Affirmed by published per curiam opinion. Chief Judge Wilkinson wrote an opinion concurring in the judgment, in which Judge Widener, Judge Wilkins, Judge Niemeyer, Judge Williams, and Judge Traxler joined. Judge Murnaghan wrote a dissenting opinion, in which Judge Michael, Judge Motz, Judge King, and Senior Judge Hamilton joined. Judge Luttig wrote a dissenting opinion.

This case came to be argued before the en banc court on October 25, 1999. The judgment of the district court is hereby affirmed by an equally divided en banc court. Separate opinions follow seriatim.

AFFIRMED.

WILKINSON, Chief Judge, with whom Judges Widener, Wilkins, Niemeyer, Williams, and Traxler join, concurring in the judgment:

We concur in the court's judgment dismissing this case against North Carolina police officers Delma Collins and James Shingleton. Plaintiff Lesly Jean contends that Officers Collins and Shingleton violated his Fourteenth Amendment due process rights by failing to turn over exculpatory evidence to the prosecutor. Because plaintiff alleges at most a negligent miscommunication between these officers and the prosecutor, we would conclude that the officers have not deprived Jean of any Fourteenth Amendment right. As a result, we do not believe there can be any § 1983 liability. For the facts of this case we would rely on our earlier en banc opinion. See Jean v. Collins, 155 F.3d 701, 703-05 (4th Cir. 1998) (en banc). In that case, we held that as of "1982, a reasonable police officer would not have known that his failure to turn over such evidence violated a criminal defendant's clearly established constitutional rights." Id. at 708. The Supreme Court then granted certiorari, vacated the judgment, and remanded to this court for further consideration in light of Wilson v. Layne, 526 U.S. 603 (1999). See Jean v. Collins, 526 U.S. 1142 (1999). It is in light of Wilson then that we address the question of when police officers are liable under § 1983 for allegedly withholding exculpatory evidence from the prosecution and by extension a criminal defendant.

Wilson's directions are straightforward ones. "A court evaluating a claim of qualified immunity `must first determine whether the plaintiff has alleged the deprivation of an actual constitutional right at all....'" Wilson, 526 U.S. at 609 (quoting Conn v. Gabbert, 526 U.S. 286, 290, 143 L. Ed. 2d 399 (1999)); accord Siegert v. Gilley, 500 U.S. 226, 232 (1991). We must initially ask, therefore, if Jean has alleged a Fourteenth Amendment due process violation by Officers Collins and Shingleton.1 Deciding this issue first can save a defendant from having "to engage in expensive and time consuming preparation to defend the suit on its merits." Siegert, 500 U.S. at 232. It"also promotes clarity in the legal standards for official conduct, to the benefit of both the officers and the general public." Wilson, 526 U.S. at 609 (citing County of Sacramento v. Lewis, 523 U.S. 833, 841-42 n.5 (1998)).

In Jean v. Rice, 945 F.2d 82 (4th Cir. 1991), we held that Jean's due process rights had been violated. Specifically, we noted that "the government's failure [to turn over material impeachment evidence to the defense] was a violation of the principles announced in Brady and its progeny." Id. at 87. As a result of the prosecutor's Brady violation, Jean's request for a writ of habeas corpus was granted. See id. The question before us now is whether there was an additional constitutional violation in this case -- a due process violation by Officers Collins and Shingleton for withholding from the prosecutor the hypnosis recordings and reports.

The Supreme Court decisions establishing the Brady duty on the part of prosecutors do not address whether a police officer independently violates the Constitution by withholding from the prosecutor evidence acquired during the course of an investigation. See, e.g., Brady v. Maryland, 373 U.S. 83 (1963); Giglio v. United States, 405 U.S. 150 (1972); United States v. Agurs, 427 U.S. 97 (1976); United States v. Bagley, 473 U.S. 667 (1985). Recent cases, including some from this circuit, have pointed toward such a duty. This court has noted that, "[a] police officer who withholds exculpatory information from the prosecutor can be liable under... section 1983," Goodwin v. Metts, 885 F.2d 157, 162 (4th Cir. 1989), but only where "the officer's failure to disclose the exculpatory information deprived the § 1983 plaintiffs of their right to a fair trial," Taylor v. Waters, 81 F.3d 429, 436 n.5 (4th Cir. 1996). And in Carter v. Burch, the court noted that a police officer's actions in failing to turn over materially exculpatory evidence to a prosecutor "violate[d] [the § 1983 plaintiff's] constitutional rights." 34 F.3d 257, 264 (4th Cir. 1994). Other circuits have also suggested that there may be a duty here. See Brady v. Dill, 187 F.3d 104, 114 (1st Cir. 1999); Walker v. City of New York, 974 F.2d 293, 298-99 (2d Cir. 1992); Geter v. Fortenberry, 849 F.2d 1550, 1559 (5th Cir. 1988); Sanders v. English, 950 F.2d 1152, 1162 (5th Cir. 1992); Jones v. City of Chicago, 856 F.2d 985, 993-96 (7th Cir. 1988); McMillian v. Johnson, 88 F.3d 1554, 1566-70 (11th Cir.), amended by 101 F.3d 1363 (11th Cir. 1996).

These cases have left unclear the exact nature of any duty that the law imposes on police with regard to exculpatory evidence. Several characteristics of this duty, however, seem evident. First, alleged failures to disclose do not implicate constitutional rights where no constitutional deprivation results therefrom. In this context, the constitutional deprivation must be defined as a deprivation of liberty without due process of law. In the absence of a cognizable injury, such as a wrongful criminal conviction, police suppression of evidence might still give rise to claims under state law. But unless the § 1983 plaintiff can point to a constitutional injury caused by the suppression, no § 1983 remedy will lie. See Albright v. Oliver, 510 U.S. 266, 270-71 n.4 (1994) (plurality opinion) (substantive due process cannot transform a state law malicious prosecution claim into a constitutional claim); id. at 281-86 (Kennedy, J., joined by Thomas, J., concurring in judgment) (same); Taylor, 81 F.3d at 436 & n.5 ("To the extent that Goodwin bases its holding on a conclusion that the officer's failure to disclose exculpatory evidence deprived the § 1983 plaintiffs of a liberty interest in avoiding prosecution on less than probable cause, that reasoning has been rejected in Albright.").

Second, to speak of the duty binding police officers as a Brady duty is simply incorrect. The Supreme Court has always defined the Brady duty as one that rests with the prosecution. See, e.g., Brady, 373 U.S. at 87 ("suppression by the prosecution of evidence favorable to an accused upon request violates due process"); Giglio, 405 U.S. at 154 (satisfying Brady "is the responsibility of the prosecutor"); Moore v. Illinois, 408 U.S. 786, 794 (1972) ("The heart of the holding in Brady is the prosecution's suppression of evidence...."); Agurs, 427 U.S. at 108 ("the prosecutor's constitutional duty to disclose"); Bagley, 473 U.S. at 676 ("the prosecutor failed to disclose evidence"); Kyles v. Whitley, 514 U.S. 419, 437 (1995) ("the prosecution, which alone can know what is undisclosed, must be assigned the consequent responsibility to gauge the likely net effect of all such evidence and make disclosure when the point of `reasonable probability' is reached").

