALDUS?
Q. I MEAN THAT IN THIS TABLE OR ANYWHERE ELSE IN YOUR ANALYSIS,
DC YOU TAKE YOUR FINDING OF HIGH LEVELS OF AGGRAVATION IN WHITE
VICTIM CASES ACROSS THE BOARD AND CONTROL FOR THAT. AND ANALYZE
THEM, WHETHER THERE‘S DISCRIMINATION BETWEEN SIMILARLY SITUATED
BLACK AND WHITE VICTIM CASES? DO YOU EVER DO THAT?
A. NO.
MR. FORD: YOUR HONOR, I DON’T THINK WE-RE EVEN CLOSE
TO THE THRESHOLD OF RELEVANCY. THAT’S WHY IM TAKING THIS UF IN
YOIR DIRE.
THE COURT: I DON-T THINK THAT THESE TABLES DIRECTLY —-
rN
b
W
4
Tr
)
14
17
18
KATZ - VOIR DIRE
i
|
DIRECTLY DEMONSTRATE ANYTHING ABOUT RACIAL EFFECTS AND DEATH
SENTENCING RATES.
1 UNDERSTAND THEM TO BE OFFERED PRIMARILY TO SUPPORT
THE FROPOSITION THAT BLACK VICTIM CASES ARE GENERALLY MORE
MITIGATED AND WHITE VICTIM CASES ARE GENERALLY MORE AGGRAVATED,
WHICH IS AN IMPORTANT PROPOSITION IN DECIDING HOW MUCH VALUE TO
GIVE TO THE CO-EFFICIENTS.
I THINK THAT IT IS MADE ENTIRELY RELEVANT, ALTHOUGH
PERHAPS REPETITIVE. BY DOCTOR WOODWORTH’S TESTIMONY WHEREIN HE
THREW OUT THOSE NUMBERS OF CASES WHICH WERE MOST INFLUENCING THE
SYSTEM AND THERE WAS AN APPRECIABLE DROP IN THE RACIAL EFFECT.
I THINK 1 HAVE THE POINT, BUT I DON'T THINK THAT THEY
DO ANYTHING MORE THAN ESTABLISH PART OF THAT WHICH PROFESSOR
BALDUS REPETITIVELY SAID, AND THAT WAS THAT WHITE VICTIM CASES
WERE MORE AGGRAVATED. I DONT REMEMBER HIM SAYING THAT BLACK
VICTIM CASES WERE GENERALLY MORE MITIGATED, BUT HE MAY HAVE.
1 UNDERSTAND THAT’ S THE POINT AND IT WEAKENS THE DATA
SOMEWHAT. BUT AT LEAST IN ONE TABLE YOU'VE MADE AN ATTEMPT TO
ADJUST FOR THAT PROPENSITY WHICH HAS THE TENDENCY TO TILT THE
SLOFES, AND PERHAPS ARTIFICIALLY.
I THINK I UNDERSTAND IT. I THINK IT GOES TO THE WEIGHT
OF THE, OF YOUR EVIDENCE AND NOT TO DEMONSTRATE A CONTRARY
HYPOTHESIS, IF THATS WHAT YOU/RE CONCERNED ABOUT. AS TQ THE —-
MR. FORD: MY CONCERN IS WITH REGARD TO YOUR HONORS
PURPOSE FOR ADMITTING IT. WE HAVE A SELECTED GROUP OF CASES AND
KATZ - VOIR DIRE
1 WE ONLY HAVE ONE GROUP, BUT MY MAIN CONCERN IS WHERE YOUR HONOR
2 WILL KNOW WHERE WE“RE GOING.
3 THE COURT: WELL, ONE OF THE CONTENTIONS I UNDERSTAND
4 YOU TO BE ASSERTING UNDER YOUR EQUAL PROTECTION THEORY IS THAT
bo’ THERE IS RACIAL EFFECT, THAT IS TO SAY, RACIAL DISCRIMINATION AT
& EVERY LEVEL OF THE SYSTEM AND TO THAT EXTENT THIS IS RELEVANT TO
7 ADDRESS THAT CLASS OF CASES WHERE LIFE SENTENCES ARE IMPOSED
2 WITH OR WITHOUT A PENALTY TRIAL. |
g ILL ADMIT IT, BUT I THINK I COMMENTED ON THE WAY I
0 VIEW IT.
11 MS. WESTMORELAND: I MIGHT MENTION TO THE COURT AT THIS
12 TIME THAT I HAVE BEEN INFORMED THAT OUR OTHER EXPERT IS IN TOWN.
13 BUT I“M KEEPING HIM OUT OF THE COURTROOM. MY UNDERSTANDING FROM
13 THE COURT-S EARLIER COMMENTS. THAT HE IS GOING TO TESTIFY
15 BASICALLY ABOUT HIS COLLABORATION WITH DOCTOR KATZ, AND HIS
1& GENERAL OPINIONS AS TO THE WORK THAT HAS BEEN DONE BY ROTH
17 FROFESSOR BALDUS, DOCTOR WOODWORTH AND DOCTOR KATZ, AS WELL, AND
13 BASED ON THE EARLIER COMMENTS OF THE COURT, I FELT IT WAS
wo 17 PROBABLY MOST APPROPRIATE OF HIM TO WAIT OUTSIDE THE COURTROOM
20 UNTIL WE GOT TO THE TIME THAT WE WERE READY FOR HIS TESTIMONY.
21 AM 1 CORRECT —--—
22 THE COURT? THAT WOULD MAKE ME FEEL BETTER ABOUT HIS
23 CREDIBILITY IF FOR NO QTHER REASON.
24 MS. WESTMORELAND: THAT WAS ONE OF MY MAIN PURPOSES,
3 YOUR HONOR.
1315
KATZ - DIRECT
THE COURT: ALL RIGHT.
BY MS. WESTMORELAND:
@. DOCTOR KATZ,» WE“VE PREVIDUSLY BEEN DISCUSSING THE PROCEDURAL
REFORM STUDY.
AT THIS TIME I WOULD LIKE TO GO INTO THE EXAMINATION
YOU DID OF THE CHARGING AND SENTENCING STUDY.
DID YOU IN MAKING AN EVALUATION OF THAT PARTICULAR
STUDY EXAMINE THE SAMPLING TECHNIGUES THAT WERE UTILIZED?
A. YES, I DID.
@. AND COULD YOU EXPLAIN WHAT PURPOSES ARE GENERALLY. THE
STATISTICAL PURPOSES FOR USING THE STRATIFIED SAMPLE?
A. A STRATIFIED SAMPLE IS GENERALLY UTILIZED WHERE THERE IS
SOME JUDGMENT OR UNDERLYING KNOWLEDGE OF THE STATISTICIAN THAT
HE HAS AS TO THE RELATIONSHIP BETWEEN THE VARIABLE THAT IS BEING
MEASURED AND THE VARIABLE THAT HE CAN STRATIFY ON.
IF THERE INDEED [OES EXIST A RELATIONSHIP THAT HE CAN
DETERMINE, HE CAN FORM HIS STRATA APPROPRIATELY AND LIMIT HIS
SAMPLE SIZE WITHIN EACH STRATUM AND INCREASE THE PRECISION OF
HIS ESTIMATE AS TO THE PROPORTION OF. OF MEMBERS OF THE
POPULATION THAT HAVE THE PARTICULAR VARIABLE THAT S PEING
MEASURED.
@. AND WHAT ARE THE, I GUESS, THE UNDERLYING PRINCIPLES BEHIND
STRATIFICATION?
A. THE UNDERLYING PRINCIPLE IS UTILIZING ADDITIONAL INFORMATION
ABOUT THE SAMPLING UNITS TO HELP DESIGN YOUR EXPERIMENT TO
KATZ - DIRECT
1 INCREASE THE PRECISION OF THE ESTIMATES THAT RESULT.
2 AN EXAMPLE OF A RELATIONSHIP MIGHT BE, FOR INSTANCE.
3 JUST AS AN EXAMPLE. A RESEARCHER MAY BELIEVE THAT THE LEVEL OF
4 AGGRAVATION IN A CRIME OR IN A MURDER IS RELATED TO THE
3 RESULTING SENTENCE. AND MIGHT BELIEVE IT’S STRATIFYING ON
O
-
SENTENCE IN MEASURING CERTAIN VARIABLES RELATING TO AGGRAVATING
FACTORS, MIGHT PROVIDE AN IMPROVED ESTIMATE ON THE PROPORTION OF 7
8 PEOPLE IN THIS POPULATION THAT HAD THIS PARTICULAR AGGRAVATING
? CIRCUMSTANCE.
10 GENERALLY. THE RESEARCHER USES HIS OWN JUDGMENT AS TO
11 WHAT VARIABLES ARE RELATED TO WHAT, AND, IN DESIGNING HIS
12 STRATIFIED DESIGN.
12 @. WHAT CONSIDERATIONS ARE UTILIZED IN DETERMINING WHAT. WHAT
14 WILL BE THE BASIS FOR THE STRATIFICATION?
13 A. THERES ANOTHER REASON FOR STRATIFYING, WHICH I SUSPECT IS
16 WIDELY EMPLOYED IN THIS CASE, AND THAT 1S, ONE MAY WANT TO DO
37 COMPARISONS ON SPECIFIC SUB-POPULATIONS OF THE TOTAL POPULATION.
1a IN THIS CASE, I BELIEVE THAT STRATIFICATION WAS PERFORMED ON
w 19 JUDICIAL CIRCUITS BECAUSE THERE WAS SOME INDICATION OR SOME
20 DESIRE TO TEST HYPOTHESES OR DO ANALYSES BETWEEM JUDICIAL
2: CIRCUITS.
22 BY STRATIFYING ON JUDICIAL CIRCUITS ONE CAN INSURE THAT
rd
0)
SAMPLES ARE OBTAINED FROM EACH JUDICIAL CIRCUIT AT A CERTAIN
EJ
BH LEVEL THAT MIGHT ALLOW IT TO BE ANALYZED. IF SIMPLY RANDOM
25 SAMPLING IS DONE, IT IS POSSIBLE THAT SOME JUDICIAL CIRCUITS
C
R
N
(B
EE
JE
S
E
¢
RE
KATZ - DIRECT
WOULD BE COMPLETELY OMITTED FROM THE SAMPLES THAT ARE SELECTED.
4. ARE THERE ANY PROBLEMS STATISTICALLY WITH STRATIFYING IN
THIS MANNER?
A. STATISTICALLY ONE WOULD WANT TO LIMIT THE NUMBER OF STRATA
THAT ARE SELECTED BECAUSE THAT REDUCES THE PRECISION OF THE
ESTIMATE.
BY HAVING A LOT OF STRATA. ONE CAN COMPLETELY NEGATE
WHATEVER PRECISION WAS GAINED BY HAVING STRATIFICATION BASED ON
A RELATED VARIABLE. |
IN THIS CASE. THERE WERE 4 SENTENCING DECISIONS THAT
WERE STRATIFIED. AND 42 JUDICIAL CIRCUITS, WHICH GIVE 1468 TOTAL
STRATA.
FURTHERMORE, IN SOME OF THEIR ANALYSES, THEY SUBDIVIDE
EACH OF THESE STRATA IN TERMS OF THE NUMBER OF CASES THAT HAD
THE DIFFERENT DEFENDANT-VICTIM RACIAL COMBINATION.
50, ONE IS TALKING ABOUT A LOT OF SUBDIVISION IN THE
STRATIFICATION AND THAT CAN LEAD TO PROBLEMS IN THE PRECISION OF
THE ESTIMATES THAT RESULT.
G. IN DECIDING WHAT STRATA TO USE. OR WHAT, HOW YOURE GOING TO
MAKE THE PARTICULAR STRATIFICATION, ON WHAT BASIS, WHAT ARE THE
CONSIDERATIONS THAT ARE USUALLY TAKEN INTO ACCOUNT AS FAR AS THE
DIFFERENT STRATA THEMSELVES, AND THEIR RELATIONSHIPS WITH EACH
OTHER?
A. ONE CAN INCREASE THE PRECISION OF THE ESTIMATION PROCESS
WITH STRATIFICATION.
fa
vs
o
S
TR
EY
EN
E
«T
ha
FR
Se
1519
KATZ - DIRECT
HOWEVER. THERE‘S A COST ASSOCIATED WITH STRATIFYING,
AND THAT IS. ONE NEEDS TO BE ABLE TO DETERMINE THE RELATIVE
WEIGHTS THAT SHOULD BE APPLIED TO EACH STRATA. EXCUSE ME, EACH
STRATUM.
THESE WEIGHTS REPRESENT THE EFFECT THAT THE PARTICULAR
STRATUM SHOULD HAVE IN THE TOTAL ESTIMATE FOR THE UNDERLYING
VARIABLE THAT 1S BEING ESTIMATED.
@. AND HOW WOULD CERTAIN THINGS, SUCH AS UNKNOWNS, AFFECT THE
WEIGHTS DEVELOPED BY THE STRATIFIED SAMPLE?
A. IT WOULD DEPEND ON THE TREATMENT OF THOSE UNKNOWNS. IF
THOSE UNKNOWNS WERE RECODED THEN IT WOULD NOT HAVE AN EFFECT ON
THE WEIGHTS.
HOWEVER, IF UNKNOWNS FOR THE PARTICULAR VARIABLE ARE
DISREGARDED. THEN THE RELATIVE WEIGHTS WOULD HAVE TO BE ADJUSTED
TO ACCOUNT FOR THOSE UNKNOWNS NOT BEING USED IN THE ESTIMATION
PROCESS.
THE COURT: HAVE YOU GOT ANY UNKNOWNS IN THE SENTENCING
OUTCOME BY JUDICIAL CIRCUIT?
THE WITNESS: 1 BELIEVE. YOUR HONOR, THAT THERE WERE
SOME UNKNOWNS IN DISTINGUISHING BETWEEN WHETHER OR NOT A PENALTY
TRIAL WAS HELD.
FOUR SENTENCING OUTCOMES. WERE DEATH SENTENCE, AS I
BELIEVE. DEATH SENTENCE, LIFE SENTENCE, AND A PENALTY TRIAL.
LIFE SENTENCE NO PENALTY TRIAL, AND A VOLUNTARY MANSLAUGHTER
SENTENCE.
151%
KATZ ~ DIRECT
1 I HAVE REASON TD BELIEVE THAT THERE WERE AT LEAST 76
2 PENALTY TRIAL CASES AT ONE POINT IN WHICH IT WAS INDICATED THAT
3 IT WAS UNKNOWN WHETHER A PENALTY TRIAL HAD OCCURRED OR NOT.
4 FURTHERMORE, I AM. I SUSPECT THAT THERE ARE OTHER LIFE
= SENTENCE CASES THAT WERE NOT SAMPLED IN WHICH THEY DID NOT HAVE
& INFORMATION ABOUT WHETHER A PENALTY TRIAL CCCURRED IN THOSE
7 CASES OR NOT.
a BY M3. WESTMORELAND:
2? @. HOW WOULD THAT AFFECT THE UNKNOWNS IN THAT PARTICULAR AREA?
10 HOW WOULD THAT AFFECT THE WEIGHTS?
11 A. WELL, THE UNKNOWNS IN TERMS OF THE PENALTY TRIAL CASES COULD
12 AFFECT THE WEIGHTS FOR THOSE CASES THAT HAVE PENALTY TRIALS AND
13 THOSE CASES THAT DID NOT HAVE PENALTY TRIALS. S0 POTENTIALLY
14 HALF THE WEIGHTS THAT ARE USED MAY BE IN ERROR, BECAUSE OF THE
13 INABILITY TO DISTINGUISH BETWEEN CERTAIN CASES AS TO WHETHER A
146 PENALTY TRIAL HAD OCCURRED OR NOT.
17 0. IN YOUR ANALYSIS OF THE CHARGING AND SENTENCING STUDY. HOW
18 DID YOU TREAT THE SAMPLE?
Ww 1% A. I WAS UNABLE TO DETERMINE WHAT THE PROPER WEIGHTS WERE. S50 I
20 USED A MORE CONSERVATIVE APPROACH AND JUST LISED THAT SAMPLE THAT
21 I HAD OF 1082 CASES AS A RANDOM SAMPLE. I BELIEVE THAT THERE
22 CAN BE SOME EFFECT IN MY RESULTS IN TERMS OF SENTENCE QUTCOME.
23 IF IT’S UNDERSTOOD THAT A HIGHER SENTENCE GENERALLY CORRELATES
<4 WITH HIGHER LEVELS OF AGGRAVATION AND MITIGATION.
2% HOWEVER, MOST OF MY ANALYSIS IS DONE IN TERMS OF CASES
1320
KATZ - DIRECT
1 WHICH HAD THE SAME SENTENCE, SO FOR THOSE ANALYSES. I DO NOT
J BELIEVE. THAT ASSUMING THAT THE SAMPLE REPRESENTS WITHIN A
3 PARTICULAR SENTENCE, A RANDOM SAMPLE IS INAPPROPRIATE.
3 2. DID YOU USE WEIGHTED NUMBERS IN YOUR ANALYSES?
| A. NO» I DIDN’T. BECAUSE I DID NOT KNOW WHAT THE APPROPRIATE
2 64 |WEIGHTS WERE. THEY CHANGE, DEPENDING ON THE PRECISE NUMBER OF
7 |PENALTY TRIAL CASES THAT HAVE OCCURRED. AND I KNOW THAT THERE’S
& |AT LEAST 76 IN WHICH IT’S UNKNOWN.
9 I MADE NO ASSUMPTIONS ABOUT IT.
0 MR. FORD: I OBJECT TO THAT, YOUR HONOR. AND MOVE TO
11 |STRIKE ON THE GROUNDS THERES NO FOUNDATION FOR THAT STATEMENT
12 |BY DOCTOR KATZ. IT’S CERTAINLY NOT A STATISTICAL ESTIMATE. I
13 |WOULD LIKE TO HAVE THAT STRICKEN FROM THE RECORD.
14 THE COURT: 1IT‘S IN AND YOU MAY CROSS—EXAMINE ON IT.
15 MR. FORD: THANK YOU, YOUR HONOR.
14 |BY MS. WESTMORELAND:
17 a. DOCTOR KATZ. IF WE COULD LOOK AT YOUR, THE ANALYSIS THAT YOU
18 |DID IN RELATION —-
W 19 THE COURT: NOW, WAIT A MINUTE. HOW DO YOU KNOW. WHAT
20 [MAKES YOU THINK THERE ARE 76 CASES WHERE A PENALTY TRIAL WAS
21 UNKNOWN?
=a THE WITNESS: THERE ARE SEVERAL WAYS.
23 IN THE DATA THAT WAS GIVEN TO ME, THERE WERE VARIABLES
24 CODED FOR WHETHER OR NOT A PENALTY TRIAL HAD QCCURRED. AND THE
25 UNKNOWN VALUE WAS INDICATED FOR MANY OF THESE CASES.
KATZ - DIRECT
I BELIEVE THERE WERE 63 CASES. OR &2 CASES IN THE
CHARGING AND SENTENCING STUDY. AND 23 OR 24 CASES IN THE
PROCEDURAL. REFORM STUDY.
ALSO. DURING THE PROCESS OF DISCOVERY, WE RECEIVED A
LIST OF CASES IN WHICH WE WERE ASKED TO IDENTIFY WHETHER A
PENALTY TRIAL HAD OCCURRED OR NOT, AND THESE WERE THE 74 CASES
FROM BOTH STUDIES, IN WHICH UNKNOWN PENALTY TRIALS HAD BEEN
INDICATED.
WE DID NOT KNOW THE STATUS OF THOSE PENALTY TRIALS.
THAT MADE ME SUSPECT THAT THE TRUE STATUS WAS UNKNOWN.
IN EACH OF THOSE 74 CASES THAT WAS REQUESTED. WERE
SPECIFIC CASES IN EITHER THE CHARGING AND SENTENCING STUDY OR
THE PROCEDURAL REFORM STUDY.
THE COURT: WELL, DOCTOR KATZ, HE DIDN‘T WEIGH THE
PROCEDURAL REFORM STUDY, HE ONLY WEIGHED THE CHARGING AND
SENTENCING STUDY. HOW MANY OF THOSE WERE UNKNOWN?
THE WITNESS: I BELIEVE THERE WERE 13 OR 12 OR 13
ADDITIONAL PENALTY TRIAL CASES IN THE PROCEDURAL REFORM STUDY IN
WHICH IT WAS UNKNOWN WHETHER A PENALTY TRIAL HAD OCCURRED. AND I
BELIEVE THAT’ S IMPORTANT BECAUSE THESE WEIGHTS ARE DEVELOPED IN
TERMS OF THE UNIVERSE OF CASES THAT HE DEFINED, AND THESE CASES.
ALTHOUGH NOT UTILIZED IM THE CHARGING AND SENTENCING STUDY. ARE
CASES FROM THE UNIVERSE. IN WHICH TO DETERMINE THE PROPER
WEIGHTS, ONE WOULD NEED TO KNOW WHETHER A PENALTY TRIAL HAD
OCCURRED OR NOT IN THOSE CASES.
KATZ ~- DIRECT
1 THE COURT: WOULDN“T THE CASES IN THE PEMALTY. WHATEVER
r)
THAT STUDY, THE FIRST STUDY. WOULDN/T THE UNKNOWNS IN THAT CASE
3 BE CO-EXTENSIVE WITH AND GREATER THAN THE NUMBER IN THE CHARGING
4 AND SENTENCING STUDY?
3 THE WITNESS: FOR THE CRIME PENALTY TRIAL CASES?
& - THE COURT: UH HUH?
7 THE WITNESS: NO, YOUR HONOR, THERE WERE. I BELIEVE,
a 23 OR 24 CASES IN THE PROCEDURAL REFORM STUDY THAT WAS INDICATED
? TO BE UNKNOWN. AND 62 OR &3 IN THE CHARGING AND SENTENCING
10 STUDY.
11 I BELIEVE THE CHARGING AND SENTENCING STUDY HAD MORE
12 LIFE SENTENCE CASES CONSIDERED THAN THE PROCEDURAL REFORM STUDY.
13 THE COURT: WHAT I°M ASKING YOU IS, ARE YOU COUNTING
14 THE TWENTY-0DD IN THE SECOND STUDY TWICEY
13 THE WITNESS: NO. YOUR HONOR, IM TRYING TO NOT DOUBLE
1& COUNT THOSE. THE 74 NUMBER THAT I REFERRED TO WERE THE
17 74 CASES THAT WERE SENT DURING DISCOVERY REQUESTING WHAT THE
18 STATUS OF THOSE PENALTY TRIAL CASES, EXCUSE ME, WHAT THE STATUS
Ww i? OF THE PENALTY TRIAL FOR THOSE LIFE SENTENCE CASES WERE. I WENT
20 THROUGH THAT LIST AND CHECKED TO SEE THAT THEY REPRESENTED REAL
21 CASES IN BOTH THE CHARGING AND SENTENCING STUDY AND THEN THE
<2 ADDITIONAL CASES AT THE END THAT HAD BEEN ADDED WERE CASES FROM
23 THE PROCEDURAL REFORM STUDY.
249 THE COURT: ALL RIGHT. BASICALLY. THE WAY I SEE IT,
25 MR. FORD IS. THAT HE HAS A REASONABLE BASIS ON WHICH TO CONCLUDE
KATZ - DIRECT
i THAT YOUR RESEARCHERS DID NOT KNOW PRECISELY HOW MANY PENALTY
2 TRIALS THERE WERE.
NOW. BECAUSE OF THE WAY THIS THING HAS EVOLVED, IT
CGULD BE THAT THAT HAS SUBSEQUENTLY BEEN CORRECTED. BUT I DON’T
KNOW.
3
a
5
® & HE HAS SOME 2ASI3 ON WHICH TO TESTIY. AND DO YOU WANT
7 TO TELL ME THAT YOU FINALLY ENDED UP WITH WEIGHTS BASED ON AN
3 ACCURATE UNDERSTANDING OF HOW MANY PENALTY TRIALS THERE WERE OR
9 WEREN”T?
10 MR. FORD: I“M NOT EXACTLY SURE WHAT THE WITNESS IS
11 BASING IT ON.» BUT I DO UNDERSTAND THE COURTS RULING AND WELL
12 INQUIRE FURTHER ON TO MAKE SURE IT“S CLEARED UP,
13 THE COURT: ALL RIGHT. ALL RIGHT. WE‘LL STOP AND GO
14 TO EAT LUNCH. WE‘LL BE IN RECESS, UNTIL QUARTER OF TWO.
13 ey
1&6 (COURT RECESSED FOR LUNCH.)
17 THE COURT: IS THERE A MEMBER OF THE PRESS PRESENT?
13 WERE YOU THE ONE THAT WAS ASKING ABOUT THE EXHIBITS?
w 1% MEMBER OF THE PRESS: YES. SIR.
20 THE COURT: ALL RIGHT. YOU’RE WITH —-7?
2) MEMBER OF THE PRESS: CONSTITUTION AND JOURNAL. YOUR
22 HONOR.
23 THE COURT: THIS GENTLEMAN HAS ASKED TO BE ABLE TO SEE
24 THE EXHIBITS, AND I CAN‘T THINK OF ANY REASON FOR HIM NOT TO.
28 AND HAVE TOLD MRS. PAGE, I FURTHER DIDNT SEE ANY REASON FOR HIM
1324
KATZ - DIRECT
1 NOT TO HAVE A XEROX OF THE EXHIBIT LIST. SO HE CAN FIGURE OUT
2 WHAT HE NEEDED AND WHAT HAVE YOU.
3 BUT BEFORE I SAY THAT. I THOUGHT I WOULD ASK YOU ALL IF
A YOU HAVE ANY OBJECTION.
= MR. FORD: NONE. YOUR HONOR.
é MS. WESTMORELAND: I DON'T KNOW OF ANY, YOUR HONOR.
7 YOUR HONOR. IF I MIGHT HAVE JUST ONE MOMENT. IM TRYING
& . |TO FIND A COPY OF AN UPDATED TABLE THAT WAS HERE THIS MORNING,
FJ AND I CAN‘T PUT MY HANDS ON IT.
10 THE COURT: ALL RIGHT. COUNSEL, THIS DOES NOT MEAN I
11 HAVE SANCTIONED A VIOLATION OF THE LOCAL RULE WHICH SAYS THAT
12 YOU“RE NOT TO TRY YOUR CASES IN THE NEWSPAPER. IM SIMPLY
13 LETTING HIM LOOK AT THE EXHIBITS.
14 MR. FORD: YOUR HONOR, FOR THE COURT’S INFORMATION, I
15 BELIEVE THE GENTLEMAN DID ASK ME ABOUT WHAT HAD COME IN, AND I
1&6 REFERRED HIM TO THE COURT RECORD AT THAT POINT.
17 THE COURT: ALL RIGHT.
12 DO ANY OF YOU PERCHANCE KNOW IF JUDGE QWENS HAS RULED?
Ww 12 MR. FORD: I DO NOT. YOUR HONOR.
20 MS. WESTMORELAND: YOUR HONOR, -—
23 THE COURT: DO YOU KNOW IF MR. BOGER RAISED THIS ISSUE
22 WITH HIM?
23 MR. FORD: 1 BELIEVE ——
24 MS. WESTMORELANDt HE HAS.
23 MR. FORD: -—- THAT ISSUE HAS BEEN RAISED.
1325
KATZ - DIRECT
1 THE COURT: ALL RIGHT.
MS. WESTMORELAND: MY UNDERSTANDING WAS, I THINK THE I
3 PARTIES WERE TO PRESENT EVERYTHING BY FIVE 0°CLOCK THIS
4 AFTERNOON, SO I DOUBT ANY RULING WILL BE FORTHCOMING UNTIL AFTER
= THAT POINT FROM JUDGE OWENS.
& THE COURT: IF MR. BOGER CAN PRESENT THIS EVIDENCE BY
7 FIVE O-CLOCK THIS AFTERNOON. THEN I WANT TO KNOW HOW COME HE
a AIN’T DONE IT IN MY CASE. ALL RIGHT.
