The only eyewitness had told the FBI that the murderer was light-skinned and had a moustache. (1) Calvin Bibb had never worn a moustache, and even the U.S. Attorney prosecuting him for murder-for-hire wouldn't call him "light-skinned" in open court. Bibb needed the FBI report recording the eyewitness's statements, and he needed to know where he could find the eyewitness so his attorney could interview her. Invoking Brady v. Maryland, (2) Bibb moved to compel the government to disclose this evidence.
The U.S. Attorney acknowledged the government's constitutional duty under Brady to disclose evidence favorable to Bibb. But the U.S. Attorney minimized that duty, emphasizing that such evidence only needs to be disclosed if material. And evidence is only material if it has a "reasonable probability" of changing the trial's outcome. (3) Not only did the Department of Justice have plenty of circumstantial evidence showing that Bibb had murdered Gerald McCoy for $2500; his co-conspirator was testifying against him as well. The U.S. Attorney did not think there was the slightest chance, much less a "reasonable probability," that impeaching the eyewitness--who was drunk and scared when she saw Bibb shoot McCoy--would change his "slam-dunk" guilty verdict into an acquittal. Hence, the evidence was not material. The government had no duty to disclose it under Brady.
The district court found the "reasonable probability" test problematic. Developed for appellate review, the test was too speculative to use in an ongoing trial. It is impossible to predict how witnesses will testify, how the judge will rule on the parties' motions, and whether the jury will ultimately find guilt. (4) And if Bibb's verdict was unpredictable, how Bibb's verdict might be affected by eyewitness
statements about the murderer's moustache and his light skin was even more so. Hence, the district court held that the government had to disclose all evidence favorable to the accused. After a two-week trial, Bibb was acquitted.
Few trial courts have adopted the Bibb court's approach. (5) Most follow one of two other approaches. Trial courts taking the "postconviction" approach hold that Brady violations can only be raised after a conviction. (6) The government has a duty to disclose favorable, material evidence during the trial, but that duty is only enforceable if the accused is convicted and raises the issue on appeal. Trial courts using the "speculative" approach hold that the "reasonable probability" test applies during trial itself: if the evidence has a "reasonable probability" of changing the verdict, the court orders disclosure. (7) But when the duty to disclose is made to rely on a postconviction remedy or a speculative trial outcome, Brady's constitutional principles are diminished, impairing the accuracy and legitimacy of the criminal process. The massive scale of American criminal justice--more than 1.5 million adults are incarcerated (8)--underscores the seriousness of these consequences.
But none of this is inevitable. The fairness of our system of criminal justice depends, in part, on enforcing a duty that already exists. When the accused has not been convicted, courts must enforce Brady by compelling the government to disclose all evidence favorable to the accused, without reference to probabilities, "reasonable" or otherwise, and without regard to its potential impact on the outcome of the proceedings. Brady's constitutional duty applies with equal force during plea bargaining, after indictment, and during the trial itself.
My argument is as follows: Part I examines the Brady v. Maryland decision. "Materiality" in Brady does not entail speculation; it is defined by the basic laws of evidence. And taking the power to decide what must be disclosed away from prosecutors and giving it to judges so they can make trials fairer is part of Brady's design. Part II looks at decisions interpreting Brady. The development of the "reasonable probability" test, when considered next to the doctrines that it was derived from and developed alongside--harmless-error review and prejudice-- shows that, like those doctrines, it applies only during postconviction review.
Part III fast-forwards to the current state of criminal justice in the United States, which is dominated by plea bargaining. At first it seemed like Brady's passageway to plea bargaining was obstructed by United States v. Ruiz. (9) But recently the Court has taken a more realistic approach to plea bargaining, moving away from Ruiz's efficiency-oriented, contract-based model and towards treating the plea bargaining process as an informal, pervasive mode of adjudication that must be regulated. And plea bargaining will never discard its unconscionable reputation without the help of Brady's duty to disclose, so Brady's broad application is inevitable.
