United States v. United States Coin & Currency

Mr. Justice Brennan,

concurring.

I join the opinion of the Court. The dissent would have us hold that the Government may continue indefinitely to enforce criminal penalties against individuals who had the temerity to engage in conduct protected by the Bill of Rights before the day that this Court held the conduct protected. Any such holding would have no more support in reason than it does in our cases.

*725 I

Frank recognition of the possible impact of retroactive application of constitutional decisions on the administration of criminal justice has led this Court to establish guidelines to determine the retroactivity of “constitutional rules of criminal procedure.” Stovall v. Denno, 388 U. S. 293, 296 (1967). Since “[e]ach constitutional rule of criminal procedure has its own distinct functions, its own background of precedent, and its own impact on the administration of justice,” the “retroactivity or non-retroactivity of a rule is not automatically determined by the provision of the Constitution on which the dictate is based.” Johnson v. New Jersey, 384 U. S. 719, 728 (1966). But although “[t]he extent to which a condemned practice infects the integrity of the truth-determining process at trial is a ‘question of probabilities,’ ” Stovall v. Denno, 388 U. S., at 298 , quoting Johnson v. New Jersey, 384 U. S., at 729 , as a general matter “[w]here the major purpose of new constitutional doctrine is to overcome an aspect of the criminal trial which substantially impairs its truth-finding function and so raises serious questions about the accuracy of guilty verdicts in past trials, the new rule has been given complete retroactive effect.” Williams v. United States, ante, at 653. “Neither good-faith reliance by state or federal authorities on prior constitutional law or accepted practice, nor severe impact on the administration of justice has sufficed to require prospective application in these circumstances.” Ibid *

*726 The reasoning that underlies these guidelines is clear. The States and the Federal Government have, of course, a legitimate interest in the evenhanded enforcement of such sanctions as they desire to impose upon any conduct that they may constitutionally prohibit. By definition a “new rule of criminal procedure” casts no doubt upon the power of government to punish certain conduct, but only upon the legitimacy of the process by which persons were found to have engaged in that conduct. Of course a government has no legitimate interest in upholding an unconstitutional system of criminal procedure. But accepting the results that an unconstitutional procedure has reached in the past does not uphold such a system for the future. Notwithstanding the new procedural rule the government retains a legitimate interest in sanctioning conduct that it may constitutionally prohibit. Accordingly, when a new procedural rule has cast no substantial doubt upon the reliability of determinations of guilt in criminal cases, we have denied the rule retroactive effect where a contrary decision would “impose a substantial burden [of retrials] upon the . . . judicial system . . . while serving neither to redress knowing violations of [constitutional rights] nor to protect a class of persons the government has no legitimate interest in punishing.” Williams v. United States, ante, at 664 (Brennan, J., concurring in result); see Desist v. United States, 394 U. S. 244 (1969). But since the government has no legitimate interest in punishing those innocent of wrongdoing, cf. Thompson v. Louisville, 362 U. S. 199 (1960), when a new procedural rule casts doubt upon the reliability of a substantial proportion of past convictions obtained without its protections, we have required the new rule be given full retroactive effect. Williams v. United States, ante, at 653. From this it follows a fortiori that a decision holding certain conduct beyond the power of government *727 to sanction or prohibit must be applied to prevent the continuing imposition of sanctions for conduct engaged in before the date of that decision. For the decision does far more than cast doubt upon the reliability of the guilt-determining process. It makes the question of reliability irrelevant, for it establishes beyond peradventure that the government has no legitimate interest in punishing such conduct at all. See Ex parte Siebold, 100 U. S. 371, 376-377 (1880). Accordingly, it may no longer continue to punish it.

II

The dissent seeks to explain its view of this case on the ground that even after this Court has declared certain individual conduct beyond the power of government to prohibit, the government retains an “interest in maintaining the rule of law and in demonstrating that those who defy the law do not do so with impunity” by punishing those persons who engaged in constitutionally protected conduct before it was so declared by this Court. Post, at 735. This argument, of course, has nothing whatever to do with the rule of law. It exalts merely the rule of judges by approving punishment of an individual for the lése-majesté of asserting a constitutional right before we said he had it. In light of our frequent reiteration that the usual mode of challenging an unconstitutional statute is expected to be violation of the statute and adjudication of the constitutional challenge in a criminal proceeding, see, e. g., Douglas v. City of Jeannette, 319 U. S. 157, 163 (1943); Dombrowski v. Pfister, 380 U. S. 479, 484-485 (1965), it is difficult to see how this argument amounts to more than a flat statement that those who assert their constitutional rights before we have declared them may not do so with impunity.