? MS. WESTMORELAND: THANK YOU, YDUR HONOR. I APOLOGIZE
0 FOR THE DELAY. OUT OF ALL THE VARIOUS DOCUMENTS, ONE JUST HAPPENS
11 TO HAVE GOTTEN MISPLACED THIS MORNING AND I CAN‘T SEEM TO PUT MY
12 HANDS ON IT.
13 ---
14 JOSEPH LAUREN KATZ,
15 BEING FREVIDUSLY DULY SWORN, RESUMED THE WITNESS STAND AND
16 TESTIFIED FURTHER AS FOLLOWS: |
17 : DIRECT EXAMINATION (CONT‘D)
18 BY MS. WESTMORELAND:
Ww 1% @. DOCTOR KATZ, TO CONTINUE YOUR TESTIMONY, I BELIEVE WE WERE
20 |BEGINNING TO MOVE INTO THE CHARGING AND SENTENCING STUDY AND
24 ANALYSIS OF THE DATA ON THE CHARGING AND SENTENCING STUDY.
2 DID YOU CONDUCT AN ANALYSIS IN RELATION TO THE CHARGING
23 AND SENTENCING STUDY SIMILAR TO THAT DONE FOR THE PROCEDURAL
24 REFORM STUDY?
2% A. YES, I DID.
KATZ - DIRECT
0. AND WOULD YOU LOOK AT RESPONDENTS EXHIBIT 44, PLEASE. AND
IDENTIFY THAT DOCUMENT, IF YOU WOULD. THAT'S TABLE 287
A. YES. IN RESPONDENT’S EXHIBIT 44% IS PRESENTED VARIABLES THAT
WERE SELECTED FROM THE CHARGING AND SENTENCING STUDY THAT WERE
ORIGINALLY DEFINED BY PROFESSOR BALDUS. AND I ATTEMPTED AGAIN TO
CLASSIFY THEM ACCORDING TO WHAT I BELIEVE WAS HIS UNDERSTANDING
OF WHAT WERE THE AGGRAVATING AND MITIGATING FACTORS.
Gl. AND ARE THESE, HOW DID YOU UTILIZE THESE VARIABLES, OR THESE
FACTORS LISTED IN THIS TABLE?
A. 1 USED THESE VARIABLES IN THE SAME MANNER AS I UTILIZED THE
VARIABLES DEFINED IN RESPONDENTS EXHIBIT 23 FOR THE PROCEDURAL
REFORM STUDY, AND PERFORMED A SIMILAR ANALYSIS.
MS. WESTMORELAND: YOUR HONOR, AT THIS TIME I WOULD
LIKE TO SUBMIT RESPONDENT’S EXHIBIT 44 INTO EVIDENCE TO SHOW THE
BASIS FOR THE SUBSEQUENT ANALYSES TO BE DISCUSSED BY DOCTOR KATZ
AND TO SHOW THE PARTICULAR VARIABLES THAT HE UTILIZED IN MAKING
THAT ANALYSES.
MR. FORD: NO OBJECTION TO THAT, YDUR HONOR.
THE COURT: THEY WILL BE ADMITTED.
1 MIGHT OBSERVE. MR. FORD. WITH REFERENCE TO YOUR
OUTSIDE OF EXPERTISE EXAMPLE —-— OBJECTION, THAT MY UNDERSTANDING
OF THE TESTIMONY TODAY 13 THAT HE HAS NOT MADE ANY INDEPENDENT
SOCIAL SCIENCE JUDGMENT, BUT IS INTERPRETING, PROFESSOR BALDUS”
JUDGMENTS IN THAT RESPECT. TALKING ABOUT WHATS HAPPENED TO
DATE, I°M NOT TALKING ABOUT WHATS COMING, I DON'T KNOW WHAT "S
Cr
B
G
T
E
e
H
E
1
R
R
1
E
R
1327
KATZ -~ DIRECT
COMING.
MR. FORD: TO DATE, HE HASN'T LEAPT OVER THE LINE. BUT
CERTAINLY BUMPED UP AGAINST IT. BETWEEN THOSE TWO THINGS.
THE COURT: UNDER EITHER VIEW, HE HAS NOT EXCEEDED HIS
EXPERTISE, THEN.
GO AHEAD.
BY MS. WESTMORELAND:
Q. DOCTOR KATZ, DID YOU EXAMINE IN THE CHARGING AND SENTENCING
STUDY FOR AN, AND EXAMINE THAT STUDY BASED ON UNKNOWNS IN LIGHT
OF TABLE, IN LIGHT OF RESPONDENTS EXHIBIT 467
A. YES. IN ORDER TO PERFORM A SIMILAR ANALYSIS TO WHAT I DID
ON THE PROCEDURAL REFORM STUDY. I HAD AN OPPORTUNITY TO EXCLUDE
UNKNOWNS FROM THE ANALYSIS IN MY ANALYSIS OF THE CHARGING AND
SENTENCING STUDY.
IN THE PROCEDURAL REFORM STUDY. IT WAS. BECAUSE OF THE
FOIL DESIGN FOR THE QUESTIONNAIRES, THERE WAS SOME UNCERTAINTY
IN MY MIND, AS TO WHETHER NON-INDICATED ITEMS WERE ACTUALLY
UNKNOWN OR NON-INDICATED BECAUSE THEY DID NOT OCCUR. AND I WAS
FRUSTRATED IN MY ATTEMPT TO TRY AND SEPARATE OUT UNKNOWNS FOR
THOSE PARTICULAR ITEMS.
IN THIS OEDRGIA CHARGING AND SENTENCING STURDY, THERE
ARE QUESTIONNAIRE ITEMS IN WHICH THERE ARE FOUR POSSIBLE
RESPONSES WHICH ALLOW ME TO DISTINGUISH SPECIFICALLY FOR EACH OF
THESE ITEMS WHETHER THE “U" WAS PRESENT CR NOT, SO ON THAT
BASIS. 1 WAS ABLE TD PERFORM AN ANALYSIS THAT DID NOT INCLUDE
o
n
Ww
D
d
KATZ ~ DIRECT
THE UNKNOWNS.
@. AND DID YOU DO A SEPARATE TABLE ON THAT BASIS?
A. YES. RESPONDENTS EXHIBIT 47 —-
@. DO YOU HAVE RESPONDENT’S EXHIBIT 47A7
A. YES.
MS. WESTMORELAND: LET ME NOTE AT THIS POINT, THE
RESPONDENTS EXHIBIT 47 WAS TABLE 29, WHICH WAS ONE OF THOSE
THAT WAS REQUESTED BE RERUN. AND WE’RE SUBSTITUTING
RESPONDENTS EXHIBIT 47A, WHICH I BELIEVE WAS PREVIOUSLY
PROVIDED FOR FOR THE COURT. |
THE COURT: LET ME DO WHAT YOU YOU WERE DOING AND SEE
IF I CAN FIND IT.
MS. WESTMORELAND: IT WAS ONE OF THE SEPARATE DOCUMENTS
THIS MORNING, YOUR HONOR.
THE COURT: ALL RIGHT.
MS. WESTMORELAND: YOU HAD BETTER SUCCESS THAN I DID, I
CAN SEE.
BY MS. WESTMORELAND:
@. DOCTOR KATZ. WOULD YOU IDENTIFY RESPONDENTS EXHIBIT 47A,
PLEASE?
A. YES. IN RESPONDENT’S 47A. I TRIED TO IDENTIFY TO THE BEST
OF MY ABILITY THE NUMBER OF UNKNOWNS THAT WERE PRESENT FOR EACH
OF THE VARIABLES THAT I AM ABOUT TO USE IN MY ANALYSIS OF THE
CHARGING AND SENTENCING STUDY.
THERE WERE SEVERAL FRUBLEM: THAT I RAN INTO BECAUSE OF
fo
7
&G
KATZ — DIRECT
THE QUESTIONNAIRE DESIGN IN THE CHARGING AND SENTENCING STUDY.
SOME OF THESE ITEMS ARE INDICATED TO BE UNDETERMINED IN THE
TOTAL NUMBER NUMBER OF UNKNOWNS,
THIS AGAIN IS THE RESULT OF THE FOIL METHOD IN WHICH ITE]
FOR THESE QUESTIONS DID NOT INCLUDE AN ITEM FOR UNKNOWN.
I DID, HOWEVER, FIND IN THESE QUESTIONS, THERE WERE
SOME SITUATIONS WHERE IMPERMISSIBLE CODES WERE PRESENT WHICH I
THREW OUT AND INDICATED AS UNKNOWN. SO I DID NOT CONSIDER THEM
FOR THOSE PARTICULAR VARIABLES.
ALSO FOR VARIABLES IN WHICH THEYRE DEFINED AS
INTERACTION VARIABLES OR COMPOSITE VARIABLES OR SOME KIND OF
COMPLEX APPLICATION OF THE QUESTIONNAIRE ITEMS, 1 ATTEMPTED TO
THE BEST OF MY MATHEMATICAL ABILITY TO DETERMINE PRECISELY WHAT
IS KNOWN AND WHAT IS UNKNOWN, AND THAT 3 REFLECTED IN THIS
TABLE.
2. IS THIS A TABLE THAT YOU UTILIZED IN EXCLUDING PARTICULAR
VARIABLES DURING YOUR LATER ANALYSES, THEN?
A. THIS IS AN UPDATED TABLE OF THE NUMBER OF UNKNOWNS. THE
ACTUAL NUMBER HAS CHANGED FOR CERTAIN OF THESE ITEMS BECAUSE I
WAS INSTRUCTED TO UPDATE THIS TABLE, USING THE NEW FILE THAT I
HAD.
20 THIS IS A REFLECTION OF THE NUMBER OF UNKNOWNS FOR
THE NEW FILE. IT DOESN‘T PRECISELY CORRESPOND TO THE NUMBER OF
UNKNOWNS 1 PRESUMED IN THE ANALYSIS I DID ON THE CHARGING AND
SENTENCING STUDY.
AS
1530
KATZ - DIRECT
@. AND I BELIEVE EARLIER WE IDENTIFIED AN EXHIBIT LABELED
I.
RESPONDENT”S 18A, WHICH ALSO WAS A COUNTING OF THE NUMBER OF
UNKNOWNS FOR THE CHARGING AND SENTENCING STUDY.
HOW ARE THESE TWO TABLES DIFFERENT, OR WHAT IS THE
DIFFERENCE BETWEEN THEM?
A. YES. IN THE, IN REPONDENT“S EXHIBIT 18. IT’S A LIST OF
VARIABLES THAT REPRESENT QUESTIONNAIRE ITEMS. AND THE TOTAL
NUMBER OF UNKNOWNS FOR EACH ITEM.
W
B
N
O
L
d
e
ly
o
N
RESPONDENTS EXHIBIT 47A IS A LIST OF THE UNKNOWNS FOR
10 THE VARIABLES THAT WERE DEFINED FROM THESE QUESTIONNAIRE ITEMS.
11 NOW THERE ARE TIMES WHERE THE ITEMS AND THE VARIABLE
12 ARE EXACTLY THE SAME. THERE ARE OTHER TIMES WHERE THE VARIABLE
13 IS DEFINED AS A COMPOSITE OR A COMPLEX DEFINITION OF THE
i4 UNDERLYING VARIABLES IN THE QUESTIONNAIRE ITEMS FOR THE CHARGING
15 AND SENTENCING STUDY.
146 R. DID YOU UTILIZE INFORMATION FROM PROFESSOR BALDUS IN
17 DETERMINING THE DEFINITION OF THESE VARIABLES, THEN. THAT ARE IN
13 THIS TABLE?
w 19 A. YES. THE DEFINITIONS ARE ASSOCIATED AS CLOSELY AS I COULD
20 TO THE ITEMS AS THEY ARE INDICATED ON THE QUESTIONNAIRE.
21 MS. WESTMORELAND: YOUR HONOR. WE SUBMIT RESPONDENT 3
22 EXHIBIT 47A AT THIS TIME FOR TWO PURPOSES.
23 ONE TO SHOW UNKNOWNS IN RELATION TO PARTICULAR
24 VARIABLES CODED IN THE COMPUTER.
23 AND ALSO AS AN ILLUSTRATION OF THE PROCESS DOCTOR KATZ
1531
KATZ - DIRECT
1 UNDERWENT IN EXCLUDING CERTAIN ITEMS FROM HIS LATER ANALYSES IN
2 REGARD TO THE CHARGING AND SENTENCING STUDY.
MR. FORD: FOR THE SECOND PURPOSE. YOUR HONOR, I ONLY
HAVE MY PREVIOUS OBJECTION. FOR THE FIRST PURPOSE, I“M STILL
UNCLEAR AS TO WHAT THIS WORD "UNDETERMINED" MEANS. AS FAR AS I
THE COURT: WELL, I KNOW THAT IT DOES MEAN THAT IF HE
3
4
3
5 couLD TELL IT MAY VERY WELL MEAN ZERO.
7
3 FOUND ANY WILD CODES, HE THREW THEM OUT.
2? I BELIEVE I KNOW THAT HE DID NOT THROW OUT ANY OTHERS
10 ON THAT CATEGORY.
13 IS THAT RIGHT?
12 THE WITNESS: YES, YOUR HONOR.
13 THE COURT: ALL RIGHT. IF I UNDERSTAND HIS TESTIMONY.
14 IT 1S THAT "UNDETERMINED" MEANS THAT BECAUSE OF THE WAY THE
13 QUESTIONNAIRE WAS SET UP, IT WAS NOT POSSIBLE TO GIVE AN UNKNOWN
1&6 RESPONSE AND THEREFORE HE CANNOT DETERMINE HOW MANY UNKNOWNS
17 THERE WERE.
138 DO YOU HAVE AN OBJECTION TO THAT?
@ 19 : MR. FORD: NONE OTHER, I GUESS, THAN THE RELEVANCE
20 QUESTION WE RAISED BEFORE. YOUR HONOR.
21 THE COURT: IT WILL BE ADMITTED.
22 BY M3. WESTMORELAND:
23 @. ALL RIGHT, DOCTOR KATZ, IF YOU WOULD REFER. THEN. TO
24 RESPONDENTS EXHIBIT 48, TABLE 30. AND IDENTIFY THAT EXHIBIT,
25 PLEASE?
KATZ - DIRECT
1 A. THIS IS AN EXHIBIT I PREPARED. RESPONDENTS EXHIBIT 48, TO
2 COMPARE, FOR THE CASES IN THE CHARGING AND SENTENCING STUDY. THE
3 COUNTS FOR THE NUMBER OF EACH VARIABLE THAT WAS DEFINED IN
RESPONDENTS EXHIBIT 46.
THE NUMBER OF CASES THAT FELL INTO EACH CATEGORY»
VOLUNTARY MANSLAUGHTER, LIFE SENTENCE AND DEATH SENTENCE.
PROCEDURAL REFORM STUDY, THEN?
4
5
&
7 @. IS THIS A SIMILAR TABLE TO THAT DONE IN REGARD TO THE
8
9 A. YES, EXCEPT IN THE PROCEDURAL REFORM STUDY. I ONLY
10 CONSIDERED THE TWO SENTENCING OPTIONS, LIFE SENTENCE AND DEATH
11 SENTENCE.
12 FURTHERMORE IN MY CONSTRUCTION OF THIS TABLE, I HAVE
13 EXCLUDED THE UNKNOWN RACE OF VICTIM CASES. EXCUSE ME. I ACTUALLY
14 INCLUDED THE UNKNOWN RACE OF VICTIM CASES, BECAUSE THEY DID NOT
15 AFFECT THE CATEGORIZATION OF EITHER A VOLUNTARY MANSLAUGHTER
146 SENTENCE. A LIFE SENTENCE OR A DEATH SENTENCE.
17 THESE COUNTS ALSO REFLECT THE UNKNOWNS THAT I ATTEMPTED
18 TO ELIMINATE FROM CONSIDERATION.
19 G. ARE THE UNKNOWNS INCLUDED IN THESE COUNTS OR EXCLUDED FROM
20 THESE COUNTS?
21 A. THEY ARE EXCLUDED FROM THESE COUNTS. SO THE NUMBER OF CASES
ad WILL CHANGE DEPENDING ON THE VARIABLES DEFINED.
23 THE COURT: FOR THE PURPOSES OF ILLUSTRATION, IF I WENT
24 BACK TO THE PREVIOUS TABLE AND FOUND "AVENGE." AND I ADDED 13 TO
23 507, THEN I WOULD GET THE TOTAL NUMBER OF CASES.
1533
KATZ — DIRECT
[a
y THE WITNESS: PRESUMABLY, YES. YOUR HONOR. UNLESS THAT
2 NUMBER HAS CHANGED FROM THE PREVIOUS DATA FILE. I WAS DIRECTED
3 TO UPDATE THOSE NUMBER OF UNKNOWNS.
4 THE COURT: IT HASNT.
3 MR. KATZ: OKAY. SO THAT SHOULD ADD UP TO 1082 FOR
é THIS TABLE.
7 BY MS. WESTMORELAND:
3 ©. FOR THE TABLE JUST, UNDER THE COUNT OF VOLUNTARY
9 MANSLAUGHTER OR IF YOU ADD UP ALL THE SENTENCE. ALL THE
10 DIFFERENT SENTENCING OPTIONS?
11 A. IF YOU ADD UP ALL THE DENOMINATORS FOR THE DIFFERENT
12 SENTENCING OPTIONS UNDER "AVENGE." AND ADD IN 13, IT SHOULD GET
13 TO 1082.
14 MS. WESTMORELAND: YOUR HONOR. WE SUBMIT RESPONDENTS
1S EXHIBIT 48 AT THIS TIME.
16 ~ MR. FORD: SAME RELEVANCE OBJECTION, YOUR HONOR.
17 THE COURT: I TAKE IT ITS A FOUNDATION OR PROCESS
13 TABLE AND NOT OFFERED --
19 MS. WESTMORELAND: THAT’S CORRECT. YOUR HONOR.
20 THE COURT: ~-- AS EVIDENCE OF ANY ULTIMATE FACT TO BE
21 DETERMINED IN THE CASE, IS THAT RIGHT?
22 MS. WESTMORELAND: IT IS A FOUNDATION TABLE. BASICALLY
23 SHOWING THE PROCESSES THAT DOCTOR KATZ TOOK AND THE STEPS THAT
24 HE TOOK IN HIS ANALYSIS.
23 THE COURT: ALL RIGHT. ILL ADMIT IT FOR THOSE
1534
KATZ - DIRECT
PURPOSES.
BY MS. WESTMORELAND:
@. DOCTOR KATZ,» AFTER YOU MADE THIS PARTICULAR TABLE, MADE THAT
PARTICULAR, THOSE PARTICULAR CALCULATIONS, DID YOU DO A
BREAKDOWN ON RACE OF VICTIM IN REGARD TO THE CHARGING AND
SENTENCING STUDY?
A. ved, T.DIND,
@. AND WOULD YOU EXAMINE RESPONDENT'S EXHIBIT 49. TABLE 317
A. YES. IN RESPONDENTS EXHIBIT 49, I COMPARED THE CASES. ALL
THE CASES IN THE CHARGING AND SENTENCING STUDY THAT WERE
RELEVANT, GIVEN MY RULE OF EXCLUDING UNKNOWN CASES IN TERMS OF
THE RACE OF THE VICTIM, WHERE ALL THE VARIABLES THAT WERE
DEFINED EARLIER IN RESPONDENT’S EXHIBIT 46.
IN THIS TABLE. THE UNKNOWN RACE OF THE VICTIM CASES
HAVE BEEN EXCLUDED AND I BELIEVE THEY WILL BE EXCLUDED FROM THE
REMAINDER OF THE CASES.
@. THEN WOULD YOU REFER TO RESPONDENT’S EXHIBIT 50, TABLE 32
AND IDENTIFY THAT DOCUMENT, PLEASE?
A. YES. IN RESPONDENT’S EXHIBIT $0. I SELECTED THE LIST OF
VARIABLES IN RESPONDENTS EXHIBIT 49, THOSE VARIABLES THAT SHOW
STATISTICAL SIGNIFICANCE OF .05 LEVEL OF SIGNIFICANCE AND
CLASSIFIED THEM ACCORDINGLY.
@. AND WHAT OBSERVATIONS DID YOU MAKE BASED ON THESE TWO
TABLES?
A. BASED ON THESE TWO TABLES. WHITE VICTIM CASES HAVE A HIGHER
fn
DR
E
S
E
R
E
N
E
f
T
E
O
L
E
Pe
Po
o
h
N
v
.
a
ul
16
17
KATZ - DIRECT
\
PERCENTAGE AT THE .0S LEVEL SIGNIFICANCE AND MORE AGGRAVATING
FACTORS THAN BLACK VICTIM CASES; AND BLACK VICTIM CASES HAVE A
HIGHER PERCENTAGE OF MORE MITIGATING FACTORS AT THE .0S LEVEL OF
SIGNIFICANCE THAN THE WHITE VICTIM CASES.
MS. WESTMORELAND: YOUR HONOR, AT THIS TIME. WE WOULD
SUBMIT RESPONDENTS EXHIBITS 49 AND s0.
MR. FORD: NO ADDITIONAL OBJECTION, YOUR HONOR.
THE COURT: ALL RIGHT. THEY WILL BE ADMITTED.
YOU SAY A HIGHER PERCENTAGE. HIGHER PERCENTAGE OF
WHAT?
THE WITNESS: THE VARIABLES ARE CLASSIFIED IN TERMS OF
THE PROPORTION DIFFERENCE. S0., FOR EXAMPLE, ON RESPONDENT'S
EXHIBIT 49, WE LOOK AT THE ORIGINAL VARIABLE, OR THE FIRST
VARIABLE, ARMED ROBBERY, WHICH REPRESENTS ARMED ROBBERY,
REPRESENTS ARMED ROBBERY WAS A CONTEMPORANEOUS OFFENSE TO THE
HOMICIDE, WHITE VICTIM CASES HAVE A PERCENTAGE OF 33.3 PERCENT:
BLACK VICTIM CASES HAVE A PERCENT OF 7.4 PERCENT. THE PERCENT
DIFFERNCE IS 25.9. THE "ZI" VALUE IS 10.179.
SO THE POSITIVE SIGNIFICANCE OF THE "Z" VALUE INDICATES
THAT WHITE VICTIM CASES HAVE A HIGHER PERCENTAGE OF THAT
VARIABLE THAN BLACK VICTIM CASES.
THE COURT: ALL RIGHT.
BY MS. WESTMORELAND:
@. DOCTOR KATZ, IN PROCEEDING WITH YOUR ANALYSIS OF THIS STUDY.
DID YOU DO ANY ANALYSIS BASED ON THE SENTENCING CUTCOMES?
1336
KATZ - DIRECT
A. YES. I DID.
@. AND WOULD YOU THEN REFER TO RESPONDENTS EXHIBIT 51, TABLE
33, AND IDENTIFY THAT DOCUMENT. PLEASE?
A. YES. IN RESPONDENT’S EXHIBIT S51 I DID A SIMILAR ANALYSIS TO
THE EARLIER EXHIBITS IN WHICH I LOOKED AT ALL CASES IN WHICH A
VOLUNTARY MANSLAUGHTER SENTENCE HAD BEEN GIVEN AND BROKE DOWN
THESE CASES IN TERMS OF THE RACE OF THE VICTIM AND COMPARED THE
PERCENT DIFFERENCES FOR EACH OF THE VARIABLES DEFINED IN
RESPONDENTS EXHIBIT 46.
@. AND THEN WOULD YOU IDENTIFY RESPONDENT ’S EXHIBIT
S52, PLEASE?
A. YES, IN RESPONDENT’ S EXHIBIT 52, I SELECTED THOSE
VARIABLES FROM RESPONDENTS EXHIBIT S51 WHICH SHOWED SIGNIFICANCE
AT THE .0S LEVEL. AND CLASSIFIED THEM ACCORDINGLY.
THE COURT: I DON’T KNOW WHAT S51 IS YET. LET ME -—-
ACCORDING TO YOUR COUNT THERE WERE 134 WHITE VICTIM CASES WHERE
ARMED ROBBERY WAS A CONTEMPORANEOUS OFFENSE —-- NO, THERE WERE
134 WHITE VICTIM CASES.
THE WITNESS: YES, YOUR HONOR.
THE COURT: EIGHT OF THEM HAD ARMED ROBBERY AS A
CONTEMPORANEOUS OFFENSE OR IT WAS PRESENT IN SIX PERCENT OF THE
CASES?
THE WITNESS: YES. YOUR HONOR.
THE COURT: SIX OF THE BLACK VICTIM CASES, OUT OF A
TOTAL OF 326, HAD THAT?
S
B
E
L
E
8)
1537
KATZ - DIRECT
THE WITNESS: YES, YOUR HONOR.
THE COURT: AM I READING IT RIGHT?
THE WITNESS: YES. YOUR HONOR.
THE COURT: ALL RIGHT.
MS. WESTMORELAND: AND THIS IS IN VOLUNTARY
MANSLAUGHTER CASES. YOUR HONOR.
THE COURT: YES.
BY MS. WESTMORELAND:
@. AND DOCTOR KATZ, DID YOU MAKE ANY OBSERVATIONS BASED ON
THESE TWO TABLES?
A. 1 OBSERVED THAT ACCORDING TO TABLE, RESPONDENT’S EXHIBIT 52
THAT THERE ARE MORE AGGRAVATING CIRCUMSTANCES IN WHICH WHITE
VICTIM CASES HAVE A HIGHER PROPORTION OF THAT CIRCUMSTANCE THAN
BLACK VICTIM CASES. AND I AM UNABLE TO COME TO ANY STATISTICAL
CONCLUSION ABOUT THE MITIGATING FACTORS. AS THERE ARE MITIGATING
FACTORS ON BOTH SIDES.
@. BY MITIGATING FACTORS ON BOTH SIDES. WHAT DO YOU MEAN?
A. ACCORDING TO THE RESPONDENTS EXHIBIT S52, THERE ARE SEVERAL
MITIGATING FACTORS THAT RELATE TO WHITE VICTIM CASES AND THERE
ARE SEVERAL MITIGATING FACTORS THAT RELATE TQ BLACK VICTIM
CASES. AND THAT'S WHAT TABLE 52 SHOWS. EXCUSE ME RESPONDENTS
EXHIBIT 352 SHOWS.
BY THE COURT:
R. LET ME ASK YOU A QUESTION ABOUT THAT.
AS I UNDERSTOOD YOUR PREVIOUS TESTIMONY, YOU HAVE MADE
1538
KATZ - DIRECT
1 THE OBSERVATION ABOUT AGGRAVATING AND MITIGATING BASED ON
SOMETHING OF A RAW COUNT OF THE NUMBER OF AGGRAVATING VARIABLES [8]
PRESENT AND THE NUMBER OF MITIGATING VARIABLES PRESENT.
IN THIS PARTICULAR TABLE. THE NUMBER OF MITIGATING. IF
YOU’RE DOING IT ON A RAW COUNT, THE NUMBER OF MITIGATING FACTORS
ON THE WHITE SIDE IS GREATER THAN ON THE BLACK SIDE, SO WHY
COULDNT WE CONCLUDE. ON THE BASIS OF THIS ROUGH ANALYSIS, THAT
WHITE VICTIM CASES ARE BOTH MORE AGGRAVATED AND MITIGATED THAN
BLACK VICTIM CASES?
©
.
HH
o
N
R
S
W
P
y
A. THATS DEFINITELY A POSSIBLE CONCLUSION.
11 @. WELL. WHY DIDNT YOU CONCLUDE THAT?
12 A. WELL. TO CONCLUDE THAT KIND OF, OR MAKE THAT KIND OF
13 CONCLUSION, I WOULD REALLY NEED MORE INFORMATION ABOUT THE
14 RELATIVE IMPORTANCE OF THESE VARIABLES, AND HOW TO WEIGHT EACH
15 OF THESE VARIABLES.
16 @. THE POINT IS THIS: HAVEN‘T YOU BEEN DRAWING THAT SORT OF
17 CONCLUSION BASED ON THAT SORT OF COUNT IN THE EARLIER TABLES?
13 A. IN THE EARLIER TABLES. THERE WASN‘T GENERALLY ANY POINT AT
é 19 WHICH ONE COULD COME TO A CONCLUSION OTHER THAN WHITE VICTIM
20 CASES ARE WORSE THAN BLACK VICTIM CASES WITHOUT GIVING VERY
21 EXTREME WEIGHTS TO THE FEW VARIABLES IN WHICH BLACK VICTIM CASES
22 WERE EITHER MORE AGGRAVATED OR THE VARIABLES IN WHICH WHITE
23 IVICTIM CASES WERE MORE MITIGATED.
24 NOW WE SEE THAT IN COMPARING WHITE AND BLACK VICTIM
23 CASES ACROSS VOLUNTARY MANSLAUGHTER THAT WE GET MIXED RESULTS.
Oo
~N
6
4
H+
»
O
W
N
KATZ — DIRECT
THE COURT: ALL RIGHT. GO AHEAD. MS. WESTMORELAND.
