delivered the opinion of the Court.
This case presents an issue of the appropriateness of an extension of the judicially created exclusionary rule: Is evidence seized by a state criminal law enforcement officer in good faith, but nonetheless unconstitutionally, inadmissible in a civil proceeding by or against the United States?
I
In November 1968 the Los Angeles police obtained a warrant directing a search for bookmaking paraphernalia at two specified apartment locations in the city and, as well, on the respective persons of Morris Aaron Levine and respondent Max Janis. The warrant was issued by *435a judge of the Municipal Court of the Los Angeles Judicial District. It was based upon the affidavit of Officer Leonard Weissman.1 After the search, made pursuant *436to the warrant, both the respondent and Levine were arrested and the police seized from respondent property consisting of $4,940 in cash and certain wagering records.2
Soon thereafter, Officer Weissman telephoned an agent of the United States Internal Revenue Service and informed the agent that Janis had been arrested for bookmaking activity.3 With the assistance of Weissman, who was familiar with bookmakers’ codes, the revenue agent analyzed the wagering records that had been seized and determined from them the gross volume of respondent’s gambling activity for the five days immediately preceding the seizure. Weissman informed the agent that he had conducted a surveillance of respondent’s activities that indicated that respondent had been engaged in book*437making during the 77-day period from September 14 through November 30, 1968, the day of the arrest.
Respondent had not filed any federal wagering tax return pertaining to bookmaking activities for that 77-day period. Based exclusively upon its examination of the evidence so obtained by the Los Angeles police, the Internal Revenue Service made an assessment jointly against respondent and Levine for wagering taxes, under § 4401 of the Internal Revenue Code of 1954, 26 U. S. C. § 4401, in the amount of $89,026.09, plus interest. The amount of the assessment was computed by first determining respondent’s average daily gross proceeds for the five-day period covered by the seized material and analyzed by the agent, and then multiplying the resulting figure by 77, the period of the police surveillance of respondent’s activities.4 The assessment having been made, the Internal Revenue Service exercised its statutory authority, under 26 U. S. C. § 6331, to levy upon the $4,940 in cash in partial satisfaction of the assessment against respondent.
Charges were filed in due course against respondent and Levine in Los Angeles Municipal Court for violation of the local gambling laws. They moved to quash the search warrant. A suppression hearing was held by the same judge who had issued the warrant. The defendants pressed upon the court the case of Spinelli v. United States, 393 U. S. 410 (1969), which had been decided just three weeks earlier and after the search warrant had been issued. They urged that the Weissman affidavit did not set forth, in sufficient detail, the underlying circumstances to enable the issuing magistrate to determine in-*438dependency the reliability of the information supplied by the informants. The judge granted the motion to quash the warrant. He then ordered that all items seized pursuant to it be returned except the cash that had been levied upon by the Internal Revenue Service. App. 78-80.
In June 1969 respondent filed a claim for refund of the $4,940. The claim was not honored, and 18 months later, in December 1970, respondent filed suit for that amount in the United States District Court for the Central District of California. The Government answered and counterclaimed for the substantial unpaid balance of the assessment.5 In pretrial proceedings, it was agreed that the “sole basis of the computation of the civil tax assessment . . . was . . . the items obtained pursuant to the search warrant . . . and the information furnished to [the revenue agent] by Officer Weissman with respect to the duration of [respondent’s] alleged wagering activities.” 6 Id., at 18. Respondent then moved to suppress the evidence seized, and all copies thereof in the possession of the Service, and to quash the assessment. Id., at 23-24.
At the outset of the hearing on the motion, the District Court observed that it was “reluctantly holding that *439the affidavit supporting the search warrant is insufficient under the Spinelli and Aguilar [v. Texas, 378 U. S. 108 (1964)] doctrines.” Id., at 47. It then concluded that “[a] 11 of the evidence utilized as the basis” of the assessment “was obtained directly or indirectly as a result of the search pursuant to the defective search warrant,” and that, consequently, the assessment “was based in substantial part, if not completely, on illegally procured evidence ... in violation of [respondent's] Fourth Amendment rights to be free from unreasonable searches and seizures.” 73-1 USTC ¶ 16,083, p. 81,392 (1973). The court concluded that Janis was entitled to a refund of the $4,940, together with interest thereon, “for the reason that substantially all, if not all, of the evidence utilized by the defendants herein in making their assessment . . . was illegally obtained, and, as such, the assessment was invalid.” Ibid. Further, where, as here, “illegally obtained evidence constitutes the basis of a federal tax assessment,” the respondent was “not required to prove the extent of the refund to which he claims he is entitled.” Id., at 81,393. Instead, it was sufficient if he prove “that substantially all, if not all, of the evidence upon which the assessment was based was the result of illegally obtained evidence.” Accordingly, the court ordered that the civil tax assessment made by the Internal Revenue Service “against all the property and assets of . . . Janis be quashed,” and entered judgment for the respondent. Ibid. The Government's counterclaim was dismissed with prejudice. The United States Court of Appeals for the Ninth Circuit, by unpublished memorandum without opinion, affirmed on the basis of the District Court's findings of fact and conclusions of law. Pet. for Cert. 12A.
Because of the obvious importance of the question, we granted certiorari. 421 U. S. 1010 (1975).
*440II
Some initial observations about the procedural posture of the case in the District Court are indicated. If there is to be no limit to the burden of proof the respondent, as “taxpayer,” must carry, then, even though he were to obtain a favorable decision on the inadmissibility-of-evidence issue, the respondent on this record could not possibly defeat the Government’s counterclaim. The Government notes, properly we think, that the litigation is composed of two separate elements: the refund suit instituted by the respondent, and the collection suit instituted by the United States through its counterclaim. In a refund suit the taxpayer bears the burden of proving the amount he is entitled to recover. Lewis v. Reynolds, 284 U. S. 281 (1932). It is not enough for him to demonstrate that the assessment of the tax for which refund is sought was erroneous in some respects.
