dissenting.
The Court holds today that without violating Title VII of the Civil Rights Act of 1964, 42 U. S. C. § 2000e et seq., a private employer may adopt a disability plan that compensates employees for all temporary disabilities except one affecting exclusively women, pregnancy. I respectfully dissent. Today’s holding not only repudiates the applicable administrative guideline promulgated by the agency charged by Con*147gress with implementation of the Act, but also rejects the unanimous conclusion of all six Courts of Appeals that have addressed this question. See Communications Workers v. American Tel. & Tel., 513 F. 2d 1024 (CA2 1975), cert. pending, No. 74-1601; Wetzel v. Liberty Mut. Ins. Co., 511 F. 2d 199 (CA3 1975), vacated on jurisdictional grounds, 424 U. S. 737 (1976); Gilbert v. General Electric Co., 519 F. 2d 661 (CA4 1975) (this case); Tyler v. Vickery, 517 F. 2d 1089, 1097-1099 (CA5 1975); Satty v. Nashville Gas Co., 522 F. 2d 850 (CA6 1975), cert. pending, No. 75-536; Hutchison v. Lake Oswego School Dist. No. 7, 519 F. 2d 961 (CA9 1975), cert. pending, No. 75-1049.
I
This case is unusual in that it presents a question the resolution of which at first glance turns largely upon the conceptual framework chosen to identify and describe the operational features of the challenged disability program. By directing their focus upon the risks excluded from the otherwise comprehensive program, and upon the purported justifications for such exclusions, the Equal Employment Opportunity Commission, the women plaintiffs, and the lower courts reason that the pregnancy exclusion constitutes a prima facie violation of Title VII. This violation is triggered, they argue, because the omission of pregnancy from the program has the intent and effect of providing that “only women [are subjected] to a substantial risk of total loss of income because of temporary medical disability.” Brief for EEOC as Amicus Curiae 12.
The Court’s framework is diametrically different. It views General Electric’s plan as representing a gender-free assignment of risks in accordance with normal actuarial techniques. From this perspective the lone exclusion of pregnancy is not a violation of Title VII insofar as all other disabilities are mutually covered for both sexes. This reasoning relies primarily upon the descriptive statement borrowed from *148Geduldig v. Aiello, 417 U. S. 484, 496-497 (1974): “There is no risk from which men are protected and women are not. Likewise, there is no risk from which women are protected and men are not.” Ante, at 138. According to the Court, this assertedly neutral sorting process precludes the pregnancy omission from constituting a violation of Title VII.
Presumably, it is not self-evident that either conceptual framework is more appropriate than the other, which can only mean that further inquiry is necessary to select the more accurate and realistic analytical approach. At the outset, the soundness of the Court’s underlying assumption that the plan is the untainted product of a gender-neutral risk-assignment process can be examined against the historical backdrop of General Electric’s employment practices and the existence or nonexistence of gender-free policies governing the inclusion of compensable risks. Secondly, the resulting pattern of risks insured by General Electric can then be evaluated in terms of the broad social objectives promoted by Title VII. I believe that the first inquiry compels the conclusion that the Court’s assumption that General Electric engaged in a gender-neutral risk-assignment process is purely fanciful. The second demonstrates that the EEOC’s interpretation that the exclusion of pregnancy from a disability insurance plan is incompatible with the overall objectives of Title YII has been unjustifiably rejected.
II
Geduldig v. Aiello, supra, purports to be the starting point for the Court’s analysis. There a state-operated disability insurance system containing a pregnancy exclusion was held not to violate the Equal Protection Clause. Although it quotes primarily from one footnote of that opinion at some length, ante, at 134-135, the Court finally does not grapple with Geduldig on its own terms.
Considered most favorably to the Court’s view, Geduldig established the proposition that a pregnancy classification *149standing alone cannot be said to fall into the category of classifications that rest explicitly on “gender as such,” 417 U. S., at 496 n. 20. Beyond that, Geduldig offers little analysis helpful to decision of this case. Surely it offends common sense to suggest, ante, at 136, that a classification revolving around pregnancy is not, at the minimum, strongly “sex related.” See, e. g., Cleveland Board of Education v. LaFleur, 414 U. S. 632, 652 (1974) (Powell, J., concurring). Indeed, even in the insurance context where neutral actuarial principles were found to have provided a legitimate and independent input into the decisionmaking process, Geduldig’s outcome was qualified by the explicit reservation of a case where it could be demonstrated that a pregnancy-centered differentiation is used as a “mere pretext . . . designed to effect an invidious discrimination against the members of one sex. . . .” 417 U. S., at 496-497, n. 20.
