TANG, Circuit Judge:
We grant en banc review in this Title VII race discrimination case to decide two questions. First, we decide the procedure a panel should follow when faced with an irreconcilable conflict between the holdings of controlling prior decisions of this court. Second, we resolve that irreconcilable conflict, by deciding that disparate impact analysis may be applied to subjective employment practices. The district court declined to apply disparate impact analysis on the authority of Heagney v. University of Washington, 642 F.2d 1157 (9th Cir.1981) (practice of hiring without well-defined criteria cannot be subjected to disparate impact analysis) and chose to disregard the later decision in Wang v. Hoffman, 694 F.2d 1146 (9th Cir.1982) (lack of objective criteria for promotion can be analyzed for disparate impact). The Ninth Circuit panel that heard the appeal from the judgment for the employers in the instant case noted our conflicting decisions but held it was bound by Heagney because it expressed the “correct view” or, alternatively, because it was the decision “first in line.” Atonio v. Wards Cove Packing Co., Inc., 768 F.2d 1120, 1132 and n. 6 (9th Cir.1985), withdrawn, 787 F.2d 462 (9th Cir.1985).
The panel’s approach did not resolve the broader question of how future panels should decide a case controlled by contradictory precedents. We now hold that the appropriate mechanism for resolving an ir[1479]*1479reconcilable conflict is an en banc decision. A panel faced with such a conflict must call for en banc review, which the court will normally grant unless the prior decisions can be distinguished. Despite the “extraordinary” nature of en banc review, United States v. American-Foreign Steamship Corp., 363 U.S. 685, 689, 80 S.Ct. 1336, 1339, 4 L.Ed.2d 1491 (1960), and the general rule that en banc hearings are “not favored,” Fed.R.App.P. 35(a), en banc review is proper “when consideration by the full court is necessary to secure or maintain the uniformity of its decisions.” Fed.R.App.P. 35(a)(1); see also American-Foreign Steamship, 363 U.S. at 689-90, 80 S.Ct. at 1339-40.
Turning to the substantive question which produced our conflicting prior decisions, we note that this case arises out of the cannery workers’ allegations of both disparate treatment and disparate impact. Thus it affords us the opportunity to refine the analytic tools for the identification and eradication of unlawful discrimination. Specifically, we now determine that disparate impact analysis may be applied to subjective employment practices.
I. BACKGROUND
Former salmon cannery workers brought a class action suit charging three companies with employment discrimination on the basis of race in violation of Title VII of the Civil Rights Act of 1964, 42 U.S.C. § 2000e et seq. (1982) and the Civil Rights Act of 1866, 42 U.S.C. § 1981 (1982). The class alleged both disparate treatment and disparate impact claims on behalf of minority persons. It alleged that the pronounced concentration of Asian and Alaska Native employees in the lowest paying cannery worker and laborer positions and the relative scarcity of such minority employees in the higher paying positions proved disparate treatment of minority people. It also alleged that certain specific employment practices of the companies proved both disparate treatment of and disparate impact on minority people. The cannery workers challenged the companies’ use of separate hiring channels for cannery workers from those used for the higher paying, at-issue jobs, as well as word-of-mouth recruitment, nepotism, rehire policies, and the lack of objective job qualifications.
The' majority of cannery workers are hired from native villages in Alaska and through a local union of primarily Filipino members of the International Longshoremen’s and Warehousemen’s Union (ILWU) in Seattle. Consequently, cannery workers are almost all members of these ethnic groups. All other positions are filled through applications received during the off-season at the mainland home offices, through rehiring previous employees and through word-of-mouth recruitment. These positions are held predominantly by white people. Another challenged practice, of particular relevance in our en banc review of this case, is the apparent lack of objective qualifications for many job classifications, and the resultant use of subjective criteria in hiring and promoting. When filling most job positions, the respective hiring officers generally seek to hire the individuals who are, in the hiring officer’s opinion, the best for the job.
In addition to the racial stratification of jobs, the cannery workers complain that even those nonwhites who obtain positions with the companies are treated differently from whites. They allege that nonwhites are segregated from whites in housing and messing, and that the bunkhouses and food provided for nonwhites are far inferior to those provided for whites.
