This appeal presents a recurring issue concerning application of the exclusionary rule barring use of evidence unlawfully seized in violation of the Fourth Amendment: whether the rule applies to bar use of such evidence in a proceeding different from the one for which the search was conducted. In Pizzarello v. United States, 408 F.2d 579 (2d Cir.), cert. denied, 396 U.S. 986, 90 S.Ct. 481, 24 L.Ed.2d 450 (1969), we decided that evidence unlawfully seized by agents of the Internal Revenue Service for use in criminal tax proceedings was barred from use in a subsequent IRS civil tax proceedings. In United States v. Janis, 428 U.S. 433, 96 S.Ct. 3021, 49 L.Ed.2d 1046 (1976), the Supreme Court decided that evidence unlawfully seized by local police officers investigating local wagering offenses was not barred from use in a subsequent federal civil tax proceeding. The Court left open the issue “whether the exclusionary rule is to be applied in a civil proceeding involving an intrasovereign violation.” Id. at 456 n. 31, 96 S.Ct. at 3033. In this appeal from a July 26, 1981 judgment of the United States Tax Court (Theodore Tannenwald, Judge), we must decide whether evidence allegedly seized unlawfully by federal narcotics agents for use in a narcotics prosecution is barred by the exclusionary rule in a subsequent federal civil tax proceeding. We thus encounter an alleged intrasovereign violation, which, unlike Pizzarello, has not been committed by officers of the agency bringing the civil proceeding. We conclude that the deterrence rationale of the exclusionary rule is not served by applying the rule to exclude evidence from a proceeding where the evidence was not seized with the participation or collusion of, or in contemplation of use by, agents responsible for the proceeding in which the evidence is presented. We therefore hold that the exclusionary rule is inapplicable to this case and affirm the judgment of the Tax Court.
I.
On August 3, 1972, pursuant to a search warrant, two agents of the Federal Bureau of Narcotics and three officers of the New York City Police Department entered the Manhattan apartment of appellant Jacques Tirado. The five investigators were members of the New York Drug Enforcement Task Force, a joint law enforcement unit drawn from the Federal Bureau of Narcotics, the New York state police, and the New York City police. Their warrant, issued by a justice of the New York Supreme Court, authorized a search of the apartment for “narcotics — -the means of committing a crime or offense, and the means of preventing a crime or offense from being discovered.”
In the apartment, in addition to cocaine and various drug adulterants and paraphernalia, the agents discovered and seized the following items relevant to this appeal: $38,865 in cash, rent receipts and notices for the apartment, a lease for the apartment in Tirado’s name, a notice to him regarding rental of a safe deposit box at a bank, several bank statements, a savings account passbook, parking tickets concerning two Rolls Royce automobiles, and several safe deposit box keys. Based largely on items seized during the search, the federal agents subsequently obtained search warrants for two safe deposit boxes registered to Tirado from which they seized $25,000 in cash and $10,000 worth of jewelry; they also learned the registration numbers of Tirado’s two Rolls Royces. Tirado was indicted and convicted in New York state court of possessing narcotics. His conviction was upheld by the Appellate Division and by the New York Court of Appeals. People v. Tirado, *30947 A.D.2d 193, 366 N.Y.S.2d 140 (1st Dep’t 1975), aff’d mem., 38 N.Y.2d 955, 348 N.E.2d 608, 384 N.Y.S.2d 151 (1976).
Shortly after the searches of Tirado’s apartment and safe deposit boxes, the federal agents involved in the narcotics case met with an agent of the Internal Revenue Service and disclosed to him the results of the searches. The IRS then used this information to reconstruct Tirado’s income for 1972 and to issue a Notice of Deficiency based on its calculations. Tirado petitioned the Tax Court to redetermine the deficiency on the ground that the deficiency was based wholly on information illegally obtained or derived from the search of his apartment. As subsequently refined, his claim was that the items seized were beyond the scope of the warrant.1 In an opinion filed on April 10, 1980, Judge Tannenwald ruled that the language of the warrant was sufficiently broad to authorize seizure of all the items Tirado sought to suppress; the ruling did not reach the Government’s alternative argument that the exclusionary rule was inapplicable to this case. Tirado v. Commis-
sioner, 74 T.C. 14 (1980). The Court subsequently sustained the deficiency determination and entered a judgment for the Commissioner, from which Tirado appeals. We agree with the Tax Court’s conclusion — that it was proper to base the deficiency determination on the seized information — but reach it by a different path, holding the exclusionary rule inapplicable in the circumstances of this civil tax proceeding. We therefore have no occasion to determine the legality of the various seizures involved.2
II.
