United States Court of Appeals
                         FOR THE EIGHTH CIRCUIT
                                 ___________

                                 No. 04-2060
                                 ___________

John M. Kohlbek,                      *
                                      *
            Plaintiff/Appellant,      *
                                      *
William Schrack,                      *
                                      *
            Plaintiff,                * Appeal from the United States
                                      * District Court for the
Michael Pritchard,                    * District of Nebraska.
                                      *
            Plaintiff/Appellant,      *
                                      *
      v.                              *
                                      *
City of Omaha, Nebraska,              *
a Municipal Corporation,              *
                                      *
            Defendant/Appellee.       *
                                 ___________

                           Submitted: May 12, 2005
                              Filed: May 1, 2006
                               ___________

Before WOLLMAN, BYE, and COLLOTON, Circuit Judges.
                           ___________

WOLLMAN, Circuit Judge.

    John Kohlbek and Michael Pritchard appeal from the district court’s grant of
summary judgment on their reverse discrimination claim seeking declaratory and
injunctive relief and damages against the City of Omaha (Omaha).1 We reverse and
remand.

                                             I.

       In 1971, fourteen years after Omaha declared that the fire department would be
racially integrated, it instituted the first of a series of affirmative action plans for the
department.2 The most recent plan, and the one at issue here, is the 2002 Affirmative
Action Plan (2002 Plan).3 Omaha developed the 2002 Plan consistent with its
interpretation of the Office of Federal Contracting Compliance Programs Guidelines
on Affirmative Action Programs (Guidelines). See 41 C.F.R. § 60-2.1 et seq. (2000).
The Guidelines are not binding on municipalities. See 

60-1.5(4). The
Guidelines specify that “availability,” which is “an estimate of the number of qualified
minorities . . . available for employment in a given job group, expressed as a
percentage of all qualified persons available for employment in the job group,” must
be determined for each job group. 41 C.F.R. § 60-2.14.

        The Guidelines require that the determination of availability in a given job
group be based on at least two factors: (1) the percentage of minorities “with requisite
skills in the reasonable recruitment area”; and (2) the percentage of minorities among
those “promotable, transferable, and trainable” within the organization. 41 C.F.R. 60-


        1
        Plaintiff William Schrack was voluntarily dismissed by the district court and
is not part of this appeal. D. Ct. Order of March 30, 2004, at 1 n.1.
        2
       The instant case is the latest chapter in a series of disputes about the Omaha
Fire Department’s hiring and promotion practices. We previously rejected a legal
challenge to one of these affirmative action plans in Warsocki v. City of Omaha, 

(8th Cir. 1984).
        3
         The 2002 Plan, by its terms, expired on December 31, 2005. (Joint App. at
421).

                                            -2-
2.14(c). Omaha characterized these factors respectively as the “external availability
percentage” and the “internal availability percentage.” Omaha Br. at 11. Omaha
based its analysis of job groups containing promotion positions solely on the internal
availability percentage “because all promotions in the Department are limited to
internal candidates.” D. Ct. Order of March 30, 2004, at 7. Omaha calculated the
internal availability percentage for African-Americans to be 11.4% at the fire captain
position and 5.8% at the battalion chief position. The actual percentages of African-
Americans in those positions were 7.7% and 3.1%, respectively.

       A placement goal must be established under the Guidelines “[w]hen the
percentage of minorities . . . employed in a particular job group is less than would
reasonably be expected given their availability percentage in that particular job
group.” 41 C.F.R. § 60-2.15. Omaha described the circumstances requiring a
placement goal as “underutilization.” Omaha determines that an underutilization
exists whenever the minority representation in a particular position is less than the
goal. If the number of individuals needed to meet a particular goal does not equate
exactly to a whole number, the number is rounded, so that 0 to 0.4 does not constitute
a person and 0.5 and above equates to a full additional person. This rule was referred
to by Omaha’s expert as a “half person rule,” and is a hybrid of the “any difference”
and “whole person” rules.