In holding for the defendant companies, the district court evaluated the evidence introduced by both sides, including conflicting statistical data. The court analyzed all the cannery workers’ claims for intentional discrimination, and concluded that the companies had successfully shown nondiscriminatory motivations for their practices. Despite the cannery workers’ contrary arguments, the court, relying on Ninth Circuit authority, refused to evaluate all of the claims under the disparate impact model of Title VII. The court subjected a few [1480]*1480claims to disparate impact analysis and again found for the defendants.
II. ANALYSIS
A. Title VII Liability
Section 703(a)(2) of Title VII, 42 U.S.C. § 2000e-2(a)(2) (1982), provides that:
It shall be an unlawful employment practice for an employer—
(2) to limit, segregate, or classify his employees or applicants for employment in any way which would deprive or tend to deprive any individual of employment opportunities or otherwise adversely affect his status as an employee, because of such individual’s race, color, religion, sex, or national origin.
An employee may prove an employer’s Title VII liability through a theory of disparate treatment or a theory of disparate impact. Proof of disparate treatment requires a showing that the employer intentionally “treats some people less favorably than others because of their race, color, religion, sex, or national origin.” International Brotherhood of Teamsters v. United States, 431 U.S. 324, 335 n. 15, 97 S.Ct. 1843, 1854-55 n. 15, 52 L.Ed.2d 396 (1977). An illicit motive may be inferred in an individual discrimination claim when the plaintiff shows he is a member of a protected class who applied for, and failed to get, a job for which he was qualified and which remained open after his rejection. McDonnell Douglas Corp. v. Green, 411 U.S. 792, 802, 93 S.Ct. 1817, 1824, 36 L.Ed.2d 668 (1973). An illicit motive may be inferred in a class-wide discrimination claim from a sufficient showing of disparity between the class members and comparably qualified members of the majority group. Segar v. Smith, 738 F.2d 1249, 1265-66 (D.C.Cir. 1984), cert. denied, 471 U.S. 1115, 105 S.Ct. 2357, 86 L.Ed.2d 258 (1985) (citing Teamsters, 431 U.S. at 335 n. 15, 97 S.Ct. at 1854-55 n. 15).
A disparate impact claim challenges “employment practices that are facially neutral in their treatment of different groups but that in fact fall more harshly on one group than another and cannot be justified by business necessity.” Teamsters, 431 U.S. at 336 n. 15, 97 S.Ct. at 1854-55 n. 15. Illicit motive is irrelevant because impact analysis is designed to implement Congressional concern with “the consequences of employment practices, not simply the motivation.” Griggs v. Duke Power Co., 401 U.S. 424, 432, 91 S.Ct. 849, 854, 28 L.Ed.2d 158 (1971) (emphasis in original). In a class action suit, commonly known as a “pattern or practice” case, plaintiffs typically assert claims both of disparate treatment occasioned by an employer’s practices and of disparate impact produced by those practices. Segar, 738 F.2d at 1266. As the Supreme Court noted in Teamsters, a pattern and practice class action case, “[ejither theory may, of course, be applied to a particular set of facts.” 431 U.S. at 336 n. 15, 97 S.Ct. at 1854-55 n. 15.
B. Impact Analysis in the Ninth Circuit
1. Conflict
Disparate treatment and disparate impact are but two analytic tools which may be used in the appropriate Title VII case to resolve the ultimate question, whether there has been impermissible discrimination by an employer. See, e.g., Goodman v. Lukens Steel Co., 777 F.2d 113, 130 (3d Cir.1985). Despite the Teamsters language stating that either theory may be applied to a set of facts, courts have not uniformly interpreted the scope of impact analysis.1 [1481]*1481Differences have arisen from the conflicting views of whether impact analysis can be applied to evaluate employment procedures or criteria different from the objective test and diploma requirement scrutinized in the seminal Griggs decision or the height and weight requirements analyzed in Dothard v. Rawlinson, 433 U.S. 321, 97 S.Ct. 2720, 53 L.Ed.2d 786 (1977).2
This circuit has clearly held that subjective practices and decisions are not illegal per se. Heagney v. University of Washington, 642 F.2d 1157, 1163 (9th Cir.1981). At the same time, we have stated that subjective practices are particularly susceptible to discriminatory abuse and should be closely scrutinized. Kimbrough v. Secretary of United States Air Force, 764 F.2d 1279, 1284 (9th Cir.1985); Nanty v. Barrows Co., 660 F.2d 1327, 1334 (9th Cir. 1981). The conflict in our decisions has developed because prior panels have not all agreed that the close scrutiny of subjective practices can or should take the form of a disparate impact analysis.