BY MS. WESTMORELAND:
Q. DOCTOR KATZ. IN YOUR OBSERVATIONS FROM PRIOR TABLES. HAVE
You ATTEMPTED TO MAKE ANY JUDGMENT ABOUT THE PARTICULAR
AGGRAVATING OR MITIGATING FACTORS PRESENT?
A. IN THE PRIOR TABLES IVE TRIED TO NOTE THAT THEY WERE
PRESENT BUT THERE ARE MANY DIFFERENT WAYS THAT THESE NUMBERS CAN
BE ANALYZED. AND IT WOULD DEPEND ON HOW THESE WEIGHTS WERE
APPLIED.
ALSO IT WOULD DEPEND ON THE DEGREE OF SIGNIFICANCE.
MIGHT BE A FACTOR IN DETERMINING WHETHER OVERALL, WHITE VICTIM
CASES ARE WORSE THAN BLACK VICTIM CASES.
THE GENERAL INDICATION THAT 1 MADE WAS THAT WHITE
VICTIM CASES HAD MORE AND USUALLY MANY MORE AGGRAVATING |
CIRCUMSTANCES IN WHICH THEY WERE SIGNIFICANT, AND BLACK VICTIM
CASES HAD MORE MITIGATING CIRCUMSTANCES.
AND GENERALLY THERE WAS REALLY NO QUESTION UNLESS VERY
EXTREME WEIGHTS WERE USED IN, IN THE CONCLUSION.
IN TABLE. EXCUSE ME, RESPONDENTS EXHIBIT 352. THERE CAN
ARISE SOME QUESTIONS AS TO WHICH CASES ARE MORE AGGRAVATED OR
WHICH CASES ARE MORE MITIGATED. AND I DON’T THINK 1 CAN FAIRLY
MAKE THOSE KINDS OF DECISIONS.
MS. WESTMORELAND: YOUR HONOR, AT THIS TIME, WE WOULD
SUBMIT RESPONDENT“3 EXHIBITS S51 AND 352.
MR. FORD: NO ADDITIONAL --
13540
KATZ - DIRECT
1 THE COURT: NOTHING ADDITIONAL, THEY’LL BE ADMITTED.
2 BY MS. WESTMORELAND:
3 Q. DOCTOR KATZ, DID YOU CONDUCT THE SAME SORT OF ANALYSIS BASED
4 ON LIFE SENTENCE CASES?
A. YES, I DID.
@. WOULD YOU THEN REFER TO RESPONDENTS EXHIBIT 3537?
@. AND IDENTIFY THAT DOCUMENT. WHICH IS TABLE 237
S
&
7 A. YES.
3
? A. YES. IN RESPONDENT’S EXHIBIT 53, I DID A SIMILAR KIND OF
10 |ANALYSIS, EXCEPT USING LIFE SENTENCE CASES. COMPARING
11 WHITE-BLACK VICTIM CASES IN TERMS OF ALL THE VARIABLES THAT WERE
12 |DEFINED IN RESPONDENT’S EXHIBIT 46.
13 |@. AND THEN IF YOU WOULD LOOK AT RESPONDENT'S EXHIBIT S54 AND
14 IDENTIFY THAT DOCUMENT, WHICH IS TABLE 367
15 A. YES. IN RESPONDENTS EXHIBIT .S4, I SELECTED THOSE VARIABLES
16 IN RESPONDENT-S EXHIBIT 53 THAT WERE SIGNIFICANT AT THE .0S
17 |LEVEL AND LISTED THEM.
13 |. AND WHAT OBSERVATIONS DID YOU MAKE BASED ON THESE TWO
, 19 |TABLES?
20 A. HERE THERE ARE MORE AGGRAVATING FACTORS FOR WHICH WHITE
21 |VICTIM CASES HAVE A HIGHER PERCENTAGE THAN BLACK VICTIM CASES
22 |AND THERE ARE MORE MITIGATING FACTORS IN WHICH BLACK VICTIM
23 CASES HAVE A HIGHER PERCENTAGE THAN WHITE VICTIM CASES.
24 MS. WESTMORELAND: YOUR HONOR, WE SUBMIT THESE TWO
25 EXHIBITS, S53 AND 54 AT THIS TIME.
1541
KATZ ~ DIRECT
1 THE COURT: STANDARD OBJECTION?
2 MR. FORD: I KNOW IT IS, BY NOW IT IS STANDARD, YOUR
3 HONOR.
4 THE COURT: ALL RIGHT. NOTHING ADDITIONAL. I WILL
3 ADMIT IT -— ADMIT THEM.
& BY MS. WESTMORELAND:
®@. DOCTOR KATZ, I REFER YOU TO RESPONDENTS EXHIBIT S55 WHICH IS
7
a TABLE 37. AND ASK YOU TO IDENTIFY THAT DOCUMENT. PLEASE?
9 A. YES. IN THIS EXHIBIT. I BROKE DOWN THE COMPARISONS IN TERMS
0 OF THOSE CASES IN WHICH A LIFE SENTENCE WAS GIVEN AND NO PENALTY
11 TRIAL WAS CONDUCTED.
12 THOSE CASES IN WHICH IT WAS UNKNOWN WHETHER A PENALTY
13 TRIAL WAS CONDUCTED WERE OMITTED FROM THIS ANALYSIS.
14 @. AND THEN DID YOU DO A SIMILAR TABLE AS TO THE LAST SEVERAL
13 TABLES, BREAKING OUT THE SIGNIFICANCE, THE ITEMS OF SIGNIFICANT
16 VALUES?
17 A. YES. THAT WOULD BE RESPONDENT’S EXHIBIT 36.
18 Gl. WHICH IS TABLE 387
a 19 A. YES. AND RESPONDENT’S EXHIBIT S& SELECTS THOSE VARIABLES IN
20 RESPONDENTS EXHIBIT 55 WHICH SHOWED A SIGNIFICANCE AT THE .035
21 LEVEL, AND CLASSIFIES THEM.
22 @. AND WHAT OBSERVATIONS DID YOU MAKE BASED ON THOSE TWO
=23 TABLES?
24 A. BASED ON RESPONDENT’S EXHIBIT 5&4. IT APPEARS THAT THERE ARE
23 A FEW MORE WHAT. A FEW MORE AGGRAVATING FACTORS WHICH WHITE
1542
KATZ - DIRECT
VICTIM CASES HAVE A HIGHER PROPORTION OF THAN BLACK VICTIM
CASES. AND IT APPEARS THAT BLACK VICTIM CASES HAVE A FEW MORE
MITIGATING FACTORS IN WHICH THEY REPRESENT A HIGHER PROPORTION
THAN WHITE VICTIM CASES.
@. DID YOU ATTEMPT TO MAKE ANY QUALITATIVE JUDGMENTS ON THIS
INFORMATION?
A. NO. THE, THE JUDGMENT AS TO WHAT IS, WHICH SIDE IS MORE
AGGRAVATED OR MORE MITIGATED RESTS ON ADDITIONAL CONSIDERATIONS
IN TERMS OF THE WEIGHTING AND THE DEGREE OF SIGNIFICANCE. AND I
DID NOT TRY AND GO THAT FAR IN MY ANALYSIS.
MS. WESTMORELAND: YOUR HONOR. AT THIS TIME. WE WOULD
SUBMIT RESPONDENTS EXHIBITS 35 AND 56.
THE COURT: ALL RIGHT. SUBJECT TO THE STANDARD
OBJECTION. I WILL ADMIT IT.
MR. FORD: THANK YOU, YOUR HONOR.
BY MS. WESTMORELAND:
@. AND DOCTOR KATZ, ILL THEN ASK YOU TO LOOK AT RESPONDENTS
EXHIBIT 57, WHICH IS TABLE 39 AND ASK YOU TO IDENTIFY THAT
DOCUMENT?
A. YES. IN RESPONDENT’S EXHIBIT 57, I COMPARED THOSE CASES IN
WHICH A LIFE SENTENCE WAS IMPOSED AND THESE CASES ADVANCED TO A
PENALTY TRIAL.
I COMPARED THEM IN TERMS OF WHITE VICTIM AND BLACK
VICTIM CASES FOR ALL THE VARIABLES THAT WERE DEFINED IN
RESPONDENTS EXHIBIT 46.
E
o
H
R
BE
|
EE
R
4
S
R
C
S
E
R
EE
f
e
o
Pa
y
P
b
KATZ - DIRECT
2. AND THEN WOULD YOU IDENTIFY RESPONDENTS EXHIBIT 58, TABLE
407
A. YES. IN RESPONDENT’S EXHIBIT 58. I THEN SELECT THOSE
VARIABLES FROM RESPONDENT'S EXHIBIT S57 THAT SHOW SIGNIFICANCE AT
THE .0S5 LEVEL OF SIGNIFICANCE.
@. WERE YOU ABLE TO MAKE ANY CONCLUSIONS FROM THESE TWO TABLES?
A. NO.
Q. AND WHY WERE YOU UNABLE TO MAKE ANY SUCH CONCLUSIONS?
A. WELL, IM UNABLE TO CONCLUDE ONE SIDE AS HAVING OR BEING
MORE AGGRAVATED THAN ANOTHER SIDE IN THE SENSE OF WHITE VICTIM
CASES BEING MORE AGGRAVATED THAN THE BLACK VICTIM CASES, BECAUSE
THERE SEEM TO BE AGGRAVATING FACTORS ON BOTH SIDES AND
MITIGATING FACTORS. ESPECIALLY IN TERMS OF BLACK VICTIM CASES.
HOWEVER, I NOTED THAT AS WE COMPARED WHITE AND BLACK
VICTIM CASES ON EACH SENTENCE. THAT THE NUMBER OF SIGNIFICANT
VARIABLES FOR BOTH THE AGGRAVATED AN MITIGATED, HAS BEEN REDUCED
IN TERMS OF THE ORIGINAL COMPARISON, SIMPLY BETWEEN WHITE AND
BLACK VICTIM CASES FOR ALL CASES.
MS. WESTMORELAND: YOUR HONOR, THEN, AT THIS TIME WE
WOULD SUBMIT RESPONDENT’S EXHIBITS S57 AND S53.
THE COURT: ALL RIGHT. SUBJECT TO MR. FORD-S STANDING
OBJECTION, I‘LL ADMIT IT.
BY MS. WESTMORELAND:
@. DOCTOR KATZ,» THEN, DID YOU MAKE ANY COMPARISON BASED ON THE
DEATH SENTENCE CASES?
1544
KATZ - DIRECT
1 A. YES.
2 @. AND WOULD YOU TURN TO RESPONDENTS EXHIBIT 59, TABLE 41. AND
IDENTIFY THAT DOCUMENT, PLEASE?
A. YES. IN RESPONDENTS EXHIBIT 59, I COMPARED DEATH SENTENCE
CASES FOR WHITE AND BLACK VICTIM CASES FOR EACH OF THE VARIABLES
3
4
3
il é DEFINED IN RESPONDENT “3 EXHIBIT 44.
7 @. AND THEN, IF YOU“D LOOK AT RESPONDENT-S EXHIBIT 40 AND
3 IDENTIFY THAT DOCUMENT? THAT'S TABLE 42.
7 A. YES. IN RESPONDENTS EXHIBIT 4&0 I THEN SELECTED THOSE
10 VARIABLES FROM RESPONDENT’ S EXHIBIT 59 THAT SHOWED SIGNIFICANCE
AT THE .0S LEVEL OF SIGNIFICANCE.
12 Q. AND WHAT OBSERVATIONS DID YOU MAKE FROM THOSE TWO TABLES?
13 A. IN RESPONDENT“ S EXHIBIT 460 IT APPEARS THAT WHITE VICTIM
14 CASES HAVE MORE AGGRAVATING CIRCUMSTANCES IN WHICH THEY SHOW
13 SIGNIFICANCE COMPARED TO BLACK VICTIM CASES. AND BLACK VICTIM
16 CASES SHOW MORE MITIGATING FACTORS IN WHICH THEY ARE SIGNIFICANT
17 THAN WHITE VICTIM CASES.
13 MS. WESTMORELAND: YOUR HONOR. WE WOULD THEN SUBMIT
a 19 |RESPONDENT-S EXHIBIT S9 AND 40.
20 THE COURT: ANYWHERE THE DEATH PENALTY WAS IMPOSED?
21 THE WITNESS: YES, YOUR HONOR.
22 THE COURT: IS THERE OBJECTION?
23 : MR. FORD: THERE IS THE SAME OBJECTION, YOUR HONOR.
24 THE COURT: ALL RIGHT, I ADMIT IT.
23 BY MS. WESTMORELAND:
po
te
P
o
d
e
o
a
[w
y
Po
y 8)
po
t ry
F
y
I
Ty
U
a
y
oo
pa
ws
~
18
19
v
O
0
O
N
6
+
W
O
N
1543
KATZ ~ DIRECT
@R. DOCTOR KATZ. DID YOU DO ANY ——
BY THE COURT:
QR. LET ME ASK YOU A QUESTION.
IS THIS, THIS IS A MEASUREMENT OF THE AVAILABILITY. NOT
AVAILABILITY. MEASUREMENT OF THE PRESENCE OF THESE CASES WHICH
GOT THE DEATH PENALTY?
A. YES.
el. PART OF THIS ORIGINAL HYPOTHESIS. IF I REMEMBER CORRECTLY,
HAD TO DO WITH THE SYSTEM'S HANDLING OF MITIGATING
CIRCUMSTANCES.
DOES THIS TABLE TEND TO DISPROVE OR PROVE THE PORTION
OF THE HYPOTHESIS DEALING WITH MITIGATING CIRCUMSTANCES?
A. I‘M SORRY. I“M NOT SURE I UNDERSTAND THE QUESTION.
@. I“M NOT SURE I CAN ASK THE QUESTION. I DON‘T KNOW HOW TO
ASK THE QUESTION.
WHAT I“M GETTING AT IS THIS: AS A VISCERAL REACTION IT
STRIKES ME THAT ANY TIME A CASE IS AGGRAVATED, YOU START
THINKING ABOUT IT IN TERMS OF A DEATH PENALTY CASE.
AND SO IN TERMS OF APPROPRIATENESS OF THE RESPONSE IN
THE MORE AGGRAVATED CASES: I.E.» THE WHITE VICTIM CASES. THEN
THERE MAY NOT BE ANY GQUARREL WITH THAT RESPONSE. NOT THAT I
FOUND THIS, BUT FOR PURPOSES OF ARGUMENT.
BUT IF THE DEATH PENALTY IS BEING IMPOSED IN CASES THAT
ARE MORE MITIGATED THAN THE HIGHLY AGGRAVATED CASES. WHAT DOES
THAT TELL YOU ABOUT HOW THE SYSTEM IS WORKING?
wo
S
r
E
R
Je
e
1
S
e
1
B
l
a
1
1346
KATZ - DIRECT
A. WELL. I BELIEVE THE VARIABLES ARE MORE MITIGATED IN THE
SENSE THAT THE UNDERLYING CIRCUMSTANCE THAT WAS PRESENT THAT
PRECIPITATED THE, THE MURDER, WAS A MITIGATING CIRCUMSTANCE.
BUT THE AMOUNT OF AGGRAVATION THAT WAS DONE IN THE
BLACK VICTIM CASES WAS PROBABLY THOUGHT TO BE SUFFICIENT TO SEEK
THE DEATH PENALTY, AND TO IMPOSE A DEATH PENALTY.
@. ALL RIGHT. JUST A SECOND.
THE COURT: ALL RIGHT, GO AHEAD, MS. WESTMORELAND.
BY MS. WESTMORELAND:
@. DOCTOR KATZ, DID YOU DO A BREAKDOWN OF THESE CASES IN THE
CHARGING AND SENTENCING STUDY BASED ON THE DEFENDANT-VICTIM
RACIAL COMBINATION?
A. YES.
@. WOULD YOU THEN EXAMINE RESPONDENTS EXHIBIT é1. TABLE 437
A. YES. |
@. AND IDENTIFY THAT DOCUMENT. PLEASE?
A. IN RESPONDENT“S EXHIBIT 41, I BROKE DOWN THE PERCENTAGES FOR
EACH OF THE VARIABLES DEFINED IN RESPONDENT-S EXHIBIT 44, FOR
ALL CASES. BY DEFENDANT-VICTIM RACIAL COMBINATION.
AGAIN. CASES IN WHICH THE RACE OF THE VICTIM WAS
UNKNOWN WERE EXCLUDED FROM THESE TABLES.
MS. WESTMORELAND: YOUR HONOR. WE WOULD SUBMIT
RESPONDENT’S EXHIBIT NUMBER 41 TO SHOW THE SUBSEQUENT PROCEDURES
THAT WERE UTILIZED AND AS A BASIS FOR THE SUBSEQUENT TABLES.
THE COURT: WHERE ARE YOU GOING, RACE OF THE DEFENDANT
13547
KATZ - DIRECT
i
OR RACE OF THE VICTIM EFFECT?
2 MS. WESTMORELAND: IT’S LOOKING AT BOTH, WELL IT“S RACE
3 OF DEFENDANT EFFECT IN THE SENSE THAT WE‘RE LOOKING AT
4 DEFENDANT-VICTIM RACIAL COMBINATIONS.
3 THE COURT: IF THERE“S NOTHING MORE THAN A STANDARD
8 3 OBJECTION I“LL ADMIT IT. ;
7 MR. FORD: THANK YOU, YOUR HONOR.
8 BY MS. WESTMORELAND:
2 A. DOCTOR KATZ, WOULD YOU THEN LOOK AT RESPONDENTS EXHIBIT 62
10 AND IDENTIFY THAT EXHIBIT, PLEASE
11 A. YES. IN RESPONDENT’S EXHIBIT 42. I THEN BROKE DOWN THE
2 DEFENDANT-VICTIM RACIAL COMBINATIONS IN TERMS OF ALL THE
13 VARIABLES IN WHICH A VOLUNTARY MANSLAUGHTER SENTENCE WAS
14 IMPOSED.
15 @. AND DID YOU DO THE SAME ANALYSIS FOR THE OTHER SENTENCING
14 OUTCOMES?
17 A. YES, I DID.
13 Q. WOULD YOU EXAMINE RESPONDENT’S EXHIBIT 63 AND IDENTIFY THAT?
Wu 19 A. YES. IN RESPONDENTS EXHIBIT 43 I THEN COMPARED CASES IN
20 WHICH LIFE SENTENCE AND NO PENALTY TRIAL HAD BEEN CONDUCTED.
2 BROKEN DOWN BY DEFENDANT-VICTIM RACIAL COMBINATION IN TERMS OF
22 ALL THE VARIABLES DEFINED EARLIER IN RESPONDENTS EXHIBIT 46.
23 @. AND WOULD YOU THEN EXAMINE RESPONDENT’S EXHIBIT 44, AND
24 IDENTIFY THAT DOCUMENT. PLEASE?
25 A. YES, IN RESPONDENT-S EXHIBIT 44, I THEN BROKE DOWN ALL THE
1348
KATZ - DIRECT
VARIABLES DEFINED IN RESPONDENTS EXHIBIT 44 IN TERMS OF
DEFENDANT-VICTIM RACIAL COMBINATION, FOR THOSE CASES IN WHICH A
LIFE SENTENCE AND A PENALTY TRIAL WAS HELD.
MS. WESTMORELAND: I“M SORRY, COUNSEL, THAT'S TABLE 46.
THE COURT: WHAT HAVE YOU GOT 80 FAR, MANSLAUGHTER. IS
| THAT RIGHT?
MS. WESTMORELAND: YES, YOUR HONOR, RESPONDENT’S
EXHIBIT 62. | | E
THE COURT: YOU‘VE GOT LIFE WITH PENALTY TRIAL.
MS. WESTMORELAND: 63 WAS LIFE, NO PENALTY TRIAL.
THE COURT: 64 IS LIFE. PENALTY TRIAL.
MS. WESTMORELAND: THAT‘S CORRECT. YOUR HONOR.
THE COURT: AND I ASSUME &6 IS DEATH SENTENCE.
MS. WESTMORELAND: 65, YOUR HONOR.
BY MS. WESTMORELAND:
0. DOCTOR KATZ,» WOULD YOU LOOK AT RESPONDENT’S EXHIBIT &57
A. YES.
Q. AND TABLE 47 AND IDENTIFY THAT DOCUMENT?
A. IN RESPONDENT’S EXHIBIT 4S I COMPARE ALL CASES IN WHICH A
DEATH SENTENCE WAS IMPOSED, BROKEN DOWN BY DEFENDANT-VICTIM
RACIAL COMBINATION IN MATERIALS OF ALL THE VARIABLES DEFINED IN
RESPONDENTS EXHIBIT 46.
MS. WESTMORELAND: YOUR HONOR. AT THIS TIME, WE WOULD
SUBMIT RESPONDENT‘S EXHIBITS 42. 43, &4 AND 645.
THE COURT: ANYTHING NEW?
-
0
N
o
b
d
a
E
N
[S
Y Co
Fo
rs
Fo
ry
12
13
1549
KATZ - DIRECT
MR. FORD: NOTHING NEW. YOUR HONOR.
THE COURT: WE-“LL ADMIT THEM, AND SEE WHERE WE GO.
BY MS. WESTMORELAND:
@. DOCTOR KATZ, IN RELATION TO THE CHARGING AND SENTENCING
STUDY AND THE INFORMATION IN THE CHARGING AND SENTENCING STUDY.
DID YOU MAKE AN EXAMINATION OF THE CASES PARTICULARLY IN
RELATION TO FULTON COUNTY CASES?
A. YES, 1 DID...
THE COURT: NOW. WAIT A MINUTE. DID HE MAKE ANY
CONCLUSIONS ON THAT LAST SERIES OF CHARTS, SINCE YOU DIDNT GIVE
ME ANY NICE LITTLE SUMMARY CHARTS, GIVE ME SOME CONCLUSIONS.
MS. WESTMORELAND: ALL RIGHT, YOUR HONOR.
BY MS. WESTMORELAND:
@. DOCTOR KATZ, IF YOU WOULD REFER BACK THEN. AND EXAMINE, I
BELIEVE, BEGINNING WITH RESPONDENT'S EXHIBIT 62, IF YOU WILL
EXAMINE THIS PARTICULAR TABLE AND TELL US WHAT OBSERVATIONS YOU
HAVE FROM THIS TABLE?
THE COURT: PARTICULARLY IM INTERESTED IN WHAT IF ANY
OBSERVATIONS HE HAS ON RACE OF DEFENDANT EFFECT.
THE WITNESS: THAT I BELIEVE, WILL BE RESFONDENT’S
EXHIBIT 41.
IN RESPONDENT'S EXHIBIT 4&1, IN WHICH ALL CASES ARE
BROKEN DOWN BY DEFENDANT-VICTIM RACIAL COMBINATION, THERE
APPEARS TO BE A PATTERN BASED ON THE DEFENDANT-VICTIM RACIAL
COMBINATION EFFECT. AND IF I CAN CITE SOME OF THESE CATEGORIES.
9g
0
~N
6
A
Ww
13350
KATZ - DIRECT
SUCH AS CONTEMPORANEOUS OFFENSE, AND CITE SOME OF THE NUMBERS.
WHEN A BLACK DEFENDANT KILLS A WHITE VICTIM, 47.1 FERCENT OF THE
TIME ARMED ROBBERY IS PRESENT.
WHEN A WHITE DEFENDANT KILLS A WHITE VICTIM, 20.2
PERCENT OF THE TIME ARMED ROBBERY IS PRESENT.
4.4 PERCENT OF THE TIME WHEN A BLACK DEFENDANT KILLS A
BLACK VICTIM.
AND 22.2 PERCENT WHEN A WHITE DEFENDANT KILLS A BLACK
VICTIM.
I COMPUTED SOME MEASURES OF SIGNIFICANT DIFFERENCES.
BUT THAT WOULD HAVE REQUIRED MANY MORE TABLES AND I“VE AVOIDED
THEM HERE, S50 I“VE LEFT IT OVERALL AS THE PERCENTAGES.
THEN TO CITE A MITIGATING VARIABLE. PERHAFS PAGE. PAGE
4 OF THIS EXHIBIT, UNDER THE CATEGORY OF SPECIAL PRECIPITATING
EVENTS, INDICATION - OF THE RELATIONSHIP BETWEEN
DEFENDANT-VICTIM RACIAL COMBINATION.
THE FIRST VARIABLE REPRESENTS BARROOM-LOVER QUARREL.
AND IT’S SPELLED B-L-V-I-C-M-0-D, AND IN IN THIS CASE. WHEN
BLACK DEFENDANTS KILL WHITE VICTIMS, ONLY SEVEN AND A HALF
PERCENT OF TIME IS THIS PRECIPITATING CIRCUMSTANCE PRESENT.
WHEN WHITE DEFENDANTS KILL WHITE VICTIMS, IT’S 43.6
PERCENT.
WHEN BLACK DEFENDANTS KILL BLACK VICTIMS, 65 PERCENT.
WHEN WHITE DEFENDANTS KILL BLACK VICTIMS, IT'S 45.8
PERCENT.
KATZ - DIRECT
THE COURT: I CAN READ THE TABLES. BUT AS TO THE DEATH
PENALTY ON THIS ANALYSIS, WHICHEVER TABLE THAT ONE IS, DID YOU
OBSERVE ANY RACE OF THE DEFENDANT EFFECT?
THE WITNESS: OVERALL, IT APPEARS TO ME THAT IN
REVIEWING THIS TABLE THAT BLACK DEFENDANTS KILLING WHITE
VICTIMS TEND TO HAVE MORE AGGRAVATING CIRCUMSTANCES AND LESS
MITIGATING CIRCUMSTANCES THAN WHITE DEFENDANTS WHO KILL WHITE
VICTIMS. AND IN TURN THEY HAVE MORE AGGRAVATING CIRCUMSTANCES
AND LESS MITIGATING CIRCUMSTANCES IN CASES WHERE BLACK
DEFENDANTS KILL BLACK VICTIMS.
THE COURT: NOW WHAT? WHERE THE VICTIM IS BLACK, WHICH
SERIES IS MORE AGGRAVATED?
THE WITNESS: WHEN THE VICTIM IS BLACK, THAT GENERALLY
REFERS TO A CASE WHERE A BLACK DEFENDANT KILLS A BLACK VICTIM
AND OVERALL THIS TABLE SUGGESTS THATS THE MOST MITIGATED TIME
OF MURDER. THAT IS SUGGESTED BY THE 471 CASES WHERE A BLACK
DEFENDANT KILLS A BLACK VICTIM. TO 27 CASES WHERE A WHITE
DEFENDANT KILLS A BLACK VICTIM.
THE COURT: YOU SAID MITIGATING, YOU DIDNT SAY
INFREQUENT. WHAT ARE YOU LOOKING AT?
THE WITNESS: I“M SORRY, YOUR HONOR, I THOUGHT I WAS,
YOU WERE ~~
THE COURT: I DON’T KNOW WHAT YOU“RE VIEWING. 617
THE WITNESS: YES, RESPONDENT’S EXHIBIT 61.
THE COURT: PAGE?
o
N
O
v
A
0
N
e
Pr
y oO
Po
ry
pt
12
13
1352
KATZ - DIRECT
THE WITNESS: WHAT PAGE?
MS. WESTMORELAND: WHICH PAGE NUMBER OF THE EXHIBIT IS
IT. DR. KATZ?
THE WITNESS: OH. EXCUSE ME. I GUESS ALL THE PAGES I
FIND GIVE AN OVERALL IMPRESSION OF WHAT THE PATTERNS OF, ARE
REPRESENTED BY THESE NUMBERS.
THE COURT: ARE YOU SAYING THAT A KILLING ACROSS RACIAL
LINES USUALLY RESULTS IN A MORE AGGRAVATED LESS MITIGATED CASE.
AND KILLING ALONG RACIAL LINES USUALLY RESULTS IN A LESS
AGGRAVATED MORE MITIGATED CASE?
THE WITNESS: I DON‘T THINK IT
GENERALIZES QUITE LIKE THAT.
I BELIEVE WHEN A BLACK DEFENDANT KILLS A WHITE VICTIM,
THAT TENDS TO BE THE MOST AGGRAVATED CIRCUMSTANCE AS COMPARED TO
THE OTHER DEFENDANT-VICTIM RACIAL COMBINATIONS.