The Brady duty is framed by the dictates of the adversary system and the prosecution's legal role therein. Legal terms of art define its bounds and limits. The prosecutor must ask such lawyer's questions as whether an item of evidence has "exculpatory" or "impeachment" value and whether such evidence is "material." It would be inappropriate to charge police with answering these same questions, for their job of gathering evidence is quite different from the prosecution's task of evaluating it. This is especially true because the prosecutor can view the evidence from the perspective of the case as a whole while police officers, who are often involved in only one portion of the case, may lack necessary context. To hold that the contours of the due process duty applicable to the police must be identical to those of the prosecutor's Brady duty would thus improperly mandate a one-sizefits-all regime.

Third, it would be impermissible to hold the police liable for due process violations under § 1983 where they have acted in good faith. In Daniels v. Williams, the Supreme Court stated "that the Due Process Clause is simply not implicated by a negligent act of an official causing unintended loss of or injury to life, liberty, or property." 474 U.S. 327, 328 (1986). The Fourteenth Amendment mandates, "nor shall any State deprive any person of life, liberty, or property, without due process of law." U.S. Const. amend. XIV,§ 1. Daniels holds that, as a matter of plain constitutional text, no "deprivation" occurs on account of official negligence. 474 U.S. at 330-33. Indeed, negligent conduct cannot by definition establish the "affirmative abuse of power" necessary to constitute a due process deprivation. See id. at 330-32. Under Daniels, then, police officer negligence or inadvertence in failing to turn over evidence cannot be actionable under § 1983.

In an analogous case to the present, the Supreme Court refused to find that police officers violated the Due Process Clause in the absence of evidence that they acted in bad faith. In Arizona v. Youngblood the Court addressed that area of the law that "might loosely be called... constitutionally guaranteed access to evidence." 488 U.S. 51, 55 (1988) (internal quotation marks omitted). Youngblood involved police who failed to refrigerate clothing which contained semen stains and to perform tests on other semen samples. Id. at 53-55, 58. The defendant argued that properly preserved evidence might well have shown that he was innocent of any sexual assault. The Youngblood Court held, however, that "unless a criminal defendant can show bad faith on the part of the police, failure to preserve potentially useful evidence does not constitute a denial of due process of law." Id. at 58.

While Youngblood dealt with the failure to preserve evidence, its principles are certainly applicable to the present situation. Here, as in Youngblood, the prosecutor and ultimately the defense allegedly failed to receive exculpatory evidence from the police. Here, as in Youngblood, the police officers' actions were alleged to constitute a due process violation. The Youngblood Court stressed its "unwillingness" to read the Due Process Clause to impose"on the police an undifferentiated and absolute duty" in that context. Id.

We similarly decline to impose a sweeping duty on police in the instant situation and note the obvious drawbacks of doing so.2 For instance, such a duty would widen the legal gulf between prosecutors and police to such an extent as to make scapegoats of police for every item of evidence discovered post-trial. Prosecutors plainly enjoy absolute immunity in the exercise of their prosecutorial duties, of which the disclosure of Brady material to the defense is clearly one. See Kalina v. Fletcher, 522 U.S. 118, 123-29 (1997); Burns v. Reed, 500 U.S. 478, 486 (1991); Imbler v. Pachtman, 424 U.S. 409, 413-16, 430-31 (1976). To confer on prosecutors absolute immunity while denying to police the right to argue even bona fides would multiply exponentially litigation against even conscientious officers.

Further, the law has already placed ultimate responsibility upon the prosecutor for disclosing Brady material to the defense. When Brady violations occur, criminal defendants may have their convictions overturned. Because police knowledge is plainly imputed to the prosecution for purposes of the prosecutor's Brady duties, see Kyles v. Whitley, 514 U.S. 419, 437-38 (1995), the prosecutor bears the responsibility for implementing procedures designed to ensure that police officers turn over all evidence to him, see Giglio, 405 U.S. at 154. To hold officers responsible under § 1983 for internal miscommunications that Kyles and Giglio charge the prosecution with preventing is to have § 1983 suits and Brady doctrine heading in diametrically opposed directions. Moreover, the§ 1983 suit could well set up a continual exercise in finger-pointing between prosecutors and police over whose fault it was that the evidence never reached the defendant. Making internal communications between prosecutors and police the customary subject of§ 1983 litigation would thrust the federal courts deep into the operations of state prosecutors' offices, a breach of federalism principles for which the Due Process Clause of the Constitution provides no warrant.

This danger is plainly illustrated by this action. The prosecutor in Jean's case, Walter Vatcher, made clear via two affidavits that Officers Collins and Shingleton had indeed disclosed substantial amounts of information to him with regard to the hypnoses. Vatcher stated that the officers had "informed [him] of the existence of the hypnoses and identification procedures" used in the Jean investigation. Vatcher was further informed that "there were some changes in[Shingleton's] description after hypnosis," specifically "that under hypnosis [Shingleton] no longer recalled any facial hair and his description of the [suspect's] shirt may have changed somewhat." Vatcher also reported that the two officers had told him of the hypnosis information months before trial and that he received "complete updates on the investigation, including facts concerning the identification procedures used." Vatcher, in fact, commended Officers Shingleton and Collins for turning over "all [of the] evidence that [he] requested from them." There is even substantial evidence via affidavits from Smith, Collins, Shingleton, and Wilson that Vatcher had been told of the existence of the hypnosis recordings and reports. Vatcher, however, stated in one of his affidavits that he did not "recall being made aware of recordings and written records."

The difficulty of trying to sort out such everyday communications between prosecutor and police underscores the need to insist at a minimum that an actual bad faith deprivation of due process rights be alleged. The conduct of Shingleton and Collins lies well below this threshold. While Jean contends that the officers acted in a "willful, wanton and reckless manner," conclusory allegations will not suffice. See, e.g., Suarez Corp. Indus. v. McGraw, 202 F.3d 676, 684-91 (4th Cir. 2000); Hartley v. Parnell, 193 F.3d 1263, 1268 (11th Cir. 1999); Wilkinson v. Russell, 182 F.3d 89, 105-06 (2d Cir. 1999); Torres v. United States, 200 F.3d 179, 186 (3d Cir. 1999). Jean simply is unable to provide evidence of a bad faith deprivation carried out by Collins and Shingleton -- a matter on which the burden plainly rests with the plaintiff.3 For example, Jean does not claim that Collins and Shingleton destroyed or otherwise failed to preserve the evidence that is now at issue -- a fact which alone tends to negate any inference of bad faith. And while Jean again claims in conclusory fashion that the hypnosis recordings and reports were "patently exculpatory," he does not point to any evidence showing that the officers actually knew of the significance of these items. Moreover, Jean conceded both in his brief and at oral argument that Vatcher at the very least was made aware of the existence of the two hypnoses. In sum, Jean points to nothing that resembles the kind of affirmative misuse of power that the Supreme Court has indicated would implicate due process protections.4

