Fragmentation of Disclosure Responsibility

The Brady Doctrine is often spoken of as if it imposes a single, unified duty on a single, unified actor. In formal language, that is roughly true: the prosecution must disclose favorable evidence material to guilt or punishment. But in operational reality, the duty is distributed across a fragmented institutional landscape. Facts are generated by officers, investigators, dispatch personnel, forensic analysts, technicians, confidential-source handlers, records custodians, supervisors, and prosecutors. Credibility information may sit in personnel files, disciplinary records, internal-affairs systems, laboratory documentation, or agency counsel files. Digital evidence may be held by specialized units or third-party vendors. The constitutional duty is singular in doctrine, but plural in administration. That mismatch is one of the central engines of Brady failure.

Kyles v. Whitley made the structural point unmistakable. The Court held that the individual prosecutor has a duty to learn of favorable evidence known to others acting on the government’s behalf in the case, including the police. That rule rejects the notion that a prosecutor may avoid responsibility by remaining personally unaware of what investigators know. It also confirms that Brady operates on an imputed-knowledge model: the state is treated as a constitutionally integrated actor even when its actual institutions are divided into bureaucratic compartments. The doctrine therefore assumes coherence at precisely the point where the system most often exhibits fragmentation.

This fragmentation matters because disclosure failures rarely occur through a single dramatic decision. More often they arise from diffusion. One actor does not record a fact; another stores a note in a separate file; another fails to appreciate impeachment significance; another assumes someone else has already disclosed it; another treats personnel misconduct as confidential rather than constitutionally relevant; another lacks access to the relevant database; and the trial prosecutor proceeds on the mistaken belief that the file is complete. In such a system, no single participant may perceive the full constitutional failure as it develops. The omission becomes systemic before it becomes visible.

That is why fragmentation of responsibility is more dangerous than simple negligence. Where responsibility is divided without a reliable mechanism of consolidation, accountability becomes diluted. Each participant may control only one segment of the informational chain, while the defense confronts the state as though it were a unified litigant. The constitutional right belongs to the accused, but the burden of honoring it is scattered among multiple offices, units, and record systems. The result is a familiar institutional pathology: everyone bears some responsibility, yet no one bears enough of it in practice to ensure full disclosure.

Giglio v. United States exposes another dimension of the same problem. There, a promise of leniency made to a key witness was not disclosed, and the prosecutor who tried the case was unaware of the promise. The Supreme Court nonetheless treated the failure as attributable to the government. The significance of Giglio extends beyond impeachment doctrine. It demonstrates that fragmentation within a prosecutor’s office does not dissolve the government’s constitutional obligation. A promise made by one government lawyer cannot be insulated from Brady scrutiny merely because another lawyer handled the trial without knowledge of it. In structural terms, Giglio rejects internal compartmentalization as a defense.

Yet that doctrinal answer does not solve the administrative problem. It simply identifies the constitutional endpoint. The practical question remains how offices are supposed to gather, verify, and disclose favorable information across fragmented systems. A line prosecutor may not control law-enforcement personnel files. Police agencies may resist turning over credibility records absent formal requests. Forensic material may be housed in separate evidence-management systems. Notes may exist outside official reports. Audio, video, and metadata may be collected by units that are not in routine communication with the charging prosecutor. The more technologically and bureaucratically complex the system becomes, the less plausible it is to rely on informal memory and ad hoc inquiry.

The doctrine itself intensifies the coordination challenge because Brady material is not limited to obviously exculpatory proof. Favorable evidence includes impeachment evidence, and impeachment information is often institutionally dispersed. It may arise from prior inconsistent statements, inducements, benefits, disciplinary histories, findings of dishonesty, forensic irregularities, or undisclosed constraints on witness reliability. Such material is rarely stored in a single place. To comply with Brady in any serious sense, the prosecution team must not merely possess a legal duty; it must maintain an architecture capable of aggregating fragments into knowledge. Without that architecture, the doctrine demands unity from a system built on separation.

