Agency.

Equal Employment Opportunity CommissionSep 27, 2002
01A04725errata (E.E.O.C. Sep. 27, 2002)

01A04725errata

09-27-2002

Agency.


Clifford Herron et. al v. Department of Agriculture

01A04725

September 27, 2002

.

Clifford Herron et. al.,

Class Agent,

v.

Ann M. Veneman,

Secretary,

Department of Agriculture,

Agency.

Appeal Nos. 01A04725; 01A02597; 01A05202; 01A04644; 01A05031;01A05034;

01A04941;01A20731; 01A20732;01A20733; 01A20734;01A20735;<1>

Agency Nos. 97-1290, 99-0055, 99-0326, 99-0509, 99-0630, 99-0148, 98-0416,

95-0817,95-0706, 96-0607, 97-0707, 97-0089, 96-0712;000514; 000517; 000518

Hearing Nos. 100-98-7120X, 100-98-7658X; 100-99-7837X; 100-99-7838X;

100-99-7839X;100-99-7840X; 100-99-7841X; 100-99-7842X; 100-99-7843X;

100-99-7844X; 100-99-7845X; 100-99-7851X; 100-99-7852X; 100-99-7853X;

100-99-7282X

DECISION

Class agents timely initiated an appeal<2> from the agency's final order

concerning their equal employment opportunity (EEO) complaint of unlawful

employment discrimination in violation of Title VII of the Civil Rights

Act of 1964 (Title VII), as amended, 42 U.S.C. � 2000e et seq. The appeal

is accepted pursuant to 29 C.F.R. � 1614.405. For the following reasons,

the Commission VACATES AND REMANDS the agency's final order.

This is a class complaint filed on February 27, 1997 against the

Department of Agriculture, Farm Services Agency (FSA) offices in

Washington, D.C. and throughout the nation. The complaint alleges that

the agency engaged in a pattern and practice of discrimination by failing

to promote its African American employees to the GS-13, 14 and 15 levels

from 1995 and thereafter, and that it employed certain facially neutral

practices which had a disparate impact on African American employees at

these levels.

The agency accepted the recommendation of the Administrative Judge

(AJ) to certify the class which was defined as all African American FSA

employees denied promotion or opportunity for promotions to positions

to the GS-13, 14, and 15 grade levels since February 27, 1995. After a

hearing, the AJ concluded that the class failed to prove that the agency's

promotion practices have a disparate impact on African Americans because

the statistical evidence did not demonstrate a statistically significant

under-representation of African Americans in FSA. She also found that

the class did not establish a prima facie case of pattern and practice

discrimination based on this same statistical evidence.

The AJ then considered the individual claims of discrimination and

determined that in the case of class agent, Clifford Herron, the agency

discriminated against him in not selecting him for the position of Program

Manager GS-15, Consolidated Farm Service Agency<3>, in the division of

Emergency and Non-insured Assistance Program (ENAP). The AJ awarded back

pay with interest, compensatory damages and attorney's fees associated

with Herron's successful claim.

She also found the agency had discriminated against class member Harold

Connor in not selecting him for the position of Director, Audits and

Investigations Group, FSA, GS-14/15. The AJ awarded Connor back pay

with interest, compensatory damages and attorney's fees.

The AJ concluded that, in the remainder of the individual complaints

filed, the agency gave legitimate non-discriminatory reasons for

not selecting them which complainants failed to demonstrate were a

pretext for race discrimination. As a result, as to these individual

complainants, the AJ found no proof of discrimination by a preponderance

of the evidence.

The agency's final order dated May 30, 2000 fully implemented the AJ's

decision.

CLASS CONTENTIONS ON APPEAL

The class contends that the AJ erred in finding no class-wide

discrimination because she based her decision on what she knew to

be unreliable and inadequate data provided by the agency. The class

claims that the AJ erred in instructing the expert not to consider 216

non-competitive promotions for District Director in his statistical

analysis. This was error, the class argued, because it accounted for

46.3% of the relevant total promotions during the time period in issue

and based on the analysis done by the class, indicated a 4.04 standard

deviation, a statistically significant shortfall of African American

promotions at the GS-13 level.