Because there was no threshold bad faith deprivation, the precise contours of any duty on the police in situations such as these is something we need not explore in detail5 Several points, however, are clear. A Brady violation that resulted in the overturning of the § 1983 plaintiff's conviction is a necessary, but not a sufficient, condition for § 1983 liability on the part of the police. It is a necessary condition because the Brady violation establishes the requisite threshold of constitutional injury (a conviction resulting in loss of liberty) below which no § 1983 action can lie. It is not a sufficient condition, however, because the Brady duty is a no fault duty and the concept of constitutional deprivation articulated in both Daniels and Youngblood requires that the officer have intentionally withheld the evidence for the purpose of depriving the plaintiff of the use of that evidence during his criminal trial. This is what is meant by"bad faith." And that must be established on the basis of evidence, including among other things the nature of the withheld material, that would negate any negligent or innocent explanation for the actions on the part of the police. Of course the bad faith manipulation of evidence on the part of the police cannot be countenanced. Constitutional absolution for the concealment, doctoring, or destruction of evidence would fail to protect the innocent, fail to assist the apprehension of the guilty, and fail to safeguard the judicial process as one ultimately committed to the ascertainment of truth. But what occurred here was at worst a negligent miscommunication among Vatcher, Collins, and Shingleton -an incident for which Vatcher under Brady, Giglio, and Kyles bears ultimate responsibility. Any decision on Vatcher's part not to request additional hypnosis information cannot possibly be recast seventeen years later as a bad faith constitutional deprivation perpetrated by these officer defendants.

As Jean has failed to allege a constitutional violation on the part of Collins and Shingleton, we would affirm the judgment.

MURNAGHAN, Circuit Judge, with whom Circuit Judges MICHAEL, MOTZ, and KING, and Senior Circuit Judge HAMILTON join, dissenting:

The fate of Lesly Jean's § 1983 action is lamentable, though scarcely surprising. What is surprising is the tenuous methodology the concurrence employs to extinguish his civil rights action. In recent years, the Supreme Court has undeniably restricted the availability of § 1983, particularly as a mode of redressing due process violations. Nonetheless, the Court still recognizes some situations, however circumscribed, where § 1983 damages remain viable.

The concurrence's analysis of Jean's § 1983 claim reflects a fundamental misunderstanding of the Supreme Court's jurisprudence in this area. While the opinion is rhetorically consonant with the Court's skepticism about civil rights litigation, it is substantively at odds with the Court's caselaw on § 1983 and with accepted understandings of the Due Process Clause. Accordingly, I dissent.

From the outset, I have been of the view that Brady v. Maryland, 373 U.S. 83 (1963), controls the analysis of Jean's civil rights claim. It is true that Brady involved suppression of evidence by a prosecutor, whereas Jean's claim alleges suppression of evidence by police officers. However, the post-Brady case of Barbee v. Warden, Maryland Penitentiary, 331 F.2d 842 (4th Cir. 1964), clarified that the State's disclosure obligation applies to police officers as well as prosecutors:

[I]t makes no difference if the withholding[of exculpatory evidence] is by officials other than the prosecutor. The police are also part of the prosecution, and the taint on the trial is no less if they, rather than the State's Attorney, were guilty of the nondisclosure. If the police allow the State's Attorney to produce evidence pointing to guilt without informing him of other evidence in their possession which contradicts this inference, [police] officers are practicing deception not only on the State's Attorney but on the court and the defendant.

Id. at 846 (emphasis added) (footnote omitted).

As these lines from Barbee suggest, the Brady disclosure regime is about getting exculpatory evidence into the hands of the defendant, not about disaggregating the State's prosecutorial team into discrete sub-groups and having a defendant's constitutional rights turn on such rigid formalisms. This is why Brady's compulsory disclosure requirement applies to all officials working in furtherance of the State's prosecution.

Of course, the manner in which prosecutors and police officers comply with Brady is different, reflecting their different functions in the criminal justice system. Police officers do not disclose evidence to criminal defendants directly. Instead, the police accumulate evidence and then ministerially deliver it to the prosecutor. The prosecutor then makes a discretionary legal judgment about whether the evidence is material and exculpatory, such that Brady compels its disclosure to the defendant. This functional differentiation, however, should not obscure the fact that Brady creates a singular constitutional duty, which prosecutors and police officers are capable of breaching in factually different ways.

The concurrence does not seriously dispute that a Brady violation occurred. Instead, the dispute is whether Jean can redress this Brady violation in a § 1983 damages action against Officers Collins and Shingleton.

Section 1983 is not a repository of substantive rights but is simply a remedial mechanism for vindicating rights with independent constitutional foundations. See Baker v. McCollan, 443 U.S. 137, 144 n.3 (1979) (holding that § 1983 "is not itself a source of substantive rights, but a method for vindicating federal rights elsewhere conferred").

The constitutional right in the instant case could not be any clearer. In Jean v. Rice, 945 F.2d 82 (4th Cir. 1991), we held that the State of North Carolina violated Jean's due process rights, as recognized in Brady, when Officers Collins and Shingleton withheld hypnosisrelated evidence from the prosecutor -- evidence that could have assisted Jean in impeaching the credibility of key government witnesses. See id. at 87. Because we have already established that a Brady violation occurred, there is a core constitutional offense that forms the basis for a § 1983 action. The availability of § 1983 as a remedial measure would seem to follow inexorably from the very fact of the underlying Brady violation.

Not so to the concurrence. In order to block what it regards as a frivolous lawsuit, the concurrence has labored assiduously to divorce Jean's § 1983 action from the underlying Brady violation that engendered it. See ante at 659 ("In Jean v. Rice... we held that Jean's due process [Brady] rights had been violated.... The question before us now is whether there was an additional constitutional violation in this case -- a due process violation by Officers Collins and Shingleton for withholding from the prosecutor the hypnosis recordings and reports.") (emphasis added). The challenge for the concurrence has been coming up with a way to say two seemingly contradictory things: that while Jean's Brady rights were clearly violated, entitling him to reversal of his conviction, Jean can not vindicate his Brady rights against Collins and Shingleton in a separate § 1983 damages action.

Since the inception of this case, the concurrence has had considerable difficulty explaining its way out of this paradox. In our first en banc opinion dealing with Jean's case, the majority relied on qualified immunity to shut down Jean's § 1983 action.1 The majority began by holding that the obligation of police officers to disclose exculpatory evidence to prosecutors was now a matter of settled law. See Jean v. Collins, 155 F.3d 701, 710 n.3 (4th Cir. 1998) ("More recently this circuit has recognized that the failure of police officers to turn over evidence to a prosecutor may violate a criminal defendant's constitutional right to receive such evidence.") (citing Taylor v. Waters, 81 F.3d 429, 436 n.5 (4th Cir. 1996); Carter v. Burch, 34 F.3d 257, 264 (4th Cir. 1994); and Goodwin v. Metts, 885 F.2d 157, 162-63 (4th Cir. 1989)); see also id. ("[T]he decisions in Taylor, Carter, and Goodwin now provide notice to police officers that they can be subject to monetary damages under section 1983 for failure to disclose exculpatory evidence to the prosecutor.").

