Especially helpful if:

• Your agency appears to be approaching or has reached an impasse with the recipient, i.e., negotiations are breaking down, but you believe it may not be in your best interests or that you are prepared to issue formal findings or proceed to enforcement.
• It is important to develop and/or maintain a constructive working relationship with the recipient for the future, but this is unlikely without the intervention of a skilled third party to move beyond the impasse you have reached.

The "aggressive" use of ADR as an approach to processing complaints can result in an extensive saving of staff time and, perhaps more importantly, in a good working relationship with your recipients. The fact that you are saving your staff time probably also means that you are requiring that less staff time be expended by the recipient. It allows your agency to appear less confrontational and to provide technical assistance up front, where it may be most helpful.(22)

However, we recommend that agencies only use ADR if they have staff who are experienced in doing complaint investigations, as the ability to "jump" from the initial allegations to an appropriate resolution requires a thorough understanding of the issues, legal concepts, and scope of the relief that would be involved.

A disadvantage of using ADR to resolve a case through a Settlement Agreement prior to issuing findings is that, if the recipient fails to meet the terms of the Settlement Agreement, DOJ can only sue the recipient based on its failure to comply with the terms of the Agreement, and not based on an actual violation of Title VI. In order to sue for a violation of Title VI, a formal violation Letter of Findings must have been issued, the recipient must have been given an opportunity to come into compliance, and a determination must have been made that voluntary compliance cannot be achieved. Also, your agency may prefer, for policy or programmatic reasons, to issue a violation Letter of Findings rather than to resolve the complaint through ADR.

For example, the complainant and other victims have faced discrimination for an extended period of time with this recipient and specifically request that a violation Letter of Findings be issued to make clear the recipient's obligations under Title VI; even without the complainants' request, you may believe a violation LOF is appropriate. Or, the case raises important policy issues that your agency wants to address and clarify in formal findings, both for this recipient and as a way of clarifying your position for other recipients that may have the same problems.

In the types of situations described above, you would not discuss settlement with the recipient until after the findings have been issued. Likewise, a recipient who has cooperated in an investigation may wish a formal compliance Letter of Findings to "clear its name."

If your agency determines that it will use ADR methods to resolve complaints, it may find that it will rarely issue Letters of Findings, which are written after investigations are completed. This is, of course, dependent upon the nature of the complaints you receive and the cooperation of the recipients with which you are dealing. For a more thorough discussion of the provisions that should be included in Settlement Agreements, see Section VIII.

You should keep in mind that the use of ADR does not mean that you can ignore the relief that would be appropriate if you conducted a full investigation. Rather, ADR is a means of resolving cases with basically the same relief you would get after a full investigation, while avoiding the expenditure of staff time the full investigation requires. Remember the following when considering whether to use ADR at any point during your case processing:

• you can/should be open to negotiate a resolution to a case at any point during the processing of a complaint;
• the type of ADR (your "settlement approach") should be selected carefully, based upon the allegations, number of persons affected, type and extent of relief involved, cooperation of the recipient, and other factors;
• the case file should include an explanation of how the resolution was determined to constitute adequate relief;
• the resolution should provide for monitoring whenever appropriate; and,
• you may always reopen a complaint if you learn that the recipient has not complied with its commitments.

Back to top

If the Complainant does not Agree to Settlement

In some cases, investigators are concerned that the complainant will not agree with the resolution of their complaint through ADR. Unless there are factors necessitating that the complainant be a party to the resolution (e.g., in a case in which the recipient wants the complainant to agree not to sue for additional relief), it is your agency's responsibility to determine what constitutes "full relief." Your agency does not represent the complainant, but rather the interests of the Federal government in ensuring nondiscrimination by its recipients. Therefore, the relief you seek is dictated by the facts of the case and not by the complainant.

If the recipient has agreed to provide what you have determined would constitute full relief (in changes to policies/procedures, relief to the complainant and other victims, etc.) and the complainant disagrees with the policy changes or refuses to accept individual relief, you may complete the agreement with the recipient and close the complaint on that basis. If the recipient has already offered full relief to the complainant and no other relief is appropriate (e.g., change in policies or practices, or relief for other victims), and the complainant refuses to accept it, you may administratively close the complaint for "failure to accept full relief."

Back to top

B. Pre- Investigation Case Closure through ADR

If a complaint is resolved without an investigation, the complaint resolution letter to the complainant and recipient should contain:

1) the basis for the complaint (race, color, national origin, sex, disability, religion, and/or age);

2) a brief statement of the allegations over which your agency has jurisdiction;

3) a brief statement of your agency's jurisdiction over the recipient;

4) an explanation of the basis for your agency's determination that the complaint has been resolved;

5) the protection from retaliation and the Freedom of Information Act/Privacy Act paragraphs;

6) a copy of any written agreement should be attached;

7) a notation for the recipient as to when the first monitoring report will be due, if applicable; and,

8) if appropriate, that the complainant has the right to file his/her own private lawsuit in Federal District Court.(23)

Back to top

C. Full Investigation Approach

Those cases that do not appear to be appropriate for early ADR processing should be handled according to the more standard investigative procedures. As indicated above, you may still be able to resolve the complaint successfully prior to issuance of formal findings. This will become evident as you proceed with the investigation. The following section discusses the steps involved in a full investigation of a complaint of discrimination.

Back to top

V. COMPLAINT INVESTIGATION

This section discusses the steps involved in a complete investigation of a complaint of discrimination. These steps would also occur when you conduct a compliance review (see Chapter X for a discussion of Compliance Reviews). As discussed in the previous section, Approaches to Complaint Resolution, you may decide that it is appropriate to initiate settlement negotiations at any time during an investigation. However, you should ensure that you have enough information when you do discuss settlement to be certain that the relief you are agreeing to is sufficient for the case at hand.

In cases where a full investigation is needed to make legally sufficient findings, identify all aggrieved victims, and determine appropriate relief, you will be the most efficient if you have carefully planned your investigation in advance. Identify the legal approach you will take up front to the extent possible, determine what kind and how much evidence will be needed, and remember that you will need to meet your burden of proof.

