IN THE UNITED STATES COURT OF APPEALS

                       FOR THE FIFTH CIRCUIT

                       _____________________

                            No. 91-6027
                       _____________________


     EQUAL EMPLOYMENT OPPORTUNITY COMMISSION,

                                    Plaintiff-Appellant,
                                    Cross-Appellee,

                              versus

     OLSON'S DAIRY QUEENS, INC.,

                                    Defendant-Appellee,
                                    Cross-Appellant,

     DAVID T. LOPEZ,

                                    Appellee-Cross-Appellant.

     _______________________________________________________

         Appeal from the United States District Court for
                  the Southern District of Texas
                           (CA H 86 3777)
     _______________________________________________________
                          (March 9, 1993)

Before REAVLEY, KING and WIENER, Circuit Judges.

PER CURIAM:1

     The Equal Employment Opportunity Commission (EEOC) appeals

the district court's judgment that Olson's Dairy Queens, Inc.

(Olson's) had not committed unlawful employment discrimination


     1
        Local Rule 47.5 provides: "The publication of opinions
that have no precedential value and merely decide particular
cases on the basis of well-settled principles of law imposes
needless expense on the public and burdens on the legal
profession." Pursuant to that Rule, the court has determined
that this opinion should not be published.
and awarding Olson's attorney's fees.         We reverse and render as

to Olson's liability and remand for determination of damages.

                               I. BACKGROUND

     The relevant facts underlying EEOC's complaint are outlined

in the district court's opinion.         

, 1217-18.

Our departure from the district court's recitation, and

ultimately its opinion, is based largely upon the testimony of

the EEOC's expert witness, Dr. Mahlon Straszheim, and Olson's

expert witness, Dr. Ira Chorush.

A.   DR. STRASZHEIM'S STUDY.

     Dr. Straszheim analyzed the extent to which Olson's actual

hiring patterns produced a different black-nonblack employee mix

than would be expected if Olson's hiring policies were entirely

race-neutral.   He did so by two distinct means.

     1.   External Availability Analysis.

          The first approach, which was the focus of the district

court's opinion, was to compare Olson's hiring history with the

percentage of black food preparation and service workers in the

relevant labor market from which Olson's draws its work force.

Dr. Straszheim, relying on years of experience in labor,

transportation, and urban economic analysis, determined that the

relevant labor market was the metropolitan Houston area -- more

specifically, the Houston Standard Metropolitan Statistical Area,

or "SMSA," as defined by the United State Bureau of the Census.

          Using detailed census data for the Houston SMSA, Dr.

Straszheim determined that blacks comprise roughly 25.2% of the


                                     2
food preparation and service workers in the Houston SMSA.    By

comparison, only 8.1% of employees of known race hired by Olson's

between 1978 and 1987 were black.

          Dr. Straszheim refined the Houston SMSA figures to

account for travel time to and from Olson's locations and the

average travel times for black food preparation and service

workers reported in the census data.   He also distinguished

between the Spring Branch (6 locations), Bellaire (2 locations),

and Katy (1 location) labor markets.   Based upon the demographics

of each of these distinct markets and the relevant travel times,

Dr. Straszheim concluded that blacks comprised 19.8% of the

relevant labor pool for Olson's Spring Branch and Bellaire

locations, and 8.1% for the Katy store.   By comparison, blacks

comprised 6.5% of hires of known race at Olson's six Spring

Branch stores, 12.3% of hires of known race at Olson's two

Bellaire-area stores, and 9.4% of hires of known race at Olson's

Katy location.

          Employing standard statistical techniques, Dr.

Straszheim concluded that there was less than one chance in

100,000 (.00001) that Olson's observed hiring patterns in the

Spring Branch stores could have resulted from truly race-neutral

hiring practices, and less than three chances in one thousand

(.0026) that Olson's observed hiring patterns in the Bellaire

stores could have resulted from truly race-neutral hiring

practices.   Dr. Straszheim found no statistically significant

difference between the number of blacks hired in the Katy store


                                 3
and the number which would be expected based upon black

representation in the relevant labor market.

     2.   Applicant Flow Analysis.

          As a separate and distinct means of assessing the race-

neutrality of Olson's hiring practices, Dr. Straszheim compared

the percentage of blacks among Olson's applicants of known race

to the percentage of blacks among Olson's hired employees of

known race.    The results of this analysis were completely

disregarded by the district court's opinion and largely ignored

by Olson's own expert, as well as by Olson's counsel in his

argument to this court.

          Between 1984 and 1987, the period for which rejected

applications were available, blacks constituted 29.6% of the

roughly 1,800 applicants of known race.    In the Spring Branch

market, 30.1% of the applicants of known race for the relevant

period were black; 39.5% in the Bellaire market; and 27.6% in the

Katy market.    By comparison, roughly 13.2% of the persons of

known race hired by Olson's Spring Branch stores during the same

time period were black, while blacks constituted 27.3% and 11.1%

of the hires of known race for Olson's Bellaire and Katy

locations, respectively, for the same period.