This Court has not spoken with respect to the burden of proof in a tax collection suit. The Government argues here that the presumption of correctness that attaches to the assessment in a refund suit must also apply in a civil collection suit instituted by the United States under the authority granted by §§ 7401 and 7403 of the Code, 26 U. S. C. §§ 7401 and 7403. Thus, it is said, the defendant in a collection suit has the same burden of proving that he paid the correct amount of his tax liability.
The policy behind the presumption of correctness and the burden of proof, see Bull v. United States, 295 U. S. 247, 259-260 (1935), would appear to be applicable in each situation. It accords, furthermore, with the burden-of-proof rule which prevails in the usual preassessment proceeding in the United States Tax Court. Lucas v. Structural Steel Co., 281 U. S. 264, 271 (1930); Welch v. Helvering, 290 U. S. 111, 115 (1933); Rule 142(a) *441of the Rules of Practice and Procedure of the United States Tax Court (1973). In any event, for purposes of this case, we assume that this is so and that the burden of proof may be said technically to rest with respondent Janis.
Respondent, however, submitted no evidence tending either to demonstrate that the assessment was incorrect or to show the correct amount of wagering tax liability, if any, on his part. In the usual situation one might well argue, as the Government does, that the District Court then could not properly grant judgment for the respondent on either aspect of the suit. But the present case may well not be the usual situation. What we have is a “naked” assessment without any foundation whatsoever if what was seized by the Los Angeles police cannot be used in the formulation of the assessment.7 The determination of tax due then may be one “without rational foundation and excessive,” and not properly subject to the usual rule with respect to the burden of proof in tax cases. Helvering v. Taylor, 293 U. S. 507, 514-515 (1935).8 See 9 J. Mertens, Law of Federal Income Taxation § 50.65 (1971).
There appears, indeed, to be some debate among the *442Federal Courts of Appeals, in different factual contexts, as to the effect upon the burden of proof in a tax case when there is positive evidence that an assessment is incorrect. Some courts indicate that the burden of showing the amount of the deficiency then shifts to the Commissioner.9 Others hold that the burden of showing the correct amount of the tax remains with the taxpayer.10 However that may be, the debate does not extend to the situation where the assessment is shown to be naked and without any foundation. The courts then appear to apply the rule of the Taylor case. See United States v. Rexach, 482 F. 2d 10, 16-17, n. 3 (CA1), cert. denied, 414 U. S. 1039 (1973); Pizzarello v. United States, 408 F. 2d 579 (CA2), cert. denied, 396 U. S. 986 (1969); Suarez v. Commisioner, 58 T. C. 792, 814-815 (1972). But cf. Compton v. United States, 334 F. 2d 212, 216 (CA4 1964).
Certainly, proof that an assessment is utterly without foundation is proof that it is arbitrary and erroneous. For purposes of this case, we need not go so far as to accept the Government’s argument that the exclusion of the evidence in issue here is insufficient to require judgment for the respondent or even to shift the burden to the Government. We are willing to assume that if the District Court was correct in ruling that the evidence seized by the Los Angeles police may not be used in formulating the assessment (on which both the levy and the counterclaim were based), then the District Court was also correct in granting judgment for Janis in both *443aspects of the present suit. This assumption takes us, then, to the primary issue.11
Ill
This Court early pronounced a rule that the Fifth Amendment’s command that no person “shall be compelled in any criminal case to be a witness against himself” renders evidence falling within the Amendment’s prohibition inadmissible. Boyd v. United States, 116 U. S. 616 (1886). It was not until 1914, however, that the Court held that the Fourth Amendment alone may be the basis for excluding from a federal criminal trial evidence seized by a federal officer in violation solely of that Amendment. Weeks v. United States, 232 U. S. 383. This comparatively late judicial creation of a Fourth Amendment exclusionary rule is not particularly surprising. In contrast to the Fifth Amendment’s direct command against the admission of compelled testimony, the issue of admissibility of evidence obtained in violation of the Fourth Amendment is determined after, and apart from, the violation.12 In *444Weeks it was held, however, that the Fourth Amendment did not apply to state officers, and, therefore, that material seized unconstitutionally by a state officer could be admitted in a federal criminal proceeding. This was the “silver platter” doctrine.13
In Wolf v. Colorado, 338 U. S. 25 (1949), the Court determined that the Due Process Clause of the Fourteenth Amendment reflected the Fourth Amendment to the extent of providing those protections against intrusions that are “ 'implicit in the concept of ordered liberty.’ ” Id., at 27. Nonetheless, the Court, in not applying the Weeks doctrine in a state trial to the product of a state search, held:
“Granting that in practice the exclusion of evidence may be an effective way of deterring unreasonable searches, it is not for this Court to condemn as falling below the minimal standards assured by the Due Process Clause a State’s reliance upon other methods which, if consistently enforced, would be equally effective.” 338 U. S., at 31.
Not long thereafter, the Court ruled that means used by a State to procure evidence could be sufficiently offensive to the concept of ordered liberty as to make admission of the evidence so procured a violation of the Due Process Clause, Rochin v. California, 342 U. S. 165 (1952), but that such a violation would exist only in the most extreme case, Irvine v. California, 347 U. S. 128 (1954).
*445Thus, as matters then stood, the Fourth Amendment was applicable to the States, but a State could allow an official to engage in a violation thereof with no judicial sanction except in the most extreme case. In addition, federal authorities, if they happened upon a State so inclined, could profit from the State’s action by receiving on a silver platter evidence unconstitutionally obtained. The federal authorities, profiting thereby, had no judicially created reason to discourage unconstitutional searches by a State, and the States, having no judicially mandated controls, were free to engage in such searches.14
Elkins v. United States, 364 U. S. 206, was decided in 1960. Invoking its “supervisory power over the administration of criminal justice in the federal courts,” id., at 216, the Court held that
“evidence obtained by state officers during a search which, if conducted by federal officers, would have violated the defendant’s immunity from unreasonable searches and seizures under the Fourth Amendment is inadmissible over the defendant’s timely objection in a federal criminal trial.” Id., at 223.