The standard of proof applied in making a determination of noncompliance, absent contrary statutory command, should be one of "preponderance of the evidence."(24) The primary reason for this is because a formal noncompliance finding may be challenged at an administrative hearing, and the evidentiary standard that will be applied by the hearing examiner will be a preponderance of the evidence.(25) Thus, formal findings of noncompliance should not be issued unless the preponderance standard is met.

Back to top

A. Complainant and Recipient Notification

If you decide to proceed with investigation of the complaint, you should notify the complainant and the recipient that you have accepted the complaint for investigation. (See TAB 12 for Letter to Recipient and TAB 13 for Letter to Complainant.) Your notification letter to the complainant and recipient should contain:

1) the basis for the complaint;

2) a brief statement of the allegations over which your agency has jurisdiction;

3) a brief statement of your agency's jurisdiction over the recipient to investigate the complaint; and

4) an indication of when the parties will be contacted.(26)

If you are prepared to do so at this point, you can consider the following two options in complaints that raise limited (usually individual) allegations:

1) Request a position statement from the recipient in response to the allegations. You would ask the recipient to include with its response all appropriate policies, procedures, and documents relating to the complaint.

2) Include in your notification an offer to engage in ADR (settlement negotiations) to resolve the complaint. This should only be done if you believe you have a good idea of the nature of the relief you are seeking and you already have some of the information you need to determine relief.

Again, these options would not be appropriate if the complaint includes class or complex allegations that must be fully investigated and that could require extensive relief. However, the request for an initial position statement can help you determine how you should approach a complaint because it gives you information about the recipient's defense. If you determine it is not appropriate to call or write a recipient and ask it to enter into settlement negotiations, in most

cases you should be open to settlement at any point during the investigative process.

Back to top

B. Planning for the Investigation

Complaint investigation and resolution should be preceded by planning, and the approach should be developed based on the nature and complexity of the issues involved. Whether or not an Investigative Plan (IP) is prepared, all case files should set out the specific allegations to be resolved and, if appropriate, the expected internal time frames to be adhered to by the investigator or investigative team.(27)

The extensiveness of an IP depends on the complexity of the issues involved. Some investigations may require revisions to the IP, or a supplemental IP, after receipt of information from the recipient or after an onsite investigation is conducted.

The basic IP will help the investigator focus on the principle issues to be explored in the course of the investigation, as well as the sources of evidence available to resolve them. An IP should include at least the following:

1) Jurisdictional information;

2) Identification of bases and issues;

3) Identification of the applicable legal theories;

4) Conclusions drawn from the analysis of the data or other evidence already gathered;

5) Description of the documentary, testimonial, and statistical evidence required to complete the investigation and the best sources and means of obtaining each type of evidence;

6) Anticipated sequence of case activities, including onsite visits if needed;

7) Anticipated timeframes for obtaining and analyzing evidence (if appropriate); and,

8) Statement of likely or enunciated recipient defenses and a description of the evidence required to test their validity.

A discussion of each of these aspects of the investigation follows. As a method of setting out information that is already known and that which is needed, you may wish to use a "chart" format. (See TAB 14 for sample IP charts for hypothetical complaint(s).)

Back to top

1. Jurisdictional Information

Any jurisdictional questions should be resolved at the outset, prior to investigating the allegations. Basically, this means determining that a covered basis of discrimination (e.g., race, sex, national origin) has been alleged in a timely fashion against an agency's recipient, giving the agency jurisdiction to investigate. (See discussion of jurisdiction in Section III.A.)

Back to top

2. Identification of Bases and Issues

The investigator should determine if the complainant alleges that discrimination against one or more members of a protected class -- because of race, color, national origin, sex, religion, disability or age -- is wholly, or at least in part, responsible for the complaint: this is the basis for the complaint.(28)

The investigator should also identify the specific action, policy, or practice responsible for the alleged discrimination (e.g., denial of educational or health services, harassment, retaliation for filing a complaint or giving testimony in an investigation, provision of unequal services, etc.). Even if intentional discriminatory treatment cannot be ascertained, does the practice, procedure, or service identified have a disparate effect on a certain protected class?

Back to top

3. Identification of the Applicable Legal Theories

Regardless of whether you develop a formal Investigative Plan or not, you must know what theory or theories of discrimination you are using in your investigation in order to understand the standards of proof needed to establish a violation. Two primary legal theories are used to establish a case of prohibited discrimination: intentional discrimination/disparate treatment and disparate impact/effects.(29) In thinking about the potential legal theories raised by a complaint, it is important to remember two things: (1) intentional discrimination may take many forms (some of the most common of which are discussed below), and (2) the complaint may raise allegations based on more than one legal theory, especially if it involves allegations that an applicant or potential participant did not meet entry or participation requirements for a federally assisted program.

Back to top

a. Intentional Discrimination/Disparate Treatment

As noted above, intentional discrimination may take many forms. One of the most common forms of intentional discrimination is disparate treatment. Simply put, disparate treatment means that similarly situated persons are treated differently (i.e., less favorably) than others because of their race, color, national origin, sex, etc. Disparate treatment cases can involve either "individual" or "class"(30) discrimination (or both). For example, if two individuals apply to participate in a federally funded program and one is rejected because the interviewer dislikes members of the rejected applicant's race, this constitutes disparate treatment. If the interviewer repeatedly rejects members of a particular race, this may indicate class discrimination or a "pattern and practice" of discriminatory conduct by the recipient.

Another type of intentional discrimination involves the use of policies or practices that explicitly classify individuals on the basis of their membership in a particular group. Such "classifications" may constitute unlawful discrimination if they are based on characteristics such as race, color, sex, etc. For example, the Supreme Court has held in a Title VII case that a policy which required female employees to make larger contributions to the pension fund than male employees created an unlawful classification based on sex. See City of Los Angeles, Department of Water and Power v. Manhart, 435 U.S. 702 (1978).