       Pursuant to the fire department’s promotion process, candidates who pass both
a written and a practical test are placed on an eligibility list in rank order of their
scores. When a promotion position opens, the fire department’s personnel director
sends the names of the top five candidates on the eligibility list to the fire chief. If
there are multiple openings, the personnel director refers three times as many
candidates as there are openings. In the event that there are several candidates tied for
the last referral position, the personnel director refers them all. If a position is
underutilized, the personnel director is required to refer minority candidates from the
eligibility list who are not among the top three candidates. The fire chief individually

                                          -3-
reviews each of the candidates referred to him by the personnel director based on test
scores, ranking, seniority, educational background, discipline record, job performance,
and attendance. When there is an underutilization, race is also taken into account by
the fire chief.

       Kohlbek passed the August 11, 2000, promotion exam for battalion chief. He
was ranked eleventh on the eligibility list, and Anthony Curtis, an African-American,
was ranked twentieth. On November 9, 2002, following promotions of other higher-
ranked candidates, Kohlbek was ranked second on the eligibility list, and Curtis was
ranked below him. The fire chief selected Curtis to fill a recently vacated battalion
chief position. The chief stated that he based his decision to promote Curtis on “a
variety of factors” including the fact that “at the time Omaha had only one black
Battalion Chief out of 28 and the Department was underutilized in that position.” Aff.
of Joe Napravnik at 2 (Joint App. at 670). Six days later, another battalion chief
position opened, and the chief selected the top-ranked candidate on the eligibility list
for the position. Had the chief made all of his promotion decisions by rank order only,
the top-ranked candidate would have been promoted on November 9, 2002, and
Kohlbek would have been promoted six days later.

      Pritchard and fifty-one other candidates passed the written exam for the position
of department captain. Pritchard was ranked eighth on the eligibility list. Two
African-American candidates, Ron Estes and Michael Andrews, were ranked thirty-
eighth and thirty-ninth, respectively. Subsequent to being placed on the list, Pritchard
was offered promotion to and declined the position of paramedic captain.4 On


      4
       The fire chief testified that regular captains and paramedic captains have the
same rank and are filled from the same eligibility list. Omaha argued below that
Pritchard’s decision to decline the offer of promotion to paramedic captain precluded
him from claiming injury from the later race-conscious promotions to the regular
captain positions. The district court assumed for purposes of deciding the summary
judgment motion that the promotions of Estes and Andrews were to Pritchard’s

                                          -4-
November 18, 2002, following promotions of other higher-ranked candidates,
Pritchard was ranked second on the eligibility list. Estes and Andrews, who were then
ranked thirty-second and thirty-third, were promoted out of rank order. Had the chief
made all of his promotion decisions by rank order only, Pritchard and the top-ranked
candidate on the list would have been promoted in lieu of Estes and Andrews.

       The fire chief testified that he probably would not have promoted out of rank
order absent the affirmative action plan. Dep. of Joe Napravnik at 39, 67-68. The
district court agreed. D. Ct. Order of March 30, 2004, at 8 (“Plaintiffs contend that
they would have been promoted but for the consideration of the race of some other
promotional candidates. All other things remaining the same, that is a contention that
appears to be supported by the record.”). Thus, to determine the legality of the
promotion decisions at issue here, it is necessary to evaluate whether the racial
classifications used in making promotional decisions under the 2002 Affirmative
Action Plan are constitutional.

                                          II.

       We review de novo the district court’s summary judgment decision. Bockelman
v. MCI Worldcom, Inc., 

, 531 (8th Cir. 2005). We apply strict scrutiny
to all governmental distinctions on the basis of race. See Grutter v. Bollinger, 

, 326 (2003) (“We have held that all racial classifications imposed by
government ‘must be analyzed by a reviewing court under strict scrutiny.”); cf.
Hirabayashi v. United States, 

, 100 (1943) (“Distinctions between citizens
solely because of their ancestry are by their very nature odious to a free people whose
institutions are founded upon the doctrine of equality.”). Applying strict scrutiny, a
racial classification is constitutional, under the Equal Protection Clause of the




detriment.