WHEREAS IT APPEARS THAT THE 27 CASES WHERE A WHITE
DEFENDANT KILLS A BLACK VICTIM. THEY DON’T APPEAR TO BE AS
AGGRAVATED AS WHEN A BLACK DEFENDANT KILLS A WHITE VICTIM. OR AS
NUMEROUS.
BY THE COURT:
QR. WELL, OBVIOUSLY THEY RE NOT AS NUMEROUS. BUT YOUR
OBSERVATION IS IT’S NOT AS AGGRAVATED?
@. HOW ABOUT MITIGATION? WHICH IS MORE MITIGATED?
A. THE MOST MITIGATED OF THESE DEFENDANT-VICTIM RACIAL
Og
O
N
B
P
N
fo
e ©
fy
[w
y
12
13
135333
KATZ - DIRECT
COMBINATIONS IS WHEN A BLACK DEFENDANT KILLS A BLACK VICTIM.
BY MS. WESTMORELAND:
QR. DOCTOR KATZ, IF YOU’D TURN NOW TO RESPONDENT S EXHIBIT 66»
TABLE 48, AND ASK IF YOU CAN IDENTIFY THAT DOCUMENT FOR ME,
PLEASE?
A. YES. THIS IS AN EXHIBIT I PREPARED TO ANALYZE THE DATA IN
TERMS OF FULTON COUNTY.
Q@. WERE YOU AGAIN EXCLUDING UNKNOWNS IN THESE PARTICULAR
INSTANCES IN THESE TABLES?
A. YES.
R. AND WHAT WAS DONE IN THIS PARTICULAR TABLE?
A. IN THIS PARTICULAR TABLE, I COMPARED ALL THE CASES THAT. IN
WHICH A VOLUNTARY MANSLAUGHTER, LIFE SENTENCE OR DEATH SENTENCE
WAS IMPOSED IN FULTON COUNTY IN TERMS OF ALL THREE SENTENCE
CATEGORIES, IN TERMS OF ALL THE VARIABLES THAT WERE DEFINED IN
RESPONDENTS EXHIBIT 46.
MS. WESTMORELAND: YOUR HONOR, WE SUBMIT RESPONDENTS
EXHIBIT NUMBER 464 AS A FOUNDATION TABLE FOR THE LATER —
THE COURT: WE’RE BEGINNING A SIMILAR SEQUENCE THAT
WE“VE SEEN BEFORE?
MS. WESTMORELAND: THATS CORRECT. THIS TIME IN
RELATION TO FULTON COUNTY.
THE COURT: ALL RIGHT, IT WILL BE ADMITTED.
BY MS. WESTMORELAND:
RA. DOCTOR KATZ, THEN AFTER YOU COMPARED THE INITIAL TABLE. DID
KATZ - DIRECT
Pe
rs
YOU MAKE A BREAKDOWN BASED ON THE RACE OF VICTIM?
2 A. YES, 1 DID, AND THIS IS RESFONDENT’S EXHIBIT 6&7.
3 QR. THIS IS TABLE 4%.
4 AND WOULD YOU IDENTIFY THIS TABLE, PLEASE?
S A. YES, IT BREAKS DOWN FOR THE RACE OF VICTIM ALL CASES IN
A & TERMS OF THE VARIABLES DEFINED IN RESPONDENTS EXHIBIT 46.
7 |@. THEN IF YOU COULD REFER TO RESPONDENT'S EXHIBIT 68, TABLE
3 S0, AND IDENTIFY THAT DOCUMENT? |
9 A. YES. IN RESPONDENT’S EXHIBIT 48 IVE SELECT THOSE VARIABLES
10 IN RESPONDENT“S EXHIBIT &7 THAT HAVE SIGNIFICANCE AT THE .0S
11 LEVEL OF SIGNIFICANCE.
12 |@. AND WHAT OBSERVATIONS DID YOU MAKE BASED ON THESE TWO
13 |TABLES?
14 |A. 1 OBSERVED THAT WHITE VICTIM CASES TENDED TD HAVE A HIGHER
15 | PROPORTION OF THE AGGRAVATING CIRCUMSTANCES FOR MANY VARIABLES
16 |AND BLACK VICTIM CASES HAD A HIGHER PROPORTION OF THE
17 |AGGRAVATED, EXCUSE ME, OF THE MITIGATING CIRCUMSTANCES, FOR THE
13 MITIGATING VARIABLES.
@ 19 THE COURT: PRETTY CLOSE, ISNT IT?
20 THE WITNESS: NO. YOUR HONOR.
21 THE COURT: I MAY HAVE STARTED TOQ SOON.
22 MS. WESTMORELAND: YOUR HONOR, AT THIS TIME. THEN, WE
23 WOULD SUBMIT RESPONDENTS EXHIBIT 4&7 AND 48.
24 THE COURT: ALL RIGHT. IF ONLY THE STANDING OBJECTION
23 IS MADE. THEY“LL BE ADMITTED.
fe
y
M
N
B
e
W
N
KATZ - DIRECT
MR. FORD: YES, YOUR HONOR.
BY MS. WESTMORELAND:
@. DOCTOR KATZ. DID YOU THEN DO AN ANALYSIS BASED ON SENTENCING
QUTCOMES IN FULTON COUNTY?
A. YES.
@. AND WOULD YOU REFER TO RESPONDENT-S EXHIBIT 6%. TABLE 3517
A. YES.
@. CAN YOU IDENTIFY THAT DOCUMENT, PLEASE?
A. YES. IN RESPONDENT’S EXHIBIT 49, I COMPARED WHITE AND BLACK
VICTIM CASES THAT RECEIVED A VOLUNTARY MANSLAUGHTER SENTENCE IN
TERMS OF ALL THE VARIABLES DEFINED IN RESPONDENTS EXHIBIT 46.
@. AND THEN IF YOU-D IDENTIFY RESPONDENT’S EXHIBIT 70. TABLE
327?
A. IN RESPONDENTS EXHIBIT 70 I THEN SELECTED THOSE VARIABLES
IN RESPONDENT’S EXHIBIT 6% THAT WERE SIGNIFICANT AT THE .03
LEVEL OF SIGNIFICANCE.
@. AND WHAT OBSERVATIONS DID YOU MAKE IN THESE TWO TABLES?
A. 1 OBSERVED THAT WHITE VICTIM CASES HAVE MORE AGGRAVATING
CIRCUMSTANCES THAN BLACK VICTIM CASES. AND THERE SEEM TO BE VERY
FEW MITIGATING VARIABLES THAT HAVE ANY SIGNIFICANCE.
@. BY IN ANY SIGNIFICANCE. WHAT DO YOU MEAN?
A. IN WHICH THEY SHOW A SIGNIFICANT "ZI" VALUE AT THE .03 LEVEL
OF SIGNIFICANCE.
Q. DID YOU NOTICE ANY DIFFERENCE AS TO THOSE MITIGATING
CIRCUMSTANCES BETWEEN WHITE AND BLACK VICTIM CASES?
135356
KATZ - DIRECT
A. IT APPEARS THAT BLACK VICTIM CASES HAVE FAR MORE MITIGATING
VARIABLES THAN WHITE VICTIM CASES.
THE COURT: THIS VARIABLE THAT KEEPS SHOWING UP THAT
SAYS COPERP, DOES THAT MEAN THAT THAT“S, THE DEFENDANT IS THE
COPERPETRATOR? |
THE WITNESS: YES, YOUR HONOR.
THE COURT: OR WAS A CO-PERPETRATOR?
THE WITNESS: YES. IN THE -——
THE COURT: ALL RIGHT.
THE WITNESS: -- CATEGORY OF DEFENDANT-VICTIM STATUS
VARIABLES, 1 LISTED VARIABLES THAT DEALT WITH THE STATUS OF THE
DEFENDANT OR THE VICTIM, AND THE VARIABLE COPERP INDICATES THAT
FOR VOLUNTARY MANSLAUGHTER CASES THERE WERE A SIGNIFICANT NUMBER
OF WHITE VICTIM CASES IN WHICH CO-PERPETRATORS WERE INVOLVED IN
THE HOMICIDE.
THE COURT: ALL RIGHT.
BY MS. WESTMORELAND:
@. DOES THAT REFER TO THE FACT. DOES THAT HAVE ANY REFLECTION
ON WHETHER THIS DEFENDANT WAS THE TRIGGER MAN OR DOES THIS JUST
REFLECT THERE WAS MORE THAN ONE PERSON INVOLVED IN THE CRIME?
A. 1 BELIEVE THIS VARIABLE JUST REFLECTS WHETHER OR NOT THERE
WAS MORE THAN ONE PERSON INVOLVED IN THE CRIME.
MS. WESTMORELAND: YOUR HONOR. WE SUBMIT RESPONDENT'S
EXHIBIT 49 AND 70 AT THIS TIME.
THE COURT: IF THERE ARE ONLY THE STANDARD OBJECTIONS,
1357
KATZ - DIRECT
1 THEY‘ LL BE ADMITTED.
2 THE REASON I WAS SMILING, SOMETIMES THESE COMPUTER
3 SLANG TERMS ARE MORE APT THAN THE MORE ELOGUENT TERMS, I WAS
4 LOOKING AT MAD LOVER OVER THERE.
= BY MS. WESTMORELAND:
-) @. DOCTOR KATZ, DID YOU DO AN ANALYSIS OF THE LIFE SENTENCE
7 CASES SIMILAR TO THESE PREVIOUS ANALYSES?
3 A. YES. IN RESPONDENT’S EXHIBIT 71.
? @. WOULD YOU IDENTIFY RESPONDENT’S EXHIBIT 71. TABLE 537
10 A. VES. IN RESPONDENT’S EXHIBIT 71 I COMPARED FOR ALL THE
31 VARIABLES DEFINED IN RESPONDENT’S EXHIBIT 44, WHITE AND BLACK
12 VICTIM CASES IN WHICH A LIFE SENTENCE WAS GIVEN.
13 @. AND THEN WOULD YOU IDENTIFY RESPONDENT”3 EXHIBIT 72? TABLE
14 S47?
15 A. YES. IN RESPONDENT-S EXHIBIT 72 I SELECTED THOSE VARIABLES
16 FROM RESPONDENT’S EXHIBIT 71 THAT WERE SIGNIFICANT AT THE .03
37 LEVEL OF SIGNIFICANCE.
13 @. IN LOOKING AT THESE TWO TABLES. WERE YOU ABLE TO MAKE ANY
o 19 OBSERVATIONS BASED ON THESE TWO TABLES?
20 A. IN TERMS OF LIFE SENTENCE CASES, WHITE VICTIM CASES HAVE
21 MORE AGGRAVATING FACTORS THAN BLACK VICTIM CASES. AND BLACK
R22 VICTIM CASES HAVE A FEW, SEVERAL MORE MITIGATING FACTORS THAN
23 WHITE VICTIM CASES.
24 MS. WESTMORELAND: YOUR HONOR, AT THIS TIME. WE WOULD
23 SUBMIT RESPONDENTS EXHIBITS 71 AND 72.
fo
b
(W
y o
Fo
y
pt
ga
0
N
B
B
b
W
N
13552
KATZ - DIRECT
THE COURT: ALL RIGHT. WITH THE STANDING OBJECTION
NOTED THEY WILL BE ADMITTED.
MS. WESTMORELAND: YOUR HONOR, I HAVE JUST HANDED THE
CLERK A DOCUMENT LABELED RESPONDENT’S EXHIBIT 72A. I DON'T
THINK THE COURT HAS A COPY OF THAT. UNFORTUNATELY I ONLY HAVE
TWO COPIES AT THIS TIME.
BY MS. WESTMORELAND:
@. DOCTOR KATZ, COULD YOU IDENTIFY WHAT’S BEEN MARKED AS
RESPONDENTS EXHIBIT 72A7
A. YES. 1 PROCEEDED TO CONDUCT THE ANALYSIS FOR FULTON COUNTY
AND COMPARING THE PROPORTION DIFFERENCE FOR ALL THE VARIABLES
DESCRIBED IN RESPONDENT’S EXHIBIT 46 FOR THOSE CASES IN FULTON
COUNTY WHERE A LIFE SENTENCE WAS IMPOSED AND NO PENALTY TRIAL
WAS HELD.
RESPONDENT’S EXHIBIT 72A, PAGE 1, REPRESENTS THOSE
VARIABLES THAT SHOWED SIGNIFICANCE AT THE .0S5 LEVEL IN THAT
EXPERIMENT. AND I CLASSIFIED IT ACCORDINGLY.
@. AND DID YOU, WOULD YOU IDENTIFY THE SECOND PAGE, THEN?
A. ON THE SECOND PAGE OF RESPONDENT’S EXHIBIT 72A, -—
@. AND LET ME NOTE AT THIS TIME. THERE WAS A TYPO UNDER THE
LEFT-HAND COLUMN WHERE THE VARIABLE "KIDNAP" IS WRITTEN. AND AT
DOCTOR KATZ” INSTRUCTIONS THAT HAS BEEN MARKED, I BELIEVE, IN
ALL COPIES.
A. 1 COMPARED THE CASES IN FULTON COUNTY WHICH IN WHICH A LIFE
SENTENCE WAS IMPOSED AND A PENALTY TRIAL WAS HELD. HOWEVER. IN
i
a
vg
BH
NN
G
G
bd
W
N
10
11
155%
KATZ - DIRECT
THIS CIRCUMSTANCE, THERE WERE ONLY TEN TOTAL CASES TO CONSIDER.
AND I SELECTED THOSE VARIABLES THAT WERE SIGNIFICANT AT
THE .0S LEVEL OF SIGNIFICANCE AND CLASSIFIED THEM.
@. WHAT SHOULD THE VALUE BE FOR KIDNAP, DOCTOR KATZ?
A. I BELIEVE IT’S 1.6737.
Q. WERE YOU ABLE TO MAKE ANY CONCLUSIONS BASED ON THIS EXHIBIT?
A. FOR THE LIFE SENTENCE, NO PENALTY TRIAL CASES. IT APPEARS
THAT WHITE VICTIM CASES HAVE A FEW MORE AGGRAVATING FACTORS.
HOWEVER, POTENTIAL WEIGHTING MIGHT, FOR EACH OF THESE FACTORS
MIGHT COME INTO PLAY TO MAKE A TRUE JUDGMENT AS TO WHETHER
THEY RE MORE SIGNIFICANT OR NOT.
AND IT APPEARS THAT THE BLACK VICTIM CASES HAVE SEVERAL
MITIGATING FACTORS.
@. AND THEN. WERE YOU ABLE TO MAKE ANY CONCLUSIONS AS TO THE
LIFE PENALTY TRIAL?
A. IN THE LIFE PENALTY TRIAL SITUATION, THERE ARE VERY FEW
CASES AND I BELIEVE THAT. IVE RECORDED THE NUMERICAL ANSWERS.
BUT THERE“S TOO FEW CASES TO REALLY BE CONFIDENT THAT ANY OF
THESE COMPARISONS ARE MEANINGFUL.
I BELIEVE THERE WERE A TOTAL OF 32 WHITE VICTIM CASES
AND 7 BLACK VICTIM CASES TO DO THE COMPARISONS, AND THIS IS THE,
THE MEASURES OF STATISTICAL SIGNIFICANCE ARE NOT GENERALLY
APPROPRIATE FOR THESE SMALL NUMBERS OF CASES, BECAUSE THEY ARE
GENERALLY PRESUMED TO APPLY WHEN YOU HAVE A LARGE NUMBER OF
CASES. SO I WOULD NOT REALLY COME TO ANY CONCLUSION DUE TO THE
—— —— eb itn, Seb, ett, _ Stren Re
oO
W
m
N
O
po
W
w
N
os
wr
y
pt
1560
KATZ ~ DIRECT
VERY SMALL NUMBER OF CASES THAT ARE FRESENT.
MS. WESTMORELAND: YOUR HONOR, WE SUBMIT RESPONDENTS
EXHIBIT 72A AT THIS TIME.
THE COURT: STANDING OBJECTION. MR. FORD?
MR. FORD: EXCUSE ME, YOUR HONOR. IF I“M CORRECT. THIS
SERIES THE DEATH SENTENCE. NO CONTROLS. AND IS INCOMPLETE. NO
CONCLUSION, BUT, IF IM RIGHT. MAYBE I'M ——
THE COURT: I THOUGHT WE HAD HAD 10 DEATH PENALTIES
IMPOSED IN FULTON COUNTY IN THE STUDY PERIOD.
MS. WESTMORELAND: I BELIEVE THAT'S AN ACCURATE NUMBER.
YOUR HONOR.
THE COURT: HUH?
MS. WESTMORELAND: I BELIEVE THAT'S ACCURATE.
THE COURT: OUT OF 20 TRIALS?
MS. WESTMORELAND: I THINK THAT SOUNDS ABOUT RIGHT.
I“M NOT CERTAIN.
THE COURT: I WAS THINKING DOCTOR KATZ JUST TESTIFIED
HE HAD TEN AND FIVE.
THE WITNESS: I HAD TEN LIFE SENTENCE PENALTY TRIAL
CASES AND TEN DEATH SENTENCE CASES IN FULTON COUNTY. I HAVEN'T
INCLUDED AN EXHIBIT TO REPRESENT THE DIFFERENCES FOR THE DEATH
SENTENCES.
THE COURT: YOU HAVE?
THE WITNESS: I HAVE NOT. I DO NOT HAVE AN EXHIBIT
PRESENT HERE TO REPRESENT THAT. PART OF THE REASON WAS THERE
KATZ ~ DIRECT
fe
y WERE SO FEW CASES THAT VERY LITTLE MEANING CAN BE ATTACHED TQ
2 THOSE COMPARISONS.
3 THE COURT: I THINK OUT OF COMPLETENESS. IF YOU’VE GOT
4 IT YOU OUGHT TO PUT IT IN.
= THE WITNESS: YES. YOUR HONOR.
i 8 THE COURT: ALL RIGHT. I“LL ADMIT 72 AND 772A.
7 BY MS. WESTMORELAND:
3 QR. DOCTOR KATZ. AGAIN IN RELATION TO FULTON COUNTY. DID YOU
@ MAKE A BREAKDOWN BASED ON DEFENDANT-VICTIM RACIAL COMBINATION?
10 A. YES.
11 R. AND WOULD YOU EXAMINE RESPONDENT“S EXHIBIT 73, TABLE 35 AND
12 IDENTIFY THAT DOCUMENT?
13 A. YES. IN RESPONDENT’S EXHIBIT 73. I BROKE DOWN ALL THE CASES
14 IN FULTON COUNTY BY DEFENDANT-VICTIM RACIAL COMBINATION IN TERMS
13 OF ALL THE VARIABLES DEFINED IN RESPONDENT S EXHIBIT 46.
146 @. AND WERE YOU ABLE TO MAKE ANY CONCLUSIONS FROM THIS TABLE?
17 A. AGAIN, EVEN IN FULTON COUNTY, IT APPEARED THAT BLACK
1a DEFENDANTS KILLING WHITE VICTIMS REPRESENT MORE AGGRAVATED CASES
wo 19 OVERALL THAN WHITE DEFENDANTS KILLING WHITE VICTIMS. AS COMPARED
20 TO BLACK DEFENDANTS KILLING BLACK VICTIMS. AND THERE WERE ONLY
21 TWO CASES IN WHICH A WHITE DEFENDANT KILLED A BLACK VICTIM.
22 SIMILARLY FOR MITIGATING CIRCUMSTANCES, OVERALL, IT
23 APPEARS THAT BLACK DEFENDANTS KILLING BLACK VICTIMS ARE THE MOST
24 MITIGATED. WHITE DEFENDANTS KILLING WHITE VICTIMS ARE SOMEWHAT
23 LESS MITIGATED OVERALL. AND BLACK DEFENDANTS KILLING WHITE
1562
KATZ - DIRECT
VICTIMS ARE THE LEAST MITIGATED. AND AGAIN ONLY TWO CASES ARE
PRESENT FOR WHITE DEFENDANTS KILLING BLACK VICTIMS.
R. AND ONCE AGAIN, WERE THE UNKNOWNS OMITTED IN YOUR
CALCULATIONS IN THIS TABLE?
A. TO DO THE TABLES FOR FULTON COUNTY. I ATTEMPTED TO USE THE
MOST RECENT DATA THAT I HAD, AND THAT WAS THE DATA THAT WAS
PROVIDED TO ME ON JULY 30. AND THERE WAS ONE OF THE FILES
THERE, "ZI" TASK FILE. WHICH HAD THE UNKNOWNS CODED IN FOR, FOR
ALL THESE VARIABLES.
HOWEVER. THERE WERE SOME PROBLEMS THAT I DISCOVERED
WITH THAT FILE IN THE CODING OF CERTAIN UNKNOWNS, AND SO I HAD
TO RESORT TO USING SOME OF THE EARLIER DATA WHICH I DEEMED MORE
RELIABLE, BECAUSE THE UNKNOWNS WERE CODED WRONG IN A
MATHEMATICAL SENSE, AND USE THAT INFORMATION TO COMPLETE THESE
TABLES.
R. $S0, THE QAUESTION BASICALLY IS DID YOU EXCLUDE THE UNKNOWNS
IN DOING THESE TABLES?
A. YES.
MS. WESTMORELAND: YOUR HONOR, AT THIS TIME, WE WOULD
SUBMIT RESPONDENTS EXHIBIT 73.
THE COURT: STANDING OBJECTION, ILL ADMIT THEM.
IS ANYONE WARM BESIDE ME?
LET“S TAKE A BRIEF BREAK. MAYBE MRS. PAGE HAS SOME
INFLUENCE WITH GSA, AND THEN WELL COME BACK.
13463
KATZ - DIRECT
fa
. (RECESS TAKEN.)
THE COURT: GO AHEAD.
JOSEPH LAUREN KATZ.
BEING PREVIOUSLY DULY SWORN, RESUMED THE WITNESS STAND AND
TESTIFIED FURTHER AS FOLLOWS:
[J
DIRECT EXAMINATION (CONT“D)
BY MS. WESTMORELAND:
Nv
00
NN
0
3
+
Ww
NW
N
@. DOCTOR KATZ. IN REGARD TO YOUR ANALYSIS AND THE ANALYSIS OF
1.
oO
PROFESSOR BALDUS, WHEN WE DISCUSSED REGRESSIONS EARLIER, YOU
11 DISCUSSED REGRESSIONS EARLIER, HAVE YOU EXAMINED THE REGRESSIONS
12 USED AND MADE ANY ATTEMPT TO SEE HOW THOSE REGRESSIONS FIT IN
13 WITH THE ANALYSES THAT YOU HAVE DONE?
14 A. YES.
13 QR. AND WHAT HAVE YOU DONE IN THAT REGARD?
16 A. ALTHOUGH PROFESSOR BALDUS HAS GENERALLY DEFINED UNKNOWNS TO
17 BE ZERO, THE REGRESSION ANALYSIS APPEARS TO REFLECT THE ATTEMPT
18 BY THE REGRESSION MODEL TO EXPLAIN ALL THE SIGNIFICANT
w 19 AGGRAVATING AND MITIGATING CIRCUMSTANCES THAT HAVE BEEN POINTED
20 OUT IN MY ANALYSIS,
23 IT SEEMS LIKE WHITE VICTIM CASES HAVE MORE AGGRAVATING
22 CIRCUMSTANCES AND MANY AGGRAVATING CIRCUMSTANCES IN WHICH
23 THEY“RE SIGNIFICANT AT THE .05 LEVEL OF SIGNIFICANCE, AND BLACK
24 VICTIM CASES HAVE MANY MITIGATING CIRCUMSTANCES.
23 ~ IN THE REGRESSION ANALYSIS. I BELIEVE PROFESSOR BALDUS
—
15464
KATZ - DIRECT
1 WAS ATTEMPTING TO CONTROL FOR THESE BACKGROUND VARIABLES.
IT APPEARS THERE ARE MANY BACKGROUND VARIABLES THAT MAY
NEED TO BE CONTROLLED FOR. I RAN AN EXPERIEMENT WHERE I DID A
REGRESSION, WHERE I USED DEATH SENTENCE VARIABLE, GIVEN MURDER
INDICTMENT, AND TRIED TO PREDICT THAT OUTCOME. BASED ON SIMPLY
THE RACE OF VICTIM RECODE VARIABLE.
AND I RAN THAT REGRESSION AND I OBTAINED A POSITIVE
O
w
d
N
CO-EFFICIENT FOR THE RACE OF VICTIM RECODE VARIABLE, AND I
ES BELIEVE THE R-SQUARED FOR THAT REGRESSION WAS SOMEWHERE IN THE
10 NEIGHBORHOOD OF .07. AND I DETERMINED THE SIGNIFICANCE LEVEL
11 FOR THE CO-EFFICIENT FOR THE RACE OF VICTIM RECODE.
12 NOW, THE SIGNIFICANCE LEVEL THAT APPEARS ON THE
13 COMPUTER PRINTOUT IS NOT NECESSARILY THE ACTUAL SIGNIFICANCE
14 LEVEL BECAUSE THE COMPUTER PACKAGES ASK, GIVES "P" VALUES ON
135 ONLY TO FOUR DECIMAL PLACES. THEREFORE IF A "P" VALUE I3
146 POTENTIALLY MUCH SMALLER THAN FOUR DECIMAL PLACES, IT WILL
17 SIMPLY PRINT IT OUT AS .0001, AND GO ON.
18 I LOOKED AT THE “P" VALUE, WHICH IS A STATISTICAL
w 1% MEASURE. AND FROM THAT APPROXIMATED WHAT THE ACTUAL “P" VALUE OF
20 LEVEL OF SIGNIFICANCE FOR THE CO-EFFICIENT OF THE RACE OF VICTIM
21 FOR THAT SIMPLE LINEAR REGRESSION.
22 AND THE CO-EFFICIENT THAT I OBTAINED WAS A "T" VALUE OF
23 A LITTLE MORE THAN EIGHT. AND THAT WOULD APPROXIMATE TO "P*
24 VALUE OF IN THE NEIGHBORHOOD OF POINT NINETEEN ZEROES AND A
23 FIVE. S50 THAT WOULD BE A VERY VERY SMALL "P" VALUE, WHICH WOULD
bp
W
O
N
a
1565
KATZ - DIRECT
MAKE THAT VARIABLE HIGHLY SIGNIFICANT.
@Q. HOW WOULD THAT PARTICULAR "P" VALUE BE PRINTED OUT FROM THE
STATISTICAL PACKAGES USED ON THE COMPUTER?
A. THE "P" VALUE WAS PRINTED OUT WAS SIMPLY THE .0001. I
LOOKED AT THE "T" VALUE WHICH IS A MEASURE, WHICH IS A
STATISTICAL MEASURE, AND FROM THAT APPROXIMATED THE "P" VALUE.
S50, IN TRYING TO RUN A REGRESSION MODEL TO CONTROL FOR
BACKGROUND VARIABLES. IT APPEARS THAT ONE HAS A LONG WAY TO GO
BEFORE THIS LEVEL OF SIGNIFICANCE CAN BE COMPLETELY ELIMINATED.
I NOTICE IN ONE OF PROFESSOR BALDUS” REPORTS IN THE
REGRESSION THAT WAS RUN WITH. I BELIEVE OVER 230 VARIABLES. THAT
THE R-SQUARED HAD RISEN TO. I BELIEVE, IN THE NEIGHBORHOOD OF
.60, AND THE "P" VALLE FOR THE CO-EFFICIENT OF THE RACE OF THE
VICTIM WAS NOW SOMEWHERE IN THE NEIGHBORHOOD OF .01 OR .0Z.
THE COURT: WHAT DOES ALL THAT MEAN, THERE IS A RACE OF
VICTIM EFFELT?
THE WITNESS: MY CONCLUSION IS THAT THERE ARE POSSIBLY
MANY BACKGROUND VARIABLES THAT NEED TO BE INCLUDED IN THE
REGRESSION MODELS TO COMPLETELY EXFLAIN THE RACE OF VICTIM
EFFECT AND WITH AN R-SQUARED OF APPROXIMATELY .460, ONE CAN STILL
ADD ADDITIONAL VARIABLES. HOPEFULLY SELECTIVELY, THAT MIGHT
AFFECT THE MODEL SO THAT THE "P" VALUE FOR THE RACE OF THE
VICTIM CO-EFFICIENT COULD COMPLETELY DISAPPEAR AND BECOME NOT
SIGNIFICANT.