Strickler v. Greene is also important here because it frames Brady claims through favorability, suppression, and prejudice. In a fragmented system, the suppression element becomes especially difficult to evaluate cleanly. Information may have existed somewhere within the government’s reach but not within the trial prosecutor’s file. A court may later describe the evidence as “suppressed by the prosecution,” but institutionally the failure may have occurred much earlier, at the level of record creation, retention, indexing, or interagency transmission. This means that Brady suppression is often the end-stage legal description of a much earlier breakdown in administrative responsibility.

The materiality standard in Bagley further complicates fragmented disclosure. Because the legal test asks whether there is a reasonable probability that the result would have been different, with “reasonable probability” meaning a probability sufficient to undermine confidence in the outcome, offices may be tempted to let fragmented custody become fragmented judgment. One custodian views a fact as minor. Another sees a personnel issue as collateral. Another treats a witness benefit as routine. But the cumulative force of multiple modest items may be substantial. Kyles emphasized that the prosecution must gauge the likely net effect of all favorable evidence and evaluate it collectively, not item by item. Fragmentation is therefore not only a problem of storage; it is a problem of aggregation. A system that cannot combine pieces cannot reliably assess materiality.

This is where the architecture of responsibility becomes decisive. If agencies maintain separate records without mandatory cross-reporting, if prosecutors rely on informal oral assurances rather than documented requests, if impeachment material is treated as a discretionary personnel matter, or if disclosure logs do not identify the provenance of information, then the prosecution team may appear constitutionally functional while remaining operationally blind. The state will still be charged with knowledge under Kyles and Giglio, but that imputed knowledge will exist largely as a judicial fiction unless the system is designed to make it real.

Fragmentation also affects timing. When responsibility is distributed, disclosure is often delayed not because anyone affirmatively rejects Brady, but because no one controls the full process from collection to production. A police agency may respond slowly to a prosecutor’s inquiry. A laboratory may not transmit notes with the final report. An internal-affairs record may require supervisory approval before release. A digital extraction may be completed after charging but before trial without being flagged as potentially favorable. These timing defects matter because Brady is not satisfied by disclosure in the abstract; disclosure must occur with enough time for meaningful defense use. Administrative fragmentation therefore converts substantive rights into scheduling hazards.

The ethical layer does not eliminate the problem. Professional rules such as ABA Model Rule 3.8(d) require timely disclosure of evidence or information known to the prosecutor that tends to negate guilt or mitigate the offense. But the phrase “known to the prosecutor” reveals the structural vulnerability. Knowledge is not self-generating. It depends on reporting channels, searchable systems, office policy, and institutional seriousness. If the architecture of knowledge is fragmented, then the ethical command attaches to an information state that may itself be artificially narrow. In such circumstances, exhortations to disclose are inadequate unless accompanied by mechanisms that make relevant knowledge actually reachable.

The fragmentation of disclosure responsibility is therefore not a secondary implementation issue. It is one of the core reasons Brady collapses in practice. The doctrine assumes a prosecution team able to collect and synthesize favorable evidence across institutional boundaries. Modern criminal administration often operates instead through specialized silos, partial files, divided custodianship, and layered discretion. When constitutional duty is centralized but institutional knowledge is decentralized, failure becomes predictable. The law speaks in the language of one actor, while the system behaves as many.

The obvious implication is that disclosure responsibility must be structurally consolidated even if operational tasks remain distributed. That means documented interagency requests, mandatory reporting duties for law enforcement and forensic units, centralized impeachment repositories, auditable disclosure workflows, supervisory signoff, and judicial expectations that reject “someone else had it” as an answer to constitutional nondisclosure. Without such measures, Brady remains formally unitary while practically fragmented. And where responsibility is fragmented, concealment need not be intentional to become systemic.

Resources

Brady v. Maryland, 373 U.S. 83 (1963)

Giglio v. United States, 405 U.S. 150 (1972)

United States v. Bagley, 473 U.S. 667 (1985)

Kyles v. Whitley, 514 U.S. 419 (1995)

Strickler v. Greene, 527 U.S. 263 (1999)

Treatment of Brady v. Maryland Material in United States District and State Courts Rules, Administrative Office of the U.S. Courts