*728 If the dissent today means what it says, it would appear to follow that Virginia might keep in jail interracial married couples whose only offense was cohabitation within the State, so long as the cohabitation was prior to Loving v. Virginia, 388 U. S. 1 (1967); or that Arkansas could still discharge school teachers who taught evolution before we struck down the relevant statute in Epperson v. Arkansas, 393 U. S. 97 (1968). Of course the dissenters would never uphold such action. But if there is any distinction between these cases and the case at bar, it can only be that Angelini is asserting his privilege against self-incrimination, rather than a right under the First or Fourteenth Amendment. Whatever may be the relevance of the source of a new constitutional rule in determining the extent to which it affects the reliability of the factfinding process at trial, however, there is no justification for allowing the government greater power to vindicate its nonexistent interest in enforcing an unconstitutional statute that punishes assertion of the privilege against self-incrimination than to vindicate its interest in enforcing a statute that punishes the assertion of any other constitutional right.

APPENDIX TO OPINION OF BRENNAN, J., CONCURRING

Our cases show little deviation from the principle that new constitutional rules of criminal procedure that affect the integrity of the factfinding process will, in general, be retroactively applied. In Tehan v. Shott, 382 U. S. 406 (1966), we denied retroactive effect to Griffin v. California, 380 U. S. 609 (1965), despite our recognition that the privilege against self-incrimination which Griffin protected did in some circumstances serve as an adjunct to truth. 382 U. S., at 414-415, n. 12 , quoting Murphy v. Waterfront Comm’n, 378 U. S. 52, 55 (1964). But in Tehan we noted specifically that the privilege *729 against self-incrimination is not primarily "an adjunct to the ascertainment of truth,” 382 U. S., at 416 , and emphasized as well that retroactive application of Griffin would, in the States concerned, “have an impact upon the administration of their criminal law so devastating as to need no elaboration.” Id., at 419. Similarly, in Johnson v. New Jersey, 384 U. S. 719 (1966), we denied retroactive effect to Escobedo v. Illinois, 378 U. S. 478 (1964), and Miranda v. Arizona, 384 U. S. 436 (1966), notwithstanding our recognition that the principles announced in those cases would in some circumstances guard against the possibility of unreliable confessions. 384 U. S., at 730. But we emphasized in Johnson that strict -pre-Miranda standards were available to those desiring to test the admissibility of confessions, ibid., as well as pointing out the severe impact that retroactivity would have on state criminal processes. Id., at 731-732 . In Stovall v. Denno, 388 U. S. 293 (1967), we denied retroactive effect to United States v. Wade, 388 U. S. 218 (1967), and Gilbert v. California, 388 U. S. 263 (1967), because of uncertainty about the frequency with which violation of the rule there announced would actually result in injustice, the availability of a due process standard to remedy at least the more serious injustices, and the “unusual force of the countervailing considerations.” Stovall v. Denno, 388 U. S., at 299 . Finally, in DeStefano v. Woods, 392 U. S. 631 (1968), we denied retroactive effect to Duncan v. Louisiana, 391 U. S. 145 (1968), and Bloom v. Illinois, 391 U. S. 194 (1968), holding respectively that the States must afford criminal defendants a jury trial on demand in serious criminal cases, and that the right to jury trial extends to trials for serious criminal contempts. As to Duncan , retroactivity was denied because we considered that there was little likelihood that bench trials, as a whole, would be unfair, and because retroactive application could in some States *730 reopen every conviction for serious crime. 392 U. S., at 633-634 . As to Bloom , we recognized that one ground for the result was “the belief that contempt trials, which often occur before the very judge who was the object of the allegedly contemptuous behavior, would be more fairly tried if a jury determined guilt.” Id., at 634. But the firm tradition of nonjury trials in contempt cases, combined with the adverse impact of retro-activity on the administration of justice, combined to persuade us that Bloom should be applied prospectively only. Id., at 634-635. In addition, it should be noted that this Court has not been hesitant to reverse contempt convictions because of the possibility of involvement on the part of the judge. See Mayberry v. Pennsylvania, 400 U. S. 455 (1971), and cases cited.

Examination of these cases, therefore, indicates that in all cases save DeStejano /Bloom, we regarded as relatively small the likelihood that noncompliance with the new rule would have resulted in serious injustice in any past cases. Moreover, in all cases save Tehan and DeStejano /Duncan, alternative methods were still available to those who could demonstrate that the feared injustice had in fact resulted. Taken in combination, these factors lead me to conclude that the cases discussed in this Appendix do not undercut the force of the proposition at issue.

The few cases in which we have recognized that a new constitutional rule may in some circumstances improve the accuracy of the factfinding process, while at the same time denying retroactive application to that rule, do not in my view undercut the force of these statements. The relevant cases are collected and discussed in an Appendix to this opinion, infra, p. 728.