Appendix — Elder v. City of Houston
Supreme Court brief2003
Ask Donna
What actually matters in this document.
Text
la
APPENDIX A — OPINION OF THE UNITED STATES
COURT OF APPEALS FOR THE FIFTH CIRCUIT
DATED AND FILED AUGUST 19, 2002
UNITED STATES COURT OF APPEALS
FOR THE FIFTH CIRCUIT
No. 00-20414
GREGORY P. COUNTIE; ET AL.,
Plaintiffs,
GREGORY P. COUNTIE, RENE CALDERON,
MICHAEL MACHA, RICK HOLTSCLAW, NATIVIDAD
VASQUEZ, ROBERT C. NELSON, AND CHRIS YORLOFF,
Plaintiffs-Appellants
VERSUS
CITY OF HOUSTON,
Defendant-Appellee.
No. 01-20519
DOROTHY A EDWARDS; AFRO-AMERICAN POLICE
OFFICERS LEAGUE; HOUSTON POLICE ORGANIZATION
OF SPANISH SPEAKING OFFICERS; LIONEL AARON;
BENNIE CONWAY, CLEMENT B CROSBY, JR; JOSE A
GARCIA; RICHARD C GARCIA; MARIA L GUILLORY;
ANTHONY R JAMMER; CHARLES A MCCLELLAND;
2a
Appendix A
SILAS MONTGOMERY, JR; CLYDE PHILPOTT;
CARL WAYNE REED; RICHARD M SPENCER;
BRUCE D WILLIAMS,
Plaintiffs-Appellees
VERSUS
CITY OF HOUSTON; ET AL.,
Defendants
CITY OF HOUSTON,
Defendant-Appellee
VERSUS
HOUSTON POLICE PATROLMEN’S UNION, AND
THE INDIVIDUAL PEACE OFFICERS IDENTIFIED
IN APPENDIX A, an affiliate of the International Union
of Police Associations AFL-CIO, Local 109; HARIL
WALPOLE; FRANK L ADAMEK; JOE M ALDACO;
WILLIAM E BAKER; T BARANKOWSKI; JERRY A
BRISCOE; RONNIE P BROOKS; GREGORY P COUNTIE;
J DEVEREUX; RUSSELL FEUSSEL; BARBARA
GASTMYER; JAMES KLEIN; DONALD KLEPAC;
STEVEN MCCREARY; DONNIE PARDUE; JAMES
PRITCHARD; L N RACKLEY; J R ROBERTS; JACKIE
SHALLINGTON; DENNIS SPRADLIN; STANLEY
STEPHENS; B G WILLOUGHBY; THOMAS ZIELINSKI;
JEFFREY E BICKEL; MONTY T BRADNEY; JEFFREY L
3a
Appendix A
HATFIELD; SCOTT A MORROW; ROY P MOODY;
ARTHUR OSBORNE; CHERI A PAGE; VINCENT C
RUSSO; W J WISSEL, JR; HOUSTON POLICE
OFFICERS’ UNION,
Intervenors-Plaintiffs-Appellants
Appeals from the United States District Court
for the Southern District of Texas
(98-CV-1600 & 92-CV-2510)
Before JOLLY, DeMOSS, and PARKER, Circuit Judges
PER CURIAM:*
In 1975 and 1976, two groups of plaintiffs brought suit
against the City of Houston, Texas, under the captions Kelley
v. Hofheinz and Comeaux v. City of Houston. The suits
alleged, inter alia, that the promotional examinations used
by the Houston Police Department (“HPD”) were racially
discriminatory. Comeaux was consolidated into Kelley and
settlement discussions continued, unsuccessfully, until 1983,
when the suit became dormant for several years. In 1992, a
group of African-American and Hispanic-American police
officers moved to intervene in Kelley, alleging that they had
been harmed by racially discriminatory promotional
examinations for the ranks of Sergeant and Lieutenant in the
HPD. Ultimately, the district court directed those officers to
file a new lawsuit. They did so on August 19, 1992, under
* Pursuant to 5TH Cir. R. 47.5, the Court has determined that
this opinion should not be published and is not precedent except
under the limited circumstances set forth in StH Cir. R. 47.5.4.
4a
Appendix A
the caption Edwards v. City of Houston. The remainder of
the original Kelley plaintiffs were then consolidated therein.
The Edwards action alleged that the examinations had
the effect of disproportionately excluding African-Americans
and Hispanic-Americans from promotion to Sergeant, and
African-Americans from promotion to Lieutenant, from 1982
onward. The plaintiffs sued on their own behalf and on behalf
of others similarly situated or who would be competing for
promotions to each rank in the future. Settlement negotiations
between the plaintiffs and the City of Houston began in the
fall of 1992 and culminated in a proposed Consent Decree.
Before a fairness hearing could be conducted by the district
court, various other police officer groups sought to intervene,
including the Houston Police Patrolmen’s Union and others.
The district court denied the motion to intervene but allowed
the putative intervenors to cross-examine witnesses, present
evidence, and raise objections to the proposed Consent
Decree during a fairness hearing.
On March 25, 1993, the district court certified a class
consistent with the plaintiffs’ complaint and entered a final
Consent Decree. The Consent Decree included, inter alia,
the provisions that (1) African-Americans and Hispanic-
Americans who took the Sergeant exam from January 1, 1982,
to that date, and who passed at least one such exam, would
receive a total of 96 remedial promotions; (2) African-
Americans and Hispanic-Americans who took the Sergeant
exam from January 1, 1982, to that date, and who were
promoted after a discriminatorily long period delaying their
ability to compete for Lieutenant promotions, would receive
five remedial promotions to Lieutenant; and (3) African-
TE SS NT TO
_ - i - =e
Sa
Appendix A
Americans who took the Lieutenant exam from January 1, 1982,
to that date, and who passed at least one such exam, would
receive a total of five remedial promotions.
Several of the putative intervenors appealed the Consent
Decree and the denial of their motion to intervene. A panel of
this court affirmed the district court. We then re-heard the case
en banc and concluded that the intervention should have been
allowed. See Edwards v. City of Houston, 78 F.3d 983 (5th Cir.
1996) (en banc). On remand, the district court converted the
Consent Decree into an interim order which permitted, but did
not require, the City of Houston to make remedial promotions
in an “acting capacity,” leading to full pay and recognition
for the promotions, but not seniority or retirement benefits.
On February 19, 1997, the City of Houston created, by ordinance,
18 Sergeant and two Lieutenant positions. In accordance with
the Consent Decree, 16 of the Sergeant positions were reserved
for and filled by African-American and Hispanic-American
police officers.
In May 1998, non-promoted HPD members filed a new
suit in Countie v. City of Houston, claiming reverse
discrimination. The district court' granted summary judgment
to the City of Houston concluding that the Countie plaintiffs
could not prove their prima facie case of discrimination because
they were not qualified for the positions and, alternatively,
that their claims were barred by 42 U.S.C. §§ 2000e-2(n).
See Countie, No. H-98-CV-1600 (S.D. Tex. Mar. 30, 2000)
(Order on cross-motions for summary judgment).
; 1. The Countie suit was also filed in the Southern District of
Texas, Houston Division, but was presided over by a different district
judge than in the Edwards proceedings.
6a
Appendix A
The Countie plaintiffs appealed. Before we heard oral
arguments, the district court in Edwards entered a final Consent
Decree, which made permanent the City of Houston’s 16 interim
promotions. See Edwards, No. H-92-2510 (S.D. Tex. Apr. 2,
2001) (Final Consent Decree). The Houston Police Patrolmen’s
Union and other interested parties then appealed that ruling.
We consolidated the various actions and heard oral argument
from each of the parties.
We review a district court’s determination of a Title VII
Consent Decree for abuse of discretion. Williams v. City of New
Orleans, 729 F.2d 1554, 1559 (Sth Cir. 1984). This is particularly
true where the district court has been involved in the negotiations
arriving at the proposed Consent Decree, such as by conducting
fairness hearings and hearing evidence presented by the parties
and intervenors. Jd. at 1558-59. Although voluntary settlement
of Title VII employment discriminations suits is preferable, a
district court cannot summarily approve such a sett!ement
without making an independent decision concerning the fairness
of every provision in the Decree. Jd. at 1559. “This requires a
determination that the proposal represents a reasonable factual
and legal determination based on the facts of the record, whether
established by evidence, affidavit or stipulation.” Jd. When third
parties are involved, the effect of the Decree on those parties
can be “neither unreasonable nor proscribed.” /d. at 1560.
Based on our review of the opinion of the district court, the
briefs, and the record, and having heard oral argument, we are
of the opinion that the district court did not abuse its discretion.
For the reasons stated by the district court, we AFFIRM the
Consent Decree in Edwards. This affirmance renders MOOT
any matter pending in Countie.
7a
APPENDIX B — FINAL CONSENT DECREE OF THE
UNITED STATES DISTRICT COURT FOR THE SOUTH-
ERN DISTRICT OF TEXAS, HOUSTON DIVISION,
DATED MARCH 30, 2001 AND
ENTERED APRIL 2, 2001
IN THE UNITED STATES DISTRICT COURT FOR
THE SOUTHERN DISTRICT OF TEXAS
HOUSTON DIVISION
C. A. NO. H-92-2510
Judge Lynn Hughes
DOROTHY A. EDWARDS, AFRO-AMERICAN POLICE
OFFICERS LEAGUE, THE HOUSTON POLICE
ORGANIZATION OF SPANISH SPEAKING OFFICERS,
LIONEL AARON, BENNIE CONWAY, CLEMENT BOYD
CROSBY, JR., JOSE GARCIA, JR., RICHARD C. GARCIA,
MARIA L. GUILLORY, ANTHONY R. JAMMER, CHARLES
A. MCCLELLAND, SILAS MONTGOMERY, JR., CLYDE
PHILPOTT, CARL WAYNE REED, RICHARD M. SPENCER,
and BRUCE D. WILLIAMS,
individually and on behalf of all others similarly situated,
Plaintiffs,
V.
CITY OF HOUSTON,
Defendant.
and
————— 4
HOUSTON POLICE PATROLMEN’S UNION et al., and
HOUSTON POLICE OFFICERS’ UNION, et ai.,
Third-Party Intervenors.
8a
Appendix B
Final Consent Decree
A. Introduction and History of This Case
1. This is a class action employment discrimination suit
brought by African-American and a. Hispanic-American
police officers, the Afro-American Police Officers League,
and the Houston Police Organization of Spanish Speaking
Officers [“plaintiffs”] against the City of Houston [“‘the City’”’]
pursuant to Title VII of the Civil Rights Act of 1964,
42 U.S.C. §§ 2000e et seg., as amended by the Civil Rights
Act of 1991 and the Equal Employment Opportunity Act of
1972 [“Title VII]. As used herein, the term “plaintiffs” does
not include any of the third-party intervenors.
2. This lawsuit has been consolidated with two earlier
lawsuits, Comeaux v. City of Houston, Civil Action No. 76-
H-1754, and Kelley v. Hofheinz, Civil Action No. H-75-1536,
filed in this Court in 1976 and 1975, respectively. These
lawsuits challenged a variety of allegedly discriminatory
practices, and the Comeaux action specifically challenged
discrimination in promotional tests. By the Order entered on
June 18, 1992 in Comeaux and Kelley, all claims of
discrimination not involving promotional examinations have
been dismissed, and the claims of discrimination involving
promotional examinations have been limited to promotions
for the ranks of Sergeant and of Lieutenant from January 1,
1982 on.
3. In this action, plaintiffs have challenged as
discriminatory the promotional examinations for the rank of
Lieutenant and for the rank of Sergeant in the Houston Police
9a
Appendix B
Department [“HPD”] which were administered during the
period from 1982 through 1991 and the system pursuant to
which these examinations are prepared. Plaintiffs alleged in
this lawsuit that the challenged examinations had the effect
of disproportionately excluding African-Americans and
Hispanic-Americans from promotion to Sergeant from 1982
to date, and of disproportionately excluding African-
Americans from promotion to Lieutenant from 1982 to date,
but were neither job-related nor consistent with business
necessity.
4. The plaintiffs sued on their own behalf, on behalf of
the African-American and Hispanic-American members of
the Police Department who took a Sergeant examination from
1982 through 1991 or who will compete for promotions to
Sergeant in the future, and on behalf of African-American
members of the Police Department who took a Lieutenant
examination from 1982 through 1991 or who will compete
for promotions to Lieutenant in the future.
5. Plaintiffs’ proposed class did not include Hispanic
Americans who took Lieutenant examinations from 1982
through 1991, because the records of the City show that the
Lieutenant examinations did not operate to exclude Hispanic-
Americans from promotion to Lieutenant during this period
of time. However, plaintiffs alleged that the Sergeant
examination discriminatorily delayed the promotions of both
African-Americans and Hispanic-Americans to the rank of
Sergeant, and a delay in their promotion to the rank of
Sergeant is necessarily a delay in their ability to compete for
promotion to Lieutenant.
10a
Appendix B
6. The plaintiffs and the City have consented to the
settlement of this action and to the relief provided herein.
The third-party intervenors have not consented to the
provisions of this Consent Decree.
7. This updated Consent Decree reflects and describes
the actions taken prior to October 16, 2000, including
amendments to the version of the Consent Decree granted
preliminary approval on Feb. 3, 1993, the actions taken
pursuant to the Interim Order of Feb. 5, 1997, and the
completion of all remedial promotions although some are
still acting promotees.
B. Certification of the Plaintiff Class, and Finding
of Jurisdiction
8. Based upon the stipulations of the plaintiffs and the
defendant and the information of record, the Court certifies
and defines the following class:
a. All African-Americans who are employed, or at
any time since January 1, 1982 were employed, as Class
A peace officers by HPD and who took a promotional
examination for the rank of Lieutenant or for the rank of
Sergeant which was administered at any time from
January 1, 1982 to the present, and those who will
compete for such promotions in the future; and
b. All Hispanic-Americans who are employed, or
at any time since January 1, 1982 were employed, as
Class A peace officers by HPD and who took a
promotional examination for the rank of Sergeant which
lla
Appendix B
was administered at any time from January 1, 1982 to
the present, and those who will compete for such
promotions in the future.
9. This Court has jurisdiction over the claims alleged
herein by virtue of 28 U.S.C. §§ 1331 and 1343 and § 706(f)
of the Civil Rights Act of 1964, 42 U.S.C. § 2000e-5(f).
All of the administrative and procedural requirements to suit
under Title VII have been satisfied.
C. Basis of the Consent Decree
10. Plaintiffs and the defendant agreed to enter the relief
provided in the Consent Decree:
a. on evidence that the promotional examinations
for Sergeant from 1982 through 1991 had a racially
disparate impact on blacks;
b. on evidence that the promotional examinations
for Sergeant from 1983 through 1991 had a racially
disparate impact on Hispanics;
c. on evidence that the promotional examinations
for Lieutenant from 1982 through 1991 had a racially
disparate impact on blacks;
d. on evidence that there was a second type of
racially disparate impact in the promotional examinations
for Sergeant as to both blacks and Hispanics, in that the
examinations substartially delayed the promotions
of both blacks and Hispanics to Sergeant and thus
substantially postponed their ability to compete for
promotion to Lieutenant;
12a
Appendix B
e. on ample evidence that the City’s promotional
examinations for the ranks of Sergeant and Lieutenant from
1982 through 1991 were not job-related, or that there was
substantial doubt as to their job-relatedness; and
f. on the experience of the Houston Fire Department
in having promotional examinations prepared in the
same manner as the Police Department promotional
examinations, with the s2.ne personnel in the Houston
Civil Service Departmer working on both examinations,
in having performed a criterion-related validation study
on the promotional examinations to the first two
promotional ranks in the Fire Department, Chauffeur and
Junior Captain, on seeing plaintiffs’ expert study, and in
concluding that there was substantial risk of being found
liable if the case were not settled.
11. The evidence received in the 1999 fairness hearing,
which included evidence adduced during the 1993 fairness
hearing, and the findings of this Court, show that the
defendant City had, and continues to have, a strong basis in
evidence as to each of the above factors.
D. Retrospective Injunctive Relief for the African-
Americans and Hispanic-Americans Harmed by
the City’s Use of the Challenged Examinations
12. Members of the plaintiff class have received 96
remedial promotions to Sergeant, and 10 remedial promotions
to Lieutenant, as set forth below.
13. The plaintiffs and class members [hereafter,
collectively termed “class members”’] receiving remedial
. sta tod ‘ . Pore aha ets
ti a aS ANE ATMOS LSE stage
Pie
l3a
Appendix B
promotions shall also receive full back seniority in rank, for
both competitive and entitlement or “benefits” purposes other
than pension benefits, as if they had actually been promoted on
the test in question. The back seniority in rank shall not apply
to pension benefits. No plaintiff or class member shall receive
back seniority back to the seniority date of a test he or she did
not take. The seniority dates shall be the dates six calendar
months after the establishment of the register pursuant to which
they were promoted, as set forth in the following table:
Table 1. Seniority Dates for Class Members Receiving
Remedial Promotions
Date of Seniority in Rank,
for Remedial Promotions
Sergeant Register From this Register
9/23/82 3/23/83
11/17/83 $/17/84
1/31/85 7/31/85
7/17/86 1/17/87
10/29/87 4/29/88
4/27/89 10/27/89
10/31/91 4/31/92
Date of Seniority in Rank,
for Remedial Promotions
Lieutenant Register From this Register
9/27/84 3/27/85
3/3/88 9/3/88
10/12/89 4/12/90
5/23/91 11/23/91
14a
Appendix B
14. In order to minimize any effect of these remedial
promotions on members of the Police Department (including
plaintiffs or class members) competing for promotions in the
usual course, these promotions were phased in over a five-
year period as follows:
Table 2. Schedule of Making Remedial Promotions
Year of Promotion Sergeant Lieutenant
1993 22 3
1994 21 2
1995 19 2
1996 18 2
1997 16 l
The 1996 remedial promotions were stayed by order of the
court of appeals. The 1996 remedial promotions, along with
the 1997 remedial promotions, were ultimately made on an
acting basis in 1997 under the Feb. 5, 1997, Interim Order.
