Is Amendment 2 really a bill of attainder? Some questions about Professor Amar's analysis of Romer.

• Author: Hills, Roderick M., Jr.
• Position: Response to article in this issue by Akhil Reed Amar, p. 203

As I first discovered as a law student in Professor Amar's classes on legal history and federal courts, it is generally an intellectual treat to listen to Professor Amar's legal analysis, even when he is attacking one's own arguments. So my pleasure at reading Professor Amar's analysis of the Court's decision in Romer v. Evans was only partly dampened by his disapproval of the respondents' brief that I and other plaintiffs' counsel filed with the Court. According to Amar, this respondents' brief provided the Court with "so little help" that it had to rely on an entirely different and much sounder argument - an argument rooted in the U.S. Constitution's prohibition on attainder,(1) contained in Article I, sections 9 and 10.

Amar maintains that (1) contrary to Justice Scalia's vituperative dissent, the attainder argument provides an intellectually compelling basis for believing that Amendment 2 is unconstitutional, and (2) the Romer decision, correctly interpreted, adopted precisely this argument. Amar's revival of the Attainder Clauses is classically Amaresque: it takes constitutional text and structure seriously and it provides an original and sensitive reading of specific constitutional clauses and a careful understanding of their structural relationships.

However, as much as I appreciate his elegant and astute reading of the Attainder Clauses, I think in the end that his application of these clauses to Amendment 2 and his reading of Romer are unconvincing. The difficulty with his argument is that, as Amar notes, the Attainder Clauses prohibit state and federal legislation from "naming persons and singling them out for distinctive treatment.112 As explained below, a law "names" persons only if it defines a closed class of persons with some fixed characteristic - a class the entire membership of which could be known (at least in theory) by the legislators at the moment when the law is enacted.

Amar provides no persuasive argument that the term "homosexual, lesbian, or bisexual orientation" denotes such a closed class. To make such an argument, he would have to show that sexual orientation is not merely a status, but also an irreversible status, a characteristic that does not change over time and that thereby defines a closed class from which members cannot exit and nonmembers cannot enter. Rather than attempt such an argument, Amar argues that Amendment 2 discriminates on the basis of "status" rather than "conduct" and thus "targets persons for who they are, not what they have done."(3) But this status-conduct distinction is irrelevant to the issue raised by the Attainder Clauses - the issue of whether Amendment 2 names persons by designating them as a closed class, the entire membership of which could be known by the legislature.

Amar's emphasis on the status-conduct decision is mischievous not merely because it misconstrues the Attainder Clauses but also because the distinction is, in a larger sense, deeply misguided: although the distinction repeatedly surfaces in gay rights litigation,(4) it is practically trivial and intellectually incoherent. Indeed, this is why neither the respondents' brief, nor Professor Tribe's amicus brief, nor - as I shall explain below - the Romer Court relied on such a distinction. Amendment 2 would be a deprivation of equal protection - although not an attaint - even if the term "orientation" were omitted from its text. For, as Romer and respondent's brief repeatedly state (and as Amar curiously ignores), the central flaw in Amendment 2 is not its ambiguous and probably severable mention of "orientation" but rather its breadth, its imposition of a "broad and undifferentiated disability on a single named group,"(5) regardless of whether the group is defined by conduct or orientation. It is the sheer breadth of Amendment 2 that makes it constitutionally suspect, and not some burden on "orientation" as opposed to "conduct."

In the following pages, I will first attempt to explain the scope and purposes underlying the Constitution's clauses forbidding bills of attainder. Then I will show that the distinction between "orientation" and conduct" - or "status" and "conduct" - really has nothing whatsoever to do with this principle. Finally, I will try to show that, like respondent's brief, Romer depends crucially on the breadth of Amendment 2 - the wide category of antidiscrimination laws that Amendment 2 preempted. It is this breadth, and not any use of the term "orientation" that led the Court to invalidate Amendment 2.

I

First, let me start where I think that Amar and I agree: What is attainder, and why is it suspicious?