Thus, Geduldig itself obliges the Court to determine whether the exclusion of a sex-linked disability from the universe of compensable disabilities was actually the product of neutral, persuasive actuarial considerations, or rather stemmed from a policy that purposefully downgraded women’s role in the labor force. In Geduldig, that inquiry coupled with the normal presumption favoring legislative action satisfied the Court that the pregnancy exclusion in fact was prompted by California’s legitimate fiscal concerns, and therefore that California did not deny equal protection in effectuating reforms “ ‘one step at a time.’ ” Id., at 495. But the record in this case makes such deference impossible here. Instead, in reaching its conclusion that a showing of purposeful discrimination has not been made, ante, at 136, the Court simply disregards a history of General Electric practices that have served to undercut the employment opportunities of women who become pregnant while employed.1 More*150over, the Court studiously ignores the undisturbed conclusion of the District Court that General Electric’s “discriminatory attitude” toward women was “a motivating factor in its policy,” 375 F. Supp. 367, 383 (ED Va. 1974), and that the pregnancy exclusion was “neutral [neither] on its face” nor “in its intent.” Id., at 382.2
Plainly then, the Court’s appraisal of General Electric’s policy as a neutral process of sorting risks and “not a gender-based discrimination at all,” ante, at 136, cannot easily be squared with the historical record in this case. The Court, *151therefore, proceeds to a discussion of purported neutral criteria that suffice to explain the lone exclusion of pregnancy from the program. The Court argues that pregnancy is not “comparable” to other disabilities since it is a “voluntary” condition rather than a “disease.” Ibid. The fallacy of this argument is that even if “non-voluntariness” and “disease” are to be construed as the operational criteria for inclusion of a disability in General Electric's program, application of these criteria is inconsistent with the Court's gender-neutral interpretation of the company's policy.
For example, the characterization of pregnancy as “voluntary” 3 is not a persuasive factor, for as the Court of Appeals correctly noted, “other than for childbirth disability, [General Electric] had never construed its plan as eliminating all so-called 'voluntary' disabilities,” including sport injuries, attempted suicides, venereal disease, disabilities incurred in the commission of a crime or during a fight, and elective cosmetic surgery. 519 F. 2d, at 665. Similarly, the label “disease” rather than “disability” cannot be deemed determinative since General Electric's pregnancy disqualification also excludes the 10% of pregnancies that end in debilitating miscarriages, 375 F. Supp., at 377, the 10% of cases where pregnancies are complicated by “diseases” in the intuitive sense of the word, ibid., and cases where women recovering from childbirth are stricken by severe diseases unrelated to pregnancy.4
*152Moreover, even the Court’s principal argument for the plan’s supposed, gender neutrality cannot withstand analysis. The central analytical framework relied upon to demonstrate the absence of discrimination is the principle described in Geduldig: “There is no risk from which men are protected and women are not . . . [and] no risk from which women are protected and men are not.” 417 U. S., at 496-497, quoted, ante, at 138. In fostering the impression that it is faced with a mere underinclusive assignment of risks in a gender-neutral fashion — that is, all other disabilities are insured irrespective of gender — the Court’s analysis proves to be simplistic and misleading. For although all mutually contractible risks are covered irrespective of gender, but see n. 4 supra, the plan also insures risks such as prostatectomies, vasectomies, and circumcisions that are specific to the reproductive system of men and for which there exist no female counterparts covered by the plan. Again, pregnancy affords the only disability, sex-specific or otherwise, that is excluded from coverage.5 Accordingly, the District Court appropriately re*153marked: “[T]he concern of defendants in reference to pregnancy risks, coupled with the apparent lack of concern regarding the balancing of other statistically sex-linked disabilities, buttresses the Court’s conclusion that the discriminatory attitude characterized elsewhere in the Court’s findings was in fact a motivating factor in its policy.” 375 F. Supp., at 383.
If decision of this case, therefore, turns upon acceptance of the Court’s view of General Electric’s disability plan as a sex-neutral assignment of risks, or plaintiffs’ perception of the plan as a sex-conscious process expressive of the secondary status of women in the company’s labor force, the history of General Electric’s employment practices and the absence of definable gender-neutral sorting criteria under the plan warrant rejection of the Court’s view in deference to the plaintiffs’. Indeed, the fact that the Court’s frame of reference lends itself to such intentional, sex-laden decisionmaking makes clear the wisdom and propriety of the EEOC’s contrary approach to employment disability programs.