15. Aclass member who received a remedial promotion,
or who was bypassed for a remedial promotion during those
years pursuant to the provisions of § 18, was able to take a
promotional examination for the next higher rank after having
served in the lower rank for one year, without regard to his
or her retroactive seniority in that rank. For example, a class
member receiving a remedial promotion to Sergeant with ten
years of retroactive seniority as a Sergeant was nonetheless
required to work as a Sergeant for one year before being
eligible to compete for promotion to Lieutenant.
13a
Appendix B
16. Although plaintiffs do not agree that the results of
the challenged examinations have anything to do with
qualifications, both plaintiffs and the defendant recognized
that there was no presently available means of determining
qualifications except for whatever information may be
provided by the test scores. In order to ensure that each of
the class members receiving remedial promotions was
qualified (to the extent that the challenged test scores can so
indicate), no class member received a remedial promotion
for a particular rank unless he or she had passed at least one
of the promotional examinations for that rank during the
1982-1992 period.
17. Subject to the provisions of 4§ 16 and 27, the
remedial promotions for a particular rank were made from
among those plaintiffs and class members who passed one
or more of the promotional examinations for that rank from
1982 to date, and who had not yet been promoted to that
rank.
18. A potential conflict arose when a class member who
was otherwise eligible to receive a remedial promotion was
also on an eligibility list for a regular promotion, and was
reached or was reachable during the life of that eligibility
list. In such a situation, the class member was required to
take the regular promotion, and was not allowed to take a
remedial promotion.
19. To ensure equity for class members subject to the
provisions of { 16 above, a class member who was bypassed
for a remedial promotion pursuant to § 16 received back
seniority in rank at the time of his or her regular promotion,
as if he or she had received a remedial promotion instead.
l6a
Appendix B
The provisions of this paragraph were given priority over
the provisions of other paragraphs of the Consent Decree on
the allocation of remedial promotions.
20. The remedial promotions to Sergeant were allocated
as follows, based on the shortfall figures on which the
plaintiffs and the City had agreed:
Table 3. Allocation of the Remedial Promotions to Sergeant
Number of Vacancies
Sergeant Examination Blacks Hispanics
9/23/82 5 0
11/17/83 8 7
1/31/85 10 7
7/17/86 6 2
10/29/87 7 7
4/27/89 10 ]
10/31/91 16 10
Total 62 34
21. African-Americans suffered 64.6% of the shortfall
in Sergeant promotions, are receiving 64.6% of the remedial
promotions to Sergeant, and received this proportion of the
remedial promotions to Sergeant made in each of these years:
14 remedial promotions in 1993, 14 in 1994, 12 in 1995,
12 in 1996 that were stayed and were ultimately received in
1997, and 10 that were scheduled and received in 1997.
22. Hispanic-Americans suffered 35.4% of the shortfall
in Sergeant promotions, are receiving 35.4% of the remedial
promotions to Sergeant, and received this proportion of the
remedial promotions to Sergeant made in each of these years:
17a
Appendix B
8 in 1993, 7 in 1994, 7 in 1995, 6 in 1996 that were stayed
and were ultimately received in 1997, and 6 that were
scheduled and received in 1997.
23. Five of the ten remedial promotions to Lieutenant
were allocated as follows:
Table 4. Allocation of Five Remedial
Promotions to Lieutenant
Number of Vacancies
Lieutenant Examination Blacks
9/27/84
3/3/88
10/12/89
5/23/91
24. In addition to the five remedial promotions to
Lieutenant described above, there were five other remedial
promotions to Lieutenant as a form of remedy for the delay
in promotion to Sergeant which occurred as a result of the
Sergeant examinations, and which in turn delayed the ability
of those class members promotees to compete for promotion
to the rank of Lieutenant. African-Americans shall receive
two of these promotions, and Hispanic-Americans shall
receive three of these promotions. The class members to
receive these promotions shall be those class members
promoted to Sergeant who have waited the longest times
between taking their first Sergeant examination after January
1, 1982 and their promotions, and who have taken and passed
the examination for promotion to Lieutenant, but who have
not yet been promoted. One such promotion was scheduled
18a
Appendix B
to be made per year over the five-year phase-in period, with the
first such promotion going to an Hispanic-American and the
further promotions alternating between A frican-Americans and
Hispanic-Americans. The promotion scheduled for 1996 was
stayed, and occurred in 1997 instead. The Sergeants promoted
pursuant to this paragraph received back seniority in the rank
of Lieutenant to a date six months after the date of the next
Lieutenant examination which was given at least two years after
their promotion to Sergeant, subject to the conditions and
limitations of § 13 of this Consent Decree. The provisions of
{| 18 and 19 above apply to these remedial promotions the
same as to the others.
25. Counsel for plaintiffs developed the list of the class
members to receive remedial promotions, and presented it to
the City for approval. The Court would have been able to resolve
any disagreements, but plaintiffs and the defendant are able to
resolve all questions by themselves.
26. In compiling the list, the individual named plaintiffs
had priority for receiving the remedial promotions if they
otherwise qualified under the standards set forth herein.
The next priority was to use rank order on the test in question.
27. If for any reason there were not enough still-employed-
but-unpromoted class members of a particular racial or ethnic
group who took and passed one of the challenged tests to make
up the numbers of remedial promotions for that group from
that test, but there are class members of that group who failed
the test in question but passed another test for the same rank
within the 1982-1992 time period, such class members were
given remedial promotions for the test notwithstanding their
failure of the test.
19a
Appendix B
28. If for any reason these numbers still could not be
made up for a particular test by the class members of the
appropriate group in question who took the test, they were
made up from among the unpromoted class members of that
group passing another test, starting with the earliest test from
1982 to date with unpromoted but passing class members.
29. If any class member declined a remedial promotion
for any reason, that promotion was given to the next-highest-
ranking (or scoring) plaintiff or class member.
30. The order in which remedial promotions were
phased in was as follows: (1) the individual named plaintiffs
satisfying the conditions set forth herein were promoted first,
subject to the allocation of promotions as between A frican-
Americans and Hispanic-Americans; (2) remedial promotions
were thereafter made in rank order from among test-passers
and in order of test score (if known) or total score (where
test score was not separately stated) among those class
members who did not pass this test but passed another test
for the same HPD rank; (3) and the tests within the 1982-
1992 time period were reached in chronological order, with
the earliest test first.
E. Limitations on Remedial Seniority in Rank
31. On. February 5, 1999, the Court directed the
plaintiffs and the defendant to reduce the number of plaintiff
class members with remedial seniority in rank to the number
of remedial promotions. On July 12, 1999, plaintiffs’ and
the defendant City filed their Joint Motion for Approval of
their Joint Report to the Court on the Restriction of the
20a
Appendix B
Number of Class Members with Remedial Seniority to 106
Persons. Their Report set forth two principles to govern the
question of which plaintiff class members were to lose their
remedial seniority in rank:
a. The first principle proposed by the parties is that
the class member originally scheduled to receive a
remedial promotion, but who was reachable on a regular
list and was therefore required to delay any promotion
for up to two years until he or she was reached on the
regular list, while another class member took the
remedial promotion, should be the person to keep the
constructive rank seniority for that promotion.
Promotions result in pay increases of several thousands
of dollars a year. The mandatory delay in promotion may
therefore have cost the original class member several
thousand dollars, and retention of the constructive rank
seniority is the only countervailing benefit to that person.
The “extra” remedial rank seniority should be taken from
the class member who received an earlier promotion
because of that provision of the Consent Decree.
b. Second, where a class member who received
constructive rank seniority has died or left the Police
Department ending that grant of constructive rank
seniority, that end should be taken into account as
reducing the total number of grants of constructive
rank seniority towards the goal of 106. Otherwise, the
trimming of grants of remedial seniority in rank ordered
by the Court will reduce the extant total to fewer than
the 106 grants to which the plaintiffs and defendants were
ordered to conform.
21a
Appendix B
32. The Motion is well taken, and it is granted.
The Plaintiffs’ and Defendant’s Joint Motion for Approval
of their Joint Report to the Court on the-Restriction of the
Number of Class Members with Remedial Seniority to 106
Persons is approved. The defendant shall remove the remedial
seniority in rank of the following persons, and shall correct
its records to show seniority in rank only from the date of
their promotion to that rank:
Name Employee and Required Adjustment
Auzenne, Adam 87559 (must lose remedial rank
seniority as Sergeant)
Sanchez, Richard 61115 (must lose remedial rank
seniority as Sergeant)
Rodriguez, Ralph 81759 (must lose remedial rank
seniority as Sergeant)
Carradero, Julio 79433 (must lose remedial rank
seniority as Sergeant)
Rios, William 82381 (must lose remedial rank
seniority as Sergeant)
Munoz, Roberto 59098 (must lose remedial rank
seniority as Sergeant)
Salazar, James 77805 (must lose remedial rank
seniority as Sergeant)
Calix, Elvin
Castaneda, Rodolfo
Picazo, William
Ong, Cynthia
Conway, Bennie
Freddy Guidry
Trevino, John P.
Trevino, Elma A.
22a
Appendix B
82345 (must lose remedial rank
seniority as Sergeant)
57866 (must lose remedial rank
seniority as Sergeant)
88981 (must lose remedial rank
seniority as Sergeant)
80764 (must lose remedial rank
seniority as Sergeant)
57680 (remedial rank seniority as
Lieutenant must be changed from
March 27, 1985, to April 12, 1990)
31177 (must lose remedial rank
seniority as Lieutenant)
58748 (must lose remedial rank
seniority as Lieutenant)
81000 (would have lost remedial rank
seniority as Lieutenant, but resigned)
These Sergeants and Lieutenants shall have seniority in rank
based on the dates of their acting or permanent promotions.
They shall lose any take-home cars they now possess solely
as a result of constructive seniority in rank, but shall not be
removed from their present assignments and shifts for the
purpose of allowing re-bidding for those positions.
23a
Appendix B
F. Waiver of Back Pay
33. Plaintiffs and their class waive all claims for back
pay and interest under all Federal and State anti-
discrimination laws arising from the City’s use of the
challenged examinations.
34. Plaintiffs and their class do not waive any right they
may have to seek appropriate relief against any person who,
or organization which, takes any action which has the effect
of unreasonably delaying any remedial promotions under this
Consent Decree. This paragraph is not intended, and shall
not be construed, to apply to any person or organization,
whether or not a member of the plaintiff class, for making
an adequately founded objection to this settlement under the
procedures set forth below for determining whether this
Consent Decree should be given final approval, or for
pursuing that objection on appeal.
G. Prospective Relief
35. To the extent that the City develops new promotional
examinations for Sergeant and Lieutenant through its Civil
Service Department and does not use alternative examining
procedures pursuant to Meet and Confer agreements, the City
defendants shall continue the procedures for the development
and use of promotional examinations for Sergeant and
Lieutenant which were used for the development of the
promotional examinations for Sergeant and Lieutenant from
1982 to date, except as follows:
a. A firm date for the administration of the selection
procedure and the final list of books (or parts thereof) or
24a
Appendix B
other texts shall be announced to members of the HPD
at least ninety days in advance of the administration of
the selection procedure.
b. The City may continue to use its present system
allowing protests of particular test items, the multiple-
keying of some items where appropriate, and the
discarding of other test items where appropriate.
c. The remaining items on the test shall be scored.
Test-takers must achieve a score of 70% of the remaining
items correct in order to pass the test and to be eligible
for promotion.
d. Seniority points shall be added to the score on
the written test, as is done at present, in order to obtain a
rank-ordered list. The City shall post an eligibility list
as soon as possible. The City shall continue its present
procedures for making promotions on a rank ordered
basis. The City shall consult with the plaintiffs herein
before making any change in these procedures in the
future.
e. Promotional registers of test-passers resulting
from use of the new selection procedures shall remain
in effect for a period of two and one-half years unless:
earlier exhausted.
f. The September 30, 1992 Lieutenant promotional
register was extended for one year, so that it expired on
September 30, 1994. The extension of this list benefitted
all of the persons who passed this test, regardless of race
or national origin.
25a
Appendix B
g. To the extent reasonably necessary to comply with
the former provisions of paragraph with respect to log-
linear analysis, the City may sometimes have taken a
longer period of time to make promotions than the time
allowed under State law, without incurring any liability
for back pay for such additional period of time.
h. The provisions of this paragraph supersede any
provisions of The Fire and Police Civil Service Act,
Texas Local Government Code chapter 143, as amended,
to the contrary.
36. Nothing in this Consent Decree shall be construed
to prevent any amendment of The Fire and Police Civil
Service Act, Texas Local Government Code chapter 143, to
create a statutory two-year life, or longer life, for promotional
registers for any position in the HPD.
37. Nothing in this Consent Decree shall be construed
to prevent the defendant from adopting an alternative
selection procedure for promotion to the ranks of Sergeant
or Lieutenant, if (1) the promotions from such procedure have
no adverse impact against African-Americans or Hispanic-
Americans or (2), if the use of the procedure in question
results in adverse impact against either African-Americans
or Hispanics, if the procedure in question meets the
requirements of the Uniform Guidelines on Employee
Selection Procedures, 29 C.F.R. Part 1607 (2000), is job-
related and consistent with business necessity, and if all
reasonable steps consistent with job-relatedness and business
necessity have been taken to eliminate such adverse impact
or reduce it to the greatest extent practicable.
26a
Appendix B
a. Subject to the other provisions of this paragraph,
the Court notes that the plaintiffs have expressly approved
the concept of the current Meet and Confer Agreement
between the City and the Houston Police Officers’ Union
as the bargaining representative for officers of the Police
Department on this subject, which was approved by an
overwhelming vote of officers, combining a written test of
some type and a properly conducted assessment center
of some type, and using the test to determine which
promotional candidates will be processed in the assessment
center. While counsel for plaintiffs and the plaintiffs’ expert
may make suggestions as to the development and use of
such procedures, they are not required to attend any of the
negotiations between the City and bargaining agent over a
new Meet and Confer agreement.
b. The defendant shall keep counsel for plaintiffs, as
well as other affected employee groups or their counsel,
informed of the details of its progress in considering,
evaluating, and adopting such procedures, and before any
such procedure is put into effect the defendant shall so report
to the Court and shall obtain and make available to counsel
for plaintiffs all information regarding the likelihood that
making promotions under the new procedure will result in
adverse impact against African-Americans or Hispanic-
Americans, and all information regarding the job-
relatedness of the procedure and its consistency with
business necessity.
38. Plaintiffs may make comments and suggestions to the
defendant regarding such an alternative selection procedure,
through their counsel or by themselves, but shall not be bound
27a
Appendix B
by any such comment or suggestion, and shall remain free to
challenge the new procedure herein in the event that it has
adverse impact against African-Americans or Hispanic-
Americans in practice, unless plaintiffs and the defendant
enter into a formal agreement and stipulation to that effect,
which is submitted to the Court for approval and which is
approved by the Court. Other affected employee groups may
also make comments and suggestions.
H. Attorneys’ Fees and Costs
39. (a) The City shall pay the reasonable attorneys’ fees,
costs and expenses, in an amount to be negotiated by the
parties or determined by the Court, for all services performed
by the Lawyers’ Committee for Civil Rights Under Law, by
co-counsel, and by all attorneys and paralegals employed by
them, in this case and in connection with plaintiffs’ effort to
intervene in the Comeaux and Kelley litigation, and forall
other reasonable services performed and expenses incurred
through the conclusion of all judicial proceedings on the grant
of final approval to the Consent Decree and/or the propriety
of the relief provided, including the defense of such relief
against collateral attack. The parties shall attempt to negotiate
in good faith over the amount of such recovery.
(b) The procedure for resolving fee claims as
between the plaintiffs and the City is that counsel for
plaintiffs shall informally submit to the defendant their
claim for attorneys’ fees and expenses, with supporting
breakdowns. Within thirty days after the receipt of this
submission, or within such further time as the plaintiffs and
the defendant may informally agree, the defendant shall
28a
Appendix B
provide its response, which shall include (1) an itemized list
of each of its objections to the submission, if any, indicating
the particular items to which it objects and the reasons for
the objection; (2) its proposed resolution of each objection;
and (3) the amount which it does not contest, and contends
would be a reasonable amount for the award of attorneys’
fees and expenses herein. The defendant shall forthwith pay
the uncontested amount. In the event that the amount of fees
and expenses is litigated and plaintiffs recover an amount
additional to the uncontested amount, plaintiffs shall also
recover prejudgment interest on the amount of their recovery,
at the rate of 10% per annum, from the thirtieth day after the
receipt of plaintiffs’ submission until the date of payment of
the amount recovered.
40. (a) The City shall pay the reasonable attorneys’ fees
up to a maximum of $ 20,000 annually, plus costs and
expenses (including reasonable expert fees in the amounts
actually charged to counsel for plaintiffs), of plaintiffs for
future services in monitoring the defendants’ compliance with
this Consent Decree. Plaintiffs shall have the right to seek
relief from this limit in the discretion of the Court, in the
event of a substantial failure by the City to comply with the
provisions of this Consent Decree or a substantial effort
required in connection with a proposed alternative selection
procedure.
(b) The procedure and mechanism described in
4 39(b) shall be applied to awards of attorneys’ fees under
{ 40(a) for monitoring compliance with the Consent Decree.
Plaintiffs shall decide when to submit their statements.
29a
Appendix B
I. Reporting and Record-Keeping
41. At quarterly intervals after the effective date of this
Consent Decree, the City shall report to plaintiffs on the steps
it has taken to fulfill the provisions of this Consent Decree,
including the name of each promotee and the date of each
promotion to the ranks of Sergeant and Lieutenant.