In Heagney, the plaintiff challenged the University’s power to classify certain jobs as “exempt” from state personnel laws, which, in turn, gave the school more discretion in setting salaries. We held that the crux of the complaint was an objection to the lack of well-defined criteria, which could not be equated with practices such as personnel tests or minimum physical requirements. Thus, although we had previously noted that both treatment and impact analysis may be applied, we held that impact analysis was inappropriate. Heagney, 642 F.2d at 1163. We followed Heagney in O’Brien v. Sky Chefs, 670 F.2d 864, 866 (9th Cir.1982) and refused to apply impact analysis to an employer’s lack of well-defined promotion criteria, noting that the lack of such criteria does not per se cause an adverse impact.
On the other hand, this court has applied impact analysis to subjective criteria in at least two cases. In Wang v. Hoffman, 694 F.2d 1146, 1148 (9th Cir.1982), which challenged the hiring and promotion policies of [1482]*1482the Army Corps of Engineers, the panel held that a promotion system lacking objective criteria could be challenged for its disparate impact, and in Peters v. Lieual-len, 746 F.2d 1390, 1392 (9th Cir.1984), the panel held that impact analysis could be applied to subjective criteria used during interviews to screen candidates, but that the plaintiff must show that use of the criteria caused the adverse impact. See also Yartzoffv. Oregon, 745 F.2d 557, 558 (9th Cir.1984) (impact analysis of subjective promotion criteria appropriate in age discrimination case, but plaintiff failed to offer proof of disparate impact).
In subsequent cases we have recognized the conflict between Heagney and Wang, but felt it unnecessary to resolve the question. See Moore v. Hughes Helicopters, Inc., 708 F.2d 475, 481 (9th Cir.1983) (noting that “[t]he law in this court is unsettled” stated disparate treatment focus well suited to analysis of subjective decision making); Spaulding v. University of Washington, 740 F.2d 686, 709 (9th Cir.) (lack of well defined criteria facilitating wage discrimination better presented under disparate treatment model on the authority of Heagney, followed by a “but cf.” citation to Wang), cert. denied, 469 U.S. 1036, 105 S.Ct. 511, 83 L.Ed.2d 401 (1984).
2. Resolution
We now hold that disparate impact analysis may be applied to challenge subjective employment practices or criteria provided the plaintiffs have proved a causal connection between those practices and the demonstrated impact on members of a protected class. The three elements of the plaintiffs’ prima facie cáse are that they must (1) show a significant disparate impact on a protected class, (2) identify specific employment practices or selection criteria and (3) show the causal relationship between the identified practices and the impact. We are persuaded that this holding comports with the express language of the statute, the intent of Congress as revealed in its discussions of the 1972 amendments, the enforcement agencies’ interpretation, and the broad prophylactic purposes of Title VII.
3. Rationale
a. Statutory Language
We begin with the observation that Title VII proscribes all forms of employment discrimination. It does so without reference to either objective or subjective practices. Title VII states that it is an unlawful employment practice “to limit, segregate, or classify ... employees or applicants for employment in any way.” 42 U.S.C. § 2000e-2(a)(2)(1982) (emphasis added). The Supreme Court construed this language as proscribing “not only overt discrimination but also practices that are fair in form, but discriminatory in operation.” Griggs, 401 U.S. at 431, 91 S.Ct. at 853. The Court developed the disparate impact model for proving discrimination in recognition of Congress’ intent to remove “artificial, arbitrary, and unnecessary barriers to employment.” Id. Although Griggs involved requirements of a high school diploma and an objective test, the opinion did not expressly limit impact analysis to such criteria.
