My chief criticism of Obergefell was its failure to substantiate its many claims of knowledge regarding the similarity of same-sex and opposite-sex relationships with respect to the purposes of marriage. The one place in which the Court descended into the particular details–on the immutability issue–it cited evidence which actually contradicted its claim, as well as ignoring the bulk of the arguments and evidence before it. Does that make Obergefell "activist"? Ryan Anderson calls it "as clear of an example of judicial activism as any we have seen in recent years."
In an article last year, I suggested that we define activism and restraint epistemically–that is, in terms of knowledge and the standards it requires. This distinction is independent of the ontological question at issue in debates over originalism, i.e., the temporal extent–intergenerational, or not?–of constitutional truthmakers. I coined the word "activismometer" for the result of applying these epistemic principles and norms of conversation, describing five levels of judicial activism and passivity. Level 1 ignores relevant evidence or enforces knowably-unconstitutional laws, level 2 requires proof beyond reasonable doubt, level 3 requires clarity, level 4 requires mere preponderance, and level 5 does not even require knowledge of relevant considerations or–as in Obergefell, though I did not mention this sort of example in my article–ignores relevant counter-evidence and counter-arguments.
The basic norm of judicial statements about constitutionality is the basic norm of human discourse in general–stick to what you know. Don't make stuff up–follow "the moral rule that persons who tell what they do not know to be true, falsify as much as those who knowingly tell falsehoods." To admit ignorance of relevant considerations about the Constitution, but impose requirements based on one's declarations about it anyway, violates this norm most straightforwardly. I call this level 5 on my activismometer. Central examples are Roe and Justice Jackson in Youngstown. Some of the lower court opinions on the same-sex-marriage question likewise claimed that they need not resolve questions about the whether same-sex and opposite-sex relationships are similarly situated with respect to issues like child-rearing; I place them as well at level 5.
Levels 2, 3, and 4 represent courts who accept that knowledge is required for judicial statements about the Constitution, but who use different standards for assessing what counts as knowledge in the context of judicial review. Many contemporary epistemologists hold that as the stakes increase, the amount of evidence requisite for knowledge likewise increases. If we've got a lot riding on whether the bank is open on Saturday, we are more likely to be agnostic in the face of evidence that might suffice for knowledge in a low-stakes context. Whether we require proof beyond reasonable doubt, clarity, or mere preponderance of the evidence (levels 2, 3, and 4 on the activismometer) depends on our assessments of the stakes involved in judicial review. (I agree with John McGinnis that the right level is 3.)
Level 1 represents the abandonment of judicial review even in the face of knowledge that a statute is unconstitutional or the failure to investigate constitutional claims fully. For a court, this is appropriate only if it lacks jurisdiction over a case, or it is otherwise not in a conversational context appropriate to judicial declarations about the Constitution. (In Paul Grice's terms, this is the violation of a "quantity" norm, as opposed to the "quality" norm requiring knowledge for assertions.) A court in a conversational context where statements about the Constitution are appropriate, however, ought to consider all relevant evidence. In particular, it ought neither to (a) ignore evidence and arguments that have emerged since the legislature passed the law that that a court could use to clarify the constitutional issue, even if it might not have been clear to the legislature itself (James Bradley Thayer's chief mistake), nor (b) ignore the state's failure to supply a justification for its laws (the mistake of O'Gorman in 1931, Williamson in 1955, and generally modern rational-basis-review law). (This adverse inference from silence would impose a burden of production on the state, but not a burden of persuasion.) These mistakes are level 1 on the activismometer, which is too low.
Level 1, then, results from the failure to consider relevant evidence that could support an assertion about the Constitution. Evidence can, however, either support or undermine constitutional claims. Epistemologists call evidence that undermines a claim a "defeater"; they distinguish undercutting defeaters, which show that one's other evidence is unreliable, but without suggesting the opposite belief, from rebutting defeaters, which point the other way from one's other evidence.
To ignore either undercutting defeaters (e.g., criticisms of studies' methodology) or rebutting defeaters (e.g., alternative studies with the opposite conclusion) is to make a mistake precisely akin to the Level 1 mistake in O'Gorman of ignoring relevant evidence. To the extent that Obergefell failed to adequately explain its rejection of contrary evidence, as I have argued here and here, its activismometer level should be set at 5. Roe claimed it didn't need knowledge at all, while Obergefell claimed implicitly it didn't need to rebut criticisms of its view in order to have knowledge. These correspond to the two Level-1 errors of (a) knowing that a statute is unconstitutional, but enforcing it anyway, and (b) ignoring evidence of a statute's unconstitutionality, and enforcing it anyway. Obergefell is thus a new kind of level-5 activism.
Posted at 4:00 AM