42. With respect to each examination for Sergeant and
Lieutenant within the ten-year period after the grant of final
approval of this Consent Decree, promptly after the
completion of any proceeding on the protest of the scoring
of any item under the City’s normal civil service procedures
or any alternative procedures, the City shall provide counsel
for plaintiffs with the following information:
a. a computer-readable data file containing the
name, race, seniority points, score on each part of the
promotional procedure, and scored answers to each
examination item (correct or incorrect);
b. the final rank-ordered register of persons passing
the test, showing the final written score, any assessment
center score, seniority points, total score, rank (if any),
and race of each test-taker. Persons failing the test shall
be listed alphabetically.
c. Plaintiffs may make reasonable requests of the
City for further information which would be helpful in
determining any question of compliance with the Consent
Decree, or which would assist in achieving its goals.
30a
Appendix B
J. Retention of Jurisdiction, and Duration of the
Consent Decree
43. This case shall be dismissed as settled, subject to
the parties’ rights of appeal and to any remand proceedings
required by a superior court. In the event of any subsequent
dispute among the parties as to any of the matters resolved
by this Consent Decree, this case shall upon motion be
automatically reinstated on the docket of this Court as an
active case, until such matter shall have been resolved.
44. The provisions of the Consent Decree shall be in
force for ten years after the conclusion of all judicial
proceedings on the grant of final approval to the Consent
Decree and/or the propriety of the relief provided. Those
portions of the Consent Decree that allow a deviation from
the requirements of The Fire and Police Civil Service Act,
Texas Local Government Code chapter 143, as amended, may
continue to be utilized for a maximum period of ten years
unless extended on motion of any party for good cause shown,
provided that the expiration of the ten-year period shall not
limit the life of a promotional register under { 35(a) above if
the examination leading to that register was administered
during the ten-year period. The period for retention of
jurisdiction and/or continuation of these Consent Decree
provisions may in the discretion of the Court be extended on
motion by any party for good cause shown.
K. The Former Log-Linear Analysis
45. The form of proposed Consent Decree originally
presented to the Court and given preliminary approval
3la
Appendix B
contained provisions for a log-linear analysis of test results,
to identify biased test questions. All parties agreed to end
the use of log-linear analysis and to substitute for that form
of relief an additional six months’ life, to a total of thirty
months, for promotional registers for the ranks of Sergeant
and Lieutenant. The Court approved a Consent Order to that
effect on November 24, 1997.
46. Only three tests were subject to log-linear analysis:
the 1993 Sergeant examination, the 1994 Lieutenant
examination, and the 1995 Sergeant examination.
a. The 1993 Sergeant examination resulted in 94
promotions, and all of the promotees passed the original
version of the test prior to log-linear analysis.
b. The 1994 Lieutenant examination resulted in 12
promotions, and all of the promotees passed the original
version of the test prior to log-linear analysis.
c. The 1995 Sergeant examination resulted in 57
promotions, and all of the promotees passed the original
version of the test without log-linear analysis.
47. There shall be no changes in past promotions or rank
seniority as a result of the previous use of log-linear analysis.
L. Interim Promotions
48. Pursuant to the Interim Order of February 5, 1997,
the parties estimate that approximately 120 acting promotions
have been made from promotional registers, and another 37
32a
Appendix B
acting remedial promotions have been made. These promo-
tions shall be made permanent, and the promotees shall be
considered in all respects as if they had received permanent
regular promotions, or permanent remedial promotions, except
for purposes of retrospective pension contributions. The City
. shall not be responsible for any additional payments to fund
the promotees’ pension benefits for this period, if any.
49. In the event that the grant of final approval to the
Consent Decree is reversed on appeal, all promotions based
on promotional registers that complied with the provisions
of the Consent Decree, but did not comply with the provisions
of State law, and all remedial promotions under this Consent
Decree, shall be undone, and all such promotees shall be
demoted to their prior ranks, regardless of whether the
promotions were permanent or acting, unless all parties
present an agreed proposal for a different course of action.
N. Pending motions
50. All outstanding motions are denied as moot.
M. Final Grant of Approval
51. The Court grants final approval to this modified form
of the Consent Decree.
Signed at Houston, Texas, March 30, 2001.
s/ Lynn Hughes
LYNN HUGHES
United States District Judge
33a
Appendix B
WE SO STIPULATE, AND WE ASK FOR THIS:
s/ Richard T. Seymour
RICHARD T. SEYMOUR
Teresa A. Ferrante
Lawyers’ Committee for Civil Rights Under Law
1450 G Street N.W., Suite 400
Washington, D.C. 20005
(202) 662-8350
Fax: 202-783-5131
JO KATHERINE SCHNEIDER D’AMBROSIO
Sroufe, Payne & Lundeen
1700 West Loop South, Suite 1230
Houston, Texas 77027
(713) 840-8008
Fax (713) 840-8088
Attomeys for Plaintiffs and for the class they represent
s/ John E. Fisher
JOHN E. FISHER
Senior Assistant City Attorney
Attomey-in-Charge
Admissions ID No. 5774
City of Houston
Legal Department
P.O. Box 1562
Houston, Texas 77251
(713) 247-2059
Attorney for Defendant City of Houston
34a
APPENDIX C — FINDINGS OF FACT ON APPROVAL
OF CONSENT DECREE FROM THE UNITED STATES
DISTRICT COURT FOR THE SOUTHERN DISTRICT
OF TEXAS DATED SEPTEMBER 12, 2000
UNITED STATES DISTRICT COURT
SOUTHERN DISTRICT OF TEXAS
CIVIL ACTION H-92-2510
DOROTHY A. EDWARDS. et al.
Plaintiffs,
versus
CITY OF HOUSTON,
Defendant.
FINDINGS OF FACT ON APPROVAL
OF CONSENT DECREE
1. African-American and Hispanic-American police
officers sued the City of Houston because the City’s
promotional examinations for Sergeant and Lieutenant
discriminated against them. The dispute between the
plaintiffs and defendants was genuine and vigorously
contested.
tN
The plaintiffs and defendants settled their dispute.
In January 1993, the parties submitted a proposed consent
decree. The court gave notice of the proposed settlement
to all interested persons and entities and scheduled a
fairness hearing on March 24, 1993.
Rey IERIE SOY BIE PY
3.
35a
Appendix C
After the fairness hearing was scheduled, various police
unions sought to intervene in the case to object to the
proposed consent decree. The court denied the unions’
request to intervene but permitted them to participate in
the fairness hearing by (a) cross-examining witnesses;
(b) proffering evidence, and (c) raising objections to the
consent decree.
The unions were permitted to intervene after their appeal
to the United States Court of Appeals for the Fifth
Circuit. The case was remanded to permit the unions “to
intervene in this suit, to allow them appropriate
discovery, and to conduct a new fairness hearing.”
Edwards v. City of Houston, 78 F.3d 983, 1006 (Sth Cir.
1996) (en banc). The unions were permitted to intervene
to conduct discovery and to object to the settlement, not
to assert new claims for affirmative relief or to assert
veto power over the settlement.
The court scheduled a second fairness hearing, which
the unions expanded into a three-day trial. The unions
presented evidence in support of their objection to the
consent decree. The plaintiffs and defendants presented
evidence in support of the settlement.
Pro se plaintiffs Willie Fields and Bennie Green did not
appear at the fairness hearing and their claims are
dismissed.
The unions’s experts were completely unpersuasive.
Dr. Antel’s report did not meet even minimum standards
required for expert testimony. Dr. Landy’s written report
and expanded testimony at trial assumed that Dr. Rathjen
10.
36a
Appendix C
accurately described her work to develop the challenged
tests, an assumption shown to be wrong. Dr. Landy
persisted in mischaracterizing Dr. Lundquist’s analysis
to the point of losing any appearance of objectivity.
The experts testifying in support of the settlement were
well-qualified and credible. Dr. Mann conducted a
thorough review of the City’s documentation and
determined that the examinations adversely impacted the
plaintiffs under the 80% rule or statistical significance,
or both. His “odds ratio” test was appropriate for the
aggregate pools. Dr. Friedland testified credibly that the
police department tests were structured like the fire
department tests. Both tests were developed by
Dr. Rathjen. The fire department tests were previously
determined in other litigation not to be job related.
Dr. Lundquist and Dr. Friedland are highly qualified and
presented credible testimony about the flaws in
Dr. Rathjen’s development of the police department tests.
The promotional examinations for Sergeant and
Lieutenant had a discriminatory impact on African-
Americans and Hispanic-Americans. The promotional
examinations were not job-related.
The City created 96 new Sergeant positions and 10 new
Lieutenant positions for remedial promotions under the
consent decree. The remedial promotions are precisely
tailored to the injury caused to Plaintiffs by the
promotional examinations. The remedial promotees were
correctly chosen as the individuals most likely to have
been specific victims.
11.
13.
14.
‘.
37a
Appendix C
No remedial promotion delayed the promotion of any
other police officer. The only delays were caused by the
unions. All police officers will benefit from the future
opportunity to compete for these promotional positions
as they become vacant.
. The use of log-linear analysis was a proper incremental
remedy. The promotions have been made and the
program has been successfully completed. Although the
plaintiffs and defendants agree to replace the log-linear
analysis with an extension of the life of promotional
registers from 24 to 30 months, there is no basis to alter
the promotions made from the three test registers
compiled under the log-linear analysis: the 1993 and
1995 Sergeant examinations and the 1994 Lieutenant
examination. Any police officer who was denied a
promotion under the consent decree was denied only the
opportunity to compete for a position created by the
consent decree.
The relief in the amended consent decree is necessary to
remedy the harm shown.
The settlement is fair and appropriate. The court approves
the. consent decree as previously modified in the final
judgment entered March 25, 1993, and in the consent
order entered November 25, 1997.
The Court adopts its earlier findings of fact and
conclusions of law, entered March 25, 1993.
38a
Appendix C
Signed September 12, 2000, at Houston, Texas.
s/ Lynn N. Hughes
Lynn N. Hughes
United States District Court
39a
APPENDIX D — FINDINGS OF FACT AND CONCLU-
: SIONS OF LAW OF THE UNITED STATES DISTRICT
' COURT FOR THE SOUTHERN DISTRICT OF TEXAS,
HOUSTON DIVISION, DATED MARCH 25, 1993
IN THE UNITED STATES DISTRICT COURT FOR
THE SOUTHERN DISTRICT OF TEXAS
HOUSTON DIVISION
C. A. NO. H-92-2510
Judge Lynn Hughes
DOROTHY A. EDWARDS, et ai.,
individually and on behalf of all others similarly situated,
Plaintiffs,
v.
CITY OF HOUSTON,
Defendant.
FINDINGS OF FACT
and
CONCLUSIONS OF LAW
I. Findings of Fact
A. The Proceedings in This Case
1. The original Complaint in this action was filed
on August 19, 1992, under Title VII of the Civil Rights Act
of 1964, 42 U.S.C. §§ 2000e et seg., as amended by the Civil
40a
Appendix D
Rights Act of 1991 and the Equal Employment Opportunity
Act of 1972 [“Title VII’’].' The plaintiffs in this action had
filed timely charges of racial discrimination with the U.S.
Equal Employment Opportunity Commission, alleging that
the City’s promotional examinations for the ranks of Sergeant
and Lieutenant in the Houston Police Department
discriminated against African-Americans and Hispanic-
Americans, on the following dates:?
Name of Plaintiff Date Charge Filed with EEOC
Sergeant Dorothy A. Edwards August 29, 1991
African-American Police
Officers’ League August 29, 1991
Houston Police Organization
of Spanish Speaking Officers March 26, 1992
Officer Lionel Aaron March 26, 1992
Detective Bennie Conway August 29, 1991
Officer Clement Boyd Crosby March 26, 1992
Officer Jose A. Garcia March 26, 1992
Officer Richard C. Garcia March 26, 1992
Officer Maria L. Guillory March 26, 1992
Sergeant Anthony R. Jammer August 29, 1991
Sergeant Charles A. McClelland August 29, 1991
Officer Silas Montgomery, Jr. March 26, 1992
Officer Clyde Philpott March 26, 1992
Officer Carl Wayne Reed March 26, 1992
Sergeant Richard M. Spencer August 29, 1991
Sergeant Bruce D. Williams March 26, 1992
1. Docket entries.
2. Copies of the charges and the Notices of Right to Sue are
attached to the Complaint in this action as Exhibits B and C.
4la
Appendix D
2. On November 11, 1991, Richard T. Seymour,
attorney-in-charge for plaintiffs, wrote to the City to notify
the City directly that he and Ms. Acosta represented the
charging parties, requested information on the questions of
the adverse impact of the challenged tests and their job-
relatedness, and suggested that the charging parties and the
respondent City use the time while the matter was before the
EEOC to gather the necessary information and to see if they
could settle the matter. The City provided some of the requested
information, but was not interested at that time in pursuing
settlement.’
3. Efforts to obtain redress for the Police Department’s
allegedly discriminatory tests began in 1975 and 1976, when
Kelley v. Hofheinz, Civil Action No. H-75-1536, and Comeaux
v. Cite of Houston, Civil Action No. 76-H-1754, were filed.
Among other claims, these cases raised the claim that the
promotional examinations of the Houston Police Department
discriminated against A frican-American police officers because
of their race, in violation of Title VII of the Civil Rights Act
of 1964, as amended. These cases’ challenges to the promo-
tional tests were based upon 1975 and 1976 EEOC charges
of racial discrimination in the promotional tests. The Comeaux
action was consolidated into Kelley. The consolidated cases
are hereinafter referred to collectively as Kelley.
4. Ten years ago, in 1983, there were unsuccessful
settlement discussions between the Kelley plaintiffs and
the City.*
3. Plaintiffs’ exhibit 17, Affidavit of Mr. Seymour, § 2.
4. Defendant City of Houston’s Motion to Dismiss for Want of
Prosecution, filed April 17, 1992 in Comeaux, at 2.
42a
Appendix D
5. On April 16, 1992, the City refused to consent to
the intervention in Kelley of the Afro-American Police
Officers League, the Houston Police Organization of Spanish
Speaking Officers, and a group of African-American and
Hispanic-American police officers.°
6. On April 17, 1992, the Afro-American Police
Officers League, the Houston Police Organization of Spanish
Speaking Officers, and a group of African-American and
Hispanic-American police officers moved for leave to intervene
in Kelley, alleging that they had been harmed by racially
discriminatory promotional examinations for the ranks of
Sergeant and Lieutenant in the Houston Police Department,
that the disposition of Kelley could impair their interests,
and that in light of the passage of time their interests were
not being represented effectively in Kelley.°
7. On the same day, the defendant City moved
to dismiss Kelley for want of prosecution.’ Its Motion recited
at p. 3 that counsel for the city had on April 16, 1992 “conferred
by telephone with Richard Seymour of the Lawyers’
Committee for Civil Rights Under Law, who had previously
indicated he represented several individuals who intended
to intervene in this litigation”, and that the motion was
opposed.
5. Certificate of Conference to Motion of Applicants for
Intervention for Leave to Intervene as Plaintiffs-Intervenors, filed
herein on April 17, 1992, at 6.
6. Complaint and docket entries.
7. Docket entries.
Be SPENSER ARLES LOL ORDO, LISS IE HR ss i ai
43a
Appendix D
8. Prior to a hearing on the motions, counsel for the
applicants for intervention unsuccessfully tried to interest
the City in settlement discussions. The applicants suggested
that, if meaningful settlement discussions could begin, the
Court could be notified of that fact and asked to defer ruling
on applicants’ motion for leave to intervene and the City’s
motion to dismiss for a reasonable time, to allow settlement
discussions to proceed. The City was not interested.°
9. Ahearing was held on June 15, 1992, on the City’s
Motion to Dismiss Kelley, and on the motion for leave to
intervene. At the hearing, the City argued that Kelley should
be dismissed, that the intervention should be denied, that
if the applicants for intervention wanted to challenge
promotional tests they should do so in a new case, that any
such challenge should be limited to the 1991 tests as to which
the applicants for intervention had filed their EEOC charges,
and that in any event if the applicants were allowed to
challenge any earlier tests Hispanic-Americans should not
be included in any such challenge. In sum, the City did
everything possible to bar or limit the relief sought by
plaintiffs. After the Court expressed reluctance to allow the
continued maintenance of testing claims going back to 1975,
the applicants for intervention proposed a compromise under
which they would only assert testing claims as to tests given
after January 1, 1982. The City refused to enter into such a
compromise.
10. The Court accepted the City’s arguments in part,
and rejected them in part. The Order of June 17, 1992 in
Kelley dismissed all claims in Kelley for want of prosecution
8. Plaintiffs’ exhibit 17, Affidavit of Mr. Seymour, § 3.
44a
Appendix D
except for test-promotion related claims after January 1,
1982, denied the application for leave to intervene, ordered
the applicants for intervention to file a new lawsuit to be
transferred to the same Court, directed that the remainder of
Kelley be consolidated into the new lawsuit, and ordered that
the new class consist of blacks and Hispanics. This case was
timely filed on August 19, 1992, after the receipt of Notices
of Right to Sue issued by the Attorney General of the United
States, and the remnant of Kelley has been consolidated into
this action.
11. In this action, plaintiffs have challenged as
discriminatory the promotional examinations for the rank of
Lieutenant and for the rank of Sergeant in the Houston Police
Department [““HPD”’] which were administered during the
period from 1982 to date and the system pursuant to which
these examinations are prepared. Plaintiffs alleged in this
lawsuit that the challenged examinations had the effect
of disproportionately excluding African-Americans and
Hispanic-Americans from promotion to Sergeant from 1982
to date, and of disproportionately excluding African-
Americans from promotion to Lieutenant from 1982 to date,
but were not job-related or corisistent with business necessity.
12. The plaintiffs sued on their own behalf, on behalf
of the African-American and Hispanic-American members
of the Police Department who took a Sergeant examination
from 1982 to date or who will compete for promotions to
Sergeant in the future, and on behalf of African-American
members of the Police Department who took a Lieutenant
examination from 1982 to date or who will compete for
promotions to Sergeant in the future.
45a
Appendix D
13. Plaintiffs’ proposed class did not include
Hispanic Americans who took Lieutenant examinations from
1982 to date, because the records of the City show that the
Lieutenant examinations did not operate to exclude Hispanic-
Americans from promotion to Lieutenant during this period
of time. However, plaintiffs alleged that the Sergeant
examination discriminatorily delayed the promotions of both
African-Americans and Hispanic-Americans to the rank of
Sergeant, and a delay in their promotion to the rank of
Sergeant is necessarily a delay in their ability to compete for
promotion to Lieutenant.
