What specific evidence would a defendant need to prove selective prosecution in federal court?
Executive summary
To prevail on a federal selective-prosecution claim a defendant must produce clear, specific evidence showing the prosecution had a discriminatory effect — that similarly situated persons of a different, protected class were not prosecuted — and a discriminatory purpose motivating the charging decision; courts start with a presumption of prosecutorial regularity and require “some evidence” to permit discovery or dismissal [1] [2] [3]. The required proof typically includes comparative data, specific non-prosecution examples, or internal government communications that reveal improper motive, and courts routinely deny broad discovery absent a colorable showing because most of the relevant evidence is in the government’s hands [4] [5] [6].
1. The controlling legal standard and burden of proof
The Supreme Court in United States v. Armstrong set the baseline: prosecutors enjoy broad discretion and defendants must present “clear evidence” that the prosecutorial policy both had a discriminatory effect and was motivated by discriminatory purpose; absent that showing courts presume prosecutors acted properly [1] [2]. Lower courts and practitioner guides restate that threshold as “some evidence” or a “colorable basis” sufficient to justify limited discovery, but they stress the defendant bears the burden of producing specific facts, not mere speculation [5] [6] [7].
2. Element one — proving discriminatory effect with comparators and statistics
To show discriminatory effect a defendant must identify others “similarly situated” who engaged in the same conduct but were not prosecuted; this typically requires concrete comparator examples and, when available, statistical analyses showing a disparity in charging patterns by race, religion, political alignment, or other protected characteristics [8] [3]. Courts have found small-n comparator lists sufficient in some cases when the nonprosecution pattern is stark, but mere anecdote or general assertions of unfairness are routinely rejected [3] [6].
3. Element two — proving discriminatory purpose with documents and statements
Discriminatory intent is harder to prove and normally requires evidence such as internal prosecutorial or DOJ communications, emails or memoranda revealing motive, public statements by decisionmakers linking enforcement to impermissible criteria, or witness testimony that connects the charging decision to bias; without such direct or strongly inferential material courts are reluctant to infer illicit motive [4] [7]. Because the government controls much of this material, defendants often try to obtain discovery by first meeting the “some evidence” threshold showing both effect and potential intent [5] [6].
4. What types of evidence actually move courts — and why discovery is rare
Practically, winning or even getting discovery requires a package: named comparators, statistical or empirical studies, affidavits from knowledgeable witnesses (e.g., defense counsel or third parties), and any public statements or documents suggesting bias; isolated assertions or timing coincidences usually fail [8] [3] [6]. The Supreme Court and many commentators warn that wide-ranging discovery into prosecutorial files is disfavored because it would disrupt prosecutorial discretion and reveal strategy, so judges will only order targeted disclosure when the defendant has established a colorable claim [2] [5].
5. Procedural realities, deadlines, and strategic considerations
Selective-prosecution claims must generally be raised early, often via a pretrial motion under Rule 12, and statutory or local scheduling deadlines can be short; failing to timely assert the claim can forfeit the issue [9] [10]. Given the evidentiary and procedural hurdles, defense strategy often couples selective-prosecution motions with conventional attacks on the government’s evidence, targeted discovery requests narrowly tailored to alleged bias, and expert or statistical support to make the initial “some evidence” showing convincing to a judge [7] [8] [11].