The rule thus announced apparently served two purposes. First, it assured that a State, which could admit the evidence in its own proceedings if it so chose, *446nevertheless would suffer some deterrence in that its federal counterparts would be unable to use the evidence in federal criminal proceedings. Second, the rule discouraged federal authorities from using a state official to circumvent the restrictions of Weeks.
Only one year later, however, the exclusionary rule was made applicable to state criminal trials. Mapp v. Ohio, 367 U. S. 643 (1961). The Court ruled:
“Since the Fourth Amendment’s right of privacy has been declared enforceable against the States through the Due Process Clause of the Fourteenth, it is enforceable against them by the same sanction of exclusion as is used against the Federal Government.” Id., at 655.
The debate within the Court on the exclusionary rule has always been a warm one.15 It has been unaided, unhappily, by any convincing empirical evidence on the effects of the rule. The Court, however, has established that the “prime purpose” of the rule, if not the sole one, “is to deter future unlawful police conduct.” United States v. Calandra, 414 U. S. 338, 347 (1974). See United States v. Peltier, 422 U. S. 531, 536-539 (1975). Thus,
“[i]n sum, the rule is a judicially created remedy designed to safeguard Fourth Amendment rights generally through its deterrent effect, rather than a personal constitutional right of the party aggrieved.” United States v. Calandra, 414 U. S., at 348.
*447And
“[a]s with any remedial device, the application of the rule has been restricted to those areas where its remedial objectives are thought most efficaciously served.” Ibid.16
In the complex and turbulent history of the rule, the Court never has applied it to exclude evidence from a civil proceeding, federal or state.17
IV
In the present case we are asked to create judicially a deterrent sanction by holding that evidence obtained by a state criminal law enforcement officer in good-faith reliance on a warrant that later proved to be defective shall be inadmissible in a federal civil tax proceeding. Clearly, the enforcement of admittedly valid laws would be hampered by so extending the exclusionary rule, and, as is nearly always the case with the rule, con-cededly relevant and reliable evidence would be rendered unavailable.18
*448In evaluating the need for a deterrent sanction, one must first identify those who are to be deterred. In this case it is the state officer who is the primary object of the sanction. It is his conduct that is to be controlled. Two factors suggest that a sanction in addition to those that presently exist is unnecessary. First, the local law enforcement official is already “punished” by the exclusion of the evidence in the state criminal trial.19 That, necessarily, is of substantial concern to him. Second, the evidence is also excludable in the federal criminal trial, Elkins v. United States, supra, so that the entire criminal enforcement process, which is the concern and duty of these officers, is frustrated.20
Jurists and scholars uniformly have recognized that the exclusionary rule imposes a substantial cost on the societal interest in law enforcement by its proscription *449of what concededly is relevant evidence. See, e. g., Bivens v. Six Unknown Fed. Narcotics Agents, 403 U. S. 388, 411 (1971) (Burger, C. J., dissenting); Amsterdam, Perspectives on the Fourth Amendment, 58 Minn. L. Rev. 349, 429 (1974). And alternatives that would be less costly to societal interests have been the subject of extensive discussion and exploration.21
Equally important, although scholars have attempted to determine whether the exclusionary rule in fact does have any deterrent effect, each empirical study on the *450subject, in its own way, appears to be flawed.22 It would not be appropriate to fault those who have attempted empirical studies for their lack of convincing data. The *451number of variables is substantial,23 and many cannot be measured or subjected to effective controls. Record-keeping before Mapp was spotty at best, a fact which *452thus severely hampers before-and-after studies. Since Mapp, of course, all possibility of broad-scale controlled or even semi-controlled comparison studies has been eliminated.24 “Response” studies are hampered by the *453presence of the respondents’ interests.25 And extrapolation studies are rendered highly inconclusive by the changes in legal doctrines and police-citizen relationships that have taken place in the 15 years since Mapp was decided.26
We find ourselves, therefore, in no better position than the Court was in 1960 when it said:
“Empirical statistics are not available to show that the inhabitants of states which follow the exclusionary rule suffer less from lawless searches and seizures than do those of states which admit evidence unlawfully obtained. Since as a practical matter it is never easy to prove a negative, it is hardly likely that conclusive factual data could ever be assembled. For much the same reason, it cannot positively be demonstrated that enforcement of the criminal law is either more or less effective under either rule.” Elkins v. United States, 364 U. S., at 218.
If the exclusionary rule is the “strong medicine” that its proponents claim it to be, then its use in the situations in which it is now applied (resulting, for example, in this case in frustration of the Los Angeles police officers’ good-faith duties as enforcers of the criminal laws) must be assumed to be a substantial and efficient deterrent. Assuming this efficacy, the additional marginal deterrence provided by forbidding a different sovereign from using the evidence in a civil proceeding surely does not out*454weigh the cost to society of extending the rule to that situation.27 If, on the other hand, the exclusionary rule does not result in appreciable deterrence, then, clearly, its use in the instant situation is unwarranted. Under either assumption, therefore, the extension of the rule is unjustified.28
In short, we conclude that exclusion from federal civil proceedings of evidence unlawfully seized by a state criminal enforcement officer has not been shown to have a sufficient likelihood of deterring the conduct of the state police so that it outweighs the societal costs imposed by the exclusion. This Court, therefore, is not justified in so extending the exclusionary rule.29
*455Respondent argues, however, that the application of the exclusionary rule to civil proceedings long has been recognized in the federal courts. He cites a number of cases.30 But respondent does not critically distinguish between those cases in which the officer committing the unconstitutional search or seizure was an agent of the sovereign that sought to use the evidence, on the one hand, and those cases, such as the present one, on the other hand, where the officer has no responsibility or duty to, or agreement with, the sovereign seeking to use the evidence.31
*456The seminal cases that apply the exclusionary rule to a civil proceeding involve “intrasovereign” violations,32 a situation we need not consider here. In some cases the courts have refused to create an exclusionary rule for either intersovereign or intrasovereign violations in proceedings other than strictly criminal prosecutions. See United States ex rel. Sperling v. Fitzpatrick, 426 F. 2d 1161 (CA2 1970) (intrasovereign/parole revocation); United States v. Schipani, 435 F. 2d 26 (CA2 1970), cert. denied, 401 U. S. 983 (1971) (intersovereign/sentenc-ing).33 And in Compton v. United States, 334 F. 2d 212, 215-216 (1964), a case remarkably like this one, the Fourth Circuit held that the presumption of correctness given a tax assessment was not affected by the fact that the assessment was based upon evidence unconstitutionally seized by state criminal law enforcement officers. Only one case cited by the respondent squarely holds that there must be an exclusionary rule barring use in a civil proceeding by one sovereign of material seized in violation of the Fourth Amendment by an officer of another sovereign.34 In Suarez v. Commissioner, 58 T. C. 792 *457(1972) (reviewed by the court, with two judges dissenting), the Tax Court determined that the exclusionary rule should be applied in a situation similar to the one that confronts us here. The court concluded that
“any competing consideration based upon the need for effective enforcement of civil tax liabilities (compare Elkins v. United States . . .) must give way to the higher goal of protection of the individual and the necessity for preserving confidence in, rather than encouraging contempt for, the processes of Government.” Id., at 805.