14. Plaintiffs have obtained substantial discovery
from the City, pursuant to both formal and informal requests.
The discovery included Registers of Eligibles for the period
from January 1, 1982 through 1992 on which the City
provided racial identifications for the promotional candidates
for whom it could identify race, vacancy announcements,
and other materials. Settlement discussions began in earnest
in the Fall of 1992, after several boxes of materials had been
provided to counsel for plaintiffs. Further responses to
discovery were suspended pending the outcome of the
negotiations.
B. Certification of the Class, and Finding of Jurisdiction
15. Based upon the stipulations of the parties and
the information of record, the Court certifies and defines the
following class:
a. All African-Americans who are employed, or at
any time since January 1, 1982 were employed, as Class
46a
Appendix D
A peace officers by HPD and who took a promotional
examination for the rank of Lieutenant or for the rank
of Sergeant which was administered at any time from
January 1, 1982 to the present, and those who will
compete for such promotions in the future; and
b. All Hispanic-Americans’ who are employed,
or at any time since January 1, 1982 were employed,
as Class A peace officers by HPD and who took a
promotional examination for the rank of Sergeant which
was administered at any time from January 1, 1982
to the present, and those who will compete for such
promotions in the future.
16. This Court has jurisdiction over the claims
alleged herein by virtue of 28 U.S.C. §§ 1331 and 1343
and § 706(f) of the Civil Rights Act of 1964, 42 U.S.C.
§ 2000e-5(f). All of the administrative and procedural
requirements to suit under Title VII have been satisfied.
C. The Rank Structure Within the Police Department
17. The entry-level uniformed position within the
Houston Police Department is Police Officer. The first-level
promotional position is currently the rank of Sergeant. Police
Officers with two years’ service as Police Officers are allowed
to compete for the rank of Sergeant.
18. At the time Kelley was filed, there were two
promotional positions for which entry-level Police Officers
could compete: Detective and Sergeant. The Detective rank
has since been merged with the Sergeant rank. No separate
examination for Detective has been given for several years.
47a
Appendix D
19. The second-level promotional position within
the Houston Police Department is the rank of Lieutenant.
Sergeants with two years’ service as Sergeants are allowed
to compete for the rank of Lieutenant.
20. There are higher ranks within the Houston Police
Department, but this case does not present any independent
claim of discrimination in promotion to such ranks because
of the selection procedures used for such ranks.
D. The Disparate Impact of the C. hallenged Examinations
1. Racial Disparities in Selection Rates
21. Sec. 4(D) of the Uniform Guidelines on Employee
Selection Procedures, 43 Fed.Reg. 38297 (1978), 29 C.F.R.
§ 1607.4(D), defines “adverse impact” as either a selection rate
for the racial or ethnic group in question which is less than
four-fifths (4/5) of the selection rate for the best-performing
group, or a difference in selection rates which is both Statistically
and practically significant. The tables below use the term
“standard deviation”. The number of standard deviations is
a measure of how unusual it would be for a difference of
at least this magnitude to occur by chance. Social scientists
commonly use a .05 significance level (probability) — or,
equivalently, one time in twenty — as sufficient to rule out
chance. This corresponds to 1.96 standard deviations.
The larger the number of standard deviations, the less likely
it would be to get the observed result by chance.’ It is difficult
9. For example, a .01 level of probability — one time in a
hundred — corresponds to approximately 2.54 standard deviations.
Three standard deviations corresponds to a .0027 level of probability,
or 27 times in ten thousand. Four standard deviations corresponds to
a .000063 level of probability, or sixty-three times in a million.
48a
Appendix D
for a test of statistical significance to detect that a modest
difference is not due to chance when the size of the sample
is small. For this reason, it is useful to aggregate statistical
information over a period of years.
22. Paragraphs 23-42 below were stipulated
as correct by the parties in the proposed Consent Decree.
Various objectors have challenged the relevancy of these
calculations, but no objector has presented evidence drawing
these stipulations into question. The Court therefore accepts
the stipulations as correct. The objections as to relevancy
are discussed below.
23. The challenged examinations for Sergeant have
had adverse impact against both African-Americans and
Hispanic-Americans. Moreover, African-American and
Hispanic-American officers who have been promoted to
Sergeant have because of these examinations had to wait for
promotion substantially longer than non-Hispanic whites
have had to wait. The tables in the following paragraphs
show the detailed information. The term “shortfall” used in
the tables is the difference between the number of promotions
actually received by persons of a particular racial or ethnic
group and the number they would have received if they had
been promoted at the same rate at which whites were promoted.
24. In the September 23, 1982 Sergeant Examination,
the rate of promotion for African-Americans was only 42.9%
of the rate of promotion for whites, resulting in a shortfall of
5 African-American promotions. There was no adverse
impact against Hispanic-American officers on this test.
*
£
x
A
‘
:
x
q
4 ‘
$
Q
>
8
;
%
ik,
49a
Appendix D
Table 1: September 23, 1982 Sergeant Examination
Other Whites Blacks Hispanics
Test-Takers: 0 85 33 14
Promoted: 0 30 5 5
% Promoted: N.A. 35.3% 15.2% 35.7%
Promotion Rate as % of White Rate: 42.9% N.A.
Availability of Blacks, Hispanics: 25.0% 10.6%
Total Promotions Among All Groups: 40 40
Actual Promotions of Blacks and
Hispanics 3 5
Expected Promotions of Blacks
and Hispanics 4.2
—Difference Between
Actual and Expected
Promotions (“Shortfall”): -5.0 0.8
25. In the November 17, 1983 Sergeant Examination,
the rate of promotion for African-Americans was only 34.2%
of the rate of promotion for whites, resulting in a shortfall of
8 African-American promotions. The rate of promotion for
Hispanic-Americans was only 28.5% of the rate of promotion
for whites, resulting in a shortfall of 7 Hispanic-American
promotions. These differences were also Statistically signi-
ficant at the .05 level, meaning that there is less than one
chance in twenty that such a difference, or a larger difference,
could have occurred by chance.
OEE
See Re
50a
_ Appendix D
Table 2: November 17, 1983 Sergeant Examination
Other Whites Blacks Hispanics
Test-Takers: 2 295 75 60
Promoted: l 69 6 4
% Promoted: 50.0% 23.4% 8.0% 6.7%
Promotion Rate as % of White Rate: 34.2% 28.5%
Availability of Blacks, Hispanics: 17.4% 13.9%
Total Promotions Among All Groups: 80 80
Actual Promotions of Blacks and
Hispanics 6 4
Expected Promotions of Blacks
and Hispanics 13.9 11.1
—Difference Between
Actual and Expected
Promotions (“Shortfall’’): -7.9 -7.1
Standard Deviation: 3.4 aoe
Number of Standard Deviations
Between Expected
and Actual Promotions -2.329 -2.299
26. In the January 31, 1985 Sergeant Examination,
the rate of promotion for African-Americans was only 38%
of the rate of promotion for whites, resulting in a shortfall of
11 African-American promotions. The rate of promotion for
Hispanic-Americans was only 42.2% of the rate of promotion
for whites, resulting in a shortfall of 7 Hispanic-American
promotions. The difference between white and African-
Sla
Appendix D
American promotion rates was statistically significant for
African-Americans at the .01 level, meaning that there is
less than one chance in a hundred that such a difference,
or a larger difference, could have occurred by chance. The
difference between white and Hispanic-American promotion
rates was statistically significant for Hispanic-Americans at
the .05 level, meaning that there is less than one chance
in twenty that such a difference, ora larger difference, could
have occurred by chance.
Table 3: January 31, 1985 Sergeant Examination
Other Whites
Blacks Hispanics
Test-Takers: 5 408
Promoted: ] 11]
% Promoted: 20.0% 27.2%
Promotion Rate as % of White Rate:
Availability of Blacks, Hispanics:
Total Promotions Among All Groups:
Actual Promotions of Blacks and
Hispanics
Expected Promotions of Blacks
and Hispanics
—Difference Between
Actual and Expected
Promotions (“Shortfall”):
Standard Deviation:
Number of Standard Deviations
Between Expected
and Actual Promotions
87 61
9
10.3%
38.0%
15.5%
128
9
52a
Appendix D
27. Inthe July 17, 1986 Sergeant Examination, none
of the 79 African-American test-takers was promoted, so the
rate of promotion for African-Americans was 0% of the rate
of promotion for whites, resulting in a shortfall of 6 African-
American promotions. The rate of promotion for Hispanic-
Americans was only 36.7% of the rate of promotion for
whites, resulting in a shortfall of 2 Hispanic-American
promotions. The difference between white and African-
American promotion rates was statistically significant for
African-Americans at the .01 level, meaning that there is less
than one chance in a hundred that such a difference, or a
larger difference, could have occurred by chance.
Table 4: July 17, 1986 Sergeant Examination
Other Whites Blacks Hispanics
Test-Takers: l 352 79 58
Promoted: 0 33 0 2
% Promoted: 0.0% 94% 0.0% 3.4%
Promotion Rate as % of White Rate: 0.0% 36.7%
Availability of Blacks, Hispanics: 16.1% 11.8%
Total Promotions Among All Groups: 35 35
Actual Promotions of Blacks and
Hispanics 0 2
Expected Promotions of Blacks
and Hispanics 5.6 4.1
—Difference Between
Actual and Expected
Promotions (“Shortfall”): -5.6 -2.1
53a
Appendix D
Standard Deviation: 1.9
Number of Standard Deviations
Between Expected
and Actual Promotions -2.594 -1.121
28. In the October 29, 1987 Sergeant Examination,
the rate of promotion for African-Americans was only 41.2%
of the rate of promotion for whites, resulting in a shortfall of
7 African-American promotions. The rate of promotion for
Hispanic-Americans was only 31.3% of the rate of promotion
for whites, resulting in a shortfall of more than 6 Hispanic-
American promotions. These differences were also Statistically
significant at the .05 level, meaning that there is less than one
chance in twenty that such a difference, or a larger difference,
could have occurred by chance.
Table 5: October 29, 1987 Sergeant Examination
Other Whites Blacks Hispanics
Test-Takers: 3 362 97 73
Promoted: 1 63 7 4
% Promoted: 33.3% 17.4% 7.2% 5.5%
Promotion Rate as % of White Rate: 41.5% 31.5%
Availability of Blacks, Hispanics: 18.1% 13.6%
Total Promotions Among All Groups: 75 75
Actual Promotions of Blacks and
Hispanics 7 4
Expected Promotions of Blacks
and Hispanics
54a
Appendix D
—Difference Between
Actual and Expected
Promotions (‘Shortfall’): -6.6 -6.2
Standard Deviation: 3.3 3.0
Number of Standard Deviations
Between Expected
and Actual Promotions -1.978 -2.097
29. In the April 27, 1989 Sergeant Examination,
the rate of promotion for African-Americans was only 20.8%
of the rate of promotion for whites, resulting in a shortfall of
11 African-American promotions. The rate of promotion for
Hispanic-Americans was only 74% of the rate of promotion
for whites, resulting in a shortfall of 1 Hispanic-American
promotion. The difference between white and African-
American promotion rates was statistically significant for
African-Americans at the .01 level, meaning that there is less
than one chance in a hundred that such a difference, or a
larger difference, could have occurred by chance.
Table 6: April 27, 1989 Sergeant Examination
Other Whites Blacks Hispanics
Test-Takers: 7 410 129 109
Promoted: l 61 4 12
% Promoted: 14.3% 14.9% 3.1% 11.0%
Promotion Rate as % of White Rate: 20.8% 74.0%
Availability of Blacks, Hispanics: 19.7% 16.6%
Total Promotions Among All Groups: 78 78
PR TLRS Bah 2 OF, See mer se, aha Yn a ee tt RAIS a 6
Arg ae Ee ey RE
55a
Appendix D
Actual Promotions of Blacks and
Hispanics 4 12
Expected Promotions of Blacks
and Hispanics 15.4 13.0
—Difference Between
Actual and Expected
Promotions (“Shortfall’’): -11.4 -1.0
Standard Deviation: 3.5 a0
Number of Standard Deviations
Between Expected
and Actual Promotions -3.235 -0.298
30. In the October 31, 1991 Sergeant Examination,
the rate of promotion for African-Americans was only 27.8%
of the rate of promotion for whites, resulting in a shortfall of
17 African-American promotions. The rate of promotion for
Hispanic-Americans was only 31.8% of the rate of promotion
for whites, resulting in a shortfall of 10 Hispanic-American
promotions. Each of these differences was also statistically
Significant at the .01 level, meaning that there is less than
one chance in a hundred that such a difference, or a larger
difference, could have occurred by chance.
Table 7: October 31, 1991 Sergeant Examination
Other Whites Blacks Hispanics
Test-Takers: 7 415 175 111
Promoted: 0 94 11 8
% Promoted: 0.0% 22.7% 6.3% 7.2%
Promotion Rate as % of White Rate: 27.8% 31.8%
a aa a RD TR LAS TR
56a
Appendix D
Availability of Blacks, Hispanics: 24.7% 15.7%
Total Promotions Among All Groups: 113 113
Actual Promotions of Blacks and
Hispanics 1] 8
Expected Promotions of Blacks
and Hispanics 27.9 i a
—Difference Between
Actual and Expected
Promotions (“Shortfall’’): -16.9 -9.7
Standard Deviation: 4.6 3.9
Number of Standard Deviations
Between Expected
and Actual Promotions -3.692 -2.514
31. In the aggregate, the rate of promotion for
African-Americans in the Sergeant Examinations from
1982 through 1992 was only 31.4% of the rate of promotion
for whites, resulting in a shortfall of 64 African-American
promotions. The rate of promotion for Hispanic-Americans
was only 34% of the rate of promotion for whites, resulting
in a shortfall of 34 Hispanic-American promotions.
Table 8. 1982-1991 Aggregate Sergeant Examinations
Other Whites Blacks Hispanics
Test-Takers: 25 2,326 675 486
Promoted: 4 461 42 42
% Promoted: 16.0% 19.8% 6.2% 8.6%
Promotion Rate as % of White Rate: 31.4% 43.6%
Saba alana Geocache Miseacekalaal it sahiiessuategl eRe
S7a
Appendix D
Availability of Blacks, Hispanics:
Total Promotions Among All Groups:
Actual Promotions of Blacks and
Hispanics 42 42
Expected Promotions of Blacks
and Hispanics 105.5 76.0
—Difference Between
Actual and Expected
Promotions (“Shortfall”): -63.5 -34.0
Standard Deviation: 9.2 8.1
Number of Standard Deviations
Between Expected
and Actual Promotions -6.880 -4.199
However, when adding the shortfall figures from the
individual examinations, the shortfall in African-American
promotions to Sergeant from 1982 through 1992 was
62 promotions, rather than 64, and the shortfall in Hispanic-
American promotions to Sergeant from 1982 through 1992
remained 34 promotions. For the purposes of settlement,
the lower figure of 62 African-American promotions, based
on the individual examinations, shall be used. The promotion
rate for Asian-Americans, Philippine Americans and Others
was 80.8% of the promotion rate for whites, a figure falling
just short of the cutoff for determining adverse impact under
the “4/5” rule. One additional promotion would have brought
their promotion rate to 20%, which is higher than the white
rate of promotions.
32. The challenged examinations for the rank
of Lieutenant have had adverse impact against African-
Americans.
58a
Appendix D
33. Many fewer persons took the individual exami-
nations for promotion to Lieutenant than did so for promotion
to Sergeant. Rather than set forth the test-by-test results for
each test, the following paragraphs show information for only
those tests with adverse impact against African-Americans
under the “4/5 rule” and a shortfall of at least one black.
34. In the September 27; 1984 Lieutenant Exam-
ination, the rate of promotion for African-Americans was
only 36.6% of the rate of promotion for whites, resulting in
a shortfall of 1.4 African-American promotions.
Table 9. September 27, 1984 Sergeant Examination
Other Whites Blacks Hispanics
Test-Takers: l 123 13 13
Promoted: 0 28 2 |
% Promoted: 0.0% 22.8% 15.4% 7.7%
Promotion Rate as % of White Rate: 67.6%
Availability of Blacks: 8.7%
Total Promotions: 31
Observed Promotions of Blacks: 2
Expected Promotions of Blacks: 2.7
—Difference Between
Observed and Expected
Promotions (“Shortfall”): -0.7
LESS ERED TLL ITS ROPER RELI te 5 NAA ROEM OME ep)
TRNE MS O18
:
S,
z
a
F
&
a
Ps
|
=
2.
=
&
4
?
k
.
2
59a
Appendix D
35. In the March 3, 1988 Lieutenant Examination,
the rate of promotion for African-Americans was 0% of the
rate of promotion for whites, resulting in a shortfall of
| African-American promotion.