b. Congressional Intent
There is considerable evidence that Congress endorsed the Griggs decision during discussion of amendments to Title VII in 1972. H.R.Rep. No. 238, 92d Cong., 1st Sess. 19, 24 (1971), reprinted in 1972 U.S. Code Cong. & Ad.News 2137, 2164; S.Rep. No. 415, 92d Cong., 1st Sess. 1, 14-15 (1971); Connecticut v. Teal, 457 U.S. 440, 447 n. 8, 102 S.Ct. 2525, 2531 n. 8, 73 L.Ed.2d 130 (1982); see Helfand and Pem-berton, The Continuing Vitality of Title VII Disparate Impact Analysis, 36 Mercer L.Rev. 939, 948-54 (1985). The section-by-section analyses of the 1972 amendments submitted to both houses of Congress expressly stated that in areas not addressed by the amendments, existing case law was intended to continue to govern. 118 Cong. Rec. 7166, 7564 (1972); Teal, 457 U.S. at 447 n. 8, 102 S.Ct. at 2531 n. 8. Thus, although Title VII was not amended specif[1483]*1483ically to extend disparate impact analysis to subjective practices, decisional law incorporated at that time included not only Griggs, but such cases as United States v. Bethlehem Steel Corp., 446 F.2d 652, 657-58 (2d Cir.1971), which applied Griggs to invalidate subjective hiring standards and procedures.
c. Enforcement Agencies’ Interpretation
Additional authority for our decision to apply the disparate impact model is found in the announcement of the four agencies charged with enforcement of Title VII — the Equal Employment Opportunity Commission, the Office of Personnel Management, the Department of Justice and the Department of Labor — that the law requires application of the disparate impact model to all selection procedures whether subjective or objective. Griffin v. Carlin, 755 F.2d 1516, 1525 (11th Cir.1985). The Uniform Guidelines on Employee Selection Procedures, adopted in 1978, define the procedures to which impact analysis applies as:
[a]ny measure, combination of measures, or procedure used as a basis for any employment decision. Selection procedures include the full range of assessment techniques from ... physical, educational, and work experience requirements through informal or casual interviews.
29 C.F.R. § 1607.16(Q) (1985).
Because the statutory language and legislative history support the administrative interpretation, the guidelines are “entitled to great deference,” and can be treated as “expressing the will of Congress.” Griggs, 401 U.S. at 434, 91 S.Ct. at 855.
d. Purpose of Title VII
Applying the tool of disparate impact analysis to subjective practices and criteria is necessary to fully implement the prophylactic purpose of Title VII to achieve equal employment opportunity and remove arbitrary and unnecessary barriers which have operated to favor white male employees over others. Teal, 457 U.S. at 451, 102 S.Ct. at 2532-33; Teamsters, 431 U.S. at 364, 97 S.Ct. at 1869; Griggs, 401 U.S. at 431, 91 S.Ct. at 853. It is essential precisely because such practices will quite often lack any discriminatory animus. Subjective practices can operate as “ ‘built-in headwinds’ ” for minority groups as readily as can objective criteria, Griggs, 401 U.S. at 432, 91 S.Ct. at 854, and these practices should likewise be exposed and eradicated when they cause adverse impact without proof of a redeeming business necessity. The Supreme Court has not held otherwise.
e. Fumco
There has been considerable discussion about the meaning of Fumco Construction Corp. v. Waters, 438 U.S. 567, 98 S.Ct. 2943, 57 L.Ed.2d 957 (1978). Some courts and commentators suggest the Supreme Court restricted the application of Griggs impact analysis to objective criteria.3 The majority of circuits, however, do not subscribe to this reading of Fumco and have applied impact analysis to subjective practices.4
The employment practice challenged in Fumco was the refusal to accept jobsite applications for bricklayers to reline blast furnaces with firebrick. Instead, the job superintendent hired only bricklayers he knew were experienced or who had been recommended by his foremen. Fumco, 438 U.S. at 570, 98 S.Ct. at 2946. In apply[1484]*1484ing the McDonnell Douglas formula of disparate treatment the Court noted the case did not implicate employment tests previously treated in Griggs and Albe-marle Paper Co. v. Moody, 422 U.S. 405, 412-13, 95 S.Ct. 2362, 2369-70, 45 L.Ed.2d 280 (1975) {Moody), or particularized physical requirements such as those discussed in Dothard, 433 U.S. at 329, 97 S.Ct. at 2726-27, and that it was not a pattern and practice case as was Teamsters, 431 U.S. at 358, 102 S.Ct. at 1866, Furnco, 438 U.S. at 575 n. 7, 98 S.Ct. at 2948-49 n. 7.