The analysis of intentional discrimination under Title VI is equivalent to the analysis of disparate treatment under the Equal Protection Clause of the Fourteenth Amendment. To prove intentional discrimination under Title VI you must show that "a challenged action was motivated by an intent to discriminate." Elston v. Talladega County Board of Education, 997 F.2d 1394 (11th Cir. 1993). This requires a showing that the recipient was not only aware of the complainant's race, color, or national origin, but that the recipient acted, at least in part, because of the complainant's race, color, or national origin.

However, it is important to remember that the record need not contain evidence of "bad faith, ill will or any evil motive on the part of the [recipient]." Elston, 997 F.2d at 1406 (quoting Williams v. City of Dothan, Alabama, 745 F.2d 1406, 1414 (11th Cir. 1984)). For example:

• A recipient official denies a woman's application to participate in a training program for construction workers. He does not select her because he thinks that it's not safe for women to do this type of work and he's concerned for her safety. Although this official did not act with any ill will toward this woman or women in general, he has considered this female applicant differently than male applicants and denied her admission to the training program because of her sex.

Evidence of discriminatory intent may take many forms and may be found in various sources, including statements by decisionmakers, the historical background of the events at issue, the sequence of events leading to the decision at issue, a departure from standard procedure (e.g., failure to consider factors normally considered), legislative or administrative history (e.g., minutes of meetings), a past history of discriminatory or segregated conduct, and evidence of a substantial disparate impact on a protected group.

In some cases, you will find that there is direct evidence of discrimination. For example, the facts may show that a member of the recipient's management staff was overheard commenting that the complainant was not selected for a particular program because of his or her race. More frequently, however, the investigation will be based on circumstantial or indirect evidence. In such cases, the analytic framework used by courts evaluating Title VII claims may provide a useful guide. See McDonnell Douglas Corp. v. Green, 411 U.S. 792 (1973).

Applying this framework to a Title VI claim, you must first determine whether the complainant can raise an inference of discrimination by establishing a prima facie case. The elements of a prima facie case may vary depending on the facts of the complaint, but such elements often include the following:

1. that the aggrieved person was a member of a protected class;

2. that this person applied for, and was eligible for, a federally assisted program that was accepting applicants:

3. that despite the person's eligibility, he or she was rejected; and,

4. that the recipient selected applicants of a different race, color, or national origin than the complainant -- or that the program remained open and the recipient continued to accept applications from applicants of a different race, color, or national origin than the complainant.

If the record contains sufficient evidence to establish a prima facie case of discrimination, you must then determine whether the recipient can articulate a "legitimate, nondiscriminatory reason" for the challenged action. See McDonnell Douglas, 411 U.S. at 802. If the recipient can articulate a nondiscriminatory explanation, you must determine whether the record contains sufficient evidence to establish that the recipient's stated reason was a pretext (i.e., an excuse) for discrimination. Id. In other words, the evidence must support a finding that the reason articulated by the recipient was not the true reason for the challenged action, and that the real reason was discrimination based on race, color, or national origin.

Similar principles may be used to analyze claims that a recipient has engaged in a "pattern or practice" of unlawful discrimination. Such claims may be proven by a showing of "more than the mere occurrence of isolated or 'accidental' or sporadic discriminatory acts." See International Brotherhood of Teamsters v. United States, 431 U.S. 324, 336 (1977). The evidence must establish that a pattern of discrimination based on race, color, or national origin was the recipient's "standard operating procedure the regular rather than the unusual practice." Id. Once the existence of such a discriminatory pattern has been proven, you may assume that every disadvantaged member of the protected class was a victim of the discriminatory policy, unless the recipient can show that its action was not based on its discriminatory policy. Id. at 362.

Finally, you should also remember that some cases of intentional discrimination may involve the use of "classifications" based on race, sex, or some other prohibited characteristic. If the facts of a case reveal that the recipient utilizes a policy or practice that explicitly treats members of a protected group differently from others, your investigation should focus on the recipient's reasons for utilizing the challenged classification policies. Most such policies will be deemed to violate Title VI, unless the recipient can articulate a lawful justification for classifying people on the basis of race, color, or national origin.

Back to top

b. Disparate Impact/Effects

The second primary theory for proving a Title VI violation is based on Title VI regulations and is known as the discriminatory "effects" or disparate impact theory. In contrast to disparate treatment, the disparate impact/effects theory does not require proof of discriminatory intent. Rather, disparate impact cases involve claims that a recipient is violating Title VI regulations by utilizing a neutral policy or practice that has the effect of disproportionately excluding or adversely affecting members of a protected group, and the recipient's policy or practice lacks a "substantial legitimate justification."

The Supreme Court has held that Title VI regulations may validly prohibit practices having a disparate impact on protected groups, even if the actions or practices are not intentionally discriminatory (Guardians, supra; Alexander v. Choate, supra.), and many subsequent cases have also recognized the validity of Title VI disparate impact claims. See Villanueva v. Carere, 85 F.3d 481 (10th Cir. 1996); Chicago v. Lindley, 66 F.3d 819 (7th Cir. 1995); Georgia State Conf. v. Georgia, 775 F.2d 1403 (11th Cir. 1985); Larry P. v. Riles, 793 F.2d 969 (9th Cir. 1984). In addition, by memorandum dated July 14, 1994, the Attorney General directed the Heads of Departments and Agencies to "ensure that the disparate impact provisions in your regulations are fully utilized so that all persons may enjoy equally the benefits of Federally financed programs." (See TAB 15 for a copy of the Attorney General's memorandum.)

Pursuant to Title VI regulations, all entities that receive Federal funding enter into standard agreements or provide assurances that require certification that the recipient will comply with the implementing regulations under Title VI. Guardians, 463 U.S. 582, 642 n. 13. You should carefully examine the assurances signed by the recipient in question when investigating any claim.

The principles used to analyze a Title VI disparate impact claim are similar to those used to analyze a Title VII disparate impact claim. New York Urban League, Inc. v. State of New York, 71 F.3d 1031, 1036 (2nd Cir. 1995). In a disparate impact case, the focus of your investigation will center on the consequences of the recipient's practices, rather than the recipient's intent.