No appeal was taken.
We disagree with the broad implications of this statement of the Tax Court for two reasons. To the extent that the court did not focus on the deterrent purpose of the exclusionary rule, the law has since been clarified. See United States v. Calandra, 414 U. S. 338 (1974); United States v. Peltier, 422 U. S. 531 (1975). Moreover, the court did not distinguish between intersover-eign and intrasovereign uses of unconstitutionally seized material. Working, as we must, with the absence of convincing empirical data, common sense dictates that *458the deterrent effect of the exclusion of relevant evidence is highly attenuated when the “punishment” imposed upon the offending criminal enforcement officer is the removal of that evidence from a civil suit by or against a different sovereign. In Elkins the Court indicated that the assumed interest of criminal law enforcement officers in the criminal proceedings of another sovereign counterbalanced this attenuation sufficiently to justify an exclusionary rule. Here, however, the attenuation is further augmented by the fact that the proceeding is one to enforce only the civil law of the other sovereign.
This attenuation, coupled with the existing deterrence effected by the denial of use of the evidence by either sovereign in the criminal trials with which the searching officer is concerned, creates a situation in which the imposition of the exclusionary rule sought in this case is unlikely to provide significant, much less substantial, additional deterrence. It falls outside the offending officer’s zone of primary interest. The extension of the exclusionary rule, in our view, would be an unjustifiably drastic action by the courts in the pursuit of what is an undesired and undesirable supervisory role over police officers.35 See Rizzo v. Goode, 423 U. S. 362 (1976).
*459In the past this Court has opted for exclusion in the anticipation that law enforcement officers would be deterred from violating Fourth Amendment rights. Then, as now, the Court acted in the absence of convincing empirical evidence and relied, instead, on its own assumptions of human nature and the interrelationship of the various components of the law enforcement system. In the situation before us, we do not find sufficient justification for the drastic measure of an exclusionary rule. There comes a point at which courts, consistent with their duty to administer the law, cannot continue to create barriers to law enforcement in the pursuit of a supervisory role that is properly the duty of the Executive and Legislative Branches. We find ourselves at that point in this case. We therefore hold that the judicially *460created exclusionary rule should not be extended to forbid the use in the civil proceeding of one sovereign of evidence seized by a criminal law enforcement agent of another sovereign.
The judgment of the Court of Appeals is reversed, and the case is remanded for further proceedings consistent with this opinion. R fi0 or<W-
Me. Justice Stevens took no part in the consideration or decision of this case.Officer Weissman’s affidavit, App. 69-74, stated: He and Sergeant Briggs of the Los Angeles Police Department each had received information from an informant concerning respondent Janis and Levine and concerning telephone numbers the two men used for bookmaking. Police investigation disclosed that Janis had two telephones with unpublished numbers, including the number given by Weissman’s informant, and that there was a third published number at the same address in the name of Nancy L. Janis. The unpublished numbers given by Weissman’s informant as being used by Levine were found to be maintained by Levine at a different address, and that address was the one given by Briggs’ informant as being Levine’s base of operations. Both informants stated that Levine and Janis were working in concert. Each officer regarded his informant as reliable; the informant had given information in the past that led to arrests for bookmaking and, in the case of Briggs’ informant, to convictions as well. Preliminary hearings and trials were pending for persons arrested with the aid of Weissman’s informant. Each officer and his informant believed that it was necessary for the informant’s safety, and his future usefulness to law enforcement officers, that his identity be kept secret.
Weissman further stated:
“Prom the nature and context of the information supplied by the informant to this affiant, and from the nature and context of the information which was supplied to Sgt. Briggs, as told to this affiant, it is believed that the informants ... at all times mentioned in this affidavit, unless otherwise specified, were speaking with personal knowledge.” Id., at 73.
The affidavit, taken in its entirety, bears some similarity to the affidavit the Court later considered in Spinelli v. United States, 393 U. S. 410, 420-422 (1969). Spinelli was a 5-3 decision handed down two months after the Los Angeles warrant in the present case had been issued. Mr. Justice White joined the opinion in Spinelli, id., at 423-429, but, in doing so, referred, id., at 427, to the “tension between Draper [v. United States, 358 U. S. 307 (1959)],” on the one hand, and Nathanson v. United States, 290 U. S. 41 (1933), and Aguilar v. Texas, 378 U. S. 108 (1964), on the other, and, “[p] ending full-scale reconsideration” of Draper “or of the *436Nathanson-Aguilar cases,” joined “the opinion of the Court and the judgment of reversal, especially since a vote to affirm would produce an equally divided Court.” 393 U. S., at 429.