I DO NOT BELIEVE THAT THE TIME CO-EFFICIENT OF THE RACE
1564
KATZ — DIRECT
OF THE VICTIM VARIABLE WILL EVER STOP BEING POSITIVE, BECAUSE
EITHER THE COMPUTER HAS, EXCUSE ME.» THAT IN DETERMINING THE
CO-EFFICIENTS FOR THE REGRESSION. IT AGAIN IS TRYING TO MAKE THE
PREDICTED OUTCOME AS CLOSE AS POSSIBLE TO THE ACTUAL OUTCOME.
AND THE RACE OF THE VICTIM VARIABLE IS A CONVENIENT VARIABLE FOR
THE REGRESSION TO USE, BECAUSE OUT OF THE 140 DEATH SENTENCE
CASES IN THE CHARGING AND SENTENCING STUDY, 119 WERE WHITE
VICTIM CASES.
SO, THE ACTUAL POSITIVE CO-EFFICIENT WILL PROBABLY
REMAIN, REGARDLESS. HOWEVER, THE LEVEL OF SIGNIFICANCE ATTACHED
TO THAT CO-EFFICIENT WILL PROBABLY BE EVENTUALLY ELIMINATED AS
MORE AND MORE VARIABLES ARE INTRODUCED INTO THE MODEL.
THE COURT: I GUESS IT”S YOUR TESTIMONY ON LEVEL OF
SIGNIFICANCE THAT I“M HAVING TROUBLE WITH. I THOUGHT THE
SMALLER THE "P" VALUE WAS THE MORE SIGNIFICANT IT WAS.
THE WITNESS: YES. "P" VALUE FROM THE SIMPLE LINEAR
REGRESSION MODEL THAT I RAN FOR THE RACE OF THE VICTIM WAS VERY
SMALL. POINT NINETEEN ZEROES AND A FIVE, APPROXIMATELY. AND
IT’S INCREASING AS MORE AT MORE VARIABLES ARE INTRODUCED INTO
THE MODEL TO THE POINT WHERE IN THE MODEL WHERE I BELIEVE
PROFESSOR BALDUS RAN OVER 230 VARIABLES, IT’S NOW THE UP TO
APPROXIMATELY .02.
I DO NOT THINK THAT ITLL STAY AT .02 IF MORE AND MORE
VARIABLES ARE ADDED, ALTHOUGH I HAVENT PERFORMED ANY ADDITIONAL
TESTS.
KATZ - DIRECT
i THE COURT: YOU THINK IT WOULD BECOME STATISTICALLY
2 INSIGNIFICANT?
THE WITNESS: YES, ALTHOUGH I DON‘T,. THAT IS BASED ON
THE FACT THAT THE R-SQUARED IS STILL ONLY .40., AND SO THERES
MUCH MORE VARIATION TO BE EXPLAINED AND ONE COULD POSSIBLY FIND
3
4
5
ip 6 ADDITIONAL VARIABLES THAT WOULD ELIMINATE THE RACE OF VICTIM
7 EFFECT.
8 | 1 BELIEVE THE ASSUMPTION THAT“S MADE IN THOSE MODELS IS
9 THAT THE REMAINDER OF THE EFFECT IS SIMPLY RANDOM ERROR THATS
10 RANDOMLY DISPERSED IN TERMS OF WHITE AND BLACK VICTIM CASES, BUT
11 THERE IS NO MORE VARIABLES THAT COULD CONSEIVABLY COULD BE TAKEN
12 OUT AND EXPLAIN PARTIALLY SOME OF THE SIGNIFICANCE OF THE RACE
13 OF THE VICTIM EFFECT.
14 THE COURT: YDU HAVE OBSERVED THAT AS THE NUMBER OF
15 VARIABLES INCREASES THE STATISTICAL SIGNIFICANCE OF THE RACE OF
16 |THE VICTIM EFFECT INCREASES?
17 THE WITNESS: YES.
18 BY MS. WESTMORELAND:
Ww 19 G. DOCTOR KATZ. DID YOU SAY THAT THE STATISTICAL SIGNIFICANCE
20 OF THE RACE OF VICTIM INCREASES?
21 |A. THE "P" VALUE INCREASES.
22 THE COURT: "P" VALUE?
23 THE WITNESS: AND I DON‘T KNOW WHERE ONE WOULD DRAW THE
24 DIVIDING LINE AS TO WHAT THE APPROPRIATE MODEL IS, IN TERMS OF
23 MEASURING PRECISE RACE OF VICTIM EFFECT.
1368
KATZ - DIRECT
BY MS. WESTMORELAND:
@. WHAT IS THE SIGNIFICANCE TO YOU, NOT STATISTICAL
SIGNIFICANCE, BUT JUST THE GENERAL SIGNIFICANCE OF THE FACT THAT
THE "P* VALUE BEGAN AT APPROXIMATELY TWENTY DECIMAL PLACES AND
SUBSEQUENTLY BECOMES ONLY TWO DECIMAL PLACES, AS 1 UNDERSTAND
YOUR TESTIMONY. WHAT IS THE SIGNIFICANCE OF THAT IN RELATION TO
THE REGRESSTION?
A. THAT COULD BE INTERPRETED -—-
MR. FORD: I OBJECT TO THAT, YOUR HONOR. I THINK
COUNSEL SPECIFICALLY PLACED THIS WHERE I THINK IT IS, AND THATS
BEYOND THE REALM OF HIS STATISTICAL EXPERTISE. I BELIEVE
COUNSEL ASKED THE QUESTION AS TO OTHER THAN STATISTICAL
SIGNIFICANCE. SO I THINK IT CALLS FOR A CONCLUSION.
THE COURT: "NO. THAT IS NOT WHAT SHE ASKED.
MR. FORD: THANK YOU.
THE COURT: IF I UNDERSTAND WHAT YOUU ARE BASICALLY
ASKING, IS IT FAIR TO EXTRAPOLATE THE LINE HAVING GONE FROM
POINT. WITH NINETEEN ZEROES AND THEN A 5. TO .02, CAN THAT LINE
BE FAIRLY EXTRAPOLATED AS MORE -- |
MS. WESTMORELAND: AS MORE VARIABLES ARE ADDED.
THE COURT: AS MORE VARIABLES ARE ADDED.
MS. WESTMORELAND: YES, YOUR HONOR.
THE WITNESS: WELL. THAT WOULD DEPEND ON WHICH
VARIABLES WERE ADDED. AND I BELIEVE THE IMPORTANCE OF THE
SIGNIFICANCE FOR THE, THE LAST REMAINING REGRESSIONS WHERE ALL
10
3
“vv
0
~~
°6
o
a
»
13569
KATZ - DIRECT
THE VARIABLES WERE INCLUDED, DEPENDS ON HOW THE REMAINING
VARIATION IS INTERPRETED. IF THERE ARE NO ADDITIONAL VARIABLES
TO BE ADDED SHALL OTHER THAN RANDOM ERROR THATS EVENLY
DISTRIBUTED FOR WHITE-BLACK VICTIM CASES. POSSIBLY THAT WOULD
SHOW SOME EFFECT. BUT THERE ARE MANY MORE VARIABLES THAT CAN BE
DEFINED THAN WHAT PROFESSOR BALDUS HAS DEFINED. IN TERMS OF
INTERACTION VARIABLES AND COMPOSITE VARIABLES THAT ARE
ADDITIONAL VARIABLES THAT CAN BE UTILIZED.
WHETHER THEY WOULD HAVE ANY EFFECT OR NOT, I WOULD NOT
KNOW. BUT I DON’T THINK ONE COULD SAY THAT ONE HAS EVER IN
REGRESSION ANALYSIS ALL THE VARIABLES AND HAS THE MODEL THAT
EXFLAINS THE SENTENCING OUTCOMES.
MS. WESTMORELAND: THAT’S ALL THE QUESTION I HAVE, YOUR
HONOR.
BY THE COURT:
@. LET ME ASK YOU ONE QUESTION BEFORE MR. FORD BEGINS.
DOCTOR WOODWORTH WAS TESTIFYING THE QTHER DAY THAT
ALTHOUGH THE R-SRUARED FACTOR WAS RELATIVELY LOW, THE MODEL. AND
I DON’T KNOW WHICH MODEL HE WAS TALKING ABOUT, PREDICTED THE
DEATH PENALTY IN 93 PERCENT OF THE CASES.
DO YOU HAVE ANY COMMENT ON THAT STATEMENT. IN TERMS OF
VALIDATING, IN TERMS OF WHETHER OR NOT IT ACTUALLY VALIDATES THE
MODEL AS BEING SELECTING ONLY THE DEATH PENALTY CASES OR ANY
OTHER KIND?
A. WELL, THE PREDICTED OUTCOMES THAT APPEAR GENERALLY ARE NOT
o
n
.
~
Oo
p
o
n
1570
KATZ - DIRECT
EITHER ZERO OR 1. THEY’RE USUALLY SOMEWHERE IN BETWEEN. LIKE .3
OR .78. AND THE, THE ATTEMPT IS TO TRY AND MAKE THE PREDICTED
OUTCOMES AS CLOSE AS POSSIBLE TO THE ACTUAL OUTCOMES.
I AM NOT SURE 1 UNDERSTAND WHAT IS MEANT BY THE
STATEMENT THAT 93 PERCENT OF THE DEATH SENTENCE OUTCOMES HAVE
BEEN PREDICTED, BECAUSE I DON’T BELIEVE HIS MODEL HAD ONES FOR
93 PERCENT OF THE DEATH SENTENCE CASES.
@. YOU HEARD THE TESTIMONY IM REFERRING TO?
A. YES.
@. CERTAINLY. THAT TESTIMONY MIGHT BE INTERPRETED -— WOULD BE
THAT IF YOU HAD A MODEL WHERE YOU CONTROLLED ONLY FOR MURDER
INDICTMENT. IT WOULD PREDICT 93 PERCENT OF ALL THE DEATH
PENALTIES IMPOSED BUT IT WOULD NOT EXCLUDE ANY OTHER, I DON’T
KNOW WHETHER THATS WHAT HE IS TALKING ABOUT, BUT I“M TRYING TO
FIND OUT IF YOU‘RE FAMILIAR WITH THE MEASURE HE WAS USING?
A. NO. IM NOT.
THE COURT: ALL RIGHT. MR. FORD. YOU MAY PROCEED.
MR. FORD: FOR THE RECORD, COULD I RENEW MY OBJECTION ON THE
BASIS, NOW AT THE CLOSE OF THE TESTIMONY, IT“S NOT BEEN TIED UP
WITH ANY CONTROLS OR ANY CONCLUSIONS SUBMITTED IN THIS CASE WITH
REGARD TO THE NUMEROUS TABLES WE HAD OF RELATIVE AGGRAVATION
LEVELS, IF I COULD RENEW MY MOTION?
THE COURT: ARE YOU TALKING ABOUT THESE TABLES --
MR. FORD: YES.
THE COURT: —-— ADMITTED LATELY? I VIEW THAT EVIDENCE
B
l
E
R
Y
BE
S
e
f
S
E
ee
E
L
ST
O
E
»
oO
[Wr
y
nN
1371
KATZ — CROSS
AS PRIMARILY A COMMENT UPON THE POSSIBILITY OF BIAS, AND NOT
DEMONSTRATING THE ABSENCE OR PRESENCE OF THE RACIAL EFFECT SO WE
DON‘T HAVE CONTROL. THEREFORE. I WILL LEAVE IT ADMITTED.
MR. FORD: THANK YOU, YOUR HONOR.
THE COURT: LET ME MAKE SURE IM RIGHT ABOUT THAT.
DOCTOR KATZ. 1 UNDERSTAND BASICALLY ALL THESE TABLES
YOU’VE BEEN GOING THROUGH SUGGEST THAT BECAUSE OF THE OBSERVED
PHENOMENA IN THAT A BLACK VICTIM’S CASE TENDING TO BE
MITIGATIVE. AND WHITE VICTIM CASES TENDING TO BE AGGRAVATED.
THAT MIGHT SKEW ANY REGRESSION WHICH TRIES TO FIT THE DATA. IS
THAT ABOUT WHAT YOURE SAYING?
THE WITNESS: YES, YOUR HONOR.
THE COURT: OKAY.
CROSS-EXAMINATION
BY MR. FORD:
G. DOCTOR KATZ, I UNDERSTAND FROM YOUR TESTIMONY YOU-VE DONE A
NUMBER OF ANALYSES ON THESE DATA THAT WERE PROVIDED TO YOU BY
PROFESSOR BALDUS AND PROFESSOR WOODWORTH, IS THAT RIGHT?
A. THAT’S CORRECT.
@. AND THATS BEEN ON A COMPUTER THAT YQU’VE BEEN USING
YOURSELF. OPERATING YOURSELF?
A. YES.
@. AND HAVE YOU PRINTED OUT YOUR RESULTS FROM ALL THOSE
OPERATIONS AND ANALYSES THAT YOU‘VE DONE?
A. YES, I HAVE.
-
vw
0
N
c
a
ep
W
N
1572
KATZ ~ CROS:
@. AND ARE THEY ALL INCLUDED IN THE EXHIBITS BEFORE THE COURT?
A. NO, THEY’RE NOT.
@. THE, ARE THEY ALL INCLUDED IN A REPORT PROVIDED?
MS. WESTMORELAND: YOUR HONOR, ILL OBJECT TO THE
REPORT. IT’S NOT A PART OF THE RECORD IN THE CASE, PREPARED
ONLY FOR COUNSEL FOR DISCOVERY PURPOSES. AND HAS NO RELEVANCY TO
THIS HEARING.
THE COURT: WELL, IF MR. FORD”S TRYING TO FIND OUT IF
HE FOUND OUT ANYTHING THAT‘LL HELP MR. FORD®S CASE, THEN I“LL
LET HIM ASK. I PRESUME THAT’S WHAT IT IS HE’S ASKING IT FOR.
MR. FORD: THAT’S CERTAINLY ONE PURPOSE. YOUR HONOR.
THE WITNESS: COULD YOU PLEASE REPEAT THE QUESTION?
BY MR. FORD:
@. HAVE YOU DONE ANY ANALYSES AND COMPUTER PRINTOUTS ON THIS
DATA THAT ARE NOT INCLUDED IN THE REPORT YOU PROVIDED TO MS.
WESTMORELAND?
A. YES.
@. HOW MANY? CAN YOU ESTIMATE?
A. RIGHT NOW, 1 HAVE SOMETHING THAT APPEARS TO LOOK LIKE AN
OFFICE FULL OF COMPUTER PRINTOUTS ON VARIOUS ANALYSES THAT IVE
DONE SINCE I STARTED THIS COMPUTER ASPECT IN EARLY FEBRUARY.
@. AND THOSE THAT ARE BEFORE THE COURT NOW AND THOSE THAT WERE
INCLUDED IN YOUR REPORTS WERE SELECTED FROM AMONG THOSE MUCH
LARGER NUMBER, IS THAT FAIR TO SAY?
A. YES.
1373
KATZ - CROSS
@. DO YOU KNOW IF THOSE, ALL THE REST OF THOSE COMPUTER
PRINTOUTS HAVE BEEN MADE AVAILABLE TO COUNSEL THROUGH THE
DISCOVERY PROCESS IN THIS CASE?
A. THEY WERE ALL MADE AVAILABLE AS OF THE TIME MY DEPOSITION
WAS TAKEN.
@. AND HAVE YOU DONE ANY SINCE THEN?
A. YES.
@. ABOUT HOW MANY? CAN YOU ESTIMATE?
A. ARE YOU ASKING FOR THE NUMBER OF COMPUTER PROGRAMS THAT I“VE
RUN SINCE THEN?
QR. YES?
THE COURT: I THINK WHAT HE WANTS TO KNOW ABOUT IS THE
NUMBER OF EXPERIMENTS. THATS REALLY WHAT YOU ARE AIMING AT.
MR. FORD: THAT’S CORRECT. YOUR HONOR. THE NUMBER. I
GUESS, OF ANALYSES IS THE WORD I WOULD USE.
THE WITNESS: I DON’T PARTICULARLY RECALL AN EXACT
NUMBER, BUT I BELIEVE I“VE DONE IN THE TENS OF EXPERIMENTS.
BY MR. FORD:
R. BY IN THE TENS, YOU MEAN LESS THAN A HUNDRED?
A. I’M SURE I HAVE MORE THAN A HUNDRED COMPUTER PRINTOUTS. BUT
1 BELIEVE THE KINDS OF EXPERIMENTS THAT I HAVE DONE CAN BE
CLASSIFIED IN THE TENS.
THE COURT: TEN OR LESS?
THE WITNESS: TEN OR TWENTY, YOUR HONOR.
BY MR. FORD:
1374
KATZ - CROSS
@. DID YOU ALSO DO SOME ANALYSES OR MANIPULATIONS OF THE DATA
ON THE COMPUTER SCREEN THAT YOU DIDN‘T PRINT OUT?
A. YES.
Gl. MANY?
A. YES.
@. AND HOW DID YOU DECIDE WHICH ONES TO PRINT OUT AND WHICH
ONES NOT TO?
A. EITHER I COMMITTED A COMPUTER ERROR. OR I PRINTED OUT A
PRINTOUT THAT WAS MORE COMPLETE THAN AN EARLIER PRINTOUT.
@. DID YOU DECIDE NOT TO PRINT OUT ANY ANALYSES BECAUSE THEY
DID NOT SUPPORT YOUR CONCLUSIONS OR YOUR TESTIMONY IN THIS CASE?
A. NO.
@. IS THERE ANY WAY TO RECOVER THOSE THAT YOU DIDN-T PRINT ouT
OR ARE THEY JUST GONE. ERASED?
A. I HAVE A COMPUTER FILE THAT HAS APPROXIMATELY A HUNDRED
PROGRAMS. I BELIEVE I HAVE TWO FILES. PART OF THE DIFFICULTY
THAT I RAN INTO WAS THAT AS I WAS GETTING UPDATES AND TAPES AND
THINGS, 1 HAD TO REVISE CERTAIN PROGRAMS TO ACCOMMODATE THESE
TAPES, AND I DID NOT SPECIFICALLY KEEP THESE PROGRAMS AS
COMPUTER PROGRAMS. 1 REVISED CERTAIN COMPUTER PROGRAMS.
HOWEVER, I DO BELIEVE I HAVE THE ORIGINAL PRINTOUT
SOMEWHERE IN MY OFFICE TO THE PREVIOUS COMPUTER PROGRAM.
@. DO I UNDERSTAND THERE‘S SOMEWHERE IN THE TENS OF PRINTOUTS
THAT HAVE NOT, THAT WERE DONE AFTER THE DEPOSITION, THAT HAVE
NOT BEEN PROVIDED TO COUNSEL TO YOUR KNOWLEDGE?
a
d
e
l
H
R
C
EA
R
T
R
E
E
E
10
11
1375
KATZ - CROSS
A. THAT’S NOT CORRECT.
®@. I MISUNDERSTOOD. I“M SORRY.
A. THERE ARE TENS OF DIFFERENT EXPERIMENTS. SOME EXPERIMENTS
HAD MANY MANY PRINTOUTS.
@. THAT HAVE BEEN DONE SINCE THE DEPOSITION AND NOT PROVIDED TO
COUNSEL, IS THAT RIGHT?
A. YES.
@. DOCTOR KATZ, LET ME GO AT THE QUESTION OF HOW, CAN YOU
ESTIMATE HOW MUCH TIME YOU SPENT ON THIS CASE, PREPARING FOR
YOUR TESTIMONY HERE?
A. SINCE I RECEIVED THE DATA IN LATE JANUARY, LITERALLY
CONSTANTLY FROM EARLY FEBRUARY UP TO THIS POINT.
THE COURT: YOU MEAN 40 HOURS A WEEK?
THE WITNESS: MORE THAN 40 HOURS A WEEK. THAT WOULD
INCLUDE WEEKENDS.
I DID HAVE TEACHING RESPONSIBILITIES DURING THAT TIME.
BUT I HAVE NOT BEEN ABLE TO ENGAGE IN ANY OTHER RESEARCH
ACTIVITY DURING THIS TIME. THIS TIME HAS BEEN OCCUPIED
ANALYZING THIS DATA SINCE EARLY FEBRUARY.
BY MR. FORD:
@. IF MY SIMPLE ARITHMETIC IS CORRECT. IT”S ABOUT SIX MONTHS.
IS THAT WHAT WE“RE TALKING ABOUT, SIX MONTHS?
A. SIX AND A HALF MONTHS. I DID SPEND SOME TIME NOT IN THE
ANALYSIS, BUT IN THE INVESTIGATION PRIOR TQ THE TIME I RECEIVED
IT.
10
1576
KATZ - CROSS
Q. SOMETHING OVER 40 HOURS A WEEK, THEN. FOR 26 WEEKS. OVER A
THOUSAND HOURS, IS THAT FAIR?
A. AT LEAST A THOUSAND HOURS, YES.
G4. AND DO YOU HAVE ANY IDEA OF HOW MUCH COMPUTER TIME YOU
UTILIZED IN THE ANALYSES?
A. 1 BELIEVE FOR ONE MONTH, I BELIEVE THAT WAS MARCH, I BELIEVE
THE BILL CAME TO SOMETHING LIKE TWELVE THOUSAND DOLLARS FOR THAT
MONTH. | | =
| BUT I HAVE DONE MANY MORE THAN THAT. SO I DO NDT KNOW
WHAT THE TOTAL COST HAS BEEN FOR THE COMPUTER TIME.
Q. THIS COMPUTER TIME I TAKE IT IS ON A STATE COMPUTER. IS THAT
RIGHT?
A. YES.
Q. AND IS YOUR TIME COMPENSATED HOURLY OR ON A FLAT FEE OR HOW
DOES THAT WORK?
A. I’M COMPENSATED HOURLY.
@. WHATS YOUR HOURLY RATE ON THAT?
A. $37.50 AN HOUR.
4. SO YOUVE BEEN PAID SOMEWHERE IN THE AREA OF FORTY THOUSAND
DOLLARS PLUS IN THIS CASE AT THIS POINT?
A. 1 HAVE NOT BILLED THE STATE ATTORNEY GENERAL‘S OFFICE FOR
ALL THE TIME I“VE PUT IN. I“VE BEEN VERY CONSERVATIVE IN MY
BILLING.
Q. DO YOU HAVE ANY IDEA HOW MUCH YOU HAVE BEEN PAID TD DATE?
A. I THINK IVE BEEN PAID IN THE NEIGHBORHOOD OF THIRTEEN,
1577
KATZ - CROSS
1 FOURTEEN THOUSAND DOLLARS AT THIS POINT.
@. DO YOU HAVE A BILL CUTSTANDING IN ADDITION?
A. 1 HAVE MANY HOURS OUTSTANDING WHICH I WILL BILL THE STATE
2
3
4 ATTORNEY GENERALS OFFICE FOR.
S @. WAS THERE A CUTOFF DATE ON THE THIRTEEN, FOURTEEN THOUSAND
6 DOLLARS TIME?
7 A. 1 BELIEVE THAT TIME WAS FOR THE MIDDLE OF JULY. UP TO THE
8 MIDDLE OF JULY.
? @. NOW, HAVE YOU, YOU“VE NEVER BEEN HIRED AS A CONSULTANT,
10 EITHER AS AN EXPERT WITNESS IN COURT OR IN ANY OTHER FASHION
11 SIMILAR TO THIS BEFORE, IS THAT A CORRECT UNDERSTANDING OF YOUR
12 TESTIMONY?
13 A. AS A DIRECT CONSULTANT. NOQ.
14 @. AND AS AN INDIRECT CONSULTANT. IS THERE AN INDIRECT
15 CONSULTANCY YOU“VE BEEN INVOLVED IN ON A PAID BASIS?
146 A. IVE WORKED UNDER PROFESSOR BURFORD BEFORE.
17 @. AND THAT WAS WITH REGARD TO A STUDY ON GREYHOUND BUS LINES?
18 A. YES.
1? Q. AND THAT WAS WHERE YOU INDICATED YOU HAD DONE SOME EMPIRICAL
20 RESEARCH, IS THAT RIGHT?
21 A. YES. |
22 ®. AND THAT WAS LIBRARY RESEARCH FROM THE CENSUS DATA?
23 A. YES. PARTLY. |
24 R. THEN YOU ALSO DID REGRESSION ANALYSIS WITH THAT DATA. IS
23 THAT RIGHT?
1573
KATZ - CROSS
1 A. YES.
2 @. NOW, WITH REGARD TO THIS CASE, YOU‘VE ALSO PREVIOUSLY GIVEN
3 A DEPOSITION IN THIS CASE, IS THAT RIGHT?
4 A. YES, I HAVE.
=) @. DID YOU HAVE THE OPPORTUNITY TO REVIEW THAT DEPOSITION AFTER
a & YOU GAVE IT AND AFTER IT WAS TYPED UP BY THE COURT REPORTER?
7 A. YES.
8 @. AND DID YOU MAKE CORRECTIONS WHERE THERE WERE ERRORS IN THAT
S TYPING?
10 A. TO THOSE ERRORS THAT I FOUND. YES.
11 @. AND DID YOU SIGN THAT DEPOSITION?
12 A. YES.
13 Q. WAS IT CORRECT TO THE BEST OF YOUR KNOWLEDGE?
14 A. TO THE BEST OF MY KNOWLEDGE. YES.
13 @. HAVE ANY OTHER CHANGES OCCURRED TO YOU SINCE YOU SIGNED IT?
14 A. NO.
17 Q. HAVE YOU WRITTEN ANYTHING ELSE IN THIS CASE BESIDES YOUR
18 REPORT? 1 BELIEVE YOUR REPORT WENT THROUGH ONE DRAFT. THAT WAS
@ i? AT LEAST, OR THAT -—— \
20 A. MY REPORT WENT THROUGH SEVERAL DRAFTS.
23 R. AND ASIDE FROM THAT REPORT AND ITS SEVERAL DRAFTS HAVE YOU
22 WRITTEN ANYTHING ELSE REGARDING YOUR OPINIONS AND FINDINGS IN
23 THIS CASE?
24 A. NO.
23 2. SHOWING YOU A DOCUMENT ENTITLED "RESPONDENTS THIRD
—— — ——— ——— — — —— . r—
137%
KATZ - CROSS
Jo
o SUPPLEMENTAL REPONSE TO INTERROGATORIES AND SUBMISSIONS." ET
2 |CETERA, MARKED JK NUMEER 3.
3 DO YOU RECOGNIZE THAT?
4 |A. I BELIEVE THIS IS THE FIRST TIME I‘VE EVER SEEN THIS
Ss |DOCUMENT.
il & |@. DID YOU WRITE THAT DOCUMENT IN TOTAL OR IN PART?
7a. NO.
8 |@ DOCTOR KATZ, DO YOU STILL HAVE YOUR COPY OF THE EXHIBITS
9 |THAT YOU TESTIFIED TO THIS MORNING —-
10 ph. YES. 1 DO.
11 @. -- BEFORE YOU?
12 CALLING YOUR ATTENTION TO, I BELIEVE. RESPONDENT’S
13 EXHIBIT NUMBER 17A. COUNTS OF THE NUMBER OF UNKNOWNS FOR
14 VARIABLES IN THE PROCEDURAL REFORM STUDY. HAVE I GOT THE RIGHT
13 DOCUMENT?
16 A. YES.
37 @. NOW, AM I CORRECT THAT IN COMPILING THIS, YOU GAVE THE
13 COMPUTER A COMMAND TO PRINT OUT ALL THE LINE ITEMS OR ALL THE
® 19 VARIABLE NUMBERS THAT WERE CODED IN THE COMPUTER AS "U", IS THAT
20 RIGHT?
21 A. NO.
22 R. CAN YOU EXPLAIN TO ME WHAT YOU DID DO TO GET THIS COUNT?
23 A. THERE ARE DIFFERENT CODES THAT WERE UTILIZED TO REPRESENT
24 UNKNOWNS, FOR THE DIFFERENT QUESTIONS AND I PRINTED OUT EACH
23 VARIABLE, A LIST OF THE ITEMS, AND A COUNT OF THE NUMBER OF
1380
KATZ - CROSS
ITEMS REPRESENTED BY EACH. EACH POSSIBLE CODE.
THEN I PICKED THOSE CODES WHICH WERE INDICATED TO BE
UNKNOWNS, AND COUNTED THOSE, INCLUDING THOSE IN WHICH THERE WAS
NO INDICATION AS TO WHAT THAT CODE REPRESENTED. AN
IMPERMISSIBLE CODE.