While well-settled today, the first en banc majority found that the "police to prosecutor" disclosure obligation was not clearly established in 1982 when Collins and Shingleton withheld the hypnosis evidence. The majority distinguished Brady-- a pre-1982 case that seemed to establish the applicable disclosure requirements -- by holding that it only imposed a disclosure obligation on the State generically. The disclosure duty's specific application to police officers, according to the majority, was a post-1982 doctrinal development. See Jean, 155 F.3d at 710 n.3. Thus, the officers were entitled to a defense of qualified immunity.

The Supreme Court's remand, which instructed us to reconsider Jean's case in light of Wilson v. Layne, 526 U.S. 603 (1999), gave the majority a chance to revisit its first en banc opinion. Wilson held that a court considering a defense of qualified immunity should first ask whether the disputed state action would violate present-day constitutional law, before reaching the backward looking question of whether the illegality of the state conduct was "clearly established" at the time it occurred. See id. at 609.

Pursuant to Wilson's directive, the majority (reconstituted as a "concurrence") now asks whether the failure of police officers to deliver exculpatory evidence to prosecutors would violate due process guarantees under year 2000 jurisprudence. See ante at 659-60. In its first en banc opinion, the majority held that the police-specific disclosure obligation was clearly established today, just not when Collins and Shingleton withheld the hypnosis evidence in 1982. Repudiating the position the majority took in the first en banc opinion, the concurrence now answers this question differently, holding that police officers have no independent duty (at least no independent Brady duty) under year 2000 due process principles to disclose exculpatory evidence to prosecutors. See ante at 659-60.

The concurrence defends this Janus-faced maneuver by seeking refuge in the Supreme Court's directive to revisit Jean's case in light of Wilson. See ante at 659 n.1 ("It would be surprising if the court had not reviewed its earlier approach in light of Wilson's directive. Indeed, it would be irresponsible for us not to do so."). This appeal to Wilson, however, is unavailing. All Wilson said was: "Answer the year 2000 question first," not "Answer the year 2000 question differently."

The real explanation for the concurrence's vacillation has to do with post hoc misgivings about the impact of its first en banc opinion. Dismissing Jean's case on a qualified immunity rationale got rid of one irksome civil rights plaintiff; but by recognizing a police-specific disclosure obligation under year 2000 law, the concurrence unwittingly opened the flood gates, inviting future criminal defendants to extort money damages from "cops on the beat" every time a problem with exculpatory evidence arises. In order to seal the rupture its first en banc opinion created, the concurrence now holds that police officers who withhold exculpatory evidence from prosecutors are not independently liable for Brady violations, even under year 2000 jurisprudence. Thus, under the concurrence's new formulation, Brady can never furnish the doctrinal basis for a § 1983 action against police officers who fail to disclose exculpatory evidence. This was true in 1982, when Collins and Shingleton withheld evidence from Lesly Jean (see en banc #1) and it remains true today (see en banc #2).

The concurrence has several theories about why criminal defendants who are the victims of police non-disclosure may not rely on Brady to generate § 1983 actions against the offending police officers. The theories fall roughly into two categories. The first group relies on restrictive interpretations of Brady and the substantive constitutional rights it creates. The second group relies on a restrictive interpretation of § 1983 by grafting heightened culpability requirements onto the § 1983 apparatus. These culpability requirements exceed what is required to prove a violation of the underlying constitutional right; but according to the concurrence, they must be established in order to trigger the heavy remedial machinery of a § 1983 damages action.

The first way the concurrence deflects Jean's § 1983 claim is by manipulating the substantive basis of his claim: the right to exculpatory evidence established in Brady. The concurrence frames Jean's Brady claim in police-specific terms, as premised on the failure of police officers to disclose exculpatory evidence to prosecutors. See ante at 659. The concurrence then describes the Brady duty in such a restrictive way that police officers (and thus Jean's claim) fall outside its ambit. See ante at 660.

The concurrence produces this perverse result by articulating the State's Brady duty in prosecution-specific terms. This narrow construction of Brady derives from the concurrence's adherence to an elaborate legal fiction called the "imputed to" theory. See ante at 661-62. Under this theory, police officers are indirectly subject to Brady, in the sense that they are expected to furnish prosecutors with evidence they accumulate. Police officers, however, are not directly subject to Brady because the disclosure duty it creates is ultimately the prosecutor's responsibility. Thus, when police officers fail to disclose exculpatory evidence, Brady "imputes" these non-disclosures to the prosecution for purposes of assessing the State's overall compliance with Brady.

Because the prosecution is charged with "constructive" knowledge of all evidence in the possession of its sub-agents, failure of the police to hand over exculpatory evidence can imperil the integrity of the State's prosecution, often leading to reversal of criminal convictions (as happened in the instant case). Non-disclosures, however, cannot subject the police to civil liability because the police have no independent duty to criminal defendants under Brady, requiring them to supply the prosecutor with exculpatory evidence. Instead, the State's Brady obligation to criminal defendants is discharged through the instrumentality of the prosecutor.

At its simplest level, the concurrence's "imputed to" theory is an attempt to shift blame for the Brady violation in this case onto the prosecutor (Mr. Vatcher), who conveniently enjoys absolute immunity from damages actions. See Imbler v. Pachtman, 424 U.S. 409, 424-27, 431 n.33 (1976).

The concurrence also attempts to buttress the "imputed to" theory by resting it on more principled foundations -- some precedential, some policy-based. Neither precedent nor policy, however, support this attempt to circumscribe the State's Brady duty.

First, the concurrence claims to derive the "imputed to" theory from the language of Brady itself, along with the language of several post-Brady opinions, which describe the State's disclosure duty in prosecution-specific terms. See Brady, 373 U.S. at 87 ("[S]uppression by the prosecution of evidence favorable to an accused upon request violates due process...."); Moore v. Illinois, 408 U.S. 786, 794 (1972) ("The heart of the holding in Brady is the prosecution's suppression of evidence...."); United States v. Agurs, 427 U.S. 97, 108 (1976) (referring to "the prosecutor's constitutional duty to disclose").

Conspicuously absent from the concurrence's discussion, however, is the important post-Brady case of Barbee, which dispelled the notion that the police are exempt from the Brady disclosure duty. In Barbee, police officers failed to disclose exculpatory ballistics and fingerprint tests that tended to show a revolver other than the defendant's was responsible for the crime in question. See Barbee, 331 F.2d at 844. We found that the ballistics and fingerprints tests were indeed exculpatory because they undermined the testimony of several witnesses who had identified the defendant's revolver as the weapon used in the shooting. Thus, we held that the police should have disclosed the exculpatory evidence and issued a writ of habeas corpus. See id. at 847.