          In light of the racial mix of actual applications made

to each of the stores, Dr. Straszheim concluded that the

likelihood that Olson's observed hiring patterns resulted from

truly race-neutral hiring practices was less than one chance in

ten thousand (.0001) for the Spring Branch stores, less than


                                  4
seven chances in one thousand (.0070) for the Bellaire stores,

and less than two chances in one thousand (.0020) in the Katy

store.

B.   DR. CHORUSH'S STUDY.

     Dr. Chorush testified that he had requested data from

Olson's Spring Branch stores for April 1990.    He found that, of

the 60 employees working at the six Spring Branch locations in

April 1990, more than one-half lived within one mile of the store

at which they worked, and more than 80 percent lived within three

miles.   He testified that many of Olson's Spring Branch area

employees were high school students and that many were employed

part time.    Dr. Chorush did not quarrel with Dr. Straszheim's

depiction of Olson's Spring Branch area employees as

predominantly nonblack.     Based upon his observations, Dr. Chorush

concluded that "most persons willing to accept positions at

Olson's are young, seeking part-time employment and residing

within a very short distance of the restaurant."     

A.   EEOC'S COMPLAINT.

     To prevail on its claim of disparate treatment, the EEOC

must establish by a preponderance of the evidence that a pattern

of intentional discrimination existed in Olson's hiring of black

applicants.    That is, the EEOC must show that racially

discriminatory hiring was Olson's regular, rather than unusual,

practice.    International Brotherhood of Teamsters v. United

States, 

, 360, 

, 1867 (1977)


                                   5
("Teamsters").    If the EEOC establishes a prima facie violation,

it is incumbent upon Olson's to articulate a legitimate,

nondiscriminatory reason for its hiring patterns.    Id.; McDonnell

Douglas Corp. v. Green, 

, 802, 

, 1824

(1973).    If Olson's articulates an acceptable rationale, the EEOC

bears the burden of showing that Olson's explanation is a pretext

for unlawful discrimination.    

n.50, 

n.50; McDonnell 

, 93 S. Ct.

at 1825.

     1.     Prima Facie Violation.

            The EEOC may establish a prima facie violation of Title

VII through statistical evidence, evidence of Olson's treatment

of individual job applicants and employees, or both.    See

Hazelwood Sch. Dist. v. United States, 

, 

(1977); see also Pouncy v. Prudential Ins. Co., 

, 802 (5th Cir. 1982) ("When the statistical showing is

sufficiently strong in a disparate treatment action, the

plaintiffs' prima facie case can be made without additional

evidence establishing that the defendant purposefully treated

minorities protected under Title VII less favorably than other

persons.").    EEOC presented both statistical and anecdotal

evidence.    While we do not dispute the district court's

assessment of the anecdotal testimony of rejected applicants

Kathy Richie, Angela Burks, Ruby Cantu, Lillie Lewis, and Jessica

J. Jones, we hold that the district court erred both in its

assessment of the statistical evidence offered by the EEOC and in


                                     6
its conclusion that the EEOC failed to establish a prima facie

violation of Title VII.

           The district court correctly observes that "[t]he

usefulness of statistical data in assessing discriminatory

practices depends . . . on the validity of the basic reference

population as the pole star being compared to the work force of

the 

, and that, "[i]n a

disparate treatment case, the statistical evidence must be

`finely tuned' to compare the employer's relevant workforce with

the qualified populations in the relevant labor market."     

(quoting Krodel v. Young, 

, 709 (D.C. Cir.

1984)).   However, we disagree with the district court's

conclusion that the EEOC's statistical evidence fails to raise a

claim of intentional discrimination.   First, Dr. Straszheim's

"external availability" methodology is sufficiently similar to

that approved by the court in United States v. Pasadena Indep.

Sch. Dist., 43 Fair Emp. Prac. Cas. (BNA) 1319, 

(S.D. Tex. Apr. 18, 1987) (DeAnda, C.J.) ("Pasadena I.S.D."), to

beg the question why the court found it so lacking here.

           Second, the travel times which the district court found

"simply 

, were confirmed by the

census data, which was, in turn, legitimized by the actual

applications received by Olson's.    We do not understand how the

district court can completely discount the possibility that

prospective employees will travel further than a few blocks to

work at Olson's when it was presented with evidence of hundreds


                                 7
of applications from job seekers not residing in the immediate

vicinity of an Olson's location.