The class argued that despite the AJ's finding that the most appropriate

statistical analysis to be used in a promotion case required the use of

applicant flow data, the AJ's expert<4> did not perform such an analysis

and such data was not used. The class claimed it was error for the AJ not

to require the agency to produce more reliable data once the EEOC expert

apprised her of the deficiencies in the data produced. According to

the class, because the agency did not keep and produce adequate records

concerning selections it made pending an EEO complaint and in general as

required by 29 C.F.R. �� 1602.14 and 1607.4, the Commission should draw

an adverse inference that the records contained information favorable

to the class.

The class claimed that its expert produced the most reliable statistical

analysis because he included all promotions including promotions of

216 District Directors in 1996 and because his analysis factored in

past discriminatory influences on the available applicant pool. In the

alternative, the class urged the Commission to disregard the statistical

analysis altogether and to base a finding of class-wide discrimination

on the testimony and documentary evidence.

Assuming that the Commission found no class-wide discrimination, the

class argues that the AJ erred in not deciding many of the claims of

non-selection brought by the individual complainants and in not ultimately

finding discrimination as to the claims of individual complainants.

The class also argues that the AJ erred in her calculation of the relief

and the amount of the attorney's fees awarded to those class agents

found to have been victims of discrimination.<5>

AGENCY REPLY

The agency argued that the AJ was correct in finding no prima facie case

of disparate treatment because out of 18 claims of discrimination,

the class was only able to prove discrimination in 2 instances.

The agency contended that the AJ's conclusion that there was no proof of

a disparate impact was supported by substantial evidence in the record.

According to the agency, even assuming that the additional 216 promotions

the class alleged should have been included in the statistical analysis

were considered, there was still no showing of a disparity in promotions

of African Americans if such were compared to the percentage of African

Americans in the available work force.

The agency argued that the Class' claim that it did not keep adequate

records was not true and there was no evidence to support the contention.

The agency stated that the Class never requested additional information

regarding applicant flow data during the 8 months of discovery and that

the Class expert testified that he realized there was information he

did not have but he failed to request it. Even so, the agency argued

that courts have recognized that alternative statistics may be used,

such as that used by the AJ's expert, in instances where labor market

statistics will be difficult to obtain. There was no showing, according

to the agency, that the alternative statistics were not appropriate.

Addressing the AJ's findings on the individual claims, the agency contends

that the Class did not contest the AJ's rulings on the individual claims

of non-selection aside from attacking her credibility findings and

that her findings of no discrimination should be affirmed. The agency

contends that those individual claims not specifically addressed by the

AJ either did not fall within the parameters of the Class claims or were

not addressed because the complainants failed to file formal complaints.

ANALYSIS AND FINDINGS

Pursuant to 29 C.F.R. � 1614.405(a), all post-hearing factual findings by

an AJ will be upheld if supported by substantial evidence in the record.

Substantial evidence is defined as �such relevant evidence as a reasonable

mind might accept as adequate to support a conclusion.� Universal

Camera Corp. v. National Labor Relations Board, 340 U.S. 474, 477 (1951)

(citation omitted). A finding regarding whether or not discriminatory

intent existed is a factual finding. See Pullman-Standard Co. v. Swint,

456 U.S. 273, 293 (1982). An AJ's conclusions of law are subject to a

de novo standard of review, whether or not a hearing was held.

After a careful review of the record, the Commission finds that the AJ's

findings regarding the Class claims are not supported by substantial

evidence in that they are based on unreliable and inadequate or

inconclusive data and because the most reliable statistical analysis

typically used in a non-promotion/hiring case was not performed.

The Commission further finds that there is no indication that the AJ's

decision considered the anecdotal testimony of Class agents and other

witnesses regarding specifically alleged discriminatory incidents

and hiring practices in reaching her conclusion that there was no

discrimination perpetuated against the class. See International

Brotherhood of Teamsters v. United States, 431 U.S. 324 (1977).

First, we address the unreliability of the data presented. According to

the AJ's decision, much of the information generally referred to as

applicant flow data, usually analyzed in promotion cases, was missing

or not available. She noted that the agency did not provide applicant

flow or machine readable data on the vacancy announcements or on the

successful candidates' prior position. Because the applicant flow data

was not made available and because she expressed no confidence in the

reliability of the parties' expert witness reports, the AJ retained her

own expert witness.

The AJ's expert witness (E1) report indicated that he was instructed to

analyze the FSA workforce and promotion data already available to the

other experts. The agency was not directed to produce applicant flow

data or any other additional data. Regarding the data he analyzed,

E1 testified that there were problems in interpreting the data

that the agency did produce and that he made �educated guesses� in

reaching assumptions about the agency's data. Testimony at p. 1974.