The Supreme Court has made it plain that the principal, if not the only, justification for excluding illegally seized evidence from governmental proceedings is to deter future governmental misconduct. United States v. Janis, supra, 428 U.S. at 446, 96 S.Ct. at 3028; see Stone v. Powell, 428 U.S. 465, 486, 96 S.Ct. 3037, 3048, 49 L.Ed.2d 1067 (1976); United States v. Calandra, 414 U.S. 338, 347-48, 94 S.Ct. 613, 619-20, 38 L.Ed.2d 561 (1974).3 The exclusionary rule *310“is calculated to prevent, not to repair”; its purpose is “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, 80 S.Ct. 1437, 1444, 4 L.Ed.2d 1669 (1960). Current doctrine does not, however, take this principle to its furthest implication, which would require the exclusion of evidence whenever even a remote prospect of deterrence exists. Since use of the exclusionary rule impairs the search for truth even as it aids observance of constitutional limitations, standards for use of the rule must balance public needs against the claims of individual liberty.
The fact that evidence was seized in violation of the Fourth Amendment does not mean that it will be suppressed for every purpose in every proceeding. The Supreme Court has restricted application of the exclusionary rule to those circumstances where its deterrent effect would most likely be “substantial and efficient,” United States v. Janis, supra, 428 U.S. at 453, 96 S.Ct. at 3031, and has cautioned that any extension of the rule beyond its core application — normally, barring use of illegally seized items as affirmative evidence in the trial of the matter for which the search was conducted — must be justified by balancing the “additional marginal deterrence” of the extension against the cost to the public interest of further impairing the pursuit of truth. United States v. Janis, supra, 428 U.S. at 453-54, 96 S.Ct. at 3031-32; see Stone v. Powell, supra, 428 U.S. at 493-94, 96 S.Ct. at 3051-52; United States v. Calandra, supra, 414 U.S. at 351-52, 94 S.Ct. at 621-22.4 As the Court has said in the slightly different context of the standing required to assert the exclusionary rule, “[Ujnbending application of the exclusionary sanction to enforce ideals of governmental rectitude would impede unacceptably the truth-finding functions of judge and jury.” United States v. Payner, 447 U.S. 727, 734, 100 S.Ct. 2439, 2445, 65 L.Ed.2d 468 (1980).
Determining when the likelihood of substantial deterrence justifies excluding evidence requires some assessment of the motives of the officials who seized the challenged evidence. This inquiry into the officers’ motivation is the fundamental issue in translating the idea of deterrence into practical decisions, for deterrence means modifying individual behavior. “In evaluating the need for a deterrent sanction, one must first identify those who are to be deterred,” United States v. Janis, supra, 428 U.S. at 448, 96 S.Ct. at 3029, those whose “conduct ... is to be controlled,” ibid. If it is not *311likely to occur to potential wrongdoers as they seize the challenged evidence to care about its use for the particular purpose later in issue, then removing the possibility of that use is unlikely to deter them from unlawful conduct. As we have recently said, “[T]he deterrent purpose of the exclusionary rule would not be significantly advanced by . . . suppressing] . . . illegally seized evidence in a [proceeding] in which the offending police officers could not possibly have had an interest at the time they conducted the illegal search.” United States v. Rea, 678 F.2d 382, 389 (2d Cir. 1982). Thus, in order to decide whether application of the exclusionary sanction is likely to have a significant deterrent effect, the key question is whether the particular challenged use of the evidence is one that the seizing officials were likely to have had an interest in at the time — whether it was within their predictable contemplation and, if so, whether it was likely to have motivated them. And if law enforcement officers are already deterred from Fourth Amendment violations by a prohibition against using illegally seized evidence to secure convictions for the offenses they are investigating, the further question is whether some significant incremental deterrence is achieved by prohibiting use of the evidence for additional purposes.