@. OKAY. THIS INCLUDES BOTH IMPERMISSIBLE CODES AND THINGS
THAT THE COMPUTER PRINTED OUT AS "U’S", IS THAT RIGHT?
A. NO.
R. WHAT DOES THE COMPUTER PRINT OUT ON THAT RUN?
A. THE COMPUTER PRINTS OUT VARIOUS DIFFERENT CODES TO REPRESENT
UNKNOWNS, DEPENDING ON WHICH CLASSIFICATION OF VARIABLE. THERE
ARE, FOR THE "W" VARIABLES, THE UNKNOWNS WERE GENERALLY
REPRESENTED WITH ZEROES AS ONE OF THE CODES FOR THE FOIL
QUESTIONS.
THE "US." THERE MAY HAVE BEEN SOME "U“3" THAT CREPT IN
THERE, TOO, BUT THE "U’/S" ARE FOR THE CHARGING AND SENTENCING
STUDY.
@. WHEN YOU DID THE SAME THING WITH REGARD TO THE CHARGING AND
SENTENCING STUDY, YOU DID FOLLOW THE PROCEDURE I JUST OUTLINED
OF GIVING A COMMAND TO PRINT OUT COUNTS OF ALL THE "U’S" FOR
EACH VARIABLE, IS THAT RIGHT?
A. THAT’S NOT PRECISELY RIGHT. BECAUSE THERE ARE SOME VARIABLES
IN WHICH THE UNKNOWNS AREN‘T REPRESENTED BY "U-S".
RA. NOW, I THINK YOU INDICATED THAT. IT’S TRUE, ISNT IT.
PERHAPS I CAN JUST ASK IT THIS WAY. THAT DIFFERENT CODING WITH
1381
KATZ - CROSS
REGARD TO UNKNOWN INFORMATION WAS FOLLOWED IN THE FILLING QUT OF
WHAT WE“VE CALLED THE QUESTIONNAIRES, THE CODING INSTRUMENT. ON
THE CHARGING AND SENTENCING STUDY AND THE QUESTIONNAIRE ON THE
PROCEDURAL REFORM STUDY, ISN“T THAT RIGHT?
A. YES.
Q. NOW, WITH REGARD TO THE CHARGING AND SENTENCING STUDY, I
BELIEVE YOU TESTIFIED YOU WERE FAMILIAR WITH THE FORMAT OF THE
QUESTIONNAIRE OR THE CODING INSTRUMENT ITSELF, IS THAT RIGHT?
A. YES.
@. AND DO YOU RECALL THEN THAT THERE WAS A SPECIFIC DEFINITION
FOR CERTAIN VARIABLES WITH REGARD TO WHAT "U" DESIGNATED?
A. YES.
@. AND DO YOU RECALL THAT WAS THAT A "1" COULD DESIGNATE
EXPRESSLY STATED IN THE FILE, "2" COULD DESIGNATE SUGGESTED BY
THE FILE BUT NOT SPECIFICALLY INDICATED; "BLANK" WOULD BE
INCONSISTENT WITH INFORMATION IN THE FILE: AND "U" WOULD THEN BE
SIMPLY UNABLE TO CLASSIFY IN ANY OF THE OTHER CATEGORIES. IS
THAT RIGHT?
A. YES.
A. 1 TAKE 1T THEN BOTH ON THE COMPUTER TAPE AS ON THE
QUESTIONNAIRE IN OTHER VARIABLES, SUCH AS DATES. PERHAPS, THAT
MIGHT BE USED THAT WOULD BE CODED SLIGHTLY DIFFERENT TO MEAN
SOMETHING MORE LIKE THE WORD UNKNOWN, RATHER THAN UNABLE TO
CLASSIFY IN THREE OTHER DEFINED CATEGORIES. IS THAT RIGHT?
A. I WOULD NOT KNOW. I HAVEN’T ANALYZED THE VARIABLES THAT
KATZ - CROSS
1 INCLUDE DATES. THAT I IMAGINE COULD BE A POSSIBILITY.
2 @. THERE WAS SOME VARIATION. AND SPECIFIED VARIATION WITH
3 REGARD TO THE CODING INSTRUMENT IN TERMS OF THE EXACT MEANING
THAT CIRCUMSTANCES UNDER WHICH A "U" WOULD BE CODED, IS THAT
RIGHT? :
A. YOU MEAN IT WAS DEFINED AS UNABLE TO CLASSIFY?
@. IN SOME CIRCUMSTANCES, AND IN SOME CIRCUMSTANCES IT WAS
DEFINED AS UNKNOWN. NO INFORMATION?
a
0H
N
o
a
p
A. BELIEVE SO, YES.
10 |(@. YOU DIDN‘T DISTINGUISH IN YOUR COUNTS ALONG THOSE LINES,
11 THOUGH. IS THAT RIGHT?
12 la. ND. I DID NOT.
13 |@. NOW. DOCTOR KATZ, YOU TESTIFIED. I BELIEVE, THAT IT’S
14 |STANDARD PROCEDURE TO ELIMINATE VARIABLES WITH UNKNOWNS IN
15 |STATISTICS, IS THAT CORRECT?
16 |A. 1 DO NOT BELIEVE I SAID THAT.
17 |@. THERE ARE VARIOUS TECHNIQUES FOR DEALING WITH UNKNOWN
18 INFORMATION IN STATISTICAL. AMONG STATISTICIANS, ARE THERE NOT?
@ 19 |A. YES.
20 |@. THERE ARE IMPUTATION TECHNIQUES. IS THAT RIGHT?
21 A. TECHNIQUES SUCH AS THAT HAVE BEEN SUGGESTED.
22 |@. LET ME ASK YOU THIS. DO YOU READ, YOU INDICATED YOU READ
23 |THE AMERICAN STATISTICIAN, YOU REMEMBER, THE MAGAZINE OF THE
24 |AMERICAN STATISTICAL ASSOCIATION?
23 A. 1 AM A MEMBER OF THE AMERICAN STATISTICAL ASSOCIATION, YES.
13583
KATZ - CROSS
f
o
@. AND YOU READ THE AMERICAN STATISTICIAN, IS THAT WHAT I
< UNDERSTOOD YOUR TESTIMONY TO BE?
3 A. I DID UP UNTIL FEBRUARY WHEN I GOT INVOLVED IN THIS PROJECT.
“4 I HAVEN’T BEEN KEEPING CURRENT IN ANY OF THOSE AREAS.
3 @. BUT THAT IS THE. WHEN YOU REFERRED TO READING THE MAGAZINE
A & OF THE ASSOCIATION, THAT’S WHAT YOU“RE REFERRING TO, IS THAT
7 RIGHT? |
8 A. 1 DON’T READ IT FROM COVER TO COVER. I READ ARTICLES OF
9 INTEREST.
10 @. THERE‘S WHAT YOU MEANT THAT YOU DID READ PARTS OF, IS THAT
: RIGHT?
12 A. YES.
13 R. THERE’S ALSO A JOURNAL OF OF THE AMERICAL STATISTICAL
14 ASSOCIATION THAT CONTAINS TECHNICAL ARTICLES. ISN‘T THAT RIGHT?
15 A. YES.
16 @. YOU CONSIDER THAT TO BE AN AUTHORITATIVE JOURNAL IN THE
17 STATISTICAL PROFESSION?
18 A. YES.
w 19 @. HAVE YOU READ ANY OF THE ARTICLES THAT HAVE APPEARED IN THE
20 LAST COUPLE OF YEARS ABOUT SURVEYING CERTAIN IMPUTATION
21 TECHNIQUES THAT ARE AVAILABLE WHEN THE INFORMATION IS UNKNOWN?
22 A. YES.
23 @. SO YOU KNOW THERE ARE A NUMBER OF DIFFERENT PRACTICES THAT
24 ARE CONSIDERED TO HAVE MORE OR LESS MERIT IN DIFFERENT
23 CIRCUMSTANCES, IS THAT RIGHT?
1384
KATZ - CROSS
A. THE GENERAL CONCLUSIONS THAT I HAVE FOUND FOR THOSE ARTICLES
po
se
5 | 1S PROCEED WITH CAUTION IN USING ANY OF THESE TECHNIQUES.
2 |@. NOW» WHEN WERE TALKING ABOUT UNKNOWN INFORMATION IN THAT
4 |CONTEXT. LET‘S SAY, USE FOR EXAMPLE. YOUR STUDY OF GREYHOUND, FOR
Ss |GREYHOUND.
i] & I BELIEVE IN THAT YOU GOT SOME POPULATION
7 |CHARACTERISTICS OF DIFFERENT POPULATIONS?
2 iA. vES.
5 |@. AND STATISTICS WITH REGARD TO TRANSPORTATION MEANS IN
10 |DIFFERENT SOCIO-ECONOMIC GROUPS, IS THAT —-
11 |A. I RECOLLECT SOMETHING LIKE THAT.
12 |@. =-- GENERALLY ACCURATE?
13 WHEN YOU“RE TALKING ABOUT UNKNOWNS THERE, PERHAPS YOU
14 MIGHT BE TALKING ABOUT YOU WERE UNABLE TO GET INFORMATION ON ONE
15 CERTAIN BUS ROUTE OR UNABLE TO GET INFORMATION FOR A YEAR OR
14 JUST SIMPLY NOTHING ABOUT THAT, IS THAT RIGHT. IS THAT THE KIND
17 OF UNKNOWN WE“RE TALKING ABOUT?
18 A. 1 DON’T BELIEVE I HAD AN UNKNOWN PROBLEM FOR THAT PARTICULAR
ho 19 STUDY, BECAUSE GREYHOUND APPLIED THE INFORMATION IN TERMS OF THE
20 TOTAL SALES, AND THE CENSUS INFORMATION WAS FAIRLY COMPLETE. SO
23 I DON’T BELIEVE THAT ARDSE AS A PROBLEM IN THAT STUDY.
22 @. SO YOUR TESTIMONY IS THAT CENSUS INFORMATION DOES NOT
23 CONTAIN UNKNOWN INFORMATION?
24 A. NO. MY TESTIMONY IS THAT I BELIEVE THE INFORMATION I USED
23 WAS SUCH INFORMATION IN WHICH THERE WERE NOT ANY UNKNOWNS, BUT I
Stiinta, | eit re—— et—— —— —————————— —— —— ety. See a ——— — — ; ——— —— — —— ———
13585
KATZ - CROSS
CAN‘T RECALL AT THIS POINT. fy
2 Q. ARE YOU FAMILIAR WITH THE IMPUTATION TECHNIQUES USED BY THE
3 UNITED STATES CENSUS IN DEALING WITH UNKNOWN INFORMATION
4 THAT IT GATHERS IN ITS SURVEYS?
5 A. I’M FAMILIAR THAT THEY DO UPDATE THEIR CENSUS INFORMATION,
iA 6 USING IMPUTATION TYPE OF TECHNIQUES. YES.
7 @. AND THAT’S METHODS OF FILLING IN UNKNOWNS THAT THEY GATHER
3° |OR INFORMATION THAT THEY’RE UNABLE TO GATHER FROM THEIR SURVEY
4 TECHNIQUES. IS THAT RIGHT?
10 A. SINCE THE CENSUS IS ONLY DONE ONCE EVERY TEN YEARS, THEY TRY
4 TO UPDATE THE INFORMATION AS BEST THEY CAN.
12 THE COURT: ARE WE TALKING ABOUT APPLES AND ORANGES? I
13 UNDERSTAND THE CENSUS HAS GOT TO MAKE SOME IMPUTATIONS OF THINGS
14 THAT ARE UNKNOWN, FORECASTING, PROJECTIONS OR WHAT HAVE YOU.
135 BUT I UNDERSTOOD THE ISSUE IN THE CASE TO BE THE PROPRIETY OF
16 THE DECISION TO CLASSIFY A "U" AS A NEGATIVE RESPONSE AND
17 LEAVING IT IN THE UNIVERSE OR SAMPLE AS OPPOSED TQ THROWING IT
13 OUT.
w 19 I DONT KNOW THAT THE QUESTION YOU-VE ASKED TOUCHES ON
20 SUCH A PRACTICE. IT MAY, BUT I WANT TO MAKE SURE WE’RE TALKING
5 ABOUT THE SAME PRACTICE.
22 MR. FORD: I DO AS WELL, YOUR HONOR, AND I HAVE
23 QUESTIONS ALONG THAT LINE, AS WELL.
24 BY MR. FORD:
25 RQ. LET’S TURN TO THAT. DOCTOR KATZ.
S
a
d
+
S
R
E
G
E
(
I
1
RE
|
SE
B
E
10
11
1326
KATZ - CROSS
YOU“VE HEARD THE TESTIMONY FROM DOCTOR BALDUS IN THIS
CASE. IS THAT RIGHT?
A. IVE HEARD THE TESTIMONY OF PROFESSOR BALDUS.
@. AND YOU HAVE HEARD PROFESSOR BALDUS TESTIFY THAT WHAT HE’S
ATTEMPTING TO MODEL IS THE INFORMATION AVAILABLE TQ THE DECISION
MAKERS IN THE PROCESS THAT RESULTS IN DIFFERENTIAL SENTENCING IN
THIS STATE. IS THAT FAIR SUMMARY?
A. I BELIEVE THAT’S WHAT HE’S TRYING TQ DO. YES.
@. AND IS THE QUESTION OF WHETHER OR NOT THAT ISSUE OF THE
INFORMATION AVAILABLE TO THOSE PEOPLE RATHER THAN THE ACTUAL
FACTS OF THOSE CASES. IS THAT WITHIN YOUR REALM OF EXFERTISE?
A. I DON’T THINK I UNDERSTAND WHAT THE QUESTION IS.
@. DO YOU HAVE ANY EXPERT TRAINING OR ANY TRAINING AT ALL BY
WHICH TO EVALUATE WHETHER THE CORRECT QUESTION IN APPROACHING
THE UNKNOWN PROBLEM IN THIS CASE IS WHAT WAS KNOWN TO THE
DECISION MAKERS, WHAT ACTUALLY OCCURRED OUT IN THE REAL WORLD,
OR WHAT APPEARED IN THE PAROLE BOARD FILE?
IS ANYTHING IN YOUR TRAINING THAT QUALIFIES YOU TO
SPEAK TO THE QUESTION OF WHETHER THE CORRECT ISSUE IN THIS
ANALYSIS IS ONE OF THOSE THREE?
A. MY TRAINING SUGGESTS THAT ONE SHOULD GO TO THE BEST DATA
SQURCES THAT ARE AVAILABLE.
@. DO YOU KNOW ANYTHING, DO YOU HAVE ANY EXPERIENCE WITH REGARD
TQ THE RELATIVE QUALITY OF DATA SOURCES IN A CRIMINAL JUSTICE
SYSTEM IN THE STATE OF GEORGIA OR ANYWHERE ELSE?
YT
EU
NE
T
|
G
E
N
L
l
NE
SE
SU
E
A
RE
L
R
-
oO
11
z2
1587
KATZ ~- CROSS
A. NO.
MS. WESTMORELAND: YOUR HONOR, I DON‘T BELIEVE DOCTOR
KATZ HAS TESTIFIED CONCERNING THE APPROPRIATE DATA SOURCES TO BE
UTILIZED OR ANYTHING OF THE SORT. I DON’T SEE THE FURPOSE OF
THIS CROSS-EXAMINATION IN THAT REGARD.
THE COURT: IS THAT AN OBJECTION OR AN OBSERVATION?
MS. WESTMORELAND: YOUR HONOR, ITS AN OBJECTION IF
WERE GOING TO CONTINUE ALONG THIS LINE OF QUESTIONING.
THE COURT: I UNDERSTAND WHAT MR. FORD IS GETTING AT.
ALTHOUGH HIS LAST QUESTION MAY NOT HAVE BEEN AS ARTFULLY PUT AS
SOME OF THE OTHERS HAVE BEEN.
WHAT HE WANTS TO KNOW IS IF DOCTOR KATZ HAS HAD ANY
PRACTICAL EXPERIENCE WITH THE DATA SOURCE THAT PROFESSOR BALDUS
IS USING, SO AS TO HAVE AN OPINION BEYOND THAT OF A STATISTICIAN
AS TO THE RATIONALITY OF CODING SOMETHING AS A "U" OR NOT. AND
ITS FAIRLY OBVIOUS HE HASN’T. I“M NOT SURE I“VE HEARD ANYTHING
ABOUT THE BACKGROUND OF DOCTOR BALDUS THAT WOULD INDICATE THAT
HE DID.
IM RELYING PRIMARILY ON ED GATES LOOKING AT THE DATA.
IM NOT RELYING VERY HEAVILY ON HIM FOR THAT DECISION. EITHER.
SO) LET”S MOVE AHEAD.
MR. FORD: THANK YOU, YOUR HONOR.
BY MR. FORD:
2. LET ME ASK YOU PERHAPS A HYPOTHETICAL QUESTION, WHICH I HOPE
I CAN STATE CLEARLY, BUT WITH REGARD TO THIS UNKNOWN QUESTION.
O
0
N
T
R
D
N
-
o
TW
O
p
b
. 1388
KATZ - CROSS
LETS SAY YOU WERE TRYING TO DO A SURVEY OF THE
AVAILABILITY OF INFORMATION IN CERTAIN SOURCES. AND. WHETHER
THAT INFORMATION WAS PRESENT. OR ABSENT IN SAY A PARTICULAR SET
OF RECORDS.
IN THAT CONTEXT. IF YOU HAD SOMEONE GO OUT AND REVIEW
THOSE RECORDS, AND CODE ALONG A NUMBER OF VARIABLES. WHETHER
PARTICULAR INFORMATION WAS PRESENT OR ABSENT. IF THEY CODED
"ABSENT", THAT WOULDN’T BE AN UNKNOWN IN THE SENSE YOURE
SPEAKING OF, IS THAT RIGHT?
A. THE SENSE THAT I USE "UNKNOWN" IS I CANNOT DETERMINE IF THE
CORRECT VALUE IS ZERO OR ONE.
S0, IN THAT INSTANCE, IF THERE‘S NO ABILITY TO
DETERMINE WHETHER IT’S PRESENT OR NOT, I DON’T SEE HOW I CAN
MAKE ANY INFERENCES ABOUT ASSIGNING IT A ZERO OR ONE, SO IN MY
UNDERSTANDING, THAT’S UNKNOWN.
@. DOESNT IT DEPEND, DOCTOR KATZ, ON WHAT THE ULTIMATE
QUESTION YOURE ASKING IS? IF YOU'RE ASKING AS TO WHETHER THE
DATA. THE INFORMATION IS PRESENT IN THE SOURCE. ISN’T THAT A YES
OR NO ANSWER, ISN’T A YES OR NO ANSWER APPROPRIATE?
A. IF THE QUESTION IS SPECIFICALLY. IS THIS STATED IN THE
SOURCE, THATS A YES OR NO ANSWER.
@. SO IF THE SIGNIFICANCE IN THIS SURVEY OF A "U" IS TO
INDICATE THERES NOTHING IN THE FILE TO INDICATE THAT SOMETHING.
ALONG THE DATA SOURCE THERE, THAT IS A FACT THAT CAN BE CODED
AND EVALUATED, IS IT NOT. THE ABSENCE OF DATA ON THAT SUBJECT?
138%
KATZ - CROSS
1 A. PLEASE REPEAT THE QUESTION.
@. IF THE QUESTION THAT IS TO BE ANALYZED IS WHETHER OR NOT A
[3
CERTAIN FACT IS REFLECTED IN THE DATA SOURCE, THE ANSWER "NO" IS
W
w
W
A DATUM THAT YOU CAN ANALYZE IN EVALUATING WHAT'S IN THOSE DATA
SOURCES AND WHAT IS NOT? a
6 THE COURT: MR. FORD, MAY I OBSERVE THAT THAT QUESTION
7 ASSUMES FACTS WHICH ARE NOT ONLY NOT IN EVIDENCE, BUT ARE
8 CONTRARY TO THE EVIDENCE.
? IF YOU’RE GOING TO TRY TO EXAMINE HIM ABOUT IT. LET'S
0 DEAL WITH WHAT WE HAVE. AND LET ME TELL YOU WHY IT CONCERNS ME.
11 THE EXAMPLE THAT WE TALKED WITH MR. GATES ABOUT HAD TO
12 DO WITH A FELLOW WHO MENTIONED THE CRIME TO SEVERAL PEOPLE. AND
13 THE DECISION WAS WHETHER TO CODE BRAGGED ABOUT THE CRIME. AND
14 IF I REMEMBER GATES” TESTIMONY, IT WAS, IT WAS POSSIBLE THAT HE
15 BRAGGED ABOUT THE CRIME BECAUSE HE TALKED TO OTHER PEOPLE. BUT
146 THERES NOTHING IN THE FILE THAT SPECIFICALLY SAYS WHAT HAPPENED
17 DURING THAT CONVERSATION, SO IM GOING TO CODE IT A “U". ITS
13 NOT INCONSISTENT.
ao 32 I DON’T KNOW, WHAT I“M TRYING TO SUGGEST TQ YOU, I
20 THINK YOU MAY HAVE GOTTEN OUT OF THIS WITNESS ALL HE KNOWS AND
21 PERHAPS OUT OF DOCTOR WENTWORTH AND PROFESSOR BALDUS ALL THEY
22 KNOW.
23 IT STRIKES ME AS A PURE JUDGMENT CALL. LOOKING AT IT
24 FROM THE LAWYERS POINT OF VIEW, I THINK THE. YOU COULD INFER
235 EITHER WAY. I COULDN‘T FAULT PROFESSOR BALDUS AS A LAWYER AS
Co
S
E
R
T
t
U
R
EE
SE
R
B
i
E
E
a
Fo
rs
[Y
Y
Fo
y
KATZ - CROSS
OPPOSED TO A SOCIAL SCIENTIST. FOR SAYING THAT SINCE THE
POSSIBILITY EXISTED, I“LL COUNT IT AS BEING THERE.
I LIKEWISE COULD NOT. SINCE IT DOES NOT APPEAR BY ANY
LEVEL OF SIGNIFICANT EVIDENCE WHETHER IT’S PROBABLE CAUSE,
PREPONDERANCE OR WHATEVER. SEE HOW HE COULD CODE IT THE OTHER
WAY.
BUT I THINK WHAT THIS FELLOW IS SAYING AND PERHAPS WHAT
OTHER STATISTICIANS MAY SAY, IS THAT IN THE ABSENCE OF ANY
INFORMATION ON THE SUBJECT ONE WAY OR THE OTHER. IT IS UNFAIR TO
PUT IT IN THE BASE, WHICH THEN TENDS TO REDUCE THE SIGNIFICANCE
OF WHAT OCCURS IN THAT VARIABLE.
AND THATS WHAT 1 UNDERSTAND THIS WITNESS” TESTIMONY TO
BE, AND I THINK YOU HAVE EXHAUSTED THE EXTENT TO WHICH HE CAN
TESTIFY BEYOND THAT.
HAVE I STATED YOUR VIEW OF THE MATTER?
THE WITNESS: YES, YOUR HONOR.
THE COURT: ALL RIGHT.
BY MR. FORD:
@. DOCTOR KATZ. WITH REGARD TO EXHIBIT 19, COUNTS OF "OTHER"
ITEMS IN THAT PROCEDURAL REFORM STUDY, THE ITEMS THAT ARE CODED
"OTHER", AS I UNDERSTAND IT. YOU HAD DATA IN THE DATA SET
AVAILABLE TO YOU THAT INDICATED HOW MANY CASES AND WHICH CASES
SHOWED THAT CODE. IS THAT CORRECT?
A. YES.
@. DID ANY OF THE ANALYSES YOU MADE ATTEMPT TO INCLUDE THAT
1591
KATZ - CROSS
TN
DATA, THAT THERE WAS SOMETHING ELSE THERE, EVEN THOUGH IT WASN-T
SPECIFICALLY DEFINED?
A. IT’S DECEPTIVE. I ONLY HAVE THE INDICATION THAT THERE®S
OTHER INFORMATION THAT APPLIES TO THAT CASE. I DON’T KNOW WHAT
THAT INFORMATION IS. I ONLY HAVE THE INFORMATION ON A COMPUTER
FILE. I DO NOT HAVE THE QUESTIONNAIRES IN WHICH THE ACTUAL
CIRCUMSTANCE IS WRITTEN IN ON THE QUESTIONNAIRE, EXCEPT FOR A
FEW QUESTIONNAIRES THAT WE COLLECTED AT SYRACUSE.
“0
w
NN
0
Oe
W
N
@. BUT AS I UNDERSTAND IT, YOU HAVE INFORMATION WHICH SHOWS YOU
10 WHICH CASES HAVE SOMETHING ELSE IN THESE CATEGORIES. AN "OTHER"
11 CATEGORY, MARKED IN THESE CATEGORIES. THAT FACT. NO OTHER IS
12 AVAILABLE TO YOU, IS THAT RIGHT?
13 A. YES.
14 QR. IS THERE ANYTHING IN THESE ANALYSES YOU MADE INDICATE THAT
135 INCLUDING THAT DATUM MAKES ANY DIFFERENCE?
16 A. IN MY ANALYSIS. I USED THE VARIABLES THAT WERE DEFINED BY
17 PROFESSOR BALDUS, AND IF HE INCLUDED "OTHER" INFORMATION IN
i3 THOSE VARIABLES. THEN IN MY ANALYSIS BY INCLUDING THOSE
@ 19 VARIABLES, I INCLUDED THE "OTHER" INFORMATION. I DID NOT MAKE A
20 SPECIAL EFFORT TO TRY TO INCLUDE THE "OTHER" INFORMATION. SINCE
21 I DID NOT DEFINE NEW VARIABLES. I TRIED TO RESTRICT MYSELF TO
22 THE VARIABLES THAT HAD BEEN DEFINED BY PROFESSOR BALDUS.
23 R. AS FAR AS YOU KNOW, IF SOMEONE WERE ABLE TO SOMEHOW TO TAKE
24 EVERY ONE OF THOSE "OTHER" THINGS THAT DIDNT FIT IN THE 23. 27
25 OR EVER HOW MANY CATEGORIES THERE WAS ON A GUESTION. IF SOMEBODY
vw
0
N
N
Oo
A
Pb
10
11
12
1392
KATZ - CROSS
TOOK THE TROUBLE OF DEFINING ALL THOSE AND MAKING NEW
SUBCATEGORIES AND TURNING THEM INTO ADDITIONAL FQILS, YOU DONT
HAVE ANYTHING INFORMATION THAT INDICATES THAT WOULD MAKE ANY
DIFFERENCE. DO YOU?
A. I HAVE NOT DONE THAT EXPERIMENT, SO I DON’T KNOW WHAT THE
EFFECTS OF THAT EXPERIMENT WOULD BE.
@. THE ANSWER IS NO?
A. THE ANSWER IS NO.
Re WITH REGARD TO WHAT I HAVE MARKED HERE AS TABLE 3. I GUESS
IT’S NUMBER. WELL I GUESS IT’S NO LONGER TABLE 3, IT’S NUMBER
20A. A TABLE OF NON-MATCH COUNTS.
AS I UNDERSTAND IT, THESE ARE PLACES WHERE IN YOUR
COMPUTER ANALYSIS OF THE TWO FILES. THE PROCEDURAL REFORM AND
CHARGING AND SENTENCING STUDIES, YOU FOUND DIFFERENCES IN THE
CODING OF CASES THAT YOU HAD ATTEMPTED TO MATCH, IS THAT RIGHT?
A. THESE ARE COUNTS OF THE, OUT OF THE 361 CASES THAT ARE
CONSIDERED. FOR THE PARTICULAR VARIABLES UNDER QUESTION. IT“S A
COUNT OF NUMBER OF CASES IN WHICH THEY ARE NOT GIVEN THE SAME
VALUE. GENERALLY. A ZERO OR A 1.
@. AND WITH REGARD TO, DOES THIS PROVE ANYTHING OTHER THAN THE
CHARGING AND SENTENCING STUDY MAY BE A CONSIDERABLE IMPROVEMENT
OVER THE PROCEDURAL REFORM STUDY IN TERMS OF THE INFORMATION IT
GETS?