More specifically, E1 noted that there was no explanation about how the

work force files were created or how the term �promotion� was defined

in the data he examined. AJ Exhibit 2 at p. 9. He also testified

that the lack of computer documentation created some ambiguity in his

interpretations of the variables the agency used. Testimony at p. 1976.

He further observed that he was unable to specifically determine the

change in organizational units over time in order to track the movement

of some employees.<6>

Further, we cannot glean from the record whether the analysis E1 performed

gave sufficiently reliable results given the problems noted above.

According to his report, E1 compared the proportion of African American

promotions to various estimates of �availability� of African Americans

provided by the parties' experts. Even with this method, E1 stated that

the workforce data supplied by the agency was not sufficient to justify

a statistical availability analysis, particularly at different points

in time. Thus, although the agency urged the Commission to accept this

alternative method, there were numerous problems noted in the record

that called into question the reliability of the data analyzed.

In addition, the class argues that it was error for the AJ to exclude over

200 �promotions� of District Directors in E1's analysis. Although the

AJ did not specifically rule that these promotions should be included it

is clear that they were not part of E1's analysis. According to E1, he

was instructed to assume that �promotions should be defined as permanent

promotions to grades 13, 14, and 15 during the period October 1, 1995 to

the present�. The class counted the total number of promotions including

the promotion of District Directors in 1996 to be 467. E1 testified

that he concluded there were a total of 243 promotions roughly 200 less

than the class count. The agency's Director of the Human Resources

Division testified that District Director promotions were an upgrade

of their positions from GS-12 to GS-13 through the merit promotion

process and that many of them, in fact, became permanent. The agency

gave no persuasive argument why these promotions should not be counted

as permanent promotions, therefore, we conclude that they should have

been counted and considered in E1's analysis. Their omission from his

analysis casts further doubt on the reliability of his results and his

ultimate conclusions.

In addition, the AJ was aware on review of E1's report and during E1's

cross-examination, that there were deficiencies in the data he analyzed.

His report specifically outlines his observations about �the adequacy

and sufficiency of the available data sources.� AJ Exhibit 2 at p. 3.

The AJ's decision makes specific reference to E1's statements about the

lack of adequate data. Our regulations provide the AJ with specific

authority to require the parties to produce such documentary evidence

as deemed necessary, and may even direct an investigation of facts by

another agency certified by the Commission. 29 C.F.R. ��1614.204(f)(3)

and(h); 1614.109a-f. The AJ had the authority to order the agency to

address and remedy any deficiencies or to supplement the record as needed.

Thus, we find it was an abuse of discretion for the AJ to proceed without

addressing the deficiencies and shortcomings noted in the evidence.

Since we conclude that the analysis performed was based on unreliable

and incomplete data and that another analysis based on a more widely

acceptable method should be done, we draw no conclusion whether these

omitted promotions are relevant to an analysis based on applicant

flow data.

Applicant Flow Data

As the AJ aptly noted,�the most appropriate analysis [in this case would

be] the composition of the employees promoted and the applicant pool

from which potential promotees come.�<7> Decision p. 5. Even with

this observation, the AJ did not ensure that the record was adequately

developed in terms of determining how the agency's hiring process operated

during the time period in question. There is no indication that the

AJ required the agency to produce the data that was missing from the

record and deemed by her to be the most probative in the case before her.

The AJ explained that neither party's expert requested additional data.

However, we conclude that this was an insufficient basis for not requiring

that the appropriate data be produced. Relying solely on the failings

of the parties as a basis for not pursuing the best method of analyzing

the agency's hiring practices, was error because the hearing process

is intended to be an extension of the investigative process in a class

complaint as in any other complaint brought pursuant to our regulations.

See 29 C.F.R. � 1614.204(h) referencing �1614.109(a) through (f);

See also EEOC Management Directive (MD) 110, as revised, November 9,

1999, Chapter 7, page 7-1. We decline to impose a sanction as the

class urges, for the agency's failure to produce applicant flow data.

Our review of the record indicates that E1 testified that the class did

not specifically request the data.