To establish liability under a disparate impact scheme, you must first ascertain whether the recipient utilized a facially neutral practice that had a disproportionate impact on a group protected by Title VI.(31) Larry P. v. Riles, 793 F.2d 969, 982; Elston, 997 F.2d at 1407 (citing Georgia State Conference, 775 F.2d 1403, 1417 (11th Cir. 1985)). This showing requires a comparison of the effects of the policy or practice on members within the protected class relative to the effect on persons outside the protected class. For example, if a recipient has a policy that applicants for its entry-level vocational training program for plumbers must have a high school diploma, and you show that this policy results in the elimination of 90 percent of the Hispanic applicants but only 15 percent of the white applicants, then the recipient's policy has an adverse impact on Hispanics.

If your investigation proves that there is a statistically significant adverse impact on members of a protected class, you must then determine whether the recipient can articulate a "substantial legitimate justification" for the challenged practice.Georgia State Conference, 775 F.2d at 1417. "Substantial legitimate justification" is similar to the Title VII concept of "business necessity," which involves showing that the policy or practice in question is related to performance on the job. Griggs v. Duke Power, 401 U.S. 424 (1971).

To prove a "substantial legitimate justification," the recipient must show that the challenged policy was "necessary to meeting a goal that was legitimate, important, and integral to the [recipient's] institutional mission." Sandoval, supra, 1998 WL 295891, at *36 (M.D.Ala.), F.Supp. , (quoting Elston, 997 F.2d at 1413). The justification must bear a "manifest demonstrable relationship" to the challenged policy or program. Georgia State Conference, 775 F.2d. At 1418. See, e.g., Elston v. Talladega County Board of Education, 997 F.2d 1394 (11th Cir.), reh'g denied, 7 F.3d 242 (11th Cir. 1993). In an education context, the practice must be demonstrably necessary to meeting an important educational goal, i.e., there must be an "educational necessity" for the practice. See Larry P., supra.

If the recipient articulates a "substantial legitimate justification" for the challenged policy, you must carefully examine that justification in order to determine whether it is a pretext for discrimination. You must also consider whether there are any less discriminatory alternatives that would still accomplish the recipient's program objectives. If you find that the "substantial legitimate justification" is a pretext for discrimination, or if there are any less discriminatory alternatives, the evidence will support a finding of a violation.

In the example given above, you would critically evaluate the recipient's asserted "substantial legitimate justification" for its policy of requiring a high school diploma. The burden is now on you to either show that the claimed need for the diploma is merely a pretext for discrimination (e.g., by showing that it is not related to the training program at all, for example, because many plumbers in that State or county never finish high school) or that there are other less discriminatory alternatives to attaining the level of education needed to participate in the program (e.g., by earning a G.E.D., through experience in internship programs, or by performing certain types of work in a related field, etc.).(32)

Courts have often found Title VI disparate impact violations in cases where recipients utilize policies or practices that result in the provision of fewer services or benefits, or inferior services or benefits, to members of a protected group. Larry P. v. Riles, 793 F.2d 969, 983 (9th Cir. 1984) (Use of IQ tests for placing school children in special classes had discriminatory effect); Sandoval, 1998 WL 295891, *46 (M.D.Ala.) (Discrimination on the basis of language, in the form of an English-only policy, had an unjustified disparate impact on the basis of national origin); Meek v. Martinez, 724 F.Supp. 888 (S.D.Fla. 1987) (Florida's use of funding formula in distributing aid resulted in a substantially adverse disparate impact on minorities and the elderly); Campaign for Fiscal Equity, Inc. v. State of New York, 86 N.Y.2d 307, 1995 N.Y. Lexis 1145 (New York. Ct. App. June 15, 1995) (Allocation of educational aid had a racially disparate impact).

Other examples of the types of policies/standards that might result in a disparate impact on protected groups include:

• imposing financial or other requirements (e.g., home ownership, residence at one address or in a certain city for a specified time period, ability to read and write English, ownership of a car, education credentials) for which there are no legitimate program justifications;
• insisting on experience or training of a certain type when experience or training in a related field would be equally useful; or
• using personal information (e.g., number of children, marital status, potential lack of child care, arrest records) as a basis for making decisions on who can participate in a program, when these factors have not been shown to have any relevance to the program requirements or are not required by the program statute or regulations.

The most important thing to remember when investigating a disparate impact claim is that the primary focus of the analysis will be on the results of the policy or practice as opposed to the intent of the decisionmaker.

Back to top

c. Retaliation)

A complainant may bring a retaliation claim under Title VI or under a Title VI regulation that prohibits retaliation. For example, most agency Title VI regulations provide that "[n]o recipient or other person shall intimidate, threaten, coerce, or discriminate against any individual for the purpose of interfering with any right or privilege secured by [Title VI], or because he has made a complaint, testified, assisted, or participated in any manner in an investigation, proceeding or hearing under this subpart." 28 C.F.R. � 42.108(e) (Department of Justice Regulation).

To establish a case based upon a claim of retaliation, the elements of proof utilized in a Title VII retaliation complaint again provide a useful guide:

1) that the person allegedly retaliated against engaged in a protected activity (i.e., filed a complaint, participated in an investigation, testified in a hearing, etc.);

2) that the person retaliating was aware of the protected activity;

3) that the party alleging the violation suffered adverse treatment after engaging in the protected activity; and

4) that there was a causal connection between the protected activity and the adverse action.

See Grant v. Bethlehem Steel Corp., 622 F.2d 43, 46 (2nd Cir. 1980). See also Davis v. Halpern, 768 F.Supp. 968, 985 (E.D.N.Y. 1991) (Defendant's summary judgment motion to dismiss Title VI retaliation claim was denied because plaintiff established evidence of prima facie case). It is important to remember that it is not necessary that any underlying discrimination complaint be proven in order to make a prima facie case, and ultimately prove, a subsequent retaliation complaint. DeCintio v. Westchester County Medical Ctr., 821 F.2d 111, 116, n. 8 (2d Cir.1987), cert. denied, 484 U.S. 965 (1987).

Back to top

4. Conclusions Drawn From the Analysis of the Data or Other Evidence Already Gathered

You should carefully analyze what you have received from the complainant and from the recipient (if, for example, you have received a position statement or the complainant has provided you with letters written to him/her by the recipient). This information will be important in deciding how much and what kind of additional information you will need to request and which theory of discrimination may apply. You may also wish to conduct some preliminary interviews with the complainant's witnesses and include this information in your analysis of how to proceed.