In answering these questions, some estimates must be made. Since we cannot read the minds of the officers and lack the guidance of sound empirical models, we must “rel[y], instead, on [our] own assumptions of human nature and the interrelationship of the various components of the law enforcement system.” United States v. Janis, supra, 428 U.S. at 459, 96 S.Ct. at 3034. The primary consideration is the relationship between the law enforcement responsibilities and expertise of the seizing officials and the type of proceeding at which the seized material is being offered. The closer the nature of the proposed use for the evidence is to the seizing officers’ “zone of primary interest,” id. at 458, 96 S.Ct. at 3034, the stronger the inference that the officers had this use in mind when they made the seizure. For example, the Supreme Court concluded in Janis that as a general matter local police looking for evidence of gambling violations are probably not actively motivated by a desire to aid the federal tax authorities in winning a deficiency judgment against the person from whom the evidence is seized; as a result, “informing” them that anything they seized illegally would be excluded from federal tax proceedings would have added little to the already existing deterrent effect of threatening to exclude the material from the gambling prosecution. Other exclusionary rule cases also illustrate this point, confining the rule to a relatively narrow range of proceedings closely related to the goals of the investigation that occasioned the search. Thus, courts routinely prohibit governmental authorities from using illegally seized evidence in the proceedings for which the search was conducted, not only in a criminal prosecution, e.g., Mapp v. Ohio, 367 U.S. 643, 81 S.Ct. 1684, 6 L.Ed.2d 1081 (1961), but also in the variety of civil proceedings, e.g, Wong Chung Che v. INS, 565 F.2d 166, 168-69 (1st Cir. 1977) (documents seized by immigration officials investigating illegal aliens excludable from subsequent deportation hearing); Knoll Associates v. FTC, 397 F.2d 530, 534-35 (7th Cir. 1968) (documents seized for purposes of FTC investigation excluded from resulting hearing); Smyth v. Lubbers, 398 F.Supp. 777, 786 (W.D. Mich. 1975) (drugs seized from dormitory rooms by state college officials investigating violations of college rules excluded from resulting disciplinary proceeding); Iowa v. Union Asphalt & Roadoils, Inc., 281 F.Supp. 391, 406-09 (S.D. Iowa 1968), aff’d sub nom. Standard Oil Co. v. Iowa, 408 F.2d 1171 (8th Cir. 1969) (business records seized by state attorney general excluded from resulting civil antitrust proceeding).5
*312When the proceeding at which the seized evidence is offered is different from (normally, subsequent to) the proceeding for which the search was conducted, the exclusionary rule is still applied when there is a close relationship between the search and the secondary proceeding. Frequently the requisite relationship will be found when the search and the secondary proceeding were initiated by the same agency. Thus, a governmental authority is barred from using seized evidence not only in a criminal prosecution that was the purpose of the search but also in a civil suit later brought against the same defendant by the same agency. See, e.g., Pizzarello v. United States, supra (records seized by IRS for purposes of criminal tax prosecution for evading wagering taxes also excluded in subsequent civil tax proceeding); Vander Linden v. United States, 502 F.Supp. 693 (S.D. Iowa 1980) (records seized by IRS agent, participating in joint investigation of narcotics trafficking for purposes of criminal prosecution for tax evasion, also suppressed in subsequent civil proceeding for assessments and fraud penalties). The rationale for these results is that agents are likely to have all the responsibilities of their agency in mind as they go about their investigations.