The Internal Revenue Service’s Certificate of Assessments and Payments, App. 81, shows a credit of $5,097, the amount actually seized by the police and subjected to the Service’s subsequent levy. The Government acknowledges, however, that $157 of this amount was money belonging to Levine. It was applied upon the joint assessment made against both Janis and Levine. Levine has not sought a refund of the $157. Brief for United States 5 n. 1. The present case, therefore, concerns only the $4,940 taken from respondent Janis.
Officer Weissman testified that there was no departmental policy to call the Internal Revenue Service in a situation of this kind. He did it “as a matter of police procedure.” He would not do it, he said, on what he “would consider a small-size book, but I considered this one a major-size book. So, I, therefore, did it.” App. 42. He further stated that some of his fellow officers had acted similarly, but that he did not think “that they all have done it.” Ibid. The District Court did not rest its conclusion on any federal involvement in, or encouragement of, the search. We therefore must assume, for purposes of this opinion, that there was no federal involvement. See n. 31, infra.
The wagering excise tax at the time was 10% of the amount of the wagers. § 4401 (a) of the Internal Revenue Code of 1954, 26 U. S. C. §4401 (a). The rate was reduced to 2%, effective December 1, 1974, by Pub. L. 93-499, §3 (a), 88 Stat. 1550.
The Government advises us that, in order to avoid multiple litigation, its policy is to counterclaim in. a refund suit, just as it did here, where there is an outstanding unpaid assessment and the refund suit and the counterclaim involve the same facts. Brief for United States 17 n. 4.
The Certificate of Assessments and Payments was stipulated "to be admissible without objection.” App. 20. The Government did not seek to introduce the wagering records obtained by the Los Angeles police.
The Government has not asserted that, absent the seized materials, it would have had grounds for an assessment against respondent and Levine.
The situation may be described as having some resemblance to that for which the Court has developed an exception to the Anti-Injunction Act, § 7421 (a) of the Code, 26 U. S. C. § 7421 (a). See Enochs v. Williams Packing Co., 370 U. S. 1 (1962); Bob Jones University v. Simon, 416 U. S. 725 (1974); Commissioner v. “Americans United” Inc., 416 U. S. 752 (1974); Laing v. United States, 423 U. S. 161 (1976); Commissioner v. Shapiro, 424 U. S. 614 (1976).
Taylor, although decided more than 40 years ago, has never been cited by this Court on the burden-of-proof issue. The Courts of Appeals, the Court of Claims, the Tax Court, and the Federal District Courts, however, frequently have referred to that aspect of the case.
E. g., Foster v. Commissioner, 391 F. 2d 727, 735 (CA4 1968); Herbert v. Commissioner, 377 F. 2d 65, 69 (CA9 1967). See Bar L Ranch, Inc. v. Phinney, 426 F. 2d 995, 999 (CA5 1970).
E. g., United States v. Rexach, 482 F. 2d 10, 15-17 (CA1), cert. denied, 414 U. S. 1039 (1973); Psaty v. United States, 442 F. 2d 1154, 1158-1161 (CA3 1971); Ehlers v. Vinal, 382 F. 2d 58, 65-66 (CA8 1967). See Bar L Ranch, Inc. v. Phinney, 426 F. 2d, at 998.
Although the present case presents only the issue whether such evidence may be used in the formulation of the assessment, there appears to be no difference between that question and the issue whether the evidence is to be excluded in the refund or collection suit itself. We perceive no principled distinction to be made between the use of the evidence as the basis of an assessment and its use in the case in chief.
“[T]he raptured privacy of the victims’ homes and effects cannot be restored. Reparation comes too late.” Linkletter v. Walker, 381 U. S. 618, 637 (1965). “The rule is calculated to prevent, not to repair. Its purpose is to deter — to compel respect for the constitutional guaranty in the only effectively available way — by removing the incentive to disregard it.” Elkins v. United States, 364 U. S. 206, 217 (1960). See United States v. Calandra, 414 U. S. 338, 347-348 (1974); Mapp v. Ohio, 367 U. S. 643, 656 (1961); Tehan v. United States ex rel. Shott, 382 U. S. 406, 413 (1966); Terry v. Ohio, 392 U.S. 1, 29 (1968).
In Elkins v. United States, 364 U. S., at 207 n. 1, the Court noted that the appellation stems from Mr. Justice Frankfurter’s plurality opinion in Lustig v. United States, 338 U. S. 74 (1949):
“The crux of that doctrine is that a search is a search by a federal official if he had a hand in it; it is not a search by a federal official if evidence secured by state authorities is turned over to the federal authorities on a silver platter.” Id., at 78-79.
The absence of this Court’s imposition of controls did not mean, of course, that the States were running unchecked in their pursuit of evidence. Not only were there tort remedies and internal disciplinary sanctions available, but, as the Court noted in Elkins: “Not more than half the states continue totally to adhere to the rule that evidence is freely admissible no matter how it was obtained. Most of the others have adopted the exclusionary rule in its entirety; the rest have adopted it in part.” 364 U. S., at 219 (footnote omitted).
See also id., at 224r-225 (Appendix to opinion).
Except for the unanimous decision written by Mr. Justice Day in Weeks v. United States, 232 U. S. 383 (1914), the evolution of the exclusionary rule has been marked by sharp divisions in the Court. Indeed, Wolf, Lustig, Bochin, Irvine, Elkins, Mapp, and Calandra produced a combined total of 27 separate signed opinions or statements.
Thus, the Court has held that the exclusionary rule may be invoked only by those whose rights are infringed by the search itself, and not by those who are merely aggrieved by the introduction of evidence so obtained. Alderman v. United States, 394 U. S. 165, 174-175 (1969).
The Court has applied the exclusionary rule in a proceeding for forfeiture of an article used in violation of the criminal law. Plymouth Sedan v. Pennsylvania, 380 U. S. 693 (1965). There it expressly relied on the fact that “forfeiture is clearly a penalty for the criminal offense” and “[i]t would be anomalous indeed, under these circumstances, to hold that in the criminal proceeding the illegally seized evidence is excludable, while in the forfeiture proceeding, requiring the determination that the criminal law has been violated, the same evidence would be admissible.” Id., at 701. See also Boyd v. United States, 116 U. S. 616, 634 (1886), where a forfeiture proceeding was characterized as “quasi-criminal.”