Back to top

5. Description of the Evidence Required to Complete the Investigation and the Best Sources and Means of Obtaining Each Type of Evidence

Documentary evidence is in written form, and may consist of business records, memoranda, letters , applications, charts , logs, handwritten notes, etc; virtually any material or format. You should also include computerized data within the general category of documentary evidence, and consider whether data should be requested in computerized form and "hard copy." If voluminous records are needed from the recipient, it will be useful and timesaving to obtain data in computer form in order that assessments and calculations can be made more readily. Documentary evidence is essential from the recipient, and in many cases from the complainant, in order to fully investigate a complaint of discrimination.

Testimonial evidence refers to oral evidence. To obtain testimonial evidence, you should develop interview questions based on oral and written information and any other available data, and conduct interviews with the complainant, recipient's staff, and witnesses, as appropriate. Remember that, in gathering evidence to investigate and prove your case, you should not only look for evidence to support a prima facie case, but also to test the validity or truthfulness of any stated or anticipated defenses that the recipient has or may assert in your case. Therefore, you will want to include a statement of the likely or enunciated defenses of the recipient and describe the evidence you will need to test their validity. By addressing asserted or anticipated defenses "up front" when you plan your investigation and identify the evidence you will want to obtain, you will save yourself (and the recipient) a lot of time and the possible aggravation of additional requests for data and interviews.

Back to top

a. Types of Evidence

It is important that, during the planning and investigation/resolution process, you keep in mind the types of evidence you will need and how each type of evidence relates to the theory or theories of discrimination that apply to your case. Evidentiary proof is an inductive process where demonstrable facts (e.g., items of evidence) serve as building blocks to structure a determination of compliance or noncompliance. This "structure" can only be as sound as the evidence that is its foundation. You must keep in mind that different types of evidence can contribute in different ways to the proof for your findings; a little of one kind of evidence may be just as good or better than a lot of another type of evidence, if the latter is weak and unreliable.

Back to top

(1) Direct Evidence

Direct evidence is evidence of the actual, subjective intent of the person(s) charged with discrimination. It may take the form of an admission of discriminatory purpose, although this will rarely occur. You will most often find such admissions during an interview, when a person is explaining or justifying his or her actions (e.g., "... but you know that women don't really want that kind of work, so why train them for it?").(33)

Direct evidence encompasses more than just admissions; it includes any facts tending to establish the subjective motives of persons involved in the alleged discrimination. This might include any of the following:

• public statements or speeches
• minutes of hearings
• facially discriminatory actions/legislation; and
• contemporaneous statements (e.g., attributed by third parties).

Back to top

(2) Circumstantial Evidence

Circumstantial evidence includes facts from which one may infer intent or discriminatory motive; it generally forms the bulk of investigative findings in disparate treatment cases. Circumstantial evidence proves intent by using objectively observable data (i.e., focusing on the results or effects of an action or on the action itself). It does not, however, prove anything directly about actual subjective intent (i.e., why an action was taken). Circumstantial evidence collectively leads to an inference of discriminatory motive. For example, historical information on how members of the protected group at issue have been treated by the recipient, and the extent of similar complaints (and corrective actions taken, if any) can aid in developing an inference of discriminatory intent and the context for the claim at issue.

Back to top

(3) Comparative Evidence

Comparative evidence is most often thought of as that which identifies difference(s) in treatment accorded similarly situated individuals or groups based on their identification with or membership in a protected class. Comparative evidence may also focus on program or employment results. For example, this might involve comparing the quality and quantity of services provided one group with another. Properly presented, comparative evidence constructs, through rules of logic, the following rebuttable proposition:

In order to be probative, the comparative evidence used must demonstrate two things: first, that persons who are similarly situated are being compared(34) and, second, that the comparisons being made are significant and inclusive. You must show that the sample of "similarly situated" is sufficient from which to draw a conclusion and that all or a statistically representative sample of those persons who are "similarly situated" have been included in the universe from which you selected your sample.

Back to top

(4) Statistical Evidence

Statistical evidence is an important category of comparative evidence. It generalizes about the experience of a class rather than focusing on the experience of an individual member of that class. This type of evidence is important because, if the proper statistical techniques are rigorously applied, the resulting inference has a certain and predictable validity (e.g., there is a 95 percent probability that the results are not due to chance). The statistical proposition is very similar to the comparative proposition, varying only in its scope:

It is important to note that statistics alone will not prove an individual claim of disparate treatment. Statistical evidence, however, can be helpful in proving individual cases of disparate treatment because it can be used as circumstantial evidence to establish the presence of discriminatory motive.

Statistics are most often used under the disparate impact model to demonstrate the adverse effect of a procedure, policy, rule, selection criteria, or method of administration. Where evidence of disparate impact reaches significant levels, statistics alone may establish a prima facie case of discrimination.(35) In order to do this, you must present statistical evidence that demonstrates that a rule, procedure, policy, practice, or method of administration has a substantial disparate impact and that it is statistically improbable that the observed pattern of events occurred by chance. As a rule of thumb, the investigator must show that the observed (i.e., the actual) results had a chance occurrence probability of less than five percent or were at what statisticians call the ".05 level of confidence." In such a case, your statistics will meet two tests:

1) that a substantial disparate impact exists, and

2) that the observed impact is statistically significant (i.e., not due to chance).

It is important to keep in mind that statistics are not irrefutable. Like any other kind of evidence, they may be rebutted. In addition, while statistics are the central evidence for a disparate impact case, most successful cases also include anecdotal and/or other types of evidence to support the statistics. For example, witnesses who have been denied services because of the criteria at issue bring the statistics "to life" and help establish a more compelling case.