Though the relationship between the search and the proceeding at which the evidence is presented provides a general indication of the likelihood of achieving marginal deterrence by applying the exclusionary rule, the specific facts of an investigation may reveal that an initial assumption about the officers’ likely motivation was incorrect. In Janis, for example, the exclusionary rule probably would have been invoked if evidence had shown that, contrary to general expectations, the police had planned their search in conjunction with, or received encouragement beforehand from, the IRS, or that the police had realized during the search that the IRS would be interested in their discoveries and for that reason had gone out of their way to aid the IRS. Any indication of an explicit and demonstrable understanding between the two law enforcement bodies would be decisive. Such collusion might also be inferred from the officers’ conduct, perhaps indicated by an incongruity between the official objective of the search (as announced in the warrant, for example) and the items actually seized. Thus, if in Janis the evidence seized had been only tenuously relevant to the gambling investigation but obviously important to tax investigations, the Court might have had to revise its assessment that the investigators were not acting out of an interest in revenue violations. By contrast, the fact that in that case the evidence seized was directly related to the gambling investigation was consistent with the Court’s expectation that the local police were not acting to aid the IRS.
In sum, we think that the proper way to assess the likelihood of marginal deterrence is to examine the relationship between the purposes of the search and those of a subsequent proceeding — both as a general matter and on the facts of the case. This approach, we believe, is sounder than those adopted by some courts, drawing on isolated language in Janis, that would either presume in favor of deterrence whenever a search and a subsequent proceeding are initiated by agents of the same sovereign, or presume against it whenever the subsequent proceeding is civil.
*313Courts that have read Janis to imply an intrasovereign rule of exclusion point to language in the opinion emphasizing the fact that the seizure there was made by agents of one sovereign (police of a California city), while the proceeding was brought by an agency of a different sovereign (the IRS of the United States), 428 U.S. at 455-56, 460, 96 S.Ct. at 3032-33, 3036; these cases also note that the Court in Janis specifically observed that its decision not to apply the exclusionary rule in that case did not foreclose the question of the applicability of the rule in proceedings “involving an intrasovereign violation,” id. at 456 n. 31, 96 5. Ct. at 3033 n. 31. See, e.g., Vander Linden v. United States, supra, 502 F.Supp. at 697; Guzzetta v. Commissioner, 78 T.C. 173 (1982). We see no reason to think that the Supreme Court intended to resolve the intrasovereignty issue, which it explicitly left open. Nor do we think, after our consideration of the question, that the fact of intrasovereignty holds promise as a reliable predictor of deterrence. The key issue, implicit in Janis as in other exclusionary rule decisions, is still what concerns the seizing officers had in mind. Intrasovereignty does not adequately reflect the complexity of such a prediction. Thus, a test based on sovereignty would represent a prediction that officials of one sovereign are likely to be deterred from illegal seizures by the threat that the evidence will be excluded from any proceeding conducted by the same sovereign, but will be indifferent to the same threat aimed at the proceedings of a different sovereign. Such a prediction sweeps too broadly. The fact that a federal law enforcement officer, attached to an agency and trained to investigate a limited class of unlawful acts, receives his pay check from the same treasury that pays all other federal agents neither charges him with concern for all the multiple and unrelated law enforcement activities of the federal government, nor warrants the suspicion that his motivations include a desire to help other agencies succeed. If anything, the rivalries and jealousies among many federal law enforcement agencies point to precisely the opposite conclusion. It would be unsound to invoke the exclusionary rule on the assumption that officers of one federal agency have such a strong motivating interest in all federal law enforcement concerns that broad application of the rule will achieve significant marginal deterrence.6 And we would not rule out the possibility that in some circumstances the exclusionary rule should be applied in intersovereign violations, perhaps if state narcotics or environmental officials unlawfully seize evidence later used by their federal counterparts.