              Third, Dr. Chorush's analysis, which the district court

found "persuasive," 

unsound.      Dr. Chorush's

analysis considers only a portion of Olson's work force at only

one point in time, presuming that what was true for the Spring

Branch stores in April 1990 must be true for all Olson's

locations over the entire period under dispute.     Dr. Chorush

begins with the presumption that one can describe Olson's labor

market by describing Olson's work force; thus, he concludes,

since most of Olson's Spring Branch employees are white teenagers

living a short distance from the store, then white teenagers

living a short distance from the store constitute Olson's

available labor force.     This is wholly at odds with the

fundamental premise of employment discrimination law.     In order

to test for discriminatory hiring, we evaluate an employer's work

force in terms of the available labor pool, not the other way

around.   The fact that Olson's April 1990 Spring Branch work

force was predominantly white teenagers living close to the store

does not mean that there were not qualified applicants who were

not white teenagers living close to the store.

              Finally, the district court's assessment of the EEOC's

statistical evidence completely disregards the "applicant flow"

analysis conducted by Dr. Straszheim.     Dr. Chorush "opin[ed]"

that Olson's could "expect" to draw its work force from a given

area.   

Dr. Straszheim analyzed the actual


                                    8
applications.    The district court found, based upon Dr. Chorush's

testimony, that "[a]pplicants for employment [at Olson's] are

therefore likely to be substantially different from those

actually holding employment in the food preparation and service

classification [of the Census]."       

own

applications indicate that blacks not living within the immediate

vicinity of Olson's locations comprise a higher percentage of

applicants than was suggested by Dr. Straszheim's census-based

analysis.

            Guided by this circuit's previous admonition that the

"most direct route to proof of racial discrimination in hiring is

proof of disparity between the percentage of blacks among those

applying for a particular position and the percentage of blacks

among those hired," Hester v. Southern Ry., 

, 1379

(5th Cir. 1974), we conclude that the district court clearly

erred when it held, without fully considering the "applicant

flow" analysis offered by the EEOC's expert, that the EEOC had

failed to provide ample statistical evidence to establish a prima

facie violation of Title VII.2   To the contrary, we find the

record replete with evidence to establish such a violation.

     2.     Olson's Rationale.

            The district court summarily accepted, without

description or explanation, Olson's articulated nondiscriminatory


     2
          We also express concern for the short shrift which the
district court gave the EEOC's "external availability" analysis,
especially when we consider the dearth of countervailing evidence
offered by Olson's expert.

                                   9
reasons for its hiring and found that the EEOC failed to show

that those articulated reasons were a pretext disguising

.   We disagree with the

district court's assessment of Olson's proffered explanation.

The record clearly demonstrates that any explanation which the

district court may have perceived to be facially

nondiscriminatory is, in fact, mere pretext.

          Discarding Mr. Watson's statement that Olson's

customers prefer to be served by persons of their own "culture,"

the only other "reasons" which may be gleaned from Olson's case

are (1) the proximity of an applicant's residence to the

restaurant, and (2) the racial make-up of the Spring Branch

school district.    While the former might conceivably satisfy the

McDonnell Douglas-Teamsters test if there was a showing that

proximity to the restaurant was either a critical factor or even

a stated criteria in Olson's hiring guidelines, that showing was

not made or even attempted.    This leaves only the intimation that

people from nearby were hired because only people from nearby

would apply.   However, we know that is not true, based upon Dr.

Straszheim's review of Olson's applications.     As for the second

explanation, the racial make-up of the Spring Branch school

district explaining the racial make-up of the employees, aside

from ignoring conditions at the Bellaire and Katy stores and in

their surrounding neighborhoods, presumes that Olson's potential

work force is composed of area high school students.     However,

while it may be true that Olson's employees are predominantly


                                 10
area high school students, the applications make it clear that

Olson's available labor force includes many persons who are not

area high school students.

B.   OLSON'S ATTORNEY'S FEES.

     42 U.S.C. § 2000e-5(k) allows the district court to grant

the prevailing party in a Title VII action to recover reasonable

attorney's fees.    Because we render judgment for the EEOC on

liability, it is the EEOC who prevails and not Olson's.

Notwithstanding that, we are compelled to express our puzzlement

at how the district court could look at this record and find that

the EEOC's complaint was "frivolous, unreasonable, or without

foundation," Christiansburg Garment Co. v. EEOC, 

,

421, 

, 700 (1978), particularly in light of the

district court's denial of Olson's two pre-trial motions for

summary judgment and of Olson's Rule 41(b) motion for dismissal,

which was offered at the close of the EEOC's case-in-chief.

                            III. CONCLUSION

     We REVERSE and RENDER judgment in favor of the EEOC on the

question of Olson's liability.    We return this matter to the

district court in order to proceed to the damages stage of this

employment discrimination class action, see 

, 97 S. Ct. at 1867-68; Richardson v. Byrd, 

, 1021 (5th Cir. 1983).

     REVERSED AND RENDERED IN PART; CAUSE REMANDED.




                                  11