The Supreme Court has stated that statistical analysis using applicant

flow data is a �very relevant� statistical model in discrimination

cases involving promotion and hiring. Hazelwood School District

v. United States, 433 U.S. 299, 309, n. 13 (1977); Wards Cove Packing

Co. Inc. v. Atonio, 490 U.S. 642, 650-1 (1989). The analysis requires

a comparison of the racial composition of persons who applied for a

position with the racial identity of persons who hold the at issue jobs.

Bullington v. United Airlines, 186 F.3d 1301 (10th Cir. 1999). In cases

involving hiring and promotion, an analysis of applicant flow data is

�the most direct route� to proof of discrimination. EEOC v. Olson's

Dairy Queens, 989 F.2d 165,169 (5th Cir. 1993). Mister v. Illinois

Central Gulf Railroad Co., 832 F.2d 1427, 1435 (7th Cir. 1987) (Applicant

studies are preferable as a rule because Title VII governs the treatment

of applicants.) Such data has probative force because it reflects

how the employer's hiring procedures actually operated as well as the

actual percentage of Caucasian versus African American applicants for

the positions in question. Bullington supra at 1313. Wards Cove Packing

Co. Inc. v. Atonio, 490 U.S. 642, 651 (1989) (The Supreme Court recognized

that where labor market statistics will be difficult if not impossible to

ascertain, certain other statistics � such as the racial composition of

�otherwise qualified applicants� are equally probative.); See Watson

v. Fort Worth Bank & Trust, 487 U.S. 977, 995 n. 3,997 (1988).(The

reliability or usefulness of any particular analysis will depend on the

surrounding facts and circumstances.); see also New York City Transit

Authority v. Beazer, 440 U.S. 568 (1979); and Dothard v. Rawlinson,

433 U.S. 321 (1977) (Court recognized certain alternative statistics).

As discussed above, E1 observed that there was no data provided about the

applicants to FSA vacancy announcements or on the vacancy announcements

themselves. Also not produced was data indicating the successful

candidates' prior position such that it could be determined whether

candidates were hired or promoted across organizational units. As we

stated before, based on the expert's own expressed reservations about

the data, the reliability of the analysis of that data is diminished.

We see no reason to find E1's analysis more reliable since it was based

on the very same unreliable data used by the parties' experts.

On remand, the parties as well as the AJ should be guided by the Supreme

Court's decision in Bazemore v. Friday, 478 U.S. 385 (1986). There,

the court set forth the parameters for evaluating the probative value of

statistical evidence in deciding the question whether discrimination was

proven by a preponderance of the evidence. That is, even a statistical

analysis (in Bazemore, multiple regression analyses) that includes less

than all measured variables may serve to prove the complainant's case

because a Title VII suit need not prove discrimination with scientific

certainty only by a preponderance of the evidence. Id. at 400. In

Bazemore, there was proof that the employer engaged in discriminatory

conduct prior to the enactment of the statute which the court found might

support the inference that such discrimination continued where the certain

aspects of the decision making process had undergone little change.

Id. at 402 citing Hazelwood supra at 309-310, n.15. The Court also held

that it was error not to consider the evidence in the record as a whole,

in addition to the statistical evidence, in determining whether there was

sufficient proof of a pattern and practice of discrimination. Id at 387

& 403.<8> In Bazemore, the court cited the probative value of evidence

of individual comparisons between salaries of African Americans versus

Caucasians and witness testimony which confirmed the continued existence

of such disparities. Id. at 403.

Because we remand this case for the statistical analysis to be redone

based on the more probative applicant flow data, we reach no decision

at this time on the class claim of disparate impact except to emphasize

that the burden of proof requires that the class identify a neutral

practice which resulted in a disparate impact on the protected class.

If a statistical disparity is shown, it must also be demonstrated that

there is a causal connection between the statistical disparity and

the specific employment practice alleged to have created a disparate

impact. Watson v. Fort Worth Bank & Trust, 487 U.S. 977 (1988).

Individual Complaints

We also remand each individual appeal decided by the AJ as well as those

not specifically decided by the AJ but presently before us on appeal,

pending a decision on the class claims. Individual complaints filed

before or after the class complaint is filed and that come within the

definition of the class claim(s) will be subsumed within the class

complaint. MD 110 at Chapt. 8-4.

Individual complaints alleging reprisal must be addressed on an individual

basis regardless of the AJ's decision on the class complaint. MD 110 at

Chapt. 8-6. As there is no decision in the record addressing Mr. Herron's

claim of reprisal or any other individual claim of reprisal properly

raised, we remand the issue for a decision on the merits.