Back to top

Evidence to Prove Disparate Treatment

You must prove that the recipient had a discriminatory motive to establish a disparate treatment violation. Evidence of discriminatory intent may be shown through direct evidence, or more commonly, through circumstantial evidence.(36)

(1) Direct Evidence of Motive/Intent

Direct evidence directly establishes the recipient's discriminatory motive and is one element of a prima facie case of disparate treatment. Most often, this will be statements by recipient officials or decision-makers expressing a biased or stereotypical view. For example, "...we send most of our Black patients to the City hospital; you know, they never have health insurance...".

Back to top

(2) Circumstantial Evidence of Motive/Intent

Circumstantial evidence includes facts from which one may infer intent or discriminatory motive; it generally forms the bulk of the evidence supporting investigative findings in disparate treatment cases. Circumstantial evidence can include, but is not limited to, comparative evidence regarding the treatment of similarly-situated individuals, and statistical evidence of an adverse impact on a protected group.

Direct and circumstantial evidence can be found in various sources, including:

• statements by decision-makers (e.g., stereotypical or biased views towards members of the complainant's class);
• statistical data;
• the historical background of the events at issue;
• the sequence of events leading to the decision at issue (e.g., the recipient's failure to respond to complaints of denial of services or racial harassment by members of its staff, or changes in eligibility requirements after protected class members became a significant proportion of those eligible for a program);
• a departure from standard procedure (e.g., failure to consider factors normally considered);
• legislative or administrative history (e.g., minutes of meetings); and,
• a past history of discriminatory or segregated conduct.

See Arlington Heights v. Metropolitan Housing Authority, 429 U.S. at 266-68 (evaluation of intentional discrimination claim under the Fourteenth Amendment); Elston, supra, 997 F.2d at 1394, 1406.

Back to top

(3) Recipient's Defense of a Prima Facie Intent Case

In a disparate treatment/intentional discrimination case, the recipient can rebut the prima facie case by showing that the finding is factually incorrect or demonstrating that the discrimination was required by law, for example by showing that actions were taken pursuant to a consent decree or remedial plan. The most common defense is that the treatment at issue was due to an alternative nondiscriminatory reason.(37) Defenses also may attempt to undermine the credibility of the evidence establishing intent. In addition, the defendant recipient may also present other members of the complainant's class who were not treated like the complainant or a person not of the complainant's class who was treated in the same manner as the complainant.(38)

Back to top

(4) Overcoming the Recipient's Defense in an Intent Case

If the recipient's defense is not based upon a Federal requirement, the investigator may show that the rebuttal evidence presented by the recipient was merely a "pretext" for discrimination. Often the determination of whether or not the asserted nondiscriminatory "reason" provided by the recipient is pretextual is based on the investigator's judgment, rather than documentary evidence. Where facts are in dispute, the investigator should attempt to corroborate the facts independently.

Rebuttal evidence may challenge the credibility of the recipient's nondiscriminatory reasons for its actions. More often, however, rebuttal evidence may not be any different than the evidence of intent presented to establish the prima facie case. Since the ultimate evaluation in disparate treatment cases is often one of assessing witness credibility, the strength of evidence to prove discriminatory motive is critical. Types of evidence that may be helpful in proving pretext are:

• the recipient failed to follow its own rules, policies, and procedures in making the decision;
• the recipient acted inconsistently with its own stated legitimate nondiscriminatory reason;
• testimony or documentary evidence obtained in the investigation contradicts the nondiscriminatory reasons; or
• the reason offered now was not offered to support the challenged decision at the time it occurred, suggesting the reason was offered as an afterthought.

Back to top

c. Evidence to Prove Disparate Impact

Rather than seeking to prove that the recipient had a discriminatory motive, you are seeking to prove that a policy or practice has an adverse impact. The evidence you will need to gather in an investigation of a case involving disparate impact will likely include both statistical and comparative evidence.(39) You will initially want to determine whether there is a disproportionate representation of protected class members participating in the program in question (e.g., four percent of participants in a housing program are African American, while 43 percent of the statutorily eligible population is African American). If this is the case, you will want to look at the application process and other aspects of program administration to determine if there is evidence that a policy or practice is causing the disparity. In complex cases, this could involve a separate analysis of each step or requirement in the process.(40)

You will need to develop evidence that will allow you to test for adverse impact by making a comparison of the effects of the policy, requirement, or practice in question on members of the complainant's protected class with persons not in the protected class. You will need to determine the representation of people within the protected class as a whole prior to the application of the policy/requirement (e.g., the number of statutorily eligible low income African Americans in a metropolitan area), as well as the proportion of protected class members who remain after you apply the policy/requirement in question (e.g., those low income African Americans who can meet a residency requirement). You must then determine the same numbers for persons outside the protected class in both circumstances. If there is a statistically significant disparity between the proportion of protected class members remaining after application of the policy or practice when compared with that proportion of persons not in the protected class, you have established a prima facie case.

For example, assume that 200 African Americans and 300 whites apply for benefits from a Federally funded disaster relief program after a hurricane destroys a town. The African Americans make up 40% of the applicants, and the whites make up 60% of the applicants. However, even though all of the applicants are found to be eligible for the relief in question, the local recipient administering the program provides benefits to only 100 of the African American applicants, while providing relief to all of the white applicants. Of those receiving relief, African Americans comprised 25 % and the whites comprised 75%. Stated another way, 50% of the African Americans received assistance, while 100% of the whites received assistance.

By establishing a prima facie case in this manner, the burden shifts to the recipient to defend its policy or practice.

If the number of protected class members in the universe of potential applicants for or participants in the program is small, you will most likely not be able to develop statistical evidence that is reliable because of the limited numbers. In that instance, you will want to develop your case based upon comparative evidence to determine whether essentially all of the protected class members are eliminated based upon the policy or procedures in question. For example, a town to which a small population of Native Americans moved within the past five years requires seven years of residency in the town in order to qualify for certain services. Even though you may not be able to develop a statistical case because the number of Native Americans is small, you may be able to prove disparate impact by showing that virtually all Native Americans will be excluded from services based upon the application of this rule, yet nonminorities are not similarly affected. (Note: You would also want to look for evidence of disparate treatment/discriminatory intent in a case such as this. For example, proof that the town changed its residency requirements only after Native Americans moved in may provide circumstantial evidence to prove intentional discrimination.)