Ill
Of course, the demonstration of purposeful discrimination is not the only ground for recovery under Title VII. Notwithstanding unexplained and inexplicable implications to the contrary in the majority opinion,6 this Court, see Washing*154ton v. Davis, 426 U. S. 229, 238-239 (1976); Albemarle Paper Co. v. Moody, 422 U. S. 405, 422 (1975); McDonnell Douglas Corp. v. Green, 411 U. S. 792, 802 (1973); Griggs v. Duke Power Co., 401 U. S. 424, 432 (1971), and every Court of Appeals7 now have firmly settled that a *155prima facie violation of Title VII, whether under § 703 (a)(1) or §703 (a)(2), also is established by demonstrating that a facially neutral classification has the effect of discriminating against members of a defined class.
General Electric’s disability program has three divisible sets of effects. First, the plan covers all disabilities that mutually afflict both sexes. But see n. 4, supra. Second, the plan insures against all disabilities that are male-specific or have a predominant impact on males. Finally, all female-specific and female-impacted disabilities are covered, except for the most prevalent, pregnancy. The Court focuses on the first factor — the equal inclusion of mutual risks — and therefore understandably can identify no discriminatory effect arising from the plan. In contrast, the EEOC and plaintiffs rely upon the unequal exclusion manifested in effects two and three to pinpoint an adverse impact on women. However one defines the profile of risks protected by General Electric, the determinative question must be whether the social policies and aims to be furthered by Title VII and filtered through the phrase “to discriminate” contained in § 703 (a)(1) fairly forbid an ultimate pattern of coverage that insures all risks except a commonplace one that is applicable to women but not to men.
As a matter of law and policy, this is a paradigm example of the type of complex economic and social inquiry that Congress wisely left to resolution by the EEOC pursuant to its Title VII mandate. See H. R. Rep. No. 92-238, p. 8 (1972). And, accordingly, prior Title VII decisions have consistently acknowledged the unique persuasiveness of EEOC *156interpretations in this area. These prior decisions, rather than providing merely that Commission guidelines are “entitled to consideration,” as the Court allows, ante, at 141, hold that the EEOC’s interpretations should receive “great deference.” Albemarle Paper Co. v. Moody, supra, at 431; Griggs v. Duke Power Co., supra, at 433-434; Phillips v. Martin Marietta Corp., 400 U. S. 542, 545 (1971) (Marshall, J., concurring). Nonetheless, the Court today abandons this standard in order squarely to repudiate the 1972 Commission guideline providing that “[disabilities caused or contributed to by pregnancy . . . are, for all job-related purposes, temporary disabilities . . . [under] any health or temporary disability insurance or sick leave plan____” 29 CFR § 1604.10 (b) (1975). This rejection is attributed to two interrelated events: an 8-year delay between Title VII’s enactment and the promulgation of the Commission’s guideline, and. interim letters by the EEOC’s General Counsel expressing the view that pregnancy is not necessarily includable as a compensable disability. Neither event supports the Court’s refusal to accord “great deference” to the EEOC’s interpretation.
It is true, as noted, ante, at 143, that only brief mention of sex discrimination appears in the early legislative history of Title VII. It should not be surprising, therefore, that the EEOC, charged with a fresh and uncharted mandate, candidly acknowledged that further study was required before the contours of sex discrimination as proscribed by Congress could be defined. See 30 Fed. Reg. 14927 (1965). Although proceeding cautiously, the Commission from the outset acknowledged the relationship between sex discrimination and pregnancy, announcing that “policies would have to be devised which afforded female employees reasonable job protection during periods of pregnancy.” EEOC First Annual Report to Congress, Fiscal Year 1965-1966, p. 40 (1967). During the succeeding seven years, the EEOC worked to develop a coherent policy toward pregnancy-oriented employment practices *157both through the pursuit of its normal adjudicatory functions8 and by engaging in comprehensive studies with such organizations as the President’s Citizens’ Advisory Council on the Status of Women. See, e. g., Address of Jacqueline G. Gutwillig, Chairwoman, Citizens’ Advisory Council, cited in App. 1159. These investigations on the role of pregnancy in the labor market coupled with the Commission’s “review . . . [of] its case decisions on maternity preparatory to issuing formal guidelines,” id., at 1161, culminated in the 1972 guideline, the agency’s first formalized, systematic statement on “employment policies relating to pregnancy and childbirth.”