In defending the non-disclosure, the State of Maryland argued that there was no constitutional infirmity because the duty of disclosure fell on the prosecution, not the police. See id. at 844. Because the prosecution had no knowledge that the police were in possession of the exculpatory evidence, the State argued that no Brady violation had occurred. See id. We rejected this formalistic distinction, holding that when "the police allow the State's Attorney to produce evidence pointing to guilt without informing him of other evidence in their possession which contradicts this inference, [police] officers are practicing deception not only on the State's Attorney but on the court and the defendant." Id. at 846 (emphasis added) (footnote omitted).

It is difficult to reconcile the concurrence's "imputed to" theory with Barbee's plain statement that the police misconduct practiced a deception "not only" on the prosecutor, but on the court "and the defendant." Of course, the police officer's duty is to disclose to the prosecution rather than to the defendant directly; but according to Barbee, the constitutional injury arising from an officer's nondisclosure runs directly from the police to the defendant, unmediated by the prosecutor. See id. ("Failure of the police to reveal... material evidence in their possession [to the prosecution] is equally harmful to a defendant whether the information is purposefully, or negligently, withheld.") The factual interposition of the prosecutor between the police and the defendant does not transform the prosecutor into a supervening legal cause of the constitutional injury.

The Supreme Court confronted a similar issue years later in Kyles v. Whitley, 514 U.S. 419 (1995). There, the State of Louisiana made the same argument the State of Maryland made in Barbee: namely, that the State should not be held accountable under Brady for evidence known only to police investigators and not to the prosecutor. See id. at 438. In language echoing Barbee, the Court criticized the State's attempt to draw a formalistic dichotomy between police and prosecutors and rejected the State's "argument for excusing a prosecutor from disclosing what he does not happen to know." Id. The State's disclosure obligation under Brady applies even when the prosecutor is completely in the dark, the Court held, because the ultimate injury to the defendant is the same, regardless of whether the police or the prosecutor is to blame. Thus, like Barbee, the Kyles opinion made clear that the goal of Brady was getting exculpatory evidence to the defendant, not creating an analytical patchwork of legally distinct disclosure duties for every sub-group working in furtherance of the State's criminal prosecution.

Some of the language in Kyles did suggest that, as a practical matter, the prosecutor's office is the fulcrum of the Brady disclosure obligation. See id. at 437 ("[T]he individual prosecutor has a duty to learn of any favorable evidence known to the others acting on the government's behalf in the case, including the police."). The insight here, however, was pragmatic rather than doctrinal. Because Kyles involved a habeas petition rather than a § 1983 damages action, the Court was obviously more concerned with the pragmatic question of how to keep the State from evading its Brady obligation than it was with the legal question of how to carve up the State's investigativeprosecutorial team, either as an abstract doctrinal exercise or for purposes of allocating distinct spheres of civil liability. Thus, the fact that Kyles made the prosecutor vicariously responsible when police officers fail to disclose exculpatory evidence does not mean that Kyles also placed police officers beyond constitutional reproach for their personal role in subverting the State's Brady duty.

The concurrence also advances a policy argument in support of its "imputed to" theory. See ante at 661-62. The concurrence argues that only prosecutors are trained to make the nuanced and quintessentially legal judgment of whether a piece of evidence is truly exculpatory, such that disclosure to the defendant is constitutionally compelled. Because police officers are unfamiliar with the "[l]egal terms of art," ante at 660, necessary to make this complex determination, it would be unfair to hold them monetarily liable when they blunder. On this view, the "imputed to" theory, which places ultimate constitutional responsibility for exculpatory evidence problems in the hands of the prosecutor, is merely a recognition of the prosecutor's special institutional competence -- and of the unfairness involved in asking a police officer to pay money damages for mistakes he was not trained to avoid.

This observation is a strawman that confuses the crucial issue. It presupposes that when a police officer discloses evidence to a prosecutor, the act is functionally identical to the discretionary legal judgment prosecutors make when disclosing evidence directly to criminal defendants. In reality, the two acts are incommensurable. Requiring police officers to disclose evidence to prosecutors does not require technical legal expertise because the act is essentially ministerial, not discretionary. The police officer's duty is not to determine whether the evidence is material and exculpatory. His duty is simply to collect the evidence and to disclose all of it to the prosecutor, who then makes the discretionary legal judgment about its material, exculpatory attributes. Thus, there is no reason to "impute" the misconduct of police officers to prosecutors when police officers are perfectly capable of understanding their ministerial function and still choose to disregard it -- not out of ignorance but out of dereliction.

By cabining Brady so that the duty it creates applies only to prosecutors, the "imputed to" theory helps the concurrence extinguish Jean's § 1983 claim. It does so by eroding the constitutional theory on which his § 1983 claim is based: the theory that police officers independently violate Brady when they withhold evidence from prosecutors.

The concurrence also attacks Jean's § 1983 action from another angle. The strategy behind the "imputed to" theory was to cripple Jean's § 1983 claim by eroding its underlying constitutional foundation. The concurrence's second strategy is to target the § 1983 remedial apparatus itself: specifically, by creating heightened culpability requirements that plaintiffs must meet in order to convert constitutional injuries they have suffered into the basis for money damages.

The concurrence resorts to this second strategy, rather than resting on the "imputed to" theory, in order to resolve a doctrinal problem created by the "bad faith" requirement the concurrence inartfully derives from Daniels v. Williams, 474 U.S. 327 (1986), and Arizona v. Youngblood, 488 U.S. 51 (1988). According to the concurrence, citizens who allege violations of the Due Process Clause must make a threshold showing that the due process violation occurred in a culpable, "bad faith" manner. Because Jean failed to prove that Collins and Shingleton withheld the hypnosis evidence with the requisite degree of culpability, the concurrence holds that Jean's claim fails on that ground alone. See ante at 660-62.

The problem with the concurrence's "bad faith" thesis, however, is that Brady makes the non-disclosure of exculpatory evidence a violation of the Due Process Clause "irrespective of the good faith or bad faith" of the non-disclosing official. Brady, 373 U.S. at 87; see also Barbee, 331 F.2d at 846 ("If the police silence as to the existence of the [exculpatory] reports resulted from negligence rather than guile, the deception is no less damaging."). The concurrence is therefore placed in the untenable position of having to reconcile this language with the central proposition of its opinion: that only "bad faith" failures to withhold exculpatory evidence violate due process.

Much of the concurrence's opinion is a misguided search for the theory that explains why "bad faith" is not required to prove due process violations under Brady, but is required for every other due process violation --including the due process claim Jean has brought against Collins and Shingleton (which apparently rests on some other strand of due process jurisprudence, independent of Brady, that the majority never fully specifies).

