Over the last decade, an obvious cause for exceptional public interest in immigration law presidentialism has been the aggressiveness of both Presidents Barack Obama and Donald Trump in using “executive action” to pursue vastly different visions of appropriate immigration enforcement. The most prominent Obama Administration initiatives were Deferred Action for Childhood Arrivals (DACA) and Deferred Action for Parents of Americans (DAPA). I discussed DAPA at some length in a recent article on a phenomenon I call “faithful nonexecution,” programmatic decisions to implement statutes in ways other than their most comprehensive enforcement, but which are nonetheless consistent with a president’s “faithful execution” obligations. With regard to DAPA, I argued that the Administration made strategic errors in framing its deferred action programs as a response to interbranch conflict and purporting to ground DAPA in a constitutionally rooted prosecutorial power of the president. The Administration could have cast both DACA and DAPA not as in conflict with Congress, but as fulfilling the purposes of existing law in a manner consistent with the principal-agent model. 

            With apologies for largely quoting what I wrote in that article, the starting points for my analysis were the express terms by which Congress empowered the Secretary of Homeland Security to implement immigration law. The Immigration and Nationality Act places the Secretary in charge of “the administration and enforcement of . . . all . . . laws relating to the immigration and naturalization of aliens.” It gives him or her a wide range of tools to carry out that authority: “He shall establish such regulations; prescribe such forms of bond, reports, entries, and other papers; issue such instructions; and perform such other acts as he deems necessary for carrying out his authority under the provisions of this chapter.” And, in case the point needed reinforcement, the Homeland Security Act of 2002 includes among the Secretary’s authorities “[e]stablishing national immigration enforcement policies and priorities.” Because of this last explicit grant of authority, I find it puzzling that Cox and Rodríguez repeatedly call the executive’s power over immigration enforcement “de facto delegation.” The delegation, I would argue, is de jure. On at least one occasion, in the Consolidated Appropriations Act of 2014, Congress had even directed DHS to “prioritize the identification and removal of aliens convicted of a crime by the severity of that crime.” The Obama Administration priorities were thus consistent with Congress’s wishes, and nothing in the Immigration and Nationality Act or in the Homeland Security Act prohibits the creation of a deferred action program as a way of implementing the Secretary’s enforcement priorities. Highlighting both the national security implications of the Secretary’s implementation scheme and a statutory text that is, at worst, ambiguous with regard to the scope of authority delegated to the Secretary, arguably situates any resulting legal controversy at a point where the executive branch would be entitled to considerable deference to its policy judgment. 

            DAPA could also have been shown to advance a series of purposes consistent with Congress’s broad public policy objectives: strengthening local law enforcement, supporting the national economy, and preserving family unity. A number of amicus briefs from police organizations supporting DAPA argued that the program would strengthen community policing by promoting trust and cooperation with local law enforcement, facilitating access to identification, and helping to protect a vulnerable population from crime and exploitation. A supportive brief filed on behalf of sixteen states and the District of Columbia stressed that removing the risk of deportation would prompt recipients of deferred action to “come out of the shadows,” apply for work authorization, and increase their earnings, thus simultaneously growing the states’ tax base. Immigrant employees who work legally are also less susceptible to wage exploitation, which should reduce the downward pressure that undocumented immigrants working for low wages otherwise exert on the incomes of U.S. citizens and permanent resident aliens. DAPA would also reduce the economic, emotional, psychological, and educational harms that result from the persistent threat of splitting up families. Had DAPA gone through conventional notice-and-comment rulemaking, all of these points would surely have been documented. The Administration could also have stressed the importance to effective administration of relative consistency. With or without DAPA, deferred action would be available to its applicants on a case-by-case basis. What DAPA would have helped to assure is that all parts of the immigration enforcement bureaucracy would apply the same rules, so that adjudicative results would not be random. Of course, implementing DAPA through notice-and-comment rulemaking and defending it in conventional administrative law terms might not have sufficed to satisfy the Fifth Circuit of its legality. My point, however, is that an administrative law framework entirely hospitable to the defense of DAPA was available. No re-conception of the separation of powers was necessary to argue for even so bold an executive branch innovation. The principal-agent model still works descriptively for agencies to which a great deal of discretion has been delegated. 

            In addition to thinking the Cox-Rodríguez re-characterization of the separation of powers is unnecessary, I have two worries about unintended consequences that could follow from their conceptualization. Although Cox and Rodríguez seem to reject unitary executive theory as a constitutional mandate, the way in which they urge the legitimacy of having political officials rather than line bureaucrats determining general enforcement policy (as opposed to individual enforcement decisions) could be read to feed claims for aggressive presidentialism. This would be tragic. Among Trump’s gravest injuries to public administration have been his attempts to re-engineer the executive branch to be a reliable instrument of pure presidential will. Not only has he sought to extend political control over all administrative adjudicators, including immigration judges. He has systematically sought to sidestep the Senate’s advice-and-consent role in staffing the immigration bureaucracy. At this moment, of the top dozen positions in the DHS order of succession, all of which require Senate advice and consent, exactly two are filled by individuals duly nominated, confirmed, and appointed. The duties of every other position are being performed either by an “acting official” or by a senior official formally holding a different post, but to whom the duties of the vacant line have been delegated by administrative order. As Trump’s performance shows, unitary executive theory predictably feeds a White House psychology of presidential entitlement that is deeply antidemocratic and antagonistic to the rule of law. The president should, of course, have oversight with regard to the policy making of the agencies that report to the president. But to use the terminology of Peter Strauss, the president’s role should be understood as “overseer,” not “decider” with regard to matters—such as immigration enforcement—that Congress has explicitly delegated to other agencies. 

            The second unintended consequence I fear is that the call by Cox and Rodríguez for a “new” understanding of the relationship between Congress and the president will ironically help to bolster the false idea among the Supreme Court’s right wing that the formalist view of separation of powers is actually the proper view. It is not just in immigration law, but also in environmental law, trade law, communications law, civil rights law, criminal law, tax law—perhaps all public law—that most administrative law scholarship has long embraced a view of the separation of powers in which the principal-agent relationship is highly dynamic, and discretion in the agent is deemed a feature, not a bug of the system. Although agency expertise has also been a classic argument in favor of such discretion, it has not been the exclusive argument. Locating fundamental policy responsibility in Congress’s chosen agents rather than the White House has been defended, and is defensible, on grounds of transparency, legal accountability, and democratic responsiveness. Presidential oversight will inevitably be important, but it should not transmogrify into unchecked powers of command. I doubt that Cox and Rodríguez disagree.