MS. WESTMORELAND: YOUR HONOR, ILL OBJECT TO THAT
AUESTION. GOES BEYOND THE WITNESS” EXPERTISE AND THE PURPOSE
te e—— Smet. bnieiiet | Yee. Yor. Sh. Mobb. ented Set. foment Nene S—— Sh———— ———— —————
1393
KATZ - CROSS
1 FOR WHICH HE“S TESTIFYING.
2 MR. FORD: I AGREE ABOUT THE FORMER. BUT NOT THE
3 |LATTER. YOUR HONOR.
4 THE COURT: THE QUESTION IS ESSENTIALLY NOTHING BUT
Ss |ARGUMENTATIVE. BUT IF YOU CAN THINK OF SOME WAY TQ ASK IT. I“LL
i 5. [LET vou ASK IT.
7 |BY MR. FORD:
8 |@. THIS TABLE. AS I TAKE IT. DOCTOR. DEMONSTRATES DIFFERENCES,
9 |NOT ERRORS, IS THAT RIGHT?
10 A. I“VE ONLY COUNTED DIFFERENCES. COMPARING THE TWO PARTICULAR
11 VARIABLES THAT APPLY TO A SINGLE VARIABLE. OUT OF THE 3é41 CASES.
12 @. ALTHOUGH IF I RECALL, YOU DID NOT USE THE CODING
13 INSTRUCTIONS FOR CHARGING AND SENTENCING STUDY TO DETERMINE
14 WHETHER THEY WOULD MAKE ANY DIFFERENCE IN THE CODING. DID YOU?
13 A. I USED -- THAT’S NOT CORRECT. |
1&6 I ALSO USED THAT INFORMATION IN DETERMINING THE CORRECT
17 CODING, EXCEPT SOME OF THESE CODES ARE BASED ON, OR MOST OF
13 | THESE CODES ARE BASED. FOR THE CHARGING AND SENTENCING
& iv VARIABLES, ARE BASED ON THE UNDERLYING QUESTIONNAIRE ITEM. SO
20 THERE ARE ONLY A FEW CASES WHERE I HAD TO RESORT TQ THAT
21 INFORMATION.
22 I DID CONSULT EVERY INFORMATION TO TRY AND BE SURE THAT
23 WHAT I WAS COMPARING, AS I OBSERVED, STATED THE SAME AND
24 DIFFERENCES ARE NOT BEING INCLUDED.
23 G. I GUESS WHAT I“M ASKING YOU, DO YOU HAVE ANY INFORMATION
1594
KATZ - CROSS
THAT SHOWS YOU THAT THIS TABLE OF INCONSISTENCIES REFLECTS
®
ANYTHING OTHER THAN DIFFERENCES AND POSSIBLY IMPROVEMENTS
BETWEEN THE METHODOLOGIES OF THE TWO STUDIES IN GETTING
INFORMATION AND DEFINING IT?
A. THIS TABLE DOES NOT SAY MORE THAN THESE NON-MATCHES OCCUR
BETWEEN THE VARIABLES THAT ARE COMPARED FOR THE 361 CASES.
@. DOCTOR KATZ, YOU TESTIFIED WITH REGARD, I THINK, AT A NUMBER
OF POINTS, WITH REGARD TO THE COMPUTATION OF "P" VALUES WHICH IS
A STATISTICAL SIGNIFICANCE MEASURE. IS THAT RIGHT?
[e
y Cc
0
©
~
0
ga
»
WW
O
N
A. YES.
PY
Te @. AND YOU COMPUTE THOSE YOURSELF, OR DOES THE COMPUTER DO IT
-
[¥)
OR CAN YOU DO IT ON A HAND CALCULATOR? HOW IS THAT DONE?
13 A. PERTAINING TO WHAT SPECIFIC COMPUTATION? IN TERMS OF THE
14 |TABLE THAT 1 PRESENTED ON "P" VALUES?
15 BQ. YES?
16 A. THAT WAS, THOSE ARE APPROXIMATE "P" VALUES. BASED ON AN
17 EXPERIMENT IN WHICH I ASSUMED THE NORMAL APPROXIMATION TO THE
13 BINOMIAL. AND IT“S AN EXPERIMENT BASED ON TOSSING COINS. AND
Ko ig BEING ABLE TO DETERMINE EXACTLY THE PROBABILITY OF GETTING A
20 CERTAIN OUTCOME. AND EXTRAPOLATING THAT TO AN APPROXIMATE
21 |ESTIMATE OF THE ASSOCIATED "P" VALUE FOR THE --
22 @. WHAT THE "P" VALUE INDICATES IS THAT, THE LIKELIHOOD THAT
23 THAT OUTCOME COULD BE THE RESULT OF CHANCE, IS THAT RIGHT?
24 A. IN THIS CIRCUMSTANCE... YES, APPROXIMATELY.
235 @. CAN YOU COMPUTE WHAT THE "P" VALUE WOULD BE FOR THE CHANCE
EE
1523
KATZ - CROSS
OF COMING UP WITH ONE HEADS IN TWELVE COIN FLIPS?
A. YES.
@. WOULD THAT BE SOMETHING YOU COULD DO HERE. OR WOULD YOU HAVE
TO GO TO THE COMPUTER FOR THAT?
A. CAN WE ASSUME THAT THE PROBABILITY OF GETTING. FIRST OF ALL,
DO WE CONSIDER THE COIN IS FAIR?
@. ASSUME THAT THE COIN IS FAIR. FIFTY FIFTY CHANCE?
A. AND WE“RE ASSUMING NO UNDERYLING INFORMATION ABOUT HOW THE
COIN IS TOSSED, IT’S JUST FIFTY FIFTY?
QR. FIFTY FIFTY CHANCE?
A. AND TWELVE COIN TOSSES, EACH COIN TOSS HAS THE SAME
PROBABILITY, ONE-HALF OF BEING HEADS.
Q. LET'S ASSUME THAT, THAT WAS THE GENERAL ASSUMPTION?
A. YES. ONE HEADS OUT OF TWELVE TOSSES.
QA. ONE OR FEWER, ISNT THAT THE CONVENTION FOR "P" VALUE?
A. WELL, BE SPECIFIC. DO YOU WANT EXACTLY ONE, ZERO OR ONE,
SOME OTHER COMBINATION?
@. I GUESS I WANT IT DONE ACCORDING TO -- IS THERE A DIFFERENT
“P* VALUE CONVENTION FOR THE DIFFERENT QUESTIONS? LETS SAY ONE
OR FEWER. IM NOT SURE I UNDERSTAND WHAT YOU“ RE SAYING.
A. ONE OR FEWER?
R. YES.
A. ALL RIGHT. I BELIEVE THAT‘S APPROXIMATELY .04.
@. SO THAT WOULD BE STATISTICALLY SIGNIFICANT? THAT'S NOT.
ASSUMED TO NOT BE DUE TO CHANCE?
NO
RE
R
E
T
h
e
a
a
a
Pa
y (a)
Te
y
pew
s
1396
KATZ - CROSS
A. LET ME TRY THIS AGAIN TO BE SURE.
YEAH, I BELIEVE I MADE AN ERROR THE FIRST TIME. 17'S
SOMETHING IN THE ORDER OF .003. I CAN GIVE YOU THE FORMULA FOR
IT. 1 DON’T KNOW IF I“M OPERATING THE CALCULATOR PROPERLY.
@. THAT WOULD BE HIGHLY STATISTICALLY SIGNIFICANT IS. THAT
RIGHT?
A. YES.
@. YOU DID THAT ON A HAND CALCULATOR. I DON’T KNOW IF IT MAKES
ANY DIFFERENCE. BUT COULD YOU COMPUTE THE PROBABILITY OF A FAIR
CHANCE DRAW OF ONE OR FEWER BLACK JURORS FROM A POOL OF 48
PERCENT BLACKS, ADULTS IN A COUNTY? COULD YOU DO THAT ALONG THE
SAME LINES? |
A. IS THIS A SAMPLE WITH REPLACEMENT OR WITHOUT REPLACEMENT?
IS THIS ASSUMING ~-
THE COURT: MR. FORD, I DON’T UNDERSTAND THAT TO BE AN
ISSUE IN THE CASE. AND I DON’T KNOW WHAT THESE DEMONSTRATIONS
HAVE TO DO.
MR. FORD: PERHAPS I AM MIXING DEMONSTRATIONS WITH
SUBSTANTIVE MATTERS, YOUR HONOR. WE DO BELIEVE THAT THAT IS AN
ISSUE IN THE CASE THAT WE HAVE PUT SOME TESTIMONY IN ON, BUT
PERHAPS, RATHER THAN WASTE TIME, IF I COULD ASK DOCTOR KATZ, IF
HE, HE CAN ESTIMATE WHETHER THAT WOULD MAKE A SIGNIFICANT
DIFFERENCE. THE DIFFERENCE BETWEEN .48 AND A .35 CHANCE. WOULD
THAT BE APPROXIMATELY THE SAME RESULT?
PERHAPS IF THE COURT, I WOULD COMPLETE THE TESTIMONY
0
.
A
P
W
N
—
_
po
s
oO
[w
ry
nN
1397
KATZ - CROSS
WITH DOCTOR KATZ, NOT MEANING TO GO --
THE COURT: I DON‘T WISH TO HAMPER YOUR
CROSS-EXAMINATION, MR. FORD, BUT I DON’T KNOW WHAT YOU“RE
TRYING TO DO. ARE YOU TRYING TQ SHOW HE DOESN'T KNOW HOW TO
COMPUTE PROBABILITIES. OR ARE YOU TRYING TO ESTABLISH A “P*
VALUE. OR WHAT IS IT YOURE TRYING TO DO THAT HAS TO DO WITH HIS
DIRECT EXAMINATION?
MR. FORD: I THINK IM TESTING THE DOCTOR’S COMPUTATION
OF "P" VALUES ON A SUBJECT THAT IS OF NOT, NOT INSIGNIFICANT IN
THIS CASE. YOUR HONOR. IT”S A SIGNIFICANT SUBJECT THAT’S BEEN
TESTIFIED TO HERE BEFORE.
BUT PERHAPS I AM MIXING TWO THINGS THAT —-
THE COURT: ARE YOU CHALLENGING THE ONLY ASSUMPTION OR
CALCULATION HE HAS MADE THAT HAS ANY CONSEQUENCE WITH THE
BALANCE OF HIS TESTIMONY, IS THAT A "ZI" FACTOR OR WHATEVER HE
CALLS IT OF 1.4 SOMETHING IS THE EQUIVALENT OF A "P" FACTOR AT
THE .0S5 SIGNIFICANCE LEVEL. IS THAT RIGHT OR WRONG. ACCORDING
TO YOUR VIEW?
| MR. FORD: I DON’T HAVE ANY INFORMATION, ESPECIALLY IN
LIGHT OF DOCTOR KATZ” RECENT COMPUTATIONS THAT HE IS
CALCULATING "P" VALUES IN ANYTHING OTHER THAN WHAT I UNDERSTAND
TO BE THE STANDARD MANNER. I WAS TESTING THAT QUESTION --
THE COURT: LET”S MOVE ALONG.
MR. FORD: THANK YOU.
BY MR. FORD:
vv
0
NN
Oo
A
ob
10
a
1398
KATZ - CROSS
QR. DOCTOR KATZ, I WOULD LIKE TO ASK YOU SOME QUESTIONS. PERHAPS
WE CAN MOVE AHEAD TO THE REGRESSION EXPERIMENTS. FIVE SETS OF
REGRESSION EXPERIMENTS THAT YOU DID.
THE COURT: WHAT’S YOUR QUESTION?
MR. FORD: I“M TRYING TO FIND IT AT THIS POINT, YOUR
HONOR.
BY MR. FORD:
@. FIRST QUESTION, DO YOU RECALL WHAT EXHIBIT NUMBER THAT IS?
A. RESPONDENT’S EXHIBIT 3%.
GB. THANK YOQU,
THE COURT: IF YOU DIDN‘T GIVE THE COURT OR MR. FORD
ONE WITH THAT NUMBER, GIVE HIM THE TABLE NUMBER. TOO, SO HE CAN
FIND IT.
MS. WESTMORELAND: IT-S TABLE 221A, YOUR HONOR.
THE COURT: ALL RIGHT.
BY MR. FORD:
@. NOW. DOCTOR KATZ, I NOTICE IN ALL FIVE OF THESE REGRESSIONS,
ARE ANY OF THESE REGRESSIONS THAT DOCTOR BALDUS —-- PROFESSOR
BALDUS OR DOCTOR WOODWORTH DID?
A. NO, THEYRE NOT.
@. NOW, EACH ONE OF THESE. CONTAINS SOME BUT CERTAINLY NOT ALL
OF THE AVAILABLE VARIABLES THAT HAVE BEEN CODED ON THE DATA BASE
IN THIS CASE. IS THAT RIGHT?
A. THAT'S CORRECT.
2. HOW WAS THAT SUM, WHO WERE YOU. PERHAPS I SHOULDNT USE
be
S
E
”
BE
RS
E
E
L
S
E
E
S
0
S
R
E
L
e
B
y
0
BT
F
o
=
—
-
ro
y nN
13
15979
KATZ - CROSS
WORDS THAT SOUND LIKE OTHER WORDS, HOW DID YOU DECIDE WHAT
VARIABLES TO USE?
A. 1 CHOSE THE VARIABLES TO INSURE THAT MCCLESKEY’S CASE, WHICH
IS L-16, BE AS. HAVE A PREDICTED VALUE AS HIGH AS POSSIBLE.
@. DID YOU DO THAT THROUGH A COMPUTER MANIPULATION. OR DID YOU
JUST EYE THAT, OR —-—
A. 1 DID THAT WITH THE UNDERSTANDING IF I PUT IN VARIABLES IN
WHICH MCCLESKEY HAD, HAD, HAD A "i" VALUE FOR THOSE VARIABLES,
THAT WOULD ALLOW THE REGRESSION TO USE THAT VARIABLE TO MAKE
MCCLESKEY“S PREDICTED OUTCOME VERY VERY HIGH.
@. SO IN OTHER WORDS, THESE ARE SELECTED ACCORDING TO WHAT.
ACCORDING TO ALL THE VARIABLES THAT YOU COULD FIND THAT
MCCLESKEY HAD A "1" VALUE ON, IS THAT RIGHT?
A. NO. THESE VARIABLES ARE TQ DEMONSTRATE HOW THE INDEXES WORK.
THE PARTICULAR VARIABLES THAT I CHOSE IN SOME OF THESE CASES WAS
SO THAT IN MCCLESKEY“S CASE, 1 WOULD NOT HAVE THE QUESTION
BROUGHT UP. WELL HE HAS A LOW PREDICTED OUTCOME VALUE. THAT
WASN’T THE PURPOSE OF IT, THAT WAS JUST SOMETHING I BUILT IN. I
USED TO RUN THIS EXPERIMENT.
NOT THAT I BELIEVE ANY OF THESE INDEXES ARE MEANINGFUL.
THAT WAS JUST PART OF MY RATIONALE IN PICKING THOSE VARIABLES.
THE COURT: THE R-SQUARES ARE APPLIED TO ALL CASES OR
ONE CASE?
THE WITNESS: THE R-SQUARED IS A MEASURE.
THE COURT: I KNOW WHAT AN R-SQUARE IS GENERALLY. BUT
A ——— _ c——— J —— ——. AO——._ F———— SA —— . A—————" RAPE. _ AT, PEA SANTA AAA AA, SNA OT IAN A APIA OO. SONI SO. pit
B
E
R
e
E
E
D
O
R
R
E
E
S
C
LE
E
NR
F
u
="
SE
©
J
*
M
E
E
A
B
E
Be
PRE
S
ti
16
17
1600
KATZ —- CROSS
WHAT IS. WHAT DID YOU APPLY THE REGRESSION TO?
THE WITNESS: TO THE FIRST HUNDRED CASES.
THE COURT: OKAY.
THE WITNESS: IN THE PROCEDURAL REFORM STUDY.
THE COURT: OKAY. I REMEMBER THAT.
BY MR. FORD:
@. THIS AS I UNDERSTAND IT WAS A CLOSED SET?
A. VES.
Q. YOU APPLIED THIS REGRESSION TO.
NOW. DID YOU GIVE THE COMPUTER SOME KIND OF COMMAND
UTILIZING THESE VARIABLES THAT WOULD TRY AND MAXIMIZE THE
REGRESSION FOR THE R-SQUARED CO-EFFICIENT? IS THAT HOW YOU DID
THIS?
A. 1 BELIEVE FOR THE FIFTH REGRESSION, I USED A STEPWISE
PROCEDURE TO ALLOW THE COMPUTER A GREAT DEAL OF FLEXIBILITY TO
BUMP UP THE R-SQUARED, SELECTING THE VARIABLES.
8. SO BASICALLY THE COMPUTER SELECTED THESE VARIABLES IN
REGRESSION NUMBER FIVE. IN ORDER TO SATISFY THE COMMAND TO
INCREASE THE R-SGQUARED TO THE MAXIMUM AMOUNT POSSIBLE WITHIN
THIS SUB-GROUP?
A. PLUS. 1 ADDED SOME VARIABLES THAT MCCLESKEY HAD, TO INSURE
THAT HE WOULD HAVE AN AGGRAVATED, PREDICTED OUTCOME.
Q. DID YOU ADD THEM ALL OUT OF MCCLESKEY’S CASE, EVERYTHING
THAT WAS IN THERE OR JUST SOME OF THEM?
A. JUST SOME OF THEM. I WANTED TO MAKE SURE THAT HE WOULD NOT
14601
KATZ - CROSS
HAVE A LOW PREDICTED OUTCOME.
HOWEVER, ONCE A REGRESSION HAS AN R-SOUARED THAT HIGH,
ALL OR MOST OF THE DEATH SENTENCE CASES WILL NATURALLY BE VERY
CLOSE TO 1. BECAUSE THAT IS WHAT THE REGRESSION IS TRYING TO DO,
MAKE THE PREDICTED OUTCOME AS CLOSE AS POSSIBLE TO THE ACTUAL
OUTCOME.
Q. WAS MCCLESKEY IN THIS SET OF ONE HUNDRED CASES?
A. YES.
@. DID YOU DO ANYTHING TO APPLY THESE REGRESSION CO-EFFICIENTS
OR THESE REGRESSION MODELS TO OTHER GROUPS OF CASES --
A. NO.
@. =-- TO SEE IF THEY HELD UP?
A. NO, THIS IS ONLY FOR ILLUSTRATION PURPOSES.
@. SO THIS DOESN‘T. DOES THIS PROVE ANYTHING OTHER THAN YOU
COMMAND A COMPUTER TO DEVELOP A MODEL FOR A SMALL SUBSET THAT
HAS A HIGH R-SQUARED?
A. WHAT THIS SHOWS IS THAT AS THE R-SQUARED IN THE MODEL
INCREASES THE PREDICTED OUTCOMES GET VERY CLOSE TO THE ACTUAL
OUTCOME. THE PREDICTED OUTCOMES WERE USED BY PROFESSOR BALDUS
AS AN INDEX. I MAINTAIN THAT IF HE HAD ENOUGH VARIABLES S50 THAT
THE R-SAUARED IS HIGH. ALL THE DEATH SENTENCE CASES WILL HAVE
PREDICTED OUTCOMES VERY CLOSE TO 13 ALL THE LIFE SENTENCE CASES
WILL HAVE PREDICTED OUTCOMES VERY CLOSE TO ZERO, BUT THAT WOULD
NOT MEAN THE SYSTEM WAS NON-DISCRIMINATORY. THAT'S SIMPLY A
PROPERTY OF HOW REGRESSIONS OPERATE.
—
TU
RE
TR
I
a
1602
KATZ - CROSS
S0 1 CONTEND THAT THE INDEXES THAT HAVE BEEN
CONSTRUCTED DON‘T NECESSARILY MEAN ANYTHING.
@. DOCTOR KATZ. DID YOU TRY AND DO ANYTHING. RUN A PROGRAM THAT
WOULD CREATE AN R-SQUARED OF .95 WITH A NUMBER OF VARIABLES. AND
IN IN A NUMBER OF CASES. ANYTHING APPROACHING WHAT WE HAVE IN
THIS CASE?
A. NO, I DID NOT TRY.
@. YOU DIDN’T TRY?
THE COURT: LET ME ASK YOU A QUESTION.
BY THE COURT:
@. I UNDERSTOOD THE LAST TESTIMONY THAT YOU GAVE TO MS.
WESTMORELAND TO BE THAT IF YOU COULD ADD MORE VARIABLES YOU
WOULD GET A HIGHER R-SQUARED FACTOR. AND AT THAT POINT IN TIME.
YOU FELT THAT THE RACIAL EFFECT TO DISAPPEAR.
AS A PRACTICAL MATTER USING HIS MODEL OR HIS DATA. AT
LEAST ON A HUNDRED CASES, WHICH IS NOT AN INSIGNIFICANT NUMBER,
YOU HAVE CREATED THAT. YOU HAVE CREATED AN R-SQUARE OF 93.
HOW, IS THAT TESTIMONY INCONSISTENT OR CONSISTENT?
A. 1 BELIEVE THEY RELATE TO TWO DIFFERENT METHODS IN WHICH
PROFFESSOR BALDUS USED REGRESSION. IN THIS INSTANCE, PROFESSOR
|BALDUS BENERATED A REGRESSION AS AN INDEX BY USING THE PREDICTED
OUTCOME OF THE REGRESSION.
ANOTHER METHOD THAT WAS USED WAS RUNNING REGRESSIONS
AND LOOKING AT WHETHER THE CO-EFFICIENT OF RACE OF THE VICTIM
TERM WAS SIGNIFICANT, AND MY COMMENTS APPLIED TO THAT PARTICULAR
14603
KATZ ~ CROSS
ASPECT OF THE USE OF REGRESSION.
THE COURT: ALL RIGHT. LETS TAKE A TEN-MINUTE BREAK
AND WE-LL COME BACK AND RUN TILL ABOUT 5:30.
(RECESS TAKEN.)
JOSEPH LAUREN KATZ,
BEING PREVIOUSLY DULY SWORN, RESUMED THE WITNESS STAND AND
TESTIFIED FURTHER AS FOLLOWS:
CROSS-EXAMINATION (CONT’D)
BY MR. FORD:
@. DOCTOR KATZ, WITH REGARD TO THESE OTHER ANALYSES OF THE DATA
THAT YOU'VE DONE THAT ARE NOT INCLUDED IN YOUR EXHIBITS HERE,
YOURE NOT RELYING ON THEM. I TAKE IT. IN YOUR TESTIMONY HERES
YOU‘RE RELYING ON WHAT IS BEFORE US, IS THAT RIGHT?
A. NO.
@. THERE ARE OTHER RUNS, THEN. THAT ARE RELEVANT THAT HAVE NOT
BEEN PRESENTED TO THE COURT. OR YOU THINK THAT ARE RELEVANT. BUT
THAT ARE NOT HERE BEFORE US?
A. 1 BELIEVE SO. VES.
@. HAVE YOU DISCUSSED THOSE AT ALL IN YOUR TESTIMONY TO THIS
POINT?
A. NO.
RA. THAT’S NOT BEFORE THE COURT AT THIS TIME?
A. THAT’S CORRECT.
J
Oo
a
bd
O
N
14604
KATZ - CROSS
@. NOW. WITH REGARD TO THESE VARIOUS CHARTS OR LISTS OF
AGGRAVATING FACTORS. I GUESS THEY START WITH EXHIBIT 25, AS 1
UNDERSTAND IT. DID I UNDERSTAND YOU TO SAY THAT CASES THAT ARE
CODED "U" ON ANY OF THOSE VARIABLES ARE ELIMINATED FROM THE
DENOMINATOR?
A. THIS IS THE PROCEDURAL REFORM STUDY. NO UNKNOWNS WERE
ELIMINATED IN THE STUDY, IN THE ANALYSIS THAT I PRESENT.
@. FROM THE CHARTS THAT ARE RELEVANT TO THE CHARGING AND
SENTENCING STUDY THAT WAS DONE, DID I UNDERSTAND YOU CORRECTLY?
A. YES. AMONG OTHER THINGS, WERE ELIMINATED.
R. NOW, ALONG THOSE LINES OF USING THOSE METHODS OF
COMPUTATION. AS I UNDERSTAND IT. WHAT YOU FOUND IS THAT IN
GENERAL, AGGRAVATING OR WHITE VICTIM CASES ARE MORE AGGRAVATED
THAN BLACK VICTIM CASES. HAVE MORE AGGRAVATING FACTORS CODED
THAN BLACK VICTIM CASES IN GENERAL. IS THAT RIGHT?
A. PURSUANT TO WHICH EXHIBIT?
R. IT SEEMS TO ME THAT WAS THROUGHOUT. I GUESS, BUT ISN’T THAT
WHAT YOU GENERALLY TESTIFIED TO. THE SIGNIFICANCE OF VIRTUALLY
ALL OF THESE EXHIBITS HERE?
A. 1’VE TESTIFIED THAT THERE ARE MORE AGCRAVATING FACTORS,
GENERALLY REPRESENTED FOR WHITE VICTIM CASES AND MORE MITIGATING
FACTORS GENERALLY REPRESENTED FOR BLACK VICTIM CASES.
IG. YOU DON‘T PRETEND TO HAVE ANY EXPERTISE IN TERMS OF THE
SOCIAL OR LEGAL IMPLICATIONS AS TO WHY THAT WOULD HAVE OCCURRED
THAT THE RECORDS EXAMINED IN THIS CASE MIGHT REFLECT MORE
1605
KATZ - CROSS
AGGRAVATING CIRCUMSTANCES IN A WHITE VICTIM CASE?
A. 1 HAVE A STATISTICAL OFINION OF THAT.
@. DO YOU HAVE ANY EXPERTISE THAT WOULD TELL YOU, GIVE YOU ANY
BASIS FOR CONCLUDING WHY DATA ARE IN THAT ONE GROUP OF CASES,
RATHER THAN IN ANOTHER. SYSTEMATICALLY. OTHER THAN THE ABILITY,
AS OPPOSED TO ABILITY TO DETERMINE THAT THEY ARE SYSTEMATICALLY
IN THAT ONE GROUP OF CASES?
A. 1 CAN GIVE YOU A POSSIBLE EXPLANATION THAT HAS OCCURRED TO
ME AS TO WHY WE SEE THAT.
R. IS IT ONE THAT'S WITHIN YOUR REALM OF TRAINING OR NOT?
THE COURT: WELL, WHY DON'T WE -——
THE WITNESS: MARGINALLY.
BY MR. FORD:
QR. WHAT'S THAT?
A. WHITE VICTIM CASES TEND TO INVOLVE CO-PERPETRATORS, MANY
MORE THAN BLACK VICTIM CASES, AND IT SEEMS THAT SINCE THE METHOD
OF CODING HAS BEEN TO GIVE CO-PERPETRATORS ALL THE AGGRAVATING
CIRCUMSTANCES OF THE CRIME WITHOUT TRYING TO SEPARATE OUT THEIR
INDIVIDUAL EFFECT. WE MAY HAVE SOME CASES WHICH RECEIVED LESSER
SENTENCES DUE TO THE FACT THAT THEIR LEVEL OF INVOLVEMENT IN THE
ACTUAL HOMICIDES WAS MUCH REDUCED. ALTHOUGH THEY GET ALL THE
SAME AGGRAVATING FACTORS THAT OCCURRED AT THE TIME OF THE CRIME.
THEREFORE. BECAUSE. AS 1 INDICATE IN THIS TABLE.
CO-PERPETRATOR CRIMES SEEM TO OCCUR MORE IN WHITE VICIM CASES
THAN IN BLACK VICTIM CASES. AND SINCE I DID SOME INVESTIGATION
T
p
N
N
Sy
WA
C
E
L
RE
B
E
G
R
E
pe
C
E
od
[oN
]
a
oO
p
d
Ww
14
13
16046
KATZ - CROSS
IN THIS MATTER IN TERMS OF VOLUNTARY MANSLAUGHTER SENTENCES AND
LIFE SENTENCE NO PENALTY TRIAL SENTENCES. EXCUSE ME, LIFE. YES,
LIFE SENTENCE, NO PENALTY TRIAL CASES. AND I OBSERVED THAT THERE
SEEMED TO BE A PATTERN ACCORDING TO THE DATA WHERE THE FINAL
RESULT COULD HAVE BEEN AN OUTCOME BASED ON THE FACT THAT IT WAS
A LIMITED ROLE IN THE MURDER. AND THAT WOULD SKEW THE
DISTRIBUTION OF AGGRAVATING FACTORS.
THE COURT: LET ME MAKE SURE I UNDERSTAND THAT.