Rather than working with the limits of the caselaw, the concurrence invents a theory out of whole cloth. According to the concurrence, the "bad faith" requirement does not apply to all due process cases. Nor does it apply to every § 1983 action. It only applies to that subset of cases that involve both § 1983 and the Due Process Clause -- that is, where criminal defendants are dissatisfied with the conventional due process remedy of a new trial and turn to § 1983 as a supplemental remedial mechanism. When criminal defendants seek nothing more than a new trial to redress a due process infirmity, Brady's "no fault" principles apply, and the defendant gets a new trial regardless of the good faith or bad faith manner in which the due process violation occurred. When criminal defendants seek money damages under § 1983, however, the "fault-based" doctrines articulated in Daniels and Youngblood apply. Unless the due process violation occurred in a culpable, "bad faith" manner, no damages remedy will lie. See ante at 662-63.

Thus, what the concurrence contemplates is a two-tiered regime, with one set of "no fault" standards for proving garden-variety violations of the Due Process Clause and a heightened, more stringent set of "fault-based" standards for redressing those same due process violations in § 1983 damages actions. In so holding, the concurrence invents a new theory of civil rights litigation that departs from established understandings of § 1983 and the Due Process Clause.

The Supreme Court has rejected the concurrence's two-tiered formulation in Parratt v. Taylor, 451 U.S. 527 (1981). There, the Court held unequivocally that § 1983 does not create a heightened culpability requirement above and beyond the culpability required to establish a violation of the underlying constitutional right. See id. at 534 ("Nothing in the language of § 1983 or its legislative history limits the statute solely to intentional deprivations of constitutional rights."); see also id. ("Section 1983, unlike its criminal counterpart, 18 U.S.C. § 242, has never been found by this Court to contain a state-of-mind requirement.").

The concurrence presumably thinks that Parratt is inapposite because the concurrence believes (falsely) that it has derived the heightened culpability requirement from the Due Process Clause, as interpreted in Daniels and Youngblood, rather than from § 1983 itself. On this theory, all Parratt forecloses is the grafting of a scienter requirement onto the whole § 1983 apparatus; but if the scienter requirement comes from the underlying constitutional right (here, the Due Process Clause), there is no Parratt problem.

There are two responses to this specious argument. One is that the concurrence's two-tiered formulation, while ostensibly derived from the Due Process Clause, still violates Parratt because its effect is to endow § 1983 with a heightened scienter requirement. Reading a culpability requirement into the Due Process Clause and then applying it only to due process claims brought under§ 1983 is an end run around Parratt. Such differential treatment of "ordinary" constitutional claims, on one hand, and "§ 1983" constitutional claims, on the other hand, is the very thing that Parratt prohibits.

The second, and more important, response is that the two-tiered formulation the concurrence purports to derive from the Due Process Clause has no credible foundation in due process jurisprudence. The concurrence has simply misunderstood the two due process cases on which it relies: Daniels and Youngblood.

Daniels is a difficult case because the Court, along with judicial commentators, often describe the opinion as having"overruled" Parratt. This is only true in part, however. Parratt is a case with two separate, but easily confused, holdings. One holding was reaffirmed by Daniels; the other holding Daniels overruled.

Parratt dealt with a § 1983 action where the underlying constitutional claim involved a violation of the Due Process Clause. Parratt first addressed the question of whether § 1983 itself imposed a threshold scienter requirement on every constitutional claim brought under its auspices. The Court framed the inquiry as "whether mere negligence will support a claim for relief under § 1983." See Parratt, 451 U.S. at 532.

The Court began by quoting the language of § 1983, which speaks of creating a damages remedy for "deprivations" of constitutional rights effected under color of state law. See 42 U.S.C. § 1983. The question in Parratt was whether "deprivation" in this context connoted a heightened state-of-mind requirement, based on the theory that § 1983 damages remedies should be reserved for only the most egregious, willful constitutional violations. The Court said "No." See Parratt, 451 U.S. at 534 ("Nothing in the language of § 1983... limits the statute solely to intentional deprivations of constitutional rights"); see also id. at 535 ("[Section] 1983 affords a civil remedy for deprivations of federally protected rights... without any express requirement of a particular state of mind.") (emphasis added).

Having decided that § 1983 contains no scienter requirement, the Parratt court went to the underlying constitutional claim the plaintiff had asserted (a violation of the Due Process Clause) and asked if that constitutional provision contained its own internal scienter requirement, similar to other constitutional provisions like the Equal Protection Clause (which requires proof of "purposeful" discrimination). See id. at 547-48 (Powell, J., concurring) ("[This case] requires the Court to determine whether intent is an essential element of a due process claim, just as we have done in cases applying the Equal Protection Clause....").

If the Due Process Clause were to contain a scienter requirement, the Court thought it would be found in the term"deprivation" as it appears in the text of the Due Process Clause.2 See id. at 548 (Powell, J., concurring) ("In the due process area, the[scienter] question is whether intent is required before there can be a`deprivation' of life, liberty, or property."). The Parratt Court held that the term "deprivation" in the Due Process Clause contained no scienter requirement -just as the term "deprivation" in the text of§ 1983 was devoid of a state-of-mind requirement. See id. at 536-37.

Justice Powell's concurrence, however, raised serious concerns about the majority's failure to recognize a scienter requirement as part of the Due Process Clause. According to Justice Powell, "`deprivation' connotes an intentional act denying something to someone, or, at the very least, a deliberate decision not to act to prevent a loss." See id. at 548. Thus, in Justice Powell's view, merely negligent invasions of life, liberty or property by the State, "causing unintended loss of or injury to [protected liberty or property interests]," do not work "a deprivation in the constitutional sense." Id. Consequently, according to Powell, "[t]he most reasonable interpretation of the Fourteenth Amendment would limit due process claims to such active deprivations." Id. Under this formulation, states should only be required to provide citizens with "due process of law" when they invade a life, liberty, or property interest in a culpable manner.

Several years later, the Court revisited Parratt in Daniels v. Williams. The Daniels court reaffirmed the first holding of Parratt. See Daniels, 474 U.S. at 329-30 ("In Parratt v. Taylor, we granted certiorari... to decide whether mere negligence will support a claim for relief under § 1983.... [W]e concluded that § 1983, unlike its criminal counterpart, 18 U.S.C. § 242, contains no state-of-mind requirement independent of that necessary to state a violation of the underlying constitutional right. We adhere to that conclusion.") (emphasis added) (internal quotations and citations omitted).

Daniels, however, did modify Parratt in another sense. Having determined that § 1983 contains no independent state-of-mind requirement, the Daniels Court addressed the legally distinct question of whether the Due Process Clause of the Fourteenth Amendment contains a state-of-mind requirement. See id. at 330 ("[I]n any given § 1983 suit, the plaintiff must still prove a violation of the underlying constitutional right; and depending on the right, merely negligent conduct may not be enough to state a claim.")

Parratt had held that the Due Process Clause, unlike the Equal Protection Clause, contains no culpability requirement. Daniels overruled Parratt on this one point alone, adopting the position of Justice Powell, who had criticized the reasoning of Parratt on the ground that the word "deprive" in the Due Process Clause connotes more than a merely negligent act. Thus, Daniels "overrule[d] Parratt to the extent that it state[d] that mere lack of due care by a state official may `deprive' an individual of life, liberty, or property" under the Due Process Clause. See Daniels, 474 U.S. at 330-31.