At the same time, we think it equally unsound to reject the exclusionary rule in all civil proceedings. A few courts have tended in that direction, relying primarily on the Supreme Court’s observation in Janis that it “never has applied [the exclusionary rule] to exclude evidence from a civil proceeding, federal or state.” 428 U.S. at 447, 96 S.Ct. at 3028. See Todd Shipyards Corp. v. Secretary of Labor, 586 F.2d 683, 689 (9th Cir. 1978) (rule inapplicable in OSHA proceedings) (dictum); Morale v. Grigel, 422 F.Supp. 988, 1000-01 (D.N.H. 1976) (rule inapplicable in school disciplinary proceeding). We think such decisions read too much into what was only an observation of fact. A test for the exclusionary rule that turns on the civil or criminal character of *314the proceeding does not comport with an objective of achieving substantial deterrence. In the first place, all uses of the exclusionary rule do not necessarily have a significant deterrent effect in every criminal proceeding. For example, the rule’s use has been refused in prosecutions that no one could reasonably have anticipated at the time of the seizure because they were the products of the defendant’s own intervening voluntary acts. See, e.g., Walder v. United States, 347 U.S. 62, 74 S.Ct. 354, 98 L.Ed. 503 (1954) (evidence seized by federal officials and suppressed in state narcotics prosecution admissible in second prosecution on cross-examination in order to impeach defendant’s direct testimony); United States v. Paepke, 550 F.2d 385 (7th Cir. 1977) (evidence seized by state officials and suppressed in state narcotics prosecution admissible in federal prosecution for making false statements to IRS); United States v. Raftery, 534 F.2d 854 (9th Cir.) (evidence seized by state and federal officials and suppressed in state narcotics prosecution admissible in federal prosecution for perjury before grand jury), cert. denied, 429 U.S. 862, 97 S.Ct. 167, 50 L.Ed.2d 141 (1976). More important, the rule can have a deterrent effect in appropriate civil proceedings, as numerous federal courts have recognized. See, e.g., Wong Chung Che v. INS, supra; Knoll Associates v. FTC, supra; Vander Linden v. United States, supra; Smyth v. Lubbers, supra; Iowa v. Union Asphalt & Roadoils, Inc., supra. And if it is not used in such circumstances where it is needed, “the Government would be free to undertake unreasonable searches and seizures in all civil cases without the possibility of unfavorable consequences,” Pizzarello v. United States, supra, 408 F.2d at 586. We therefore agree that the exclusionary rule can be properly and beneficially applied in those civil proceedings where it has a realistic prospect of achieving marginal deterrence.
III.
Applying our approach to the facts of this case, we think that the deterrent purpose of the exclusionary rule would not have been served by suppressing in this civil Tax Court proceeding the seized evidence that formed the basis of the deficiency notice. Tax deficiency proceedings are too remote from the “zone of primary interest” of the narcotics agents who made the seizures in Tirado’s apartment. As in Janis, it is not reasonable to suppose that a rule barring use of the evidence in a civil tax proceeding would have materially influenced those agents in their decision whether to make the particular seizures. Obviously, collecting taxes was not the official objective of the search for “narcotics — -the means of committing a crime or offense, or the means of preventing a crime or offense from being discovered,” as stated in the warrant. Nor would agents of the Drug Enforcement Agency be likely to harbor a general motivating interest in assisting the enforcement of civil tax obligations.
No special circumstances in this case undermine our general expectation concerning the likely motivation of the agents. There is no indication that the agents made the seizures other than as part of a narcotics investigation and solely in order to obtain a narcotics conviction. Agent Panessa, one of the two federal agents involved, testified at the Tax Court suppression hearing that he considered everything that was seized, including financial records, to be evidence or fruits of possessing or trafficking in narcotics, that seizing financial records for that purpose was not uncommon, and that he had in fact seized them for that purpose. The Tax Court credited this testimony in ruling that the seizures did not exceed the scope of the narcotics warrant. Tirado v. Commissioner, supra. Agent Panessa testified further, without contradiction, that he had had no preexisting agreement with the IRS regarding the search of Tirado’s apartment. And despite Tirado’s claim that the narcotics agents “worked in close conjunction with the [IRS],” Appellant’s Reply Brief at 7, nothing in the record suggests that the task force for which the agents worked had any connection with the IRS, either in the form of direct IRS participation, see Vander Linden v. United States, *315supra (investigation by members of interagency “Drug Abuse Law Enforcement” project in which IRS played major role), or by means of a specific agreement, cf. Guzzetta v. Commissioner, supra (although “select liaison” existed between New York police and IRS for purpose of channeling evidence of tax violations to IRS, evidence not excluded where decisions to share information pursuant to liaison arrangement were made administratively and after the fact). There is, in short, no evidence that the IRS agents actually participated in the search, or that the Service encouraged it either directly, through a particular agreement regarding Tirado, or indirectly, through a general policy or liaison arrangement.