BY THE COURT:
A. WHITE VICTIM CASES GENERALLY HAVE MORE CO-PERPETRATORS THAN
BLACK VICTIM CASES.
A. YES.
RQ. THE AGGRAVATING CIRCUMSTANCES THAT YOU ADDED, YOU ADDED
WHETHER IT WAS TRUE OF THE CO-PERPETRATOR FOR THE CRIME
DEFENDANT?
A. I BELIEVE THATS HOW THE QUESTIONNAIRES WERE CODED IN TERMS
OF THE INSTRUCTIONS GIVEN AS TO HOW TO TREAT CO-FERPETRATORS.
R. AND THENCE THAT IS WHAT YOU DID?
A. NO, YOUR HONOR. THAT WAS, THAT’S MY UNDERSTANDING OF HOW
THE CODING WAS DONE ON THE QUESTIONNAIRES.
@. YOU DIDN‘T DO ANYTHING TQ INTERFERE WITH THAT?
A. NO, I JUST COUNTED WHAT WAS PRESENT ON THE -—-—
@. IS WHAT YOU ARE SAYING, THEN, AS IT COMES DOWN TO THE
INDIVIDUAL DEFENDANT. THERE MAY HAVE BEEN FEWER AGGRAVATING
CIRCUMSTANCES ATTRIBUTABLE TO HIM?
1607
KATZ - CROSS
1 A. YES, IF THAT”S CONSIDERED IN THE FINAL SENTENCING QUTCOME.
2 THEN THERE MAY BE FEWER AGGRAVATING CIRCUMSTANCES DIRECTLY
3 ATTRIBUTABLE TO THAT DEFENDANT. BUT, IN TERMS OF CODING THESE
AUESTIONNAIRES, HE WOULD HAVE BEEN GIVEN. THAT DEFENDANT WOULD
HAVE BEEN GIVEN ALL THE OTHER AGGRAVATING CIRCUMSTANCES THAT
RELATED TO THE ACTUAL HOMICIDE.
0
BY MR. FORD:
@. NOW. DO ANY OF THE EXHIBITS YOU PRESENTED HERE TEST FOR THAT
E
E
S
G
R
.
RE
BE
R
Re
OR TRY AND MAKE ANY ADJUSTMENT ACCORDING TO THAT. OR CONTROLS
10 ACCORDING TO THAT. TO TEST THAT HYPOTHESIS?
11 A. I INVESTIGATED THAT QUESTION. AND I FOUND THAT IT“S
12 DIFFICULT TO SEPARATE OUT THE ACTUAL INVOLVEMENT FOR EACH
13 CO-PERPETRATOR. BECAUSE THERE IS VERY LITTLE INFORMATION GIVEN
14 AS TO THE ACTUAL INVOLVEMENT.
13 THERE ARE QUESTIONS IN EACH STUDY THAT RELATE TO THIS
16 AREA, BUT I FIND THE INFORMATION THERE INCOMPLETE. AND THERE ARE
17 SQ MANY POSSIBLE DIFFERENT CIRCUMSTANCES THAT CAN ARISE, WHERE A
13 PARTICULAR CODING REPRESENTS OR SEEMS TO INDICATE THAT THE
Kos 19 DEFENDANT HAD NO INVOLVEMENT BUT IN FACT, THE DEFENDANT HAD SOME
20 OTHER INVOLVEMENT IN AN AGGRAVATED WAY.
21 THAT CANT BE INDICATED IN THE CODING, BUT I WAS UNABLE
22 TO DO ANY SEPARATION OF THE IMPACT FOR EACH CD-PERPETRATOR IN
23 TERMS OF THE AGGRAVATING FEATURES OF THE OFFENSE.
24 Q. SO YOU HAVE NO ANALYSIS THAT TESTS THAT HYPOTHESIS OR
25 POSSIBILITY, IS THAT RIGHT?
1608
KATZ - CROSS
1 A. WHAT I HAVE DONE IS I“VE INVESTIGATED CLOSELY --
2 @. DOCTOR KATZ, DO YOU HAVE ANYTHING —-
3 A. YES.
4 @. IS THERE ANYTHING BEFORE THE COURT?
S A. NO.
é @. NOW, IN TERMS OF THE IMPLICATIONS OF THE GENERAL FINDING
7 THAT ACROSS THE BOARD, WERE YOU HERE IN COURT, AND DID YOU HEAR
a DOCTOR BALDUS TESTIFY WITH REGARD TO HIS FINDINGS AS TO WHETHER
¥ WHITE VICTIM CASES WERE GENERALLY MORE AGGRAVATED THAN BLACK
10 VICTIM CASES?
11 A. I BELIEVE $0, YES.
12 @. AND DID HE TESTIFY ANYTHING TO THE CONTRARY OF THAT?
13 A. I RECOLLECT THAT I BELIEVE HE TESTIFIED THAT WHITE VICTIM
14 CASES ARE MORE AGGRAVATED.
13 @. DO YOU RECALL HIM TESTIFYING THAT HE THEN CONTROLLED FOR
146 LEVELS OF AGGRAVATION IN MAKING HIS DETERMINATION OF THE
17 EXISTENCE OF DISCRIMINATION BETWEEN SIMILARLY SITUATED
13 DEFENDANTS?
é 19 A. I RECALL THAT WAS HIS EXPERIMENTAL ATTEMPT AT CONTROLLING
20 FOR THESE SITUATIONS. YES.
21 @. BUT YOU’VE NOT DONE THAT. AFTER YDU MADE THIS FINDING, IS
22 THAT RIGHT?
23 A. I HAVE NOT RUN SIMILAR EXPERIMENTS AS PROFESSOR BALDUS DID
24 IN TRYING TO CONTROL AND MATCH SIMILAR CASES. ELIMINATING THE
25 LEVELS OF AGGRAVATION AND MITIGATION. NO.
2 . spt oy " TORR TIS Cr, Sl LF is ne oy sui ee Tt lie Vial wataimial iain wit bord daAtly ny TT, ual
bw
Se
le
+
o
R
E
Y
Se
i
1
B
I
a
a
t
e
p_
[ar
y
12
13
1609
KATZ - CROSS
@. HAVE YOU DONE ANY REGRESSION ANALYSES ON THE FULL SET OF
DATA FROM THE CHARGING AND SENTENCING STUDY THAT MAKE THE
SIGNIFICANCE OF THE RACE OF VICTIM EFFECT GO AWAY?
A. NO.
QR. DO YOU HAVE ANY SETS OF VARIABLES YOU COULD PROPOSE THAT
MIGHT MAKE THE RACE OF VICTIM GO AWAY, THAT YOU“VE CONSTRUCTED
IN CONJUNCTION PERHAPS WITH THE ATTORNEY CENERAL?
A. IVE NOT CONSTRUCTED ANY ADDITIONAL VARIABLES. BUT IF YOU
WANT ADDITIONAL VARIABLES, I CAN DESCRIBE SOME FOR YOU.
@. ADDITIONAL VARIABLES TO THOSE INCLUDED IN THE STUDY OR
ADDITIONAL VARIABLES THAT ARE WITHIN THE STUDY?
A. ADDITIONAL TO THOSE THAT WERE INCLUDED IN THE STUDY.
EXACTLY WHICH STUDY ARE YOU PROPOSING. THE CHARGING AND
SENTENCING STUDY OR THE PROCEDURAL REFORM STUDY?
@. I’M JUST ASKING IF YOU‘VE DONE ANYTHING IN A CONTROLLED WAY,
TO TEST WHETHER THERE“S A REASONABLE HYPOTHESIS. TO TEST
REASONABLE HYPOTHESIS THAT WOULD MAKE THIS RACE OF THE VICTIM
EFFECT GO AWAY IN TERMS OF REGRESSION ANALYSIS, ON THE FULL SET.
NOT JUST A HUNDRED SELECTED?
A. MY VIEW FOR THE REGRESSION ANALYSIS HAS BEEN THAT THERE IS A
SUBSTANTIAL ERROR INTRODUCED IN THE USE OF THOSE, OF CODING
THOSE UNKNOWNS TO BE ZERO. THE REGRESSION ANALYSIS THAT I RAN
INCLUDED THOSE UNKNOWN CODES FOR THE VARIABLES AND GENERATED THE
NUMBER OF OBSERVATIONS ALMOST COMPLETELY DISAPPEARED, SO NO
LEGITIMATE REGRESSION COULD HAVE BEEN RUN FOLLOWING THAT
vv
®
NN
6
O
s
1610
KATZ - CROSS
APFROACH. :
@. DID YOU USE ANY OF THE IMPUTATION TECHNIQUES THAT YOU‘VE
READ ABOUT IN THE LITERATURE TO SEE IF THEY DID ANYTHING?
A. NO.
@. DID YOU TRY ANY SETS OF INTERACTION VARIABLES TO SEE IF THAT
WOULD MAKE ANY DIFFERENCE?
A. NO.
@. EARLIER ON, DOCTOR KATZ, CAN I SEE JK-2. I SHOWED YOU A
PIECE OF PAPER WITH THE PARAPHRASE OF YOUR HYPOTHESIS ON IT.
THATS MARKED JK-2, DO YOU STILL HAPPEN TO HAVE THAT?
MS. WESTMORELAND: YOUR HONOR, I OBJECT. I BELIEVE
DOCTOR KATZ” TESTIMONY IS THAT IS NOT HIS HYPOTHESIS. THE
HYPOTHESIS FOR HIS METHODOLOGY.
THE COURT: I DON’T BELIEVE HE TESTIFIED EITHER. I
THINK HE SAID THATS. THATS HIS RESTATEMENT OF WHAT PROFESSOR
BALDUS SAID, BUT IN ANY EVENT, WHAT ABOUT IT?
BY MR. FORD:
RQ. YOU STILL HAVE THAT DOCUMENT WITH YOU?
A. YES, 1 DO.
MR. FORD: DOES THE COURT HAVE ITS COPY.
THE COURT: I HAD IT A MOMENT AGO. YES.
BY MR. FORD:
@. NOW. IN YOUR REPORT SETTING THIS APPROACH UF, OF LOOKING AT
THE RELATIVE AGGRAVATION LEVEL, THIS IS HOW YOU STATE IT, THE
REASONING PROCESS THAT LED YOU TO RUN THOSE CROSS TABULATIONS
14611
KATZ - CROSS
1 BASED ON AGGRAVATION LEVEL, ISN‘T THAT CORRECT?
2 A. I BELIEVE THIS IS THE EXPLANATION I GAVE IN THE REPORT FOR
THE METHODOLOGY THAT WAS USED.
@. AND THEN YOU WENT ON TO TEST WHETHER MANY DEFENDANTS OF
AGGRAVATED BLACK VICTIM CASES WOULD RECEIVE LESSER SENTENCES
3
4
5
id 64 |THAN DEATH. THROUGH THIS KIND OF MECHANISM, ISNT THAT RIGHT?
7 |A. 1 DON’T UNDERSTAND EXACTLY WHAT YOURE REFERRING TO.
8 |BY THE COURT:
> |g. DID YOU TEST THAT HYPOTHESIS. THAT STATEMENT, DID YOU TEST
10 IT?
11 |A. EXCUSE ME. WOULD YOU PLEASE RESTATE THE QUESTION?
12 |@. DO YOU KNOW WHAT HYPOTHESIS HE‘S TALKING ABOUT? HAVE YOU
13 |GOT A COPY OF IT?
18: las ves.
1s |@. DID YOU TEST IT?
tt. 1a, yma.
17 |BY MR. FORD:
1s |@. IT’S KIND OF DIFFICULT TO TEST THAT HYPOTHESIS WHEN THE WORD
@ 19 |"MANY" IS AN OPERATIVE WORD THERE, IS THAT NOT RIGHT? THAT'S
20 |NOT EXACTLY A PRECISE TERM IN STATISTICAL TERMINOLOGY, IS IT?
2 THE COURT: MR. FORD, REPETITIVE, MR. FORD. GET TO
22 |YOUR POINT.
23 |BY MR. FORD:
oa |g. IM SHOWING YOU A DOCUMENT THAT‘S BEEN MARKED AS PLAINTIFF’S
23 EXHIBIT JK-4. DOES THAT ACCURATELY RESTATE THE HYPOTHESIS
1612
KATZ - CROSS
1 SUBSTITUTING FOR THE WORD "MANY" THE WORDS "A HIGHER PROPORTION?"
2 MS. WESTMORELAND: COULD I SEE A COPY OF THE DOCUMENT
3 FIRST?
4 MR. FORD: IM SORRY.
= I BELIEVE COUNSEL AND I BOTH ARE WITHOUT COPIES AT THIS
POINT. YOUR HONOR —-- COUNSEL DOES HAVE A COPY NOW, YOUR HONOR.
THE WITNESS: YES, JK-4, SUBSTITUES THE WORD "A
&
7
8 HIGHER PROPORTION" FOR"MANY." AS FAR AS I CAN TELL.
9 BY MR. FORD:
1Q @. ISN’T ANOTHER WAY OF STATING THAT HYPOTHESIS THAT IF SOCIETY
11 TOLERATES A HIGHER LEVEL, PERHAPS CAN I USE COUNSEL‘S COPY, I'M
12 SORRY. IF HIGHER LEVELS OF AGGRAVATION ARE TOLERATED FOR
13 DEFENDANTS IN BLACK VICTIM CASES AS COMPARED TO DEFENDANTS IN
14 WHITE VICTIM CASES BEFORE A HIGHER SENTENCE IS IMPOSED. THAN A
13 DEFENDANT IN AN AGGRAVATED BLACK VICTIM CASE WILL HAVE LESS OF A
16 CHANCE OF GETTING THE DEATH PENALTY THAN A DEFENDANT IN A
$7 SIMILARLY AGGRAVATED WHITE VICTIM CASE? ISN‘T THAT A WAY OF
18 RESTATING THAT SAME THING?
“ 17 A. I WOULD NEED TO THINK ABOUT THAT. COULD YOU REPEAT THAT
20 STATEMENT FOR ME?
21 @. WELL, IF WE IGNORE THE FIRST, IT’S THE FIRST PHRASE BEFORE
22 THE COMMA, FOCUS ON THE SECOND PHRASE WHICH I UNDERSTAND TO BE
23 WHAT~S TESTED, ISN‘T THE RUJESTION HERE WHETHER A DEFENDANT IN AN
24 AGGRAVATED BLACK VICTIM CASE HAS LESS OF A CHANCE OF GETTING THE
23 DEATH PENALTY THAN A DEFENDANT IN A SIMILARLY AGGRAVATED WHITE
1613
KATZ - CROSS
VICTIM CASE?
ISN“T WHAT REALLY WHAT THIS SAYS? A HIGHER PROPORTION
oF —-
A. WELL, YOUR DISTINCTION IS IN TERMS OF THE SAME LEVEL OF
AGGRAVATION AND MITIGATION IN THAT PARTICULAR CASE?
Q. 1 GUESS IM TRYING TO ASK ABOUT THE WORD "MANY." SOMETIMES
"MANY" MEANS ABSOLUTE NUMBERS. YOU DIDN‘T MEAN THAT IN YOUR
HYPOTHESIS, DID YOU?
A. NO, THAT WAS NOT WHAT WAS MEANT.
@. SO SOMEHOW. "MANY", TO HAVE MEANING IN A CONTEXT OF A COMPLEX
SITUATION LIKE THIS WHERE YOUYRE NOT TALKING ABOUT ABSOLUTE NUMBEF
THAT WORD "MANY" HAS TO BE TRANSLATED INTO A PROPORTION, LNT THAT
RIGHT?
A. I BELIEVE YOU MAY BE TRYING TO TEST A DIFFERENT HYPOTHESIS
THAN WHAT I WAS TRYING TO TEST, GIVEN THE DATA THAT I HAD. 1
BELIEVE YOU'RE TRYING TO GET TO SOME SPECIFIC LEVELS OF
AGGRAVATION AND MITIGATION TO TRY AND DO COMPARISONS.
1 DON’T BELIEVE THE DATA I HAVE IS THAT DETAILED. THAT
ALLOWS THAT.
@. LET’S JUST ASSUME HYPOTHETICALLY THAT IT IS. THAT YOU CAN
DEFINE AGGRAVATION?
A. IF I HAD DETAILED DATA TO WHERE IT CAN BE SUBDIVIDED 350 I
CAN SAY. SUCH AS A CROSS TABULATION PROCESS, WHERE I CAN MATCH
WHITE AND BLACK VICTIM CASES WITH SIMILAR LEVELS OF AGGRAVATION,
MITIGATION, THEN THAT WOULD BE DESIRABLE. YES.
RE »
»
fo
s
r
i
t
R
C
T
RR
J
R
E
E
EE
BE
10
11
1614
KATZ ~ CROSS
@. BUT NOTHING THAT YOU“VE PRESENTED HERE IN COURT TODAY EVEN
ATTEMPTS TO DO THAT?
A. THAT’S CORRECT.
@. EXCEPT FOR, I BELIEVE. A COUPLE OF EXHIBITS THAT WERE DERIVED
DIRECTLY FROM WHAT PROFESSOR BALDUS DID, IS THAT RIGHT?
A. THOSE EXHIBITS. I DON’T BELIEVE, ARE PRESENTED TQ ATTEMPT TD
PERFORM THAT EXPERIMENT OTHER THAN DEMONSTRATE THAT THE CASES
THAT ARE BEING COMPARED ARE PROBABLY NOT SIMILAR IN AGGRAVATING
| AND MITIGATING CIRCUMSTANCES. IT“S NOT MEANT TO SUBSTITUTE AN
EXPERIMENT.
@. CAN WE TURN TO EXHIBIT 337
THAT TABULATION, AS I UNDERSTAND IT. COMPARES DEATH
SENTENCING RATES ALONG THE NUMBER OF AGGRAVATING CIRCUMSTANCES.
IS THAT RIGHT?
A. YES.
RA. S50 THAT DOES CONTROL FOR THE POSSIBLE DIFFERENCE BETWEEN
THOSE GROUPS IN THOSE NUMBERS, IS THAT RIGHT?
A. YES.
R@. AND ALSO TABLE 27, OR EXHIBIT 27. TABLE 107
R. THAT ONE ALSO DOES IN A DIFFERENT FASHION. CONTROLLING FOR
STATUTORY AGGRAVATING CIRCUMSTANCES?
A. THATS CORRECT.
@. IS THAT RIGHT?
AND ON BOTH OF THOSE THERE ARE STATISTICALLY
KATZ - CROSS
1 SIGNIFICANT EFFECTS. DIFFERENCES SHOWN BETWEEN THE DEATH
2 SENTENCING RATES FOR WHITE VICTIM CASES AND BLACK VICTIM CASES.
3 IS THAT RIGHT?
4 A. YES.
S @. AND NO STATISTICALLY SIGNIFICANT EFFECTS THE OTHER
5 DIRECTION, IS THAT CORRECT?
7 A. THAT’S CORRECT.
8 ®@. DOCTOR KATZ, IS IT YOUR TESTIMONY THAT IF BLACK VICTIM AND
? WHITE VICTIM CASES DIFFER IN ONE DIRECTION ALONG THE VARIABLE OF
10 AGGRAVATION. IT-S IMPOSSIBLE FOR THEM ALSO SIMULTANEOUSLY TO
11 DIFFER ALONG THE VARIABLE OF DEATH SENTENCING RATES?
i2 A. AS I UNDERSTAND THE QUESTION, I DON’T THINK IT“S IMPOSSIBLE.
13 @. ISNT THAT WHAT MULTIVARIATE ANALYSIS DOES IS MEASURE THE
14 IMPACTS OF TWO DISTINCT. VARYING VARIABLES?
135 A. THAT'S THE PURPOSE OF MULTIVARIATE ANALYSIS. CORRECT.
1&6 @. BUT YOU’VE NOT DONE THAT KIND OF MULTIVARIATE ANALYSIS IN
17 ANY OF YOUR EXHIBITS HERE TO TEST THE QUESTION OF THE SAME,
13 WHETHER AT THE SAME TIME, WHITE VICTIM AND BLACK VICTIM CASES
w 19 ARE DIFFERENT IN THEIR LEVELS OF AGGRAVATION, THEY ARE ALSO
20 DISTINCTLY DIFFERENT IN THE MANNER IN WHICH THEY RE TREATED IN
23 DEATH SENTENCING. HAVE YOU?
22 A. THAT’S CORRECT.
23 MR. FORD: IF 1 COULD HAVE A MOMENT. YOUR HONOR?
24 1 HAVE NO FURTHER QUESTIONS AT THIS TIME. YOUR HONOR.
25 THE COURT: ALL RIGHT. ANY REDIRECT?
A
1616
KATZ - CROSS
MS. WESTMORELAND: NO, YOUR HONOR, I HAVE NO QUESTIONS.
THE COURT: ALL RIGHT, THANK YOU.
(WITNESS EXCUSED)
THE COURT: CALL YOUR NEXT WITNESS. MS. WESTMORELAND.
MS. WESTMORELAND: YOUR HONOR, AT THIS TIME I WOULD
LIKE TO INDICATE TO THE COURT MY NEXT WITNESS WOULD BE DOCTOR
BURFORD. AND AS I INDICATED PREVIOUSLY HE IS IN TOWN. AND IS AT
THE HOLIDAY INN AT THIS POINT. APPARENTLY I MISCALCUATED AND
ANTICIPATED THAT THE CROSS-EXAMINATION OF DOCTOR KATZ WOULD LAST
UNTIL 35:30.
I CAN GET DOCTOR BURFORD OVER HERE VERY SHORTLY. IF THE
COURT WOULD DESIRE. I DON’T WANT TO DELAY THE PROCEEDINGS ANY.
THE COURT: I WONDER IF PROFESSOR BALDUS WOULD DO A
STUDY FOR ME ON HOW MANY JUDGES IN THIS COUNTRY WOULD SAY YOU
HAVE RESTED?
WELL, ALL RIGHT.
MS. WESTMORELAND: I APOLOGIZE TQ THE COURT.
THE COURT: PLEASE DON’T RUN OUT AGAIN. I“VE BEEN
TRYING TO DECIDE WHAT I“M GOING TO DO TO ANY LAWYER THAT WASTES
TIME IN THIS PROCEEDING THAT I THOUGHT WAS GOING TQ TAKE 2 DAYS.
SO IVE FINALLY DECIDED THAT I“M GOING TO APPOINT EACH AND EVERY
ONE OF YOU AS A SPECIAL MASTER IN A CONSTRUCTION DAMAGES CASE,
OF WHICH I HAVE MANY.
MR. FORD: PERHAPS I COULD USE A MINUTE OF THE COURT’S
TIME TO MAKE A MATTER CLEAR TO THE COURT WE SPOKE ABOUT THE
0
~
.
5.
A
Pp
0
14617
KATZ - CROSS
EARLY PART OF AFTERNOON.
MY UNDERSTANDING IS IN THIS CASE AND HAS BEEN |
THROUGHOUT MY CAREER THAT I“M VERY CAREFUL WITH REGARD TO
TALKING TO THE PRESS, AND I TRY TO FOLLOW THESE GUIDELINES AS
CLOSE AS I CAN. |
I DID SPEAK WITH THE GENTLEMAN FROM THE ATLANTA
CONSTITUTION TODAY. HE ASKED ME SEVERAL QUESTIONS, INCLUDING
WHETHER DOCTOR BALDUS‘ TESTIMONY HAD EVER BEEN PRESENTED BEFORE.
I TOLD HIM IT HAD NOT. HE ASKED ME IF IT WAS CONTAINED IN THE
COURT RECORD, I MAY HAVE GOTTEN CLOSE TO THE LINE, HE DID ASK ME
-~ I MAY HAVE GOTTEN TO THE LINE —-- WHETHER I CONSIDERED IT
SIGNIFICANT. I SAID WE DID. I BIT MY TONGUE. I SAID,
"PLEASE I“M SORRY, I REFER YOU TO THE COURT RECORD."
I JUST WANTED TO THE COURT TO KNOW WHAT HAD TRANSPIRED
AND THAT I AS OF THIS POINT. HAVE RESOLVED NOT TO ANSWER ANY
FURTHER QUESTIONS.
THE COURT: GOOD. THE LEGAL DEFENSE FUND HAS ALWAYS
ENHANCED ITS CREDIBILITY AT LEAST WITH ME BY NOT COMING TO TOWN
TRAILING CLOUDS OF FLURRY AND GIVING INTERVIEWS ON THE SIDEWALK
IMMEDIATELY AFTER THE HEARING EVERYDAY.
THE ATTORNEY GENERAL HAS ALSO BEEN CAUTIOUS IN THAT
RESPECT.
I DON‘T FEEL PUT UPON BY ANYTHING YOU-“VE TOLD ME YDU“VE
DONE, BUT I WOULD APPRECIATE IT IF YOU WOULD NOT GO ANY FURTHER
THAN WHAT YOU HAVE DONE.
1618
KATZ - CROSS
MR. FORD: THANK YOU, YQUR HONOR. fo
b
THE COURT: FINE.
MS. WESTMORELAND: YOUR HONOR, BY THE SAME TOKEN, I
WOULD NOTE THAT THE GENTLEMAN JUST ASKED ME WHERE DOCTOR KATZ,
WHAT UNIVERSITY HE WAS FROM, AND I INFORMED HIM GEORGIA STATE.
AND THAT WAS ALL. I TOLD HIM I WOULD ANSWER NO FURTHER QUESTIONS.
THE COURT: YOU KNOW, IT’S A FINE LINE. AS A PRACTICAL
ECONOMIC MATTER: THE MEDIA IS HARD PUT TO COVER EVERY
T
Y
E
E
TR
LA
UE
SE
RR
R
|
ne
i
b
d
Se
l
SIGNIFICANT COURT EVENT. AND I HAVE ALWAYS TAKEN THE POSITION
ro
y o THAT A CERTAIN AMOUNT OF BACKGROUND BRIEFING WHICH IS NOT FOR
1 ATTRIBUTION, NOT FOR PUBLICATION, BUT JUST TO UNDERSTAND WHAT'S
12 GOING ON WHEN THEY ZIP IN AND OUT, MAY BE PERMISSIBLE. WE DON’T
13 HAVE A JURY IN THIS CASE, BUT IT IS UNPROFESSIONAL TO TRY TO
14 STIR UP SENTIMENT. USING THE COURT HEARING AS A MEDIA EVENT,
13 WHICH YOU HAVEN’T DONE, AND IT IS UNPROFESSIONAL OBVIOUSLY TO
146 TRY TO BRING INFLUENCE BY GIVING ME HATE MAIL ONE WAY DOR THE
17 OTHER. BUT ALL I MEANT TO INDICATE TO YOU ALL, WAS THAT BY
18 VIRTUE OF MY SAYING THAT HE COULD SEE THE DOCUMENTS THAT WERE IN
Ww 19 EVIDENCE, I DIDN‘T MEAN FOR YOU TO TAKE THAT AS CARTE BLANCHE TO
20 GO AND COMMENT UPON THEM AT GREAT LENGTH AND ARGUE YOUR CASE.
21 IN DUE COURSE YOUR BRIEFS WITH THE ARGUMENTS THEREIN
a WILL BE PRESENTED, AND IF HE CARES, HE CAN READ THEM. AND IN
23 DUE COURSE I“LL FILE AN OPINION AND HE CAN SEE WHAT I SAID.
24 ALL RIGHT. WELL SEE YOU IN THE MORNING AT 9:30.
23 AA
161%
KATZ - CROSS
=
(COURT ADRJOURNED FOR THE DAY.)
2 C-E-R-T-1-F-I-C-A-T-E
3
4 |UNITED STATES OF AMERICA
S |NORTHERN DISTRICT OF GEORGIA
é
7 I, JIM PUGH, OFFICIAL COURT REPORTER OF THE UNITED
8 |STATES DISTRICT COURT FOR THE NORTHERN DISTRICT OF GEORGIA. DO
9 |HEREBY CERTIFY THAT THE FOREGOING ( PAGES CONSTITUTE A TRUE
10 |TRANSCRIPT OF PROCEEDINGS HAD BEFORE THE SAID COURT HELD IN THE
11 CITY OF ATLANTA. GEORGIA. IN THE MATTER THEREIN STATED.
12 IN TESTIMONY WHEREOF I HAVE HEREUNTO SET MY
13 |HAND ON THIS [ DAY OF
14
15
16
JIM PUGH
17 OFFICIAL COURT REPORTER
i NORTHERN DISTRICT OF GEORGIA