« Hyman on Equal Protection
Chris Green
| Main | Tom Bell: The Constitution as if Consent Mattered
Michael Ramsey


Originalism and judicial restraint, part III
Chris Green

The story so far (part I, part II): judicial assertions about the Constitution might be thought to be governed by a norm of knowledge that can be breached in two ways: by deliberate ignorance of relevant considerations (Level 1 on the activismometer, too low, as in O'Gorman & Young v. Hartford Fire Insurance (1931), explained below) and by assertions in the teeth of ignorance of relevant considerations (Level 5 on the activismometer, too high, as in Jackson's opinion in Youngstown Sheet and Tube v. Sawyer (1952), and other cases to be explained in further posts). What counts as "knowledge" depends on our assessments of the stakes in judicial review, leading to levels 2, 3, and 4 on the activismometer--demanding proof, respectively, beyond a reasonable doubt (e.g., as in Commonwealth ex rel. O'Hara v. Smith (Pa. 1811) and Washington's opinion in Ogden v. Saunders (1827)), by clear and convincing evidence (e.g., as in Fletcher v. Peck (1810)), or only a preponderance of evidence (e.g., as Steven Calabresi would have it).

Level 1 is the deliberate failure to consider evidence relevant to a statute's constitutionality in a case in which a court has jurisdiction to assess it. The approach to the presumption of constitutionality taken in O'Gorman & Young v. Hartford Fire Insurance (1931), in which the Court said that, in order to sustain a statute, it would assume the legislative fact-finding necessary to sustain the distinctions drawn in legislation, even where the legislature itself found no such facts, is such an example. The Court refused to make any inference at all from the legislative silence on the relevant justificatory facts. That failure, however, is profoundly relevant to whether the legislature's distinction was in fact, justified.

Compare the situation with criminal self-defense justifications. A criminal defendant who wants a jury to be able to consider the issue of whether his actions are justified (say, in killing someone) must present some evidence in order to receive a jury instruction on self-defense. A criminal defendant who supplies no evidence to show he was justified is treated as unjustified. Moreover, even if the defendant produces evidence at trial that his actions were justified, most reasonable observers would think it mightily mysterious, and suspicious, if that defendant had never told that story before trial. 

Griffin v. California (1965) and Doyle v. Ohio (1976), to be sure, protect criminal defendants against such adverse inferences. But those rules are not rooted in the fact that inferences from silence are, as a general matter, irrational or unreliable. Rather, they are rooted in the particular circumstances of an individual testifying in own defense and the ability of arrestees not to submit to implicit pressure to talk to the police.

When O'Gorman is used in the context of a legislature who adopts a legislative distinction without actually articulating any facts that could justify that distinction, it operates a bit like a Griffin or Doyle rule for the government's justifications for its statutes, but shorn of any Fifth Amendment rationale. Under O'Gorman, courts use the presumption of constitutionality as a "fact-finding" tool--really, a device for suppressing adverse inferences from the government's failure to justify a distinction at the proper time. Indeed, when O'Gorman is used to find facts when the government has failed to articulate its justification even after the fact, it goes far beyond even Griffin and Doyle, because it operates to suppress the adverse inference from the government's failure to justify its distinctions at any time, akin to a rule requiring the prosecution to disprove justification defenses on which the criminal defendant has offered no evidence at all.

The failure to draw reasonable inferences from governmental silence in defense of its own statutes is not the proper application of a high burden of proof (i.e., Level 2 activism), but instead an instance of delibrate ignorance (i.e., Level 1 excess passivity). Deliberate ignorance is of course sometimes a good idea--self-incrimination and the exclusionary rule may be instances, and respect for others' privacy is obviously another--but judicial review does not seem to be one of them.

The model of self-defense justifications in the criminal law shows why departing from O'Gorman need not take us any higher than Level 2 on our activismometer. The criminal defendant has a burden of production with respect to explaining his justification defense--that is, is subject to an adverse inference from the failure to produce such evidence. But under most states' approach, once the defendant produces some evidence on the issue, the prosecution still retains the burden of showing guilt--i.e., disproving the self-defense justification, if offered--beyond a reasonable doubt. Whether to be at Level 2, 3, or 4 is an issue of how much evidence is required to show unconstitutionality, but the O'Gorman issue is whether legislative silence itself counts as evidence that can help meet that burden. Ignoring relevant silence is staying at Level 1; we can pay attention to it but still require unconstitutionality to be shown beyond a reasonable doubt, and thus stay at Level 2.

Next up: Roe, Windsor, and Perry as Level 5 cases.

[Update: some details about Griffin and Doyle corrected.]