Yesterday, I raised an important question about the Obama Administration’s marijuana and immigration policies: now that there’s a new President, what happens to folks who relied on these policies?
For reasons explained in my last post, constitutional due process protections generally don’t protect either group from enforcement. Because executive officials cannot change substantive legal obligations on their own authority, past enforcement policies cannot bar future enforcement—even if regulated parties relied to their detriment on past assurances.
Yet this rule is not absolute. It ultimately turns on a balance between separation-of-powers considerations and fairness concerns. While separation-of-powers limits must generally win out in this balance, sometimes fairness concerns should predominate—either because those concerns are particularly acute or because separation-of-powers costs are attenuated (or both).
Drawing off case law recognizing a narrow reliance defense in some analogous contexts, my forthcoming article identifies several situations in which a reliance defense should be available:
Indirect reliance—DACA beneficiaries
President Trump remains free to cancel the “Deferred Action for Childhood Arrivals” (DACA) program if he wishes, and the immigrants who benefited from it remain subject to potential deportation, however heartless it would be to actually deport them. But the Trump Administration cannot use information from these immigrants’ own deferred action applications against them.
That is so because the unfairness to individual immigrants would be unusually sharp. After all, a programmatic request for information based on promised relief invites reliance in a far more specific way than does promised nonenforcement by itself. At the same time, any separation-of-powers cost in barring use of this information is limited, so long as the government remains free to build enforcement cases by other means.
Had DACA never existed, the government would have had to do gumshoe detective work to identify and apprehend DACA beneficiaries. It shouldn’t be spared that burden by virtue of having tricked those same immigrants with false promises of relief. To allow the government to perform such a bait and switch would be an outrageous form of entrapment that the Due Process Clause should prevent.
Indirect reliance—state marijuana records
By the same token, federal officials also should not be able to employ disclosures by state-authorized marijuana businesses to prosecute them for federal offenses. Some have argued that compelled production of state-collected information is unconstitutional across the board. Whether or not that is true in general, a due process defense should prevent the government from using information reported to state authorities in reliance on a then-effective federal nonenforcement policy.
Again, the federal government can still prove its case on its own if it wishes, but it shouldn’t come out ahead by virtue of having encouraged regulated parties to engage in the conduct they effectively confessed to state authorities.
Even apart from the Justice Department’s announced marijuana nonenforcement policy, an appropriations rider now precludes use of Justice Department funds to prosecute individuals who possess or distribute medical marijuana in compliance with state law.
Much like an executive nonenforcement policy, this appropriations restriction does not purport to alter the underlying substantive law. It simply suspends enforcement of those substantive laws for the time being. Yet because the suspension comes from Congress—which could change the law if it wished—the separation-of-powers concern that generally justifies limiting reliance defenses is absent. Accordingly, down the road, courts should be more willing to recognize a due process defense in appropriate cases.
Finally, if either the marijuana policy or the DACA deferred-action program were to persist for a sufficiently lengthy period without change, these policies should eventually harden into concrete guarantees.
A number of prominent scholars have advocated a general due process defense of desuetude. This general desuetude defense would preclude prosecution when defendants violated laws that had long gone unenforced. (Co-blogger Jamal Green argues here for recognizing such a defense generally in the immigration context.)
Whether or not due process should provide such protection in general, an overt nonenforcement policy that persists for a substantial period should eventually support a desuetude defense for conduct within the policy’s parameters.
Why? Because if an overt nonenforcement policy persists for some lengthy period without any apparent departures from the policy’s terms, the balance between fairness and separation-of-powers will eventually tip in favor of fairness. After all, reliance interests will only build up more and more over time, while the policies’ durability across administrations may suggest that democratic support for the underlying substantive laws has lapsed (even if Congress fails to amend the underlying laws).
* * *
This post and the preceding one are just a brief summary of my views on these reliance questions. For the full details, and some additional examples of valid defenses, read the full article here. For another short version of my argument on DACA beneficiaries, see this Washington Post op-ed.
Unless one party or the other manages to establish long-term control over the Presidency, I suspect this type of administrative flip-flop will continue to arise in presidential transitions. If that prediction proves true, questions of reasonable reliance will complicate federal law enforcement for years to come.