Thus, after Daniels, a violation of the Due Process Clause requires proof of two elements: (1) that the state official"deprived" (i.e., culpably denied) a citizen of life, liberty, or property and (2) that the "deprivation" occurred "without due process of law." If these two requirements are met, a due process violation is established. There are no additional culpability requirements or conditions precedent that must be satisfied in order to redress the due process violation in a § 1983 damages action.

The concurrence has considerable difficulty with the concept of "deprivation" articulated in Daniels. The concurrence seems to think that it has two separate applications in the instant case, when in reality it has only one. According to the concurrence, § 1983 plaintiffs seeking money damages for the non-disclosure of exculpatory evidence must first prove that the non-disclosure "deprived" them of their liberty interest in not going to jail. See ante at 660 ("[A]lleged failures to disclose do not implicate constitutional rights where no constitutional deprivation results therefrom. In this context, the constitutional deprivation must be defined as a deprivation of liberty without due process of law.... [There must be] a cognizable injury, such as a wrongful criminal conviction...."). Here, "deprivation" seems to mean "infringement," devoid of any "bad faith" connotation; and the "liberty" interest that is the object of the "deprivation" seems to mean freedom from incarceration.

The concurrence, however, then abruptly shifts to a new definitional framework. According to the concurrence, proof that the nondisclosure of exculpatory evidence resulted in your incarceration is just a baseline requirement. This sort of "deprivation" of "liberty" is good enough to get your conviction reversed. If a criminal defendant turns around and seeks money damages under § 1983, however, he must prove a different type of "deprivation" of a different type of "liberty." Specifically, says the concurrence, the criminal defendant must prove that the state actors who withheld the exculpatory evidence culpably "deprived" him of his "liberty" interest in unfettered access to exculpatory evidence. See ante at 663.

Two things immediately stand out. First, the concurrence has surreptitiously changed its definition of the underlying "liberty" interest. Second, the concurrence has surreptitiously changed the definition of "deprivation." Whereas "deprivation" in the earlier context just meant "infringement," without any scienter component, "deprivation" in this latter context requires the showing of "bad faith" discussed in Daniels -- a perplexing and incongruous result the concurrence produces through judicial sleight-of-hand. Cf. ante at 659 ("In this context [where only a new trial is at stake], the constitutional deprivation must be defined as a deprivation of liberty [i.e., incarceration] without due process of law"), with ante at 663 ("[Proof of incarceration] is not a sufficient condition [for § 1983 damages], however, because the Brady duty is a no fault duty... [whereas] the concept of constitutional deprivation articulated in both Daniels and Youngblood requires that the officer have intentionally withheld the evidence for the purpose of depriving the plaintiff of the use of that evidence during his criminal trial. That is what is meant by`bad faith.'").

In reality, the "bad faith" requirement discussed in Daniels attaches to the "deprivation" component of the Due Process Clause in every due process case, not just § 1983 cases. This is because Daniels is a restrictive interpretation of the Due Process Clause itself, not a restriction on the operation of § 1983.3

The crucial point the concurrence overlooks is that, in a non-§ 1983 context, where a defendant seeks only a reversal of his conviction on due process grounds, the requirement that the defendant prove a "bad faith" deprivation of some protected liberty interest is satisfied by the simple act of the State's prosecution -- a deliberate, intentional effort by the government to send a citizen to jail. It may seem strange to refer to every prosecution as being in "bad faith," unless the prosecution was executed in an unprincipled manner. But the"bad faith" requirement in Daniels, rooted in the word"deprivation," simply means a state action that is more than merely negligent -- an act that involves the willful, intentional use of government power to invade some protected life, liberty, or property interest possessed by its citizens. See Parratt, 451 U.S. at 548 (Powell, J., concurring) ("A `deprivation' connotes an intentional act denying something to someone...."); see also id. at 548 n.4 ("[T]o `deprive' is to dispossess; bereave; divest; to hinder from possessing...."). It is these types of intentional, willful exertions of government authority that raise the "abuse of power" concerns at the heart of the Due Process Clause -concerns the Due Process Clause mitigates by making the government observe certain procedural safeguards prior to effecting the deprivation of those protected interests. See Daniels, 474 U.S. at 331 (holding that the Due Process Clause was "intended to secure the individual from the arbitrary exercise of the powers of government") (quoting Hurtado v. California, 110 U.S. 516, 527 (1884)).

Because Daniels creates only one "deprivation" analysis, rather than two, the concurrence's two-tiered formulation of Jean's case is analytically indefensible. The only thing that could save the concurrence's approach is if there were two underlying liberty interests: first, a liberty interest in avoiding incarceration (which the state "deprives" when it intentionally brings a prosecution that results in a conviction); and second, a liberty interest in unfettered access to exculpatory evidence (which state officers "deprive" when they withhold evidence in a culpable, intentional manner).

To my knowledge, no court has ever recognized this type of freestanding liberty interest in exculpatory evidence; and I would be surprised if the Supreme Court, which has severely cut back on the expansion of fundamental liberty interests under substantive due process jurisprudence, would countenance the concurrence's curious effort to do so.4 See Paul v. Davis, 424 U.S. 693, 710-14 (1976).

The concurrence's approach is foreclosed by Graham v. Connor, 490 U.S. 386 (1989). In Graham, the Supreme Court held that, where a particular constitutional provision "provides an explicit textual source of constitutional protection" against a particular sort of government behavior, that specific constitutional provision and "not the more generalized notion of `substantive due process' must be the guide for analyzing these claims." Id. at 395. In other words, courts may not recharacterize state misconduct as a violation of "fundamental" liberty interests under substantive due process jurisprudence when there is a pre-existing constitutional provision that already proscribes the same conduct.

In the instant case, the right to exculpatory evidence is already protected under the procedural component of the Due Process Clause. Under this constitutional doctrine, States may not"deprive" citizens of their liberty interest in avoiding incarceration unless they first provide "due process of law" -- a term that encompasses the panoply of procedural safeguards we associate with the concept of a "fair trial." Brady is a case in this long doctrinal development elaborating the elements of a "fair trial." It held that part of the "process" the State must provide, prior to effecting the deprivation of a criminal defendant's liberty, is an unconditional right of access to exculpatory evidence. See Brady, 373 U.S. at 87.

Thus, because the right to exculpatory evidence is already protected under procedural due process, Graham prohibits the concurrence from recharacterizing the police officers' withholding of evidence as governed by substantive due process (based on the unsupportable notion that there is a "fundamental" liberty interest in exculpatory evidence).5

The errors in the concurrence's analysis all seem to derive from its belief that Jean's habeas petition and his § 1983 action rest on different legal foundations. For purposes of getting his conviction reversed, the concurrence says that Jean properly relied on Brady and traditional procedural due process principles, which delineate the elements of a "fair trial." See Jean v. Rice. To get money damages for the non-disclosure of exculpatory evidence, however, the concurrence thinks that Jean needs a different legal theory, which treats the right to exculpatory evidence as a "fundamental" liberty interest rather than a "fair trial" guarantee. As part of this legal theory, Jean must also prove that the "deprivation" of his liberty interest in exculpatory evidence occurred in "bad faith" -- a culpability requirement the concurrence lifts from Daniels and then selectively applies only in that subset of cases where due process claims are brought under § 1983.