There are studies and commentary to the effect that the exclu*448sionary rule tends to lessen the accuracy of the evidence presented in court because it encourages the police to lie in order to avoid suppression of evidence. See, e. g., Garbus, Police Perjury: An Interview, 8 Crim. L. Bull. 363 (1972); Kuh, The Mapp Case One Year After; An Appraisal of Its Impact in New York, 148 N. Y. L. J: Nos. 55 and 56 (1962); Comment, Police Perjury in Narcotics “Dropsy” Cases: A New Credibility Gap, 60 Geo. L. J. 507 (1971); Effect of Mapp v. Ohio on Police Search-and-Seizure Practices in Narcotics Cases, 4 Colum. J. L. & Soc. Probs. 87 (1968). See also People v. McMurty, 64 Misc. 2d 63, 314 N. Y. S. 2d 194 (N. Y. C. Crim. Ct. 1970).
It is of interest to note that the exclusion of this evidence from the California state trial was required by a decision of the State’s Supreme Court issued some years prior to Mapp. See People v. Cohan, 44 Cal. 2d 434, 282 P. 2d 905 (1955).
We are aware of the suggestion, made by some commentators and incorporated in some studies, that police often view trial and conviction as a lesser aspect of law enforcement. See, e. g., J. Skolnick, Justice Without Trial 219-235 (2d ed., 1975); Milner, Supreme Court Effectiveness and the Police Organization, 36 Law & Contemp. Probs. 467, 475, 479 (1971); Oaks, Studying the Exclusionary Rule in Search and Seizure, 37 U. Chi. L. Rev. 665, 720-736 (1970).
See, e. g., Bivens v. Six Unknown Fed. Narcotics Agents, 403 U. S. 388, 411 (1971) (BurgeR, C. J., dissenting); ALI Model Code of Pre-Arraignment Procedure § SS 290.2 (Proposed Official Draft 1975); Davidow, Criminal Procedure Ombudsman as a Substitute for the Exclusionary Rule: A Proposal, 4 Tex. Tech. L. Rev. 317 (1973); Davis, An Approach to Legal Control of the Police, 52 Texas L. Rev. 703 (1974); Foote, Tort Remedies for Police Violations of Individual Rights, 39 Minn. L. Rev. 493 (1955); Geller, Enforcing the Fourth Amendment: The Exclusionary Rule and Its Alternatives, 1975 Wash. U. L. Q. 621; Kaplan, The Limits of the Exclusionary Rule, 26 Stan. L. Rev. 1027 (1974); LaFave, Improving Police Performance Through the Exclusionary Rule — Part II: Defining the Norms and Training the Police, 30 Mo. L. Rev. 566 (1965); McGowan, Rule-Making and the Police, 70 Mich. L. Rev. 659 (1972); Quinn, The Effect of Police Rulemaking on the Scope of Fourth Amendment Rights, 52 J. Urb. L. 25 (1974); Roche, A Viable Substitute for the Exclusionary Rule: A Civil Rights Appeals Board, 30 Wash. & Lee L. Rev. 223 (1973); Spiotto, The Search and Seizure Problem — Two Approaches: The Canadian Tort Remedy and the U. S. Exclusionary Rule, 1 J. Police Sci. & Ad. 36 (1973); Spiotto, Search and Seizure: An Empirical Study of the Exclusionary Rule and Its Alternatives, 2 J. Leg. Stud. 243 (1973); Comment, Federal Injunctive Relief from Illegal Search, 1967 Wash. U. L. Q. 104; Comment, The Federal Injunction as a Remedy for Unconstitutional Police Conduct, 78 Yale L. J. 143 (1968); Comment, Use of § 1983 to Remedy Unconstitutional Police Conduct: Guarding the Guards, 5 Harv. Civ. Rights-Civ. Lib. L. Rev. 104 (1970).
The salient and most comprehensive study is that of Oaks, cited above in n. 20. Professor (now President) Oaks reviews at length the data in previous studies and the problems involved in drawing conclusions from those data. The previous studies include, inter alia, D. Oaks & W. Lehman, A Criminal Justice System and the Indigent: A Study of Chicago and Cook County (1968); J. Skolniek, Justice Without Trial (1st ed. 1966); Goldstein, Police Discretion not to Invoke the Criminal Process: Low-Visibility Decisions in the Administration of Justice, 69 Yale L. J. 543 (1960); Kamisar, On the Tactics of Police-Prosecution Oriented Critics of the Courts, 49 Cornell L. Q. 436 (1964); Kamisar, Public Safety v. Individual Liberties: Some “Facts” and “Theories,” 53 J. Crim. L. C. & P. S. 171 (1962); Kamisar, Wolf and Lustig Ten Years Later: Illegal State Evidence in State and Federal Courts, 43 Minn. L. Rev. 1083 (1959); Katz, The Supreme Court and the States: An Inquiry into Mapp v. Ohio in North Carolina. The Model, the Study and the Implications, 45 N. C. L. Rev. 119 (1966); Kuh, supra, n. 18; Nagel, Testing the Effects of Excluding Illegally Seized Evidence, 1965 Wis. L. Rev. 283; Paulsen, The Exclusionary Rule and Misconduct by the Police, 52 J. Crim. L. C. & P. S. 255 (1961); Comment, Search and Seizure in Illinois: Enforcement of the Constitutional Right of Privacy, 47 Nw. U. L. Rev. 493 (1952); Weinstein, Local Responsibility for Improvement of Search and Seizure Practices, 34 Rocky Mt. L. Rev. 150 (1962); Younger, Constitutional Protection oil Search and Seizure Dead?, 3 Trial 41 (Aug.-Sept. 1967); Comment, Effect of Mapp v. Ohio on Police Search-and-Seizure Practices in Narcotics Cases, 4 Colum. J. L. & Soc. Probs. 87 (1968).