Table 10. March 3, 1988 Lieutenant Examination
Other Whites Blacks Hispanics
Test-Takers: 0 98 11 7
Promoted: 0) 10 0 ]
“% Promoted: N.A. 10.2% 0.0% 14.3%
Promotion Rate as % of White Rate: 0.0%
Availability of Blacks: 9.5%
Total Promotions: 11
Observed Promotions of Blacks: 0
Expected Promotions of Blacks: 1.0
—Difference Between
Observed and Expected
Promotions (“Shortfall”): -1.0
36. In the October 12, 1989 Lieutenant Examination,
the rate of promotion for African-Americans was only 67.6%
of the rate of promotion for whites, resulting in a shortfall of
1 African-American promotion.
60a
Appendix D
Table 11. October 12, 1989 Lieutenant Examination
Other Whites Blacks Hispanics
Test-Takers: 0 76 15 9
Promoted: 0 15 2 2
% Promoted: N.A. 19.7% 13.3% 22.2%
Promotion Rate as % of White Rate: 67.6%
Availability of Blacks: 15.0%
Total Promotions: 19
Observed Promotions of Blacks: 2
Expected Promotions of Blacks: 2.9
—Difference Between
Observed and Expected
Promotions (“Shortfall”): -0.9
37. In the May 23, 1991 Lieutenant Examination,
the rate of promotion for African-Americans was only 54.8%
of the rate of promotion for whites, resulting in a shortfall of
2.5 African-American promotions.
Table 12. May 23, 1991 Sergeant Examination
Other Whites Blacks Hispanics
Test-Takers: 1 11] 19 15
Promoted: l 32 3 6
% Promoted: 100.0% 288% 15.8% 40.0%
Promotion Rate as % of White Rate: 54.8%
6la
Appendix D
Availability of Blacks: 13.0%
Total Promotions: 42
Observed Promotions of Blacks: 3
Expected Promotions of Blacks: 5.5
—Difference Between
Observed and Expected
Promotions (“Shortfall”): -2.5
38. Table 13 shows aggregate information for all
of the Lieutenant examinations from 1982 through 1991,
including the 1992 promotions from the 1991 Lieutenant
register of eligibles. The recent 1992 Lieutenant examination
is excluded from this table because only two persons from
the register compiled from that examination have been
promoted.
39. In the aggregate, the rate of promotion for
African-Americans in the Lieutenant Examinations from
1982 through 1992 (excluding the late 1992 test) was only
60.3% of the rate of promotion for whites, resulting in a
shortfall of 5 African-American promotions. There was no
adverse impact against Hispanics in rates of promotion to
Lieutenant, and no adverse impact against Asian-Americans,
Philippine Americans, and Others.
62a
Appendix D
Table 13. 1982-1991 Lieutenant Examinations in the Aggregate
Other Whites Blacks Hispanics
Test-Takers: 5 681 90 71
Promoted: ] 113 9 12
% Promoted: 20.0% 166% i0.0% 16.9%
Promotion Rate as % of White Rate: 60.3% 101.9%
Availability of Blacks: 10.6%
Total Promotions: 135
Observed Promotions of Blacks: 9
Expected Promotions of Blacks: 14.3
—Difference Between
Observed and Expected
Values (“Shortfall”): -5.3
2. Racial Disparities in Waiting Time Before Promotion
40. The Sergeant examinations from 1982 through
1992 caused additional adverse effects on members of the
class who were promoted, as well as their adverse impact on
class members who were not promoted. African-Americans
and Hispanic-Americans have tended to be clustered towards
the bottoms of the lists of eligibles, and as a result have had
to wait longer than whites to be promoted from a register.
Table 14 shows the average number of days’ wait from the
establishment of a register of eligibles for Sergeant until
promotion, for the period of time from 1982 through 1992,
for those members of each racial and ethnic group who were
promoted:
63a
Appendix D
Table 14. Number of Days’ Wait from the Establishment
of a Register to Promotion for the Average
Officer of Each Group Who Was Promoted
to Sergeant from 1982 through 1992
Whites Blacks Hispanics
Average Days’ Wait
from Establishment
of Register Until
Promotion: 171.7 212.0 222.4
41. These racial differences resulting from the Sergeant
promotional examinations are even more pronounced when
examining the delay until promotion across tests. Table 15
shows the average number of days’ wait to promotion for
the persons in each racial or ethnic group, starting with their
first competition for Sergeant in 1982 or later, until the date
of promotion:
Table 14. Number of Days’ Wait from First Com-
petition for Sergeant (1982 and later) Until
Promotion to Sergeant, for the Averacte
Officer Promoted to Sergeant from 1982
through 1992
Whites Blacks Hispanics
Average Days’ Wait
from First Effort
to Be Promoted
(in 1982 or later),
until Promotion: 530.7 810.1 780.9
64a
Appendix D
Thus, both African-Americans and Hispanic-Americans on
average had to wait six months longer than whites to be
promoted.
42. The Lieutenant examinations from 1982 through
1991 caused additional adverse effects on members of the
class who were promoted, as well as their adverse impact on
class members who were not promoted. African-Americans
have tended to be clustered towards the bottoms of the lists
of eligibles, and as a result have had to wait longer than whites
to be promoted from a register. The average successful
African-American candidate for promotion to Lieutenant had
to wait 336.4 days to be promoted from that register,
compared to 226 days for whites. The average number of
days’ wait to promotion for the persons, starting with their
first competition for Lieutenant in 1982 or later, until the
date of promotion, was 783.7 days for African-Americans,
compared to 615.6 days for whites.
3. Racial Disparities in Mean and Median Written
Scores
43. The City calculated mean test scores by race and
gender for the 1982 Sergeant examination, and calculated
both mean and median written test scores by race and gender
for the 1982, 1983, 1986, and 1987 Sergeant examinations,
and for the 1982 and 1986 Lieutenant examinations.
This information was provided to plaintiffs in response
to discovery.” It shows a generally consistent sizable racial
disparity in test scores between whites and African-Americans
10. Plaintiffs’ exhibits 11-16.
See NAA Pee chee
7 PR RACES ar gente
65a
Appendix D
of the same gender, and between whites and Hispanic-
Americans of the same gender:
Table 16. The City’s Calculation of Racial Disparities
in Mean and Median Written Test Scores
A. September 23, 1982 Sergeant Examination
Mean
Race and Gender No. Score
White Males 87 67.4]
Black Males 28 61.32
Hispanic Males 1] 62.73
Phil. Males 0 N.A.
Asian Males 0 N.A.
White Females 3 68.33
Black Females 3 47.67
Hispanic Females 0 N.A.
Phil. Females 0 N.A.
Asian Females 0 N.A.
B. November 17, 1983 Sergeant Examination
Mean Median
Race and Gender No. Score Score
White Males 301 58.46 60.66
Black Males 58 .- 52.09 50.00
Hispanic Males 46 54.16 54.00
Phil. Males 0 N.A. N.A.
Asian Males 0 N.A. N.A.
66a
Appendix D
White Females 14 59.14 60 33
Black Females 10 49.53 52.00
Hispanic Females ] 67.33 67.33
Phil. Females | 72.00 72.00
Asian Females l 44.66 44.66
C. July 17, 1986 Sergeant Examination
Mean Median
Race and Gender No. Score Score
White Males 327 62.31 61.00
Black Males 64 56.64 55.00
Hispanic Males 52 59.04 70.50
Phil. Males l 55.00 55.00
Asian Males ‘ 0 N.A. N.A.
White Females 24 61.75 57.00
Black Females 17 49.34 46.00
Hispanic Females 6 52.67 44.00
Phil. Females 0 N.A. N.A.
Asian Females 0 N.A. N.A.
D. October 29, 1987 Sergeant Examination
Mean Median
Race and Gender No. Score Score
White Males 341 75.47 76.00
Black Males 80 69.27 71.00
Hispanic Males 65 66.09 65.00
Phil. Males 3 84.66 87.00
Asian Males 0 N.A. N.A.
67a
Appendix D
: White Females 21 73.00 73.00
i Black Females 17 67.64 68.00
‘ Hispanic Females 8 73.62 75.00
: Phil. Females 0 N.A. N.A.
; Asian Females 0 N.A. N.A.
E. June 24, 1982 Lieutenant Examination
Mean
Race and Gender No. Score
White Males 493 76.00
Biack Males 15 60.00
Hispanic Males 27 71.00-
Phil. Males l N.A.
Asian Males 0 N.A.
White Females 18 61.50
Black Females 0 N.A.
Hispanic Females 0 N.A.
Phil. Females 0 N.A.
Asian Females 0 N.A.
F. December 11, 1986 Lieutenant Examination
Mean Median
Race and Gender No. Score Score
White Males 68 75.13 75.50
Black Males 12 70.58 69.00
Hispanic Males 6 68.00 66.00
Phil. Males 0 N.A. N.A.
Asian Males ] 70.00 70.00
68a
Appendix D
White Females l 86.00 86.00
Black Females 2 82.50 82.50
Hispanic Females 0 N.A. N.A.
Phil. Females 0 N.A. N.A.
Asian Females 0 N.A. N.A.
44. The information in Table 16 is consistent with
the information in Tables 1-15 showing substantial racial
disparities in promotion rates.
45. Plaintiffs have.also performed calculations of the
mean differences in test scores. Where information on the
written test score was available, this was used. Where only
information on total score was available from the test-
registers, this was used. Information was calculated for
whites, African-Americans, Hispanic-Americans, and others,
a category which includes Asian-Americans and Philippine
Americans. The following table shows the results:"'
Table 17. Plaintiffs’ Calculation of Racial Disparities
in Mean Test Scores
A. September 23, 1982 Sergeant Examination
Mean Difference
Total From Mean
Race No. Score White Score
Whites 85 73.36
Others, Asians, Phil. 0 N.A. N.A.
African-Americans 33 64.82 -8.54
Hispanic-Americans 14 69.50 -3.86
11. Affidavit of Barbara Diggs, plaintiffs’ exhibit 4.
SRE SRE MEN SERB BARES SHOR IE REN OA ERED
69a
Appendix D
B. November 17, 1983 Sergeant Examination
Mean Difference
Total From Mean
Rac No. Score White Score
Whites 295 64.93
Others, Asians, Phil. ] 78.00 13.07
African-Americans 75 58.13 -6.80
Hispanic-Americans 60 58.85 -6.08
C. January 31, 1985 Sergeant Examination
Mean Difference
Total From Mean
No. Score White Score
Whites 408 75.36
Others, Asians, Phil. ] 59.00 -16.36
African-Americans 87 67.31 -8.05
Hispanic-Americans 61 70.38 -4.98
D. July 17, 1986 Sergeant Examination
Mean Difference
Written From Mean
Score White Score
Whites 62.12
Others, Asians, Phil. 55.00 -7.12
African-Americans $5.32 -6.80
Hispanic-Americans 59.09 -3.03
70a
Appendix D
E. October 29, 1987 Sergeant Examination
Mean Difference
Written From Mean
Race No. Score White Score
Whites 362 75.33
Others, Asians, Phil. 3 84.67 9.34
A frican-Americans 97 68.99 -6.34
Hispanic-Americans 73 66.92 -8.4]
F. April 27, 1989 Sergeant Examination
Mean Difference
Written From Mean
Race No. Score White Score
Whites 410 66.66
Others, Asians, Phil. 6 63.67 -2.99
African-Americans 129 60.04 -6.62
Hispanic-Americans 109 60.16 -6.50
3. October 31, 1991 Sergeant Examination
Mean Difference
Written From Mean
Race No. Score White Score
Whites 415 74.61
Others, Asians, Phil. 6 57.50 -17.11
African-Americans 175 66.85 -7.76
Hispanic-Americans 111 67.21 -7.40
7la
Appendix D
H. June 24, 1982 Lieutenant Examination
Mean Difference
Total From Mean
Race No. Score White Score
Whites 54 83.87
Others, Asians, Phil. ] 91.00 iis
African-Americans 3 69.67 -14.20
Hispanic-Americans 4 79.50 -4.37
I. August 11, 1983 Lieutenant Examination
Mean Difference
Total From Mean
Race No. Score White Score
Whites 63 77.37
Others, Asians, Phil. | 68.00 -9.37
African-Americans 3 58.00 -19.37
Hispanic-Americans 9 69.22 -8.15
J. September 27, 1984 Lieutenant Examination
Mean Difference
Total From Mean
Race No. Score White Score
Whites 123 83.85
Others, Asians, Phil. l 76.00 -7.85
African-Americans 13 77.08 -6.77
Hispanic-Americans 13 80.85 -3.00
72a
Appendix D
K. November 7, 1985 Lieutenant Examination
Mean Difference
Total From Mean
Race No. Score White Score
Whites 88 80.19
Others, Asians, Phil. ] 70.00 -10.19
African-Americans 12 75.25 -4.94
Hispanic-Americans 8 72.25 -7.94
L. December 11, 1986 Lieutenant Examination
Mean Difference
Written From Mean
Race No. Score White Score
Whites 68 75.66
Others, Asians, Phil. 0 N.A. N.A.
African-Americans 14 72.29 -3.37
Hispanic-Americans 6 68.00 -7.66
M. March 3, 1988 Lieutenant Examination
Mean Difference
Written From Mean
Race No. Score White Score
Whites 98 68.98
Others, Asians, Phil. 0 N.A. N.A.
African-Americans 1] 65.36 -3.62
Hispanic-Americans 7 70.71 1.73
73a
Appendix D
N. October 12, 1989 Lieutenant Examination
Mean Difference
Written From Mean
Rac : Score White Score
Whites 81.73
Others, Asians, Phil. N.A. N.A.
African-Americans Fito -3.98
Hispanic-Americans 9 76.67 -5.06
O. May 23, 1991 Lieutenant Examination
Mean Difference
Written From Mean
No. Score White Score
——
Whites 111 72.31
Others, Asians, Phil. ] 84.00 10.89
African-Americans 19 73.21 0.10
Hispanic-Americans 15 tate 0.62
P. September 30, 1992 Lieutenant Examination
Mean Difference
Written From Mean
Score White Score
Whites 69.82
Others, Asians, Phil. 65.50 -4.32
African-Americans 66.00 -3.82
Hispanic-Americans 65.80 -4.02
74a
Appendix D
46. This table shows a generally consistent pattern
in which test scores obtained by African-Americans and
by Hispanic-Americans are generally lower than test scores
obtained by whites. This is consistent with the above findings
on adverse impact in promotion rates.
47. The information in this table as to Asian-
Americans, Philippine Americans and others contains
averages for very few test-takers, and the information is
correspondingly less meaningful. There is much less
consistency of results. The overall picture is consistent with
the lack of adverse impact against Asian-Americans,
Philippine Americans and Others shown in findings 31 and
39 above.
4. The Question of Adverse Impact Against Women
48. Female and Asian-American Police Officers
have also objected to the settlement on the ground that
it does not contain information on the question of adverse
impact against women and, in the event that there is such
adverse impact, does not provide relief for such adverse
impact.
49. The parties are not aware of any EEOC charge
having been filed which raises the question of gender
discrimination against women in the Police Department’s
promotional examinations. There is no evidence that such a
charge was filed, and no claim of gender discrimination has
ever been made in this lawsuit.
75Sa
Appendix D
50. The City has not provided gender-identified
Registers of Eligibles to plaintiffs for all registers, and no
information on gender was entered onto the database until
the objections of women were filed. Plaintiffs’ paralegal,
Barbara Diggs, examined the first names of test-takers in
an effort to assign gender te them, so that an approximate
analysis could be performed for the 1989 and 1991 Sergeant
examinations. She was not able to determine gender for
26 test-takers in 1989 or for 27 test-takers in 1991. Because
it was likelier that more women would be taking the Sergeant
examinations than the Lieutenant examinations, the analysis
was done only for Sergeants. The results are as follows:!?
Table 18. Approximate Adverse Impact Calculation for
the 1989 and 1991 Sergeant Examinations,
by Gender
A. April 27, 1989 Sergeant Examination
Unknown Males Females
Test-Takers: 26 562 67
Promoted: 0 69
% Promoted: 0% 12.3%
Promotion Rate as % of White Rate:
Availability of Women Among
Gender-Identified Test-takers
Total Promotions:
12. Affidavit of Ms. Diggs, plaintiffs’, exhibit 4.
76a
Appendix D
Observed Promotions of Women:
Expected Promotions of Women:
—Difference Between
Observed and Expected
Promotions (“Shortfall’’):
Standard Deviation:
Number of Standard Deviations
Between Expected and
Observed Promotions
B. October 31, 1991 Sergeant Examination
Unknown Males
Test-Takers: 27 600
Promoted: 0 101
% Promoted: 0% 16.8%
Promotion Rate as % of White Rate:
Availability of Women Among
Gender-Identified Test-takers
Total Promotions:
Observed Promotions of Women:
Expected Promotions of Women:
—Difference Between
Observed and Expected
Promotions (“Shortfall’’):
77a
Appendix D
Standard Deviation:
Number of Standard Deviations
Between Expected and
Observed Promotions -0.939
The results for the 1991 examination show a low degree of
adverse impact under the “4/5” rule, although the results are
not statistically significant. With the aggregation of enough
tests, it may be possible to show statistically significant
disparities by gender.
51. Some of these women are class members, how-
ever (41 of the 81 women taking the 1991 test), and there is
a risk of double-counting adverse impact under both race
and gender.
52. It is not clear that there has been substantial
adverse impact against women because of their gender,
distinct from adverse impact against African-American
women or Hispanic-American women because of their race.
53. The resolution of this lawsuit will not impair
women in any effort they may make to prove any disparate
impact against them as women, distinct from any adverse
impact some women share with other class members because
of their race, arising from the promotional examinations
in the Police Department. It will not prevent any woman
aggrieved by any gender discrimination in these promotional
examinations from filing an EEOC charge of discrimination,
and will not limit them in seeking remedial promotions or in
obtaining an analysis of possible gender bias in test items
comparable to the log-linear analysis to eliminate racially
biased test items.
- ’
Sa
Appendix D
S. Objections to the Showing of Disparate Impact
(a) The Contention that Plaintiffs Must Prove
Disparate Impact on a Ouestion-by-Ouestion
Basis
54. The Houston Police Patrolmen’s Union [“HPPU”’]
contends that a showing of disparate impact cannot be made
based on ractal disparities In promotion rates, but can only
be made on a question-by-question basis, or can only be made
as to those questions which are contended not to be job-
related."
55. HPPU’s proposal is not an efficient or
workablemeans of showing disparate impact. First, showings
of the disparate impact of each particular question on an
examination would require keypunching the answers of each
test-taker to each 1tem on the examination. A total of 708
candidates took the October 31, 1991 Sergeant Examination,
and keypunching !00 items for each candidate would require
keypunching 70,800 items of information in addition to name
and race for just one examination. This would be enormously
expensive. Second, the adverse impact of a particular test
item 1s not meaningful without reference to the whole test:
a. The City does not make promotions based on
responses to specific test items, but in response to the
total scores on all test items.
13. HPPU Objections, { 9 at 8-9 and { 16 at 12-13, Objections
io Proposed Consent Decree, Vol. II, Tab 59.
