concurring in the judgment.
In my opinion Justice Brennan’s criticism of the Court’s approach to this case merits a more thoughtful response than that contained in footnote 10, ante, at 176-177. Justice Brennan correctly points out that if the analysis of legislative purpose requires only a reading of the statutory language in a disputed provision, and if any “conceivable basis” for a discriminatory classification will repel a constitutional attack on the statute, judicial review will constitute a mere tautological recognition of the fact that Congress did what it intended to do. Justice Brennan is also correct in reminding us that even though the statute is an example of “social and economic legislation,” the challenge here is mounted by individuals whose legitimate expectations of receiving a fixed retirement income are being frustrated by, in effect, a breach of a solemn commitment by their Government. When Congress deprives a small class of persons of vested rights that are protected — and, indeed, even enhanced1 — for others who are in a similar though not identical position, I believe the Constitution requires something more than merely a “conceivable” or a “plausible” explanation for the unequal treatment.
1 do not, however, share Justice Brennan’s conclusion that every statutory classification must further an objective that can be confidently identified as the “actual purpose” of the legislature. Actual purpose is sometimes unknown. Moreover, undue emphasis on actual motivation may result in identically worded statutes being held valid in one State and invalid in a neighboring State.2 I therefore believe that we *181must discover a correlation between the classification and either the actual purpose of the statute or a legitimate purpose that we may reasonably presume to have motivated an impartial legislature. If the adverse impact on the disfavored class is an apparent aim of the legislature, its impartiality would be suspect. If, however, the adverse impact may reasonably be viewed as an acceptable cost of achieving a larger goal, an impartial lawmaker could rationally decide that that cost should be incurred.
In this case we need not look beyond the actual purpose of the legislature. As is often true, this legislation is the product of multiple and somewhat inconsistent purposes that led to certain compromises. One purpose was to eliminate in the future the benefit that is described by the Court as a “windfall benefit” and by Justice Brennan as an “earned dual benefit.” That aim was incident to the broader objective of protecting the solvency of the entire railroad retirement program. Two purposes that conflicted somewhat with this broad objective were the purposes of preserving those benefits that had already vested and of increasing the level of payments to beneficiaries whose rights were not otherwise to be changed. As Justice Brennan emphasizes, Congress originally intended to protect all vested benefits, but it ultimately sacrificed some benefits in the interest of achieving other objectives.
Given these conflicting purposes, I believe the decisive questions are (1) whether Congress can rationally reduce the vested benefits of some employees to improve the solvency of the entire program while simultaneously increasing the benefits of others; and (2) whether, in deciding which vested benefits to reduce, Congress may favor annuitants whose railroad service was more recent than that of disfavored annuitants who had an equal or greater quantum of employment.
*182My answer to both questions is in the affirmative. The congressional purpose to eliminate dual benefits is unquestionably legitimate; that legitimacy is not undermined by the adjustment in the level of remaining benefits in response to inflation in the economy. As for the second question, some hardship — in the form of frustrated long-term expectations— must inevitably result from any reduction in vested benefits. Arguably, therefore, Congress had a duty — and surely it had the right to decide — to eliminate no more vested benefits than necessary to achieve its fiscal purpose. Having made that decision, any distinction it chose within the class of vested beneficiaries would involve a difference of degree rather than a difference in entitlement. I am satisfied that a distinction based upon currency of railroad employment represents an impartial method of identifying that sort of difference. Because retirement plans frequently provide greater benefits for recent retirees than for those who retired years ago — and thus give a greater reward for recent service than for past service of equal duration — the basis for the statutory discrimination is supported by relevant precedent. It follows, in my judgment, that the timing of the employees’ railroad service is a “reasonable basis” for the classification as that term is used in Lindsley v. Natural Carbonic Gas Co., 220 U. S. 61, ante, at 174, and Dandridge v. Williams, 397 U. S. 471, ante, at 175, as well as a “ground of difference having a fair and substantial relation to the object of the legislation,” as those words are used in F. S. Royster Guano Co. v. Virginia, 253 U. S. 412, ante, at 174-175.
Accordingly, I concur in the judgment.
The 1974 Act provided increased benefits for spouses, widows, survivors, and early retirees. See 45 IT. S. C. § 231c (g).
Compare Rundlett v. Oliver, 607 F. 2d 495 (CA1 1979) (upholding Maine's statutory rape law), with Meloon v. Helgemoe, 564 F. 2d 602 *181(CA1 1977), cert. denied, 436 U. S. 950 (1978) (striking down New Hampshire’s statutory rape law).