CONCLUSION

Based on the foregoing and after due consideration of the parties

statements on appeal and the record transmitted by the agency, the

Commission VACATES the agency's final action and REMANDS this matter for

further proceedings consistent with this decision and as ordered below.

ORDER

The agency is hereby directed to take the following action:

1. The agency is directed to submit a copy of the complaint file to

the EEOC Washington Field Office within fifteen (15) calendar days of

the date this decision becomes final. The agency shall provide written

notification to the Compliance Officer at the address set forth below

that the complaint file has been transmitted to the Hearings Unit.

2. The agency is directed to supplement the record with applicant flow

data for a statistical analysis including but not limited to:

a) Machine readable data (hereinafter in a format deemed acceptable

by the AJ) on the applicants to FSA vacancy announcements during the

relevant time period as established by the AJ;

b) Machine-readable data on the vacancy announcements issued during

the relevant time period as established by the AJ and;

c) Machine readable data on the successful candidate's prior position

for each vacancy.

3. The agency is directed to provide standard computer documentation

describing the meaning of any variables used in any data produced and

how the agency's files were prepared.

4. The AJ may in his/her discretion reopen the period for discovery

to allow the parties to take depositions or other forms of discovery

regarding any additional analyses performed or to further clarify the

additional data produced by the agency.

5. The AJ shall take evidence and issue a finding and decision on the

issue of reprisal.

Thereafter the Administrative Judge shall issue a decision on the

complaint in accordance with 29 C.F.R. � 1614.204(h) and the agency

shall issue a final action in accordance with 29 C.F.R. � 1614.204(j).

IMPLEMENTATION OF THE COMMISSION'S DECISION (K0501)

Compliance with the Commission's corrective action is mandatory.

The agency shall submit its compliance report within thirty (30)

calendar days of the completion of all ordered corrective action. The

report shall be submitted to the Compliance Officer, Office of Federal

Operations, Equal Employment Opportunity Commission, P.O. Box 19848,

Washington, D.C. 20036. The agency's report must contain supporting

documentation, and the agency must send a copy of all submissions to

the complainant. If the agency does not comply with the Commission's

order, the complainant may petition the Commission for enforcement

of the order. 29 C.F.R. � 1614.503(a). The complainant also has the

right to file a civil action to enforce compliance with the Commission's

order prior to or following an administrative petition for enforcement.

See 29 C.F.R. �� 1614.407, 1614.408, and 29 C.F.R. � 1614.503(g).

Alternatively, the complainant has the right to file a civil action on

the underlying complaint in accordance with the paragraph below entitled

"Right to File A Civil Action." 29 C.F.R. �� 1614.407 and 1614.408.

A civil action for enforcement or a civil action on the underlying

complaint is subject to the deadline stated in 42 U.S.C. 2000e-16(c)

(1994 & Supp. IV 1999). If the complainant files a civil action, the

administrative processing of the complaint, including any petition for

enforcement, will be terminated. See 29 C.F.R. � 1614.409.

STATEMENT OF RIGHTS - ON APPEAL

RECONSIDERATION (M0701)

The Commission may, in its discretion, reconsider the decision in this

case if the complainant or the agency submits a written request containing

arguments or evidence which tend to establish that:

1. The appellate decision involved a clearly erroneous interpretation

of material fact or law; or

2. The appellate decision will have a substantial impact on the policies,

practices, or operations of the agency.

Requests to reconsider, with supporting statement or brief, must be filed

with the Office of Federal Operations (OFO) within thirty (30) calendar

days of receipt of this decision or within twenty (20) calendar days of

receipt of another party's timely request for reconsideration. See 29

C.F.R. � 1614.405; Equal Employment Opportunity Management Directive for

29 C.F.R. Part 1614 (EEO MD-110), 9-18 (November 9, 1999). All requests

and arguments must be submitted to the Director, Office of Federal

Operations, Equal Employment Opportunity Commission, P.O. Box 19848,

Washington, D.C. 20036. In the absence of a legible postmark, the

request to reconsider shall be deemed timely filed if it is received by

mail within five days of the expiration of the applicable filing period.

See 29 C.F.R. � 1614.604. The request or opposition must also include

proof of service on the other party.

Failure to file within the time period will result in dismissal of your

request for reconsideration as untimely, unless extenuating circumstances

prevented the timely filing of the request. Any supporting documentation

must be submitted with your request for reconsideration. The Commission

will consider requests for reconsideration filed after the deadline only

in very limited circumstances. See 29 C.F.R. � 1614.604(c).

