(concurring).
This is not an easy case to deal with under Washington v. Davis, 426 U.S. 229, 96 S.Ct. 2040, 48 L.Ed.2d 597 (1977). On the one hand, there can be no question about the unequal impact of this law: practically speaking, it permanently shuts off whole areas of state employment to women. On the other hand, as Judge Murray points out in his dissent, a strong initial case can be made for the proposition that it is “neutral on its face,” and not motivated in any ordinary sense by a discriminatory intent.* Arguably, therefore, the challenged statute is the kind of law which, notwithstanding its widespread impact on women’s employment opportunities, should be upheld as constitutional. The thrust of Washington v. Davis and related decisions such as Village of Arlington Heights v. Metropolitan Housing Corporation, 429 U.S. 252, 97 S.Ct. 555, 50 L.Ed.2d 450 (1977), is that we must accept that well-intentioned programs may have uneven side effects: society is too complicated for every discriminatory consequence to disqualify legitimate policies. Welfare programs, for example, foreseeably benefit minority groups disproportionately, just as tax deductions do whites. Examinations (as in Washington v. Davis) designed reasonably to weed out those unqualified for police work, may eliminate minority applicants more than others. Town and city planning laws, designed to improve community life, may because of separate economic factors, create barriers to minorities. Society would soon be in a state of paralysis if it could adopt only laws having strictly equal impact upon all groups and classes within it.
But while I fully recognize not only that Washington v. Davis is the law of the land but also that its principle reflects an essential limitation upon the sweep of the equal protection clause, I do not believe that the Massachusetts veterans preference law ac*151tually falls within its ambit. This, as Judge Tauro convincingly demonstrates, is no ordinary statute having merely an incidental unequal impact. It is a statute which goes a long way towards making upper level state employment a male preserve. Upon close inspection, the seeming “neutrality” of the veterans preference law, and even its seeming absence of intentional discrimination, are both open to serious question.
I turn first to the matter of its neutrality. While the dividing line between veterans and non-veterans is not the same as the dividing line between men and women, the ineluctable effect of this law is to confer an absolute priority upon a class that is 98% male in a sphere of employment where women, generally, should have the same access as men. What the laws does, is to take a group which has, for unique reasons, been selected almost exclusively from the male population (military service being what it was and is), and grant it an absolute preference in an entirely different sphere of public employment where male preference is not only not the rule but is constitutionally impermissible. The law may be “facially neutral” in the limited sense and it is not based overtly on selection by sex, but since the preferred class is 98% male the effect is virtually the same as if it were.
The discriminatory impact in Washington v. Davis was far less inevitable: the selection device at issue, a police examination, did not mandate the recruitment of a class made up, overwhelmingly, of whites. While past experience might have indicated that proportionately fewer blacks than whites would pass the neutral examination, this was not an inevitable outcome: a black who was determined to succeed might by dint of extra effort make up for past disadvantages; coaching and recruiting measures, as well as educational and economic improvements, might, over the years, increase the number of successful blacks. No such opportunity exists here for women. The veterans’ preference law prefers an already established class which, as a matter of historical fact, is 98% male. Because only persons who have served during wartime are eligible for the preference, the class cannot be expanded in the near future to include more women. Thus its “neutrality” is at best skin-deep. The law was sexually skewed from the outset, since the exclusionary effect upon women was not merely predictable but absolutely inescapable and “built-in”.
This same inevitability of exclusionary impact upon women also undermines the argument of no discriminatory intent. There is a difference between goals and intent. Conceding, as we all must, that the goal here was to benefit the veteran, there is no reason to absolve the legislature from awareness that the means chosen to achieve this goal would freeze women out of all those state jobs actively sought by men. To be sure, the legislature did not wish to harm women. But the cutting-off of women’s opportunities was an inevitable concomitant of the chosen scheme — as inevitable as the proposition that if tails is up, heads must be down. Where a law’s consequences are that inevitable, can they meaningfully be described as unintended? Doubtless the impact of women, if considered at all, was regarded as an acceptable “cost” of aiding veterans. But may society properly elect to aid veterans or any other group at the cost of abolishing equal employment opportunities in a major segment of public employment? In my view, the answer is “no”.
This is not to say that society may not bestow benefits upon veterans. But I think it may not construct a system of absolute preference which makes it virtually impossible for a woman, no matter how talented, to obtain a state job that is also of interest to males. Such a system is fundamentally different from the conferring upon veterans of financial benefits to which all taxpayers contribute, or from the giving to them of some degree of preference in government employment, as under a point system, as a quid pro quo for time lost in military service. The latter measures do not impose unfairly upon one segment of our society; the instant law, in contrast, forces women to pay a disproportionate *152share of the cost of benefiting veterans by sacrificing their own chance to' be selected for state employment.
Thus while it is concededly a close question whether the Massachusetts veterans’ preference is to be regarded as the sort of neutral classification with unintended effects absolved by Washington v. Davis, I feel on balance that it is not. Rather the law is more realistically viewed as substantively non-neutral. The destruction of normal female opportunities in the state employment system is too evident a consequence of the super-imposition of veterans as an absolutely preferred class upon that system. If this can be done constitutionally, the equal protection clause of the Constitution is, in this area of employment, little more than a hollow pretense, whatever it may remain in theory. As I think the unique problem posed in this case is distinguishable from any contemplated in Washington v. Davis, I adhere to our former judgment.
The statute can be called facially neutral in that it does not make a division based strictly on sex. The law provides employment preference for veterans, not males. While veterans are . 98% male, a few veterans are female, and there are many males who are not veterans.
The statute can likewise be said not to be based on a discriminatory intent, in the sense that no one thinks that it was enacted as a pretext to harm women. While the harm to female employment opportunities is extensive and, given the statutory scheme, inevitable, it was not this harm which prompted passage of the law, but rather the entirely justifiable desire to aid individuals who had served their country, often at great sacrifice.