Must Police Return Wrongfully Seized Marijuana?

Imagine the following scenario. P is a local police officer who stops D for speeding. In the course of a consensual search of D’s car, P finds a small bag of marijuana. State law bans the simple possession of marijuana by most people, so P seizes the drug. However, D subsequently convinces a judge that he was allowed by state law to possess the marijuana (say, because he’s a qualified patient under the state’s medical marijuana law). Following the judge’s ruling, D asks P to return the marijuana. Must P do so?

Many states explicitly require the police to return marijuana they have wrongfully (as a matter of state law) seized from individuals like D. But the police sometimes refuse to obey such requirements. The police claim that the act of returning marijuana constitutes a drug distribution offense under 21 U.S.C. Section 841, since federal law defines “distribution” as any transfer of a banned substance. Hence, a state law requiring police to return marijuana poses a direct conflict with and is thereby preempted by federal law.

Ah, but there is a twist. 21 U.S.C. section 885(d) expressly immunizes state police from “civil or criminal liability” under federal drug laws if they are “lawfully engaged in the enforcement of any law or municipal ordinance relating to controlled substances.” The provision was likely adopted to shield police from liability for participating in sting operations. But marijuana users have claimed that Section 885(d) also applies to other scenarios, including the return of marijuana, and thereby resolves any conflict between state and federal law.

Who is right? There are at least four possible ways of resolving these disputes, none of which is entirely satisfactory:

1. Adopt a purposivist interpretation of Section 885(d) and side with the police. Just last month, for example, the Colorado Supreme Court found that Section 885(d) would not immunize state police for returning seized marijuana; thus, state police could ignore a state law that required them to do so. Colorado v. Crouse. This interpretation is arguably consistent with the purpose of Section 885(d), but it is difficult to reconcile with the provision’s text. The Crouse court, for example, suggested that a police officer would not be “lawfully engaged in the enforcement” of state law if she violated federal law—i.e., by distributing marijuana. But the same could be said of the undercover police officer who sells marijuana during a sting operation.
2. Adopt a literal interpretation of Section 885(d) and side with users. In one case, for example, a California appellate court found that Section 885(d) plainly shielded state police from federal criminal liability for returning marijuana to a medical marijuana patient. City of Garden Grove v. Kha. The Kha court reasoned that Section 885(d) “makes law enforcement personnel immune from any civil or criminal liability arising out of their handling of controlled substances as part of their official duties. . . . There can be little question the [City] police would be acting pursuant to their official duties, were they to comply with the trial court’s order to return [the citizen’s] marijuana to him.” This interpretation arguably comports with the text of Section 885(d) (and prevents state agents from undermining state marijuana reforms). But it difficult to reconcile with the text of another provision of the Controlled Substances Act. 21 U.S.C. Section 903 expressly preempts any state law that poses a “positive conflict” with the CSA. To be sure, Congress might not care if state police return marijuana to the same person from whom they seized it. But interpreted literally, Section 885(d) might block preemption of other, more controversial state actions. Imagine, for example, that a state orders its agents to produce and sell marijuana. Indeed, in the early 2000s, the City of Oakland, California, thought it could immunize a local medical marijuana dispensary by deputizing its owner (Ed Rosenthal) and ordering him to supply the needs of the City's medical marijuana patients. When Rosenthal was later prosecuted by the federal government, the district court rejected his assertion of Section 885(d) immunity, though in so doing, it (like the Crouse court) had to rely on a questionable reading of Section 885(d). United States v. Rosenthal.
3. Find that the anti-commandeering rule empowers state agents to return marijuana. The idea is that if Congress can’t force state police to seize marijuana in the first instance, it also shouldn’t be allowed to force them to retain the drug if they no longer wish to hold it. To be sure, Congress can preempt some state actions that violate federal law – say, growing marijuana at a state-run farm. But there has to be some limit to preemption if the anti-commandeering rule is to do any work. Otherwise, as I’ve argued elsewhere (p. 1446-49), Congress could prevent states from voting to repeal their marijuana bans, releasing prisoners held on marijuana charges, etc. (both of which entail actions of some sort). So as long as state police do no more than restore the proverbial state of nature (say, by returning marijuana to its original owner), Congress may not be able to stop them. (I sketched this argument in the article above (p. 1459-60).) If a court followed this approach, it could sidestep the 885(d) issue and users would win. 
4. Find that the police lack standing to challenge state duties as preempted. The likelihood that any police officer would actually be prosecuted for returning marijuana is almost zero (because of DOJ enforcement memoranda, congressional spending restrictions, etc.). In other words, the threatened injury to the police is too speculative to satisfy federal standing requirements. If a court followed this approach, it again wouldn’t have to reach the merits of the Section 885(d) issue and users would presumably win. However, even if this standing argument would work in federal court, it wouldn’t necessarily work in state court (where most of these claims have been raised ). That’s because state courts don’t necessarily apply the same restrictive standing rules as their federal counterparts.

As I noted above, none of these approaches is entirely satisfactory to me. But I’m interested to hear what others think – and whether there might be another solution to the puzzle.