We do not read this footnote to preclude impact analysis of the claims presented in the case at bar. Clearly, the facts giving rise to allegations of discrimination may support a prima facie case of disparate treatment or disparate impact. See, Teamsters, 431 U.S. at 336 n. 15, 102 S.Ct. at 1854-55 n. 15 (“[e]ither theory may, of course, be applied to a particular set of facts.”) In other words, Furnco imposes no limitation on use of impact analysis beyond the restrictions inherent in demonstrating a prima facie case.
The Furnco plaintiffs identified a specific practice, but were unable to prove that the practice had an adverse impact on black bricklayers. 438 U.S. at 571, 98 S.Ct. at 2946. Because they failed to demonstrate disparate impact, they failed to establish a prima facie case of disparate impact, and thus, use of that analytic tool was inappropriate.
In contrast, the plaintiffs in this case contend they are consigned to lower paying jobs by a system of racial segregation implemented through a variety of specific employment practices. The statistics provide evidence of a significant disparate impact and the challenged practices are agreed to cause disparate impact. Thus, these plaintiffs are entitled to the application of impact analysis as an appropriate analytic tool to challenge the discriminatory effect of the companies’ practices because they have satisfied the elements of the prima facie case: a significant disparate impact on a protected class, the identification of specific employment practices or selection criteria and a causal relationship between the identified practice and the impact.
f. Logic Supports Impact Analysis
Although the language of the statute and Congressional discussions of Title VII, as well as Supreme Court pronouncements are sufficient authority for the application of disparate impact analysis to subjective employment practices, we should also note that we are unpersuaded by the defendants’ objections to our decision based on appeals to logic or social policy. Defendants argue that there is a logical basis for a distinction between objective and subjective practices and for the correlative categorization of the analysis of the proof of impermissible discrimination. In their view subjective practices are by nature and definition based upon intent and thus should be evaluated only for discriminatory animus. They argue that only objective practices can be evaluated for disparate impact.
We disagree. Subjective practices may well be a covert means to effectuate intentional discrimination, as the defendants point out, but they can also be engendered by a totally benign purpose, or carried on as a matter of routine adherence to past practices whose original purposes are un-discoverable. Subjective practices are as likely to be neutral in intent as objective ones.5 If, in fact, the subjective practices are a “covert means” to discriminate intentionally, by definition intent will be difficult to prove. If the practices are the cause of adverse impact, the purposes of Title VII are well-served by advancing proof of adverse impact, thereby obviating the necessity of proving intent. Proof of intent where adverse impact can be shown may be not only unnecessary but undesirable because of the animus the process generates.
[1485]*1485We also do not agree that only objective practices can be analyzed for disparate impact. When we view employment practices from the perspective of their impact on a protected class, we are unable to see a principled and meaningful difference between objective and subjective practices. There is no bright line distinction between objective and subjective hiring criteria, because almost all criteria necessarily have both subjective and objective elements. For example, while the requirement of a certain test score may appear “objective,” the choice of skills to be tested and of the testing instruments to measure them involves “subjective” elements of judgment. Such apparently “subjective” requirements as attractive personal appearance in fact include certain “objective” factors. Thus the terms merely represent extremes on a continuum, and cannot provide a line of demarcation to guide courts in choosing the appropriate analytic tool in a Title VII discrimination case.