                                         -5-
Fourteenth Amendment, only if it is a narrowly tailored measure that furthers a
compelling governmental interest. 

.

        Omaha contends, and the district court agreed, that the 2002 Plan is narrowly
tailored to further a compelling interest in remedying the effects of past racial
discrimination. The district court held, as a matter of law, that Omaha had
demonstrated a compelling governmental interest in remedying past discrimination
in its hiring of firefighters. D. Ct. Order of March 30, 2004, at 16-17. The district
court further held that “the 2002 Plan is narrowly tailored to achieve the governmental
interest in remedying past discrimination.” 

We disagree, and conclude that
the 2002 Plan was not narrowly tailored to remedy specifically identified past
discrimination.5

       “In determining whether a race-conscious remedy is narrowly tailored, we look
at factors such as the efficacy of alternative remedies, the flexibility and duration of
the race-conscious remedy, the relationship of the numerical goals to the relevant
labor market, and the impact of the remedy on third parties.” Sherbrooke Turf, Inc.
v. Minn. Dept. of Transp., 

, 971 (8th Cir. 2003) (citing United States v.
Paradise, 

, 171, 187 (1987) (plurality and concurring opinions)). In
examining these factors, we must ensure that the racial classifications are not used any
more broadly than the asserted compelling interest requires. 

(“[R]acial classifications, however compelling their goals, are potentially so
dangerous that they may be employed no more broadly than the interest demands.”);
cf. 

(“[T]o be narrowly tailored, a national program must
be limited to those parts of the country where race-based measures are demonstrably
needed.”) (emphasis removed). To ensure narrow tailoring, we must conduct “a most
searching examination” requiring “the most exact connection between justification


      5
       We do not reach the issue of whether Omaha is estopped from claiming an
pattern of past discrimination because of its prior litigation positions.

                                          -6-
and classification.” Gratz v. Bollinger, 

, 270 (2003); Adarand
Constructors, Inc. v. Pena, 

, 236 (1995).

        The relationship of Omaha’s numerical goals to the relevant labor market is not
sufficiently exact to withstand this searching examination. Cf. In re Birmingham
Reverse Discrimination Employment Litig., 

, 1549 (11th Cir. 1994)
(rejecting a plan with no relationship between the numerical goals and the relevant
market as not narrowly tailored). Because Omaha employs the use of racial
classifications in situations where there is no identified past discrimination, we hold
that its affirmative action plan, as it applies to promotional decisions, is not narrowly
tailored to further the goal of remedying past discrimination.6 See Ensley Branch,
N.A.A.C.P. v. Seibels, 

, 1577 (11th Cir. 1994) (holding that each specific
use of racial classifications in determining promotions must be necessary to cure
employment discrimination).

       For a remedial racial classification to be justified, it is necessary that a prima
facie case of discrimination be established. See City of Richmond v. J.A. Croson Co.,

, 500 (1989) (rejecting claims of a compelling interest in remediating
discrimination when no prima facie case could be established). Omaha uses statistical


      6
        Appellants have raised several substantial concerns regarding the
constitutionality of the method used to calculate the external availability percentages,
calling into question Omaha’s use of racial classifications in entry-level hiring.
Appellants, however, do not assert any cognizable injury suffered as a result of
Omaha’s entry-level hiring practices. Thus they do not have standing to challenge
those practices here. See Minnesota Citizens Concerned for Life v. Federal Election
Comm'n, 

, 131 (8th Cir.1997) (requiring plaintiff to show causal
relationship between challenged conduct and injury to establish standing). As a result,
we consider only the racial classifications used by Omaha in making promotional
decisions. Cf. Donaghy v. City of Omaha, 

, 1456 (8th Cir. 1991)
(refusing to consider racial classifications that did not result in any injury to the
plaintiffs).