79a
Appendix D
b. While a great deal of adverse Impact may occur
as tO a particular test question, it may be offset by other
test questions, with the result that the test as a whole
does not have enough adverse Impact to be legally
significant.
c. The adverse impact of a selection procedure
may have little to do with particular items having an
extraordinary degree of adverse Impact, but may simply
be the cumulation of small amounts of adverse impact
on each item.
AS a practical matter, it is the overall test score, plus seniority
points, which determines the candidates who will be
promoted, and it is the overall test score Which is the
appropriate unit for purposes of determining adverse
Impact. Third, the HPPU misconceives the function of
identifying numerous test items Which cannot readily be
defended as job-related: this is not part of a showing of
adverse impact, but is part of a showing of substantial
doubt that the tests are job-related. The fact that numerous
items are clearly not job-related Suggests the presence
of yet others which full litigation could identify as not
job-related, and demonstrates that the overall test scores
determining promotions are so substantially influenced
by non-job-related components that the overall test scores
cannot be said to be job-related.
- ~~ a! a ee eee, =— sh
SOa
Appendix D
(b) The Contention that the Parties Must Prove
Adverse Impact Based on the Numbers of
Persons Eligible to Take the Examination,
Not on the Numbers of Test-Takers
56. Doug Elder and Mark Clark, on behalf of members
of the Houston Police Officers Association,'* complain
that showings of adverse impact should not be based on the
proportions of test-takers of each race who were promoted,
but on the proportions of persons eligible to take the test
who were promoted.'* Thus, if 2,000 whites were eligible to
take the test, 1,000 took the test, and 500 were promoted,
25% of the eligible whites would have been promoted.
If SOO African-Americans were eligible to take the test
and every last one took it and only 125 were promoted,
Elder and Clark would then say that there was no adverse
impact because 25% of the eligible African-Americans had
been promoted.
57. Even if following the approach suggested by
Elder and Clark would make a difference in the determination
of adverse impact and they have offered no supporting
evidence — it would make no sense. A person who chooses
not to compete for promotion to Sergeant is not “eligible”
for promotion: one of the conditions of eligibility is that one
take the test. Indeed, Elder and Clark admit in their objection
2 that the analysis offered by the parties 1s the type “generally
used” when dealing with objective tests.
14. Objections to Proposed Consent Decree, Vol. I], Tab 70.
l
ws
Id., Objection | at 5-6.
Sla
Appendix D
(c) The Contention That There Must be an
Outside Audit
58. Elder and Clark complain in their objection 2
that there must be an “independent or third party analysis
or validation of the numbers or data’. This is incorrect. The
raw numbers of test-takers and promotions of members of
each race have been set forth in the tables, and the HPPU
can check both the arithmetic accuracy of the stated
promotion rates for each race and any standard-deviation
analysis from the raw numbers stated. The City’s registers
of eligibles are neither secret nor confidential, and a union
with all of its resources ought to be able to confirm the
accuracy of the raw numbers.
59. Contrary to the Elder and Clark statement.'°
the Court has never entered any Order blocking access to
any statistical information. Moreover, Chief of Police Nuchia
has sworn that no request for such information was ever made
out, if these objectors had requested the information. it would
have been given to them. Elder and Clark cannot be heard
to complain of not having been given information which
was freely available to them if they had simply bothered to
request it.
(d) The Contention That Entry-Level Officers
Should Be Compared with Manager
60. Plaintiffs and the City have not contended that
adverse impact can be shown by comparing the racial
16. HPPU makes a similar misstatement in @* 16-17 of its
proposed Complaint in Intervention.
82a
Appendix D
composition of entry-level police officers with the racial
composition of managers. This suggestion is only made by
Elder and Clark at 9-10.
(e) The Contention That Adverse Impact Should
Be Determined Without Regard to Candi-
dates Who Failed the Tests
61. The objections of Herman Mar et al. state in
“16 at 10-11 that the proper statistical comparison is the
number of promotions in each racial group divided not by
the number of test-takers in each group, but by the number
of test-passers in that group. They assert without support that
passing the test is necessary to show that one Is “qualified”.
This assumes the very question at issue In a testing case:
is the test valid?
62. A showing of disparate impact simply means that
the challenged practice has done sufficient harm to the
interests of a protected group to warrant an inquiry into its
justification. The persons who fail a test are just as much
harmed as those who pass but have scores too low to be
reached for promotion. The objection is not valid.
(f) The Contention That Passing Rates Are More
Probative Than Promotion Rates
63. Inconsistently, the Mar objectors also compiain
that adverse impact should be based on passing rates rather
than promotion rates. Objections, 4 15 at 9-10. Here, they
focus exclusively on persons who fail the test, and ignore
persons who pass the test but obtain scores too low to be
4.
SIa
Appendix D
reachable for promotion. This is a mirror image of their other
objection, and is just as wrong.
64. In the ordinary course, there are many more test-
takers who pass a promotional test for Sergeant or Lieutenant
than the Police Department can promote. The City ranks them
in order of their total scores, based on their written scores
and up to ten seniority points, and promotes candidates based
on their relative rank. It is not unusual for hundreds of
lest-passers to remain unpromoted on the ex piration of
the list.
65. The persons who pass a test but have scores too
low to be reached for promotion are just as much harmed as
those who fail the test. The objection is not valid.
(h) Conclusion as to the Objections to the
Determination of Adverse Impact
66. None of the objections to the showi ing of adverse
Impact are substantial. The parties have adequately demon-
strated the existence of adverse impact against both African-
Americans and Hispanic-Americans on written portions of
the Sergeant examinations from 1982 ee 1992. and
against African-Americans on the writte 1 portions of the
Lieutenant examinations during this eee
67. The ninety-six remedial promotions for A frican-
Americans and for Hispanics for the Sergeant examinations
do not exceed the promotions actually lost because of
the adverse impact caused by the written examinations
themselves.
PSS ce ween. ~_ceneteetemmnmenae
84a
Appendix D
68. The five remedial promotions for African-
Americans for the Sergeant examinations do not exceed
the promotions actually lost because of the adverse impact
caused by the written examinations themselves.
69. The five remedial promotions to Lieutenant
are an appropriate remedy for the delay in ability to compete
for Lieutenant positions caused by the written portions of
the Sergeant examinations.
70. No objector has presented any credible evidence
that any factor other than the challenged tests accounts for
any meaningful part of the racial disparities in promotion
rates between whites and African-Americans, or between
whites and Hispanics:
E. The Job-Relatedness of the Challenged Examinations
1. Problems with the Job-Relatedness of Numerous
Questions
Ths
Plaintiffs have identified numerous questions on
the examinations which they contended are on their face
extremely difficult to defend as job-related and consistent
with business necessity, while cautioning that a more
detailed inquiry could reveal many others. Plaintiffs’ counsel
provided some of this information to the City in a meeting
with the City Attorney held on December 1, 1992, prior to
the City’s decision to settle the case. On the City’s request,
counsel for plaintiffs provided the City on December 21,
1992 with a written list of such illustrative “problem
questions” for several challenged promotional examinations,
Appendix D
and offered to prepare such a list for others of the challenged
examinations if the City so desired.'’ Mr. Seymour’s letter
to counsel for the City stated:
i have also gone through a number of
the examinations in question, to add to the list of
problems | went over with you and with Mr. Hall
at our December | meeting, in connection with
the illustrative May 199] Lieutenant test.
Naturally, | cannot identify problems which
can only be discerned by examining the study
materials and comparing them with the test items.
and we have not gotten far enough to compare
the assumptions of particular questions with the
assigned duties of Sergeants and Lieutenants in
the HPD. The following are the facially detectible
problems:
1. September 23, 1982 Sergeant Examination:
For several questions, it is hard to tell which are
problem questions without looking at the text of
the study materials. The following questions,
however, appear on their face to be problema-
tic: questions 36 (question requires ability to
distinguish views of socielogical theorists by
name, where such ability has no relationship
to the job), 44 (tendentious), and 88 (concerns
unspecified person’s theory, not job-related
knowledge). In addition, knowledge of precise
standards for charging particular offenses is not
17. Plaintiffs’ exhibit 8.
ET cc NES ATIC IT AIS NN 3 RE a POD AEE sR A AEN CO REPORT SLE ROUT 0 HB RMN
SOa
Appendix D
relevant when the charging Officer and reviewing
instead of memory.
2. November 17, 1983 Sergeant Examination:
For several questions, it is hard to tell which are
probiem questions without looking at the text of
the study materials. The following questions,
however, appear on their face to be problematic:
questions 50 (familiarity with term “gestalt”);
52 (nothing to do with Sergeants); and 57, 59, 61,
65. 67, 68, 69, 72, 76, 77, 80, 82, 83, 86, 87, 88,
89. 95. 97. 99, 104, and 105 (questions require
ability to distinguish among sociological theories
by name, and to distinguish views of one author
from those of another, where such ability has
no relationship to the job). In addition, knowledge
of precise standards for charging particular
offenses is not relevant when the charging Officer
and reviewing Sergeant actually make use of
written standards instead of memory.
3. January 31, 1985 Sergeant Examination
(same general points): The first page of the exam-
ination booklet indicates that the Civil Service
Department saw a problem with question 19, but
it is hard to read the notation. There are problems
with questions 40 (on hypercube queuing model
computer program, not used by HPD), 53 and 65
(tendentious), and 60 and 62 (no answer is
accurate on its face, although the text should be
checked to see if it is also inaccurate).
S/a
Appendix D
4. July 17, 1986 Sergeant Examination.
The first page of the examination booklet indicates
. that the Civil Service Department saw problems
with questions 2, 9, 11, 14, 16, 20, 24, 25, 28, 29.
31 (the correct answer assumes that Sergeants
are responsible for staffing, but this is not true
in the HPD),'* 34, 36, 49, 50 (two problems),
51, 53, 56, 60, 76, and 85, and that typograph-
ical errors affected three questions (numbers
not fully legible, but could be 10, 48, and 87).
In light of the City’s own admission of these
problems, no purpose would be served by our
doing an independent check.
5S. October 29, 1987 Sergeant Examination.
The first page oi the examination booklet indicates
that the Civil Service Department saw problems
with several questions (up to ten), but the markings
are very faint on our copy and we cannot tell which
items or their number. The body of the test indicates
that there may have been a problem with question
14, but the remarks are illegible. There are also
facially detectible problems with questions 6
(contradicts other tests covering the same subject),
12 (has to do with National Labor Relations Act
definitions of unfair labor practices, which is
inapplicable to the HPD), 16 (has to do with job
evaluation, a matter of concern to industrial
18. This is an example of a problem which is found only by
sitting down with members of the HPD, and which we would not
have found simply by doing a desk review of the items. Other tests
may contain similar problems without our being able to detect them.
Sda
Appendix D
psychologists and economists, not to Sergeants),
17, 22 (questions require ability to distinguish
views of one author from those of another,
where such ability has no relationship to the job),
29 (has to do with manufacturing, not with police
work), and 30 (has to do with white-collar work,
not police work).
6. April 27, 1989 Sergeant Examination.
We only have question | through 92 on this test,
and could not review the remainder. The Civil
Service Department deleted questions 29 and 98
from the test after it was given. The largely
illegible markings on the front sheet of the test
indicate that the civil Service Department also
had questions about items 9, 20, and 86, but these
were retained. The body of the test indicates that
there may have been a problem with question 19
(remarks not completely legible), contains largely
illegible notes about question 86, and contains
notes about question 89 (“claims it onl) applies
to a misdemeanor”). There are also facially
detectible problems with questions 9 (incompre-
hensible), 18 (question does not deal with what 1s
correct and what is incorrect, but with what is not
listed in text: the item rewards photographic
memory and penalizes understanding of the point
by officers who may recall it from another source),
71 (no answer is correct; without seeing the study
text it is not possible to say whether the text itself
is correct), and 72 (same; the item writer — or
possibly the text author — has confused liberty
with due process).
aie
aA”. Awn Ane -_ Pa le : <i a i
sYa
Appendix D
If it would be helpful to the City for me to
identify such questions from other examinations.
| will be happy to do so. It seems to me that these.
in combination with the problems pointed out for
the 199] Lieutenant examination, should suffice.
2 le,
the City decided to settle this lawsuit. The City examined
the information, and concluded that it would in fact be
difficult to defend the job-relatedness of these. and perhaps
other, questions. Indeed, the City Attorney read some of the
non-job-related test questions to the City Council as part of
his explanation of why the city feared it might not be able to
defend this lawsuit successfully if it were litigated.
lhe Council accepted the explanation.
73. The parties have provided the Court with copies
of the challenged tests for in camera inspection. Based upon
its inspection, the Court is satisfied that the City made a
reasonable and arm’s-length judgment that it would be
difficult to defend the job-relatedness of many of the
questions on these tests.
74. By Order of March 22, 1993, the Houston Police
Patrol Union was allowed to designate two persons of its
choosing to examine the questions for itself, subject to the
requirement that the items not be disclosed.
5. The City also faced the risk that, if the litigation
continued, closer inspection of the study materials and
comparisons with actual job duties would bring to light
additional questions — themselves potentially numerous
which would also be difficult to defend as job-related.
2. This information was provided to the City before
90a
Appendix D
76. Adifference ofa single point on a test may determine
whether a particular candidate will be promoted, and how
long he or she may have to wait for a promotion.
The problem of there being numerous test questions,
the job-relatedness of which is open to substantial doubt, is
important in itself but takes on added importance because of
the adverse light it sheds on the degree of care with which
the entire examinations were constructed.
2. The City s Knowledge of its Problems in Showing
that the Fire Department Examinations, Prepared
in the Same Manner as the Police Department
Examinations At Issue Herein, Were Job-Related
78. The City’s preparation of promotional examina-
tions for both its Police and Fire Departments is governed
by Chapter 143 of the Texas Local Government Code. The
examinations for both Departments are prepared by the same
City Personnel Department officials, in the same manner.
Book committees are established by each Department to pick
the volumes which promotional candidates are to study.
Candidates are supposed to study the books over a period of
time, committing as much as possible to rote memory.
Multiple-choice test items are written, based on the
information in the books.
79. Because Fire and Police Department promotional
examinations are prepared in the same manner, by the same
persons, and under the same standards, the City’s experience
Yla
{ppendix D
P]
In 1990 and 1991 in attempting to show that the Fire
Department’s promotional tests were job-related is relevant
to the question whether the Police Department’s promotional
examinations were job-related.
S80. The City’s study of the Fire Department's
examinations occurred in-the course of litigation against
those examinations, Houston Chapter of the International
Association of Black Professional Firefighters v. City of
Houston, C.A. No. H-86-3553 (S.D.Tex.). The City’s expert,
Dr. Jeanneret, performed a criterion-related validation Study
of the promotional examinations for the two lowest-le\ el
promotional ranks, Chauffeur and Junior ( ‘aptain, and found
some statistically significant positive relationships between
test scores and research ratings of job performance which
supervisors made under his direction.
81. Plaintiffs’ experts in that case were Charles R.
Mann, Ph.D., an expert statistician, and David L. Friedland,
Ph.D., an expert industrial psychologist. Dr. Friedland has
previously helped to develop selection tests in urban public
Safety departments. Their March and April 1991 Affidavits
recounting some of their analyses were introduced in
evidence at the fairness hearing prior to final approval of the
settlement in that lawsuit. The Affidavits havé also been
received in evidence here,'’ and Drs. Mann and Friedland
testified about those analyses at the fairness hearing herein.
19 Plaintiffs’ exhibits 18, 19. and 20
Q2a
Appendix D
8? Dr. Mann’s March 29, 1991 Affidavit’? stated that
the November 1990 Report prepared by the City’s expert In
the Fire Department case described a total of 11 performance
dimensions. For example, the report states at p. 7:
A dimension is a general term used to represent
important work behaviors common to performance
of several job tasks. Performance dimensions are
the general categories of behavior often used to
distinguish workers who are successful or effective
from those workers who are less successful or less
effective on the job.
Dr. Jeanneret also stated on p. 8 of his November 1990 Report
that each performance dimension “was divided into five
blocks labeled Very Good, Good, Average, Poor, and Very
Poor.” Dr. Mann also noted from Appendix D to Dr.
leanneret’s November 1990 Report that he established the
following dividing lines for his adjectival characterizations
of employee performance:
Range of Points Characterization of Performance
1-12 Very Poor
13-24 Poor
25-36 Average
37-48 Good
49-60 Very Good
20. Plaintiffs’ exhibit 18
Appendix D
Id., 43-44. Because Dr. Mann’s field of expertise is statistics
and not industrial psychology, he expressed no opinion
(a) on the manner in which Dr. Jeanneret established his
performance dimensions, or (b) on the meaning of those
dimensions, or (c) on Dr. Jeanneret’s adjectival characterizations
of various levels of ratings. /d., 4 50.
83. Dr. Mann further explained that a Pearson
product-moment correlation may be used to measure the
strength of the linear relationship, if any, between continuous
variables such as test score and job performance ratings.”!
fd. 4 SZ.
84. A correlation of | corresponds to a perfect posi-
tive linear relationship between the variables, in which the
Increase in one variable is directly proportional to the increase
in the other variable, at every point along the continuum of
values. A correlation of -1 corresponds to a perfect negative
linear relationship between the variables, in which the
Increase in one variable is inversely proportional to the
decrease in the other variable, at every point along the
continuum of values. A correlation of 0 indicates that there
is no linear relationship between the two variables. /d., § 53.
85. A correlation can be any number between -]
and +1. The square of the correlation of two variables
(the “coefficient of determination”) is the proportion of
the variation in one variable which can be “predicted” or
“explained” by the other variable. In correlating test scores
with job ratings, for example, a correlation of .3 “predicts”
or “‘explains” 9% of the variation in job ratings. /d., 4 54.
21. Correlations between test scores and performance measures
are commonly referred to as “validity coefficients” by psychologists.
