concurring in the judgment.
In this case the Court of Appeals held that New York’s civil service veterans’ preference violated both equal protection and the right to travel, relying on Zobel v. Williams, 457 U. S. 55 (1982). Shortly after the Court of Appeals’ decision was issued, we struck down New Mexico’s property tax veterans’ preference in Hooper v. Bernalillo County Assessor, 472 U. S. 612 (1985). Both Zobel and Hooper held that the classifications used by the States to award preferences to certain citizens failed to pass a rational-basis test under the Equal Protection Clause. As a result, we had no occasion to reach the issues whether the classifications would survive heightened scrutiny or whether the right to travel was violated. See Hooper, supra, at 618, and n. 6; Zobel, supra, at 60-61, and n. 6.
The classification held invalid on equal protection grounds in Hooper was remarkably similar to the one at issue here; Hooper, therefore, would appear to be controlling. The plurality opinion, however, instead begins the analysis by addressing the “right to migrate.” Ante, at 901-905. Moreover, heightened scrutiny is employed without first determining whether the challenged New York classification would survive even rational-basis analysis. Ante, at 907-909. But as we observed in Zobel, supra, at 60, n. 6, and reiterated only last Term in Hooper, supra, at 618, n. 6, “[r]ight to travel cases have examined, in equal protection terms, state distinctions between newcomers and longer term residents.” This follows because “[i]n reality, right to *913travel analysis refers to little more than a particular application of equal protection analysis.” Zobel, supra, at 60, n. 6.
I believe the appropriate framework for reviewing New York’s preference scheme is the one dictated by Zobel and followed in Hooper — both very recent cases. Because “[t]his case involves a distinction between residents based on when they first established residence in the State,” just as in Hooper, “we [must] subject this case to equal protection analysis.” Hooper, 472 U. S., at 618, n. 6. The first question is whether the law survives rational-basis analysis under the Equal Protection Clause. “[I]f the statutory scheme cannot pass even the minimum rationality test, our inquiry ends.” Id., at 618. Under Hooper, it seems clear that New York’s provision is invalid on equal protection grounds.
The State proffers four justifications for the challenged classification. First, it claims that the preference system encourages New York residents to enlist during times of war. This justification rests entirely on the State’s characterization of the preference as being prospective and self-executing in nature. But plainly the preference is granted only retrospectively following definitive action by the legislature; legislative action is necessary to fix both the period when “war” is deemed to have commenced and when that war has ended.
In many cases a New York resident entering military service will have no idea whether he or she will be entitled to the preference following a successful tour of duty and honorable discharge. For example, the beginning of the Vietnam era was established by legislation as a “time of war” some three years after hostilities commenced. The legislature in 1973 later declared the “end” of that “war” (which was never declared a “war” by Congress) for purposes of the preference as March 29, 1973, but in 1983 it amended that finding to expand coverage retroactively again until May 7, 1975. See Brief for Appellant 9-10. The same happened in the Korean Con*914flict, and indeed, the provision at issue was adopted after the end of World War II with the clear intent to grant the benefit retroactively to veterans of that war. Id., at 7-8.
Moreover, the preference applies to all servicemen entering the military during the legislatively defined period, regardless of whether they enlisted voluntarily or were inducted as a result of the draft. Providing a “bonus” to draftees is hardly a boon intended or necessary to “encourage” them to enlist. Hence, this “encouragement to enlist” sharply “differs from the local ‘bounty’ laws enacted during the Civil War era, through which States paid residents cash bonuses for enlisting.” Hooper, 472 U. S., at 622, n. 12.
Second, the State argues that the preference provides partial compensation to residents for service during time of war. But our holding in Hooper clearly rejected any such retroactive rewards targeting only past residents. While we acknowledged that a “State may award certain benefits to all its bona fide veterans,” id., at 620, just as in that case “it is difficult to grasp how [New York] residents serving in the military suffered more than residents of other States who served, so that the latter would not deserve the benefits a State bestows for national military service.” Id., at 621.
Third, the State contends that it is permissible to encourage past-resident veterans to settle in New York after their military service ends. While such a preference might indeed encourage such veterans to return, it simultaneously has the effect of discouraging other veterans from settling in New York who are aware that civil service appointments will be hard to obtain. As we observed in Zobel and reiterated in Hooper, supra, at 619-620, “[t]he separation of residents into classes hardly seems a likely way to persuade new [residents] that the State welcomes them and wants them to stay.” Zobel, supra, at 62, n. 9. Moreover, Hooper made it clear that a “selective incentive” such as New York provides here “would encounter the same constitutional barrier faced *915by the [New Mexico] statute’s distinction between past and newly arrived residents.” Hooper, supra, at 619, n. 8.
Finally, the State asserts that the preference is targeted at a very special group of veterans who have both knowledge of local affairs and valuable skills learned in the military, and who therefore would make exceptional civil servants. But these “special attributes” are undeniably possessed by all veterans who are currently residents of New York.
Indeed, the irrationality of the New York scheme is highlighted by appellee Baez-Hernandez’ experiences. The current provision would grant a civil service hiring preference to a serviceman entering the military while a resident of New York even if he was a resident only for a day. But Baez-Hernandez, who was a resident of New York for over 10 years before applying for a civil service position — and who therefore has considerably more “local knowledge” than many returning veterans — can never receive the preference. Moreover, Baez-Hernandez was a resident of New York for two years when he was called from reserve status to active duty, where he remained until he was discharged as partially disabled. He therefore served “in time of war” after obtaining New York residency. Yet he still cannot qualify despite his being endowed with all the desired “special attributes.”
Just as in Hooper, “[stripped of its asserted justifications, the [New York] statute suffers from the same constitutional flaw as the Alaska statute in Zobel.” 472 U. S., at 622. Given our reasons for granting review, our admonition in Hooper need only be reiterated briefly to demonstrate the invalidity of the New York scheme under equal protection, rational-basis analysis:
“The State may not favor established residents over new residents based on the view that the State may take care of ‘its own,’ if such is defined by prior residence. Newcomers, by establishing bona fide residence in the State, become the State’s ‘own’ and may not be discriminated *916against solely on the basis of their arrival in the State after [a fixed date].” Id., at 623.
Appellees Soto-Lopez and Baez-Hernandez, by establishing bona fide residence in New York, have become the State’s “own” and must be treated accordingly with regard to any veterans’ preference. “Neither the Equal Protection Clause, nor this Court’s precedents, permit the State to prefer established resident veterans over newcomers in the retroactive apportionment of an economic benefit.” Ibid.
I would affirm the judgment of the Court of Appeals based on our reasoning and holdings in Hooper and Zobel, rather than adding dicta concerning the right to travel.