Therefore, while some eight years had elapsed prior to the issuance of the 1972 guideline, and earlier opinion letters had refused to impose liability on employers during this period of deliberation, no one can or does deny that the final EEOC determination followed thorough and well-informed consideration. Indeed, realistically viewed, this extended evaluation of an admittedly complex problem and an unwillingness to impose additional, potentially premature costs on employers during the decisionmaking stages ought to be perceived as a practice to be commended. It is bitter irony that the care that preceded promulgation of the 1972 guideline is today condemned by the Court as tardy indecisiveness, its unwillingness irresponsibly to challenge employers’ practices during the formative period is labeled as evidence of inconsistency, and this indecisiveness and inconsistency are bootstrapped into reasons for denying the Commission’s interpretation its due deference.
For me, the 1972 guideline represents a particularly conscientious and reasonable product of EEOC deliberations and, therefore, merits our “great deference.” Certainly, I can find *158no basis for concluding that the guideline is out of step with congressional intent. See Espinoza v. Farah Mfg. Co., 414 U. S. 86, 94 (1973). On the contrary, prior to 1972, Congress enacted just such a pregnancy-inclusive rule to govern the distribution of benefits for “sickness” under the Railroad Unemployment Insurance Act, 45 U. S. C. § 351 (k) (2). Furthermore, shortly following the announcement of the EEOC’s rule, Congress approved and the President signed an essentially identical promulgation by the Department of Health, Education, and Welfare under Title IX of the Education Amendments of 1972, 20 U. S. C. § 1681 (a) (1970 ed., Supp. V). See 45 CFR § 86.57 (c) (1976). Moreover, federal workers subject to the jurisdiction of the Civil Service Commission now are eligible for maternity and pregnancy coverage under their sick leave program. See Federal Personnel Manual, ch. 630, subch. 13, S13-2 (FPM Supp. 990-2, May 6, 1975).
These policy formulations are reasonable responses to the uniform testimony of governmental investigations which show that pregnancy exclusions built into disability programs both financially burden women workers and act to break down the continuity of the employment relationship, thereby exacerbating women’s comparatively transient role in the labor force. See, e. g., U. S. Dept. of Commerce, Consumer Income (Series P-60, No. 93, July 1974); Women’s Bureau, U. S. Dept. of Labor, Underutilization of Women Workers (rev. ed. 1971). In dictating pregnancy coverage under Title VII, the EEOC’s guideline merely settled upon a solution now accepted by every other Western industrial country. Dept, of Health, Education, and Welfare, Social Security Programs Throughout the World, 1971, pp. ix, xviii, xix (Research Report No. 40). I find it difficult to comprehend that such a construction can be anything but a “sufficiently reasonable” one to be “accepted by the reviewing courts.” Train v. Natural Resources Def. Council, 421 U. S. 60, 75 (1975).
*159The Court’s belief that the concept of discrimination cannot reach disability policies effecting “an additional risk, unique to women . . . ante, at 139, is plainly out of step with the decision three Terms ago in Lau v. Nichols, 414 U. S. 563 (1974), interpreting another provision of the Civil Rights Act. There a unanimous Court recognized that discrimination is a social phenomenon encased in a social context and, therefore, unavoidably takes its meaning from the desired end products of the relevant legislative enactment, end products that may demand due consideration to the uniqueness of “disadvantaged” individuals.9 A realistic understanding of conditions found in today’s labor environment warrants taking pregnancy into account in fashioning disability policies. Unlike the hypothetical situations conjectured by the Court, ante, at 139-140, and n. 17, contemporary disability *160programs are not creatures of a social or cultural vacuum devoid of stereotypes and signals concerning the pregnant woman employee. Indeed, no one seriously contends that General Electric or other companies actually conceptualized or developed their comprehensive insurance programs disability-by-disability in a strictly sex-neutral fashion.10 Instead, the company has devised a policy that, but for pregnancy, offers protection for all risks, even those that are “unique to” men or heavily male dominated. In light of this social expérience, the • history of General Electric's employment practices, the otherwise all-inclusive design of its disability program, and the burdened role of the contemporary working woman, the EEOC's construction of sex discrimination under § 703 (a)(1) is fully consonant with (the ultimate objective of Title VII, “to assure equality of employment opportunities and to eliminate those discriminatory practices and devices which have fostered [sexually] stratified job environments to the disadvantage of [women].” McDonnell Douglas Corp. v. Green, 411 U. S., at 800.