Part IV looks at prosecutors, confronted with their own cognitive biases, amplified by their discretionary power, thwarted by the formless "reasonable probability" test, trying but failing to satisfy their constitutional obligations under Brady, and convicting innocent people as a result. Part V solves the problem by implementing Brady with a constitutional decision rule that defines materiality according to evidence law and holds prosecutors accountable with minimal judicial effort. Constitutional norms are best implemented by rules adapted to particular decisionmaking environments--here, the plea-bargaining and trial processes. Part VI concludes that, contrary to the assertions of some commentators, applying the "reasonable probability" test before conviction is not ordained by precedent. Indeed, my solution is more consistent with Brady, more consistent with judicial economy, and more consistent with fairness.
Throughout, I am guided by the principle that "justice suffers when any accused is treated unfairly," (10) but my primary purpose is pragmatic: to give trial judges the strongest foundation, in law and in policy, to broadly implement Brady's duty to disclose, ensuring fairness for all accused of crimes, whether they are plea bargaining, awaiting trial, or being tried.
I: BRADY V. MARYLAND: "NOT TO ACHIEVE VICTORY, BUT TO ESTABLISH JUSTICE"
This Part is a close reading of the Supreme Court's Brady v. Maryland decision. Section A covers the essentials: the facts of the dispute, the legal issues lurking in the background, the basic principles staked out by the Court, and, of course: Brady's holding. Section B hones in on a single element--materiality. What's material and what's not would prove to be an enduring source of controversy, but it was not controversial to the Brady Court, which was simply referencing evidentiary laws every litigator is forced to memorize. Section C argues that Brady's holding is a constitutional rule that allocates power away from prosecutors and towards the courts to better implement the principles of fairness underlying the Brady decision.
From Deception to Fairness: Brady's Facts and Brady's Holding
The government confronts a person accused of a crime as an adversary, not an equal. While the government's authority and resources overwhelm the rich and poor alike, (11) the endemic poverty of criminal defendants (12) makes for an especially stark disparity. (13)
This imbalance is reinforced by the adversary system, which entrusts the production of evidence to the adversaries themselves. (14) Unlike a private party, the government can offer witnesses monetary rewards and promise testifying suspects immunity. (15) Along with its superior resources, this ensures that the government has more persuasive evidence. (16)
This power over evidence creates the potential for abuse. There were few safeguards against such abuse until the middle of the twentieth century, when the Supreme Court, in a series of terse opinions, prohibited the government from using testimony its prosecutors knew to be perjured or false. (17) This false testimony violated due process by making the trial a "contrivance" where the government convicted the accused through "deliberate deception." (18) These decisions, while laudable, did not create balance or fairness. While false evidence could not deliberately be used against the accused, genuine evidence that exculpated the accused could still be withheld. The focus was on the government's wrongdoing, not the nature of the evidence itself.
Brady v. Maryland reversed that focus: wrongdoing was secondary to whether the undisclosed evidence was favorable to the accused. John Brady's trial had been unfair, not because the prosecutor had done any wrong or deliberately deceived him in any way, but because the prosecutor had not disclosed evidence that was favorable to him--his accomplice's confession. (19)
John Brady and his accomplice Donald Boblit had both confessed to lying in wait for their friend William Brooks, knocking their friend unconscious, stealing their friend's brand-new Ford Fairlane, and hiding their friend's corpse deep in the woods, but each blamed the other for actually choking the life out of their friend William Brooks. (20) This finger-pointing did not stop the trial court from sentencing them both to death. Nor did it stop the Maryland Court of Appeals from upholding those sentences. (21)
But then the proceedings took an unexpected turn. Poring over trial transcripts for anything that might bolster his client's application for clemency, John Brady's lawyer uncovered Donald Boblit's confession to killing William Brooks. (22) Although Brady's only defense had been his accomplice's culpability for the actual killing, and although Brady had accordingly requested all of his accomplice's statements to the police and prosecution, the government had still withheld the confession. (23)
So Brady appealed. The Maryland Court of Appeals held that suppressing the confession violated due process. (24) Presaging the later Supreme Court decision, its holding was based on the confession's...