If the agents here had doubts about whether the seizures would withstand Fourth Amendment scrutiny, there is insufficient likelihood that those doubts would have led to a decision to forgo the seizures simply because there was a legal rule that any illegality would imperil efforts to force Tirado to pay his back taxes. Nor would knowing that the opposite were true — that the evidence could be used in a tax proceeding — by itself have dispelled those doubts, overcoming the agents’ primary concern for their own narcotics case. Under the circumstances, the deterrent purpose of the exclusionary rule would not have been served by applying the rule in this subsequent proceeding.
The judgment of the Tax Court is affirmed.
. The Deficiency Notice, issued May 9, 1974, was based in relevant part on the cash found in the apartment and the cash later seized from the safe deposit boxes, the amount in two bank accounts, the jewelry seized from one safe deposit box, the apartment rent for the year as determined by the lease and the rent notices and receipts, and various expenditures relating to the purchase and operation of the automobiles. See Tirado v. Commissioner, 74 T.C. 14, 18 n. 4 (1980). In the narcotics prosecution two and a half years before, the state court had admitted into evidence the cash found in the apartment, had suppressed the lease, the rent notices and receipts, and the bank notice about the safe deposit box rental, and had not considered the remaining items. Id. at 17-18. In the present federal proceedings, Tirado has challenged the seizure of all of these items; he apparently does not dispute the Tax Court’s conclusion that it was not bound by any of the state court’s findings regarding their admissibility, see id. at 20; Oregon v. Hass, 420 U.S. 714, 719, 95 S.Ct. 1215, 1219, 43 L.Ed.2d 570 (1975).
. We see no jurisprudential consideration that would require us first to consider whether the facts of the various seizures violate the Fourth Amendment and then, only upon a finding of a constitutional violation, proceed to consider whether the facts present a circumstance warranting application of the exclusionary rule. Each issue is of constitutional dimension, and each turns on the precise facts of the case. Focusing first on the exclusionary rule has the virtue of permitting disposition by a single ruling instead of the several rulings that would be required if each seizure were adjudicated. In electing to consider the applicability of the exclusionary rule without pausing to review the correctness of the trial court’s rulings on the lawfulness of the seizures, we follow a traditional approach. See United States v. Ajlouny, 629 F.2d 830 (2d Cir. 1980) (exclusionary rule inapplicable in light of recently enacted statute; constitutionality of foreign intelligence surveillance left undecided), cert. denied, 449 U.S. 1111, 101 S.Ct. 920, 66 L.Ed.2d 840 (1981); United States v. Caron, 474 F.2d 506 (5th Cir. 1973) (exclusionary rule inapplicable to wiretap evidence offered for impeachment; validity of wiretap left undecided).
. A separate justification for the exclusionary rule has sometimes been found in the “imperative of judicial integrity.” See United States v. Peltier, 422 U.S. 531, 536-38, 95 S.Ct. 2313, 2317-18, 45 L.Ed.2d 374 (1975); Elkins v. United States, 364 U.S. 206, 222-23, 80 S.Ct. 1437, 1446 — 47, 4 L.Ed.2d 1669 (1960). This principle dates to the first enunciation of the exclusionary rule in Weeks v. United States, 232 U.S. 383, 392, 34 S.Ct. 341, 344, 58 L.Ed. 652 (1914); see also Olmstead v. United States, 277 U.S. *310438, 470, 48 S.Ct. 564, 569, 72 L.Ed. 944 (1928) (Holmes, J., dissenting); id. at 483, 485, 48 S.Ct. at 575 (Brandéis, J., dissenting). Yet however venerable its pedigree, the independent significance of the imperative of judicial integrity “has now been discounted substantially, if not completely,” United States v. Ajlouny, 629 F.2d 830, 840 (2d Cir. 1980), cert. denied, 449 U.S. 1111, 101 S.Ct. 920, 66 L.Ed.2d 840 (1981), as a basis for invoking the exclusionary rule. As the Supreme Court has said, “While courts, of course, must ever be concerned with preserving the integrity of the judicial process, this concern has limited force as a justification for the exclusion of highly probative evidence.” Stone v. Powell, 428 U.S. 465, 485, 96 S.Ct. 3037, 3048, 49 L.Ed.2d 1067 (1976). The Court has even suggested that the idea of judicial integrity in the context of the exclusionary rule is best viewed as having been merged in the deterrence rationale, and for that reason has no independent vitality. See United States v. Janis, 428 U.S. 433, 458-59 n. 35, 96 S.Ct. 3021, 3034 n. 35, 49 L.Ed.2d 1046 (1976); Michigan v. Tucker, 417 U.S. 433, 450 n. 25, 94 S.Ct. 2357, 2367 n. 25, 41 L.Ed.2d 182 (1974).
