Justice Breyer, with whom Justice Ginsburg joins, concurring in the judgment.

I agree with the majority and with Justice Souter that United States v. Halper, 490 U. S. 435 (1989), does not provide proper guidance for distinguishing between criminal and non-criminal sanctions and proceedings. I also agree that United States v. Ward, 448 U. S. 242, 248 (1980), and Kennedy v. Mendoza-Martinez, 372 U. S. 144, 168–169 (1963), set forth the proper approach.

I do not join the Court’s opinion, however, because I disagree with its reasoning in two respects. First, unlike the Court I would not say that “ ‘only the clearest proof ’ ” will “transform” into a criminal punishment what a legislature calls a “civil remedy.” Ante, at 6. I understand that the Court has taken this language from earlier cases. See Ward, supra, at 249. But the limitation that the language suggests is not consistent with what the Court has actually done. Rather, in fact if not in theory, the Court has simply applied factors of the Kennedy variety to the matter at hand. In Department of Revenue of Mont. v. Kurth Ranch, 511 U. S. 767 (1994), for example, the Court held that the collection of a state tax imposed on the possession and storage of drugs was “the functional equivalent of a

successive criminal prosecution” because, among other things, the tax was “remarkably high”; it had “an obvious deterrent purpose”; it was “conditioned on the commission of a crime”; it was “exacted only after the taxpayer ha[d] been arrested for the precise conduct that gives rise to the tax obligation”; its alternative function of raising revenue could be equally well served by increasing the fine imposed on the activity; and it departed radically from “normal revenue laws” by taxing contraband goods perhaps destroyed before the tax was imposed. Id., at 781–784. This reasoning tracks the non-exclusive list of factors set forth in Kennedy, and it is, I believe, the proper approach. The “clearest proof” language is consequently misleading, and I would consign it to the same legal limbo where Halper now rests.

Second, I would not decide now that a court should evaluate a statute only “ ‘on its face,’” ante, at 6 (quoting Kennedy, 372 U. S., at 169), rather than “assessing the character of the actual sanctions imposed.” Halper, 490 U. S., at 447; ante, at 7. Halper involved an ordinary civil-fine statute that as normally applied would not have created any “double jeopardy” problem. It was not the statute itself, but rather the disproportionate relation between fine and conduct as the statute was applied in the individual case that led this Court, unanimously, to find that the “civil penalty” was, in those circumstances, a second “punishment” that constituted double jeopardy. See 490 U. S., at 439, 452 (finding that $130,000 penalty was “sufficiently disproportionate” to $585 loss plus approximately $16,000 in government expenses caused by Halper’s fraud to constitute a second punishment in violation of double jeopardy). Of course, the Court in Halper might have reached the same result through application of the constitutional prohibition of “excessive fines.” See ante, at 9; Alexander v. United States, 509 U. S. 544, 558–559 (1993); Halper, 490 U. S, at 449 (emphasizing that Halper was

“the rare case” in which there was an “overwhelmingly disproportionate” fine). But that is not what the Court there said. And nothing in the majority’s opinion today explains why we should abandon this aspect of Halper’s holding. Indeed, in context, the language of Kennedy that suggests that the Court should consider the statute on its face does not suggest that there may not be further analysis of a penalty as it is applied in a particular case. See 372 U. S., at 169. Most of the lower court confusion and criticism of Halper appears to have focused on the problem of characterizing—by examining the face of the statute—the purposes of a civil penalty as punishment, not on the application of double jeopardy analysis to the penalties that are imposed in particular cases. It seems to me quite possible that a statute that provides for a punishment that normally is civil in nature could nonetheless amount to a criminal punishment as applied in special circumstances. And I would not now hold to the contrary.

That said, an analysis of the Kennedy factors still leads me to the conclusion that the statutory penalty in this case is not on its face a criminal penalty. Nor, in my view, does the application of the statute to the petitioners in this case amount to criminal punishment. I therefore concur in the result.