Even in a case that lends itself to statistical analysis, you will want to conduct your investigation so that you are not only analyzing the "numbers," but also looking behind the statistics at the actual practices. You will want to include as witnesses, to the extent possible, examples of individuals who are qualified yet who are adversely affected by the policy or practice in question. This type of evidence will add "life" to your statistics and help rebut the recipient's justification for its policy or practice.

Back to top

Recipient's Defense to a Prima Facie Impact Case

In a disparate impact case, the recipient must establish a "substantial, legitimate justification" for the policy or practice in question to rebut a prima facie finding against it. In some cases, the recipient may chose to rebut the presumption in some other fashion, such as disputing the facts, presenting new or different facts, or challenging the statistical analysis. More specifically, some examples of rebuttals to a prima facie finding of disparate impact include:

• the finding is factually incorrect;
• the individuals compared were not "similarly situated";
• the comparison was not inclusive (i.e., some persons who were similarly situated were not included);
• the statistical evidence is not probative or was inaccurately calculated (e.g., irrelevant to the issue, insufficient sample, inappropriate universe for comparison, etc.);
• presentation of contrary statistical evidence;
• demonstration that actions were taken pursuant to a consent decree or remedial plan;
• proof that the rule, policy, procedure, etc., under attack did not cause the adverse impact (i.e., by questioning the "significance" of the finding); or
• demonstration that a facially neutral practice or procedure had a manifest relationship to a legitimate business or program objective of the recipient.

Back to top

(2) Overcoming the Recipient's Defense in a Prima Facie Impact Case

If the recipient shows a disputed practice is necessary to achieve a program objective, an investigator may rebut this by showing that the practice is not necessary to achieve the objective or by showing that there is a less discriminatory alternative that achieves the same objective.(41) The alternative must meet all the recipient's needs or overall objectives. The determination of whether there is a less discriminatory method to accomplish the recipient's asserted business or program necessity frequently requires that the investigator possess a thorough understanding of the recipient's program and how it operates. The more you work with your recipients, the more knowledge you will develop in how they work and the options available that could meet the specific "business" or program needs so as not to result in discrimination. You may find that, in especially important or complex cases, it is necessary to hire a consultant with the specialized knowledge and experience you need to help you determine whether a less discriminatory alternative is available.

Back to top

d. Quality and Usefulness of Evidence

Your investigation must gather sufficient good quality evidence. To gather too much or unreliable evidence is to waste valuable resources -- both your own and that of the recipient and complainant. You must keep some guidelines in mind when developing your Investigative Plan or complaint investigation/resolution approach in order to make the most of the information you gather. In order to support a determination of noncompliance, the evidence contained in the case file must be material to the allegation raised, relevant, and reliable.

Back to top

(1) Material evidence

Material evidence relates to one or more of the issues raised in an allegation. You must make sure that every piece of evidence you seek is material to one or more of the allegations raised in the complaint, or to an issue that your agency has determined must be investigated.

Back to top

(2) Relevant evidence

Relevant evidence is that which tends to either prove or disprove an issue raised in an allegation. For example, if a complainant alleged race discrimination in the provision of program benefits, information concerning the sex of the trainees would not be relevant proving race discrimination.

Back to top

(3) Reliable evidence

Reliable evidence is that which is dependable and/or trustworthy. You should always attempt to get the most reliable evidence. Some factors to consider in determining whether testimony is reliable are:

• whether the witness is disinterested in the outcome of the complaint;
• whether the witness is qualified to testify concerning matters in his or her statement; and,
• whether the statements are factual rather than conclusory.

For example, if the head of a department alleged to have discriminated states the events never happened, his/her testimony should be taken with a "grain of salt" for it is in his/her interest to deny the events at issue. The weight to be given such testimony depends on whether it is corroborated by other evidence. Similarly, the reliability of a statement by a program administrator that an alleged discriminatory practice is absolutely necessary for the functioning of the program depends on whether the administrator has the knowledge, experience, and expertise to know what is "absolutely necessary" and whether there is corroborating evidence.

Investigators will discover that, in most cases, statements will come from witnesses who do not meet all of the criteria for reliability. The weight to be given such testimony requires a balancing of all of the factors, such as, are the statements consistent or inconsistent with other witnesses, are the statements supported by objective evidence, etc. Reliability is a relative concept and requires the exercise of judgment. It is a standard to be used in determining the weight that a particular piece of evidence should be given relative to other evidence gathered during the investigation. Evidence that meets the standards of reliability should be given more weight in making a finding of discrimination than evidence that does not.

Back to top

e. Quantity of Evidence

While planning for the investigation, you must make a preliminary determination concerning the amount of evidence you will need to resolve the allegation(s) raised. This decision will be reexamined periodically throughout the investigation/resolution process, based upon the information you are gathering.

Initially, you will want to request enough information from the complainant to have a clear picture of the allegations -- the who, what, when, where, why -- the evidence that the complainant believes would help support his or her assertion that discrimination has occurred.(42) Complainants can be very helpful in providing information on the types of records a recipient keeps that will lend support to their allegations. They can also suggest important witnesses to interview who could give testimony to support their allegations. In addition, the initial data request to the recipient should ask the recipient to identify and submit documentation to support its defenses and understanding of the events at issue.

Further, before reviewing material or interviewing witnesses, ask the complainant or recipient what the documents that they want you to examine, or the person(s) whom they want you to interview, will reveal and how the documents or interview will support their respective positions. Either party may, knowingly or not, attempt to "drown" you in documents or interviews to convince you of his/her position, even though those documents or interviews will not provide relevant evidence.

It is important to note, however, that you are not required to interview all of a complainant's witnesses if the evidence you have gathered shows that additional interviews will not change your findings. Similarly, you are not required to interview all of a recipient's witnesses if you have documentation or irrefutable testimony to show that additional witness statements will not change your findings.

The investigator need only exhaust all sources likely to support the complainant and all sources likely to support the recipient; there need not be equal amounts of evidence for each party. An investigation conducted in this manner might reveal that there is no evidence to support the complainant's allegations and ample evidence to support the recipient's version of the facts. In this case, the investigation would be complete.