Oaks discusses the types of research that may be possible, and the difficulties inherent in each. His final conclusion is straightforward:
“Writing just after the decision in Mapp v. Ohio, Francis A. Allen declared that up to that time, 'no effective quantitative measure of the rule’s deterrent efficacy has been devised or applied.’ [Allen, Federalism and the Fourth Amendment: A Requiem for Wolf, 1961 Sup. Ct. Rev. 1, 34.] That conclusion is not yet outdated. The foregoing findings represent the largest fund of information yet assembled on the effect of the exclusionary rule, but they obviously *451fall short of an empirical substantiation or refutation of the deterrent effect of the exclusionary rule.” Oaks, supra, n. 20, at 709.
More recently, Canon, Is the Exclusionary Rule in Failing Health? Some New Data and a Plea against a Precipitous Conclusion, 62 Ky. L. J. 681 (1974), discusses the data collected and reviewed by Oaks, and explores the difficulties in drawing conclusions from those data. The paper also reviews studies that appeared subsequent to the Oaks article: Spiotto, supra, n. 21, at 243; and two papers by Michael Ban, The Impact of Mapp v. Ohio on Police Behavior (delivered at the annual meeting of the Midwest Political Science Assn., Chicago, May 1973) and Local Courts v. The Supreme Court: The Impact of Mapp v. Ohio (delivered at the annual meeting of the American Political Science Assn., New Orleans, Sept. 1973). Canon describes his own research, but his data and conclusions appear to suffer from many of the same difficulties and faults ■present in the prior studies, many of which are explicitly recognized. Consequently, although Canon argues in favor of retaining the exclusionary rule while Oaks argues against it, Capon’s conclusions are no firmer than are Oaks’: “Consequently, our argument is negative rather than positive; we are maintaining that the evidence from the 14 cities certainly does not support a conclusion that the exclusionary rule had no impact upon arrests in search- and-seizure type crimes in the years following its imposition.” Canon, supra, at 707. “Consequently, we cannot confidently attribute the increased use of search warrants entirely or even primarily to police reaction to the exclusionary rule.” Id,., at 713. See also id., at 724-725 and at 725-726. Canon concedes that “the inconclusiveness of our findings is real enough,” id., at 726, but argues that the exclusionary rule should be given time to take effect. “Only after a substantial amount of time has passed do trends of changing behavior (if any) become apparent.” Id., at 727. One might wonder why, if the substantial amount of time necessary for the rule to take effect is extremely relevant, the study fails to take into account the fact that over half the States have had an exclusionary rule for a significantly greater length of time than Mapp has been on the books.
Most recently, Critique, On the Limitations of Empirical Evalu*452ations of the Exclusionary Rule: A Critique of the Spiotto Research and United States v. Calandra, 69 Nw. U. L. Rev. 740 (1974), reviews the Oaks, Canon, and Spiotto papers and the studies mentioned therein. The comment discusses the design difficulties present and involved in studying the deterrent effect of the exclusionary rule in general. Although a proponent of the rule, the author concludes:
“A review of Spiotto’s research and that conducted by others does not demonstrate the ineffectiveness of the exclusionary rule. Rather, it tends to illustrate the obstacles that stand in the way of any sound, empirical evaluation of the rule. When all factors are considered, there is virtually no likelihood that the Court is going to receive any ‘relevant statistics’ which objectively measure the 'practical efficacy’ of the exclusionary rule.” Id.., at 763-764.
The final conclusion is clear. No empirical researcher, proponent or opponent of the rule, has yet been able to establish with any assurance whether the rule has a deterrent effect even in the situations in which it is now applied. It is, of course, virtually impossible to study the marginal deterrence added to Mapp by the Elkins silver platter rule because of the difficulty of controlling the effect of intersovereign exclusion.
We are aware of no study on the possible deterrent effect of excluding evidence in a civil proceeding.
For discussion of the variables involved, see Canon, supra, n. 22; Geller, supra, n. 21; Kaplan, supra, n. 21; Milner, supra, n. 20; Oaks, supra, n. 20; Wright, Must the Criminal Go Free if the Constable Blunders?, 50 Texas L. Rev. 736 (1972); Critique, supra.
Studies have attempted to compare the experience in countries without the exclusionary rule with the experience in this country. See, e. g., Oaks, supra, n. 20, at 701-706; Spiotto, The Search and Seizure Problem — Two Approaches: The Canadian Tort Remedy and the U. S. Exclusionary Rule, 1 J. Police Sci. & Ad. 36 (1973). See generally The Exclusionary Rule Regarding Illegally Seized Evidence: An International Symposium, 52 J. Crim. L. C. & P. S. 245 (1961). The difficulties in drawing conclusions from cross-cultural comparisons are self-evident. See also Canon, supra, n. 22, at 692 n. 53.
See generally id., at 713-717, 723-725; Katz, supra, n. 22; Murphy, Judicial Review of Police Methods in Law Enforcement, 44 Texas L. Rev. 939, 941-943 (1966).
We do not mean to imply that more accurate studies could never be developed, or that what statisticians refer to as “triangulation” might not eventually provide us with firmer conclusions. We just do not find that the studies now available provide us with reliable conclusions.
If the exclusionary rule is not “strong medicine,” but does provide some marginal deterrence in the criminal situations in which it is now applied, that marginal deterrence is diluted by the attenuation existing when a different sovereign uses the material in a civil proceeding, and we must again find that the marginal utility of the creation of such a rule is outweighed by the costs it imposes on society.
“[W]e simply decline to extend the court-made exclusionary rule to cases in which its deterrent purpose would not be served.” Desist v. United States, 394 U. S. 244, 254 n. 24 (1969).
“As with any remedial device, the application of the rule has been restricted to those areas where its remedial objectives are thought most efficaciously served.” United States v. Calandra, 414 U. S., at 348.
“Where the official action was pursued in complete good faith, however, the deterrence rationale loses much of its force.” Michigan v. Tucker, 417 U. S. 433, 447 (1974). See United States v. Peltier, 422 U. S., at 537-538.