COMPLAINANT'S RIGHT TO FILE A CIVIL ACTION (R0900)

This is a decision requiring the agency to continue its administrative

processing of your complaint. However, if you wish to file a civil

action, you have the right to file such action in an appropriate United

States District Court within ninety (90) calendar days from the date

that you receive this decision. In the alternative, you may file a

civil action after one hundred and eighty (180) calendar days of the date

you filed your complaint with the agency, or filed your appeal with the

Commission. If you file a civil action, you must name as the defendant in

the complaint the person who is the official agency head or department

head, identifying that person by his or her full name and official title.

Failure to do so may result in the dismissal of your case in court.

"Agency" or "department" means the national organization, and not the

local office, facility or department in which you work. Filing a civil

action will terminate the administrative processing of your complaint.

RIGHT TO REQUEST COUNSEL (Z1199)

If you decide to file a civil action, and if you do not have or cannot

afford the services of an attorney, you may request that the Court appoint

an attorney to represent you and that the Court permit you to file the

action without payment of fees, costs, or other security. See Title VII

of the Civil Rights Act of 1964, as amended, 42 U.S.C. � 2000e et seq.;

the Rehabilitation Act of 1973, as amended, 29 U.S.C. �� 791, 794(c).

The grant or denial of the request is within the sole discretion of

the Court. Filing a request for an attorney does not extend your time

in which

to file a civil action. Both the request and the civil action must be

filed within the time limits as stated in the paragraph above ("Right

to File A Civil Action").

FOR THE COMMISSION:

______________________________

Carlton M. Hadden, Director

Office of Federal Operations

September 27, 2002

__________________

Date

1Clifford Herron, Appeal No. 01A20735;

Starendal Bryant Appeal No. 01A02597; Jim Lawson Appeal No. 01A20733;

Carnell McAlpine 01A05034; Matthew Miller Appeal No. 01A04941; Harry

Milner Appeal No. 01A20732; Charles Smith Appeal No. 01A20734; Helen

Smith Appeal No. 01A20731; Johnnie Blackwell Appeal No. 01A04644; Jerome

Jeffries Appeal No. 01A05202; Ozetta Thomas Appeal No. 01A05031.

2The Commission consolidates the individual appeals filed by class members

and class agents with the Class appeal pursuant to 29 C.F.R. �1614.606.

3The Consolidated Farm Service Agency, Federal Crop Insurance Corporation

and part of the Farmers Home Administration were combined to form the

Farm Services Agency in 1995.

4The AJ decided that based on the lack of reliability of the data

presented, she would appoint an expert to assist her in her evaluation

of the case. The person she appointed is a Social Science Analyst with

the EEOC's Office of General Counsel, Research and Analytic Services.

5Since the filing of the Class appeal, the individual appeal of Harold

Connor was settled and withdrawn.

6The AJ observed that this was a problem in accepting the conclusions

of the agency's expert where various reorganizations occurred changing

the identity of various administrative units thereby making it unclear

from where an employee had been promoted.

7The AJ's expert gave a cautionary footnote in his report that �it

is possible that a well designed applicant flow study, examining

qualified applicants for each promotion separately, would reach different

conclusions.� Memorandum from E1 to Administrative Judge dated June 11,

1999 at p. 2.

8Such evidence might include inter alia Class Exhibit C-6, admitted

into evidence, consisting of a report issued by the agency entitled

�Civil Rights at the United States Department of Agriculture - A Report

by the Civil Rights Action Team� (February 1997) which refers to a

1995 Government Accounting Office report. In the agency's report it

acknowledges, in effect, that women and minorities in comparison to white

men, were still represented in lower relative numbers in the agencies'

key job categories and that white men continued to dominate the higher

ranks of USDA's top positions in 1996. Id at p. 33. The dates of the

selections at issue in this case occurred between February 1995 and 1998,

closely coinciding with the findings in the agency's report.

The agency's report also acknowledged that minority farmers and

recipients of FSA program benefits were denied ready access because of

discrimination on the part of county office staff and that minority

participation rates in the commodities and disaster relief programs

was disproportionately low. At least two of the selections at issue

in this case were in the Emergency Non-Insured Assistance Program which

administered the disaster relief programs. Id at p. 21