Back to top

6. Anticipated Sequence of Case Activities, Including Onsite Visits, if Needed

Your anticipated sequence of activities will be important in proving whether discrimination occurred or not. The burden of proof is on the investigator to show that the recipient has violated the statute. You want to conduct your investigation in such a manner that you can anticipate what you will need and when you will need to have it. The first step is generally the notification to the recipient and a request for information. As you plan your investigation, you must allow yourself sufficient time to request and analyze the initial data prior to conducting interviews either by telephone or onsite. You should plan your investigation carefully, to avoid having to make many subsequent data requests and to return onsite for additional interviews.

Back to top

7. Anticipated Timeframes for Obtaining and Analyzing Evidence

Your agency may require that you include timeframes in your IP, especially if its regulations or a court order provide such requirements. Even if you are not required to include timeframes, it may be helpful to set out some general timeframes for yourself as a guideline to keep your investigation "moving."

Back to top

8. Statement of Likely or Enunciated Recipient Defenses and a Description of the Evidence Required to Test Their Validity

This section of the IP is essential if you are to develop a legally sufficient showing that discrimination has occurred. As we discussed in the section on "Applicable Legal Theories" earlier in this Chapter, the burden of proof is initially upon the agency to establish a prima facie case through the preponderance of evidence on the record. Where this is accomplished, the burden then shifts to the recipient. The recipient may rebut the prima facie case by providing a valid, reasonable, nondiscriminatory purpose for the actions it has taken. If it does so, the burden then shifts back to the investigating agency to demonstrate that the stated reasons are merely a pretext for discrimination. A "pretext" is an excuse to disguise discrimination. The burden of "persuasion" always remains with the agency. The "preponderance of evidence" simply means that the evidence on one side outweighs the evidence on the other side.

Back to top

a. Establishing a prima facie case

In a civil rights context, a prima facie case is made by presenting sufficient evidence which, if unrebutted, would be legally sufficient to establish a discrimination violation. Mere allegations of discrimination do not constitute a prima facie case. As already discussed, what constitutes a prima facie case will change depending on the theory of proof utilized. Already discussed were the general rules for proving a prima facie case under Title VI utilizing both an disparate impact/effects and an disparate treatment/intent standard.

Back to top

b. Recipient's Rebuttal of a Prima Facie Case

In every case, after a prima facie case has been established, the recipient should be provided the opportunity to rebut the evidence presented. This is to ensure that the investigator has not overlooked anything.

In a disparate treatment/intentional discrimination case, the recipient can rebut the prima facie case by showing that the finding is factually incorrect or demonstrating that the discrimination was required by law, for example by showing that actions were taken pursuant to a consent decree or remedial plan. The most common defense is that the treatment at issue was due to an alternative nondiscriminatory reason.

In a disparate impact case, the recipient need only articulate a legitimate nondiscriminatory reason for its actions. This means that the recipient need establish the existence of a nondiscriminatory reason for the demonstrated disparate impact. The reason must be clear and reasonably specific. In some cases, the recipient may chose to rebut the presumption in some other fashion, such as disputing the facts or presenting new or different facts. Following are some examples of typical rebuttals of this type to a prima facie finding of disparate impact:

• the finding is factually incorrect;
• individuals compared were not "similarly situated";
• the comparison was not inclusive (i.e., some persons who were similarly situated were not included);
• statistical evidence is not probative (e.g., irrelevant to the issue, insufficient sample, inappropriate comparison, etc.);
• presentation of contrary statistical evidence;
• demonstration that actions were taken pursuant to a consent decree or remedial plan;
• use of rebuttal witnesses (e.g., other members of the complainant's class who were not treated like the complainant; a person not of the complainant's class who was treated in the same manner as the complainant);
• the difference in treatment was based on some other objective reason (e.g., qualifications, favoritism, nepotism, etc.);(43)
• proving that the rule, policy, procedure, etc., under attack did not cause the adverse impact (i.e., by questioning the "significance" of the finding); or
• demonstrating that a facially neutral practice or procedure had a manifest relationship to a legitimate business or program objective of the recipient.

Back to top

c. Overcoming the Recipient's Rebuttal

Under the disparate treatment theory, once you have shown intentional bias -- discriminatory motive -- the recipient has virtually no defense. Most defenses will attempt to undermine the credibility of the evidence establishing intent and attempt to show the alleged discrimination actually resulted because of nondiscriminatory reasons.

This is not the situation in a disparate impact case. A recipient need only show proof of business necessity or other legally acceptable reason to escape liability. For example, if the Federal program statute required that a program be conducted in a particular way, even if a disparate impact results, following Federal law would be a valid defense.

If the recipient's defense is not based upon a Federal requirement, the investigator may show that the rebuttal evidence presented by the recipient was merely a "pretext" for discrimination. Often the determination of whether or not the nondiscriminatory "reason" provided by the recipient is pretextual is based on the investigator's judgment, rather than documentary evidence. Where facts are in dispute, the investigator should attempt to corroborate the facts independently.

If the recipient shows a disputed practice was necessary to achieve a program objective, an investigator may rebut this by showing that the practice is not necessary to achieve the objective or by showing that there are less discriminatory alternatives that achieve the same objective. The determination of whether there is a less discriminatory method to accomplish the recipient's asserted business or program necessity frequently requires that the investigator possess a thorough understanding of the recipient's program and how it operates. The more you work with your recipients, the more knowledge you will develop in how they work and the options available that could meet the specific "business" or program needs so as not to result in discrimination. You may find that, in especially important or complex cases, it is necessary to hire a consultant with the specialized knowledge and experience you need to help you determine whether a less discriminatory alternative is available.

Back to top

C. Data Collection

The importance of diligently gathered and thoroughly analyzed data cannot be overstated. Civil rights cases can be made or broken on the basis of well thought-out data analyses that are precise, systematic, and resistant to rebuttal.

Data is gathered during the course of the investigation to answer two main questions concerning the allegations:

What happened? (Including when, where, and how.) The complaint alleges that something did or did not happen. Data must be gathered to determine whether the alleged event occurred or not.