                                          -7-
evidence to distinguish between those situations where discrimination is demonstrated
and those where no discrimination can be presumed. As described above, Omaha
defines “underutilization”—the state in which past discrimination is presumed and
racial classifications are used—based on the half person rule. This rule determines
that any time the actual number of minorities in a particular position is not within half
of a person of the goal, an underutilization exists. If an underutilization is found,
Omaha uses racial classifications in its promotion decisions. It thus becomes
necessary, in determining whether Omaha’s racial classifications are narrowly tailored
to remedy identified past discrimination, to determine whether the method of finding
underutilization truly coincides with situations where discrimination can be legally
inferred.

        Establishing a prima facie case of discrimination based simply on statistical
evidence is difficult. We have recognized that a disparity between actual hiring levels
and expected hiring levels does not necessarily demonstrate discrimination. Taylor
v. Teletype Corp., 

, 1133 (8th Cir. 1981). “Numbers must be
statistically significant before one can properly conclude that any apparent racial
disparity results from some factor other than random chance.” 

cf. Aiken v. City of Memphis, 

, 1163 (6th Cir. 1994) (holding that a
sufficient prima facie case of discrimination was present to justify remedial racial
preferences when statistical evidence showed a “gross” and “wide” disparity). Dr.
Paul White, Omaha’s statistical expert, agreed that a measure of statistical significance
is necessary when comparing actual numbers to expected numbers in an attempt to
infer the existence of discrimination. See White Report, 2-3 (Joint App. at 706-07).

       The half person rule, on the other hand, does not require a statistically
significant showing of discrimination before racial classifications are triggered. As
Omaha’s expert testified, the half person rule triggers racial classifications in more
situations than if a test of statistical significance were used to determine whether racial



                                           -8-
discrimination has occured. See White Dep. at 70 (Joint App. at 877).7 Thus, by
using the half person rule, Omaha, under the auspices of the 2002 Plan, made racial
classifications in its promotional decisions that went beyond its interest in remedying
identifiable racial discrimination. Cf. 

(noting that promotion
goals that go “beyond the elimination of the ‘vestiges of past discrimination’” would
be unconstitutional).

       We need not speak in purely hypothetical terms. The promotional decisions at
issue here were both done at times when there was no statistically significant
difference between the actual and expected number of minorities in the Fire Captain
and Battalion Chief ranks. White Report, 5 (Joint App. at 709). In fact, Dr. White
conceded that, among the thirteen years he studied, he did not find a single year with
a statistically significant difference between the actual and expected number of
Battalion Fire Chiefs, and he found only three years in which there was a statistically
significant shortfall of black Fire Captains. 

recent of these shortfalls
was in 1988, fourteen years prior to the use of racial preferences challenged by
Pritchard. The promotional decisions at issue thus were the result of racial
classifications that were not narrowly tailored to remedy identified past
discrimination.

                                         III.

       Omaha argued in the alternative to the district court, relying on Grutter v.
Bollinger, 

(2003), that its compelling interest in developing and
retaining a diverse workforce in the fire department justified the 2002 Plan. The
district court concluded that it need not decide this issue in light of its holding on


      7
       In his deposition, Dr. White was actually discussing the any difference rule and
the whole person rule, but he acknowledged that the half person rule is simply a blend
of those two standards. 

                                         -9-
Omaha’s remediation rationale and in view of the fact that the record was not
sufficiently developed on the issue. See D. Ct. Order of March 30, 2004, at 17. The
parties have not addressed this issue on appeal, and thus we leave it for the district
court to examine on remand.

       The judgment of the district court is reversed, and the case is remanded for
further proceedings consistent with this opinion.
                      ______________________________




                                        -10-