Sunday, November 22, 2015
On-Line Symposium on Texas v. United States -- Michael Kagan, The Uses and Abuses of Notice and Comment
Michael Kagan provides the latest installment in ImmigrationProf Blog's On-Line symposium on Texas v. United States.
In this Symposium post, I would like to focus on the question of whether DACA and DAPA should have gone through a notice and comment process under the Administrative Procedure Act (APA). This contention was the primary basis for Judge Hanen’s injunction against the deferred action programs, although the Fifth Circuit’s November 9 decision added an alternative ground – the claim that the Department of Homeland Security does not have the statutory authority to implement these programs.
I would like to make the case that notice and comment is not required and that notice and comment would contribute little to democratic governance in the context of President Obama’s deferred action policies. It is difficult to imagine how any executive branch policy could have been announced with more transparency or subject to more vigorous public discussion than DACA and DAPA. The main impact of requiring notice and comment would be to empower unelected and unaccountable frontline DHS employees who happen to disagree with the elected President about how the immigration enforcement discretion should be used. That would not be a good thing for American democracy.
I would also like to suggest that the Department of Justice has made a questionable strategic choice by trying to convince the Fifth Circuit that DAPA and DACA are not really binding on frontline DHS officials. DOJ has in essence argued a question of fact when the lower courts have primarily erred on a question of law. The Fifth Circuit has erred in the way it has defined “binding” rules, and in so doing it has made it more difficult for the President to perform as Chief Executive when frontline public employees disagree with the President’s policy choices. While conservative legal thinkers (and judges) may generally oppose President Obama's immigration policies, they should be concerned about an approach to administrative law that makes it difficult for a President to direct federal agencies that are staffed by people with contrary policy preferences.
As I explain in a forthcoming article, the APA challenge to DACA/DAPA involves a longstanding problem of administrative law: What is the difference between a “legislative rule” and a “nonlegislative rule?” Section 553 of the APA requires a notice and comment process for “rulemaking.” However, the notice and comment process is time consuming and burdensome. It also results in rules that significantly constrain an agency’s flexibility because the agency would have to go through a new notice and comment process to rescind or change them. But agencies need not always go through this process because notice and comment is not required for “general statements of policy,” sometimes called non-legislative rules.
The resolution of Texas v. United States depends on whether DACA/DAPA are legislative rules or not. The trouble is that the Supreme Court has said relatively little to help make the distinction – a void the Court could fill in this case. But for now, this void is filled by a muddled body of case law from the D.C. Circuit, which said in 1974 that “the distinguishing feature of a policy statement is that it …[is not] a binding norm.” In 1987, the D.C. Circuit elaborated on this definition with a two-part test in Community Nutrition Institute v. Young (CNI) to determine whether a rule is binding. First, a general statement of policy “is one that does not impose any rights or obligations.” That is not a problem for DACA and DAPA, which were established by a memorandum from Secretary Jeh Johnson that states “this memorandum confers no substantive right.” The trouble is with the second part of the CNI test: A policy statement “genuinely leaves the agency and its decision-makers free to exercise discretion” (emphasis added). The leading Fifth Circuit case on point, Professionals & Patients for Customized Care v. Shalala (1995), adopted this language verbatim. That sentence is the doctrinal foundation for the Fifth Circuit’s decision in Texas.
The Texas case highlights the fact that there are two different ways to understand this test. The correct way, in my view, is that if an agency is free to deviate from a general statement of policy at its own discretion, it is not a binding rule. Thus, so long as the Secretary of Homeland Security is free to rescind the policy at a moment’s notice or to decide not to follow it in an individual case, the rule is not binding and need not go through notice and comment. DACA and DAPA would survive under this interpretation of the APA.
But in Texas, Judge Hanen and the majority of the Fifth Circuit panel have applied a very different understanding. They have reasoned that a rule is binding if it deprives frontline officers – not the agency as a whole – of discretion. To prove this, they have relied on the low rate of denials of DACA applications, and on statements by leaders of DHS employee unions that “applications are simply rubberstamped if the applicants meet the necessary criteria.”