I would affirm the judgment of the Court of Appeals.
General Electric’s disability program was developed in an earlier era when women openly were presumed to play only a minor and temporary *150role in the labor force. As originally conceived in 1926, General Electric offered no benefit plan to its female employees because “ ‘women did not recognize the responsibilities of life, for they probably were hoping to get married soon and leave the company.’ ” App. 958, excerpted from D. Loth, Swope of G. E.: Story of Gerard Swope and General Electric in American Business (1958). It was not until the 1930’s and 1940’s that the company made female employees eligible to participate in the disability program. In common with general business practice, however, General Electric continued to pursue a policy of taking pregnancy and other factors into account in order to scale women’s wages at two-thirds the level of men’s. Id., at 1002. More recent company policies reflect common stereotypes concerning the potentialities of pregnant women, see, e. g., Cleveland Board of Education v. LaFleur, 414 U. S. 632, 644 (1974), and have coupled forced maternity leave with the nonpayment of disability payments. Thus, the District Court found: “In certain instances it appears that the pregnant employee was required to take leave of her position three months prior to birth and not permitted to return until six weeks after the birth. In other instances the periods varied .... In short, of all the employees it is only pregnant women who have been required to cease work regardless of their desire and physical ability to work and only they have been required to remain off their job for an arbitrary period after the birth of their child.” 375 F. Supp. 367, 385. In February 1973, approximately coinciding with commencement of this suit, the company abandoned its forced-maternity-leave policy by formal directive.
The Court of Appeals did not affirm on the basis of this finding, since it concluded that “the statute looks to ‘consequences,’ not intent,” and “[a]ny discrimination, such as that here, which is ‘inextricably sex-linked’ in consequences and result is violative of the Act.” 519 F. 2d 661, 664.
Of course, even the proposition that pregnancy is a voluntary condition is overbroad, for the District Court found that “a substantial incidence of negligent or accidental conception also occurs.” 375 F. Supp., at 377. I may assume, however, for purposes of this argument, that the high incidence of voluntary pregnancies and the inability to differentiate between voluntary and involuntary conceptions, except perhaps through obnoxious, intrusive means, could justify the decision-maker’s treating pregnancies as voluntarily induced.
The experience of one of the class plaintiffs is instructive of the reach of the pregnancy exclusion. On April 5, 1972, she took a pregnancy leave, delivering a stillborn baby some nine days later. Upon her return *152home, she suffered a blood clot in the lung, a condition unrelated to her pregnancy, and was rehospitalized. The company declined her claim for disability payments on the ground that pregnancy severed her eligibility under the plan. See id., at 372. Had she been separated from work for any other reason — for example, during a work stoppage — the plan would have fully covered the embolism.
Indeed, the shallowness of the Court’s “underinclusive” analysis is transparent. Had General Electric assembled a catalogue of all ailments that befall humanity, and then systematically proceeded to exclude from coverage every disability that is female-specific or predominantly afflicts women, the Court could still reason as here that the plan operates equally: Women, like men, would be entitled to draw disability payments for their circumcisions and prostatectomies, and neither sex could claim payment for pregnancies, breast cancer, and the other excluded female-dominated disabilities. Along similar lines, any disability that occurs disproportionately in a particular group — sickle-cell anemia, for example — could be freely excluded from the plan without troubling the Court’s analytical approach.