“ [I]t will not do to forget that the Weeks rule is a rule arrived at only on the nicest balance of competing considerations and in view of the necessity of finding some effective judicial sanction to preserve the Constitution’s search and seizure guarantees. The rule is unsupportable as reparation or compensatory dispensation to the injured criminal; its sole rational justification is the experience of its indispensability in ‘exert [ing] general legal pressures to secure obedience to the Fourth Amendment on the part of federal law-enforcing *455officers.’ As it serves this function, the rule is a needed, but grud[g]ingly taken, medicament; no more should be swallowed than is needed to combat the disease. Granted that so many criminals must go free as will deter the constables from blundering, pursuance of this policy of liberation beyond the confines of necessity inflicts gratuitous harm on the public interest as declared by Congress.” Amsterdam, Search, Seizure, and Section 2255: A Comment, 112 U. Pa. L. Rev. 378, 388-389 (1964) (footnotes omitted).
Suarez v. Commissioner, 58 T. C. 792 (1972); Pizzarello v. United States, 408 F. 2d 579 (CA2), cert. denied, 396 U. S. 986 (1969); Knoll Associates, Inc. v. FTC, 397 F. 2d 530 (CA7 1968); Powell v. Zuckert, 125 U. S. App. D. C. 55, 366 F. 2d 634 (1966); Rogers v. United States, 97 F. 2d 691 (CA1 1938); Anderson v. Richardson, 354 F. Supp. 363 (SD Fla. 1973); Iowa v. Union Asphalt & Roadoils, Inc., 281 F. Supp. 391 (SD Iowa 1968), aff’d sub nom. Standard Oil Co. v. Iowa, 408 F. 2d 1171 (CA8 1969); United States v. Stonehill, 274 F. Supp. 420 (SD Cal. 1967), aff’d, 405 F. 2d 738 (CA9 1968), cert. denied, 395 U. S. 960 (1969); United States v. Blank, 261 F. Supp. 180 (ND Ohio 1966); Lassoff v. Gray, 207 F. Supp. 843 (WD Ky. 1962).
The decision by the District Court to suppress the evidence did not rest upon any finding of such an agreement or participation, and from the record it does not appear that any “federal participation” existed. See Lustig v. United States, 338 U. S. 74 (1949); Byars v. United States, 273 U. S. 28 (1927). As stated above in n. 3, we decide the present case on the assumption that no such agreement or arrangement existed. Respondent remains free on remand to attempt to prove that there was federal participation in *456fact. If he succeeds in that proof, he raises the question, not presented by this case, whether the exclusionary rule is to be applied in a civil proceeding involving an intrasovereign violation.
It is well established, of course, that the exclusionary rule, as a deterrent sanction, is not applicable where a private party or a foreign government commits the offending act. See Burdeau v. McDowell, 256 U. S. 465 (1921); United States v. Stonehill, supra.
See Pizzarello v. United States, supra; Knoll Associates, Inc. v. FTC, supra; Powell v. Zuckert, supra; Iowa v. Union Asphalt & Roadoils, Inc., supra; United States v. Blank, supra. See also Hand v. United States, 441 F. 2d 529 (CA5 1971).
We express no view on the issue whether sentencing and parole revocation proceedings constitute “civil proceedings” for the purposes of the principles announced in this opinion.
In Anderson v. Richardson, 354 F. Supp. 363 (SD Fla. 1973), which otherwise might be in this category, the trial court relied on *457Pizzarello, supra, in enjoining a tax assessment based upon illegally seized evidence. The Government had conceded, however, that the jeopardy assessment upon which it relied could not ultimately succeed. 354 F. Supp., at 366. To the extent that dicta in that case might be relevant, the court failed to note that Pizzarello concerned an intra-sovereign situation.
Tn United States v. Chase, 67-1 USTC ¶ 15733 (DC 1966), the District Court relied entirely upon principles of judicial integrity in excluding from a tax proceeding evidence unconstitutionally seized by state agents. Id., at 84,477. As noted previously, the Court has since clarified the fact that the primary, if not the sole, function of the exclusionary rule is deterrence. See United States v. Calandra, supra; United States v. Peltier, supra. See also n. 35, infra.
To the extent that recent cases state that deterrence is the prime purpose of the exclusionary rule, and that “judicial integrity” is a relevant, albeit subordinate factor, we hold that in this case considerations of judicial integrity do not require exclusion of the evidence.
Judicial integrity clearly does not mean that the courts must never admit evidence obtained in violation of the Fourth Amendment. The requirement that a defendant must have standing to make a motion to suppress demonstrates as much. See Alderman v. United States, 394 U. S. 165 (1969).
The primary meaning of “judicial integrity” in the context- of evidentiary rules is that the courts must not commit or encourage *459violations of the Constitution. In the Fourth Amendment area, however, the evidence is unquestionably accurate, and the violation is complete by the time the evidence is presented to the court. See United States v. Calandra, 414 U. S., at 347, 354. The focus therefore must be on the question whether the admission of the evidence encourages violations of Fourth Amendment rights. As the Court has noted in recent cases, this inquiry is essentially the same as the inquiry into whether exclusion would serve a deterrent purpose. See United States v. Peltier, 422 U. S., at 538; Michigan v. Tucker, 417 U. S., at 450 n. 25. The analysis showing that exclusion in this case has no demonstrated deterrent effect and is unlikely to have any significant such effect shows, by the same reasoning, that the admission of the evidence is unlikely to encourage violations of the Fourth Amendment. The admission of evidence in a federal civil proceeding is simply not important enough to state criminal law enforcement officers to encourage them to violate Fourth Amendment rights (and thus to obtain evidence that they are unable to use in either state or federal criminal proceedings). In, addition, the officers here were clearly acting in good faith, see n. 1, supra, a factor that the Court has recognized reduces significantly the potential deterrent effect of exclusion. See Michigan v. Tucker, 417 U. S., at 447; United States v. Peltier, 422 U. S., at 539.