The assumption that informs the concurrence's two-tiered approach seems to be that there are two different types of due process violations, each of which is correlated with a different type of due process remedy. When the due process violation involves the state's failure to provide a procedural safeguard that comprises one element of a "fair trial" (such as the right to exculpatory evidence), the concurrence thinks the sole remedy is a new trial. A § 1983 damages remedy, however, is only available under the concurrence's analysis when the due process violation involves tortious conduct by a state official -specifically, that subset of tortious conduct discussed in Daniels, involving culpable "deprivations" of state-created liberty or property interests.

This is presumably why the concurrence refuses to analyze Jean's § 1983 action as a simple attempt to vindicate the Due Process Clause's "fair trial" guarantee. Instead, when money damages are at stake, the concurrence implies that the non-disclosure of exculpatory evidence must be treated as a tortious interference with a criminal defendant's liberty interest in exculpatory evidence, rather than a procedural defect in the criminal defendant's "fair trial." Believing that the non-disclosure of exculpatory evidence is actionable under § 1983 only on a "tort" theory, rather than a "fair trial" theory, the concurrence analyzes Jean's section 1983 action as a tort-like claim, rather than a Brady claim.

Though the concurrence never fully specifies the tort that encompasses the right to exculpatory evidence, it appears to be the tort of malicious prosecution. This tort, recognized in most states, protects criminal defendants from the malicious institution or continuation of criminal proceedings that are unsupported by probable cause. See Goodwin v. Metts, 885 F.2d 157, 160 n.1 (4th Cir. 1989). Liability for malicious prosecution attaches not only to prosecutors who pursue criminal convictions without probable cause; it also attaches to police officers who cause an unfounded prosecution to continue by failing to tell the prosecutor about new exculpatory evidence uncovered since the defendant's arrest and indictment. See id. at 161-62.

As the Supreme Court noted in Albright v. Oliver, 510 U.S. 266 (1994), "the extent to which a claim of malicious prosecution is [also] actionable under § 1983 is one on which there is an embarrassing diversity of judicial opinion." Id. at 270 n.4 (internal quotations omitted). In Goodwin, decided three years before Albright, we considered a claim of malicious prosecution brought against a police officer who failed to disclose exculpatory information to a prosecutor -- evidence that would have caused the prosecutor to drop the charges had he been informed of its existence. The plaintiff brought state common law claims for malicious prosecution and a § 1983 federal claim, based on the theory that the Due Process Clause constitutionalizes the state tort of malicious prosecution. We agreed, holding that "[a] police officer who [maliciously] withholds exculpatory information from the prosecutor can be liable [for malicious prosecution] under both section 1983 and the state common law." Id. at 162.

Recasting Jean's § 1983 claim on the malicious prosecution foundation recognized in Goodwin has two implications. First, the use of a "tort-based" theory rather than a "fair trial" theory enables the concurrence to apply Daniels' "deprivation" requirement in the selective manner it desires. Specifically, the malicious prosecution framework forces Jean to prove that the "tort" Collins and Shingleton committed when they withheld the hypnosis evidence culpably infringed Jean's liberty interest in exculpatory evidence. Second, the requirement that malicious prosecution claims be supported by evidence of "malice" also ratchets up the culpability bar, even if only rhetorically, and makes it that much harder for Jean to state a successful claim for money damages.

In Albright, however, the Supreme Court called the theory behind the concurrence's tort-based approach into question when it held that the Due Process Clause does not recognize a fundamental "liberty" interest in being free from malicious prosecution. See Albright, 510 U.S. at 269-71. Subsequently, in Taylor v. Waters, 81 F.3d 429 (4th Cir. 1996), we saved our malicious prosecution jurisprudence from invalidation by issuing a saving construction of Goodwin. Specifically, we held that "[t]o the extent that Goodwin based its holding on a conclusion that the officer's failure to disclose exculpatory evidence deprived the § 1983 plaintiffs of a liberty interest in avoiding prosecution on less than probable cause, that reasoning has been rejected in Albright.... But, to the extent that Goodwin ruled that the officer's failure to disclose the exculpatory information deprived the § 1983 plaintiffs of their right to a fair trial, its holding is not affected by Albright." Id. at 436 n.5.

Thus, as a result of Albright, we were forced to recast the constitutional basis of malicious prosecution, employing a"fair trial" due process rationale, rather than a "tort-based" due process rationale. When police officers maliciously withhold exculpatory evidence from prosecutors, this violates the Due Process Clause because it denies criminal defendants a "fair trial" -- not because it creates a tortious interference with a liberty interest in exculpatory evidence. Thus, contrary to the concurrence's suggestion, malicious prosecution can not furnish the tort on which to hang Jean's § 1983 action.6

More fundamentally, the concurrence's belief that§ 1983 claims must proceed on a tort theory, or not at all, is entirely misplaced. This would only be the case if: (1) section 1983 damages were expressly reserved for due process violations involving tortious conduct; and (2) failures to comport with the Due Process Clause's"fair trial" guarantee (as expressed in cases like Brady) entitled criminal defendants only to a new trial. Neither proposition is supportable.

The Due Process Clause's "fair trial" guarantee, usually discussed under the heading of procedural due process, prohibits the State from depriving its citizens of liberty in a criminal trial unless it first observes certain procedural safeguards. While the focus in procedural due process cases is on the State's non-compliance with certain delineated "procedures," the constitutional injury in such cases goes beyond the denial of the "process" that is due. A better description of the constitutional injury would be "the deprivation of liberty without due process of law."

Reconceptualizing the constitutional injury in procedural due process cases to encompass the underlying liberty deprivation puts considerable strain on the concurrence's suggestion that a "new trial" is the sole remedy when the State fails to observe a procedural requirement, such as the obligation to disclose exculpatory evidence. While Jean finally persuaded this court to vacate his conviction because of the due process infirmities surrounding the officers' non-disclosure of the hypnosis evidence, this belated remedy was hardly commensurate with the extent of his constitutional injuries. The harm Jean endured was the deprivation of his liberty without due process of law, not merely the denial of some freestanding procedural interest in exculpatory evidence. Consequently, the nine years he spent in jail are part of the corpus of compensable injuries he sustained.

Without a § 1983 damages action, Jean will not receive compensation for the full extent of his injuries. The concurrence, however, obstinately adheres to the view that the Due Process Clause's "fair trial" guarantee is vindicated solely through new trials -- with § 1983 damages reserved for due process violations involving tortious acts by state officials. This two-tiered interpretation of the rights and remedies created by the Due Process Clause is indefensible analytically. And in human terms, the concurrence's flawed analytical construct works a more profound injustice by denying a deserving civil rights plaintiff the relief to which he is entitled.

I respectfully dissent.