Can constitutional theory be both persuasive and determinate? A recent, must-read exchange between Michael Dorf and Larry Solum implicitly raises this question. In the exchange, Dorf ultimately posits the “Brown test,” which demands that any constitutional theory must make possible the result in Brown v. Board, but need not require that result. The exchange (which Paul Horwitz has already collected and commented on here) illustrates that there are importantly different ways of assessing a constitutional theory’s determinacy. Below, I outline relevant features of the Dorf/Solum posts, explore the relationship between theory and determinacy, and suggest a reformulation of the Brown test. In short, I suggest that evaluations of constitutional theories should emphasize what the theories most encourage, as opposed to what is possible, necessary, or reasonable under those theories.
In a provocative post, Dorf proposed an initial formulation of the Brown test: “any interpretive approach that is acceptable must actually produce Brown.” On its face, this test seems attractive. Brown is probably the most canonical constitutional case of them all, so a constitutional theory that didn’t “produce” Brown might seem alien and illegitimate. Of course, there are reasons to question this claim, as Michael Ramsey, Asher Steinberg, and others have recently noted. But I will assume that constitutional theory must accept some version of the Brown test.
In response, Solum questioned whether non-originalist theories could pass the Brown test as Dorf initially formulated it. In Solum’s words, “[I]t is quite clear that many of the major alternatives to originalism do not as a matter of necessity produce the outcome in Brown.” Solum offered examples from constitutional common law and multiple interpretive modalities.
Dorf’s reply post largely agreed with Solum (on this point) and so offered a revised formulation of the Brown test:
If a practitioner of some brand of originalism (or for that matter any other approach) cannot plausibly conclude that Brown is right, then that brand of originalism (or the other approach) fails the Brown test. To pass the Brown test, a theory need not guarantee the outcome in Brown.
This revised formulation raises interesting questions about the practical import of constitutional theory. Let’s accept the Brown test as Dorf has now formulated it. And let’s further assume that originalism and its major alternatives neither dictate nor preclude the result in Brown. Finally, let’s assume that no case is more central to modern constitutional law than Brown. (In light of the exchange between Dorf and Solum, each of these assumptions is at least plausible.)
Do those assumptions mean that constitutional theory is destined to be so indeterminate as to be otiose? I don’t think so.
Imagine that an unspecified Constitutional Theory X becomes the undisputed law of the land. (This is a simplifying assumption that we would often want to relax when evaluating constitutional theories.) To assess the implications of this event, we might ask:
- What can interpreters do via legal argument?
- What can interpreters do *and get away with it*?
- What can they get away with *in good conscience*?
- What can they get away with *and still pride themselves on adherence to the law*?
If we are asking only the first question, then no constitutional theory is likely to be particularly interesting. Constitutional practice is complicated, with many cross-cutting values, principles, and rules. Constitutional Theory X must reflect that complexity if it is to be persuasive as either an explanation or critique of constitutional practice. But that complexity necessarily creates a considerable degree of indeterminacy in the operation of constitutional theory. As a result, persuasive constitutional theories have a hard time preventing motivated interpreters from achieving whatever conclusions they might realistically desire to achieve.
So the problem of indeterminacy does not arise exclusively because of the Brown test discussed by Dorf and Solum. Rather, the Brown test illustrates that considerable and possibly unavoidable indeterminacy follows any persuasive constitutional theory. Adam Samaha made a similar point when he suggested that “there might be a negative correlation between a constitutional authority theory’s persuasiveness and its practical implications for interpretation.”
But the full implications of Constitutional Theory X can’t be measured based on question #1 alone. While leaving a great deal of discretion, Constitutional Theory X might influence the persuasiveness of various options and cast those options in a different light. In other words, even if all realistically desirable outcomes are available under any viable Constitutional Theory X, that wouldn’t mean that all such outcomes are equally persuasive.
Because Constitutional Theory X may very well change the relative strengths of various constitutional positions, its adoption may affect the answers to questions #2-4. For instance, the theory’s main result may be that interpreters discipline one another more effectively, implicating question #2. Or each interpreter might internalize the new constitutional theory and so choose not to make especially strained or unreasonable arguments, implicating question #3. Finally, the interpreters, either individually or collectively, might endorse the constitutional theory so completely that they joyfully employ the highest standards and exhibit great diligence in applying the theory, implicating question #4. So by the time we reach question #4, Constitutional Theory X seems much more likely to matter.
Of course, questions #1-4 aren’t the only relevant questions we might ask, and each might be further refined in various ways. But they make the point: a constitutional theory’s practical determinacy is a matter of degree.
The difference between question #1 and questions #2-4 suggests that debates over constitutional theory should not focus on what outcomes are merely available under any given constitutional theory. At times, the Dorf/Solum exchange could be read to involve these sorts of claims, such as when the authors discuss what is “necessary” under a theory or what a theory can “guarantee.”
Moreover, questions #2-4 suggest that constitutional theory might profitably make greater use of the familiar terms “best,” “reasonable,” and “unreasonable.” For instance, Dorf initially formulated the Brown test so as to require that any non-Brown outcome must be an unreasonable interpretation. His reformulation, by contrast, requires only that Brown be a reasonable constitutional interpretation. But there is a third salient option: perhaps we should demand a constitutional theory under which Brown is the best constitutional interpretation.
This new reformulation of the Brown test has the advantage of allowing it to do significant work in evaluating constitutional theories. Dorf’s original formulation had the power to rule out mainstream versions of originalism, but Solum correctly observed that few if any serious theories could pass that demanding test. Under Dorf’s revised formulation, by contrast, it is doubtful that the Brown test excludes any significant originalism, since there are various reasonable ways to square such theories with the conclusion reached in Brown. For instance, an original-expectations originalist might reasonably rely on Michael McConnell’s work, or an originalist who accomodated precedent might reasonably extend pre-Brown cases eroding Plessy. (Notably, Brown found historical sources “inconclusive” either way and rejected only certain “language” in Plessy, without overruling Plessy in its entirety.) So while Dorf initially suggested that “Brown is at least a prima facie problem for originalists,” his revised formulation seems to leave originalists without any Brown problem at all.
When reformulated in terms of the best interpretation, however, the Brown test can meaningfully discriminate between theories. For instance, there is a strong argument that Brown deviates from the best interpretation of the Constitution under significant forms of originalism, such as theories that focus on original expected applications. Thus, the best-interpretation formulation of the Brown test would seem to pose a genuine problem for at least some significant originalist theories. In his reply post, Dorf himself noted that while “pluralism of the sort championed by Philip Bobbitt (and Dick Fallon and others) could produce not-Brown, … I think that in 1954 it would have been more likely to produce Brown.” This focus on what is “more likely” to follow from an interpretive approach is akin to asking whether the approach is best understood to yield Brown. And Dorf also suggested that certain types of originalism could “comfortably” yield Brown, whereas that outcome would be “tougher sledding” for others. Again, this discussion suggests the value of the best-interpretation formulation.
* * *
In short, we shouldn’t just ask what a constitutional theory demands or allows. We should also ask what it most encourages.
First posted on Prawfs.