As Amar notes, at the core of the rules against attainder is the notion that [al law naming persons and singling them out for distinctive treatment is suspicious."(6) The fundamental principle underlying the rule is that the state and federal legislatures must make policy by "generally applicable rule[s]"(7) rather than by laws that single out groups or persons by name - what the Court calls "specifically designated persons or groups."(8) Thus, the most obvious violation of the rule against attainder is a statute that literally designates individuals by their proper names - for instance, a law stating, "Akhil Amar is barred from holding public office." Amar is surely correct that the Attainder Clauses would also bar the Congress from using definite descriptions for the same purpose as a proper name: it would equally be unconstitutional for Congress to declare that "all persons who wrote an article entitled Of Sovereignty and Federalism will be barred from public office," for the definite description obviously serves the purpose of singling out a specific individual and no one else. Moreover, Amar must also be correct that a law can be a bill of attainder if it designates a specific group of persons: so, for instance, a law barring "all members of the Amar family" from holding public office would be an attainder just as much as if each member of the family were individually listed in the text of the statute.

From these uncontroversial propositions, Amar makes the insightful observation that the rule against attainder is actually "a prototype of the Equal Protection Clause:"(9) the rule limits the ability of the legislature to single out disfavored groups and thereby prevents deprivation of equal legal protections. As Amar notes, a law barring "all persons of East Indian descent" from holding public office would be just as much of an attainder as a law barring "the Amar family" from holding public office. For such a law would "specifically designate[]" a group of persons and no one else for disfavored treatment just as much as a law penalizing an entire family.10 Indeed, the analogy between racial and familial classifications is extraordinarily close: both are legal relationships defined by lineal descent, ancestry, or blood line.

But Amar's astute analogy between the rule against attainder and the rule against deprivations of equal protection is not just an insight but a warning: there is a danger that the concept of attainder can become just as murky and incoherent as concepts of equal protection, bogged down in "`free-form' constitutionalism" which Amar rightly disparages but that has notoriously plagued equal protection law. The problem is that, while the rule against attainder prohibits laws that impose punishment on "specifically designated persons or groups,"(11) we do not really have a clear notion of what it means to "specifically" designate something or someone. Of course, proper names are the easiest case. 12 But, as we have seen, they are not the only sort of "specificity" that the anti-attainder rule prohibits. Laws that burden "members of the Communist Party" or "former rebels against the United States government" are also too "specific." But then what exactly is not "too specific"? What passes muster as a "generally applicable rule"? As Professor Tribe has noted, "the concept of legislative 'specification' . . . cannot be so broad as to swallow up all laws that impose some disabling limitation upon an ascertainable group."(13( Could a law impose regulatory burdens on "the catfood industry"? Fly fishermen"? "Persons under six feet in height"?

Unfortunately, Amar does not provide a criterion for defining illegal legislative specification - illegal "naming" - beyond stating that laws cannot "target[] persons for who they are" on the basis of their "Status."(14) This is a little vague: what does it mean exactly to target persons "for who they are?" All laws, after all, distinguish between persons based on their characteristics - frequently including characteristics like age and handicap that are involuntary, personal traits. And these characteristics seem to be a part of what identifies persons as "who they are."

I think that one can provide a more precise account of what it means to "name" a person or group. The essence of such "naming" - such illegal legislative specification - is that the legislation defines a closed class, a class with a membership that is permanently fixed when the class is defined, from which members can never exit and into which nonmembers can never enter, as a matter of law. Logic, precedent, and policy suggest that the Attainder Clauses forbid such closed classes and nothing else.

Consider, first, the logic of proper names. The essence of naming is to designate a unique person or group of persons to which anyone not so designated can never belong. Unlike a general term, a name functions like a telephone number or address in that it designates one item or set of items and no others, without attempting to say anything more about the items.15 When I use the name "Akhil Amar," I intend to refer to a specific person and to no one else (not even to other people who might, by coincidence, have the same name). Likewise, if I refer to "all persons who aided the Confederacy during the Civil War," I name a closed class of...