The linchpin of the argument is the D.C. Circuit’s phraseology in CNI: does a rule leave “the agency and its decision-makers free to exercise discretion.” The entire argument hinges on those three words. But as I show in my article, both the D.C. Circuit and the Fifth Circuit have been inconsistent about this phrasing, sometimes referring to the agency as a whole, while elsewhere referring to the agency “and its decision-makers.” In its most recent case on this issue, Association of Flight Attendants-CWA v. Huerta (the 2015 case that gave us all the freedom to use tablet computers during take off and landing on commercial airplanes) the D.C. Circuit found that instructions given to frontline inspectors about how to interpret a pre-existing regulation did not need to go through notice and comment. All of this suggests that it is an error to base an important decision on these three stray words in CNI.
This argument as not been emphasized by the Department of Justice so far. Instead, DO’s primary strategy has been to try to show that frontline officials do actually have some discretion in DACA and DAPA. There are valid points to be made here, but I think this should be at best a fallback argument. One of the great virtues of DACA and DAPA is that the eligibility standards are quite clear-cut, and that they do take much decision-making power away from DHS employees who might not exercise it the way the President wants. I would argue that even assuming the “rubberstamp” depiction of DACA and DAPA, notice comment would not be required.
It’s not just a question of careless phraseology. There are good reasons not to require notice and comment in order for a cabinet secretary to direct his subordinates how to exercise the agency’s discretionary enforcement powers. One of President Obama’s problems is that frontline DHS officers have opposed his immigration policies from the outset. This is evident in the Texas case itself, given that the central evidence relied on by Judge Hanen came from the unions representing ICE and USCIS employees. With the Morton Memos, the Obama Administration tried a system governed by a loose, open-ended guidance document under which the decisive decisions had to be made case-by-case. The results were disappointing, because ICE officials simply have different ideas than the President about when and how prosecutorial discretion should be exercised. The more rigid, categorical nature of DACA and DAPA corrected this problem. They maintain the discretionary power of the agency, while stripping most decision-making power away from agency employees who disagree with the way the President wants to use that power.
Conservative jurists may not like President Obama’s immigration policies, but they should be sympathetic with an elected president trying to implement policy through an agency staffed by nominal subordinates who are hostile to his policy preferences. As Linda Greenhouse recently noted, this was a significant problem for President Ronald Reagan, which led conservative thinkers to favor stronger president control over executive branch agencies. A Republican president in the future might regret it if President Obama loses this fight, because it would make it much harder for any President to direct public policy at the frontlines. As the Fifth Circuit noted in Professionals & Patients, “All statements of policy channel discretion to some degree – indeed, that is their purpose.”
Think of what it would mean if notice and comment were required in order for a head of an agency to tell his or her nominal subordinates how to exercise discretion. It would become much harder for nominal heads of agencies – and ultimately the President – to carry out their policies. More power over public policy would shift to unelected public employees and their unions, and it would become harder for the public to either hold the President to account or to change policy by electing a new President.
In a previous Symposium post, Peter Margulies argued that there is value in requiring notice and comment because “input from a spectrum of stakeholders would improve regulatory outputs.” I believe these benefits are vastly overstated and more relevant in other policy contexts. There would be a stronger argument to insist on notice and comment if President Obama had launched a large scale change in immigration policy in secret. But that is not the case here. Involvement of stakeholders through a formalized process would also have value where Congress has insisted that technical or scientific considerations take priority over political calculations. But as the Supreme Court said in Arizona v. Untied States (2012), “Discretion in the enforcement of immigration law embraces immediate human concerns.” How to weigh those concerns is ultimately subjective. Some people think certain groups of immigrants are sympathetic, while others see them as unsympathetic. The best way in a democracy for this call to be made is for it to be attributable to an elected official, rather than to an anonymous agency employee.
In Lincoln v. Vigil (1993), the Court said that general statements of policy can be useful because they “advise the public prospectively of the manner in which the agency proposes to exercise a discretionary power.” This is a call for transparency, which is a prerequisite for the voting public to be able to hold the President accountable. It is hard to think of how a series of notices in the Federal Register would have enhanced public information beyond what President Obama and Secretary Johnson actually did – high profile presidential speeches, detailed memoranda published on official websites, user-friendly web guides, and so on.
The insistence on notice and comment appears to suggest that there are some constituencies that want to speak up about deferred action who have not been able to do so far. But it would be more correct to say that there are constituencies who loudly oppose President Obama’s policy choices, that these people are frustrated because they have not gotten their way, and that some of these people work for the Department of Homeland Security. The more democratic response to these people would be to say, if you don’t like the current deferred action policies, elect a new president who will change them.
Indeed, there are a number of candidates offering to do just that.