United States v. Daryle Sellers, 906 F.3d 848

Background

Sellers was arrested in an ATF "stash house" reverse-sting operation and convicted of conspiracy to distribute cocaine and conspiracy to interfere with commerce by robbery; he appealed the denial of discovery on a selective enforcement claim.
Stash-house reverse-stings: undercover agents (with confidential informants) recruit targets to plan a robbery of a fictitious drug stash house; arrests occur before any real crime exists.
Sellers proffered statistical evidence showing a large racial disparity in stash-house prosecutions in the district (majority Black or Hispanic; very few white defendants) and Agent Carr testified similarly nationwide.
The district court denied discovery applying the Armstrong selective-prosecution discovery standard; Sellers appealed that denial.
The Ninth Circuit majority held Armstrong’s rigorous discovery standard for selective prosecution does not strictly apply to selective enforcement claims in stash-house reverse-sting cases and remanded for the district court to apply a more flexible, case-specific discovery gatekeeping analysis.
The court emphasized two material differences from selective prosecution: (1) law-enforcement agents lack the same presumption of regularity and privileges afforded prosecutors; (2) comparative statistics identifying similarly situated non-targeted individuals typically do not exist in reverse-sting contexts.

Issues

Issue	Plaintiff's Argument	Defendant's Argument	Held
Standard for discovery on a selective enforcement claim arising from stash-house reverse-stings	Sellers argued Armstrong’s selective-prosecution discovery standard is too strict because agents, not prosecutors, do the targeting and comparative data are unavailable; he sought discovery based on racial disparity evidence	Government argued Armstrong applies and Sellers’ raw statistics are insufficient to warrant discovery	Court held Armstrong’s rigorous discovery standard for selective prosecution does not strictly apply; district courts should exercise discretion and may permit discovery on a lower, case-specific showing in stash-house reverse-sting cases
Whether Sellers’ proffered statistics alone warranted discovery	Sellers argued the district statistics and agent testimony raised an inference of discriminatory effect sufficient to obtain discovery	Government argued raw, non-comparative statistics are irrelevant and insufficient	Court did not decide whether Sellers’ specific proffer was ultimately sufficient; remanded to district court to apply the flexible standard and determine if discovery is warranted
Need to show similarly situated non-targeted persons	Sellers: impossible in reverse-sting context and should not be required for initial discovery	Government: comparable individuals are required (per Armstrong/Bass) to show discriminatory effect	Court held requiring proof of similarly situated non-targeted persons at the discovery threshold is often infeasible in reverse-sting cases and therefore not a prerequisite to all discovery requests
Scope of permissible discovery	Sellers sought agent/supervisor materials on targeting practices and demographics	Government argued such discovery is unwarranted absent Armstrong-level showing	Court held district courts have discretion to fashion limited or broad discovery based on reliability/strength of the defendant’s showing; in-camera review and tailored discovery are options

Key Cases Cited

United States v. Armstrong, 517 U.S. 456 (Sup. Ct.) (rigorous discovery standard for selective prosecution claims)
United States v. Davis, 793 F.3d 712 (7th Cir. 2015) (en banc) (held Armstrong is not dispositive for selective enforcement in stash-house stings and permitted broader discovery)
United States v. Washington, 869 F.3d 193 (3d Cir. 2017) (applied flexible discovery approach for selective enforcement in reverse-sting cases)
United States v. Hare, 820 F.3d 93 (4th Cir. 2016) (discussed difficulties of obtaining comparative evidence in stash-house context)
United States v. Bass, 536 U.S. 862 (Sup. Ct.) (raw, non-comparative statistics insufficient to show discriminatory effect)