The transition from President Obama to President Trump has produced an abrupt shift in many areas of policy, particularly with respect to law enforcement.
As I discussed in an earlier post, President Obama took significant steps to moderate immigration and marijuana enforcement, even as his administration pursued aggressive regulation in other areas. Now, the priorities have flipped. Immigration enforcement is in. Regulatory enforcement is out. We don’t know yet about marijuana.
This sort of flip-flop raises hard questions of reliance. What happens now to people who opened marijuana businesses or applied for immigration relief based on President Obama’s assurances? Can due process protect their reliance and provide a defense against future enforcement? If not, haven’t they been deprived of fair notice about what conduct may be sanctioned?
After Trump leaves office, this problem could arise in quite different areas of policy. Trump might well build on Obama’s example with aggressive nonenforcement policies of his own. Even if he doesn’t, enforcement rates will likely crater in key areas of regulation, from environmental law and labor protections to tax and ethics compliance. We may then end up four or eight years from now with a very different set of parties claiming reliance.
A forthcoming article of mine, Reliance on Nonenforcement, addresses when, if ever, such defendants can claim a constitutional due process reliance defense.
In general, for reasons I’ll explain in this post, due process cannot protect reliance on nonenforcement policies. It is important, though, to understand why, because framing the problem correctly helps to identify situations in which a due process defense is valid—a question I’ll address in a future post.
Why is nonenforcement reliance generally unprotected? The reason is not that enforcement meets commonsense standards of fairness. As a practical matter, President Obama’s policies invited millions of people, many of them legally unsophisticated, to take significant legal risks in reliance on formally non-binding assurances. Marijuana entrepreneurs started businesses that amount to criminal enterprises under federal law. For their part, undocumented immigrants effectively confessed their immigration violations and exposed their whereabouts under the Deferred Action for Childhood Arrivals (DACA) program.
The policies’ fine print, to be sure, made clear that DACA and the marijuana guidance were revocable shifts in enforcement, not durable changes in the law. In real-world terms, though, these programs’ beneficiaries took the President at his word that no harm would come to them if they acted in reliance. It is inconceivable that they would have acted as they did had they anticipated that future enforcement was a real possibility.
Nevertheless, fairness cannot be the sole consideration in this context. Separation-of-powers principles also must weigh in the balance.
If regulated parties could rely on nonenforcement assurances, then executive officials could wipe away governing substantive laws simply by inviting reliance on promised forbearance. Executive officials then could disable disfavored laws simply by encouraging regulated parties to break them.
Executive officials do not have this authority. The framers called this sort of law-cancellation power a “suspending” or “dispensing” authority. Although English monarchs historically exercised it, our Constitution’s Take Care Clause repudiates this power by making clear that the executive function in our system is to “faithfully execute[]” the law, not to exercise a permanent veto over it.
An important early circuit decision, United States v. Smith (1806), illustrates this point. Presiding over a criminal trial as circuit justice, Supreme Court Justice William Patterson held that the President’s alleged authorization of criminal conduct was not even relevant in the resulting trial. Any contrary rule, Patterson explained, “would render the execution of the laws dependent on [the President’s] will and pleasure.”
That principle remains vital to the rule of law in this country. It is why President Trump cannot authorize violations of tax laws, ethics requirements, environmental protections, or any number of other statutes he might chafe against. Regulated parties may well be tempted to capitalize on this administration’s willingness to turn a blind eye. But if the substantive law remains unchanged and the political winds shift, a moment of reckoning could still come.
The same basic principle dictates that marijuana entrepreneurs and undocumented immigrants who benefitted from Obama’s policies have no general reliance defense either. The new administration could choose to prosecute marijuana sellers or seek removal of deferred-action recipients (though, for reasons I’ll address in my next post, it cannot use DACA applicants’ own submissions to target them).
This conclusion is harsh. Its harshness is another reason why nonenforcement policies are a troubling means of addressing contested policy questions.
Such policies amount to playing chicken with the next administration, daring it to incur the political cost of disrupting regulated parties’ reliance. In the past, political incentives and good-governance norms have prevented such harsh reversals. We will see whether those restraints still apply in the Trump era.
In the meantime, recognizing that the reliance inquiry in this context requires balancing separation-of-powers costs against fairness considerations can help identify isolated contexts in which the harsh general rule against reliance can be relaxed—either because separation-of-powers concerns are attenuated or because fairness concerns are unusually acute.
Tomorrow, I’ll turn to that question.