94a
Appendix D
86. The magnitude of an observed correlation 1s
not meaningful unless there is reason to believe that it did
not differ from zero merely by chance. In order to consider
the true correlation as opposed to its observed value, it 1s
necessary to perform a test of statistical significance of the
hypothesis that the true value of the correlation coefficient
is zero. /d., para, 55,
87. The closer to zero the correlation, the less useful
to the employer 1s the test and therefore the more difficult
it 1s to justify adverse *mpact against minorities or any other
cost of using the test. Whether a particular correlation
is sufficiently high to justify the use of a test, given the
test’s degree of adverse impact, requires a value judgment
outside the field of statistics. As a rule of thumb, however,
Statisticians commonly advise their clients that correlations
less than .3 (those which do not “explain” or “predict” at
least 9% of the variation) are not practically useful in many
situations. /d., 4] 56
88. There are, however, situations in which having
any information at all is extremely important, and a statis-
tically significant correlation may be important even though
the correlation is extremely low. For example, the failure
rate of a critical aircraft engine component may have a very
loose, but real, relationship with the operating temperature
of the engine. This relationship may reasonably lead the
engine builder to re-design the component or modify the engine
to have a lower operating temperature. The above information
is previded for purposes of illustration; ultimately, the value
judgment as to the utility of a low correlation must be made
outside the realm of statistics, by a person taking into account
all_of the relevant considerations. /d., 4 57.
Sa
Appendix D
89. Dr. Mann found a numbe! of problems with the
data and approach used by Dr. Jeanneret. but was asked to
ignore those problems, accept for purposes of argument the
raw data prepared by the ( ‘ity’s expert, and to carry out further
analyses using those raw data. /d.. 4] 58-71.
90. For the sake of brevity, Dr. Mann’s analysis of
the data for the 1990 Chauffeur test will be used as illustrative
of his analyses for all of the other challenged tests in that
case. A large number of candidates took this test. reducing
the risk of random fluctuations. Dr. Mann found the following
correlations between test scores and job-performance ratings
on the Chauffeur composite that set of dimensions of
job behavior which according to the City’ expert captured
performance on those parts of the Chauffeur Job which
Were most important. Dr. Mann found a .209 correlation
between test scores on the 1990 Chauffeur eXamination and
the “Chauffeur Composite” performance rating. /d., 4 72.
He concluded from the results on this and the other tests
that “These correlations are sufficiently low that [ believe
additional justification is needed before using results based
on them.”
91. Ascatterplot showing test scores and job perfor-
mance ratings for the 1990 Chauffeur test appears on the
next page. Each point shown on a scatterplot represents one
or more persons with a particular job performance rating and
a particular score on the test in question. The rank for which
the examination was given is shown on the vertical axis,
with the full scale from 0 to 60 represented. Test scores from
the examination or test aggregate In question are shown
on the horizontal axis, with the full scale from 0 to 100
96a
Appendix D
represented. A person with a Chauffeur composite rating of
50 and a 1990 Chauffeur test score of 70 will appear on the
scatterplot for the 1990 Chauffeur examination as a point at
the intersection of 50 on the vertical scale and 70 on the
horizontal scale. The scatterplot for the 1990 Chauffeur test
(Attachment 4 to Dr. Mann’s Affidavit) is shown below:
50 | |! T ‘Me ! 7
' ~« te
a 3 |
| eef
§ . % 5
‘ond = : ae Z se i ae |
/ i + s has © © Se :
- | tse abd Perens |
= 1. owe ae . “a - Me ' |
= | Ag Pee, |
c4aok jpn aS 4
| . $ e « ‘
2 | : Th) ha the Fe Mg :
OC i mcs
am | : “ot :
= SU Zz 8 ss i. 7 —
a
= 20+ :
ae
{Sd gue ond
O H J ! ! Z ! | ! l !
0 10 ZO 30 40 50 60 70 80 90 100
Chauffeur Test Score (1990)
97a
Appendix D
92. The scatterplot shows visually that there are
numerous test-takers with relatively high performance ratings
but relatively low test scores, and that there are numerous
test-takers with relatively low performance ratings but
relatively high test scores. The relationship between test score
and job performance is not a close one. This scatterplot also
confirms visually the weakness of the .209 correlation found
for this test. /d., V9 92-93.
93. Dr. Mann also prepared a set of bar charts
showing mean performance ratings for written test-score
intervals or bands of 25 persons. Where the 25th person in
a test-score band had the same test score as one or more persons
not already in the band, the additional persons were included
in the band in order to eliminate any breaking of ties. Thus.
every band but the last includes a minimum of 25 persons.
The last test-score band on each page shows information for
the from | to 25 persons remaining. These bar charts show
the size of the differences between mean performance ratings
of persons in different test-score bands. The chart for the
1990 Chauffeur test appears on the next page.
94. The mean Chauffeur Composite performance
rating for the 31 firefighters scoring from 81 to 95 (the top
1990 written test score achieved) on the 1990 Chauffeur
test was 43.72 points on the 60-point scale. This mean
performance composite rating was almost five points lower
than the mean rating of 48.46 points obtained by the 26
firefighters scoring from 76 through 80 on that test, and
was more than three points /ower than the mean job perfor-
mance rating of 47 points obtained by the 33 firefighters
22. See Attachment 5 to Dr. Mann’s Affidavit.
98a
Appendix D
Mean Performance Ratings of
1990 Chauffeur Test Score Groups
All Test-Takers
Test
Scores
81-96 ANAS 43.72
76-80 Ny
73-75 SSR «7
71-72 Hl ASA ER 47.3:
51-63 SRE 45.48
58-60 SAM «0.04
§6-57 4 42.68
53-65 43.22
51 QR «4-02
48-50 +I RAMR 42.11
0 20 40 60 80
Chauffeur Composite Rating
Note:
ne 542
Group size © 25 « ties
100
99a
Appendix D
scoring from 73 to 75 on the written test. Indeed. one has
to go down to the firefighters scoring from 67 to 70 on the
written test to find a mean job performance rating lower than
that achieved by the top scorers. Some bands of firefighters
with failing test scores had higher mean job performance
ratings than the top-scoring fire fighters. See, e.g., the
37 firefighters scoring from 65 to 66 (mean job rating of
45), the 42 firefighters scoring from 61 to 63 (mean job
rating of 45.48), and the 25 firefighters scoring exactly 5]
(mean job rating of 44.02).
95. While there is a general gross trend in which
bands of firefighters scoring lower on the test have lower
job-performance ratings, these are chiefly distinctions among
the top group of 125 or so firefi fighters with scores ranging
from 71 to 95, the large middle group of 335 or so firefighters
with scores ranging from 46 to 70, and the bottom 83
with scores ranging from 26 through 45. Even there, the lev el
of the difference is just a few points on a 60- point scale,
and it is not clear that this is a meaningful difference in job
shia This information tends to suggest that, even
f differences of a few points in performance ratings are
ies meaningful, the test may be useful only in mak-
ing gross distinctions among persons with relativ ely large
—
differences in test scores.
96. Dr. Mann’s staff also divided the Chauffeur test-
takers for a given examination into {ive equal sets or quintiles,
in order of their age geen ratings, in order to compare
the mean test scores of firefighters with the same level
of job performance. /d., 4 99. The bar chart for the 1990
Chauffeur test, from Attachment 7 to the Affidavit. is on the
next page:
100a
Appendix D
Mean 1990 Chauffeur Test Score
by Chauffeur Composite Rating
All Test-Takers
Chauffeur
Composite Rating
\ a. 62.79
NY
\\
\ \
ANY
NN
WY
Highest 20%
(51.00-60.00)
0) |
}
4th 20%
(46.33-50.83)
Middle 20% 60.18
(41.83-486.17)
A\
0 20 40 60 80
Test Score
lOla
Appendix D
There was very little difference in mean test scores among
the highest-performing 60% of firefighters.
97. The next two bar charts drawn from Attachments
9 and 11 to Dr. Mann’s Affidavit use the same job-performance
quintiles, and compare the proportions of firefighters in the
different job-performance groups who pass or who fail the
test. /d., §| 102, 105. There is very little difference among
the middle three quintiles in the proportions of firefighters
passing or failing the test. The gross differences shown in
the charts are among the top quintile, the large middle group,
and the bottom quintile.
98. One of the interesting features of the charts is
that the great majority of firefighters at all levels of job
performance fail the test. In the highest-performing group of
firefighters, close to 63% fail the test.
102a
Appendix D
1990 Chauffeur Test: Percent Passing
by Chauffeur Composite Rating
All Test-Takers
Chauffeur
Composite Rating
Highest 20% 37.04
(51.00-60.00)
ee
’ \\\ NY
4th 20% \ N \Y\ 28.57
(46.33-80.83) ‘x AY
Middle 20%
(41.83-46.17)
2nd 20%
(36.50-41.67)
Lowest 20%
(16.00-36.17)
t | i j
0 20 40 60 80 100
Percent Passing (Score 70 or Above)
n= 542
lO3a
Appendix D
1990 Chauffeur Test: Percent Failing
by Chauffeur Composite Rating
All Test-Takers
Chauffeur
Composite Rating
" \ ae AN\\
Highest 20% . Se a
(51.00-60.00) Me
\ \ oe
A
4th 20% ARAL RAR
(46.33-50.83) SATS AN Ae ARE A EN
WW
Middie 20% \ a
(41.83-46.17) a
AY \
2nd 20%
(36.50-41.67)
\ \\\
Lowest 20% \ . NW \ A \
, AAAI AK ANN TN
; \ \\ NY \
N\\ \ \ “
(16.00-36.17) EERE \\ AN \ 7
AN WAR AW \ AN \ \\ \\\ . WY) oF ’
Percent Failing (Score Below 70)
LO4a
Appendix D
99. Dr. Mann’s office then prepared bar charts with
the same job-performance quintiles as before, but showing
test-passing rates separately by race for the persons in that
quintile. /d., {4 112-113. The chart for the 1990 Chauffeur
test, taken from Attachment 15 to the Affidavit, is set forth
below.
100. For firefighters in the highest-performing group,
43.42% of whites, compared to 38.4% of Hispanic-
Americans and only 10.53% of African-Americans, passed
the test. Whites passed the test more than four times as
often as African-Americans performing on the job at the
same high level
101. For firefighters in the lowest-performing group,
16.36% of whites, compared to 0% of Hispanic-Americans
and only 3.03% of African-Americans, passed the test.
102. In every one of the five job-performance bands,
whites passed the test substantially more often than Hispanic-
Americans, and substantially more often than African-
Americans, who were performing their jobs at the same level.
103. Whites in the poorest-performing group of
fire-fighters passed the test at a rate higher than African-
Americans in the dest-performing group of firefighters.
1OSa
Appendix D
1990 Chauffeur Test: Percent Passing
by Race by Chauffeur Composite Rating
All Rated Test-Takers
RR 10.53%
Highest 20% 4T38FGtiee so) ga. sey
(51.00-60.00)
——— + in
NYY
yy
4th 20%
(46.33-50.83)
Middle 20%
(41.83-46.17)
0.00% of 18
2nd 20% 423 17.39%
NY 3.03%
Lowest 20% ~ 0.00% of 18
(16.00-36.17) as 16.36%
0.00% 20.00% 40.00% 60.00% 80.00% 100.00%
——;
Percent Passing (Score 70 or Above)
OY Biack C_) Hispanic HB White
nm * 537, includes Biacks, Hispanics
and Whites Only
4 . : ie. pe i ps
=
106a
Appendix D
104. Dr. Mann’s office also prepared a series of
charts showing the range of written test scores for firefighters
within each of the five Chauffeur-composite performance
quintiles.** The chart for the 1990 Chauffeur test 1s reproduced
below. It shows that, among the highest-performing fifth of
firefighters, the test scores obtained range from a low of
26.25 to a high of 90. Among the lowest-performing fifth
of firefighters, the range of test scores ran from 30 to 87.5.
There was a great deal of overlap in the ranges of test scores
obtained by firefighters in the five different performance
bands. Median test scores, i7.e., the points at which half
the test scores are above the point and half are below, are
indicated by a horizontal line. /d., 4/4] 2-5.
23. April 20, 1991 Supplemental Affidavit of Dr. Mann,
plaintiffs’ exhibit 19.
Buney syisodwo5 snayyneyg a
(00°09-0019) (€@'09-Ce'gr) (L\'9F-E8 IP) (£9'tr-09' 9¢) (4t°9€-00'91)
ROS [eayByy BOS Ypy OZ SIPPING B%OZ puz BOOZ 189M07}
—— Tr SS TY — CO00'0
92°92 5 heeaieee
i 00'O0€
o9'ze
lnc o9'Le [|
a + 00°'0r
- 4S
pond v 00°09
~~ Ty
Pring
| + 00°08
00°06 00 06 oe'ee ose
00°96
—_ eeyesrarineessnineeessnensineneveieinse 0'96 nents seneenesheninnierenernennenini Senebicicall 00 001
}SOL unayyneyy
SaYel YSOL payer jy
SINUIND Buney ayisodwoy snayyneyo Aq
9409S winuixew pue wnwiuiy inayyneyd O66L
108a
Appendix D
105. The same type of chart was then prepared with
the information broken down by race. /d., 4 6-7. The leftmost
vertical bar in each cluster shows information for African-
Americans, the middle vertical bar in each cluster shows
information for Hispanic-Americans, and the rightmost
vertical bar in each cluster shows information for whites.
The chart for the 1990. Chauffeur test 1s reproduced below:
(
|
Buney ssodwoy sNeyyneuy Peieeipey evipen
AlVO COLlUM F SO)WOdeIH ‘eRoRg “20geu
{00'08-0019) (C8 OS 4EC'OPr) (Zi'9r-E9 1b) (£9'lr-09'9e) (4t°9€-60'91)
BOZ JeoYB Hy BOZIYID BOZ @IPPIW OZ pug *%OZ }SBMOq
J
MH @
109a
~
™—
ais
‘——~
~~
~
~
bmg
NB)
~
“~
~
a)
—
—oor
SOy sNayyneyo
SJOHe]-]Say_ payey jy
BINHUIND Buney ayisodwoy inayyneyg Aq
9109S wnwixeyy pue wnuwiuiyy Inayjneuy OBEL
110a
Appendix D
106. Dr. Mann stated his observations as follows
in 4 7:
Across all of these eight charts, the top of the
range of written test scores obtained by whites
within a performance-rating group exceeded the
top of the range of written test scores obtained
by blacks within the same performance-rating
group in all but one combination of test and rating
group.” For all but two combinations of test and
rating group,” the top of the range of written test
scores obtained by whites within a performance-
rating group exceeded the top of the range of written
test scores obtained by Hispanics within the same
performance-rating group.
107. The characteristics described above are not
usually considered the hallmarks of valid and fair tests.
108. Dr. Mann stated in his Affidavit, 4 115, that,
based on the types of charts described above, and considering
the information provided in the analysis of the City’s expert,
he saw no statistical justification for the passing cut-off score
of 70 on the chalienged promotional tests, or for making
promotions of test-passers in rank order of their test scores
and seniority points.
24. The exception was the lowest 20% rating group on the 1990
Chauffeur test.
25. The exceptions were the 1988 Chauffeur test, 4th 20%
group, and the 1990 Chauffeur test, middle 20% group.
Llla
Appendix D
109. Plaintiffs have shown substantial doubt as to
the job-relatedness of the challenged examinations. It is not
necessary for the Court to review this evidence in further
detail.
. The objectors have not presented any credible
ev oe . iene relatedness of the challenged examinations.
F. The Remedial Promotions
111. Under the terms of the Consent Decree. A frican-
Americans and ha ee who took an examination
for Sergeant from January 1, 1982 to date. and who passed
at least one examination ah this rank, will receive a total of
96 remedial promotions: African-Americans and Hispanic-
Americans who took an examination for Sergeant from
January 1, 1982 to date, and who were promoted after a
discriminatorily long waiting period which delayed their
ability to compete for Lieutenant promotions will receive
five remedial promotions to Lieutenant: and A frican-
Americans who took an examination for Lieutenant from
January 1, 1982 to date, and passed at least one examination
for this rank, will receive a total of 5 remedial promotions.
The named plaintiffs in this lawsuit who meet the necessary
conditions shall have priority in receiv ing these promotions,
and the remainder of the promotions shall be made in rank
order from among those unpromoted A frican-Americans and
Hispanic-Americans passing at least one of these tests, in
accordance with the schedule set forth in the Consent Decree,
as modified below. There will not be any back pay.
112a
Appendix D
112. The number of remedial promotions is the
same as the number of promotions which were lost to A frican-
Americans and Hispanic-Americans because of the adverse
impact of the promotional tests from 1982 through 1991.
The Court’s ruling that this case would be limited to promo-
tional examinations given from January 1, 1982 on means
that there are no remedial promotions as a remedy for any
adverse impact of the promotional examinations prior to
1982.
113. The parties have directed the attention of the
Court to an error in Table 18 on p. 22 of the Consent Decree.
The total of 62 remedial promotions to the rank of Sergeant
for African-Americans 1s the agreed total, and represents the
aggregate shortfall in promotions for African-Americans from
1982 through 1992. However, the allocation for each
particular test is based on the shortfall for that test, which
leads to a larger number. See the explanation in 4 17 of the
Consent Decree. Accordingly, the allocations of remedial
promotions of African-Americans to Sergeant in Table 18 of
the Consent Decree must be adjusted downwards, so that
the aggregate number of such promotions remains 62. The
adjustments will have the least impact on individuals if done
on those tests with the largest allocations. Accordingly, Table
18 is revised as follows:
lL 13a
Appendix D
Table 18. Allocation of _the Remedial Promotions to
Sergeant
Number of Vacancies
Sergeant Examination Blacks Hispanics
9/23/82 5 ()
11/17/83 8 7
1/31/85 10 7
10/29/87 7 7
4/27/89 10
10/31/9] 16 10
Total 62 34
114. The Consent Decree provides the mechanism
by which the class members receiv Ing remedial promotions
are to be identified. Plaintiffs will prepare the tentative list
according to those rules. subject to checking by the City
In the event that the parties are unable to agree, the Court
can resolve any disputes. There is no indication. however, of
any reason why the parties should be unable to reach
agreement on the application of these rules.