The cryptic “but cf.” citation to McDonnell Douglas Corp. v. Green, 411 U. S. 792 (1973), ante, at 137, is perhaps the most mystifying. McDonnell involved a private nonclass action under § 703 (a) (1) of Title VII in which the plaintiff explicitly complained that he was discharged from employment for racial, rather than licit, motives. 411 U. S., at 796. In such a case, where questions of motivation openly form the thrust of an individual plaintiff's complaint, the “effects” that company policies may have had on an entire class of persons understandably are only tangentially placed in issue, see id., at 805 n. 19. Even so, the Court expressly held that a prima facie violation of Title VII could be proved without affirmatively demonstrating that purposeful discrimination had occurred. Instead, the Court concluded that such an illicit purpose *154is inferable from the interplay of four factors which together reveal that the employers’ policies have worked to disadvantage the complainant vis-á-vis other prospective employees. See id., at 802. Only if the employer then satisfies the burden of articulating "some legitimate, nondiseriminatory reason for the employee’s rejection,” ibid., must the latter actually seek to establish an intent to discriminate. Id., at 804. Even at this juncture, however, McDonnell makes clear that statistical evidence of the racial composition of the labor force — that is, a statistical showing of adverse impact on the protected group of which the individual plaintiff is part — will be persuasive evidence that the failure to rehire the particular complainant “conformed to a general pattern of discrimination against” his group. Id., at 805. Thus, McDonnell went far in allowing proof of “effect,” even in the setting of an individualized rather than group claim of discrimination.
Equally unacceptable is the implication in the penultimate paragraph of the opinion, ante, at 145, that the Fourteenth Amendment standard of discrimination is coterminous with that applicable to Title VII. Not only is this fleeting dictum irrelevant to the reasoning that precedes it, not only does it conflict with a long line of cases to the contrary, infra, at_15JL and this page, but it is flatly contradicted by the central holding of last Term’s Washington v. Davis, 426 U. S. 229, 239 (1976): “We have never held that the constitutional standard for adjudicating claims of invidious racial discrimination is identical to the standards applicable under Title VII, and we decline to do so today.”
See Boston Chapter, NAACP v. Beecher, 504 F. 2d 1017, 1020 (CA1 1974); United States v. Wood, Wire & Metal Lathers, Local Union 46, 471 F. 2d 408, 414 n. 11 (CA2 1973); Pennsylvania v. O’Neill, 473 F. 2d 1029 (CA3 1973) (en banc); United States v. Chesapeake & Ohio R. Co., 471 F. 2d 582, 586 (CA4 1972); United States v. Hayes Int’l Corp., 456 F. 2d 112, 120 (CA5 1972); United States v. Masonry Contractors Assn. of Memphis, Inc., 497 F. 2d 871, 875 (CA6 1974); United States v. Carpenters, 457 F. 2d 210, 214 (CA7 1972); United States v. N. L. Industries, Inc., 479 F. 2d 354, 368 (CA8 1973); United States v. Ironworkers Local 86, 443 F. 2d 544, 550-551 (CA9 1971); Muller v. United States Steel Corp., 509 F. 2d 923, 927 (CA10 1975); Davis v. Washington, 168 U. S. *155App. D. C. 42, 46, 512 F. 2d 956, 960 (1975), rev’d on constitutional grounds, 426 U. S. 229 (1976).
Indeed, following Griggs, Congress in 1972 revised Title VII, and expressly endorsed use of the “effect only” test outlined therein in identifying “increasingly complex” “forms and incidents of discrimination” that “may not appear obvious at first glance.” See H. R. Rep. No. 92-238, p. 8 (1972).
For synopses of the Commission’s positions regarding pregnancy and sex discrimination adopted in the course of administrative decisionmaking and litigation activities, see the EEOC’s Annual Reports to Congress.
Lau held that the failure to provide special language instruction to Chinese-speaking students in San Francisco schools violated the ban against racial or national origin discrimination contained in § 601 of the Civil Rights Act of 1964. The Court concluded that the Act, as interpreted by the administrative regulations promulgated by the Department of Health, Education, and Welfare addressed “effects] [to discriminate] even though no purposeful design is present,” and ultimately sought to further the broad goal of insuring “a meaningful opportunity to participate in the [schools’] educational program . . . .” 414 U. S., at 568. Faced with such a generalized objective, the Court repudiated the analysis of the Court of Appeals which had relied upon San Francisco’s commitment of equal educational offerings and resources to every child as the basis for concluding that Chinese students have suffered no discrimination due to the failure to adjust the school program to remedy their unique language deficiencies. Instead, the Court agreed that the anti-discrimination language fairly can be read “to require affirmative remedial efforts to give special attention to linguistically deprived children.” Id., at 571 (Stewart, J., concurring). Similarly, given the broad social objectives that underlie Title VII, see infra, at 160, and General Electric’s apparent unhesitancy to take into account the unique physical characteristics of their male workers in defining the breadth of disability coverage, see supra, at 152, ample support appears for upholding the EEOC's view that pregnancy must be treated accordingly.
See, e. g., n. 1, supra.