Finally, we think a distinction between subjective and objective practices serves no legitimate purpose. To the contrary, preserving the distinction could serve to encourage employers to abandon “objective” criteria and practices in favor of “subjective” decision making as a means of shielding their practices from judicial scrutiny. It would subvert the purpose of Title VII to create an incentive to abandon efforts to validate objective criteria in favor of purely discretionary hiring methods. See Griffin v. Carlin, 755 F.2d 1516, 1525 (11th Cir. 1985) (“Rather than validate education and other objective criteria, employers could simply take such criteria into account in subjective interviews____ It could not have been the intent of Congress to provide employers with an incentive to use such devices rather than validated objective criteria.”).
g. Policy Considerations Support Impact Analysis
The defendants argue that the burden placed on an employer in an impact case is somehow made unduly onerous when the practices identified as having a disparate impact are subjective in nature. A class claim of disparate impact is essentially an allegation that a disparity in the position of nonwhites and whites, often proved through statistical evidence, is “the systemic result of a specific employment practice that cannot be justified as necessary to the employer’s business.” Segar, 738 F.2d at 1267. As in a disparate treatment claim, the initial burden is on the plaintiffs. To establish a prima facie case of disparate impact, the plaintiffs must prove that a specific business practice has a “significantly discriminatory impact.” Teal, 457 U.S. at 446, 102 S.Ct. at 2530; Dothard, 433 U.S. at 329, 97 S.Ct. at 2726-27. To reiterate, plaintiffs’ prima facie case consists of a showing of significant disparate impact on a protected class, caused by specific, identified, employment practices or selection criteria.
Once the plaintiff class has shown disparate impact caused by specific, identifiable employment practices or criteria, the burden shifts to the employer. The crucial difference between a treatment and an impact allegation is the intermediate burden on the employer. To rebut the prima facie showing of disparate impact the employer may refute the statistical evidence as in the treatment claim and show that no disparity exists. But if the employer defends by explaining the reason for the disparity he must do more than articulate that reason. He must prove the job relatedness or business necessity of the practice. Moody, 422 U.S. at 425, 95 S.Ct. at 2375. The Supreme Court’s decision in Burdine that the burden of persuasion always stays with the plaintiff in a treatment case expressly preserved the different allocation of burdens in an impact case. The Court stated that it “recognized that the factual issues, and therefore the character of the evidence presented, differ when the plaintiff claims that a facially neutral employment policy has a discriminatory impact on protected classes.” Texas Dept, of Community Affairs v. Burdine, 450 U.S. 248, 252 n. 5, 101 S.Ct. 1089, 1093 n. 5, 67 L.Ed.2d 207 (1981).
[1486]*1486Precisely what the employer must prove will vary with the factors of different job settings, but “[t]he touchstone is business necessity.” Griggs, 401 U.S. at 431, 91 5. Ct. at 853. In our view, proving business necessity is no more onerous in a case involving subjective practices than one involving objective practices, because in either case the employer is the person with knowledge of what his practices are and why he uses the methods and criteria he does, as well as the person with superior knowledge of precisely how his employment practices affect employees. See Se-gar, 738 F.2d at 1271; Pouncy v. Prudential Insurance Co., 668 F.2d 795, 801 (5th Cir.1982). The burden of proof on the employer is commensurate with the greater burden on the plaintiff to prove impact and to establish the causal connection between the practices and the impact. Once a challenged practice which causes disparate impact is identified, it does not place an unfair burden to ask an employer to justify the challenged practice.6 We emphasize that while proving business necessity may be “an arduous task,” Bunch v. Bullard, 795 F.2d 384, 393 n. 10 (5th Cir.1986), this burden will not arise until the plaintiff has shown a causal connection between the challenged practices and the impact on a protected class.
In weighing competing policy considerations urged by the defendants, primary guidance is provided by the purpose of Title VII, namely to eradicate the existence and effects of discrimination in employment. Treatment and impact analyses are interpretive constructions intended to provide guidance in evaluating the evidence presented in discrimination cases so as best to effectuate Congressional intent. In this case, that intent is best realized by a decision to apply disparate impact analysis to subjective employment practices.
CONCLUSION
For the reasons discussed, we hold that disparate impact analysis can be applied to subjective employment practices. To the extent our prior decisions have held to the contrary they are expressly overruled.
We return this cause to the panel to reconsider the district court’s disposition of the plaintiffs’ claims in light of this decision.