115. Some objectors have raised the concern that
normal promotions will cease because of the remedial
promotions. This is not correct.
116. Some objectors have raised the concern that
police officers may be demoted in order to make remedial
promotions possible. This is not correct. The parties have
lla
ippendix D
jointly represented to the Court that there will never be any
demotion or layotfin order to make any remedial promotion
possible. The proposed Consent Decree does not contain any
“bumping” provision
117. Some objectors have complained that the
particular officers who will receive the remedial promotions
have not vet been identified and their names made public
The publicizing of the names would add nothing to the
question of the fairness of the proposed Consent Decree, and
} } 2 .7 . . .
this objection ts reyected
LL8. Some objectors have complained that it 1s not
clear at What point in the year any set of remedial promotions
will be made. The Consent Decree provides the Police
Department with the tlexibilify to make these promotions
within a given year when it sees fit. There 1s no unfairness to
any officer in allowing the Police Department to exercise
tlexrbility
119. Some objectors have complained that they do
not know the source of the positions to be filled by remedial
promotions. The source of their concern ts that the remedial
yromotions will fill some vacancies which would otherwise
a
ye filled from promotional registers
120. The parties have gone to great lengths to
accommodate the promotional expectations of officers who
are not members of the plaintiff class
a. they have not proposed a promotional goal in
which any particular percentage of future promotions
Will be reserved for Atrican-Americans or Hispanic-
\mericans;
b. they have agreed to a settlement in which there
is Stull likely to be adverse impact against them on future
examinations. The degree of the adverse impact should
be ameliorated by the provisions of the Decree, but it
is not likely to be eliminated entirely;
c. the parties have not proposed that all of the
remedial promotions be made immediately, but have
agreed to a five-year schedule, with some flexibility in
the Police Department to shorten the schedule if it sees
fit. Given that some class members have been waiting
eleven or more years for their promotions already, this
is a concession of great magnitude. This concession
allows normal promotions of other officers to proceed
on a regular schedule, even if in reduced numbers
l-1. Some officers who took some of the challenged
promotional examinations object to the remedial promotions
of Atrican-American and Hispanic-American officers who
took the same remedial examinations, on the ground that such
Atrican-American and Hispanic-American officers received
lower scores on the examinations than they received. They
object that they are being by passed. F.g., objection of Charles
Johnson, Tab 52, paragraph 6 in Vol. Il of the objections to
the settlement. The remedial promotions are intended by the
parties to provide a partial remedy for examinations which
have been shown to cause substantial adverse impact against
blacks and Hispanics, in circumstances in which plaintiffs
have presented substantial evidence that the examinations
ll6a
Appendix D
are not job-related. There 1s nothing unfair to white promo-
tional candidates on the same examinations that they do not
share in the remedial promotions, because there was never
any adverse impact against whites on these examinations such
that remedial promotions would be justified. Moreover,
plaintiffs have presented substantial evidence tending to show
that relative differences in test score are not a meaningful
indicator of comparable differences in job performance.
122. The remedial promotions are narrowly tailored
to accomplish the purpose of remedying plaintiffs’ claims of
testing discrimination. They do not unduly trammel the
interests of police officers who are not members of the
plaintiff class.
G. Remedial Seniority
123. The proposed Consent Decree provides that
class members on the remedial promotion list, including those
who will be barred from receiving remedial promotions
because they can be reached on a register of eligibles, will
receive back seniority in rank to compensate them for some
of the injury suffered because of their delays in receiving
promotions.
124. Some objectors have complained that allowing
remedial seniority to class members harmed by these tests in
the past will reduce the benefit of their own seniority. This
is a natural consequence of any seniority relief.
125. Remedial seniority 1s necessary to place class
members in the position they would have held if there had
ll 7a
Appendix D
never been any testing discrimination. This relief is narrow ly
tailored to accomplish its remedial objective, and does not
unduly trammel the interests of other officers.
H. The Reduction of Adverse Impact in the Future
126. The proposed Consent Decree seeks during the
next ten years to reduce the amount of adverse impact against
AtricanAmericans and Hispanic-Americans taking examina-
tions for Sergeant and for Lieutenant (a) by striking “biased
items’’, 1.e., those questions identified by a statistical analysis
devised by the City’s expert in the Houston ( ‘hapter case,
which have substantial adverse Impact against any race, white
as well as African-American or Hispanic-American, and
Which are not equally predictive of job performance for all
races, and (b) by extending the life of promotional registers
during this period of time to two years.
127. In the Houston Chapter case, Dr. Mann attempted
to estimate the possible effect of this procedure on the
reduction of adverse impact in the future. He applied the
procedure to the results of the 1983, 1984, 1988 and 1990
Chauffeur examinations, and to the 1982, 1984. 1988 and
1990 Junior captain examinations. The analysis was a
hypothetical analysis for each of these eight sets of examina-
tion data. Hypothetical new scores were obtained. and the
test-takers’ then actual seniority points were added to the
hypothetical scores in order to obtain a hypothetical new set
of rankings.
28. While application of the new procedure reduced
the adverse impact of these examinations. adverse impact
llSa
Appendix D
against African-Americans and Hispanic-Americans still
remained. Some questions were hypothetically eliminated
from the test because they worked to the detriment of whites
and were not valid items
129. No test item will be stricken because it has
adverse impact against any group, /.e., because the proportion
of white test-takers answering it correctly is much larger than
the proportion of African-American or Hispanic-American
test-takers who answer it correctly. A test item will be struck
only if it 1s a biased item. Test items with large racial
differences in the rates at which particular groups answer
them correctly will remain on the test if they are valid items.
130. The elimination of biased items improves the
quality, of the test, and makes it a better predictor of job
performance. In Dr. Friedland’s phrase, it helps to remove
race as a factor in the test scores.
131. Asian-Americans and women have asked to be
included in the analyses of biased test items. The City is free
to include them in the analyses whenever there are sufficient
numbers of Asian-Americans and women to make such an
analysis meaningful, but this 1s a voluntary decision for the
City. This case does not encompass discrimination against
Asian-Americans or against women.
132. The extension of the test registers to two years
benefits every officer who passes the test, because it increases
their opportunity for promotion
119a
ippendix D
133. The provisions in the Consent Decree providing
a means for the elimination of biased items are narrowly
tailored to accomplish the remedial purposes of the Consent
Decree, and do not unduly trammel the interests of any
officers who are not members of the class.
The Provisions of State Law
134. The proposed Consent Decree supersedes some
provisions of the Fire and Police Civil Service Act. Texas
Local Government Code chapter 143, as amended. [here-
inafter, “Chapter 1430°’] while lea Ing many other provisions
unaffected.
135. Plaintiffs have shown substantial doubt as to the
job-relatedness of the City’s promotional examinations for
Sergeant and-Lieutenant. Nevertheless, the proposed Consent
Decree allows the City defendants (a) to continue compiling
promotional examinations for these ranks in the manner
set forth in Chapter 143; (b) to continue using a passing
cut-off score of 70% correct, as provided by Chapter 143:
(3) to continue the use of seniority points as provided |
Chapter 143; and (4) to continue u sing test scores, in conjunc-
tion with the seniority points, to rank-order promotional
candidates, as provided in Chapter 143.
136. The Consent Decree also allows the City to
explore, and to consult with all other employee groups as
well as with plaintiffs, alternative selection procedures as
authorized by Chapter 143.
120a
Appendix D
137. Paragraph 55 of the proposed Consent Decree
makes two major changes and several minor changes trom
the operation of Chapter 143. First, it provides a mechanism
for striking from the test those items which the City identifies
as biased items, /.e., those which have a high degree of
adverse impact against any race and which are not job-related
for all races. This is an expansion of the power the city already
has under chapter 143 to strike improper test items or to score
additional responses as correct. The result should be to
improve the test and make it a ™ore accurate predictor of
job performance in the promotional rank. This is a necessary
change for the reduction of adverse impact from these tests,
and is compatible with the purposes of merit selection
underlying Chapter 143.
138. Second, 4 55(g) of the proposed Consent Decree
provides that promotional registers of test-passers resulting
from use of the new selection procedures shall remain in
effect for a period of two years unless such lists are earlier
exhausted. Because the disparate impact of the challenged
examinations results in whites scoring disproportionately
at the top of promotional registers, and passing African-
American and Hispanic-American candidates scoring
disproportionately lower down, a one-year extension in the
life of these promotional registers 1s a necessary change for
the reduction of adverse impact from these tests. This remedy,
and the striking of biased test items, are far less burdensome
to nonminorities than an injunction against the tests and a
requirement that the City develop entirely new selection
procedures.
l2la
Appendix D
139. Third, there are several minor changes.
The need to perform statistical analyses of responses and
to examine test items for bias necessitates a short post-
ponement of release of the test results. Here, the policy of
Chapter 143 has been preserved to the greatest possible
extent by the requirement in 4 55(f) of the proposed Consent
Decree that the City post an eligibility list as soon as
possible. The provision of § 55 (a) of the proposed Consent
Decree requiring ninety days’ notice of a firm date for the
examination and of the books to be studied is intended
to ensure that all officers have an adequate and definite
time period within which to study for the examination, and
is fully compatible with the purposes underlying Chapter 143.
The Airport Police
140. Terry Hughes and the Houston Airport Police
Officers’ Association have filed an objection to the Consent
Decree which is limited to the provisions of Paragraph 61
entitled “Reclassification of Peace Officers.” The essence of
their objection is that Paragraph 61 will deny the Airport
Police Officers the right to transfer and become Classified A
Officers in the Houston Police Department. This claimed
right stems from the judgment of the 280th District Court
of Harris County. Texas, allegedly granting them equal
transfer rights. (Plaintiffs’ Exhibit No. 21). The Court does
not conclude that this judgment grants any right which
Paragraph 61 curtails. However, even if it does curtail any
existing right, Paragraph 61 is essential to the viability
and practicality of the Consent Decree. The Decree strikes
a delicate balance between the competing interests of
- minorities receiving remedial promotions with retroactive
122a
Appendix D
seniority benefits and of nonminorities seeking to obtain
promotions through the regular competitive process who are
concerned with preserving their own seniority rights under
the civil service system. In light of these competing interests,
the Court finds that Paragraph 61 will be essential to maintain
that balance and avoid additional disruption by the merger
of other specialized police divisions as Class A officers. Such
a restric-tion as contained in Paragraph 61 is reasonable and
necessary to the implementation of the Consent Decree which
is Of a limited duration.
K. Objections that the Proposed Consent Decree Does
Not Provide Sufficient Relief
141. McCloy Medlock, Richard Humphrey, Willie
Fields, and Bennie L. Green, four of the plaintiffs in the
Comeaux action, have filed a Motion to Substitute Attorney,
Objections to Order of Dismissal, Objections to Consent
Decree, and First Amended Objections to Consent Decree.*®
The First Amended Complaint in Comeaux, filed on
December 29, 1973, described their individual claims at 5-6
as follows:
Plaintiff Fields would show that although he passed
the Detective’s examination, he was denied a pron tion
because of the HPD’s discriminatory promoii: al
practices. In addition, plaintiff Fields was denied a
transfer to another department for which he was fully
qualified and was denied because of his race.
..
26. Objections to Proposed Consent Decree, vol. II, tab 60
3a
Appendix D
Plaintiff Green would show that he was not treated
equally with White officers in the Department with
regard to terms and conditions of employment solely
on the basis of his race and has suffered because of
Defendants’ subjective efficiency rating system which
affects Blacks disparately and limits their mobility
within the Department.
Plaintiff Humphrey would show that although
he took the Lieutenant’s exam and scored well. he was
denied a promotion to Lieutenant because of the HPD’s
subjective testing and qualify Ing practices and
procedures.
Plaintiff Medlock would show that even though he
took the Lieutenant’s examination and scored well. he
was denied a promotion to Lieutenant because of HPD’s
discriminatory promotional practices and procedures
utilizing subjective tests and efficiency ratings.
142. Plaintiffs’ exhibits 9 and 10 show the following:
a. Willie E. Fields failed the Sergeant tests given
in 1975 and 1977, passed the 1979 Sergeant test with a
rank of 97, and failed the Sergeant tests given in 1983
and 1985.
l24a
Appendix D
b. Bennie Lee Green passed the 1979 Sergeant test
with a rank of 86, and failed the 1980, 1981, and 1983
Sergeant tests.
c. Richard C. Humphrey passed the 1976 and 1977
Lieutenant tests with ranks of 69 and 66, respectively
It is unclear whether he passed the 1978 Lieutenant test.
He failed the 1979 and 1980 Lieutenant tests.
d. McCloy Medlock passed the 1975 Lieutenant
test with a rank of 45, passed the 1976 Lieutenant test
with a rank of 43 (“Meloy Medlock” is probbaly the
same person as McCloy Medlock), and passed the 1979
Lieutenant test with a rank of 26.
143. The case in which these objectors were named
plaintiffs had had no active prosecution apparent on the
docket entries for a decade prior to the attempt by the
plaintiffs herein to intervene in the lawsuit. It would not be
reasonable to delay or cancel the proposed relief for persons
who actively sought to press their rights, in favor of persons
who did not press their rights. The Court makes no findings
herein as to the relative responsibility of client and counsel
in Kelley for the failure to press that action, but will not allow
this objection to stand in the way of final approval of the
settlement.
144. Plaintiff Jammer objects to the provision of
‘| 37 of the Consent Decree stating that no plaintiff or class
member can receive a remedial promotion unless he or
she passed one of the challenged tests. Plaintiff Jammer
was promoted to Sergeant on October 15, 1988. Thereafter,
he failed the 1991 and 1992 Lieutenant tests.
125a
Appendix D
145. Recognizing that there are serious problems with
the promotional tests and their passing level, but that there
is no presently available alternative means of demonstrat-
ing basic qualification for promotion in the eyes of other
members of the Department, only plaintiffs and class
members who have passed a test will be eligible to receive a
remedial promotion. It is an imperfect mechanism, and in its
application to an individual may incorrectly disqualify that
individual from relief. It is the only available tool for the
purpose, however. In light of the number of remedial
promotions, this is important. The compromise embodied
ad
in {| 37 of the Consent Decree is appropriate.
L. Procedural Fairness, and General Matters
146. The evidence shows that the proposed Consent
Decree was arrived at through arm’s-length bargaining among
well-represented parties.
147. Any conclusion of law which is more appro-
priately a finding of fact shall be considered a finding
of fact.
I]. Conclusions of Law
A. General
1. The Court has jurisdiction over this lawsuit.
2. The certification and definition of the plaintiff
class under Rule 23(b)(2), F.R.Civ.P., remains appropriate,
and is hereby made final.
126a
Appendix D
3. The Notice provided herein meets the standards
of the Constitution and of Rule 23, F.R.Civ.P.
4. Any finding of fact which is more appropriately
a conclusion of law shall be considered a conclusion of law.
B. The Legal Standards Applicable to the Approval of
a Consent Order Providing for Race-Conscious
Relief
5. The proposed Consent Decree involves two types
of race-conscious provisions. First, it provides for 106 remedial
promotions. Second, it provides for a ten-year period of time
in which the elimination of future test questions for the ranks
of Sergeant and Lieutenant biased against any race — whites,
African-Americans or Hispanic-Americans — according to
a methodology used by the City defendant’s expert
psychologist in the Houston Chapter case.
6. Voluntary affirmative-action programs memo-
rialized in consent decrees are reviewed under the same
standard as other voluntarily adopted affirmative-action
plans. The Supreme Court’s reasoning in Local 28 of the
Sheet Metal Workers’ Int'l Ass'n v. E.E.O.C., 478 U.S. 421,
92 L.Ed.2d 344, 106 S.Ct. 3019 (1986), compels that the
two types of voluntary plans be reviewed under the same
test. In that case, the Court held that the fact that a voluntary
affirmative action program was embodied in a consent decree
did not render the program subject to § 706(g) of Title VII
— the provision limiting the types of remedial orders coyrts
can enter in Title VII suits — and that parties could voluntarily
agree to a remedy that a court could not enter at the conclusion
127:
Appendix D
ofa Title VII litigation. The Court explicitly excluded consent
decree programs from $ 706(g) because they essentially
mirrored the type of voluntary program approved in Johnson
v. Transportation Agency, 480 U.S. 616, 94 L.Ed.2d 615.
107 S.Ct. 1442 (1987): it follows that the two kinds of
programs should be reviewed under the same legal standard.
7. All courts of appeal which have decided this
question have reviewed the two types of programs under
the same standard. See Davis v. City and County of San
Francisco, 890 F.2d 1438 (9th Cir., 1989): Howard y.
McLucas, 871 F.2d 1000, 1006 (11th Cir.. 1989): In re
Birmingham Reverse Discrimination Employment Litigation,
833 F.2d 1492, 1501 note 23 (11th Cir., 1987). But see Mann
v. City of Albany, 883 F.2d 999 (11th Cir., 1989) (declining
to comment on the issue).
8. Johnson crystallized the legal standard for deter-
mining whether a race-conscious affirmative action plan
comports with Title VII of the Civil Rights Act of 1964.
Justice Brennan, writing for the Court, first noted that the
party challenging an affirmative action plan bears the burden
of proving the plan’s invatidity. 480 U.S. at 626, 94 L.Ed.2d
at 627. The opinion then articulated a two-prong test for
determining a plan’s validity:
The first issue is ... whether the
consideration of the sex [or race or ethnicity]
of applicants ... was justified by the existence
of a “manifest imbalance” that reflected
underrepresentation of women in “traditionally
segregated job categories.” . . .
r=
* * x
128a
Appendix D
We next consider whether the Agency Plan
unnecessarily trammeled the rights of male [or
nonminority] employees or created an absolute bar
to their advancement.
480 U.S. at 631, 637-38, 94 L.Ed.2d at 630, 634 (citations
omitted). Though the decision does
This text is long and has been trimmed here. Open the source document for the complete record.
This is a copy of a public record, reproduced as it was published. It is not legal advice, and it may not be the version a court would rely on. Check the official source before you cite it.