. The presumed beneficial effect of the exclusionary rule in deterring official lawlessness has so far eluded either verification or rebuttal at the hands of empirical science. See United States v. Janis, 428 U.S. 433, 449-53, 96 S.Ct. 3021, 3029-31, 49 L.Ed.2d 1046 (1976) (reviewing available studies). This inconclusiveness may be inherent in the effect being studied. As the Supreme Court has noted, “Since as a practical matter it is never easy to prove a negative, it is hardly likely that conclusive factual data [in support of the rule’s deterrent effect] could ever be assembled.” Elkins v. United Slates, 364 U.S. 206, 218, 80 S.Ct. 1437, 1444, 4 L.Ed.2d 1669 (1960). This is a troubling fact about a legal rule whose legitimacy is based upon its practical effectiveness. Uncertainty as to the rule’s general deterrent effect makes even more hazardous any prediction as to its marginal deterrent effect when extended to secondary proceedings.
. In a handful of cases decided just after the Supreme Court’s application of the exclusionary rule to forfeiture proceedings, see One 1958 Plymouth Sedan v. Pennsylvania, 380 U.S. 693, 85 S.Ct. 1246, 14 L.Ed.2d 170 (1965), a few courts applied the exclusionary rule in civil proceedings by analogy to criminal prosecutions. These cases stretched the “quasi-crimi*312nal” rationale used in the forfeiture cases to reach, for example, a military proceeding to discharge a civilian employee, Powell v. Zuckert, 125 U.S.App.D.C. 55, 366 F.2d 634 (1966), and, interestingly, a civil tax assessment proceeding, United States v. Blank, 261 F.Supp. 180 (N.D. Ohio 1966). Powell v. Zuckert found the burdens resulting from a discharge proceeding relevantly comparable to the imposition of criminal sanctions, 366 F.2d at 640, and the District Court in United States v. Blank cited One 1958 Plymouth Sedan, supra, as authority for applying the exclusionary rule to “forfeiture and deficiency proceedings, civil in form, which inflict an onerous monetary penalty upon an accused which approximates the visitations of the criminal code,” 261 F.Supp. at 182, 183. Though Blank, like the case now before us, barred illegally seized evidence in a civil tax proceeding, the search there had been conducted by IRS agents.
. In fact, the few cases before and after Janis that have applied the- exclusionary rule when violations were intrasovereign all involved searches and subsequent proceedings initiated by officers of the same agency, see, e.g., Pizzarello v. United States, 408 F.2d 579 (2d Cir.), cert. denied, 396 U.S. 986, 90 S.Ct. 481, 24 L.Ed.2d 450 (1969); Vander Linden v. United States, 502 F.Supp. 693 (S.D. Iowa 1980); United States v. Blank, 261 F.Supp. 180 (N.D. Ohio 1966). The District Court in Vander Linden, which first ventured a broad intrasovereign rule (“[T]he ‘deterrent effect’ in an ‘intrasovereign’ situation would be furthered by excluding illegally obtained evidence in subsequent civil trial proceedings.” 502 F.Supp. at 697), explicitly emphasized the “same agency” aspect of the violation (“[TJhis is not only an intrasovereign situation but is actually the same division of the IRS that committed the illegal search and seizure that is now seeking to utilize evidence obtained by the illegal search and seizure.” Ibid.).