Do Judges Discover Or Create Law?

Yesterday the Supreme Court decided Danforth v. Minnesota. The case involved a complex question about retroactivity and the powers of state courts. At issue was the Supreme Court's interpretation of the Confrontation Clause in Crawford v. Washington and the rule's retroactivity under the Court's precedent. Since rules of this sort are not retroactive, the Court then had to decide whether state courts could make the rule retroactive in their own courts. The Court ruled that the its retroactivity doctrine is only a floor and not ceiling. State courts could decide to expand the defendant protections if they so chose.

Orin Kerr took issue with one comment from the majority's ruling. Justice Stevens argued that,

[T]he source of a "new rule" [in Crawford] is the Constitution itself, not any judicial power to create new rules of law. Accordingly, the underlying right necessarily pre-exists our articulation of the new rule. What we are actually determining when we assess the "retroactivity" of a new rule is not the temporal scope of a newly announced right, but whether a violation of the right that occurred prior to the announcement of the new rule will entitle a criminal defendant to the relief sought.

Justice Stevens is painting a common picture of the judicial method in constitutional cases. The judge looks at the relevant constitutional text, interprets it based on the given facts, and announces a rule that is derived from the Constitution itself. But is this picture accurate? Kerr asks

So if I understand Justice Stevens correctly, a case like Miranda v. Arizona was constitutionally required at the time of the Framing, but the Court just didn't see the true Constitution until the 5-4 decision by Earl Warren in 1966? Um, like, okay.

Kerr obviously believes that in cases like Miranda judges make the law rather than discover it. I think Kerr is basically correct and that Stevens' conception of the judicial role is overly simplistic.

Mitchell Berman, in a very, very long but very worthwhile paper, argues that most constitutional decisions involve two steps: first the Court must determine the actual meaning of the Constitution and then it must create doctrine that is used to implement that meaning. The decision has two outputs: constitutional meaning and constitutional doctrine. A good example of this dichotomy is the Court's three tiered Equal Protection doctrine. For the last 90 years or so the Court has reviewed different classifications based on divergent standards of review. Racial classification are reviewed under strict scrutiny, which basically assumes the classification violates the Constitution and places a heavy burden on the government to justify it. When reviewing gender classification the Court uses a more liberal standard, but still places the burden on the government. All other classifications are reviewed under a very lax rational basis standard.

Does the Constitution really require these standards? While the Court has ruled that the standards of review in cases involving the Free Exercise Clause are constitutionally required and immune from congressional intervention, a more plausible understanding of these clauses is that standards of review are merely doctrine used to implement the constitutional commands.

Why can't the Court just apply the Constitution's meaning to the case at hand without implementing via doctrine? The most common justification is epistemic uncertainty. In contrast to the other branches, the judiciary has few tools to determine empirical facts outside of the briefs and the lower court's decision. So the Court is at a disadvantage; if the federal or a state government decides to legislate a racial classification, the Court is in no position to determine whether that classification conflicts with the Equal Protection Clause. It has no way to really know if that particular classification was generated by racial animus.

But courts do know history. They are certain that previous racial classifications were rooted in prejudice and hate. So what the Court can do is assume that all racial classifications are illicit unless the government can prove otherwise. In some situations the government will be unable to sufficiently prove good faith even when its classification was not prohibited under the Constitution, but that is price we have to pay for epistemic uncertainty.

Miranda is another example of this dichotomy. The Court clearly believed that the Constitution's meaning prohibited forced confessions, even in state courts. But how can the Court know if the confession was forced? So the Court created all types of rules that presumed the confession was forced unless the defendant was told his rights at the time of the arrest. This rule was not constitutionally mandated, but it was a mechanism used to implement the Constitution given the lack of information readily before the Court.

I don't know enough about Crawford to decide whether the rule promulgated in that case was part of the Constitution's meaning or the Court's doctrine. But it's important to keep this distinction in mind, at the very least, because doctrine can be changed must easier than the Constitution's meaning itself.